Charles
                         460382012                       Associates
Final Report
SCARCITY, RECYCLING AND SUBSTITUTION OF POTENTIALLY
CRITICAL MATERIALS USED FOR VEHICULAR EMISSIONS CONTROL
Prepared for  -
U.S. Environmental Protection Agency
2565 Plymouth Road
Ann Arbor, Michigan  48105
CRA.Report No. 501
Prepared by
Charles River Associates
200 Clarendon Street
Boston, Massachusetts
February 1982

-------
                                                                  Charles
                                                                  River
                                                                  Associates
                             TABLE OF CONTENTS
                                                                       Page

Chapter 1:  INTRODUCTION AND SUMMARY	   1-1

Chapter 2:  MATERIALS CRITICALITY AND EPA POLICYMAKING	   2-1

     Critical  Materials and Strategic Materials 	   2-1
     Market Contingencies That May Be the Basis for Considering
       A Material  Critical	   2-2
          Disrupted Factors of Production 	   2-3
          Inadequacy of Reserves and Resources in the Long Run. .  .   .   2-4
          Technological Shocks	   2-4
          Disrupted Foreign Sources of Supply 	   2-5
          The Business Cycle	   2-6
          Defense	   2-6
     Principles of Measuring Materials Criticality and Implications
      for EPA Pol icymaking	   2-7
          Materials Prices and Compliance Costs 	   2-7
          Compliance Costs and the Critical ity of Materials Used
            to Meet EPA Regulations	   2-9
          The Criticality of Materials from a National Perspective.   .   2-14
          Income Redistribution and Noneconomic Dimensions of
            Materials Criticality 	   2-21
          Role of Secondary Production and Inventories	   2-22
          Adequacy of Private Adaptations to the Threat of Supply
            Disruptions and Other Contingencies 	   2-24
          Implications for EPA Policymaking 	   2-27
          Should EPA Sponsor R&D on Vehicular Emissions Control?.  .   .   2-31
          A Major Qualification and A Possible Policy Prescription.   .   2-32
     Material  Imports and Balance of Payment Problems 	   2-34
     A Simple Economic Model for Estimating Criticality Due To
       Foreign Supply Disruptions 	   2-36
          Key Parameters and Formulas	   2-37
          Treatment of Stockpiling	   2-43
          Meaningful!ness of Results from the Model 	   2-46
     Sample Estimates of the Critical ity of Platinum, Palladium,
       Rhodium, Chromium, Manganese, Nickel, and Titanium Metal .  .   .   2-51
          Normal Consumption, Production, and Prices	   2-52
          Contingency Threats and Price Elasticities	   2-54
          Price Elasticities of Consumption for Vehicular Emissions
            Control	   2-59
          Illustrative Criticality Estimates and Conclusions	   2-60
     Bibliographic Note	   2-64
     Appendix 2-A:  Availability of Materials from the U.S. National
       Stockpile	   2-67
     Appendix 2-B:  Computer Program to Calculate Average Annual
       Economic Losses from Contingencies in Material  Markets  ....   2-70
     Appendix 2-C:  Earlier Approaches to Materials Criticality .  .   .   2-79
     Chapter 2 References 	   2-117

-------
                                                                 Charles
                                                                 River
                                                                 Associates
                     TABLE OF CONTENTS (Continued)

                                                                     Page
Chapter 3:  PROJECTIONS OF MATERIALS CONSUMPTION FOR CONTROL
              OF VEHICULAR EMISSIONS	   3-1

Appendix 3A:  EPA ESTIMATES OF CONSUMPTION OF PLATINUM-GROUP
              METALS FOR CONTROL OF VEHICULAR
              EMISSIONS IN 1981	   3A-1

Chapter 4:  PLATINUM-GROUP METALS 	   4-1

     Introduction 	   4-1
     Stocks	   4-1
        U.S. National  Stockpile 	   4-2
        Refiner, Importer, and Dealer Stocks	   4-2
        Industry Shelf Stocks 	   4-7
        Industry Stocks in Use	   4-9
        Private Speculative/Investment Stocks 	   4-11
        Stocks and Increased Demand in the Short Run	   4-11
     Statistical Overview of Supply and Demand	   4-12
        Supply	   4-12
        Demand.	   4-35
        Consumption Trends in the 1970s 	   4-39
        Prices	   4-47
     Primary Producers	   4-54
        South Africa	   4-54
        The Soviet Union	   4-59
        Canada	   4-61
        United States  	   4-62
        Colombia	   4-63
        Other Countries	   4-63
        World Reserves	   4-63
     Supply Reliability	   4-64
     Consumption Elasticity and Secondary Recovery	   4-65
        Petroleum Reforming 	   4-66
        Petroleum Cracking	   4-66
        Nitric Acid Production	   4-66
        Chemical Processes Other Than Nitric Acid 	   4-67
        Telephone Switching Equipment 	   4-67
        Dental and Medical Uses	   4-68
        Electrical --  Other Than Telephone Switches 	   4-68
        Glass	   4-68
     Price Elasticities for Calculating the Critical ity of Platinum,
       Palladium, and  Rhodium 	   4-69
     Speculation and Increased Demand for Platinum-Group Metals .   .   4-69
     Appendix:  Annotated Bibliography and Guide to Sources of
       Information on  Platinum-Group Metals 	   4-72
     Chapter 4 References 	   4-76

-------
                                                                 Charles
                                                                 River
                                                                 Associates
                      TABLE OF CONTENTS (Continued)

                                                                     Page
Chapter 5:  RECYCLING OF PLATINUM-GROUP METALS FROM CATALYTIC
              CONVERTERS	   5-1

     Background on the Function of the Catalytic Converter	   5-2
     The Characteristics of Catalyst Material  	   5-2
        Cost Breakdown For New and Used Converters	   5-3
     Recycling Catalytic Converter	   5-5
        Phase I:  Rejected Catalysts	   5-5
        Phase II:  Replacement After 50,000 Miles 	   5-6
        Phase III:  Auto Catalyst From Salvaged Autos	   5-6
     Spent Catalyst PGM Refining	   5-9
        Future Availability of PGMs From Scrapped Converters. ...   5-11
     Chapter 5 References 	   5-12

Chapter 6:  SUBSTITUTES FOR PLATINUM-GROUP CATALYSTS IN VEHICULAR
              EMISSIONS CONTROL 	   6-1
     Background	   6-1
     Oxidizing Catalysts	   6-2
     NOX Removal Catalysts	   6-4
     Base-Metal  Catalyst Research 	   6-5
     Summary on Possible Replacement of Platinum-Group Metals .  .  .   6-6
     Chapter 6 References 	   6-8
                                   m

-------
                                                                 Charles
                                                                 River
                                                                 Associates
                                    H
                          INTRODUCTION AND SUMMARY
Technologies currently  being  employed  in the United States to control
vehicular emissions require large expenditures on imported platinum-group
metals, and smaller expenditures on other materials sometimes characterized
as "critical."   There is considerable  concern about the reliability of
foreign supplies of many of these materials, and also about implications for
the U.S. balance of trade.  In order to carry out its regulatory functions
efficiently, the U.S. Environmental Protection Agency must weigh these
concerns appropriately, which involves quantifying the important costs and
benefits to the United  States that are associated with materials consumption.
It is important to recognize  all the costs imposed by the potential
unreliability of foreign sources of materials supply, but it is also
important not to overestimate the importance of these costs (as some parties
may attempt to do when  it is  in their  interest).  In this study we explain
how EPA can quantitatively estimate the criticality of materials, and factor
those estimates directly into decisions about regulations controlling
vehicular emissions.

As groundwork for our analysis, subcontractor Rath and Strong projects the
quantities of platinum-group  metals and other potentially critical materials
that will be used for vehicular emissions control in the United States during
the 1980s.  In  order to obtain a full  assessment of the issue, we also
consider in some detail two further topics.  The first topic is the current
extent of recycling of  platinum-group  metals and stainless steel scrap from
                                    1-1

-------
                                                                 Charles
                                                                 River
                                                                 Associates
catalytic converters;  we also give particular attention to the extent of
recycling that is likely in the future.  The second topic is the
technological possibilities for using  emissions control technologies that
require smaller amounts of platinum-group metals.

The basic contents of this study are organized into five chapters following
this introductory chapter.  Chapter 2  develops our main conclusions and
recommendations regarding how the Environmental Protection Agency can factor
into its policy decisions their effects on material markets.  Our main
conclusion is that EPA must estimate compliance costs carefully, taking into
account likely increases in prices of  materials due to EPA-induced demands,
and also include some adjustment for the probability of large (but usually
temporary) increases in prices of materials due to future market
contingencies such as foreign supply disruptions.  By and large, however, it
is not reasonable to expect EPA to do  more than estimate likely future costs
of compliance as these depend on likely future costs of materials.
Detrimental effects on the U.S. balance of trade from increased importation
of materials to satisfy EPA regulations is a minor consideration that can
usually be given low priority by EPA.

U.S. policymaking in response to the criticality of materials is more
efficiently undertaken at the national level, if it is justified at all.
National tariffs and stockpiles are the appropriate policy instruments, not
ad hoc decisionmaking  by each individual government agency whose decisions
affect total  U.S. consumption of materials.

Chapter 2 defines a "critical  material" simply as one for which contingency
planning is worthwhile.  If the contingency is a military conflict, then the
material is also "strategic."   Chapter 2 then presents a comprehensive list
of nonmilitary contingencies that may  justify preparatory planning, that is,
nonmilitary contingencies that may be  the basis for considering a material
critical.  Contingencies that are potentially important from the perspective
of consumers include mine disasters, labor strikes, equipment failure,
depletion of reserves, new competing demands for a material, or just
processing bottlenecks due to unexpectedly large total  demand for the
material.  However, the type of contingency that most often makes a material
highly critical  is the threat of foreign supply disruptions.

Our analysis of the criticality of platinum-group metals and other materials
used for vehicular emissions control in fact concentrates upon the
unreliability of imported supplies.  The seven materials for which we
quantitatively estimate criticality are:
                                   1-2

-------
                                                                 Charles
                                                                 River
                                                                 Associates
•    Platinum;

•    Palladium;

•    Rhodium;

•    Chromium;

•    Manganese;

•    Nickel; and

t    Titanium metal.

Only in the case of titanium metal  are we more concerned about problems other
than foreign supply disruptions.

The quantitative measure of a material's criticality is the expected future
cost per year due to  the contingencies threatening the market, averaging out
years in which the contingencies  do and do not occur.  We provide
illustrative estimates for each of  the materials listed above.  In principle,
the expected cost of  contingencies  can be multidimensional, and include
noneconomic costs such as greater U.S. pollution from increased domestic
production of materials, or even  from relaxation of vehicular emissions
standards, were that  to be deemed likely.  However, for this study we
concentrate on the strictly economic costs associated with contingencies in
material markets, usually disruptions in foreign supplies.  Thus, we measure
criticality strictly  in terms of  expected dollar losses per year.

Criticality of a material  can be  measured from the perspective of a
particular end use, such as control of vehicular emissions, or from the
perspective of the nation as whole.  When criticality is calculated from a
national perspective, it must be  recognized that the same contingencies that
inflict costs on U.S. consumers of  materials will often benefit U.S.
producers of those materials.  This is generally the case where foreign
supply disruptions are the contingency of concern, so that some balancing of
criticality from the  perspectives of consumers versus producers is required.
For purposes of illustrative calculations, we assume a dollar gained by U.S.
producers of a material compensates for a dollar lost by U.S. consumers.

For a contingency such as a foreign supply disruption, the key parameters of
the criticality estimation that we  explicitly recognize in our calculations
are as follows:
                                    1-3

-------
                                                                 Charles
                                                                 River
                                                                 Associates
•    The severity of contingencies that threaten U.S. consumers (or
     producers)  of a material, as measured by price increases (decreases)
     that occur;

•    The expected time between occurrences of these contingencies;

•    The duration of these  contingencies;

•    The quantity of the  material consumed in the United States under normal
     conditions;

•    The quantity of the  material produced in the United States under normal
     conditions, both from  primary and secondary sources;

•    The extent to which  U.S. consumption can be reduced when prices rise
     (the "elasticity" of short-run U.S. demand);

•    The extent to which  U.S. primary and secondary production can be
     increased when prices  rise  (the "elasticity" of short-run U.S. supply);
     and

•    The size of normal U.S.  inventories and stockpiles.

The most important determinants  of the criticality that we estimate for the
seven materials listed earlier are subjective estimates of the severity,
frequency, and duration of  a  "representative" contingency for each market
(that is, the first three of  the items listed above).  Unlike the other
parameters of the critical ity estimates, it is  unfortunately not possible to
estimate these parameters with any precision from historical data or
engineering analysis.

Not surprisingly, from the  perspective of consumption for vehicular emissions
control it turns out that platinum is the most  critical of the materials we
consider, by more than an order  of magnitude.   However, the criticality
penalty is very small relative to the apparent  disadvantages of alternative
emissions control technologies,  as described later in the report.  Thus, the
criticality of platinum-group metals to the United States is not a strong
reason to discourage use  of these materials for emissions control.

Moreover, our simplified  methodology for calculating illustrative measures of
materials criticality does  not take into account the fact that current
consumption of platinum-group metals for emissions control creates a "rolling
stockpile" of the material  that  will increasingly be available through
recycling in the future.  If  this consideration were factored into the
analysis, using a more sophisticated economic model, the criticality of
platinum-group metals used  for emissions control would be considerably less.
(The same is true of the  alloying metals used in the 409 stainless steel of
catalytic converters, since most of that material will be recycled in the
future as well.)
                                    1-4

-------
                                                                 Charles
                                                                 River
                                                                 Associates
The analysis in Chapter 2 summarized  above draws on background information
presented in Chapters 3 through  6.  Chapter 3 summarizes projections of U.S.
materials consumption for vehicular emissions control developed by
subcontractor Rath & Strong.   Consumption of platinum and palladium is
projected to drop off somewhat from the high rates of 1980 and 1981, but
still represent a very sizable share  of total U.S. consumption of those
metals.  U.S. consumption of  rhodium  is projected to rise substantially by
the mid-1980s.  Consumption for  vehicular emissions control of other
materials is very small, both relative to total U.S. consumption of those
materials, and in terms of the share  of compliance costs for which it
accounts (or might plausibly  account  in during future times of shortages and
high prices).

Chapter 4 presents extensive  information on the markets for platinum-group
metals, particularly platinum, palladium, and rhodium.  The most important
considerations for this study are U.S. reliance on potentially unreliable
exports from South Africa and the Soviet Union, and the very large price
increases necessary to induce most users of these materials to reduce
consumption to any significant extent.  Ameliorating U.S. vulnerability to
foreign supply disruptions are substantial stockpiles maintained by most
consumers, and very high rates of secondary recovery, making most
applications of platinum-group metals interpretable as "stocks in use"
(rather than "consumption").

Chapter 5 assesses publicly available information about the young industry
recycling platinum-group metals  from  obsolete catalytic converters in the
United States.  Taking into account the cost of gathering and processing used
converters, and the losses and contamination that occur during use, recovery
of platinum-group metals from this source appears to be only moderately
profitable at 1980 prices for platinum-group metals.  Decreases in the prices
of platinum-group metals which would  be large by historical standards, but
are conceivable, could make recovery  at least temporarily uneconomical.

Finally, Chapter 6 discusses  possible substitutes for platinum-group
catalysts in vehicular emissions control.  There is no published evidence
that a catalyst system using  only base metals could match the performance of
the present three-way catalyst system, leaving aside the question of
durability.  It probably would be possible to design an oxidizing catalyst
system using only base metal  catalysts that would meet 1980 standards for
emissions of carbon monoxide  and hydrocarbons, but it would almost certainly
not meet 1981 standards.  The unit would have to be somewhat larger than
present emissions converters  using noble metals, and might well be more
costly at normal prices for materials.  Most avenues for using base metal
catalysts in place of noble metal catalysts were investigated in the early
1970s and found  (with a high  degree of assurance) to be too unpromising to
justify further  research.  General Motors contin-ies to support research on
zeolite catalysts by Professor Hall at the University of Wisconsin, but
results are much too preliminary to warrant optimism.
                                    1-5

-------
                                                                    Charles
                                                                    River
                                                                    Associates
                 MATERIALS CRITICALITY AND EPA POLICYMAKING


This chapter contains the main conclusions of our study, drawing on
background information presented in later chapters.  We first define what a
"critical" or "strategic" material  is, and then describe market contingencies
that can justify such a designation.  We explain how criticality can be
measured in terms of expected dollar losses per year, from the perspective of
consumption for vehicular emissions control or from the perspective of the
nation as a whole, and actually perform illustrative calculations for seven
elemental materials.  We describe how the criticality measure can be factored
into EPA policymaking, particularly through its role in estimation of
compliance costs for EPA regulations.  We also draw out implications for more
general EPA policy issues such as the likely adequacy of research on
emissions control undertaken by U.S. vehicle manufacturers, and possible
inadequacies in auto manufacturers' stockpiling of critical materials used
for emissions control.


                 CRITICAL MATERIALS AND STRATEGIC MATERIALS


The term "strategic and critical  material" was institutionalized in 1939 by
the original  legislation that established the U.S.  National Defense
Stockpile.  (See Appendix 2-A.)  The currently effective version of that
legislation (as amended in 1979)  defines strategic  and critical  materials as
those that "(A) would be needed to  supply the military, industrial, and
essential civilian needs of the United States during a national  emergency,
and (B) are not found or produced in the United States in sufficient
quantities to meet such a need."
                                      2-1

-------
                                                                    Charles
                                                                    River
                                                                    Associates


The more general term "critical material" has been used in a variety of
contexts by many different commentators and analysts.  Webster's New
Collegiate Dictionary defines "critical" in this context to mean:
"indispensable for the weathering, solution, or overcoming of a crisis," and
in fact gives "the stockpiling of strategic and critical  materials" as an
illustrative application of the term.  Operationally, a material tends to be
described as "critical" for national  policymaking if future events threaten
to inflict serious damage on the United States, and current planning and
policies can reduce the expected costs associated with the threat.  If the
threatening event is a military conflict, the material is also described
as "strategic."  Webster's New Collegiate Dictionary defines "strategic" in
this context to mean:  "required for the conduct of war," and again gives
"strategic material" as an illustrative application of the term.

Our usage of the terms "critical" and "strategic" is entirely compatible with
general usage, as expressed in the above dictionary definitions.  Since we
will mainly be concerned with nonmilitary contingencies in material markets,
we will focus on the "criticality" of materials.  Our primary objective is to
provide an operational, quantitative definition of materials criticality,
that is as decisive as possible for determining what materials should be used
for vehicular emissions control, and also for assessing the likelihood that
the private sector will  make the appropriate choices.  (Alternative
definitions of materials criticality, and methods of measuring it, are
briefly surveyed in Appendix 2-C and in the Bibliographic Note at the end of
this chapter.)

The reader should be warned that terminology tends to be variable and
changing in this area.  Currently, much publicity has been generated about
schemes through which private individuals can readily invest in stockpiling
of "strategic metals."  In this context, a strategic metal  tends to be an
imported metal with defense applications whose market volume is not large and
whose market price is volatile.


                 MARKET CONTINGENCIES THAT MAY BE THE BASIS
                    FOR CONSIDERING A MATERIAL CRITICAL


For the U.S. Department of the Interior, CRA is currently conducting a study
of methods for detecting and evaluating emerging problems in material
markets.*  One early output of that study was a comprehensive categorization
of issues, problems, and contingencies in material  markets that might inflict
losses on U.S. citizens, and thus be of concern to the Department of the
Interior or other federal  agencies.   The discussion that  follows draws on
this related project to describe general types of market  contingencies,
*See Charles River Associates (forthcoming).
                                    2-2

-------
                                                                   Charles
                                                                   River
                                                                   Associates


preparations for which may benefit from special  planning by private firms and
government agencies.  In other words, we now categorize and describe  market
contingencies that can be the basis for considering a material  "critical."

Despite the potentially broad applicability of the concept of materials
critical ity, there is one type of nonmilitary contingency that outstrips the
others in prominence, namely disruptions in foreign supplies of materials
that the United States imports.  This preeminence of foreign supply
disruptions as the market contingency of concern is particularly true for the
materials of central interest in this study:  platinum-group metals are
supplied to western markets predominantly by South Africa and the Soviet
Union, neither of whose reliability is unquestioned, though the nature of the
contingencies of concern is quite different for the two countries.   Other
materials used in vehicular pollution control  equipment, such as chromium for
stainless steel, are also obtained from potentially insecure foreign  sources.
(For example, South Africa is the largest exporter of chromium as well as
platinum; the Soviet Union was an important exporter of chromium before 1975,
and Albania is now an important supplier to the West.)

We now systematically review all the major types of market contingencies, the
threat of which could conceivably be the basis for considering a material
"critical."
DISRUPTED FACTORS OF PRODUCTION

Production of minerals, like other goods and services, is interpreted by
economists to depend ultimately on the use of three main types of inputs or
"factors of production":  land, labor, and capital.  In addition, other
materials, transportation services, etc. are purchased from other firms (who
themselves use land, labor, and capital).  Disruption of any essential  factor
of production can stop production or delivery of a material, though
disruption of some factors are considerably more likely than others.
(Foreign producers, particularly those in less developed countries, are
considerably more prone to disruptions of this type than are domestic
producers; we distinguish problems with foreign production as a separate
category below.)

The most notable "land" used in the production of materials is of course the
mineral deposits from which metals and other materials are produced.
Deposits can be made suddenly unavailable through natural  disaster, mine
accidents, or even sabotage.  However, this type of disruption has not been
very important historically in the United States.

By far the most common contingency affecting the availability of labor for
mineral production is the deliberate strike, often organized by labor unions.
One can conceive of other contingencies, such as disease,  having an effect,
but such events have not been important historically.
                                     2-3

-------
                                                                   Charles
                                                                   River
                                                                   Associates
Once capital equipment is installed, it can break down and disrupt
production.  For major metals, this type of risk is spread so widely  as  to
render it unimportant.  For materials produced from only a few sources,  it
can be a problem from time to time, though usually one that is only temporary
and modest in proportions.

Materials and services purchased from other firms can also disrupt producers
of materials.  Transportation routes can be severed by landslides  or
breakdown of equipment.  There may just be sudden competition for  cargo  space
from highly valued commodities that can afford to pay more for
transportation.  For example, in South Africa, platinum is sufficiently
valuable as to justify transportation of refined material  by airplane if
necessary.  However, other minerals produced in South Africa, such as
chromium ore, are much more bulky, and shipments have at times been
significantly delayed by the priority shipment of other goods, such as
agricultural commodities in season.

Sharp increases in the prices of some inputs can be as disruptive  to  mineral
production as cutoffs in other inputs.  The most prominent example in recent
years has been the sharp increase in the prices of petroleum and other forms
of energy.  OPEC has claimed at times that it was embargoing the United
States and other regions of the world, but there is little evidence that this
strategy has ever been effective (in the sense of imposing much greater  costs
on the embargoed regions compared to other importing countries).   Energy
availability may also be subject to more localized disruptions. For  example,
hydroelectric production of energy can be disrupted by low rainfall.


INADEQUACY OF RESERVES AND RESOURCES IN THE LONG RUN

For a few materials, published estimates of reserves and resources are
sufficiently low so there might appear to be some chance of "running  out" in
a decade or two, before its use to meet EPA regulations has terminated.   (A
"resource" qualifies as a "reserve" if production therefrom is economically
viable.)  This fear is usually misplaced, because exploration can  expand
reserves, and new technologies can make economical the exploitation of
previously uneconomical deposits.  Because it generally emerges so slowly,
this problem is usually not included in the calculation of materials
criticality, though it should be considered when estimating the likely future
cost of complying with EPA regulations.  Thinking in terms of "contingency
planning" is usually not that useful in this case.


TECHNOLOGICAL SHOCKS

The technological contingency of greatest concern to current consumers of a
material is usually the possibility that a very large new use will emerge,
resulting in an escalation of price.  Fortunately, consumption of  materials
                                     2-4

-------
                                                                   Charles
                                                                   River
                                                                   Associates


for new technologies tends to grow sufficiently slowly, or with  sufficient
warning, so that major price increases can usually be avoided through
development of new productive capacity.   The use of platinum as  a catalyst
for pollution-control  equipment is a case in point.  On the other hand,
rhodium is produced as a byproduct, and has experienced relatively greater
price increases because supply is limited by the amount of platinum
production.  Like inadequacy of reserves, this type of problem tends  to
emerge sufficiently slowly so that it is usually not explicitly  included in
the calculation of a material's criticality, though it should be factored
into the analysis of materials costs at some point, as we discuss further
below.

DISRUPTED FOREIGN SOURCES OF SUPPLY

Foreign producers of materials, particularly those in less developed  regions,
are subject to all the contingencies described above, often to a considerably
greater extent than domestic producers.   Localized military conflicts and
sabotage can cause major disruptions.  Also, foreign producers of materials
are often not constrained by law from acting in a glaringly monopolistic
fashion, and their governments may even attempt to use their exports  as  a
political weapon.

There was concern during the 1970s that foreign exporters of minerals and
metals would use OPEC as a model and form effective cartels.  Foreign
producers of metals such as aluminum and copper have attempted to increase
their joint monopoly power by forming producers' associations, but these
organizations have had very limited success.  (For an extensive analysis of
the difference between OPEC and mineral  producers, and the limited ability of
the latter to collude, see Charles River Associates (1976).)  If foreign
producers of a mineral or metal organize effectively, there is a theoretical
possibility of "price gouging," where prices are raised very high, very
rapidly, in order to catch consumers before they can change their consumption
patterns.  However, most foreign producers appreciate sufficiently the
detrimental long-term effects of such a policy, so that it has not been
common.

Most monopoly power in mineral markets is due to "natural" monopoly power,
stemming from the fact that one country has low production costs and  a
dominant market share.  This situation is often not too damaging to consumers
over time, since the dominant producer must keep prices low enough to
preclude entry by major competitors.  Also, this situation tends to be quite
stable, as the low-cost dominant producer is greatly interested in
encouraging consumption of his material.  Monopoly power can be exercised by
private companies or by foreign governments.  One of the neatest ways for a
foreign government to accomplish this objective is simply to impose tarrifs
on exports of a mineral, which raises the world price, and funnels monopoly
profits directly into the national treasury.  Caribbean producers of  aluminum
ores have favored this technique.  (The U.S. Constitution forbids export
tariffs in this country.)
                                     2-5

-------
                                                                   Charles
                                                                   River
                                                                   Associates


Moderate collusion among foreign producers is often tacit rather than overt.
This is probably the case in the platinum market.   South African producers
are not explicitly organized, but each realizes that an attempt to expand its
market share may invite retaliatory expansion by its competitors.   As a
group, they are better off if capacity expansions  are planned
conservatively.

If, as is often the case, monopolistic aspects of  foreign production  are
stable over time, they may not give rise to sudden contingencies (such as
price gouging) and criticality planning of the type we develop below  may not
be necessary.  Monopolistic world prices will be continually somewhat higher
than they would be in a competitive environment, but consumers may rationally
come to accept that situation as a fact of life, rather than a contingency
subject to sudden reversal that requires planning.

Petroleum exports have in recent years been used with some success as a tool
of foreign policy, but there is no analogous example among mineral markets.
The value of world petroleum imports dwarfs trade  in even major metals such
as iron, copper, and aluminum.  The Soviet Union stopped exporting metals to
the United States prior to the Korean War, but no  serious damage was
inflicted on this country.  The United Nations attempted to impose an embargo
on Rhodesian chromium during the 1960s, but it was not enforceable.


THE BUSINESS CYCLE

The rate of economic activity in western market economies fluctuates
considerably, inducing large changes in the consumptions and prices of
materials.  High prices (or even unavailability) of materials due  to  booming
competing demands can seriously affect some consumers in much the  same way as
a foreign supply disruption.  However, consumers with long-term contracts
with suppliers of materials may be somewhat protected.  Materials  whose
consumption is particularly sensitive to the rate  of economic activity are
sometimes considered "critical," but this is far from a universal  practice.
(We later consider the criticality of titanium metal from this viewpoint.)


DEFENSE

The premier contingency making a material "critical" is war.  As discussed
above, the material is called "strategic" in that  case.  The costs imposed on
the United States by most of the contingencies described above are
predominantly economic, and so the resulting criticality can be analyzed
using economic theory.  The costs of losing a war  cannot be measured  in
purely economic terms, so determining the extent to which a material  is
strategic involves other considerations with which we do not attempt  to deal.
Our analysis of materials used to meet EPA regulations will concentrate upon
peacetime contingencies.
                                     2-5

-------
                                                                   Charles
                                                                   River
                                                                   Associates
                PRINCIPLES OF MEASURING MATERIALS CRITICALITY
                    AND IMPLICATIONS FOR EPA POLICYMAKING
Regulations implemented by the Environmental  Protection Agency affect U.S.
material markets in a variety of ways.   The most widely publicized effects  on
material markets involve direct regulation, such as those telling  producers
of materials the maximum amounts of specified air pollutants they  are allowed
to emit.  However, EPA rulemaking can also affect material  producing
industries more indirectly, simply by increasing the demand for materials
which other industries need to meet EPA regulations.  This study deals with
this latter issue, in particular increased demands for platinum-group metals
and other materials used by motor vehicle manufacturers to meet EPA emissions
standards.

Obviously, some important aspects of the availability of materials for EPA
regulations are normally factored into EPA's rulemaking processes, in
particular the prices of materials as they affect projected costs  of
complying with the rules.  Once EPA standards have been set, the regulated
industry has a continuing incentive to minimize the cost of compliance,
taking  into account the prices of materials.  The key question we  consider  in
this chapter is whether there is some aspect of the availability of
materials, beyond the inclusion of their price in estimated compliance costs
(such as their "criticality"), that EPA should take into account when making
rules and regulations.  Our ultimate conclusion to this question in most
cases is "no."  However, there can be exceptions, and in any case there is
often controversy surrounding decisions on this issue.  Thus, in the
remainder of this chapter we construct a fairly elaborate rationale for our
conclusions, and also describe the possible exceptions.
MATERIALS PRICES AND COMPLIANCE COSTS

Predicting future compliance costs for a new or proposed EPA regulation, as
it depends on material prices, should not be a matter of simply referring to
the most recent price quote in Metals Week or the Chemical  Marketing
Reporter.  The current price of a material may be significantly above or
below the long-run market equilibrium, often because general economic
activity in the industrialized consuming countries (as measured by their
GNPs, for example) happens at the moment to be significantly above or below
the historical trend.  Or there may be some transient supply problem, such as
a labor strike or a transportation bottleneck, causing current prices to be
significantly above prices that are likely in the future.

Usually, the most relevant basis for predicting future compliance costs is
the long-run equilibrium price of a material, where producers are earning an
adequate, but not excessive, rate of return.  Predicting long-run equilibrium
                                     2-7

-------
                                                                    Charles
                                                                    River
                                                                    Associates


 prices for materials under normal  market conditions  is  not a  trivial  task  if
 the highest possible degree of accuracy  is required.  However,  it is  a
 manageable problem.   Here, we concern  ourselves  with  whether  these  are
 important further issues  for EPA to  consider,  beyond  simply accepting the
 normal  cost of materials  implied by  the  methods  of compliance chosen  by  the
 regulated private sector.

 Where EPA regulations will  require relatively  large  increases in  world
 production of a material,  one obvious  complication is the  effect  of EPA
 induced demands on the long-run equilibrium price of  the material.  For  most
 materials, long-run  supply is very "elastic" in  its response  to price
 incentives.   In economists'  jargon,  that means that a very  modest increase  in
 the market price of  a material  will  eventually be a sufficient  incentive  for
 its producers to increase  greatly  their  output,  allowing a  number of  years
 for unrushed capacity expansions,  and  perhaps  even additional time  to explore
 for new reserves.   It may  even  be  possible to  increase  greatly  the  production
 of a material  and eventually  to have a lower market price  than  previously, if
 expanded production  allows  increased economies of scale or  induces  advances
 in processing technologies.   The most notable exceptions to the above
 generalization  (that materials  are available in  very elastic supply in the
 long run),  occur when  a material is produced as  a byproduct of another
 material.   In  that case, increased market  prices may lead to very little
 additional  production.  This  consideration  is particularly  relevant for the
 more minor of the  platinum-group metals, most notably rhodium.

 In  addition  to  setting regulations to be met in  the "long run,"  EPA must  also
 decide  upon  how quickly to  impose  standards of a given  stringency.  More
 rapid imposition  of  a  standard  may lead  to  more  rapid increases in  the demand
 for materials  and  short-run increase in  their prices.  These short-run price
 increases  will  recede  after capacity to  produce  the materials has expanded,
 but they do  imply  temporarily higher compliance costs from  faster
 implementation  of  regulations.  For simplicity in the following  discussion,
 we  usually abstract  away from the additional difficulties associated with
 analyzing  the speed  of implementing EPA regulations,  and consider only the
 costs of regulations after they have been in effect for a while.  However,
 our  conclusions about the adequacy of considering materials criticality  only
 to  the extent that compliance costs are affected, also generally hold true
 when  evaluating the overall costs (and benefits)  of different possible speeds
 of  implementation.

 A final consideration relevant for predicting the costs  of materials used
 to comply with EPA regulations is the probability of  contingencies such  as
major foreign supply disruptions, that can greatly increase the  price of a
material for a number of months, or even years.  As  discussed  above,
 significant susceptability to such contingencies  qualifies  a material  to be
                                    2-8

-------
                                                                   Charles
                                                                   River
                                                                   Associates


considered "critical."  We now consider how such considerations should in
principle be included in the normal  course of predicting compliance costs for
EPA regulations, which requires analyzing how private firms include these
considerations in their choices among alternative methods for complying with
EPA regulations.  This same analysis will be the basis for quantifying the
criticality of a material.


COMPLIANCE COSTS AND THE CRITICALITY OF MATERIALS
USL-U TO MELT EPA REGULATTUNF	

Figure 2-1 presents a concrete example of the materials component of
compliance costs, in terms of the amount of an imaginary material
"catalystium" used to meet EPA vehicular emissions standards.  The
illustrated catalystium "demand curve" assumes that when the market is in a
normal undisrupted state, the world market price is $100 per ounce, and
10,000 ounces per year are purchased to meet EPA regulations.  However, we
suppose that there is a threat that the producers of catalystium located in
a foreign country will be disrupted by a localized military conflict.   For
the sake of simplicity, we assume we know that the world market price  rises
from $100 per ounce to $300 per ounce during such disruptions,  and that the
disruptions last exactly one year.  We further assume that the  probability of
a disruption occurring in any future year is 0.1.

If producers of emissions control equipment continued to purchase 10,000
ounces of catalystium per year during disruptions, then it is very easy to
include the effect of the supply disruptions in the calculation of average
expected compliance costs in future years.  The expected future price  of
catalystium would be 0.9($100) + 0.1($300) = $120 per ounce,  averaging out
years in which supply disruptions do and do not occur.

However, the example illustrated in Figure 2-1 is a bit more  realistic.  It
assumes that producers of emissions control  equipment can cut back somewhat
on the use of catalystium when its price rises suddenly.  A variety of
design changes may allow reductions  in the use of catalystium,  but for
simplicity we can assume here that simply using a greater proportion of
another material  in the equipment allows EPA regulations to be  met. Using
this greater proportion of the alternative material  is not economical
(i.e., does not minimize compliance  costs) when catalystium costs  $100 per
ounce, but it is economical  when catalystium costs $300 per ounce,  and (we
suppose) the switch in technologies  can be made rapidly after such  a price
increase occurs.

How can the calculation of compliance costs  take into account disruptions  in
the supply of catalystium in this more complicated case?  One straightforward
                                     2-9

-------
                                                    Charles
                                                    River
                                                    Associates
Figure 2—1


CONSUMPTION OF "CATALYSTIUM" TO COMPLY WITH EPA STANDARDS,

AND ECONOMIC LOSSES FROM SUPPLY DISRUPTIONS

(A HYPOTHETICAL EXAMPLE)
PRICE ($/ounce)
300 -
200 -
100
                         5           8       10


                        QUANTITY CONSUMED (1000 ounces/year)
                          2-10

-------
                                                                    Charles
                                                                    River
                                                                    Associates
 way to proceed  would  be  to calculate  costs  for  each  of  the  alternative
 technologies  separately,  then  weight  the  cost using  less  catalystium by 0.1,
 and the cost  of the normal  technology by  0.9.   (That is,  weight by  the
 relative frequencies  of  a  disruption  occurring  and not  occurring.)  However,
 it is  also  possible to infer the effect of  disruptions  in the supply of
 catalystium on  compliance  costs, just using  the demand  curve for catalystium
 illustrated in  Figure 2-1.

 In  order to appreciate how this is done,  consider the more  artificial demand
 "curve"  for catalystium  illustrated in Figure 2-2, which  assumes that at a
 catalystium price of  exactly $200 per ounce  it  is economical to switch
 entirely from the normal  technology to the  alternative  using less
 catalystium.  The resulting demand curve  is  a step function.  (The  more
 realistic smooth demand  curve  in Figure 2-1  assumes  that  the switch away from
 catalystium occurs gradually as prices rise  from $100 per ounce to  $300 per
 ounce.)   In Figure 2-2,  $200 per ounce of catalystium represents a
 "break-even"  price at which it is equally economical  to use either  of the two
 technologies  to meet  EPA  standards.  We can  tell from the diagram that the
 normal  technology costs  in  total ($200-$100)(10,000)  = $1,000,000 more per
year,  when  the  catalystium  price is $200  rather than  $100.  Thus, at the
catalystium price of  $200 per  ounce, the  alternative technology using less
catalystium must also cost  $1,000,000 more than the  normal technology with
catalystium at  $100 per ounce.   Of this $1,000,000, additional  costs for the
8,000  ounces  of catalystium used with the alternative technology are
 ($200-$100)(8,000) =  $800,000  per year.  Thus,  the cost of changing
technologies, apart from the effect of an increase in the price of
catalystium,  is $1,000,000  - $800,000 = $200,000 per  year.

This line of  reasoning is probably clearer in the geometric terms of
Figure 2-2.    When the catalystium price rises from $100 to just below $200,
the regulated industry continues to buy 10,000 ounces at an additional  cost
for the year  of almost $1,000,000.   This loss is represented geometrically in
Figure 2-2  as the area of the  rectangle made up of the two smaller rectangles
labeled 2a  and  1.  When the catalystium price rises from just below $200 to
just above  $200, the  total cost of compliance does not increase
significantly, but switching to the alternative technology causes the costs
to be  broken  down into

•    the additional  $200,000 cost of the alternative  technology,  represented
     as Area  1 in Figure 2-2,  plus

•    the additional  $800,000 cost of the remaining amount of catalystium
     that is purchased,  represented as Area 2a  in  Figure 2-2.
                                     2-11

-------
                                                        Charles
                                                        River
                                                        Associates
Figure 2-2



CONSUMPTION OF "CATALYSTIUM" TO COMPLY WITH EPA STANDARDS:

ARTIFICIAL CASE WHERE DEMAND IS A STEP FUNCTION
PRICE ($/ounce)
300 -
200 -
100
                                      I        I
                   2b
                   2a
                                      I
                         5            8        10


                            QUANTITY CONSUMED (1000 ounces/year)
                           2-12

-------
                                                                   Charles
                                                                   River
                                                                   Associates
If the catalystium price then rises from $200  per  ounce  to $300 per ounce,
the remaining 8,000 ounces of catalystium that are purchased will cost an
additional ($300-$200)(8,000) = $800,000 per year,  represented in Figure 2-2
as Area 2b.  (We are assuming in Figure  2-2 that no further reductions in the
use of catalystium are economical  in this price range.)  Thus, the total
increase in the cost of compliance caused by the catalystium price rising
from $100 to $300 during supply disruptions is $200,000  in adjustment costs
(represented by Area 1), plus $1,600,000 in additional costs for 8,000 ounces
of catalystium (represented by Areas 2a  and 2b).   The cost savings achieved
by switching to the alternative technology (rather than  just continuing to
buy 10,000 ounces of catalystium at $300 per ounce), is  represented in
Figure 2-2 as Area 3, amounting to $200,000 for the year of the disruption.

The above line of reasoning generalizes  readily to the case where
adjustment away from consumption of catalystium is gradual, as its price
rises from $100 per ounce to $300 per ounce during supply disruptions.  This
more realistic case is illustrated in Figure 2-1.   Again, Area 1 ($200,000)
represents adjustment costs and Area 2 ($1,600,000)  represents the additional
cost of purchasing the remaining 8,000 ounces  of catalystium during
disruptions.  Area 3 ($200,000) represents the cost savings achieved by
adjusting to alternative technologies, rather  than just  continuing to consume
10,000 ounces per year during supply disruptions.

We now have sufficient information about the much  simplified example of
catalystium to illustrate the preferred  approach to measuring its criticality
for vehicular emissions control, due to  the threat of foreign supply
disruptions.  Remember our earlier assumption  that the disruptions are
expected to occur on average in one year out of ten.  Relative to the
situation where normal price $100 occurs with  certainty, the expected
additional costs due to supply disruptions are (0.1) ($1,800,000) + (0.9)
($0) = $180,000 in each future year.  (That is, additional costs of
$1,800,000 are borne on average in one year out of ten,  and no additional
costs are borne in nine years out of ten.)  This expected economic cost per
year is the quantitative measure of the criticality of  catalystium
consumption for vehicular emissions control.

It is clear from the above catalystium example that there are at least five
basic determinants of the criticality of a material  from the point of view of
consumption for vehicular emissions control.   (We  will discuss other
considerations in more general terms later.)   The  five determinants are:

•    the severity of contingencies that  threaten U.S. consumers, as measured
     by price increases that occur;

t    the probability of the contingencies occurring;

•    the duration of the contingencies;
                                     2-13

-------
                                                                   Charles
                                                                   River
                                                                   Associates

•    the quantity of calatystium consumed in  the  United States at normal
     prices; and

•    the extent to which U.S.  consumption can be  reduced  when prices rise
     (the "elasticity" of short-run U.S.  demand).

It is interesting to note that the normal  price of catalystium does not
directly enter the calculations, except as a  base from which to calculate
plausible price increases during disruptions. Thus,  for  example, simply
examining the cost share of various materials used for vehicular emissions
control  may not be a reliable  guide to their  respective critical!'ties, though
there is some relationship between cost shares and criticality, as we explain
further below.

The above calculation of criticality was  simplified in a  number of important
respects, notably in ignoring  the effects of  inventories  and recycling.  We
allowed the producers of emissions control  equipment  to switch to more
economical alternate technologies when the price  of catalystium jumped, but
we did not allow them to accumulate an economical level of  inventories in
preparation for disruptions.  Also, we did not allow  them to use recycled
scrap to a greater extent.  These considerations  deserve  further discussion,
but it is convenient to discuss first criticality from a  national
perspective, rather than just  from the perspective of producers of emissions
control  equipment.


THE CRITICALITY OF MATERIALS FROM A NATIONAL  PERSPECTIVE

For other consumers of catalystium, criticality is measured in exactly the
same fashion as for producers  of vehicular emissions  control equipment.  In
Figure 2-3, the demand curve for producers of vehicular emissions control
equipment is reproduced on the left, and another  (also hypothetical) demand
curve for other consumers is given in the middle  of the figure.  The example
assumes that it is economical  for other consumers of  catalystium to cut back
on their consumption by 60 percent in response to tripled prices during
supply disruptions, in contrast with producers of vehicular emissions control
equipment, who find it economical to cut back consumption by only 20 percent.
As a result, adjustment costs  for other consumers during  supply disruptions,
represented as Area Ib in Figure 2-3,  are larger  than corresponding Area la
for producers of vehicular pollution control  equipment.   Triangular Area Ib
represents adjustment costs of (1/2)($300-$100)(10,000-4,000) = $600,000 per
year, while Area la represents adjustment costs of only $200,000 per year (as
previously calculated).  But,  of course,  "other"  consumers of catalystium
benefit from their greater flexibility by having  to pay much less for
catalystium during supply disruptions:  Area  2b represents additional costs
of only ($300-$100)(4,000) = $800,000 per year, as opposed to Area 2a for
producers of vehicular emissions control  equipment, which was previously
calculated to be $1,600,000 for the year  of a disruption.
                                     2-14

-------
             Figure 2—3

             U. S. CONSUMPTION OF "CATALYSTIUM" FOR ALL END USES, AND
             TOTAL ECONOMIC LOSSES FROM SUPPLY DISRUPTIONS
            CONSUMPTION FOR
            VEHICULAR EMISSIONS
            CONTROL EQUIPMENT
                                CONSUMPTION FOR
                                OTHER END USES
TOTAL U. S.
CONSUMPTION
            PRICE ($/ounce)
ro

en
300 -
            100 -
                                                                         0           12       20

                                                                      QUANTITY CONSUMED (1000 ounces/year)
                                                                                                      Charles
                                                                                                      River

-------
                                                                   Charles
                                                                   River
                                                                   Associates

On the right side of Figure 2-3,  total  U.S.  consumption  of catalystium  is
obtained by summing (horizontally, at each price)  consumption  for  vehicular
control equipment and consumption for other end users.   (Note  that the
quantity scale on the total consumption graph  has  been compressed.)  The
reader can easily confirm that costs during  disruptions  are additive across
all end uses.  That is, the national adjustment cost, represented  by Area 1,
is the sum of Areas la and Ib ($800,000),  and  the  remaining national
expenditure for catalystium consumption during disruptions, represented by
Area 2, is the sum of Areas 2a and 2b ($2,400,000).  Total  national costs
borne by U.S. consumers during disruption  years are  thus $800,000  +
$2,400,000 = $3,200,000.

If the United States produces no catalystium,  and  could  not do so  even  when
the price of imports triples for a year, then  sufficient information on  the
above example has been given to calculate  the  criticality of catalystium
from a national perspective.  Remembering  that disruptions are assumed  to
occur on average in one year out of ten, expected  U.S. losses  from
disruptions are (0.1)($3,200,000) = $320,000 per year.

However, suppose that the United States produces catalystium in normal  times,
and could expand output somewhat during disruptions  in foreign supplies.  In
that case, the criticality of catalystium  will  be  less from a  national
perspective, though consumers will still face  the  same expected losses
(assuming the probability and severity of  price increases from disruptions
are as before).  The calculation of national criticality in terms  of expected
losses yields this result, by recognizing  that U.S.  producers  benfit greatly
from foreign supply disruptions,  thus facing negative criticality  from  the
threat of this particular contingency.   We now describe  how the "criticality
calculus" described above can be extended  to yield this  result.

Figure 2-4 gives a hypothetical  U.S. supply  curve  for catalystium,  showing
production of 4,000 ounces per year at the normal  price  of $100 per ounce,
and production of 5,000 ounces during years  in which supply disruptions
occur, when the price on the world market  is assumed to  be $300 per ounce.
How much additional benefit do U.S. producers  receive as a result  of
producing 5,000 ounces at $300,  rather  than  4,000  ounces at $100?   It is
clear revenues rise from ($100)(4,000)  = $400,000  to ($300)(5,000)  =
$1,500,000, but the additional  cost of producing 1,000 more ounces  must be
netted out.

This additional cost to producers can be calculated  by estimating  how much
additional production would occur at prices  between  $100 and $300  (in a way
analogous to measuring additional costs to consumers in  Figure 2-2).
Figure 2-5 gives a more artificial U.S. supply curve, specifying that the
additional U.S. production of 1,000 ounces per year  all  kicks  in at $200 per
ounce.  According to this supply  curve, the  additional labor,  materials,
energy, and other factors of production required to  produce an additional
1,000 ounces of catalystium cost  $200 per  ounce produced,  so it is  economical
(profitable) to produce the additional  quantity when the market price is
above $200 but not when it is below $200.  Thus, the total  cost of  producing
                                     2-16

-------
                                                       Charles
                                                       River
                                                       Associates
Figure 2—4


U. S. PRODUCTION OF "CATALYSTIUM," AND ECONOMIC GAINS FROM

FOREIGN SUPPLY DISRUPTIONS
PRICE ($/ounce)
300 -
200 -
100 -
                           QUANTITY PRODUCED (1000 ounces/year)
                            2-17

-------
                                                     Charles
                                                     River
                                                     Associates
Figure 2—5


U. S. PRODUCTION OF "CATALYSTIUM": ARTIFICIAL CASE WHERE

SUPPLY IS A STEP FUNCTION
PRICE ($/ounce)

300-
200 -
100-



	
	 	
	
J

••••
1


I
1






0 45
                          QUANTITY PRODUCED (1000 ounces/year)
                          2-18

-------
                                                                   Charles
                                                                   River
                                                                   Associates

the additional  1,000 ounces Is ($200)(1,000) =  $200,000 for the year of the
disruption.

In general, even for the more realistic  smooth  supply  curve in Figure 2-4,
the cost of additional  production can  be estimated  as  the area under the
supply curve, out to the point at which  production  occurs.  Thus, in
Figure 2-4 the cost of additional  production is represented as trapezoidal
Area 1, equal to (1/2)($100 + $300)(5,000 - 4,000)  = $200,000.
U.S. producers thus gain Area 2 in Figure 2-4,  equal to
($1,500,000 - $400,000) - $200,000 = $900,000,  during  years in which the
price of catalystium rises to $300 per ounce.   (If  producers just continued
to produce 4,000 ounces during a disruption year, they would only gain
$800,000.)

We are now in a position to recalculate  U.S. losses during years in which
foreign supplies of catalystium are disrupted,  recognizing the fact that U.S.
production will expand.  Figure 2-6 combines the total U.S. demand curve of
Figure 2-3 and the U.S. supply curve of  Figure  2-4.  When the world market
price is at the normal  level  of $100 per ounce, U.S. consumption is 20,000
ounces and U.S. production is 4,000 ounces, requiring  net imports of 16,000
ounces per year (the horizontal distance between the supply curve and the
demand curve).   At the disruption price  of $300 per ounce, U.S. consumption
is 12,000 ounces and.U.S. production is  5,000 ounces,  requiring net imports
of 7,000 ounces per year.

The U.S. loss areas described in earlier figures are renumbered in
Figure 2-6.  Area 1 ($800,000) is adjustment costs  suffered by U.S. consumers
in order to reduce consumption from 20,000 ounces to 12,000 ounces.  Areas 2,
3, and 4 together ($2,400,000) represent extra  payments by U.S. consumers for
the remaining 12,000 ounces that are purchased  at the  high $300 price rather
than at the normal $100 price.  Of that  total additional transfer to (all)
suppliers by U.S. consumers,  Area 4 ($900,000)  accrues to U.S. producers as
increased profits and Area 3  ($100,000)  represents  additional revenues of
U.S. producers used to cover increased production costs (beyond the $100 per
ounce that consumers normally pay).   Loss Area  2 ($1,4000,000) accrues to
foreign producers as extra payment for the 7,000 ounces of catalystium that
are still imported.

We should mention that the method described above for  measuring the cost
imposed by a large increase in the price of a material may require
supplementation where adjustment to the  disruption  involves dismissal of
workers.  The standard assumption implicit in the above methodology is that
these workers can find alternative employment at comparable wages.  Where
this assumption is significantly overoptimistic, the additional cost of
unemployed labor should be added when  calculating the  criticality of a
material from a national  perspective.
                                    2-19

-------
                                                    Charles
                                                    River
                                                    Associates
Figure 2-6


TOTAL U. S. CONSUMPTION AND PRODUCTION OF "CATALYSTIUM," AND

NET ECONOMIC LOSSES FROM FOREIGN SUPPLY DISRUPTIONS
PRICE ($/ounce)
300	1	
200 -
100
                                  QUANTITY (1000 ounces/year)
                           2-20

-------
                                                                   Charles
                                                                   River
                                                                   Associates

It may have already occurred to the reader that  if  the  United States started
off as a net exporter of catalystium,  rather  than a net importer, then the
gains accruing to U.S producers from a foreign supply disruption will be
larger than the losses suffered by  U.S.  consumers,  as U.S. producers export
more onto the world market at much  higher  prices.   Thus, a material can be
quite critical to U.S. consumers, and yet  have "negative" criticality from a
national perspective.  This balancing of gains and  losses by various groups
within the United Sates requires closer  examination, to which we now turn.


INCOME REDISTRIBUTION AND NONECONOMIC DIMENSIONS OF MATERIALS CRITICALITY

Netting gains of U.S. producers against  losses of U.S.  consumers, each
calculated in dollar terms, is a common  procedure,  but  it may be preferable
for purposes of national policymaking to keep separate  accounts of  these
gains and losses.  For example, as  a value judgment, legislators may feel
that a dollar gained by U.S. producers is  not as important as a dollar lost
by U.S. consumers.  The implied value judgment is that  a redistribution of
income from U.S. consumers to U.S.  producers  is  undesirable, rather than
being the neutral consideration that the netting procedure would require.

If the analyst keeps separate accounts of  losses and gains by U.S.  consumers
and producers, then a "multi-dimensional"  measure of criticality results.
The user of the multi-dimensional measure  can then  apply his or her own
weights to losses suffered by various groups  of  consumers and producers, in
order to calculate a single summary measure of the  criticality of various
materials (as is generally required to make final policy decisions).  But,
of course, this summary measure will  generally be somewhat different from
that which results from applying another person's "weights"  (value
judgments).

Other, noneconomic effects of disruptions  in  material markets may make it
desirable to measure the criticality of  materials in additional dimensions
that are not even denominated in dollar  terms.   Continuing with our earlier
example, suppose that increases in  the world  price  of catalystium from $100
to $300 causes Congress to relax vehicular emissions standards.  In that
case, consumption of catalystium would decrease  during  disruptions  more than
previously, and, as our earlier diagrams indicate,  the  direct economic losses
from disruptions measured in dollar terms  would  be  less.  However,  the
noneconomic effects of increased vehicular emissions due to relaxed standards
would be considered a cost of the disruption  by  most of the U.S. populace.
Some estimate of increased air pollution would then be  an appropriate
additional dimension for a criticality measure used for national
policymaking.  The prime example of a noneconomic dimension of materials
criticality concerns its usefulness for  military contingencies (that is, the
extent to which it is "strategic").  Obviously,  the cost to the United States
of being less well prepared for war cannot be measured  entirely in  dollar
terms.
                                     2-21

-------
                                                                   Charles
                                                                   River
                                                                   Associates

It is necessary to understand that a measure  of materials criticality can be
extended to include noneconomic dimensions, but we cannot explore all such
possibly useful generalizations here.   Hereafter we will simply assume that
all costs imposed on U.S.  citizens by contingencies of concern in the
catalystium market are strictly economic,  and that a  dollar gained  by a U.S.
producer compensates for a dollar lost by  a U.S. consumer.  Under these
assumptions, we can conclude that the criticality of  catalystium in the case
allowing for U.S. production (as illustrated  in Figure 2-6) equals  expected
annual loss (0.1)($3,200,000 -$900,000) =  $230,000 per year (that is,
averaging out years in which disruptions do and do not occur).


ROLE OF SECONDARY PRODUCTION AND INVENTORIES

Two important activities occurring in U.S. material markets that we chose,
for simplicity, not to include explicitly  in  the above illustrative
calculations for the catalystium market are inventory adjustments and
secondary recovery (recycling).  It is straightforward to include secondary
recovery in the analysis in a roughly appropriate way, by simply including
secondary production with  primary production  (from mines) in the supply
curves illustrated in Figures 2-4 and 2-6. Just as for  primary U.S.
production, the criticality of a material  from a national perspective is
reduced the greater the amount of recycling in normal times, and the greater
the extent to which recycling can be expanded when prices suddenly  rise.
(This description of the role of secondary recovery is qualitatively correct,
but it ignores the linkage between past consumption and  the pool of
scrappable items from which secondary production can  come during disruptions.
More sophisticated market models that explicitly recognize this linkage
should ideally be used to  calculate the criticality of materials.)

Business firms faced with  the threat of disruptions in the supply of an input
such as catalystium normally maintain inventories or  stockpiles to  be used
when supply disruptions occur.   It is clear from the  above analysis how the
existence of normal business inventories and  stocks can  decrease expected
costs from supply disruptions (that is, decrease the  criticality of a
material), by reducing the amount of material  that must  be purchased on the
world market at very high  prices during disruptions.  On the other  hand,
holding inventories imposes costs of its own  that should also be attributed
to the disruption threat,  and included in  the measure of criticality.
Administering and maintaining a stockpile  requires the time of a firm's
managers and employees, and involves other out-of-pocket expenses as well,
even after the inventories have been acquired.

The original  cost of the stockpiled material  is not counted by economists as
a cost to the firm (or nation)  at the time of acquisition, since one kind of
asset (money)  has just been transformed into  another  kind (stockpiled
                                    2-22

-------
                                                                   Charles
                                                                   River
                                                                   Associates

catalystium, in our example).  The transformation could be reversed,  if
desired (except for transaction costs, such as transportation  and brokerage
charges).  However, holding assets in the form of catalystium  over time,
rather than in plant and equipment, in other immediately productive assets,
or even in a financial instrument earning interest,  does impose  costs on the
firm (and the nation).  These costs are usually approximated by  the estimated
interest costs of financing the stockpile (even if the stockpiler did not
actually borrow to finance his inventories).   Many metals and  other materials
do not cost much (out of pocket)  to store,  relative  to their market value, so
the dominant cost of stockpiling  is in fact the interest expense.   As a
general rule, U.S. firms increase their stockpiles,  held in anticipation  of
market contingencies, until  the next unit stockpiled is not expected  to be
salable during the next disruption for enough to cover its expected holding
costs  (recognizing that the dates, severities, and duration of disruptions
are uncertain events).

In summary, U.S. consumers of a material  generally make two major types of
adaptations to the threat of supply disruptions (and other contingencies  as
well).  Before the disruption they acquire inventories, and after it  occurs
they switch to alternative technologies,  as summarized in market demand
curves for the material (such as  Figure 2-3).  (Having the capability to
switch quickly to alternate technologies  during disruptions may, of course,
also require advance planning.)

For purposes of measuring the economic criticality of a material  from a
national perspective, we can now lengthen the list of determinants (developed
earlier for particular consumers) as follows:

t    the severity of contingencies that threaten U.S. consumers  or producers,
     as measured by price increases that  occur;

•    the expected time between these contingencies;

t    the duration of these contingencies;

•    the quantity of the material consumed in the United States  under normal
     conditions;

•    the quantity of the material produced in the United States  under normal
     conditions, both from primary and secondary sources;

t    the extent to which U.S. consumption can be reduced when  prices  rise
     (the "elasticity" of short-run U.S.  demand);

•    the extent to which U.S. primary and secondary  production can be
     expanded when prices rise (the "elasticity" of  short-run  U.S.  supply);
     and

t    the size of normal U.S. inventories  and stockpiles.
                                     2-23

-------
                                                                   Charles
                                                                   River
                                                                   Associates

In addition to privately held inventories,  the U.S.  National  Defense
Stockpile contains platinum-group metals and other materials  that  could
conceivably be used for vehicular emissions control  during  a  nonmilitary
emergency.  Appendix 2-A considers this possibility,  and concludes that an
act of Congress would probably be required  to authorize  such  releases.  We do
not assume these stockpiles will  necessarily be available during the
nonmilitary contingencies upon which we base the illustrative criticality
estimates given at the end of this chapter.


         O  PRIVATE ADAPTATIONS TO THE THREAT
OF SUPPLY DISRUPTIONS AND OTHER CONTTNGtHCTES

Recognizing the adaptations that private firms make in response  to  the  threat
of contingencies in material  markets, key questions are whether  the U.S.
government should make additional preparations and adaptations with respect
to critical materials, and, more particularly, whether the  Environmental
Protection Agency should weigh the critical ity of the material in its
decisionmaking, beyond including the effects of the underlying contingencies
on the usual calculation of expected future  compliance costs  (as we described
that process earlier).  Under certain circumstances,  which  can be
approximately satisfied in some material  markets, general government policies
(such as stockpiling, tariffs, quotas, or subsidies to domestic  producers)
are not needed.  In these circumstances,  private firms can  be expected  to do
the amount of stockpiling, and choose the production technologies,  that are
efficient from a national perspective.  In these cases EPA  only  must consider
the effect of materials prices on expected compliance costs when choosing
among policy options.

We cannot analyze in detail here the conditions under which private
adaptations to market contingencies are efficient from a national
perspective.  (Charles River Associates has  filled many volumes  analyzing
these issues, particularly for materials  criticality  stemming from  the
threat of foreign supply disruptions.  See the Bibliographic  Note and
References.)  However, we can state the most important of these  conditions
and give some indication of their relevance.

The first of the conditions under which private firms would prepare
sufficiently for market contingencies is  that there be no expectation of
price controls, material allocation, or other government interference with
the market, even during serious disruptions.  For example,  if firms expect
inventories to be reallocated from "have" firms to "  have not" firms (as
actually occurred in the post-OPEC U.S. petroleum market),  then  they will
have reduced incentive to accumulate contingency stocks, and  private holdings
will be less than is justified by benefits (and costs)  measured  from a
national perspective.  Expectation of price  controls  (perhaps instituted with
the rationale of "moderating the inflationary impact of a supply disruption")
can have the same unfortunate consequences.
                                    2-24

-------
                                                                   Charles
                                                                   River
                                                                   Associates

A second condition is as follows:   The  preparations for contingencies by
individual  U.S. firms must not reduce the  likelihood or severity of the
contingency faced by other U.S.  firms.   This  condition is likely to fail to
some extent.  For example, when  an individual  U.S. firm holds an additional
unit of inventories, it must bear  the entire  cost of holding that unit, but
some of the benefit generally accrues to other firms.  This "external"
benefit to other U.S. firms occurs during  disruptions, when the firm with the
additional  stockpiled unit needs to buy one unit less on the world market,
tending to decrease the price at which  other  U.S. firms obtain their imports.
One can imagine an artificial  situation where this mechanism would not
operate, for example where a cartel  of  foreign producers threatens to
suddenly form and double the existing market  price, no matter how much the
resulting rate of demand is reduced due to U.S. buyers consuming out of
stocks.  However, it is much more  likely that the cartel would set a lower
price, at least initially, if U.S. stocks  are higher.  This mechanism is
referred to in CRA studies as the  "price deterrence" benefit of U.S.
stockpiling.  Reductions in U.S. consumption  during disruptions due to use of
alternative technologies can also  have  price  deterrence benefits that are
external to the individual U.S.  firm actually adopting the alternative.

In general, whenever a U.S. firm bears  all the cost of some action preparing
for a disruption, or all of the  cost of an adjustment made during the
disruption, but other firms reap some  "external" benefit, private
preparations and adjustments tend  to be less  extensive than is desirable for
the nation as a whole.  All of the costs of an additional increment of
preparation or adjustment are recognized by the private decisionmaking unit,
but all of the benefits to the nation are  not. The private firm stops
preparing or adjusting when the  incremental private cost equals the
incremental private benefit, whereas, from a  national viewpoint, the firm
should continue preparing and adjusting until  the incremental private cost
equals the incremental national  benefit.

Another mechanism of the same nature can be relevant when a material market
is threatened by a disruption that is deliberately decided upon by foreign
producers of a material, or by their governments.  In that case, greater U.S.
preparations, particularly larger  U.S.  stockpiles, can decrease the
probability of a disruption occurring in the  first place.  In CRA studies
this mechanism is called the "probability  deterrence" benefit of U.S.
stockpiling.  (See for example the appendix to Klass, Burrows, and Beggs
(1980).)

Depending on the types of threats  facing a particular market, price
deterrence and probability deterrence can  make private preparations and
adjustments much less than would be desirable from a national perspective.
In these situations, a strong case can  be  made for government policies, such
as tariffs or stockpiling, that  will manipulate or augment private
preparations and adjustments.
                                     2-25

-------
                                                                   Charles
                                                                   River
                                                                   Associates

Other conditions can cause preparations for market contingencies by U.S.
firms to fall short of the national  optimum.   One  such  condition involves the
way private firms compare present costs and benefits  with  future costs and
benefits.  Generally, decisionmakers "discount"  costs and  benefits  occurring
in the future relative to current costs and benefits; that is, a dollar
gained tomorrow justifies the expenditure of something  less  than a  dollar
today.  But makers of public policy  may conclude that private firms discount
future costs of disruptions too much,  thus incurring  insufficient costs for
preparations today.  (Economic theory  does not offer  a  definitive answer to
the question of what rate of time discount is most appropriate for  national
policymaking.)

The above discussion assumed, as is  often the case, that U.S. firms consuming
a critical imported material are individually small relative to the world
market, but together account for a sizable proportion of world consumption.
On the other hand, if a single U.S.  firm accounts  for most of U.S.
consumption, then the importance of  price deterrence  effects and probability
deterrence effects, as discussed above, may be considerably  less.  Most of
the benefits of preparation for, and adjustments to,  disruption will  accrue
to that single U.S. firm, rather than  being "external,"  so the extent of
private preparations and adjustments will  tend to  be  much  closer to the
national optimum.  In Japan, it is common for firms to  coordinate decisions
about critical imported materials, thus gaining  benefits that otherwise would
be possible only in the case of a single importing firm.   However,  in the
United States such coordination would  run afoul  of antitrust regulations.
(There are many competing considerations in deciding  the desirability of the
Japanese institutional arrangement versus the U.S. arrangement, and this
consideration is very probably not the deciding  one.)

One circumstance that can lead to more private preparations  for disruption
(relative to the national optimum) rather than less,  is "risk aversion" on
the part of private firms.  This situation can be  explained  with reference to
our earlier sample calculation of criticality for  consumers  of the  imaginary
material catalystium.  We calculated the criticality  of catalystium to
consumers by averaging losses that could be expected  to occur over many
years.  Since disruptions were assumed to occur  in one  year  out of  ten, the
average expected loss to consumers was (0.1)($3,200,000) = $320,000 per year.
Contrast this expected loss with that  which would  occur if a disruption loss
of $640,000 occurred in one year out of two.   In that case,  expected  losses
are again $320,000 per year, and the criticality of the material is as
before.

This procedure for evaluating losses of different  sizes occurring with
different frequencies is entirely appropriate for  policymaking at the
national level, where losses in a given year of  $640,000 or  $3,200,000 are
miniscule relative to the size of the  entire U.S.  economy.  However,  consider
                                     2-26

-------
                                                                   Charles
                                                                   River
                                                                   Associates

a small U.S. firm for which catalystium accounts  for a  major  proportion of
production costs, and suppose there is stiff competition  from other  firms
whose product does not require catalystium.   For  that small firm, a  tripling
of the price of catalystium, sufficient to inflict costs  of $3,200,000 on
U.S. consumers over one year, might be sufficiently severe to cause
bankruptcy.  The fact that losses will "average out," over many  succeeding
years, is small comfort if the firm has gone bankrupt in  the  meantime.  In
this case, where the possible loss is catastrophic for  the decisionmaking
unit, it is rational to be "risk averse,"  and undertake more  stockpiling and
other preparations for disruptions than would be  justified by average
expected costs alone.

In addition to stockpiling on its own, a small  firm can also  reduce  the risk
of an increase in the price of catalystium by buying "forward" contracts for
future delivery of catalystium on a commodity exchange.  In that way the
risk is spread among a great many speculators,  who can  individually  diversify
their speculations so that they are not catastrophically  affected by a
disruption in the catalystium market.  In  this way private risk  aversion can
be reduced, which is beneficial to the individual  firm, but may  reduce
preparations for disruptions that they would otherwise  undertake.  Another
strategy for reducing risks of supply disruptions is to enter into long-term,
fixed-price contracts with reliable suppliers who are unlikely to be
disrupted.


IMPLICATIONS FOR EPA POLICYMAKING
We have described above a number of conditions in material  markets  that  can
cause private firms'  preparation and adjustments for market contingencies,
particularly supply disruptions, to be different from (usually  less
extensive) those that would be justified by the costs and expected  benefits
measured from a national  perspective.  In those circumstances,  a  case can be
made for government policies such as tariffs and national  stockpiles, which
modify or augment private preparations or adjustments made in response to the
threat or occurrence of a contingency such as a supply disruption.  A
material must be "critical," that is, threatened with serious contingencies
such as a major foreign supply disruption, to justify government  actions in
addition to the private actions that profit-maximizing U.S. firms undertake
naturally.  However, government actions such as tariffs and national
stockpiles are not necessarily justified for all critical  materials.  That
is, criticality is a necessary condition, but not a sufficient  condition, to
justify general government policies such as tariffs and stockpiles.

With this analysis of criticality and general government policymaking in
material markets as a background, we can now address directly the principle
question that this chapter asks:  how should the Environmental  Protection
                                     2-27

-------
                                                                   Charles
                                                                   River
                                                                   Associates

Agency factor the criticality of materials  into  its  decisionmaking, beyond
recognizing the effects of contingencies on prices of materials  when
calculating average future compliance costs?  We presume  that  any effect of
EPA decisions on the long-run equilibrium price  of a material  (roughly
speaking, its "normal" price) is taken into account.  (The  usual example is
where increased demands induced by EPA regulations raise  the normal price of
the material somewhat.)  We also presume that any operative, or  likely,
general  government policies in the material  market (such  as tariffs) are
taken into account by EPA when calculating  compliance costs.

Our answer to the question of appropriate EPA policymaking  procedures is
particularly easy to justify if general  government policies in the material
market,  such as stockpiles and tariffs,  are presumed to adequately recognize
the material's critical ity.  In that case,  the price paid for  the material by
producers of vehicular emissions control equipment respresents the material's
true cost to society, and no special  EPA policymaking, beyond  careful
forecasting of compliance costs, is required.  This  conclusion can be
justified in great detail, using elaborate  versions  of the  type  of economic
cost-benefit analysis that we described above as a basis  for measuring  the
criticality of materials.  However, the conclusion is sufficiently plausible
on its face (and we have sufficiently burdened noneconomists reading this
chapter with unfamiliar concepts) so that we forgo its full development.  It
is basically just one application of very general economic  theories showing
how the price system can efficiently allocate resources in  a free market
economy.

It is, of course, true that EPA decisions can greatly change the quantities
of a material that are consumed in the United States, as  well  as other
conditions affecting its criticality, thus  changing  the general  government
policies that are appropriate.  To take an  obvious example, increased U.S.
consumption to meet EPA regulations would presumably increase  the optimal
size of government stockpiles.  In order to facilitate better  and more  timely
government policymaking for materials markets, it certainly could be
worthwhile in principle for the EPA to inform other  government agencies,
particularly those with direct responsibility for policymaking in material
markets, concerning EPA decisions that will  significantly affect material
markets.  (We hope circulation of this study outside EPA  will  serve this
purpose to some degree.)  In return,  EPA might learn of possible changes in
general  government policies that would affect the forecasting  of compliance
costs and hence potentially affect EPA decisions.

The above line of reasoning might sound a trifle artificial to those familiar
with U.S. policies toward material  markets,  because  in fact such policies
have been designed almost exclusively in response to the  threat  of military
contingencies.  As we discuss further in Appendix 2-A, the  U.S.  National
Defense Stockpile is currently reserved exclusively  for defense  related
                                      2-28

-------
                                                                   Charles
                                                                   River
                                                                   Associates

applications, unless an act of Congress authorizes  release  for other  purposes
(which has never happened in the forty-plus years of  the  stockpile's
existence).   The United States has  modest tariffs on  the  importation  of a
number of raw and processed materials.   The policy  debate on  such tariffs
often centers around the desirability of having  secure  domestic sources of
supply during wartime (though the political power of  special  interest groups
sometimes seems the more important  consideration).  Certainly, U.S. tariffs
have not been adjusted as the types of  nonmilitary  contingencies that we
identified earlier threaten and recede,  in the way  in which economic  theory
suggests would be optimal.

On the other hand, there are a number of government policies  aimed  at the
structural causes of certain nonmilitary contingencies  in material  markets.
Government sponsored R&D is often aimed at reducing U.S.  dependence on
imports of materials, for example,  by making  it  profitable  to exploit
previously uneconomical U.S. deposits.   The U.S. Department of the  Interior
is the lead agency in this area. U.S.  government agencies  such as  the
National Labor Relations Board are  concerned  with settling  domestic labor
disputes, including those affecting material  markets.  The  U.S. Department of
State and other U.S. agencies are concerned with international relations that
may affect the conditions under which this country  imports  materials.
Nevertheless, it is still relevant  to note that  the United  States in  most
cases has not generally employed market-specific policy instruments,  such as
tariffs and "economic" stockpiles,  to counteract the  threat of nonmilitary
contingencies such as foreign supply disruptions.

Is this lack of fine-tuned U.S. policies aimed at nonmilitary contingencies
a serious problem?  In most cases,  probably not. In  many material  markets,
it can plausibly be argued that the private sector  undertakes sufficient
preparations for, and adjustments to, nonmilitary contingencies in  material
markets so that the benefit of even theoretically optimal government  policies
on tariffs and stockpiling would diminish expected  national losses  only
moderately.   Moreover, when decisions are finally made  in the real  world
about such national policy instruments  as tariffs and public  stockpiles, it
must be recognized that they are often  in practice  more costly than analysis
of their theoretical optimality would indicate.  Administering tariffs and
stockpiles can be much more costly  than originally  estimated, particularly
when vested economic interests and  political  realities  intrude into the
management process.  Recognizing these  facts  of  political life reduce
considerably the potential scope for beneficial  government  policymaking, and
makes historical practice in the United States more understandable.

Suppose then that the EPA is contemplating policy options that will greatly
affect the market for certain materials (such as increasing demand  for the
platinum-group metals), and it suspects that  national policies (such  as
                                     2-29

-------
                                                                   Charles
                                                                   River
                                                                   Associates

tariffs and economic stockpiles)  do  not adequately recognize the criticality
of the materials.   What should EPA do?   The  first observation we make, which
is easy to understand in the light of discussions above,  is that it is often
difficult to determine whether national  policies "adequately" recognize the
criticality of materials, given private efforts to ameliorate that
criticality.  To say the least, it would be  very ambitious for  EPA  to analyze
conditions in material markets in sufficient depth to make such a
determination, particularly where government agencies with direct
responsibility for these issues have not done so.  In fact, EPA can hardly be
held responsible for such policy analysis, and doubtless  it is  a
responsibility whose lack EPA does not  regret.

Furthermore, any actions EPA could take to reduce consumption of a  critical
material -- for example by requiring compliance through one technology rather
than another -- would fall into a category that economists call "second-best
solutions."  It is more efficient from  a national perspective to have
national policies that discourage all consumers from using a critical
material, if that is indeed called for  because some condition exists in the
market that makes private policies otherwise inadequate.

The two main conclusions we reach regarding  the role of materials criticality
in EPA decisionmaking are thus as follows:   first, in almost all cases the
EPA need only calculate compliance costs in  a comprehensive manner  that
recognizes the likelihood of market  contingencies temporarily raising market
prices in the future.  (The effect of EPA induced demands for materials on
compliance costs should also be recognized,  both during possibly rapid
implementation, and during the "long run" thereafter.)  Second, where the EPA
anticipates its policies will have major impacts on material markets, it
should coordinate with the Department of the Interior and other federal
agencies with more direct responsibility for policies affecting material
markets.

In order to implement these suggestions, it  can still be  useful for EPA to
estimate roughly the criticality of  materials required for compliance with
its regulations.  Estimating criticality from the perspective of consumption
for vehicular pollution control equipment is very closely related to
estimating how much expected future  compliance costs will be raised due to
the contingencies that threaten the  market,  which is something  EPA  should do
anyway.  Estimating criticality from a  national perspective suggests the
possible importance of EPA coordinating policies affecting materials markets
with other government agencies.  The last part of this chapter  pursues both
of these conclusions by developing rough estimates of the criticality of
platinum-group metals and four other materials required for vehicular
emissions control  equipment.
                                    2-30

-------
                                                                   Charles
                                                                   River
                                                                   Associates
SHOULD EPA SPONSOR R&D ON VEHICULAR EMISSIONS CONTROL?
Our discussion above considered decisions  by  private  firms  about  amounts of
critical materials to use, when there exists  a  range  of different
technologies for complying with EPA standards on  vehicular  emissions.  A
further, related question is whether private  research and development on new
compliance technologies is sufficiently  extensive,  or whether  government
sponsored research of some sort is called  for.

Research is of course an uncertain process, and there is always the chance
that a government sponsored program will discover a new technology that
private investigators overlooked.   However, there appears to be no general
evidence that government sponsored research tends to  be more productive per
dollar spent.  If anything, conventional wisdom asserts just the  opposite,
that private R&D tends to be more  efficient.  There are horror stories about
ill advised and unsuccessful government  supported research, but there are
other cases as well, where government research  was  successful  where private
research was not successful, or was deemed too  unpromising  even to pursue.  A
classic case in the history of material  markets is  the development by the
U.S. Bureau of Mines of technologies to  process deposits of the low-grade
iron ore taconite.

We reach no conclusions here concerning  the comparative cost effectiveness of
private versus government research and development.  Rather, we simply
examine the general circumstances  under  which private decisions about R&D
expenditures in this area are made, to see if there are strong reasons to
suspect that the total  amount spent would  be  inadequate.

The usual cause of inadequate private research  (and hence the  usual
justification for government sponsorship of research)  involves reasoning much
like that we described above to analyze  the adequacy  of private stockpiling
and other preparations for contingencies in material  markets.  An individual
private firm bears all  of the expenses of  research  it undertakes  on its own.
If there are large "external" benefits to  other firms from  successful
research by one firm, which that one firm  cannot  substantially capture
through licensing fees and other arrangements,  then the firm contemplating
research will tend to spend less than is justified  by the expected benefits
of its research to the nation as a whole.  So-called  "basic" research tends
to have the most extensive external benefits, which cannot  be  appropriated by
the successful researcher.  Thus,  government  sponsorship of research is
typically most justified for basic research.

How do the circumstances of private research  and  development on vehicular
emissions control stack up against the usual  justification  for government
sponsorship?  By and large it appears that the  private level of effort in
this area should be roughly appropriate.   Most  of the research is specific to
                                    2-31

-------
                                                                    Charles
                                                                    River
                                                                    Associates
emissions control, and has limited usefulness outside the motor vehicle
industry.  To be more specific, much of the benefit of research in this area
undertaken by General Motors will accrue to General Motors.  Moreover,
licensing fees for technologies patented at General Motors should allow GM to
appropriate some of the benefits received by other motor vehicle
manufacturers.  The general case for government sponsorship of research in
this area appears quite weak.

In Chapter 6 we review research on alternatives to conventional  catalytic
converters (containing platinum-group metals) that has taken place to date.
Certainly there is no direct indication that promising avenues of research to
develop more cost effective compliance technologies have been left
unexplored.


A MAJOR QUALIFICATION AND A POSSIBLE POLICY PRESCRIPTION

The relatively optimistic conclusions reached above about the adequacy of
private R&D and private preparations for contingencies such as supply
disruptions, are applicable to all industries, whether their consumption of a
critical material stems from the nature of consumers'  demands in the
marketplace or from the need to satisfy government regulations.   However,
there is one further major consideration where consumption of a  material  is
based predominantly on the need to satisfy a government regulation.

The main qualification we would make to our general case for the adequacy of
private R&D on vehicular emissions control  technologies concerns private
expectations about the stringency of future emission standards.   If  U.S.
firms do not expect a future standard to be actually in effect and enforced,
then they may well not undertake adequate research on ways to meet that
standard in the manner least costly to themselves and the nation.   EPA is
undoubtedly in a better position than we to assess the credibility of
scheduled future vehicular emissions standards.

This same type of issue can arise when assessing the adequacy of the
preparations that the U.S. auto industry makes for contingencies such as
foreign supply disruptions.  If the U. S. auto industry believes that it will
be able to arrange relaxation of emissions standards whenever the price of
platinum, palladium,  and rhodium rises greatly during a supply disruption,
then the industry will probably undertake less extensive preparations than
they would otherwise.  In particular, they would stockpile much  less of these
metals for such contingencies.
                                    2-32

-------
                                                                   Charles
                                                                   River
                                                                   Associates
It is easy to see why a politically  persuasive  case  for  relaxation  of
emissions standards could be made during a  major  supply  disruption  in the
platinum-group markets.  It would probably  be virtually  impossible  for U. S.
vehicle manufacturers to obtain  enough  of these metals on  the  spot  market if
they had no inventories and supplies from South Africa and the Soviet Union
were cut off.  As discussed further  below,  the  prices of the required
platinum-group metals could become "astronomical"  in this  case.

It would be very interesting in  this regard to  know  the  size of the
inventories of platinum, palladium,  and rhodium that the U. S. auto industry
currently maintains.  However,  as discussed further  in Chapter 4, consumers'
stocks of these metals are not officially estimated  in the United States.   If
stocks in the auto industry are quite low in comparison  with other  industries
(such as petroleum refining or glass manufacturing)  where  continuing use of
platinum-group metals is mandated by technology rather than by government
regulation, then a case could be made that  the  auto  industry has some
expectation of throwing itself on the mercy of  the regulatory  and legislative
processes in the event of a major disruption in platinum-group supplies from
South Africa.  (In fact, stocks  of these critical  materials in the  auto
industry should be considerably  greater than those "on the shelf" in the
petroleum and glass industries,  because those other  industries have control
over material obtained by secondary  recovery (after  a few  years' use),
whereas the auto industry has no special access to material obtained from
obsolete catalytic converters.)

The obvious solution to this problem, if it is  indeed a  problem, would be to
require documentation from U. S. auto manufacturers  that they  have  a
specified minimum level of inventories  on hand  at all times, unless given an
explicit exemption by EPA.  We have  not investigated the legal or practical
aspects of implementing such a new regulation.  In setting the appropriate
level for such inventories, it would be important for EPA  to decide whether
very large increases in the price of platinum-group  metals could in fact
eventually justify some relaxation of emissions standards. (One interesting
aspect of such a policy is that it would probably be about as  effective with
foreign vehicle manufacturers as with U.S.  manufacturers,  even though they
would not be directly subject to the stockholding requirement; the  reason is
that they would fully expect EPA vehicular  emissions standards to be
maintained as long as U.S. vehicle manufacturers  could consume platinum,
palladium, and rhodium out of required  inventories.)
                                     2-33

-------
                                                                   Charles
                                                                   River
                                                                   Associates
              MATERIAL IMPORTS AND BALANCE OF PAYMENT PROBLEMS


The reader may have noticed in the preceding discussions that, despite  its
references to general  policymaking in materials markets, no explicit mention
was made of the effect of increased material imports on the U.S.  balance  of
trade or balance of payments.  This omission was purposeful, since  the  issue
tends to be largely a red herring for efficient government policymaking in
markets for particular goods such as materials.  It is relatively
straightforward to estimate the increased dollar value of U.S. material
imports that would result from increased consumption for vehicular  emissions
control.  However, in themselves, increased imports usually do not  signal
significant national  losses --and certainly not losses of the same  magnitude
as the dollar value of increased imports.

It would be too time consuming to develop fully the rationale for the above
conclusion here, but the general idea is as follows.  International  trade
benefits all countries that participate, by allowing them to import goods in
whose production they are relatively inefficient, and export goods  in whose
production they are relatively efficient.  Imports are a necessary  part of
this process, and as such they benefit the United States rather than harming
it.

Nevertheless, though it may at first seem paradoxical  in view of  the above
general truths, it is possible in theory to benefit the United States by
reducing U. S. imports.  This can be done for all imports most easily by
imposing a general tariff, or for imports of a particular good (such as a
material)  by imposing a specific tariff or quota.  Where reduced  imports  of a
particular material are planned, the rationale is generally that  the United
States is a large importer on the world market, and reduced U.S.  imports  can
reduce the price at which the remaining imports are obtained.   In effect,
implementing such policies allows the United States to act monopolistically
("monopsonistically," to be more precise) with respect to its foreign
suppliers.  The United States does benefit, but foreign exporters lose  even
more than the United States gains.

The biggest problem with policies such as tariffs, particularly a general
tariff on all imports, is that they invite retaliation.  If other trading
countries also reduce their imports, then all countries will typically  be
worse off than with no tariffs at all, simply because the greater
international productive efficiency allowed by trade has been diminished.
Recognizing these facts, the United States has often been a world leader  in
attempting to reduce international trade barriers, so that all  countries
benefit from more free trade.  U. S. tariffs on material  imports  can be a
useful  response to the threat of particular market contingencies  (as
discussed earlier in this chapter), particularly where retaliation  by foreign
suppliers is not a problem.  However, simply imposing trade barriers to
                                     2-34

-------
                                                                   Charles
                                                                   River
                                                                   Associates
improve the U. S. balance of trade largely represents  a  "beggar-thy-neighbor"
policy that runs counter to the U. S.  tradition of supporting  free  trade
(continually threatened by special interest groups though  the  tradition may
be).

Some of the same effect as a tariff on imports  of materials  could be  achieved
by EPA requiring that compliance with  its regulations  be achieved with less
of an imported material, such as the platinum-group metals.  It is  unlikely
such actions would invite retaliation  in the way a tariff  can, so,  from a
narrow perspective, some small  economic gain to the United States could
accrue.  However, in order to obtain such monopolistic (monopsom'stic)
benefits, foreign prices for the remaining amount of  U.  S. imports  must be
driven down.  The first requirement is that EPA decisions  affect a  large
proportion of U. S. imports of the material.  As we show later, EPA has this
"leverage" only in the market for platinum-group metals.  Moreover, in most
major material markets, world supply is usually very  price elastic  in the
long run, which means that a reduction in U. S. consumption  (or in  the growth
of U.S. consumption) will cause only a very small decrease in  the world price
of the material, after suppliers have  had a chance to  ajust  to the  new
situation.  Thus, for example,  the monopsonistic benefit to  the United States
of reducing platinum consumption would be very  small  because of the
elasticity of supplies from South Africa.

There are exceptions to the above generalizations, however,  particularly by
product materials where demand reductions can lead to  sizable  price
decreases.  Among materials used for vehicular  pollution control in the
United States, the outstanding example is rhodium, which is  a  byproduct of
platinum.  Rhodium production is essentially at the limit  imposed by  current
world platinum production, and rhodium is usually the  most expensive
commercial metal (per unit weight).  If EPA eliminated U.S.  rhodium
consumption for vehicular emissions control, there would be  a  significant
decrease in its price on the world market.  It  would  be  possible to do a
careful calculation of the optimal extent to reduce U.S. rhodium consumption
in order to realize the maximum monopsonistic gain for the United States.
(If EPA decisions determined all of U.S. rhodium consumption,  then  the
maximal monopsonistic gain to the United States would  be obtained simply as a
result of EPA recognizing the effect of its decisions  on the price  of rhodium
when calculating compliance costs, as  we have recommended.)  However  the gain
to the United States would still be very modest, and  it  seems  almost  certain
that EPA has more important policy concerns to  occupy  its  attention.
Moreover, regulating rhodium consumption downward would  indirectly  violate
                                     2-35

-------
                                                                   Charles
                                                                   River
                                                                   Associates


the U.S. tradition advocating free trade. (An interesting further aspect of
the case of rhodium is that the parties most harmed by a reduction in U.S.
imports would be platinum producers in South Africa.)

We have attempted above to make the most reasonable case we can for EPA being
concerned with the implications of its policies for U.S. material  imports and
the balance of trade.  However, in the final analysis  we do not think it is
of sufficient importance to warrant inclusion as a separate consideration in
EPA policymaking.  It will be hard enough for EPA simply to take into account
such vagaries of international trade as changes in exchange rates among the
U.S. dollar and other world currencies, as these affect the dollar cost of
U.S. material imports, and hence compliance costs for  EPA regulations.  It is
not really reasonable to expect EPA to design optimal  monopsonistic importing
policies as part of its decisionmaking.


      A SIMPLE ECONOMIC MODEL FOR ESTIMATING CRITICALITY DUE TO FOREIGN
                             SUPPLY DISRUPTIONS
The economic model  we use in this chapter to estimate the criticality of
materials used in vehicular emissions control  equipment, is a slightly
generalized version of the simple analysis of U.S.  supply and demand curves
that we described earlier in numerical  terms for the imaginary material
"catalystium."  As in that example, we concentrate  on materials criticality
due to the threat of foreign supply disruptions, and calculate criticality
from both the national perspective and the perspective of consumption for
vehicular emmissions control.
                                    2-36

-------
                                                                   Charles
                                                                   River
                                                                   Associates
KEY PARAMETERS AND FORMULAS
Figure 2-7 displays, with general  algebraic nictation, the linear supply and
demand curves that are assumed to be relevant to each of the materials we
consider.  Key parameters of the criticality measurement, whose roles are
explicitly indicated in Figure 2-7, are as follows:


          P0 = the normal world market price of the material

         XP0 = the world market price during disruption


          Qd = U.S. consumption at normal price P0

          Qp = U.S. primary production at normal price P0

          Qr = U.S. secondary production at normal price P0


          ed = the price elasticity of U.S. consumption

          ep = the price elasticity of U.S. primary production

          es = the price elasticity of U.S. secondary production.

Other key parameters in the criticality measurement are the following:

           D = the duration of disruptions, and

           T = the time interval between (starts of) disruptions

When we consider later the possible role of stockpiling, we must also
specify

           r = the "real" rate of interest (difference between observed
               nominal rates and the rate of inflation), and

           e = the out-of-pocket expenses of stockholding, measured
               relative to the value of the material stockpiled.
                                     2-37

-------
                                                  Charles
                                                  River
                                                  Associates
Figure 2—7

MARKET MODEL FOR ESTIMATING CRITICALITY OF IMPORTED
MATERIALS THREATENED WITH FOREIGN SUPPLY DISRUPTIONS
PRICE
XP
2P
      SECONDARY,
      PRODUCTION
   0  Q
                        /    *  PRIMARY PLUS
                        1   /< SECONDARY
                              PRODUCTION
CONSUMPTION
                                             QUANTITY
                         2-33

-------
                                                                   Charles
                                                                   River
                                                                   Associates


The price elasticity of U.S. consumption is the relative decrease in
U.S. consumption that results from a doubling of the price"!   Thus,  if a
doubling of price causes a 10 percent decrease in consumption (as was the
case for producers of emissions control  equipment in the "catalystium"
example dicussed earlier), ej = 0.1.  Since we are assuming  all  demand (and
supply) curves are linear, decreases in  consumption resulting from  larger
price increases are proportionately greater.   Thus, an X-fold increase in
price during disruptions causes U.S. consumption to decrease to
Qd [l-(X-l)ed].

The price elasticity of U.S. primary or  secondary production is  the
relative increase in U.S. primary or secondary production that results from  a
doubling of price.  Along linear supply  curves, an X-fold increase  in
price causes U.S. primary production to  increase to Qs [l-(X-l)es].

Figure 2-8 reproduces the total U.S. demand and supply curves from  Figure 2-7
and labels the loss areas that were explained in numerical  terms for  the
catalystium example.  Triangle 1 is the  net cost consumers incur from
reducing consumption by fraction (X-Den.  The rectangle made up of Areas 2,
3, and 4 is the additional payment (or  transfer") consumers make to  domestic
and foreign suppliers due to the price increase.  Rectangle  2 is the
additional transfer to foreign suppliers.  The rectangle made up of areas 3
and 4 is the additional transfer to domestic producers, of which Triangle 3
represents the additional cost of production (beyond the cost of importing
the material at normal price P0).

The first column of Table 2-1 translates the loss areas in Figure 2-8 into
algebraic formulas convenient for performing the actual calculations.  These
formulas take  into account the duration  (D) and frequency (1/T)  of
disruptions, to give average expected losses per year.  All  losses  are
expressed as multiples of market values  observable in normal times.  For
example, U.S.  consumers' losses are a multiple of the value  of their
consumption in normal times (P0Qd)-  "Losses" of primary and secondary
producers are  generally negative (that is, "gains ), because the transfer (a
negative cost) is bigger than the adjustment cost by an amount represented in
Figure 2-8 as  Trapezoid 4.  The losses of primary and secondary  producers are
expressed in Table 2-1 as multiples of the values of their production in
normal tmes (P0Qp and P0Qs> respectively).

Average expected U.S. losses per year, netting gains by U.S. producers
against the larger losses by U.S. consumers, are obtained by adding up the
losses in the  first column of Table 2-1.  The result is our measure of the
criticality of the material from a national perspective.  We do not present
                                     2-39

-------
Figure 2-8



U. S. ECONOMIC LOSSES (AND GAINS) FROM

FOREIGN SUPPLY DISRUPTIONS
PRICE
XP
                                                    Charles
                                                    River  4
                                                    Associates
                     Qs [1 + (X-1)6,1
                                                QUANTITY
                            2-40

-------
River
Table 2-1 Associates
FORMULAS FOR AVERAGE EXPECTED ANNUAL U0S, ECONOMIC LOSSES FROM FOREIGN SUPPLY DISRUPTIONS*
Import-Eliminating
Economic Groups No Stockpiling Stockpile** Comprehensive Stockpile
Consumers
Adjustment (1)
Transfer (2,3 a
Primary Producers
Adjustment (3,
Transfer (3 and
Secondary Producer
Adjustment (3,
Transfer (3 and
Stock Holders
Holding Cost
Acquisition Cos
Less Revenues
2
HP 0 • •• vy - - C 11 D/T ^ 0
p
nd 4) {PQQd[tX-l)-(X-l) ed]} D/T ** '" '' 0
, .2
part; i_r LJ - — ?j — — t j u/ 1 ~ u
0
<1 Dart} f-P 0 r( X-l ^ hf X-1 ) c 11 n/T -^ - 	 0
i , (Jai u; x -i -VpL VA-iyT^A-i/ n ' 	
S
2
part) [P Q -^ii- £ 1 D/T -* 	 °
vj j C- o
4, part) {-P0QSC(X-1)+(X-1)2 es:> D/T -* 	 °
0 (r+e) Po -Qp [1+(X-1) ep] D (r+e) Po (Qd-Qp-Qs)
l-Qs [1+(X-1) esl
tf o -(X-1J Po l-l^Hx-ij^ilD/T o
-o: n+fx-n e:i!

-------
ro


ro
         Table 2-1


         FORMULAS FOR AVERAGE EXPECTED ANNUAL  U.S.  ECONOMIC  LOSSES  FROM  FOREIGN  SUPPLY  DISRUPTIONS*


         (Continued)
         SOURCE:  Charles River Associates,  1981.


         *Appendix 2-B describes a simple computer program  that  performs  these calculations.


         **Arrows indicate that formulas are unchanged  from the  column  to the  left.
                                                                                                            Charles
                                                                                                            River
                                                                                                            Associates

-------
                                                                    Charles
                                                                    River
                                                                    Associates


separate formulas for calculating economic losses from the persective of U.S.
consumption for vehicular emissions control, because they are a special  case
of the formulas in Table 2-1.  All  that is required is setting U.S.  primary
and secondary production (Qp and Qs) equal to zero, and setting U.S.
consumption (Q^) equal to consumption for vehicular emissions control
equipment, with the appropriate price elasticity


TREATMENT OF STOCKPILING

The economic model we are using is too simple to provide an adequate
explanation of the holding of inventories and stockpiles.  As discussed
earlier, U.S. stockpiles can reduce U.S. losses during a disruption  by
reducing the amount of imports required.  Stockholding of a material can be
undertaken by U.S. producers, consumers, dealers, brokers, commodity
exchanges; or in fact any private party that wishes to bear the expense  of
doing it, generally with the expectation of selling at some time in  the
future at a significantly higher price that covers holding costs.   In Table
2-1 we treat U.S. stockpiling as a separate activity,  even though  it may be
undertaken by consumers, primary producers, or secondary producers.
(Consumers holding stocks can be considered to sell to themselves  during
disruptions.)

Table 2-1 allows the calculation of U.S. economic losses using three
different assumptions about stocks.  None of these three oversimplified
assumptions about stocks will be exactly appropriate,  but they provide useful
perspective in the area where our analysis of materials criticality  would
otherwise be weakest.  The formulas in the first column of Table 2-1, which
we discussed above, calculate average expected U.S. economic losses  per  year
assuming there is no U.S. stockpiling at all.

The formulas in the second column of Table 2-1 assume  that the U.S.  stockpile
is sufficiently large so that the United States need not import any  of the
material during disruptions, but the domestic price is still  equal to the
world market price (XPp) during disruptions.  Thus, U.S. consumers and
producers experience the same losses and gains as they did with no stockpile.
(The arrows in the second column of Table 2-1 indicate where loss  formulas
are unchanged from the first column.)  The difference  between the  two cases
is that additional revenues that accrued to foreign suppliers during
disruptions when there was no U.S.  stockpile now accrue to U.S. stockpilers.

As explicitly indicated in the labeling of the quantity axis in Figure 2-8,
the amount of U.S. imports during disruptions, measured at annual  rates, is
                                     2-43

-------
                                                                   Charles
                                                                   River
                                                                   Associates
 Annual Rate of Imports
 During Disruptions      = Qd [l-(X-l)ed] - Qp [l+(X-l)ep] - Qs [l+(X-l)es]

If disruption duration D is a half year,  for example,  the actual  quantity of
imports during the disruption will  be half of the amount determined by the
above formula, which we assume to be the  quantity stockpiled in the second
column of Table 2-1.

The "real" cost of holding a dollar of assets in  the  form of stockpiles
(rather than in a form earning interest)  is the "real"  interest rate r, which
we specify to be 6 percent (r = 0.06) for results reported in this  chapter.
The real  interest rate is estimated as the difference  between the "nominal"
interest rate actually observed in money  and capital markets, and the rate of
inflation.  Our use of an estimated real  interest rate, to calculate the
costs of holding stockpiles for a year, implies the reasonable assumption
that the'value of stockpiled material will rise at the same rate as the
general rate of inflation in the U.S. economy, even in the absense  of any
supply disruption.  All prices, values, and economic  losses that we report in
this chapter are in terms of constant 1981 dollars, that is, dollars deflated
to adjust for future inflation (which raises the  "nominal," but not the
"real," prices of materials and most other goods.)  (A real interest rate of
6 percent is approximate by historical standards.  However, in 1981 real
interest rates were unusually high in the United  States.  Nominal interest
rates were nearly 20 percent per year, while general  inflation was  running at
less than 10 percent per year, implying a real interest rate closer to 10
percent.   Since the analysis we are doing here bears  on policies over many
future years, estimating future costs of  holding  stocks on a more normal
historical basis seems appropriate.)

In addition to the real interest cost of holding stocks, there are
out-of-pocket costs for management, warehousing,  etc., that should in
principle be included.  For most materials, the interest cost is considerably
larger, but we also allow for an annual out-of-pocket cost (e), per dollar
value of stockpiled material.  We generally just assume e = 0.005 for results
reported in this chapter.  The annual cost of holding  the
"import-eliminating" stockpile is thus

                              Annual Rate of Imports
                    (r+e) P0    During Disruptions     D

as entered under "Stock Holding Costs" near the bottom of the second column
in Table 2-1.
                                    2-44

-------
                                                                   Charles
                                                                   River
                                                                   Associates
 In addition to holding costs, stockpilers of course also acquire the material
 during normal times at price P0, and sell it during disruptions at price XP0.
 The cost of acquiring stocks, net of revenues received from sale during
 disruptions, is

                              Annual Rate of Imports
                   -(X-l) P0    During Disruptions     D

 which will generally be negative because prices rise during supply
 disruptions (X>1).  This negative cost, adjusted for the fact that it only
 occurs every T years, is entered under "Acquisition Cost, Less Revenues" at
 the bottom of the second column in Table 2-1.  This amount would accrue to
 foreign suppliers if there were no U.S. stockpile, as assumed in the first
 column of Table 2-1, but with an import-eliminating stockpile it accrues to
 U.S. stockholders instead.  If the acquisition costs, less revenues from
 sales during disruptions (on an annual  average basis), is more negative than
 the holding cost per year is positive,  then holding stocks in anticipation of
 the contingency is profitable, and U.S. economic losses are reduced by
 stockpiling.  (If disruptions are sufficiently mild or infrequent, it may not
 be profitable to stockpile, in which case total U.S. economic losses would
 not be reduced by stockpiling.)

 A U.S. stockpile that would be sufficiently large to replace completely
 normal imports during foreign supply discriptions, and hence prevent any
 domestic price increases, clearly could not be profitable for private
 stockpilers.  It would also be larger than the optimum stockpile that
 minimizes national losses (assuming it is not possible for the United States
 to export stockpiled material).  Nevertheless, such a stockpile may be closer
 to the appropriate size than the "import-elminating" stockpile considered in
 the second in second column of Table 2-1.  Thus, in the third column of Table
 2-1 we consider a "comprehensive" stockpile, sufficient to make up for normal
 U.S. imports of (Q(j-Qp-Qs)D during the disruption.  The cost of holding the
 stockpile is indicated in the third column of Table 2-1.   All  other entries
 in the third column are zero because there is no price increase during
 disruptions in the U.S. market.  (U.S.  consumers suffer no losses, U.S.
 producers experience no gains, and the revenue from sale of U.S. stockpiles
 just equals the acquisition cost.)

 (There would be some justification for estimating U.S. materials criticality
 using estimates of actual  U.S. stocks,  rather than the hypothetical  stock
 sizes considered in Table 2-1.  However,  available data on U.S.  stocks are
often incomplete, notably for the platinum-group metals,  as discussed in
Chapter 4.   There are also theoretical  reasons for considering the
hypothetical  stock sizes in a simple model  that does not recognize any
relationships between stock sizes and the severity of price increases during
the specified foreign supply disruptions.)
                                   2-45

-------
                                                                   Charles
                                                                   River
                                                                   Associates
A computer program to perform the simple calculations described in Table 2-1
is presented in Appendix 2-B.


MEANINGFULLNESS OF RESULTS FROM THE MODEL

It has undoubtedly occurred to the reader, before reaching this point in the
analysis, that estimating the criticality of materials is a highly inexact
science, particularly as regards specifying the severity, duration and
frequency of disruptions threatening material markets.  It can be instructive
to review the history of supply disruptions and other contingencies in the
markets of interest.  But situations and threats change too rapidly for
historical analysis to provide definitive results.  For example,  historically
South Africa has been a quite reliable source of platinum-group metals,
chromium, and other materials often identified as critical.  And yet its
reliability into the 1990s and beyond has been a source of concern to U.S.
policymakers.  Clearly we must often rely on quite subjective estimates of
the likely severity, duration, and frequency of contingencies threatening
particular markets in order to obtain the most relevant measures of materials
critical ity.

The results we report below are of course also limited by the simplicity of
our economic model, and the fact that we are only specifying one type of
market contingency.  Clearly there is in fact some probability of any of a
wide range of disruptions, characterized by different severities  and
durations.  Nevertheless, our specification of a single,  representative
disruption (severity, duration, and frequency)  for each market can still  be
a roughly valid basis for comparisons among markets.   These specifications
distill  considerable CRA experience in analyzing contingencies in material
markets using more sophisticated models.  The results we report here are
probably in the same ballpark as criticality estimates that would be obtained
with much more extensive applications of much more sophisticated  models.
Moreover, the results reported here have the advantage that they  clearly
indicate the reasons that the criticality of materials used for vehicular
emissions control  equipment differ so markedly.

For the U.S. Department of the Interior, CRA has been working on  a quite
sophisticated model specifically designed to simulate reliably much more
severe disruptions in the platinum and palladium markets  than have occurred
historically.  Elements of this model  are discussed in Chapter 4.   The model
is dynamic, so that,  for example,  the longer a  disruption lasts the more
consumers can adjust away from materials that have become more costly.   The
linkage between past consumption of plantium and palladium and current
secondary recovery is explicitly recognized.  Supply  and  demand conditions
abroad,  as well  as in the United States, are recognized so that,  for example,
it is possible to  estimate the extent to which  greater U.S.  stockpiles will
                                     2-46

-------
Table 2-2
MATERIALS USED FOR VEHICULAR EMISSION CONTROL:
CRITICALITY FROM A NATIONAL PERSPECTIVE.
Material (form)

Platinum - Group
   Platinum (metal)
   Palladium (metal)
   Rhodium (metal)
                                          Value of Normal  U.S.
                                                                                          Price Elasticity of U.S.
Consumption
($ million)
729.0
123.0
76.0
Primary
Production
PoQp
($ million)
1.4
1.3
0.2
Secondary
Production
PO°S
($ million)
315.0
47.5
23.8
Consumption
ed
0.03
0.045
0.03
Primary
Production
eP
1.3
0.3
0.4
Secondary
Production
0.015
0.015
0.015
Other Materials
   Chromium (ferro)
   Manganese (ferro)
   Nickel  (metal)
   Titanium Metal  (sponge)
686.0
792.0
1,590.0
253.0
0
14.4
200.0
0
56.4
0
100.0
11.0
0.07
0.04
0.08
0*
0
0.1
0.08
0*
0.05
0
0.02
0.01
*For titanium,  the price  elasticities of demand and primary supply  are  set  equal to zero, as an (Imperfect) adjustment for the fact that
the specified contingency of concern Involves an Increase  1n demand creating a processing bottleneck.  See the discussion 1n the text.
Table continued on following page.
                                                                                                                             Charles
                                                                                                                             River
                                                                                                                             Associates

-------
ro
CO
Table 2-2 (continued)
MATERIALS USED FOR  VEHICULAR EMISSION CONTROL:
CRITICALITY FROM  A  NATIONAL PERSPECTIVE.
Material  (form)
Platinum - Group
   Platinum (metal)
   Palladium (metal)
   Rhodium (metal)

Other Materials
   Chromium (ferro)
   Manganese (ferro)
   Nickel  (metal)
   Titanium Metal  (sponge)
                                                  Supply Disruption
                                                                                                    Expected annual  U.S.  losses  ($ million/year) with
Severity
(Proportional Increase
In World Price)
X
7
6
7
5
4
3
3
Duration
D
2.5
2.5
2.5
2
1.5
1
1
Expected Time
Between
Disruptions
T
20
20
20
20
20
15
15
No Stockpiling
246.0
36.0
32.9'
211.0
164.0
152.0
32.3
Import-Eliminating
Stockpile
103.0
17.0
11.9
96.1
77.3
84.4
15.7
Comprehensive
Stockpile
67.0
12.1
8.5
'81.8
75.8
83.9
15.7
SOURCE:   Charles  River Associates, 1981.
                                                                                                                              Charles
                                                                                                                              River
                                                                                                                              Associates

-------
ro
i
            Table 2-3
            MATERIALS USED FOR VEHICULAR  EMISSION CONTROL:
            CRITICALITY FROM THE PERSPECTIVE OF THAT END USE
Material  (form)
Platinum - Grown
   Platinum (metal)
   Palladium (metal)
   Rhodium (metal)

Other Materials
   Chromium (ferro)
   Manganese (ferro)
   Nickel  (metal)
   Titanium Metal  (sponge)
                                                    Vehicular Emission Control
                                                                                               Supply Disruption
Value of
U.S. Consumption
PoQd*
($ million)
280.0
18.7
18.0
10.0
0.3
2.0
3.8
Price Elasticity
of Consumption
*d*
.03
.03
.03
0.25
0.33
0.5
0.5
Severity
(Proportional Increase
1n World Price)
X
7
6
7
5
4
3
3
Duration
D
(years)
2.5
2.5
2.5
2
1.5
1
1
Expected Time
Between Disruptions
T
(years)
20
20
20
*
20
20
15
15
            TOTAL
                                          332.8
                 Table continued on following page.
                                                                                                                                          Charles
                                                                                                                                          River
                                                                                                                                          Assnrintec

-------
            Table 2-3  (continued)
           MATERIALS USED FOR VEHICULAR EMISSION  CONTROL:
           CRITICALITY FROM THE PERSPECTIVE OF  THAT END USE
ro
en
o
Material (form)

Platinum - Grown
   Platinum (metal)
   Palladium (metal)
   Rhodium (metal)

Other Materials
   Chromium (ferro)
   Manganese (ferro)
   Nickel  (metal)
   Titanium Metal  (sponge)
                                                        Expected Annual  Consumers Losses  ($  million/year) With
No Stockpile


   191.0
    10.8
    12.3


     2.00
     0.03
     0.13
     0.25
                                                                          Import-ElIminatlng
                                                                              Stockpile
56.2
 3.5
 3.6


 2.00
 0.03
 0.13
 0.25
                       Comprehensive
                          Stockpile
45.5
 3.0
 2.9


 1.30
 0.03
 0.13
 0.25
           TOTAL
                                         216.51
                               65.71
                            53.11
           SOURCE:  Charles River Associates,  1981.
                                                                                                                                           Charles
                                                                                                                                           River
                                                                                                                                           Associates

-------
                                                                   Charles
                                                                   River
                                                                   Associates
diminish the severity of price increases on the world market during
disruptions.  We had hoped to be able to report results from these
sophisticated platinum and palladium models in this study,  but unfortunately
appropriate results are not yet available.

On the other hand, there are definitive advantages to applying the  same
simple model to all markets.  The model used here, though simple, recognizes
all of the potentially crucial determinants of materials criticality.
Moreover, these determinants can be tabulated to indicate readily why
expected economic losses are much different in one case versus another.   For
present purposes, clarity and organized analysis of the issues are
undoubtedly more important than refined estimates of the second digit  of the
criticality measure.


         SAMPLE ESTIMATES OF THE CRITICALITY OF PLATINUM, PALLADIUM,
          RHODIUM, CHROMIUM, MANGANESE, NICKEL, AND TITANIUM METAL


As explained in some detail above, the most useful measure of the criticality
of a material is expected losses (usually predominantly economic) associated
with the threat of various market contingencies, the most important of which
often involves disruptions of foreign  sources of supply.  Expected  economic
losses can be calculated from the perspective of the nation as a whole,
netting gains and losses of domestic producers and consumers, or from  the
perspective of a particular end use, such as control of vehicular  emissions.

Table 2-2 presents sample estimates of the criticality -- from a national
perspective — of seven elemental materials used for vehicular emissions
control in the United States, using the economic model described in Table
2-1.  The required parameter estimates are presented on the first  page of
Table 2-2, and expected economic losses (calculated under three different
assumptions about stockpiles that will be available during the contingency)
are presented on the second page.

Table 2-3 presents the corresponding criticality estimates calculated  from
the perspective of consumption for vehicular emissions control.  The severity
(X), duration (D), and frequency (1/T) of the market contingency are repeated
from Table 2-2.

In the following discussion, we first  explain our projections of the values
of consumption and production for the  seven raw materials.  Then we specify
price elasticities of consumption and  production, and the severity, duration,
and frequency of a representative contingency threatening each market.  Given
these parameters of the criticality calculation, we finally interpret the
resulting estimates of expected economic losses given in the last three
columns of Tables 2-2 and 2-3, and draw out the policy implications for EPA.
                                     2-51

-------
                                                                   Charles
                                                                   River
                                                                   Associates
NORMAL CONSUMPTION,  PRODUCTION,  AND PRICES

The parameter estimates in Tables 2-2 and 2-3 are designed to  be  roughly
appropriate for the mid-1980s in the United States.   Trend projections  of
total U.S. consumption and production under normal  conditions  were  obtained
from Mineral Trends and Forecasts, U.S.  Bureau of Mines (1979).   Appropriate
estimates of other parameters are much more uncertain than these  quantities,
so any reasonable projections of total  U.S. consumption and production  are
adequate for our purposes.

Total U.S. primary production (Qp) includes only production from  domestic raw
materials.  Total U.S. secondary production (Qs) includes only production
from oT_d or obsolete scrap.  Production  from new scrap, generated by
processing before materials are sold in  final  goods (to consumers,  investors,
or the government),  is not explicitly considered because its supply can
usually not be expanded significantly, even under the incentive of  much
higher prices for the material.   Correspondingly, total U.S. consumption  (Qj)
includes consumption out of primary production and old scrap,  but consumption
of new scrap is netted out.  (For platinum, palladium, and rhodium, the
estimates of secondary production (Qs) and total consumption (Qj) from  U.S.
Bureau of Mines  (1979) have been roughly adjusted upward by CRA to  include
"toll refining" of the metal, where secondary refining of used metal is
undertaken on a fee-for-service basis, with the consuming industry
retaining ownership.)

The values of materials reported in Tables 2-2 and 2-3 are based  upon the
following market prices, projected to be roughly appropriate for  normal
market conditions in the mid-1980s:
•    Platinum

•    Palladium

•    Rhodium

•    Chromium


•    Manganese


t    Nickel

•    Titanium
     Metal
$450 per troy ounce of metal

$95  per troy ounce of metal

$475 per troy ounce of metal

$940 per short ton of elemental  chromium contained in
ferrochromium

$480 per short ton of elemental  manganese contained in
ferromanganese

$5,700 per short ton of metal


$11,000 per short ton of metal  in the form of sponge.
                                     2-52

-------
                                                                   Charles
                                                                   River
                                                                   Associates
These prices (labeled P0) are measured in constant 1981 dollars, so that
inflation has no direct effect on the values reported in Tables 2-2 and 2-3.
(For simplicity, these values (that is, price P0 times quantity produced or
consumed) are calculated as though all of U.S. production or consumption is
processed through the indicated form of the material; this assumption is
least appropriate for chromium, where a majority of U.S. consumption indeed
requires production of ferrochromium or similar ferroalloys, but a
substantial proportion of U.S. consumption does not.

The first column of Table 2-3 estimates the value of U.S. consumption of each
of the seven elemental materials for vehicular emissions control  in 1985 to
1987.  The prices specified above (P0) are applied, while quantities consumed
(Q
-------
                                                                   Charles
                                                                   River
                                                                   Associates
Clearly leverage is greatest in the market for platinum,  and is  negligible
for the four alloying elements.


CONTINGENCY THREATS AND PRICE ELASTICITIES

We now consider briefly major contingencies threatening the markets  for  the
seven elemental materials used for vehicular emissions control,  and  specify
the severity, duration,.and frequency of a representative contingency
(usually a foreign supply disruption) threatening each.  The parameter
estimates are reported in Tables 2-2 and 2-3.   As discussed earlier,
specification of a single contingency can only begin to characterize  the
range of possible events in various material markets.   Our main  concern  is
that the relative sizes of the single disruption threats  be roughly
appropriate when comparing one market with another,  and that the resulting
measure of expected U.S. economic losses be in the right  ballpark.   In the
following discussion of each market we also specify  all elasticities  that
characterize the responsiveness of U.S.  consumption  and production to  much
higher market prices.


PLATINUM, PALLADIUM, AND RHODIUM

As discussed at greater length in Chapter 4, western consumption of  the
platinum-group metals is supplied predominantly by South  Africa  and  the
Soviet Union.  There is clearly the potential  for a  very  severe  cutoff in
primary world supplies.  Almost all platinum-group mining outside South
Africa is a byproduct of nickel and copper which implies  that output  from
alternative sources would not expand significantly in  response to larger
increases in the price of platinum-group metals.

Moreover, the demand for platinum-group  metals tends to be very  unresponsive
to price increases.  We estimate in Chapter 4  that a five-fold price  increase
would only cut platinum and rhodium consumption by 12 percent, and palladium
consumption by 18 percent.  That is,


               EC! = (0.12/4) = 0.03 for platinum and rhodium
                                     and
               ed = (0.18/4) = 0.0045 for palladium
                                    2-54

-------
                                                                   Charles
                                                                   River
                                                                   Associates
The major consideration mitigating the severity  of disruptions  in  world
supplies of platinum,  palladium,  and rhodium is  secondary  recovery.   There
are very large quantities of these metals currently in use as catalysts,  and
in other applications  where a very high percentage can be  recycled.   During
disruptions in primary supplies,  the natural  operation of  the price  system
would redirect some of this recycled material  toward the most critical
end uses (that is, to  those uses  that are willing to pay the  most  for the
material).   However, because secondary recovery  is already so extensive  in
the platinum-group markets, there are only very  modest opportunities to
expand recyling during emergencies.  We specify  secondary  recovery elasticity
es= 0.015  for all three platinum-group metals under consideration.

For purposes of the sample criticality measurements in Tables 2-2  and 2-3, we
specify a seven-fold price increase (X=7) for platinum and rhodium during
disruptions.  We specify the probability of a disruption to be  sufficiently
high so that the expected time between disruptions is T =  20  years.   Because
South Africa or the Soviet Union  could be cut off from the United  States  for
a long time, we specify disruption duration D = 2.5 years. For palladium we
specify a somewhat less severe price increase during disruptions (X=6)
because, worldwide, price responsiveness is somewhat greater  on the  demand
side (and perhaps even on the supply side) of the market.

There may be significant mining of platinum-group deposits at the  Stillwater
complex in the United  States by the 1990s, but we do not factor that
possibility into our analysis.  It would not be possible to create entirely
new underground U.S. capacity quickly enough after a supply disruption has
started to produce much during the first two or three years of  the
disruption, but existing or abandoned sources, especially  placer deposits and
the old mine at Goodnews Bay, Alaska, could be expanded or activated fairly
quickly.  The Goodnews Bay deposit yields almost entirely  platinum,  and  its
reactivation justifies a large primary supply elasticity for  platinum in the
United States.  We specify ep = 1.3 for platinum.  There is actually more
rhodium than palladium in the Goodnews Bay deposits, justifying a  slightly
higher overall U.S. supply elasticity for rhodium.  We specify  ep  =  0.3  for
palladium and ep = 0.4 for rhodium.  These price elasticities of primary
supply are quite large, but they  have relatively little effect  on  the
calculations because normal U.S.  primary production is so  low.


CHROMIUM

South Africa is the dominant producer of chromium for western markets.
Neighboring Rhodesia produces substantial quantities of high-grade ores
suitable for metallurgical applications.  Turkey and the Philippines also
produce substantial quantities.  The Soviet Union formerly exported  large
amounts of chromium to the West,  but in recent years Albania  has become  the
important Communist exporter to the West.  A complete disruption in  southern
                                     2-55

-------
                                                                   Charles
                                                                   River
                                                                   Associates
Africa (South Africa and Rhodesia)  would eliminate a large proportion of
western supplies and cause substantial,  sustained price increases  on the
world market.  We specify disruption severity X = 5 for common  grades
of ferrochromium, and disruption duration D = 2 years.   We specify the time
between disruptions to be T = 20 years,  the same as for the platinum-group
metals, which are also obtained from South Africa and Communist countries.

The United States does not produce significant amounts  of primary  chromium
under normal market conditions.  A severe disruption such as that  specified
in Tables 2-2 and 2-3 would induce some production from small  U.S. deposits
in California and elsewhere, but our results would not be changed
significantly by recognizing this complication, so we simply specify £p = 0.
(A slightly generalized version of the economic model would be required to
recognize zero production at normal prices, with significant production
beginning at some higher price.)

Chromium is used predominantly for stainless steel and other steel alloys.
It is also used for refractories and in chemicals without first being
processed into ferrochromium.  Consumption for chemicals is most responsive
to price changes, while consumption for stainless steel is least responsive.
There are very limited opportunities for reducing the chromium content of
stainless steels while still retaining the corrosion resistance at higher
temperatures required for the most demanding applications of stainless steel.
However, less than 20 per cent of stainless steel consumption is for
demanding applications such as turbines, while many other end uses could
substitute coated steels, plastics, and other materials.  A five-fold
increase in the price of ferrochromium corresponds to roughly an 80 per cent
increase in the price of stainless steel, which would lead to substantial
conservation of stainless steels.  Such a large price increase would reduce
chemical consumption of chromium by considerably more than half.  We specify
the overall U.S. price elasticity of demand for chromium (valued in the form
of ferrochromium) to be e^ = 0.07.

Approximately eight per cent of U.S. chromium consumption is from secondary
sources, predominantly scrapped stainless steel used to make new stainless
steel.  There are modest opportunities for increasing stainless scrap
recovery, for example by more carefully sorting stainless scrap from the
more common types.  (A CRA pundit once remarked that there might also be
significantly increased recovery of stainless steel hubcaps through illicit
channels.)  We specify the price elasticity of secondary supply to be
£s = 0.05.


MANGANESE

The Soviet Union produces more manganese than any other country by a
considerable margin, but uses the material intensively in its domestic steel
industry, and exports little to the West.  South Africa is by far the
                                     2-56

-------
                                                                   Charles
                                                                   River
                                                                   Associates
largest exporter to the industrialized western  countries,  followed  by  Brazil,
Gabon, Australia, and India.  Southern Africa is not as important in the
manganese market as in the chromium market,  and there are  a  larger  number  of
alternative suppliers with more extensive possibilities for  expanding
production.  Thus, we specify the disruption severity to be  less  (X =  4) and
the disruption duration less (D = 1.5 years).  We keep the same disruption
frequency as for the other materials obtained from South Africa  (time
between disruptions  T = 20 years).

Well over 90 percent of U.S. consumption of  manganese is for steelmaking,
where it is an indispensable constituent of  virtually all  steels.   There is
a small amount of flexibility in the quantity of manganese used per ton of
steel, and there could be a small  amount more if manganese specifications  for
steels were modified to reflect technologically minimal  needs under the
current state of the art of steelmaking.  Also, manganese  that occurs
naturally in many iron ores and in steel scrap, and is normally lost in
processing, can be conserved by such procedures as slag recycling.  We
specify the overall U.S. demand elasticity to be ej =  0.04.

Projected 1985 U.S. primary production of manganese is very  modest, though it
could probably be expanded substantially during a sustained  serious
disruption of foreign supplies.  (The Bureau of Mines may  be assuming  some
production from ocean nodules.)  We specify  the domestic price elasticity  of
supply to be es = 0.1.


NICKEL

The most common cause of supply disruptions  in the world nickel market has
been labor strikes at Canadian mines, particularly at the  dominant  Canadian
producer INCO.  The Canadian market share has dropped greatly over  the last
two decades, to less than 40 per cent, so such labor unrest  is less critical
to the United States and other nickel importing nations than was  earlier the
case.  Newer producers, such as Australia and the island of  New Caledonia  (an
"overseas department" of France, in the southwestern Pacific Ocean) have been
more stable, and in any case represent alternative sources of supply  .
(Canadian production of nickel is discussed  briefly in Chapter 4, because
Canadian production of platinum-group metals is predominantly as  a  byproduct
of nickel.)

Strikes tend not to last for very long, in comparison with other  sorts of
contingencies, such as civil wars, so we specify disruption  duration D = 1
year.  Price increases  during strikes will  affect purchasers forced to go to
the spot market more than those under undisrupted long-term  contracts, so  we
specify the relative price increase to U.S.  consumers to be  only  X  = 3.
However, we specify somewhat more frequent disruptions occurring  on the
average of every T = 15 years.
                                     2-57

-------
                                                                   Charles
                                                                   River
                                                                   Associates
Nickel is used in stainless steels,  alloy steels,  nonferrous alloys, and is
also used for electroplating.   Limited  substitution  is possible, either
through use of different alloys,  or  use of different types of materials
altogether.  We specify the U.S.  price  elasticity  of demand to be e^ = 0.08.
(The appropriate elasticity would be somewhat  greater if the disruption
lasted longer than one year.)

U.S. primary production of nickel  is modest, normally accounting for between
10 and 15 percent of U.S. consumption.   There  has  been a fair amount of
excess capacity in recent years,  and production  could probably be expanded
significantly over the course  of  a year.  We specify the U.S. price
elasticity of supply to be es  = 0.08.

There is modest secondary production of nickel  from  obsolete scrap, with
limited possibilities for expanding  this recovery  during periods of high
nickel prices.  We specify the price elasticity  of secondary U.S. supply to
be es = 0.02


TITANIUM

The critical ity of titanium is unlike that for the other metals we have
considered, in that disruptions in supplies of the elemental raw material are
not a serious threat, but "processing bottlenecks can be.  Most titanium is
mined from deposits of the ore ilmenite, and is  used in very mundane
applications as a white pigment.   Less  than five percent of titanium is
processed  into a metallic "sponge" and  then metal, the majority for
applications in aerospace hardware.   Some is also  used in  steel alloys.
(Most metal is produced from the  ore rutile, rather  than ilmenite, though it
is possible at slightly higher cost  to  make metal  from ilmenite if the
appropriate processing capacity has  been constructed.)

The most likely bottleneck in  the production of titanium metal is in the
capacity to make sponge from rutile.  In fact, world production was seriously
constrained by sponge capacity in 1980  and 1981, causing a substantial
increase in prices over long-run  equilibrium levels. Under incentives of
very high  prices, new sponge plants  can be built in  a year or a year and a
half.  Thus, disruptions in the supply  of titanium metal are not likely to
last a long time.  We specify  duration  D = 1 year.  Consumers under long-run
contracts  with producers may not  suffer a great deal, which affects our
specification of a relatively  low disruption severity X =  3.

By far the most common cause of insufficient sponge  capacity is difficulty  in
predicting the demand for titanium metal, as determined by such factors as
the rate of macroeconomic activity in industrialized countries and decisions
about military spending.  As discussed  earlier,  there is some question
                                     2-58

-------
                                                                   Charles
                                                                   River
                                                                   Associates
whether the threat of such common commercial  contingencies  truly  qualifies  a
material to be considered "critical."   Most of the same issues  arise,  but our
very simple economic model is not really set up to analyze  situations  where
the U.S. demand curve will likely be shifting outward when  the  contingency
occurs.  We ignore this analytical complication for the present analysis,
except  (as an imperfect adjustment) we specify negligible price elasticities
of domestic demand and supply, ej = 0 = es.  In recent years the  United
States  has imported increasing amounts of titanium sponge from  Japan and the
Soviet  Union.  Disruptions in these imports could also impose costs on the
United  States.

Only modest amounts of titanium metal  scrap are generated from  obsolete
aircraft and chemical processing equipment.  In most cases, scrappage  of
obsolete titanium equipment would not be accelerated simply because of higher
scrap values, since titanium values are typically very small relative  to the
total value of the equipment.  Thus, we specify a very small price elasticity
of secondary production, es = 0.01.


PRICE ELASTICITIES OF CONSUMPTION FOR VEHICULAR EMISSIONS CONTROL

We have not investigated  in any detail the efficiency of substituting for 409
stainless steel and other alloys in emissions control systems.   Substitution
of materials such as aluminized steel  has been proposed.  The main cost
imposed by such a substitution would apparently be reduced durability,
necessitating replacement costs borne by car owners.  EPA durability
requirements might have to be relaxed to allow such a substitution.  It is
said that aluminized steels tested to date lasted less than half as long as
409 stainless steel in catalytic converters.

Under the price incentives that we assume to exist during disruptions (see
Table 2-2), we suppose for purposes of calculations that use of any of these
alloying materials can be eliminated (not necessarily simultaneously,
however).  The implied elasticities of demand for vehicular emissions control
Ud) are as follows:

                             •  Chromium:  0.25

                             •  Manganese:  0.33

                             •  Nickel:  0.5

                             t  Titanium metal:  0.5

The above elasticity estimates should be regarded as illustrative rather than
definitive.
                                      2-59

-------
                                                                   Charles
                                                                   River
                                                                   Associates
For platinum, pallodium, and rhodium,  we assume only very modest reductions
in usage would be possible in response to six-  and seven-fold  price
increases.  For purposes of illustrative calculations,  we specify price
elasticity ed = 0.03.


ILLUSTRATIVE CRITICALITY ESTIMATES AND CONCLUSIONS
Having explained all  the parameters required for  the  calculations,  we can now
compare the "criticality"  (average annual  ecoomic losses)  for  the  seven
elemental  materials described in Table 2-2 (national  perspective)  and Table
2-3 (vehicular emissions control only).   From a national  perspective,
platinum is the most critical of the materials considered,  but chromium,
manganese, and nickel  have average annual  losses  that are of the same order
of magnitude.  The severity of the supply disruption  threatening the platinum
market is ameliorated considerably by the extensive secondary  recovery of old
material that takes place in the United States.

The value of normal U.S. nickel  consumption is more than  double that of any
of the other materials, but the severity and duration of  disruptions is not
as great as for the other materials considered, and U.S.  production, both
from primary and secondary sources, is significant.   Also,  the price
elasticity of demand is greater for nickel  than for the other  materials
considered.

The value of U.S. manganese consumption is somewhat greater than the value of
U.S. chromium consumption, but the greater severity of the  disruption threat
for chromium makes it the more critical  material.

The criticality of palladium, rhodium, and titanium metal  from a national
perspective, is less than that of the other materials, largely because the
value of U.S. consumption in normal times is less.   The criticality of
palladium and rhodium is substantially reduced by secondary recovery, while
the representative contingency threatening the market for titanium metal ties
with nickel for being the least severe of those specified.

The above conclusions about the relative criticality  of the various materials
hold whether losses are calculated assuming no stockpiling  or  very large
stockpiles.  (A comprehensive stockpile is usually  only moderately larger
than an "import-eliminating" stockpile,  and both  are  large  relative to the
disruption threat.)

From the perspective of consumption for vehicular emissions control, platinum
is the most critical  material by more than an order of magnitude relative to
any other material considered here.  The value of platinum  consumption dwarfs
                                      2-60

-------
                                                                   Charles
                                                                   River
                                                                   Associates
that for the other materials, and no other material  faces a  more severe
disruption threat.  The criticality of palladium,  rhodium, and  chromium  is
potentially significant but relatively low.  The criticality of manganese,
nickel, and titanium is negligible.  The criticality of the  alloying
materials (chromium, manganese, nickel, and titanium metal)  is  lessened
considerably by large price elasticities of consumption (ej), reflecting our
assumption that alternative alloys (such as aluminized steels)  could  be  used
during such a truly severe supply disruption.

As noted earlier, the above calculations of the criticality  of  platinum,
palladium, and rhodium, do not explicitly recognize  the linkage between
current consumption of the material and possibilities for future secondary
recovery.  In the case of platinum-group metals used for vehicular  emissions
control, consumption taking place in the late 1970s  and 1980s will  develop
into a "rolling stockpile" of material that could  considerably  reduce U.S.
vulnerability to foreign supply disruptions in the 1990s. When this
consideration is fully taken into account, the criticality of platinum could
be considerably less than estimated above.  We have  not projected the size  of
the U.S. rolling stockpile of platinum-group metals  in this  study.  It will
be a simple exercise with the model mentioned earlier that CRA  has  designed
for the U.S. Department of the Interior.

The average annual losses from contingencies calculated in Tables 2-2 and 2-3
are much more illustrative than definitive, for reasons indicated throughout
the discussions above.  However, these calculations  clearly  provide the  right
types of information to guide policy decisions by  the Environmental
Protection Agency.  We finish our discussion in this chapter assuming the
criticality estimates in Table 2-3 are appropriate.   It will be clear how to
use any revised estimates that become available.

Consider in particular the menus of materials required for the  currently
projected emissions control  technology as specified  in Table 2-3.   The total
annual  cost for that menu, where all  the indicated material  markets are  in  a
normal  state, is approximately $330 million per year.   Expected annual losses
from contingencies in the various material  markets are approximately
$50 million per year, assuming large stockpiles.  (Private inventories are  in
fact typically substantial for the more critical materials;  Chapter 4
describes stockpiles for the platinum-group metals.)   Total  expected  annual
costs for the menu of materials, including disruption costs, is thus
approximately $380 million per year.

Suppose that an alternative emissions control technology presented  itself,
which was equally effective and otherwise like that  described in Table 2-3,
                                     2-61

-------
                                                                   Charles
                                                                   River
                                                                   Associates
except that total  expected annual  costs for a substitute  menu  of  materials
(including expected losses from market contingencies)  was less than  $380
million per year.   That alternative technology would have a lower overall
expected cost than the currently projected technology using platinum-group
catalytic converters.

We confirm in Chapter 6 that no new technology currently  on the horizon has
much promise of being more cost effective than platinum-group  catalytic
converters.  The critical  materials penalty of approximately $50  million  per
year constitutes only a very modest additional  expected cost over the
normal cost of $330 million for the indicated menu of materials,  particularly
when all the disadvantages of alternative control  technologies are
considered.

If a new, more cost-effective technology should be developed,  we  have
reasonable confidence that the U.S. auto industry would choose that
technology in a way that appropriately weighs the criticality  of  the required
materials.  The reasons for this conclusion were discussed earlier in  the
chapter.  They can be summarized by saying that, in this  case, there appears
to be little reason for a significant discrepancy between the  cost to  private
firms of complying with EPA regulations, and the cost to  society.  (This
fortunate situation is not always the case for other EPA  regulations.)

Moreover, assuming the U.S. auto industry believes that EPA emissions
standards of a given stringency will be in effect for many years  into  the
future, the industry should have sufficient incentive to  undertake the amount
of research and development that is appropriate to that stringency.  The
reason is analogous to that made above concerning appropriate  private
decisions about which materials to use:  The cost reductions made possible  by
a new, lower cost emissions control technology would benefit the  firm
developing the new technology to an extent comparable to  the total national
benefit from reduced control costs.  This optimistic conclusion,  that U.S.
firms will undertake the amount of research and development approximately
appropriate for a given EPA standard, will, however, not hold  if  U.S.  firms
seriously doubt that the standard will be maintained into the  future.

The  preparations that the U.S. auto industry has made for disruptions  in
foreign supplies of platinum-group metals, particularly the size  of
stockpiles they maintain, may be inadequate if they expect that relaxation  of
EPA  emissions standards can be arranged during those disruptions.  As
protection against this possibility, EPA might consider requiring U.S.
vehicle manufacturers to hold specified minimum levels of inventories  of
platinum, palladium, and rhodium.

The  fact that we can directly use our definition of materials  criticality to
choose compliance technologies having the lowest expected costs in the future
shows that we are using the appropriate concept.  The concept  of  materials
criticality is designed to facilitate decisions about what materials to use
                                     2-62

-------
                                                                   Charles
                                                                   River
                                                                   Associates


and what policies (such as stockpiling  and tariffs)  to  pursue,  to  protect
that rate of consumption of critical  materials  which is deemed  efficient.  In
this chapter we have emphasized the choice of materials to consume, but
extensions of the same methodology are  appropriate  for  choosing efficient
stockpiles and tariffs. Earlier investigations  have  summarized  appropriate
variables for determining materials criticality, but few have fashioned  a
quantitative measure of criticality appropriate for  direct inclusion in  the
decisionmaker's balancing of economic (and noneconomic)  costs and  benefits.
Appendix 2-C, which is extracted from a 1978  CRA study,  briefly surveys  some
of these alternative analyses.   Other work in this  area, particularly more
recent work, is summarized briefly in the  following  Bibliographic  Note.
                                    2-63

-------
                                                                   Charles
                                                                   River
                                                                   Associates
                             BIBLIOGRAPHIC  NOTE
To our knowledge, no other study has specifically  addressed  in any detail the
issue of how a line agency of the federal  government,  without direct
responsibility for materials policymaking,  should  factor  the criticality of
materials into its policymaking, for example where consumption of those
materials could be greatly increased by  the agencies'  decisions.  However,
many other studies have treated in great detail  more  general aspects  of the
problem, such as what contingencies threaten what  markets, and what national
policies are most effective in reducing  expected costs from  contingencies.
We briefly survey here (and in Appendix  2-C) studies  that are particularly
prominent, recent, unusual, or closely related  to  the  analysis we have
presented in this chapter.

The Office of Minerals Policy and Research  Analysis at the U. S. Department
of the Interior has taken important steps in recent years toward
implementating the type of methodology that we  recommend  for measuring the
criticality of materials.  See Adams, White, and Grichar  (1979), who  consider
the criticality of bauxite, cobalt, copper, iron,  and  nickel.  The analysis
by Adams, White, and Grichar is notable  for considering a range of disruption
severities and for formally surveying (by "modified Delphi"  techniques)
market experts to estimate disruption probabilities.   An  earlier
methodological and empirical study by the same  group  at the  Department of the
Interior estimated expected economic costs from potential disruptions in the
markets for aluminum, chromium, platinum, and palladium.  See U.S. Department
of the Interior (1975).  This work was reviewed and revised  in Charles River
Associates (1977).  Two shorter articles summarizing  basic methodologies and
conclusions are Adams (1977), and Burrows and Beggs (1977).

The CRA study that treats methodologies  for estimating the criticality of
materials in most general terms is Charles River Associates  (1978).   The
chapter from that report which summarizes alternative  methodologies for
measuring the criticality of materials is reproduced  as Appendix 2-C  to this
chapter.  The most comprehensive CRA study in this area is the multi-volume
Charles River Associates (1976), which estimated economic costs from  problems
in the markets for platinum, palladium,  chromium,  manganese, and three other
non-energy materials, as well as the market for petroleum.   Other CRA studies
treating the criticality of materials which the United States imports, and
appropriate policy responses at the national level, include: Charles River
Associates (1977a), Charles River Associates (1976b),  Charles River
Associates (1976a), and Charles River Associates (1975).  Much of the CRA
work referred to above is summarized in  the book Klass, Burrows, and  Beggs
(1980), which specifically treats the same material markets  as Charles River
Associates (1976).
                                     2-64

-------
                                                                   Charles
                                                                   River
                                                                   Associates
Since World War II, a number of landmark studies of problems  and
contingencies affecting U.S. material  markets have been  conducted  by
government-sponsored commissions.  Among the most prominent were the  U.S.
President's Materials Policy Commission (1952) and the National Commission  on
Materials Policy (1973).  Most studies by major government commissions
focused on general  issues in mineral  markets (such as resource  depletion,
possibilities for more recycling, appropriate amounts of research  and
development, and control of pollution from domestic mineral production),
rather than specifically on the types of contingencies that make a material
"critical."  In the 1970s, several  major government studies appeared  that
focused more particularly on those types of contingencies, particularly
disruptions of foreign supplies.  A short but useful  example  is the
Special Report on Critical Imported Materials by the Council  on International
Economic Policy (1974).

The most sophisticated of the government studies undertaken in  the 1970s was
that by the National Commission on Supplies and Shortages (1976),  which  went
into considerable depth on ways of improving federal  policymaking  on  material
markets at the national level.  Also notable for breadth in covering  issues
affecting material  markets, and appropriate policymaking at the national
level, are drafts produced by the recent Domestic Policy Review of Nonfuel
Minerals.  See U.S. Department of the Interior  (1979).

The third volume of U.S. Department of the Interior (1979) is entitled _A
Compendium of Issues, Options, and Recommendations Contained  in Major
Post-war Nonfuel Mineral Policy Studies^It is available through  the
Superintendent of Documents of the U.S. Government Printing Office.   This
large volume provides a comprehensive bibliography and a collection  of  key
quotes from most of the significant studies of national  policymaking  for
nonfuel minerals undertaken since 1945.  It includes many references  that  we
do not mention here because they are less directly relevant,  less  prominent,
less incisive, or representative only of special interests.  We recommend
this source if more extensive bibliographic information than  we provide  here
is desired.

The Brookings study by Til ton (1977) is to be recommended for sensible
economic analysis of a broad range of contingencies and problems  currently
facing minerals industries.

A recent study conducted at Resources for the Future, Inc. evaluates
contingencies in the markets for cobalt, chromium, manganese, aluminum,
copper, lead, and zinc, concluding that only for chromium does  the United
States clearly face "undue vulnerability ... to contingencies that might
either seriously disrupt supplies or cause sharp upward movement  of prices,
with consequent serious economic impact."  No quantitative estimate of
economic costs or the criticality of the various materials is developed,
nowever.  See Fischman  (1979), p. 1, and also an abridged version  issued in
book form, Fischman (1980).
                                     2-65

-------
                                                                   Charles
                                                                   River
                                                                   Associates
A study by Zabrowski and Lyle (1978) at the old Federal  Preparedness  Agency
(now the Federal Emergency Management Agency)  investigates  certain
characteristics of the input-output matrix for an  economy that  bear on the
criticality of its material industries, as we  would  recommend measuring  it.
However, the analysis appears incomplete as a  basis  for  choosing materials to
utilize, or policies to ameliorate supply disruptions, in part  because it
does not consider the likelihood of contingencies  of various severities.

A recent report by Kern 0. Kymn (1980) at the  U.S. Federal  Emergency
Management Agency is notable for assessing criticality of a material  (steel
is used as an example) stemming from labor strikes,  rather  than foreign
supply disruptions.

It is worth noting that planning for contingencies affecting the price and
availability of materials can be considerably  more unwieldy than the
situation faced by EPA in planning for vehicular emissions  control.   The
principal investigator of this study recently  participated  in a panel for the
National Materials Advisory Board, leading to  the  short  report  entitled
Identification of Critical and Strategic Materials for Naval Combat Systems,
NMAB (1981).Navy systems are so complex that it  is inordinately costly to
trace all the materials used back through third and  fourth  level vendors to
the suppliers of the original raw materials, despite the fact that military
purchases of components are much better documented than  most commercial
purchases.   As a result, production bottlenecks can  spring  up in unexpected
places in the chain of materials supply.

There are,  of course, a multitude of sources of information and data  on  the
markets for individual materials.  Those for the platinum-group markets  are
reviewed in Chapter 4.  Similar sources of information are  available  for
chromium, manganese, nickel, titanium, and other potentially critical
materials.   Publications of the U.S. Department of the Interior provide
particularly useful  summaries of relevant information.   The Annual Report of
the Secretary of the Interior Under the Mining and Minerals Policy Act of
1970 usually provides a useful  summary of policy issues  in  minerals markets
that are considered most pressing at the national  level.  The U.S. Federal
Emergency Management Agency (FEMA) makes quarterly (National Defense) Stock-
pile Reports to Congress, describing the operative legislation, the current
status of the stockpile, goals for the future  size of the stockpile,  and
recent activities (including a separately bound statistical supplement).
                                     2-66

-------
                                             Charles
                                             River
                                             Associates
           APPENDIX 2-A


AVAILABILITY OF MATERIALS  FROM THE
  U.S. NATIONAL DEFENSE STOCKPILE
                2-67

-------
                                                                   Charles
                                                                   River
                                                                   Associates
                               APPENDIX    2-A

                     AVAILABILITY OF MATERIALS  FROM  THE
                       U.S.  NATIONAL DEFENSE STOCKPILE

In this appendix we consider possible avenues by  which materials  can  be
released from the National  Defense Stockpile for  non-military  contingencies.
Our general  conclusion is that an act of Congress would  almost certainly  be
required to authorize release of materials required  for  vehicular emissions
control during a non-military emergency.

The U.S. National Defense Stockpile was  established  just prior to World War
II.  Its sole objective is  to "serve the interest of national  defense."   It
is "not to be used for economic or budgetary purposes."   The enabling
legislation for the stockpiling, amended in 1979, further requires that the
quantities of material stockpiled are to be  "sufficient  to sustain the United
States for a period of not less than three years  in  the  event  of  a national
emergency."  Stockpile goals are estimated by the Federal Emergency
Management Agency (FEMA) as the difference between requirements for materials
in wartime and the amount likely to be available, assuming some austerity in
non-military consumption.

The Strategic and Critical  Materials Stockpiling  Act (as amended  July 30,
1979) allows disposal of materials only  under  specified  authorities,  the  most
general of which is an act of Congress.   The President can authorize
disposals under more narrowly defined conditions.  Perishable  materials are
to be "rotated" out of the stockpile to  prevent deterioration. With  prior
notification of the Committees on Armed  Services  of  the  Senate and House  of
Representatives, the President can dispose of materials  that  are  in excess of
stockpile requirements, or that may deteriorate in value if not sold.

Beyond the above described types of stockpile releases,  which  are primarily
designed for routine management of the stockpile, Section 7 of the Strategic
and Critical Materials Stockpiling Act also  grants the President  more
discretionary authority.  In time of war declared by Congress, or in  time of
a declared national emergency, the President or his  delegate  may  release
materials specifically required for the  national  defense.  The President  also
has the power during times other than declared  emergencies or wars to release
material specifically required for purposes  of  the national defense.

Historically, there have been 28 releases authorized by  the President under
Section 7; all but three of these were made  during World War  II,  the  Korean
Conflict, or the Viet Nam Conflict.  All  releases under  Section 7, including
those made during wartime,  have been made directly by the President,  rather
than through delegated authority.  The three releases made at other times
consisted of mercury (in 1956 and 1959)  and  asbestos (in 1979).  The  critical
                                     2-68

-------
                                                                   Charles
                                                                   River
                                                                   Associates
consideration in all these releases was that the material  would be used
directly for defense purposes.  Since use of material  for control  of
vehicular emissions ostensibly does not constitute application to  national
defense, it is highly doubtful that releases under Section 7 could be made  by
the President for that purpose.

It is conceivable that platinum-group metals or other  material  required for
vehicular emissions control, might be available from the stockpile,  because
amounts held in excess of requirements are due to be sold.  (There is no
excess currently.)  However, in this case the President is required  to use
competitive negotiation and formal advertising, unless prior explanation is
made to Congress.  In the usual case of competitive bidding, producers of
emissions control equipment would have to vie with other consumers for the
materials they wish to obtain.

There is one further avenue under which the President has in the past
authorized releases from the National Defense Stockpile.  Section  101 (b) of
the Defense Production Act of 1950 (as amended August 20,  1980),  gives the
President general authority to allocate materials in the civilian  market when
he finds:

       (1) that such material is a scarce and critical material  essential to
       the national defense, and (2) that the requirements of the  national
       defense for such material  cannot otherwise be met without creating a
       significant dislocation of the normal  distribution  of such  material  in
       the civilian market to such a degree as to create appreciable
       hardships.

In two cases, historically, it has been found that operation of the  defense
priorities system has created sufficient "hardships" in the civilian market
that releases from the national stockpile were justified.   The materials
released were argon gas and titanium.  Since contingencies other than
operation of the defense priorities system seem most likely to affect
materials required for vehicular emissions control, this avenue for  arranging
releases from the National  Defense Stockpile also does not promise to be very
useful  for meeting EPA related materials consumption in a  non-military
emergency.

The general  conclusion that emerges from the above review of peacetime
releases from the national  stockpile is that almost certainly an act of
Congress would be required to respond to the types of  non-military
contingencies that appear most likely in the market for platinum-group
metals.  This conclusion is particularly clear if EPA  were to wish that the
material  be specifically allocated for use in vehicular emissions  control
equipment, rather than being put up for competitive bids among all consumers
in the market.   A FEMA official with whom we discussed the matter  concurred
with this conclusion.
                                    2-69

-------
                                                       Charles
                                                       River
                                                       Associates
                   APPENDIX  2-B


COMPUTER PROGRAM TO CALCULATE AVERAGE  ANNUAL  ECONOMIC
   LOSSES FROM CONTINGENCIES IN  MATERIAL  MARKETS
                        2-70

-------
                                                                   Charles
                                                                   River
                                                                   Associates
                               APPENDIX   2-B
        COMPUTER PROGRAM TO CALCULATE  AVERAGE  ANNUAL  ECONOMIC  LOSSES
                   FROM CONTINGENCIES  IN MATERIAL MARKETS
This appendix lists a simple computer program designed  to  calculate  the
economic losses determining the criticality  of a material.  The very  simple
economic model  upon which the program is based is  described in the text of
the chapter.   The formulas applied are described fully  in  Table 2-1.  The
required criticality calculations can be performed by hand without
difficulty, but the program is a convenience, particularly where one  is
interested in disaggregating losses according to whether they are suffered by
U.S. consumers, primary producers, secondary producers, or stockholders.  The
numbers displayed in the following listing of the  program  and its output are
for the "catalystium" example explained in the early part  of the chapter.

The program is written in the BASIC language for the Tektronix 4052 computer,
but should be readily adaptable to most other minicomputers.  The program
requires less than 8,000 bytes of core.
                                    2-71

-------
LIST
108 PEM—PROGRAM TO CALCULATE MATERIALS CRITICALITY (EXPECTED ANNUAL
118 PEM—ECONOMIC LOSSES FROM CONTINGENCIES)
128 REM—SAMPLE CALCULATIONS FOR IMAGINARY "CATALYSTIUM" MARKET
138 PEM—UERSION 1, AUGUST 15,  1981,  S. BEGGS
140 REM--PARAMETERS
158 REH—VALUE OF NORMAL U.S. CONSUMPTION FOR VEHICULAR EMISSION
160 REM--CONTROL <$ MILLION)
170 Ui=l
ISO REM—PRICE ELASTICITY OF U.S.  CONSUMPTION FOR VEHICULAR EMISSION
196 REM—CONTROL
260 Ei=8.1
210 REM—SUPPLY DISRUPTION SEVERITY (PROPORTIONAL INCREASE IN WORLD
228 REM—PRICE)
239 X=3
249 Z=X-1
258 REM—EXPECTED TIME BETWEEN (STARTS OF) DISRUPTIONS  (YEARS)
266 T=10
278 REM—EXPECTED DURATION OF DISRUPTIONS (YEARS)
288 D=l
290 REM—VALUE OF TOTAL NORMAL U.S. CONSUMPTION <* MILLION)
300 1*2=2
310 REM—PRICE ELASTICITY OF TOTAL U.S. CONSUMPTION
328 E2=0.2
338 REM—VALUE OF NORMAL U.S. PRIMARY PRODUCTION ($ MILLION)
340 V3=0.1
350 REM—PRICE ELASTICITY OF U.S.  PRIMARY PRODUCTION
360 E3=0.125
378 REM—VALUE OF NORMAL U.S. SECONDARY PRODUCTION ($ MILLION)
380 V4=6.3
390 REM--PRICE ELASTICITY OF U.S.  SECONDARY PRODUCTION
400 E4=6.125
410 REM—STOCKHOLDING COST, PER $1 MILLION OF MATERIAL  PER YEAR
420 REM—•<* MILLION, AT NORMAL PRICES)
430 C=6.065

-------
i
^j
CO
448
45Q
468
478
488
49Q
580
510
528
538
548
558
560
578
580
598
688
618
628
638
648
658
668
678
688
698
788
718
 28
            REM—CALCULATIONS FOR CONSUMERS PRODUCING UEHICULAR  EMISSION
            REM—CONTROL EQUIPMENT
            Il=Ul*El*Zt2/2/T*D
            I2=im-'T*D
            18=11+12
            REH—CALCULATIONS FOR THE UNITED STATES
            Jl=U2*E2*2t2/2--'T*D
            J2 = U2t.''T*D
            J8=J1+J2
            Ki=U3*E3*Zt2-''2/T*D
         738
         748
         750
         768
         770
         788
            K8=K1+K2
            Ll=U4*E4*ZT2/2/T*D
            L2=-U4* < Z+Zt2*E4 > /
            L8=LH-L2
            Hl=8
            H2=8
            H8=8
PAGE
PRINT "EXPECTED U.S.  ECONOMIC LOSSES FROM MARKET  CONTINGENCIES"
PRINT "         i* MILLION, AVERAGE PER YEAR;"
PRINT " "
P Ij T L I T II II
PRINT "ADDITIONAL COSTS FOR PRODUCERS OF UEHICULAR  EMISSION"
PRINT "CONTROL EQUIPMENT (NO STOCKHOLDING)"
PRINT "     TOTAL = "!10
PRINT "          ADJUSTMENT = "511
PRINT "          TRANSFER = "512
PRINT " "
PRINT " "
PRINT "U.S. LOSSES WITH NO STOCKPILE"
GOSUB 800

-------
no
i
        798
        800
        810
        820
        838
        848
    GO TO 1010
    REM—SUBROUTINE
               TO PRINT OUT LOSS DISAGGREGATIOH
        860
        870
969
918
920
930
948
950
968
978
988
950
1800
1818
1828
1830
1040
1050
I860
1870
1088
1090
1180
1110
1120
1136
PRINT "
PRINT "
PRINT "
PRINT "
PRINT "
PRINT "
PRINT "
PRINT "
PRINT "
PRINT "
PRINT "
PRINT "
PRINT "
PRINT !I
PRINT "
PRINT "
PRINT "
PRINT "
PRINT "
 RETURN
 STOP
 PAGE
 PRINT "U.S
 H2=-M2
 M2=0
                CONSUMERS"
                     SUBTOTAL  =  "JJO
                          ADJUSTMENT COST
                          TRANSFER COST =
                PR I MAR'
                                PRODUCER?"
                                        :.O
                                        II
                                       ST
                                     ST =
     SUBTOTAL = "5KO
          ADJUSTMENT CO
          TRANSFER CO
SECONDARY PRODUCERS"
     SUBTOTAL = "!LO
          ADJUSTMENT COST
          TRANSFER COST =
STOCKHOLDERS"
     SUBTOTAL = "5H0
          HOLDING CO
                                             3TS
                                       J2
                                                    K2
                                                     ";LI
                                                     L2
                                       HI
                        u.s,
                     ACQUISITION COST,  LESS REUENUES =
                                              H2
                TOTAL = "?M0
                ADJUSTMENT &
                TRANSFER (TO
                  STOCKHOLDING
                  FOREIGNERS) =
                                                                  Ml
                                                          M2
                         LOSS
               JES
        WITH IMPORT-ELIMINATING STOCKPILE"
             HO=H1+H2
             M1=J1+K1+L1+H1
                   880
GOSUB
STOP
PAGE
PRINT
                    "U.S. LOSSES WITH COMPREHENSIVE STOCKPILE"

-------
      OH3 00£T
098 anSQD 06cl
     Q=SW QlZ
    TH=TW 89c
    tH=0H RSc
     0=cH 0tc
     0=01 0cc
     0=0>l 06 1 1
     0=c!M 08TI
     0=1 X 0^H
          09U

-------
        EXPECTED U.S.  ECONOMIC LOSSES FROM MARKET CONTINGENCIES
                 U MILLION,  AVERAGE PER YEAR)


        ADDITIONAL COSTS FOR  PRODUCERS OF VEHICULAR EMISSION
        CONTROL EQUIPMENT (NO STOCKHOLDING)
             TOTAL = U.18
                  ADJUSTMENT  =0.02
                  TRANSFER =  0.16
        U.S.  LOSSES WITH NO STOCKPILE
             CONSUMERS
                  SUBTOTAL = 0.32
                       ADJUSTMENT COST = 0.08
                       TRANSFER COST = 0.24
             PRIMARY PRODUCERS
-lu                 SUBTOTAL = -0.0225
                       ADJUSTMENT COST = 0.0025
                       TRANSFER COST = -0.025
             SECONDARY PRODUCERS
                  SUBTOTAL = -0.0675
                       ADJUSTMENT COST = 0.0075
                       TRANSFER COST = -0.075
             STOCKHOLDERS-
                  SUBTOTAL = 0
                       HOLDING COSTS = 0
                       ACQUISITION COST,  LESS REVENUES = 0
             U.S.  TOTAL = 0.23
                  ADJUSTMENT & STOCKHOLDING COSTS = 0.09
                  TRANSFER (TO FOREIGNERS)  = 0.14

        STOP  IN LINE 1010 PRIOR TO LINE 1020

-------
U.S.  LOSSES WITH IMPORT-ELIMINATING STOCKPILE
     CONSUMERS
          SUBTOTAL = 0.32
               ADJUSTMENT COST = 8.88
               TRANSFER COST = 8.24
     PRIMARY PRODUCERS
          SUBTOTAL = -8.8225
               ADJUSTMENT COST = 0.8825
               TRANSFER COST = -8.825
     SECONDARY PRODUCERS
          SUBTOTAL = -8.0675
               ADJUSTMENT COST = 8.8075
               TRANSFER COST = -8.075
     STOCKHOLDERS
          SUBTOTAL = -0.0945
               HOLDING COSTS = 8.0455
               ACQUISITION COST, LESS REMENUES = -8.14
     U.S. TOTAL = 0.1355
          ADJUSTMENT fe STOCKHOLDING COSTS = 0.1355
          TRANSFER 
-------
U.S. LOSSES WITH COMPREHENSIUE STOCKPILE
     CONSUMERS
          SUBTOTAL = 0
               ADJUSTMENT COST = 0
               TRANSFER COST = 6
     PRIMARY PRODUCERS
          SUBTOTAL - 6
               ADJUSTMENT
               TRANSFER CO
     SECONDARY PRODUCERS
          SUBTOTAL = 0
ro

oo
                COS
   T
•T  =
it
= 0
3
               ADJUSTMENT COST = 0
               TRANSFER COST = 0
     STOCKHOLDERS
          SUBTOTAL = 0.104
               HOLDING COSTS = 0.104
                              3T
     U.S
     ACQUISITION
TOTAL - 0.104
ADJUSTMENT fe STOCKHOLDIHG
TRANSFER 
-------
                                                  Charles
                                                  River
                                                  Associates
               APPENDIX  2-C


EARLIER APPROACHES  TO MATERIALS CRITICALITY
                     2-79

-------
                                                            Charles
                                                            River
                                                            Associates
                             APPENDIX 2-C

                EARLIER APPROACHES TO MATERIALS CRITICALITY
 The following survey of earlier analysis of materials criticality is taken
 from Charles River Associates (1978).  A few more recent studies are
 discussed briefly in the Bibliographic Note.  The chapters preceding the
 following discussion (in its original report), discussed materials
 criticality in more general terms, including the application of more
 sophisticated economic models to the determination of materials criticality.
 However, the text of the present chapter is sufficient introduction to
 understand all important issues that are raised in this appendix.
      Earlier studies of materials criticality have in the main
had the  same concerns  which motivated the analysis we have  pre-
sented above.  In this chapter we discuss some of  the more  prom-
inent of  these earlier studies in light of the basic principles
and economic modeling  recommended in the  previous  two chapters.
Examining the strengths and weaknesses of these earlier efforts
from  such a perspective also allows  us to elaborate upon our
earlier discussions  and provide  further examples.

                          Import Dependence

      The  most prominent area of  policy analysis for which criti-
cality measurements  have been developed is undoubtedly that of
foreign supply disruptions.  This emphasis has been particularly
pronounced since the formation of OPEC and the subsequent Arab
oil embargo.
      The  typical first step toward investigating the criticality
of imported materials  is to rank  the  materials by  the  percentage
of U.S. import dependence which each  accounts for.   A  bar chart

                                  2-80

-------
                                                       CHARLES RIVER
                                                         ASSOCIATES i
                                                       INCORPORATED
of  such a ranking for the year  1972  is reproduced  below  as  Figure
4-1.l  U.S.  import dependence for platinum group metals,  chromium
and cobalt was nearly complete, whereas only a moderate  percentage
of  U.S. copper consumption had  to be imported.
     Although such a ranking of materials is useful,  it  obviously
has limitations as a measure of criticality.  The  most obvious
limitation of Figure 4-1 is the omission of information  on  the
absolute size of the market.  For example, although U.S.  import
dependence was slightly greater for  tantalum than  it  was  for
aluminum (ores and metal), the  absolute value of the  aluminum
ore imports  is so much greater  than  the absolute value of the
tantalum imports that there can be little doubt that  aluminum
is  the more  critical material.2
     Chapter 3 of the present study  is a good general guide to
the many further relevant considerations which need to be inclu-
ded in a measurement of criticality,  such as the probability of
a disruption occurring and the existence of substitutes.  It is
worth reiterating a point previously made about the use of pol-
icy models for measuring materials criticality.  If materials
markets to be compared are assumed to be identical in all re-
spects except the value of imports — the same normal inventory/
consumption  ratio, the same price elasticities, the same dis-
ruption severity, the same disruption probabilities and so on —
then the resulting criticality loss  measures are simply propor-
tional to import dependence in absolute dollar terms.

     JThis particular diagram is from National Commission on
Materials Policy, Material Needs and  the Environment  Today and
Tomorrow,  Final Report (Washington,   D.C.:  U.S. Government Print-
ing Office,  June 1973),  pp. 2-25.   Similar charts have been
produced by other investigators.
      On the other hand,  measuring import dependence  as a percent-
age of consumption does have the great advantage of indicating
whether there are domestic producers who might expand production
when foreign supplies are cut off.
                                  2-81

-------
                                                 Figure 4—1
                      PERCENTAGE  OF U.S. MINERAL REQUIREMENTS
                                      IMPORTED DURING 1972
                                                                                            CHARLES RIVER
                                                                                                ASSOCIATES
                                                                                           INCORPORATED
                             PERCENTAGE IMPORTED
                                                   MAJOR
                                             FOREIGN SOURCES
                           100%

                                           60%
                                                  2SX
                                                    i
                              0%
                               I
PLATINUM GROUP METALS
MICA »—;
CHROMIUM
STRONTIUM
COBALT
TANTALUM
ALUMINUM to—*—il
MANGANESE
FLUORINE
TITANIUM taM
                                    UK. USSR, SOUTH AFRICA. CANADA, JAPAN. NORWAY
                                    MMA. IRA2IL. MALAGASY
                                    USSR. SOUTH AFRICA. TURKEY
                                    MEXICO, VAIN
                                    ZAIRE, ICLGIUM. LUXEMBOURG. FINLAND, CANADA. MOMMA Y
                                    NIGERI A. CANADA, ZAIRE
                               l    JAMAICA, (URIHAM, CANADA. AUSTRALIA
                                    «RAZIL.GASO«, SOUTH AFRICA. ZAIRE
                               a    MEXICO. SPAIN. ITALY.SOUTH AFRICA
                               ^    AUSTRALIA
                               -'J    CANADA. lOUTX AFRICA
                                    MAl>rSlA. THAIUkNO. «OUVtA
                               H    MEXICO. JAPAN. PERU. UK. KOREA
                               a    CANADA. NORWAY
                               .^    ««A21L. NIGERIA. MAlAGAnr. THAI LAND
                               J    IOUTH AFRICA. MtXICO. UK. *OUVIA
t             m.l.l.l.1.1.1.' 11111' 11 WW]    CANADA. IWITZERLANO.UBR
I             Tftffi?*fPPfrf!rJJJJ'UMf*    CANADA
I             agiiiJAM^JMMMaM    CANADA. MEXICO
'               P,,. • ""• LU.IJJ.I. .IJU    CANADA. MEXICO. FCRU
I                IUJAIJJUAULIJUIUUUI    CANADA. PERU. MEXICO. HONOURAt. AUVTRAUA
I                tMHUWAfMHyorfm    PERU. IREUkNO. MEXICO. ORECCE
I                 BA».".I.I.I.'.IJ.I.UUW<    CANADA, MEXICO. JAMAICA
I                   KtfJ*fMf.VMm    CANADA, JAP AN. MEXICO. UK
I                   KJJJJWAMJAAM    PCRU, CANADA
f                    muuuuuuumaM    •OUTH AFRICA, CHIUE. UBR
I                    ifff^fffBoaaa    CENTRAL » SOUTH AMERICA  .CANADA.MIDDLE EACT
1                     IBPIIIflBflPflm    CANADA. VENEZUELA, JAPAN. COMMON MARKET KECI
I                      HJLU.UJJUUU    CANADA. AUTTRAUA.PERU. MEXICO
I                      •IIIIIM«    MEXICO, AUSTRALIA. itLCIUM. LUIKHO URC. CANADA. FCRU
I                         BWWVT    CANADA. PERU. CHILE
|                         JUUUWM    CANADA. AUCTRAUA
1                          FAMJUI    AUSTRALIA. MALAYSIA. INDIA
I1                          IUUW»    GREECE. ITALY
|                             m    CANADA,MEXICO. (AMAMAS
                                                                GREECE. IRELAND
                                                                CANADA
                                                                WEST GERMANY. FRANCE
                                                                CAMAOA. MEXICO. ITALY, PORTUGAL
                                           *0%    2S%      OK
 SOURCE:  National Commission on Materials Policy, Material Needs and the Environment Today and Tommorrow
           (Washington D.C.:  U.S. Government Printing Office, June 1973), p. 2—25.
                                                      2-02

-------
                                                       CHARLES RIVER
                                                         ASSOCIATES
                                                       INCORPORATED
          Measuring Criticality Without Economic Models

     Even if one's basic definition of materials criticality
 is much  less explicit and detailed than that  developed above  in
 Chapter  2,  it  is not particularly difficult to  identify many  of
 the  factors which determine  such criticality.   A number of  stud-
 ies  have compiled comprehensive lists of determinants on  a  sub-
 jective  basis, without constructing a formal  economic model.
 Typically,  cross tabulations of characteristics for various ma-
 terials  have been prepared, with summary index  values calculated
 for  each material by arbitrarily weighting the  characteristics.
     We will now review a few of the more important studies which
 have developed measures of materials criticality without  the  use
 of an economic model

 Study by King and Cameron
     One of the more detailed and formal efforts along these
 lines has been done by Alwyn King and John Cameron at the U.S.
 Army War College.1 The list of observable factors which they
 considered relevant to determining U.S. vulnerability to  for-
 eign supply disruptions is reproduced below as  Table 4-1.
     In order to compute a summary measure of criticality, King
 and Cameron first assigned arbitrary numerical  weights to large
 (L), medium (M) and small (S) effects on the materials' vulner-
 ability for the factors listed in Table 4-1.  Then the direction
 and magnitude of the factor's effect on the vulnerability of par-
 ticular materials were also characterized by arbitrary weights.
 The two weights were multiplied in each cell of Table 4-1 and then
 added,  first for "economic,  political and military" considerations
 separately,  and then for all three together.  Each material inves-
 tigated was treated in this fashion in turn.
     1Alwyn H.  King and John R. Cameron,  "Materials and the New
Dimensions of Conflict" (Carlisle Barracks,  Pa.:  U.S. Army War
College,  Strategic Studies Institute,  May 15,  1974);  Alwyn H.
King, "Materials Vulnerability of the United States — An Update"
 (Carlisle Barracks,  Pa.:   U.S. Army War College, Strategic Stu-
dies Institute, April 30,  1977).
                                2-83

-------
                                                                  CHARLES RIVER
                                                                    ASSOCIATES
                                                                 INCORPORATED
                                  Table 4-1

          FACTORS AFFECTING COMMODITY  RELATIVE VULNERABILITY  INDEX,
                       AS COMPILED BY KING AND CAMERON

                                                                       L = Large
                  Factor                     Effect on Vulnerability   M = Medium
                                             	S = Small

	Economic    Political  Military

Domestic reserves:
  Avai labiIity                                     L            L          L
  Cost of developing                               L            L          S

Domestic production  industry:
  Present capability                               L            L          L
  Cost of augmenting                               L            L          S

Substitute materials:
  Present availability                             L            L          L
  Cost of research to develop                      L            L          S
  Time required to develop                         ILL

Additional domestic  resources:
  Present availability                             ILL
  Cost to develop suitable processes               L           M          M
  Time to develop suitable processes               M           MM
  Probability of discovery if not available        M           M          S
  Cost of additional exploration                   M           M          S

Foreign suppIiers:
  Number of controlling companies                  L           S          M
  Number of supplier countries                     M           L          M
  Political  stability of suppIier countries        M           L          M
  Ideology of supplier countries                   L           L          L
  Productive capacity of supplier countries        L           L          L
  Economic sufficiency of supplier countries       L           L          S
  History of political relations with US           S           'M          S
  US dollar  involvement in supplier country        M           M          S
  Accessibility of supplier countries (supply
      routes)                                      SSL

US Stockp Me:
  Present US stockpile objective                   L           L          L
  Actual  quantity in US stockpile                  M           M          M
  Customary industry stockpile                     M           M          M


Table continued on following page.
                                       2-84

-------
                                                                  CHARLES RIVER
                                                                    ASSOCIATES
                                                                 INCORPORATED
                            Table 4-1 (Continued)
          FACTORS AFFECTING COMMODITY RELATIVE VULNERABILITY INDEX,
                       AS COMPILED BY KING AND CAMERON
                                                                       L = Large
                                             Effect on Vulnerability   M = Medium
                                             	S = Small
                                               Economic    Political  Military
Trend in usage of critical  material                M           M          S

Proportion of national  consumption directly
    related to military requirements               SSL

Importance of secondary sources (recycling)        M           M          M
SOURCE:  Alwyn H. King, Materials Vulnerability of the United States
         An Update (Carlisle Barracks, Pa.:  U.S. Army War College,
         Strategic Studies Institute, April 30,  1977), p. A-2.
                                        2-85

-------
                                                       CHARLES RIVER
                                                         ASSOCIATES
                                                       INCORPORATED
     King and Cameron have only reported results  for a truncated
version of the above methodology based on a consideration  of  the
following five factors:  availability of domestic reserves; avail-
ability of substitutes; number of foreign suppliers;  ideology  of
foreign suppliers; and U.S. stockpile objective.  Also,  they  con-
sidered only eleven of  the materials most likely to have  high
criticality ratings.  Results, in rank order with their  "vulnera-
bility index ratings," were chromium  (34), platinum group  (32),
tungsten  (27), manganese  (23), aluminum  (22),  titanium  (20),  co-
balt  (20), tantalum (16), nickel  (14), mercury (11) and  tin  (6).1
These results are intuitively plausible.  However, applying the
methodology to many more materials using all 27 factors  in Table
4-1 could easily lead to questionable results  since there  is  no
overall economic theory underlying the analysis.

Relative Inclusiveness of the CRA Policy Model
     It is interesting to compare King and Cameron's list  of  de-
termining factors and subjective approach to materials criticality
with the use of the CRA economic policy model  described  in Chap-
ter 3.  Both approaches are concerned primarily with foreign  sup-
ply disruptions.  Use of an economic model makes much heavier
demands on the analyst when new factors are to be formally in-
cluded in the methodology.  However, the economic modeling ap-
proach described in Chapter 3 in fact includes explicitly  or  im-
plicitly virtually all of the observable factors listed  in Table
4-1.  Furthermore, application of the economic model greatly
enhances the value of this basic information by utilizing  as
accurate a quantitative specification as possible, and by  trans-
lating this information into dollar loss figures, broken down by
who gains and who loses.  Thus,  groups with differing interests
and values have information in the most useful possible  form  for

     :King and Cameron,  op.  ait.,  p. 17.
                                 2-86

-------
                                                       CHARLES RIVER
                                                         ASSOCIATES |
                                                      INCORPORATED
ordering their own priorities and for participating in the
process of determining a national consensus on priorities and
policies.  The complexity and scope of the materials criti-
cality problem is reduced in the most meaningful possible way
by applications of an economic model.
     In the remainder of this section we will-consider-the-  -
factors listed in Table 4-1 and evaluate how adequately they
can be incorporated in the economic model described in Chap-
ter 3.  In .some-cases, a generalization of the model in Chap-
ter 3 would allow somewhat fuller recognition of relevant cir-
cumstances (e.g., the depletion of reserves), but in general
the available policy models can be very comprehensive  if care-
ful analyses precede specification of their parameters.

     Domestic Reserves
     U.S. reserves of the commodity under study should be
important determinants of the domestic supply curve in the
CRA. policy model; however, it is not possible to infer directly
the flow of domestic production from the stock of reserves.
If reserve estimates were accurate and inclusive, it might
be desirable to allow explicitly for backward shifts in the
U.S. supply curve as the domestic resource base is depleted.
However, reserve estimates only reflect the state of knowledge
at a particular point in time, and historically depletion has
been more than offset by discovery and technological change
in most cases.  Thus, the fact that the existing CRA policy
model ignores depletion is usually not a serious omission,
though it may be so in cases such as petroleum where the size
of domestic reserves is relatively well known.
                                2-87

-------
                                                      CHARLES RIVER
                                                         ASSOCIATES
                                                      INCORPORATED!
     Domestic Production Industry
     The present capability of the domestic production  industry
is indicated by the position of the short-run domestic  supply
curve.  The cost of (gradually) augmenting domestic production
capacity is indicated by the position of the long-run domestic
supply curve relative to the short-run curve.
     The intersection of the short-run and long-run domestic
supply curve occurs at a single rate of production which can
be referred to as the "capacity" of the domestic industry,
though capacity in this sense represents an optimal adaption
of industrial capital to the existing rate of production, not
an upper bound on production.

     Substitute Materials
     The availability of substitutes under current technology
is usually the primary determinant of the short-run and long-
run price elasticities of demand in the CRA policy model.  In
many cases important "substitutes" for a material whose criti-
cality is being investigated will be factors of production
other than alternative materials.  For example, labor may be
substituted for a material such as manganese by increasing
the attentiveness of workers controlling its consumption in
steelmaking.1
          Substitutions may also be made by consumers of final
goods and services; when the price of a potentially critical
material rises, consumers may decide to decrease their pur-
chases of goods requiring its use.  All these considerations

      When the price of a material input shifts upward, an
economist considers any factor of production which is conse-
quently used in greater amounts to be a "substitute" for the
material input.  The engineer's definition of a "substitute"
is much less general.
                                 2-88

-------
                                                       CHARLES RIVER
                                                         ASSOCIATES
                                                      INCORPORATED
affect the "derived demand curve" for a raw material which
enters the policy model.  The demand curve for  the raw material
is characterized as "derived" because it can be determined
from the demand curve for the final good and information  about
the substitutions which producers can make.
     In the standard market model underlying the CRA policy
model, changes in U.S. technology, such as development of new
substitute materials, are theoretically represented as shifts
in the demand curves.  In practice, movements along a demand
curve are often interpreted to include some minor predictable
technological innovations and some efficiencies due to "learn-
ing by doing."  The dynamic implications of such processes are
captured in their relevant form in the CRA node! by the short-
run and long-run supply and demand curves, and by solving for the
intersections of short-run and long-run curves which maximize
firms' profits over time.

     Domestic Resources
     U.S. "resources," as opposed to "reserves," would become
economical to exploit only at prices higher than normal.  Thus,
the shape of the domestic long-run supply curve is strongly
affected by the existence of resources.  However, unless  a
market disruption (such as the formation of OPEC) is expected
to endure, it may not be efficient to exploit such resources
because of the long lags between substantial investments  and
actual domestic production.  If exploitation requires signi-
ficant research and development, the same distinctions between
dynamic adjustments in short-run curves and shifts of under-
lying long-run curves must be made on the supply side of  the
domestic market as were discussed above for the demand side
of the market.
                                 2-89

-------
                                                      CHARLES RIVER
                                                        ASSOCIATES
                                                      INCORPORATED
     Foreign Suppliers
     The number and nature of foreign suppliers of an imported
material determine the probability of supply disruptions of
various lengths and severities.  There is in practice no reli-
able relationship between readily discerned characteristics
of supplier countries and the probability of disruptions.  The
characteristics listed in Table 4-1 are suggestive in particular
cases, but expert assessment of the overall disruption proba-
bility is undoubtedly much more reliable than subjective weight-
ing of the tabled characteristics.1
     Plausible disruption scenarios involving various producer
countries can often be translated into effects on the world
market price using econometric market models prior to appli-
cation of a policy model; the effect of stock releases on the
world market price during the disruption should also be esti-
mated.  The particular approach chosen depends on the market
structure both in normal times and during the disruption.  For
example, the structure of the chromite market was fairly com-
petitive in the 1950s and early 1960s, but some collusion among
producing countries has occurred during the 1970s.  A sharp
cut-off of supplies from southern Africa would today encourage
much greater collusion than that which has occurred since 1974.
Thus, predicting world prices during such a disruption involved
both constructing a cartel pricing model utilizing historical
behavior patterns for consumers and suppliers who were assumed

     !0ne characteristic not explicitly mentioned in Table 4-1
involves distinguishing sources of supply as to whether or
not they are controlled by less developed countries.  Some
investigators have made this distinction an important element
in their analysis of materials criticality, but, as with the
other characteristics, it is only part of the broader situ-
ation which determines the probability of disruptions of
various severities and durations.
                               2-90

-------
                                                       CHARLES RIVER
                                                         ASSOCIATES
                                                      INCORPORATED
to remain competitive, and utilizing hypothetical  beh'avior
patterns leased on profit maximization  for  the  suppliers  who
were assumed to collude.1
     The situation described  above  for  the  chromite  market
in which there is a competitive market  structure in  normal
times and monopolistic pricing during the supply disruption
is typical.  However, in analyzing  a cutoff of  cobalt supplies
from Zaire due to the recent  military conflict  in  Shaba Prov-
ince, it was appropriate to reverse this pattern.   Zaire  nor-
mally exercises considerable  monopolistic power as the price
setter in the cobalt market.  However,  during a complete  dis-
ruption of Zaire the "fringe" of  competitive suppliers would
become the only sources of supply and the market structure
would hence become fully competitive.   Determining a competi-
tive price which clears the disrupted market by utilizing
historical behavior patterns  captured by the CRA econometric
model then became appropriate.2
     Before applying the policy model discussed in Chapter 3,
it is appropriate to utilize  formal market  models  in the  man-
ner illustrated above to determine  the  severity of plausible
disruption scenarios in terms of  world  price increases.   This
formal modeling approach contrasts  with the approach recommended
earlier for determining the probability of  a disruption begin-
ning and continuing, when this probability  must generally rely
on a subjective evaluation of a multitude of relevant factors,

    :See Charles River Associates Incorporated, The  Report of
the U.S.  Department of the Interior on  the  Critical  Materials
Aluminum,  Chromium,  T?latinum and  'Palladium:  A  Review  and Revision
(Cambridge, Mass.:  CRA, July 1977), Chapter  6.
    2See Charles River Associates Incorporated, Implications  of
the War in Zaire for the Cobalt Market  (Cambridge, Mass.:  CRA,
June 1977).
                                2-91

-------
                                                      CHARLES RIVER
                                                         ASSOCIATES
                                                      INCORPORATED
like those characteristics of foreign suppliers which are
listed in Table 4-1.  Expert opinions on probabilities are
likely to differ significantly.  However, the probabilities
are readily interpreted intuitively and discussion of whether
the chance of a disruption is one chance in four or one chance
in forty over the next decade is much more incisive than an
arbitrary weighting of a mixed bag of relevant considerations.

     U.S. Stockpile
     The size of industry and government stocks which are
available during an economic disruption directly enters the
determination of materials criticality when the CRA policy
model is applied.  Moreover, the specified size of stocks
held by other consuming nations can be of crucial importance,
though it is not listed in Table 4-1.  In general, the pre-
paredness of foreign consumers determines how aggressively
they bid for remaining world production during a disruption,
and hence affects the price at which imports can be obtained
by the United States.
     One of the significant advantages of applying a formal
economic model to the determination of materials criticality
is simply the identification of relevant considerations which
may be overlooked if the structure of the problem is delineated
more informally.  Remembering to include the stocks of foreign
consumers is a case in point.
     In general, there is a vast economic literature on such
problems as estimating supply and demand curves which can be
drawn upon to improve criticality measurements.  For example,
the smaller the proportion of the cost of final goods which
is accounted for by a potentially critical raw material, the
less price elastic its demand will tend to be and hence the
                                2-92

-------
                                                       CHARLES RIVER <
                                                         ASSOCIATES I
                                                      INCORPORATED I
greater its criticality; formulas precisely  specifying  such
relationships are often very useful for studies  such  as those
we are recommending for determining materials  criticality.

     Trend in Usage
     As importation of a material increases, the expected
dollar loss which measures its criticality will  generally
increase as well.  If the domestic quantities  demanded,
supplied, and stocked at the normal price increase propor-
tionately, and if the price elasticities and other parameters
remain constant, then expected dollar losses calculated with
the CRA policy model will also increase proportionately.  In
general, of course, trends in various categories of end users
can affect the price elasticity of demand for  all categories
combined, which appropriately summarizes the information
required by the policy model; criticality loss measures would
then not increase precisely proportionately with the  quantity
of total consumption.
          At any time, the fact that there has recently been
an unexpected change in the trend of usage can make an  impor-
tant difference in criticality ratings.  A sharp increase in
usage can leave world producers above the capacity point (where
their short-run supply curve intersects their  long-run  supply
curve), thereby limiting their ability to respond if  other
segments of the market are disrupted.  On the  other hand, an
unexpected trend decrease in consumption can result in  world-
wide excess capacity and considerably lower estimates for
criticality loss measures.  As circumstances change,  naturally
criticality estimates should be periodically updated.
                                2-93

-------
                                                       CHARLES RIVER
                                                         ASSOCIATES
                                                      INCORPORATED!
     Military Consumption
     The U.S. military does not  consume  large  amounts  of raw
materials directly.  However,  its purchases  of fabricated
items do indirectly require sizable  quantities of  raw  materials
If the military demand is  to be  considered a fixed requirement,
e.g., because there is no  suitable substitute  for  a critical
application, then all of the elasticity  in the U.S.  demand
curve will be due to decreases in civilian demand.   Such a
demand curve can be utilized in  the  usual fashion  when the
CRA policy model is applied to analyzing foreign supply dis-
ruptions.  If a foreign supply disruption of concern is in
fact a major military conflict involving the United States,
then further issues are involved, as discussed elsewhere in
this study.

     Secondary Recovery
     Secondary recovery decreases the amount of U.S. primary
demand at each price, which is the appropriate input for the
CRA policy model.   In fact there are further ramifications
to secondary recovery which can be of great  importance in a
few cases where rates of secondary recovery  are very high and
the lags between consumption and recovery are  short.   The
prime example is platinum, which is  largely  used as  a  catalyst
in petroleum refining and in chemical production.:   In this
case the amount of platinum in use is essentially  a  sizable
stock which can be reallocated during disruptions.

     :The recovery rate for platinum used in catalytic con-
verters for automobiles and in electrical equipment  is also
very high,  but recovery lags are longer  for  this end use  than
for the other uses.
                               2-94

-------
                                                        CHARLES RIVER
                                                          ASSOCIATES
                                                        INCORPORATED
      A special policy model  has  been designed at CRA for
 application to the platinum  and  palladium markets, where the
 most important precaution  in the face of threatened foreign
 supply disruptions need be only  some extra metal held "in use."1
 This platinum or palladium held  "in  use" is "extra" in the
 sense that using more of the metal as catalysts has benefits;
 however, these benefits would normally not justify the cost
 of purchasing the metal unless there was also a significant
 probability that the market  price would increase after pur-
 chase (due to a foreign supply disruption in the present
 instance).

 A. BOM Tabulation
      The study by King and Cameron is similar in format to
 a number of other cross tabulations  of materials' character-
 istics relevant for measuring criticality.  Not all such
 tabulations attempt to push  the  analysis so far toward subjec-
 tive quantification of characteristics and computation of
 summary indices as do King and Cameron.   One of the more graphic
 and authoritative of such  cross  comparisons has been constructed
 by John Morgan of the U.S. Bureau of Mines.  It is reproduced
 below as Figure 4-2.2

     :See Charles River Associates Incorporated, The Report of
 the U.S. Department of the Interior,  op. cit.,  Chapters 9
 and 10.
      2See John D. Morgan,  "Mineral Data Improvements and Critical
 Materials R&D at the U.S.  Bureau of  Mines," Proceedings of the Work-
 shop on Government Foliates and Programs Affecting Materials Availability
 (Columbus, Ohio:  Metals and  Ceramics Information Center, Feb-
ruary,  1976), pp. 319-344.  This  same paper was  presented before
the National Symposium on Ceramics in the Service of Man in June
1976, and was reproduced as a Bureau  of Mines Publication
entitled National Considerations  of Strategic and Critical Materials.
                                 2-95

-------
                 Figure 4—2

BOM CROSS TABULATION OF MINERAL PROBLEMS
                                              CHARLES RIVER
                                                ASSOCIATES
                                             INCORPORATED
MINERAL PROBLEMS
1975 AND BEYOND
PftUIMINAIT DATA, AUOUST, 1975
BUREAU Of MINIS
AAAjXXX Ma|0« P'oblom Iran lha national
JOJSSXS? viewpoint
'-•••i^af* Moderate problem Iroia tka national
•:.•:".'.." Minor or more lacaliiad problem
_ Large itockpile eicenei prevent
' a currant probleia
i. U.i exparff contribute lubitanlialry
to our balance ol trade
ABRASIVES, NATURAL
ABRASIVES, 1ANUFACTURES,
ALUMINUM (Incl. BAUXITE t, ALUMIHA)
ANTIMONY
ARGON
ARSENIC (byproduct of Copper)
•»
>.
u
T
i
«
u
M
V
W
T9
w
3
a




"S = 0 >
u 3 > C o
• « g • o e
3 * 5 = J ^
A U 44 C
w 3 o ••* • •
i X « fi U * J — i
s -s s "• , i .
- S x ? 3
X *l U «* W*J J
STJ c '** — * u u g
« > « ^ • i a
3 ^ -• 3 • O °- ^
w^us »- ^3
« V « 6 »» 1- C J -- —
C C -O -- V 2
"*-xi- MO« c e
SS1^"0 • •=" 5 «• | |
h 4 U 4) « S CSC
3 3 *t a* ^ B X -- O O
OVO.C M-.UU
U « u u w
5 e 5 S 2 55
3 <• —
u 4 41 w
r i e s
s i s i
— o « t*
.s — u a
b e
!«
1 .

. -: ,
c c 2
1 J
=. =


S3 ^88SSS888i
%' -^ 'f.'i
L

ASBESTOS, CHRYSOTILE 8££8
SARITE
BERYLLIUM
BISMUTH (byproduct of Le*d)
BO RON
BROMINE
CADMIUM (byproduct of Zinc)
CALCIUM CHLORIDE
CEMENT
CES IUM
CHLORINE1, MANUFACTURED
CHROMIUM (tncl. FERSOCHROMIUM)
CLAYS
COAL
COBALT (byproduct of Cooper & Nickel)
COLUMBIUM (NI08IUM)
COPPER
CORUNDUM S EMERY
DIAMONDS, GEM STONES
DIAMONDS, INDUSTRIAL, STONES, NATUHAL
DIAMONDS, INDUSTRIAL, SORT, NATURAL
DIAMONDS, INDUSTRIAL, SORT, SYNTHETIC
DLATCMITE
FELDSPAR
FLUORSPAR
GALLIUM 'byproduct of Aluminum & Zinc)
GARNET
i
W«



oc^oddddf
?WVVW9




88 -S






35-'' SSSSS!

^SSSS^S^S:



SS
W" ^^^
SS «^5
r •



c -•»







^

w

j^J;


\'-^
rf '•<



•';*
'


O3 (A Wl
1 1 « 1 n

?• :
-
"'•"." r

'-•!. KS ^X
I *x"2v5
"35"-' 1VAVS
> :




?SS4««






N<

KXJ •'
sssss^ssss
Csx '-—•"-

r


•
*i» ^ 'i1*


H c

d

«
!SS

ss








t-' •

?5




r- i- S
'j&
y/i
j&&


SS


. . ?,
SS8S88S

•-.•"


Ft~ • -• /




--.

~-^
•V-.









-a-
WJ
.- ;


',•£ ^^' "




f*-*

*'^- ^ '^'"








•»
'" -•-•;•:
i- T-' ."



• ' -"V ^ i


t'Z

h| t

J k
, -.-«
v**
1 | « | n
VI
3
e
0
3
o


^










•$




W

























=«




O
                     2-96

-------
                                           Figure 4—2 (cent.)

                    BOM CROSS TABULATION OF MINERAL PROBLEMS
                                 CHARLES RIVER
                                     ASSOCIATES
                                INCORPORATED!
                 MINERAL  PROBLEMS
                    1975  AND BEYOND

                 PIUIMINARY DATA, AUGUST. 1975
                       • U«EAU OF MINES

                f   Moior problems ttom  sWe notiono!
                >    viewpoint
                  Moderate problems Iron the oolio
                    viewpoint

                  Minor or more  locolned problems
               S  Lara* stockpile eacesiet  prevent
                    o current problem

               E  US etporli contribute substantially
                    to our balance of trade
         MAGNESIUM
         HANGASESE (incl
         MERCURY
         MICA, SCRAP i FLAKE
         MICA, SHEET
         MOLYBDENUM
         NATURAL GAS
         KICKEL
         NITROGEN, ELEMENTAL










•

1





•
J


&
-5
> c
s ^

*s • U
-• 0-
rj -O


"O /
a 4,
^


w «
3 r
*j W
* *• V
4« ^
U PI
\ Z ~
i : i
3f 1-
- 0 i
; — u

n w tn r.
?
c
!




3



C
—
*
•a
W



j
5
— •*• r

r: = =





















i
3
£
V.

in
£ I
« 4.
2 '
O C
2 E
a. >

s s i
o *
s- ; »
1 » ^
00 —
>s •- U
U t X
w C U
3 b
1 S „
•- * • *J
U -* C
c ac «
— -. c -






i
•













1 C
=
c



















o g
n v
 b | c | d K. I  i | j
         K1TROGEN, FIXED
         OIL SHALE
         OWGEd
         PEA1
         PERLITE
         PETROLEUM
         PHOSPHATES
         PLATINUM GROUP  PLATINUM, PALLADIUM, OSMIUM,
                       RHODIUM, RUTHENIUM, 1R1DIUM
                                                              JS
         POTASH
         PUMICE 4 VOLCAS1C CINDER
         QUARTZ CRYSTALS  (LASCA for »ynthetlc)
         RADIUM
         (AXE EARTHS
         HHEK1W1  (byproduct of Copptr-Molybdtnun
         RUBIDIUM (byproduct of Lithium 4 Ce«tum)
                GRAVEL
                     _
          SCANDIUM  (byproduct of Uranium & Pho«ph«te)
         SELENIUM  (byproduct of Copper)
          SILICON (METAL i FERRDSILICON)
          SILVER
          SLAG - IRON & STEEL
          SODIUM CARBONATE & SULFATE
          STAUSOLITE (byproduct of Tlt»nium Biaerall)
          STONE
          STRONTIUM
          SULFUR
          TALC
          TJUflALUM
          TELLURIUM (byproduct of Copper)
          THALLIUM (byproduct of Zinc)
          THORIUM
          TIN
          T1TAK1UM Itocl. tLKENITS 4 RBT1LE)
          TUNGSTEN
          URANIUM
          VANADIUM
          VMMICULITE
          UOLLASTONITE
          TTTRIUM (byproduct of tort  Earth«)
          ZEOLITES
          ZINC
          ZIRCONIUM (incl.  ZIRCON!
                                                          «5
p     SSS
                                                           $$•
             J8S

                                                                                        to  w
                                                                                         m J
                                                                                        56666aSi
                                                                                    k   1  ~  n  o
SOURCE:  Dr. John D. Morgan, "Mineral Data Improvements and Critical Materials R & D at U.S. Bureau of Mines'
           Proceedings of the Workshop on Government Policies and Programs Affecting Material Availability
           (Columbus  Ohio: Metals and Ceramics Information Center, February 1976), pp. 342-343.
                                                   2-97    '

-------
                                                      CHARLES RIVER
                                                         ASSOCIATES
                                                      INCORPORATED
     Many of the problem areas identified in Figure  4-2 have
already been discussed in terms of our theoretical framework
and the application of policy models to determination of
materials criticality.  However, other of these problem areas
deserve a brief comment.
     A. mineral is apparently considered a "U.S. foreign ex-
change drain"  (problem area f) if the net imports of the United
States have a high value.  If there is an opportunity for
this country to develop domestic resources of a particular
mineral, say through investment in developing new recovery
technologies, then such a mineral does present an opportunity
for beneficial government R&D programs.  In cases such as
petroleum, large expenditures on imports may be partially the
result of collusion among foreign suppliers.  However, in
general international trade greatly benefits the United States,
and the theory of comparative advantage explains why it is good
that we import some goods and services in exchange for those
we export.  Thus, large U.S. expenditures for imports are not
a sufficient condition for criticality; such expenditures may
or may not be a symptom of some problem or indicate  the need
for revised government policy.
     Health, safety, and environmental problems (areas h, 1,
m, and n)  are included in Figure 4-2, in contrast to studies
restricted to criticality stemming from actual or potential
foreign supply disruptions.   As we noted in the general the-
oretical discussion of Chapter 2, a restricted viewpoint is
quite workable as long as the national policies designed to
treat each problem area are relatively distinct.  For example,
stockpiles counteract the threat of foreign supply disruptions
but have little direct impact on health, safety, and environ-
mental problems.  The reverse is true of pollutant taxes or
job safety regulations.
                               2-98

-------
                                                       CHARLES RIVER
                                                         ASSOCIATES
                                                      INCORPORATED
     Another set of problem areas identified  in Figure  4-2
relates to a fairly distinct set of national  policies,  and
thus can often be usefully treated apart from criticality
due to foreign supply disruptions.  Manpower, energy, and
transport  (problem areas i, j, and o) are inputs  into "the U.S.
minerals extraction and processing industries which  are  par-
ticularly prone to problems requiring national attention.
Examples of relevant policies would be  forced arbitration of
strikes or regulation of the price of natural gas.   Materials
which are considered a "load on the U.S. transport system"
are apparently simply voluminous and heavy, which may or may
not indicate criticality.
     "Inadequate recycling" (problem area h)  can  be  an  impor-
tant element in the analysis of many policy problems, from
foreign supply disruptions to occupational health regulations,
However, it can be an area of concern simply  on the  grounds
of economic efficiency.  Depletion allowances in  the tax system,
discriminatory freight rates,  and other institutional char-
acteristics of the U.S. economy appear  to bias consumption
toward primary sources of materials and away  from secondary
scrap.   Economic losses due to these inefficiencies  could in
principle be included in a materials criticality  rating  if
the perspective were to be broader than a concern only with
foreign supply disruptions.
     It is clear that cross tabulations of market character-
istics like that in Figure 4-2 are a useful way of organizing
information.   When the CRA policy model is applied to materials
criticality,  cross tabulations of the inputs  into the model
should be constructed as well as cross  tabulations of outputs.
Such an input tabulation would be in many ways a  particularly
incisive substitute for the type of presentation  we  have been
discussing.
                                2-99

-------
                                                        CHARLES RIVER j
                                                          ASSOCIATES !
                                                       INCORPORATED!
        Measuring Criticality  With Input-Output Models

     There have been  a  number  of studies of the effects of
shortages on the U.S. economy  based on input-output models.
Such models explicitly  consider all sectors of the economy
simultaneously; this  inclusiveness is an advantage when one
wishes to consider  simultaneous shortages of most materials, as
would occur for example during a war.  The Federal Preparedness
Agency routinely uses such models in planning stockpile objec-
tives.  Unfortunately,  adding  to an input-output model all the
features which the  CRA  policy  model incorporates for measuring
materials criticality would be an extremely complex project
which is not likely to  be  undertaken soon.  In any case, a
policy model based  on a model  of a single market implicitly
recognizes the interrelationships among markets to a greater
extent than may be  immediately apparent.
     A recent study done at the Stanford Research Institute is
representative of the results  which an input-output model
generates when shortages are analyzed.1  Moreover, the study
explicitly addressed  the issue of determining criticality of
materials in terms  of prescreening those which were worth
careful consideration using the formal model.   We will use this
report as the basis for most of our discussion of the applica-
tion of input-output  models to materials criticality.

      lSee the  following publications:  Mark D. Levine and  Irving
W.  Yabroff,  Department of Defense Materials Consumption and the Impact of
Material and Energy Resource Shortages , prepared for the Defense
Advanced Research Projects Agency by the Stanford Research
Institute, November 1975;  Evan E. Hughes, et al. ,  Strategic Resources
and National Security:  An Initial Assessment, prepared  for the Defense
Advanced Research Projects Agency by the Stanford Research
Institute, April 1975;  Mark D. Levine and Irving W. Yabroff,
"Strategic Resources  and National Security," Paper I in  Proceed-
ings of the Department  of Defense Materials Shortage Workshop,  Metals and
Ceramics Information  Center, January 1975.
                              2-100

-------
                                                       CHARLES RIVER
                                                         ASSOCIATES
                                                      INCORPORATED
Prescreening of Materials

   The prescreening of materials for the SRI study was done

on a formal basis which is similar to other subjective weight-
ing schemes we have already discussed.l   Seven criteria were
used to rate commodities on a scale of 1 to 10:

     (1)  Percent of U.S. consumption for defense purposes.2

     (2)  U.S. reserves.
     (3)  Percent of U.S. consumption imported.

     (4)  Vulnerability of sources.
     (5)  Difficulty of substitution.
     (6)  Value of consumption  ("economic importance").
     (7)  Leverage of industry  (high value of final good

          output per dollar of raw material input).3
     Jesuits for 74 potentially critical raw materials are

 reproduced in Table 4-2.  A rating of 10 for a particular

 criterion is the maximum contribution to criticality which

 is allowed; a rating of zero indicates a lack of data.  A


     :The methodology is described below only in enough detail
 that the summary table of results can be roughly interpreted.
 For a complete discussion, see Hughes and others, pp. 191-212
 Data utilized for criteria  (1) , (2), (3) and  (6) were
 explicitly for the year 1972.  Criteria (4),  (5) and  (7) were
 evaluated subjectively.
     2The first criterion reflects the fact that the study was
 done for the Department of Defense; in terms introduced in
 Chapter 2, the perspective of this study is somewhat less
 general than "national concensus," though introducing
 specialized interests in such an ad hoc fashion is open to
 criticism.
     3The "leverage" of a raw material is directly taken into
 account in the CRA policy model when the J .S. price elasti-
 city of demand for the raw material is derived from the price
 elasticity of demand for finished goods.
                              2-101

-------
                                          PRIORITIES OF MATERIALS:

                                      (1)         (2)          (3)
o
ro
                 Name of Material
DoD Use
L61*



C73»

[5]»
m*
Cio]»

ra»

DP


[8]"

M*







I.
2.
3.
4.
5.
6.
7.
8.
9.
10.
II.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Aluminum
Iron
Manganese
Graphite
Copper
Yttrium
Chromium
Platl num-Group
Tungsten
Mica-Sheet
Nickel
Antimony
Cobalt
Fluorine
Mercury
Silver
Tantalum
Tin
Lithium
Asbestos
Co 1 umb 1 urn
Cesium
Bismuth
Potassium
Cadmium
5
8
7
5
10
4
7
5
8
10
7
8
10
6
7
5
6
8
7
3
6
0
4
1
9
  U.S.
Reserves

   10
    7
   10
   10
    5

    9
   10
   10
    9
   10

   10
   10
    4
   10
    9

    9
   10
   10
    5
    9

    6
    0
    8
    7
    6
     Table 4-2
RANKING ON NATIONAL SECURITY AND ECONOMIC CRITERIA
       (4)               (5)            (6)
                      Difficulty
                         of
                     Substitution
                                                                                  (7)
                                                               10
                                                                2
                                                                8
                                                                9
                                                                2

                                                                4
                                                                7
                                                               10
                                                                5
                                                               10

                                                                7
                                                                5
                                                                7
                                                                4
                                                                8
                                                                9
                                                                0
                                                                9

                                                                7
                                                                8
                                                                5
                                                                6
                                                                4
Vulnerability
  of Sources
                                      3
                                      3
                                      6
                                      6
                                      4

                                      0
                                      7
                                      6
                                      7
                                      7

                                      4
                                      6
                                      6
                                      5
                                      7

                                      6
                                      6
                                      4
                                      3
                                      5

                                      6
                                      5
                                      3
                                      4
                                      3
                         6
                         8
                         7
                         5
                         6

                         0
                         5
                         8
                         7
                         4

                         4
                         7
                         7
                         6
                         7

                         8
                         6
                         3
                         6
                         5

                         4
                         5
                         7

Economic
Importance
6
to
1
0
6
0
1
1
1
1
2
1
1
1
1
1
1
1
0
1
1
0
1
1
1

Industrial
Leverage
5
9
8
2
7
0
5
3
4
2
4
3
4
3
2
4
2
3
2
2
2
1
3
6
3
Geometric
Mean
Columns 2-8
6.0
5.9
5.7
5.5
5.2
5.2
5.1
4.9
4.9
4.8
4.8
4.7
4.7
4.6
4.5
4.5
4.5
4.3
4.2
3.8
3.8
3.8
3.7
3.6
3.6
         Table continued on following page.
                                                                                                                                 CHARLES RIVER
                                                                                                                                    ASSOCIATES
                                                                                                                                INCORPORATED

-------
                                                                 Table 4-2 (Continued)
                                      PRIORITIES OF MATERIALS:  RANKING ON NATIONAL SECURITY AND ECONOMIC CRITERIA
                                    (1)
       DID*
ro
O
GO
       [I]"
Name of Material DoD Use
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
40.
49.
50.
Tha 1 1 1 urn
Indium
Zinc
Beryl 1 lum
Thorium
Gold
Lead
Vanadium
German 1 urn
Sul fur
Selenl urn
Arsenic
Hafnium
Strontium
Barium
Rubidium
Z 1 rcon 1 urn
Corundum
Gypsum
Tl tan lum
Chlorine
Sand and Gravel
Iodine
Stone -Crush
Te 1 1 u r 1 urn
10
0
8
10
10
4
8
5
10
5
6
1
0
5
6
0
5
4
1
4
4
2
3
3
3
  (2)

  U.S.
Reserves
  I
  7
  6
  I
  I

  9
  2
  9
  6
  8

  I
  3
  I
  I
  I

 10
  I
 10
  6
  I

  I
  I
  4
  I
  I
 (3)
(4)
(5)
(6)
(7)
                                                                  Vulnerability
                                                          Imports     of Sources
10
 0
 5
 7
 2
 2
 2
 I

 4
 0
10
 9
 4

 I
 0
 0
 4
 5
                                                           10
                                                            I
                                                            4
 0
 4
 3
 6
 7

 2
 4
 3
 5
 I

 3
 6
 3
 4
 3

 5
 3
 0
 2
 3
Difficulty
of
Substitution
5
5
2
6
4
5
5
4
3
7
7
4
5
4
4
5
6
2
3
4
7
6
3
6
6

Economic
Importance
0
0
1
1
0
1
1
1
1
1
1
0
0
1
1
1
0
1
1
1
2
3
1
3
1

Industrial
Leverage
1
1
4
3
1
1
4
2
1
5
2
2
1
1
3
1
1
1
3
2
8
7
1
7
1
Geometric
Means
Columns 2-8
3.5
3.4
3.4
3.3
3.2
3.1
3.1
3.0
2.9
2.8
2.7
2.7
2.7
2.6
2.6
2.5
2.5
2.4
2.4
2.4
2.4
2.4
2.3
2.3
2.2
       Table continued on following page.
                                                                                                                              CHARLES RIVER
                                                                                                                                 ASSOCIATES
                                                                                                                             INCORPORATED

-------
                                                   Table 4-2 (Continued)
                          PRIORITIES  OF MATERIALS:  RANKING ON NATIONAL SECURITY AND ECONOMIC CRITERIA
Name of Material
51.  RhenI urn
52.  Boron
53.  Molybdenum
54.  Nitrogen
55.  Phosphorus

56.  Clays
57.  Silicon
58.  Garnet
59.  Rare Earths
60.  Talc
                      (1)
DoD Use
  0
  0
  9
  3
  2

  3
  9
 10
  2
  4
             (2)
(3)
(4)
(5)
ro
i
o
4^

61.
62.
63.
64.
65.
Gal Hum
Magnesium
Sodium
Ca 1 c 1 urn
Bromine
1
5
5
Q
4
U.S.
Reserves Imports







.











:




5
1
1
1
1
1
1
5 1
1
1
10
1
1
1
1
1
1
0
1
1 I
1
1
1
0
GNP. See text.
Vulnerability
of Sources
4
6
3
1
2
2
1
2
8
3
2
2
1
1
2
3
2
3
2
2
1
2
2
0

Difficulty
of
Substitution
6
7
3
8
B
7
6
2
7
5
3
4
6
7
5
3
5
3
3
3
3
3
2
0

(6)
                                             Economic
                                            Importance
                                                I
                                                I
                                                I
                                                2
                                                I
(7)
                                               Industrial
                                                Leverage
                                                   I
                                                   3
                                                   3
                                                   6
                                                   6

                                                   4
                                                   3
                                                   I
                                                   I
                                                   2

                                                   I
                                                   3
                                                   5
                                                   9
                                                   2

                                                   2
                                                   2
66.  Dlatomlte         4
67.  Feldspar          3
68.  Kyanlte           0
69.  Mica-Scrap        4
70.  Pumice            3

71.  Stone-Dlmen       4
72.  Vermlcullte       3
73.  Perlite           3
74.  Scandium          0

"Ranking by effect of shortage  on  GNP.

SOURCE:  Evan E. Hughes,  et  al. , Strategic Resources and National Security:  An Initial Assessment, prepared for
         the Defense Advanced Research Projects Agency by the Stanford Research Institute, April 1975, pp. 202-205.
                                           Geometric
                                             Means
                                           Columns 2-8

                                              2.2
                                              2.2
                                              2.2
                                              2.2
                                              2.1

                                              2.1
                                              2.1
                                              2.0
                                              2.0
                                              2.0

                                              2.0
                                              2.0
                                              2.0
                                              2.0
                                              1.9

                                               .8
                                               .8
                                               .7
                                               .7
                                               .7

                                               .7
                                               .5
                                               .4
                                               .0
                                                                                                                     CHARLES RIVER
                                                                                                                        ASSOCIATES
                                                                                                                    INCORPORATED

-------
                                                      CHARLES RIVER
                                                        ASSOCIATES
                                                     INCORPORATED
summary measure for the seven criteria is calculated simply
by taking a geometric average of all positive ratings.  The
materials are listed in Table 4-2 by their rank according to
this summary measure.1
    The 12 nonfuel materials which were studied intensively by
Levine and Yabroff with their input-output model were titanium,
platinum, cobalt, tin, chromium, aluminum, copper, silver,
nickel, tungsten, zinc and lead.  These materials are indicated
on the left side of Table 4-2 by an asterisk and by an alter-
native ranking which will be explained below.  The lack of
close correspondence between the formal prescreening rank and
the materials which were eventually deemed worthy of further
study is a pointed commentary on the weakness of subjective
weighting schemes.  Considerations such as vulnerability of
supply sources and substitution possibilities are undeniably
important in determining materials criticality, but their
relative importance differs in complicated ways from material
to material; a formal economic model is generally required to
clarify such issues.
    The point is not that prescreening of materials before
detailed construction of criticality ratings is infeasible or
undesirable.  I&ther, given that there are only roughly a
hundred candidate raw materials, it is relatively easy to
select the most useful ones to analyze in depth according
to the principles presented in Chapter 2 of this study.  The
same materials are the major candidates in most studies of
criticality, assuming the results from formal weighting

    1 There is no weighting of the relative importance of the
various criteria, as was done by Cameron and King.
                             2-105

-------
                                                      CHARLES RIVER
                                                        ASSOCIATES
                                                      INCORPORATED
procedures are applied with good judgment.  The criticality of
less major materials should also be analyzed in time, but  just
using roughly conservative assumptions will often show quickly
that the criticality of a material such as dimensional stone  is
not as great as for materials of more immediate interest.
     Consider the case of sheet mica, which Levine and Yabroff
did not study intensively in spite of its having a high
rating (10) in Table 4-2.  Despite the high rating, good
substitutes exist, consumption is declining steadily, and  use
of natural sheet mica is expected to be very low by the year
2000.  The United States has very few reserves of sheet mica,
but resources which could be exploited at considerably higher
prices are sizable.1  Moreover, the U.S. strategic stockpile
contains a very large quantity of the material, some of which
would very probably be available to U.S. industry during any
major disruption of imports.  The criticality of sheet mica
would almost certainly be much lower than indicated in Table
4-2 if it were analyzed according to the principles we have
recommended in this study.

The Cost of Disruptions
     The last report from the SRI study applied a 150-sector
input-output model to gas, petroleum, coal and the twelve
nonenergy materials listed above.  Although this final stage
of the study does not explicitly consider the relative criti-
cality of materials, some basic results are generated which
indicate what one can expect from utilizing an input-output
model to determine materials criticality.
     Producing sheet mica is labor-intensive,  so the cost of
U.S.  labor (relative to that in less developed countries where
production occurs)  makes it uneconomical to exploit U.S.
resources.
                              2-106

-------
                                                      CHARLES RIVER
                                                        ASSOCIATES
                                                     INCORPORATED
     The ultimate results reported in the SRI study are repro-
duced in their graphical form in Figure 4-3.  Shortages of
individual raw materials are measured on the horizontal axis
as a percent of normal consumption, and the percent reduction
in attainable GNP is measured on the vertical axis.1  induc-
tions in gas, petroleum and coal availability quickly cause
catastrophic reductions in GNP according to results
generated by the input-output model.  However, quite con-
ceivable shortages of the 12 nonenergy materials are also
represented as having disastrous effects on GNP.
     It is important to put the results in Figure 4-3 into
perspective before critiquing them.  First, input-output models
have intrinsic limitations which are very difficult to avoid
and which were quite familiar to the authors of the study.
Thus, results for shortages beyond certain points are con-
sidered unrealistic and are so indicated by using dashed lines
in the figure; even the more reliable results represented with
solid lines are artificial enough that they probably should
not be interpreted literally (as we discuss further below).
Second, the results in Figure 4-3 are not intended by the
authors to indicate complete criticality ratings, since they do
not consider such obviously relevant factors as the probabil-
ity of a disruption occurring.2
     1A linear programming algorithm was used to maximize GNP
given the resource shortage.  Because of unrealistic fixed
coefficients production functions and unrealistic constant
values of unit outputs, artificial constraints on industrial
capacities and on final demands had to be imposed in order to
obtain results which were not obviously unreasonable.
     2Based partially on their formal analysis, the authors
do venture the tentative recommendation that chromium,
aluminum, tin and platinum may be the most critical commodi-
ties for economic stockpiling by the United States.  See
Levine and Yabroff, Department of Defense Materials Consump-
tion, op. ait., p. 12.  Additional study is also recommended.
                               2-107

-------
                                                              Figure 4—3



                                     EFFECTS OF ENERGY AND MATERIAL SHORTAGES ON GNP

                                                         ESTIMATED BY SRI
PO
o
00
                  a.
                  2
2
O

H
O

a

ct
                  UJ
                  o
                  a:
                  UJ
                  a.
30



28



26



24



22



20



18



16



14



\2.



10


 8



 6



 4



 2



 0
                         0
            /*' '   Ki    i    /
           A//   i/   /   /
          I SI  I   8    I
         '$/'         I
         &     !
15
                                        20       25      30


                                        PERCENT SHORTAGE
                                                             35
40
45
50
                  SOURCE: Mark D. Levjne and Irving W. Yabroff, Department of Defense Materials Consumption and the Impact

                          of Matmial and Energy Resources Shortages (Menlo Park, Ca.: Stanford Research Institute, November

                          1975), p. 78. Report prepared for the Defense Advanced Research Projects Agency.
                                                                                                                CHARLES RIVCR

                                                                                                                   ASSOCIATES

                                                                                                               INCORPORATED

-------
                                                      CHARLES RIVER
                                                        ASSOCIATES
                                                     INCORPORATED
     Nevertheless, the ranking of materials from left to right
in Figure 4-3 is highly suggestive of the type of results which
a full input-output analysis of criticality would generate.
This ranking of nonenergy materials was indicated above in
Table 4-2 by the bracketed numbers and asterisks at the left
side.*
     Comparing the rankings for nonmetallic titanium and
metallic titanium in Figure 4-3 is a revealing way to indicate
some of the most important limitations of input-output models.
It is suggested that a given percentage reduction in GNP can
be caused by a relatively small shortage of the nonmetallic
titanium but that the same percentage reduction requires a
much larger shortage of metallic titanium.  According to our
interpretation of Figure 4-3, nonmetallic titanium ranks first
among the nonenergy materials in what we might call poten-
tial criticality, while metallic titanium ranks last.
     Metallic titanium popularly has a high-technology image
because of its use in aerospace applications.  However, non-
metallic uses account for the bulk of titanium consumption;
most goes into white pigments for very mundane usage in paints,
paper and plastics.  The wide usage of titanium pigments as an
intermediate good throughout industry accounts for its apparent
importance in Figure 4-3.  As the availability of titanium
decreases, buildings would not be built and appliances would
not be manufactured because of a shortage of white paint, at
least according to an input-output model which allows for no
substitutions.

     *Four materials are omitted from Figure 4-3 for clarity,
but are included in Table 4-2 :  tin and chromium fall between
cobalt and aluminum; copper and silver fall between aluminum
and nickel.
                              2-109

-------
                                                      CHARLES RIVER
                                                        ASSOCIATES
                                                      INCORPORATED
     It is possible at considerable expense to modify input-
output models to take substitution possibilities into account
in a rudimentary fashion.  However, adding to an input-output
model all the features which make the policy model in Chapter
3 very well suited to determining materials criticality
would be an unrealistically complicated job at the present
time.  In particular, the input-output model is static and
does not take into account the successive adaptions which
the economy can make over time.  A related difficulty is that
the GNP loss measures generated by an input-output model are
generally less satisfactory than the economic surplus
measures used in the policy model of Chapter 3; the surplus
measures recognize that successive unit reductions in the
availability of a good inflict larger and larger economic
costs on society.

Analysis of Two or More Markets
With a Policy Model
     As mentioned earlier, the sectoral inclusiveness of
input-output models makes them an important tool when analyz-
ing simultaneous shortages in many markets, as would occur
during a war.  However, even in planning mobilization for a
war, models which are sophisticated in other ways can be
useful for analyzing certain problems.
     For analyzing disruptions in one market or several
markets,  the advantages of models like those discussed in
Chapter 3 of this study are compelling.   Most importantly,
such models allow a realistic dynamic treatment of the adjust-
ments the economy can make daring a disruption, including
substitution of alternative materials and the optimal alloca-
tion of available inventories.  The simpler structure of the
                              2-110

-------
                                                      CHARLES RIVER '.
                                                        ASSOCIATES i
                                                      INCORPORATED!
underlying microeconomic models  also means  that other  aspects
of materials criticality, such as the probability of disrup-
tions of various lengths and severities, can be treated  in a
more satisfactory manner.
     The policy model discussed  in Chapter  3 does assume in
its basic form that prices for inputs other than the poten-
tially critical material are relatively stable during  supply
disruptions.  Such an assumption may be significantly
unrealistic for certain closely  related markets.  However,
generalizing the policy model to take account of several
closely related markets, such as iron and manganese  (comple-
ments) or bauxite and copper (substitutes) would only  be
moderately costly.  More ambitious generalizations of  the
policy model would allow treating simultaneous disruptions in
multiple markets, as long as most of the markets in the economy
were relatively unaffected.  Fortunately for ease of analysis
as well as for peace of mind, it is very unlikely that dis-
ruptions of nonfuel minerals would massively disrupt the U.S.
economy to the extent that such microeconomic analysis would
be an insufficient tool for analysis.  In any case,  the
microeconomic policy model could be nested inside a macro-
economic model if this were considered necessary for analyzing
cataclysmic cases.
     Because it ignores, or at best oversimplifies the substi-
tution possibilities and other dynamic adjustments which the
economic system can make, an input-output model can easily
consider simultaneous disruptions of many materials.  How-
ever, in most cases a microeconomic approach to analyzing
widespread disruptions offers more realistic and reliable
results.
                               2-111

-------
                                                     CHARLES RIVER
                                                       ASSOCIATES
                                                     INCORPORATED
                Other Studies of Criticality

       In the remainder of this chapter we briefly discuss
interesting aspects of other studies bearing on the criticality
of materials.

NMAB Study
       The National Materials Advisory Board (NMAB) recently
completed a study of critical materials which utilized defini-
tions different from those which we have proposed:  "Critical
materials are those that are necessary to manufacture the pro-
ducts required for a national emergency and its accompanying
essential civilian needs."1  "Criticality" thus does not neces-
sarily imply that a serious disruption of supplies is probable;
however, "strategicness" is used to connote unreliable foreign
sources.  In terms of the basic concepts developed in Chapter 2
of the present study, the definition of criticality used in
the NMAB study appears to be more closely related to the total
utility gained from a material than to expected losses of util-
ity.
       In any case, the actual selection of materials in the
NMAB study is not closely tied to the formal definition of
criticality.  In fact, the selection process is only described
in general terms as a subjective weighting of many relevant
factors, with the details omitted.  The materials selected for
detailed study were chromium, germanium, iridium, rhenium,

       National Materials Advisory Board, Committee on the
Technical Aspects of Critical and Strategic Materials, A
Screening for Potentially Critical Materials for the National
Stockpile, Publication NMAB-329 (Washington, D.C.:  National
Academy of Sciences, 1977), p. iv.  Critical and strategic
materials are discussed in this study specifically in the
context of decisionmaking for the U.S. national stockpile.
                              2-112

-------
                                                     CHARLES RIVER
                                                       ASSOCIATES j
                                                     INCORPORATED i
zirconium, hafnium and vanadium.  Chromium was identified as
the most critical of these materials.

C.O.I.E.P. Study
       A task force of federal agencies under the direction of
the Council on International Economic Policy and the National
Security Council produced a useful overview of critical mate-
rials in 1974. l   The basic explicit criterion of criticality
employed was U.S. import dependence, but other considerations
relating to the threat of foreign supply disruptions were
weighted qualitatively; there was no attempt to rank materials
by their criticality.  This study illustrates the type of pre-
screening of materials which can be usefully done before the
more formal and ambitious approach we have proposed is imple-
mented .

OTA S tudy
       A recent study of stockpiling policies by contractors
for the Office of Technology Assessment has a short section on
criteria for selecting materials.2  The analysis is somewhat
unwieldy, involving consideration of separate stockpiles for
each of five objectives:  to discourage or counteract cartel or
unilateral political actions affecting price or supply; to
cushion the impact of nonpolitical import disruptions; to
assist in international materials market stabilization; to
conserve scarce domestic materials; and to provide a market for
temporary surpluses and to ease temporary shortages.  In a

       1 Council on International Economic Policy, Critical
Imported Materials (Washington, D.C.,  December 1974).
       2U.S. Congress, Office of Technology Assessment, An
Assessment of Alternative Economic Stockpiling Policies
(Washington, B.C.,  August 1976), pp. 52-57.
                              2-113

-------
                                                     CHARLES RIVER |
                                                       ASSOCIATES |
                                                     INCORPORATED!
rough way this approach does recognize that materials critical-
ity depends on the type of conditions which exist and the type
of contingencies which threaten.  Selection of materials re-
lated to the above problem areas was finally done by a survey
and consensus of expert opinion, formalized as a "modified
Delphi technique."

NCSS Study
       The National Commission on Supplies and Shortages did
not specifically recommend implementation of a system for
measuring criticality of materials.  However, the approach to
criticality which we have proposed here is logically a prelim-
inary part of the process of policy analysis, most obviously
for determination of stockpile levels.  The Commission's dis-
cussion of stockpiling in fact recognizes the usefulness of the
type of policy model discussed in Chapter 3, and earlier ver-
sions of the CRA policy model are explicitly examined.1  The
Commission's report also recognizes the importance of the cost
measurements which we have emphasized and it delves in depth
into many closely related issues, such as improving data
collection for policy analysis.

NSF Study
     The National Science Foundation sponsored a methodolog-
ical study of materials criticality by International Research
and Technology Corporation in 1974.2  This study alludes to
many of the ideas which we have earlier identified as important,

        1 National Commission on Supplies and Shortages, Govern-
ment and the Nation's Resources  (Washington, B.C.,  December
1976), pp. 131-140.
        2International Research and Technology Corporation,
Critical Materials:  A Problem Assessment, prepared for  the
National Science Foundation  (Arlington, Va.,  May 1974) .
                               2-1H

-------
                                                       CHARLES RIVER
                                                         ASSOCIATES |
                                                      INCORPORATED l
 but it has no unifying theory of criticality and in particular
 fails to take full advantage of economic concepts and theory.
 In the final analysis it simply lists many relevant considera-
 tions without relating them to an explicit economic model
 explaining losses from disruptions.

 AFL-CIO Study
         The AFL-CIO commissioned a study of imported raw mate-
 rials and their importance for U.S. workers and consumers.
         Nine commodities were considered of major signifi-
         cance in terms of use by U.S. industry, import
         dependency, vulnerability to price and supply
         manipulation, and impact on U.S. employment . .  .
         aluminum, copper, lead, zinc, tin, nickel, manga-
         nese, iron ore and chromium.1
 This study did not consider the measurement of materials criti-
 cality in any depth.  It is interesting for current purposes
 only because the materials of greatest concern to the most
 prominent U.S. labor organization are no different from  those
 selected by other investigators, despite a perspective arguably
 narrower than a national concensus.

 POD Workshops
        The proceedings of two workshops sponsored by the
 Department of Defense in 1975 and 1976 are interesting for
 expressing a broad range of concerns relating to and merging
 with the issue of measuring materials criticality.  Shortages
 of fabricated items due to inadequate capacity, OSHA and EPA
 regulations, and even scheduling difficulties are discussed
                 Gutchess and Stanley Ruttenberg, Rend Mate-rials
 for America:  A Program to Assure Meeting Future Needs, prepared for
the Industrial Union Department of the AFL-CIO (July 1975), p.10
                               2-115

-------
                                                     CHARLES RIVER
                                                       ASSOCIATES
                                                     INCORPORATED
in the various papers.1  The most frequent complaint is a lack

of timely and reliable forecasts of materials availability.

We have considerably narrowed the focus of our present inquiry

so that these important issues must be treated elsewhere.

       Admittedly, we have had relatively little to say direct-

ly about how to determine materials which are most critical for

defense purposes, beyond showing in Chapter 2 how such consid-

erations can be made conceptually compatible with criticality

stemming from nonmilitary considerations.  For some materials,

such as germanium (which is used in infrared optical technol-

ogy) ,  nonmilitary criticality may be small relative to other

materials, while military criticality is great.2
            following Proceedings volumes were published by
the Metals and Ceramics Information Center, a Department of
Defense Information Analysis Center:  Workshop on Government
Policies and Programs Affecting Materials Availability  (Feb-
ruary 1976) ; Materials Shortage Workshop  (January 1975) ,
Edward Dyckman, "Review of Government and Industry Shortages,"
Item A in the 1975 volume, is a useful overview of some earlier
work on materials criticality which we have also discussed.

       2See "Demand of New Technology on DOD Material Supply  —
Initial Findings," Item C in the 1975 DOD Workshop Proceedings
volume.
                              2-116

-------
                                                                     Charles
                                                                     River
                                                                     Associates
                            CHAPTER 2 REFERENCES
1.   Adams, Robert L.  1977.  "Evaluative Methods -Some Basic Concepts  and
     Tools of Analysis."  In conference proceedings on Contingency  Planning
     for Materials Resources.  Sponsored by The Engineering Foundation  and
     The National science Foundation, Edward B. Berman, ed.   New York:
     United Engineering Center.

2.   Adams, Robert L., Barbara A. White and James S. Grichan.  1979.
     "Developing a Critical  Minerals Index:  A Pilot Study."   Office  of
     Minerals Policy and Research Analysis, U.S. Department of the  Interior.
     July.

3.   Burrows, James C. and Steven D. Beggs.  1977.   "Policy Implications  of
     Producer Country Supply Restrictions:  Approach and Conclusions."   In
     conference proceedings  of Contingency  Planning for Materials Resources.
     Sponsored by The Engineering Foundation and The National  Science
     Foundation, Edward B. Berman, ed.   New York: United Engineering  Center.

4.   Charles River Associates.  1978.  "Measuring Materials Criticality for
     National Policy Analysis."  Prepared for the General  Accounting  Office.
     March.

5.   	.  1977a.  "Implications of the War in Zaire for the Cobalt
     Market."[Revised).  Prepared for the Office of Minerals Policy and
     Research Analysis, U.S. Department of  the Interior.   June.

6.         	.  1977b.  "The Report of  the U.S. Department of the
     Interior on the Critical  Materials Aluminum, Chromium,  Platinum  and
     Palladium: A Review and Revision."  Prepared for the  Office of Minerals
     Policy and Research Analysis, U.S. Department of the  Interior.   July.

7.   	.  1976a.  "Policy Implications of Producer  Country  Supply
     Restrictions."  Prepared for Experimental  Technology  Incentives  Program,
     National Bureau of Standards, U.S. Department of Commerce.  Vol. I-V.
     August-December.

8.   	.  1976b.  "Public Policy  in the Chromium Market."   Prepared
     for the Office of Minerals Policy  and  Research Analysis,  U.S.  Department
     of the Interior.  October.

9.    	.   1976c.  "Public and Private Stockpiling for  Future
     Shortages."   Prepared for National  Commission  on Supplies and  Shortages.
     August.
                                    2-117

-------
                                                                     Charles
                                                                     River
                                                                     Associates
                      CHAPTER 2 REFERENCES (continued)
10.   	.  1975.  "Economic Issues Underlying Supply Access
     Agreements: A General Analysis and Prospects in 10 Mineral  Markets."
     Prepared for the Bureau of International  Labor Affairs,  U.S.  Department
     of Labor.

11.  Council on International Economic Policy.  1974.   "Critical  Imported
     Materials."  Special Report, Washington,  D.C.:  December.

12.  Dyckman, Edward.  1975.   "Review of Government and Industry  Shortages."

13.  Fischman, Leonard L.  1980.  "World Mineral  Trends and U.S.  Supply
     Problems."  Prepared by Resources for the Future.   Washington,  D.C..

14.  	.  1979.  "Major Mineral Supply  Problems."  Prepared  by
     Resources for the Future.  Prepared for the  Office of Science and
     Technology Policy, Executive Office of the President. Washington, D.C.:
     September.

15.  Gutchess, Joceyln and Stanley Ruttenberg. 1975.   "Raw Materials  for
     America:  A Program to Assure Meeting Future Needs."   Prepared  for the
     Industrial Union Department of the AFL-CIO,  July.

16.  Hughes, Evan E.  et al.  1975.  "Strategic Resources and  National
     Security:   An Initial Assessment."  Prepared for  the  Defense  Advanced
     Research Projects Agency by the Stanford  Research  Institute,  April.

17.  International  Research and Technology Corporation.   1974.   "Critical
     Materials:  A Problem Assessment."  Prepared for  the  National Science
     Foundation, Arlington, Va., May.

18.  King, Alwyn H.  and John  R. Cameron.   1974.   "Materials and the  New
     Dimensions of Conflict."  Carlisle Barracks,  Pa.:   U.S.  Army  War
     College, Strategic Studies Institute, May 15.

19.  King, Alwyn H.   1977.  "Materials Vulnerability of  the United States --
     An Update."  Carlisle Barracks,  Pa.:   U.S. Army War College,  Strategic
     Studies Institute, April 30.

20.  Klass, Michael  W., James C. Burrows,  and  Steven D.  Beggs.  1980.
     "International  Minerals  Cartels  and  Embargoes:  Policy Implications for
     the United States."   Praeger:  N.Y.

21.  Kymm, Kern 0.   1980.   "Development of an  IDEAS  (Industrial Disruptions
     Economic Analysis System)  Critical Industries Strike  Impact Assessing
     System."  June.
                                     2-118

-------
                                                                    Charles
                                                                    River
                                                                    Associates
                      CHAPTER 2 REFERENCES  (continued)


22.  Levine, Mark D. and Irving W.  Yabroff.   1975.   "Department of Defense
     Materials Consumption and the  Impact of Material and Energy Resource
     Shortages."  Prepared for the  Defense  Advanced  Research Projects Agency
     by the Stanford Research Institute,  November.

23.  ^	.  1975.   "Strategic Resources and National Security:  Paper I
     in Proceedings of the Department of  Defense Materials Shortage Workshop,
     Metals and Ceramics Information Center, January.

24.  Materials Shortage Workshop.  1975.  Metals and Ceramics  Information
     Center, Department of Defense  Information and Analysis Center.
     January.

25.  Morgan, John D.  1976.   "Mineral  Data  Improvements  and Critical
     Materials R&D at the U.S. Bureau of  Mines." Proceedings  of the Workshop
     on Government Policies  and Programs  Affecting Materials Availability.
     Columbus, Ohio!Metals and Ceramics Information Center,  February:
     319-344.

26.  National Commission on  Materials Policy.  1973.  "Materials Needs and
     the Environment Today and Tomorrow."  U.S.  Government Printing Office,
     June.

27.  National Commission on  Supplies and  Shortages.  1976.  "Government and
     the Nation's Resources."  December.

28.  National Materials Advisory Board (NMAB).   1981.  "Identification of
     Critical and Strategic  Materials for Naval  Combat Systems."  Commission
     on Sociotechnical  Systems, National  Research Council.  April.

29.            .  1977.   "A  Screening for Potentially Critical Materials for
     the National Stockpile."  Committee  on  the  Technical Aspects of Critical
     and Strategic Materials.  Washington,  D.C.:  National Academy of
     Sciences.

30.  Tilton, John E.  1977.   "The Future  of  Nonfuel  Minerals."  The Brookings
     Institution, Washington, D.C.

31.  U.S.  Congress.  1976.   "An Assessment of Alternatives Economics
     Stockpiling Policies."   Office of Technology Assessment,  Washington,
     D.C.:  52-57.

32.  U.S.  Department of the  Interior.   1979a.  "Domestic Policy Review of
     Nonfuel Minerals."   Draft report, Vol.  I-II.  Volume III  prepared by the
     Department of Mineral Economics,  Pennsylvania State University for the
     U.S.  Bureau of Mines.
                                     2-119

-------
                                                                    Charles
                                                                    River
                                                                    Associates
                      CHAPTER 2  REFERENCES  (continued)


33.  U.S. Department of the Interior.   1979b.   "Mineral Trends and
     Forecasts."

34.  U.S. Department of the Interior.   1975.   "Critical Materials:  Commodity
     Action Analyses (Aluminum,  Chromium, Platinum and Palladium).  Prepared
     by the Office of Minerals Policy  Development.  March.

35.  U.S. Department of the Interior.   Mining  and Minerals Policy.  Issued
     annually.

36.  U.S. Federal  Emergency Management Agency.  Stockpile Report to the
     Congress.   Issued quarterly.

37.  Webster's  New Collegiate Dictionary. 1976.  G. & C. Merriam Co.

38.  Workshop on Government Policies and Programs Affecting Materials
     Availability.  1976.   The Metals  and Ceramics Information Center,
     Department of Defense  Information Analysis Center.  January.

39.  Workshop Proceedings,  Department  of Defense.  1975.  "Demand of New
     Technology of DOD Material  Supply — Initial Findings."  Item C.
                                   2-120

-------
                                                                 Charles
                                                                 River
                                                                 Associates
  PROJECTIONS OF MATERIALS CONSUMPTION  FOR CONTROL OF VEHICULAR EMISSIONS
Rath & Strong projected  consumption of materials for U.S. vehicular emissions
control for this study.   Only  projected consumption for the platinum-group
metals is reported in detail here, because, as confirmed in Chapter 2, these
are by far the most critical materials from the perspective of vehicular
emissions control.  However, order of magnitude estimates for other
potentially critical  materials,  as used in Chapter 2, are also presented
below.

Tables 3-1 through 3-3 project consumption of platinum, palladium, and
rhodium by the four significant  U.S. auto manufacturers, for emissions
control on cars and light trucks.  These projections are based upon assuming
a constant volume of U.S. production over the reported time horizon from 1980
to 1987.  Assumed production by  U.S. manufacturers, in thousands of vehicles
per year, is as follows:
General Motors
Ford
Chrysler
American Motors

     Total
 Cars

5,700
2,090
  950
  285

9,025
 Li ght
Trucks

 1,382
   916
   272
   154

 2,724
 Total

 7,082
 3,006
 1,222
   439

11,749
The projections in Tables 3-1 through  3-3  recognize differences in the
technologies used for emissions control  by the  four U.S. manufacturers, and
the effect of future mixes of engine types.  One  important reason for
                                    3-1

-------
                                                                   Charles
                                                                   River
                                                                   Associates
Table 3-1
U.S. CONSUMPTION OF PLATINUM FOR VEHICULAR EMISSIONS CONTROL, 1980-1987
(1000 grams per year"!
Year

1980
1981
1982
1983
1984
1985
1986
1987
 Cars

20,096
19,060
15,651
14,780
13,980
14,244
15,595
14,468
 Light
Trucks
 6.
 5,
 4.
301
976
936
 4,723
 4,
 4.
507
561
 4,857
 4,535
 Total

26,397
25,036
20,587
19,503
18,487
18,805
20,452
19,003
SOURCE:  Rath & Strong, 1981.
                                     3-2

-------
                                                                   Charles
                                                                   River
                                                                   Associates
Table 3-2
U.S. CONSUMPTION OF PALLADIUM  FOR VEHICULAR  EMISSIONS CONTROL,  1980-1987
(1000 grams per year)
Year

1980
1981
1982
1983
1984
1985
1986
1987
 Cars

8,534
8,115
6,025
4,661
4,414
4,493
4,871
4,553
Light
Trucks
  676
  543
  847
  506
  434
1,449
1,537
1,435
 Total

11,210
10,658
 7,872
 6,167
 5.
 5.
 6,
848
942
408
 5,988
SOURCE:  Rath & Strong,  1981.
                                     3-3

-------
                                                                    Charles
                                                                    River
                                                                    Associates
Table 3-3

U.S. CONSUMPTION OF RHODIUM FOR VEHICULAR EMISSIONS CONTROL,  1980-1987
(1000 grams per
Year

1980
1981
1982
1983
1984
1985
1986
1987
 Cars
 25.
 13.
339.
947.
851.
858.
964.
Light
Trucks
  7.2
  3.9
140.
291,
266,
266,
291,
.5
.7
.1
.2
.2
877.7
268.1
 Total

   32.4
   17.4
  479.6
  238.8
  118.0
  124.6
1,255.7
1,145.8
SOURCE:  Rath & Strong, 1981.
                                       3-4

-------
                                                                  Charles
                                                                  River
                                                                  Associates
 projected  declines in platinum and palladium consumption between 1980 and
 1987,  is assumed reductions in the size of gasoline engines and greater usage
 of diesel  engines.  Rath & Strong assumed that diesel  engines will  be
 equipped with a monolithic substrate contained in a clamshell  particulate
 trap  largely made of 409 stainless steel, in which particles are burned using
 hot exhaust gases.  No use of noble metals in the particulate traps is
 assumed.

 The Rath & Strong subproject report for this contract contains a detailed
 description of the assumptions underlying their projections, and a  wealth of
 data  beyond that required for the immediate purposes of this study  (such as a
 disegregation of platinum-group consumption by vehicle manufacturer).   Thus,
 their report is included with this study as a separately bound appendix.

 Appendix 3-A presents alternative estimates, by the Environmental Protection
 Agency, of consumption of platinum-group metals for control  of vehicular
 emissions  in 1981.  The EPA estimates differ considerably from those
 presented  for 1981 in Tables 3-1 through 3-3, particularly for rhodium.

 The detailed projections of platinum-group consumption provided in  Tables 3-1
 through 3-4 are of some interest in their own right.   However,  such detail
 and accuracy is not really required for the rough estimation of materials
 criticality performed in Chapter 2.   For that purpose, we simply round  off
 projected  annual consumption for the period 1985-1987  roughly as follows  (in
 thousands  of grams per year):

 •     Platinum:  20,000
 •     Palladium:  6,000
 •     Rhodium:    1,200

 Other  potentially critical  materials used in vehicular systems  for  emissions
 control, fuel  management, and related purposes,  are chromium, manganese,
 nickel, and titanium.   These materials are mostly contained  in  the  stainless
 steel   (type 409) used in the catalytic converter.   The following very rough
 estimates of annual  consumption  of these materials by  U.S. vehicle
 manufacturers, for emission control  and related  purposes  on  cars and light
 trucks, were provided by Rath and Strong (in short tons per  year):

 •    Chromium:   10,600
 t    Manganese:    730
 •    Nickel:        360
 •    Titanium:      350

 As  explained in the preceding  chapter,  these quantities are  small proportions
 of  total U.S.  consumption of these metals,  and their criticality from the
 perspective of U.S.  vehicle manufacturers  and EPA is much  less  than the
 criticality of the platinum-group metals.   (We confirmed  for this project
 that consumption for emissions control  of  several  other metals, including
molybdenum and vanadium,  is nonexistent or  entirely inconsequential.)
                                   3-5

-------
                                                                 Charles
                                                                 River
                                                                 Associates
               EPA ESTIMATES OF CONSUMPTION OF PLATINUM-GROUP
             METALS FOR  CONTROL OF VEHICULAR EMISSIONS IN 1981
Tables 3-1 through 3-3  in  the  text provide Rath & Strong estimates of U.S.
consumption of platinum, palladium and rhodium for vehicular emissions
control in 1980 and 1981,  as well as for the 1985-1987 time frame considered
by Charles River Associates in Chapter 2.  The Rath & Strong estimates for
1981 differ substantially  from best estimates produced internally by the
Control Technology Assessment  and Characterization Branch of EPA's Office of
Air, Noise and Radiation.  The EPA estimates are given in Table 3A-1.

Time did not allow us to investigate the reason for the large discrepancies
between Tables 3-1 through 3-3, and the EPA estimates for 1981.  In any case,
the Rath & Strong estimates for 1985-1987 are much more compatible with the
EPA estimates for 1981, and the 1985-1987 projections were the basis for the
work in Chapter 2.
                                   3A-1

-------
                                                                 Charles
                                                                 River
                                                                 Associates
Table 3A-1
U.S. CONSUMPTION OF PLATINUM, PALLADIUM AND RHODIUM FOR VEHICULAR
EMISSIONS CONTROL, EPA ESTIMATES FOR  1981
     Number of Light-Duty Gasoline Vehicles
     Produced in the United States:  8,864,444
     Grams of Platinum-Group Metals Consumed per Vehicle:
               Platinum:     1.957
               Palladium:    0.653
               Rhodium:      0.184
               Total:        2.794

     Implied Total  U.S. Consumption of Platinum-Group Metals,
     in Thousands of Grams:
               Platinum:    17,350
               Palladium:    5,790
               Rhodium:      1,630
     Equivalent Total  U.S. Consumption
     of Platinum-Group Metals, in Troy Ounces
               Platinum:   557,700
               Palladium:  186,000
               Rhodium:     52,400
SOURCE:   Environmental  Protection Agency,  Office of Air, Noise and Radiation,
         Control  Technology Assessment and Characterization Branch.
         August 5,  1981.
                                  3A-2

-------
                                                                   Charles
                                                                   River
                                                                   Associates
                            PLATINUM-GROUP METALS
                                INTRODUCTION


This chapter focuses on the platinum-group metals (PGMs)  used in vehicular
emissions control  systems.  As discussed in another chapter,  there  is
currently no economical alternative to the noble-metal  catalytic converter
which uses platinum, palladium, and rhodium.  Supplies  of these  materials
thus become an important issue when contemplating the costs of auto emission
regulations.

The supply elasticity of PGMs is discussed in this chapter by time-frame of
response.  Stockpiles of PGMs are first analyzed in terms of their  ability to
bridge a short-term demand/supply gap.  Following the stocks discussion is an
analysis of world primary production elasticity by country, a discussion of
total world PGM reserves, and an analysis of world supply reliability.

Following the primary production section, other PGM-using industries are
examined, focusing on the response of consumption and secondary  recovery in
each industry to PGM price rises.  The chapter concludes  with a  discussion of
the role of speculation in PGM markets.


                                   STOCKS
The most immediate possible supply response to any increase in demand for
platinum group metals would come from stocks.  The key questions then become
                                     4-1

-------
                                                                   Charles
                                                                   River
                                                                   Associates
what government and private stocks of PGMs exist, how large they are,  and
what the immediate supply elasticity of the stocks is.  The secretive  nature
of the platinum industry makes these questions difficult to answer precisely;
nevertheless, much information is public and discussions with industry
personnel have revealed additional information.   The industry secrecy  often
has good cause; one firm recently told the U.S.  Bureau of Mines  that it would
not be reporting any PGM stocks during the current period due to forced
removal by armed robbery.

There are several  different types of PGM stocks.  Table 4-1 categorizes PGM
stocks in the United States and indicates whether reliable data  are regularly
reported on each of them.

The following sections discuss each kind of stock and available  information
about their size.
U.S. NATIONAL STOCKPILE

Currently, the U.S. government maintains stockpiles of three platinum-group
metals -- platinum, palladium, and iridium.   As Table 4-2 indicates,  current
inventories of each metal  are well below stated goals of the General  Services
Administration (GSA), which is in charge of maintaining stockpiles of U.S.
critical materials.  The current total  PGM inventory of 1,725,000 troy ounces
has not changed since 1971.

There is industry speculation that the  GSA may soon purchase PGMs.  One
mechanism that would facilitate this purchase is the proposed National
Defense Stockpile Transaction Fund.  Monies entering the fund through sale  of
commodities in excess of their goals can be used to purchase other
commmodities in deficit.  Silver and tin are two excess commodities,  although
recent GSA efforts to sell  silver have  been stopped in Congress.

The anticipated military buildup under  the Reagan administration, according
to industry sources, is expected to lead to primary focus on platinum-group
metals and cobalt for stockpile acquisition.  (See American  Metal Market,
1981.)  It is too early to tell how much of the PGM stockpile deficit will  be
eliminated and, of course,  planning of  GSA activities in commodity markets  is
kept secret in order to avoid speculative reaction.


REFINER, IMPORTER, AND DEALER STOCKS

U.S. Bureau of Mines data on stocks of  PGMs held by refiners, importers,  and
dealers from 1975 to 1980 are reproduced in Table 4-3.  The  figures show  a
squeezing down of inventories starting  in 1978 due to higher metals prices
and strong demand.  Figure 4-1 plots the changing composition of  refiner,
importer, and dealer stocks among the PGMs during the 1970s.
                                     4-2

-------
                                                                   Charles
                                                                   River
                                                                   Associates
Table 4-1

PLATINUM GROUP METAL STOCKS



      Type	      Regular Data Reported?

      U.S. National  Stockpile                            Yes
      Refiner, Dealer,  and  Importer  Stocks               Yes
      Industry Shelf Stocks                             No
      Industry Stocks in  Use                             No
      Private Speculative/Investment Stocks              No



SOURCE:  Charles River Associates,  1981.
                                      4-3

-------
                                                                   Charles
                                                                   River
                                                                   Associates
Table 4-2

CURRENT STATUS OF U.S.  NATIONAL DEFENSE  STOCKPILE
(Thousands of Troy Ounces)
         Material                Goal                Current  Inventory

         Platinum               1,310                    453
         Palladium              3,000                   1,255
         Iridium                   98                      17
         Total                   3~TO                   T772T
SOURCE:  U.S. Bureau of Mines,  1981,  Mineral  Commodity  Summaries,  Washington,
D.C.:  U.S. Government Printing Office.
                                       4-4

-------
Table 4-3


REFINER, IMPORTER,  AND DEALER STOCKS
OF PLATINUM GROUP METALS,  1975-1980*
(Troy Ounces)
                                                                   Charles
                                                                   River
                                                                   Associates
         Year
Platinum
Palladium
Rhodium
1975
1976
1977
1978
1979
1980**
420,770
536,318
438,045
369,823
305,605
493,000
335,621
459,765
475,358
369,937
323,865
293,831
53,847
47,769
48,392
51,322
47,678
46,421
*Includes metal in depositories of the New York Mercantile  Exchange.


**As of December 31,  1980.
SOURCE:  U.S. Bureau of Mines, various issues,  Minerals  Yearbook  and Mineral
         Industry Surveys, various issues,  Washington, D.C.:   UTS.  Government
         Printing Office.
                                       4-5

-------
t
r
o
o
u
n
c
  Figure 4—1

  Platinum Group Metal Stocks held by Refiners, Importers, & Dealers

                     CPrepared By:  Char lee River Associates)
   1 250038
   1000000 -
    750000 -
    500000 ~
    250000 ~
             1970  1971   1972  1973  1974  1^75  1976  1977  1978  1979
            Rhod i urn
            Platinum
            Pa I I adI urn
                        Other
years
SOURCE*  U.S. Bureau of  Mines, Minerals Yearbook,
Mineral  Industry Surveys,  various  Issues, Washington, D.C.
                                                                           Charles
                                                                           River
                                                                           Associates

-------
                                                                   Charles
                                                                   River
                                                                   Associates


The U.S. Bureau of Mines (U.S.  BOM)  collects  these data by sending out a
questionnaire to all  U.S.  domestic  refiners,  dealers, and importers.  U.S.
BOM sources indicate  that  the  returned  questionnaires capture about 90
percent of the firms.   A U.S.  BOM telephone survey revealed that the less
than ten percent of firms  that did  not  respond dealt with minute quantities
of PGMs.  Hence, it can be safely concluded that the U.S. BOM reports capture
well over 95 percent of refiner, importer, and dealer PGM stocks.


INDUSTRY SHELF STOCKS
There exist no published data on the amount of platinum-group metals held in
inventory for future use by industries  in  the  United  States or abroad.  Firms
generally will not disclose information of this nature, so knowledgeable
industry observers willing to talk are  one of  the  few readily available
sources of information on industry PGM  shelf stocks.

The National  Materials Advisory Board (NMAB) in their recent report on
government PGM stockpiling strategies stated that  the chemical, petroleum,
and glass industries hold large inventories of PGMs.   (See NMAB, 1980.)
However, the NMAB did not provide any order of magnitude estimates of
industry shelf stocks.  The only bit of quantitative  information in their
report on shelf stocks is that from 1971 to 1977 the  petroleum industry added
an estimated 350,000 to 450,000 troy ounces of platinum to its shelf
stockpiles.

Discussions with knowledgeable industry observers  have tended to corroborate
the NMAB claim that the petroleum, chemical, and glass industries maintain
large PGM inventories.  One source stated  that the chemical industry probably
has over one year's supply of replacement  needs, and  the source "firmly
believed" that the glass industry has one  year's supply as well.  Another
source stated that petroleum companies  sometimes lease their PGM stocks and
usually have one year's supply as shelf stocks.

Applying this one-year estimate to data on sales to consuming industries
data, one can arrive at lower bound estimates  for  PGM shelf stocks for the
chemical, petroleum, and glass industries  (Table 4-4).  (See Tables 4-21 and
4-23 for basic data.)

The U.S. automobile industry, the largest  domestic user of platinum and
rhodium, and the third largest user of  palladium (in  1979), is completely
secretive about its PGM inventories.  It is reported, however, that the
automobile companies purchase PGMs directly from primary producers, and in
turn sell them back to catalyst manufacturers  such as Engelhard or
Johnson-Matthey.  Auto company personnel interviewed  would not divulge the
size of their PGM inventories.
                                     4-7

-------
                                                                   Charles
                                                                   River
                                                                   Associates
Table 4-4


LOWER-BOUND ESTIMATES OF PLATINUM GROUP  METAL  SHELF  STOCKS,

BASED ON 1979 RATES OF CONSUMPTION

(Troy Ounces)
       Industry
Platinum
Palladium
Rhodium
Chemical
Petroleum
Glass
98,600
170,013
88,594
199,743
24,588
1,729
11,684
1,223
15,276
SOURCE:   Order of magnitude estimates  by  Charles  River  Associates,  1981,
                                     4-8

-------
                                                                   Charles
                                                                   River
                                                                   Associates
INDUSTRY STOCKS IN USE
Platinum-group metals actually being used in production  processes  are
referred to as PGM "stocks in use,"  because in  many  industries  (e.g.,
chemical, glass, and petroleum) PGMs are used as  catalysts  or otherwise
indirectly, and loss rates are quite low.  In petroleum  refining,  for
example, platinum is used as a catalyst and thus  not consumed by the
production process, but it must be periodically recycled because of high
temperatures and contamination; the  loss rate is  around  3 percent  during
recycling.  For the chemical industry the recycling  loss rate is about 18
percent, and for the glass industry  about 3 percent.

In a 1976 report, (CRA, 1976) CRA used these loss rates, data on industry
growth, and current annual industry  PGM consumption  data to arrive at order
of magnitude estimates of stocks in  use.  Very  rough estimates  of  1979 stocks
in use can be generated from these original estimates and the ratio of
current to previous annual PGM consumption by industry.   These  estimates are
provided in Table 4-5.

There are dangers inherent in this estimation methodology,  however.  First,
Bureau of Mines consumption data are not comprehensive,  but are based on
reported sales to consuming industries.  In its annual Mineral  Commodity
Summaries, the Bureau of Mines calculates "apparent  consumption" with an
accounting formula based on imports, and this figure is  always  well above
reported sales to consuming industries.*  A second difficulty with an
estimation method of this nature is  that industries  purchase PGMs  not only to
replenish stocks in use, but also to adjust shelf stocks.  If in a given year
an industry's shelf stocks are being increased, consumption would  be high and
so would the subsequent stock-in-use estimates.

In late 1980 CRA designed, for the U.S. Department of the Interior, a new
model of platinum and palladium consumption and production  that focused on
engineering estimates of the speed and extensiveness with which consumption
of the two metals could be reduced after supply disruptions so  severe that
their price would rise to five or ten times normal levels (which is
considerably beyond the range of historical experience).

Part of this modeling effort involved estimating  the typical holding period
prior to recycling of stocks in use, and estimating  recovery rates.  This
*The accounting formula is:

     Mine Production + Secondary Refining Production  + (Imports  -  Exports)
     + (Beginning Stocks - Ending Stocks).
                                     4-9

-------
Table 4-5

ESTIMATES OF PLATINUM AND PALLADIUM STOCKS IN  USE,  1979
(Troy Ounces)
                                                                   Charles
                                                                   River
                                                                   Associates
Industry

Petroleum
Chemical
Glass
Electrical
Other*
 Platinum
1,787,500
1,158,168
1,098,900
1,653,106
1,895,528
Palladium

  849,940
1,874,896
   24,040
5,027,132
3,924,000
*Includes jewelry, medical,  dentistry,  and miscellaneous.


SOURCE:  Charles River Associates estimates,  based  on  U.S.  Bureau  of Mines,
         Minerals Yearbook,  various issues, Washington,  D.C.:
         U.S. Government Printing Office.
                                       4-10

-------
                                                                   Charles
                                                                   River
                                                                   Associates


model will provide more sophisticated estimates of platinum and palladium
stocks in use, which would largely overcome the two difficulties described
above.  Unfortunately, this platinum-palladium model  has not yet been  fully
implemented for computation.

The stock-in-use estimates shown in Table 4-5 are rough order  of magnitude
numbers, and the confidence intervals are unknown.  They serve as a  useful
reference point, however.


PRIVATE SPECULATIVE/INVESTMENT STOCKS

Platinum-group metals have recently become an object of private speculative
investment, probably due in large measure to the realization that the  United
States is extremely import dependent and industrial use is rising.  The New
York Mercantile Exchange sponsors trade in platinum futures contracts, in
response to this interest by investors outside the chain of producers  and
consumers.

Absolutely no estimates are published on the amount of PGMs in the hands of
private investors.  The Bureau of Mines omits sales to private investors from
the "miscellaneous sales"  category in its publications, because of
confidentiality problems due to the small number of buyers.  Bureau  personnel
have  indicated that these undisclosed numbers have been quite  small  to date,
however.

In commodities futures trading, only a small percent of the contracts  are
actually consummated by delivery, because an investor's long or short
position is usually nullified with an offsetting contract before the
specified delivery date.  At that point, profits or losses are counted and
the trader is out of the market, for richer or poorer.  Hence, the amount of
existing valid platinum futures contracts or "open interest" is not the
proper indication of how much metal is actually trading hands  through  the
futures market.  In general, it seems unlikely that much platinum is being
held  by individuals who fear economic or political chaos, and  consequent
depreciation of paper currency.  Such individuals seem much more likely to
hold  traditional media of exchange, such as gold or silver.  Indications are,
therefore, that the stock of PGMs in the hands of private investors is quite
small relative to other kinds of stocks discussed in this section.
STOCKS AND INCREASED DEMAND IN THE SHORT RUN

The above order of magnitude estimates of various PGM stocks allow us to
address the question of how large a gap between demand and supply they could
bridge, given a one or two year lag in any response of primary production.
                                     4-11

-------
                                                                   Charles
                                                                   River
                                                                   Associates


U.S. auto manufacturers  could perhaps buy  a  small percentage of shelf
stocks held by other industries  at a moderate  premium.  However, astronomical
prices probably would be required  to induce  sale of anything like a majority
of shelf stocks held by  other industries,  particularly at times when market
conditions were perceived to be  more uncertain than usual.

Simple arithmetic demonstrates that, except  for rhodium, U.S. government
stockpiles could help meet increased PGM demand for the auto industry
(Table 4-6).  However, such action is highly uncertain due to the fact that
government stocks are reserved for emergencies, unless compelling reasons
lead to special Congressional  action.   No  shipments from U.S. government PGM
stockpiles have occurred in recent years.

Table 4-7 presents current refiner, importer,  and dealer stocks as  a percent
of 1979 auto industry consumption. It  is  physically possible for these
stocks to meet substantial  percentage increases in auto industry PGM demands.
However, evidence indicates that flows  from  this source can be
price-inelastic.  For example, while the dealer price for platinum  increased
70.5 percent in 1978, refiner, importer, and dealer stocks of platinum fell
by only 17.4 percent during that year.   Similar price-inelastic behavior
occurred for palladium and rhodium. These figures suggest that PGM prices
might have to increase dramatically to  encourage flows from stocks  held by
refiners, importers, and dealers in the United States.  The price required to
call forth these private stocks  could be more  moderate than these numbers
suggest if the demand/supply gap is expected to be only temporary.


                  STATISTICAL OVERVIEW  OF  SUPPLY AND DEMAND
In this section, a brief outline of PGM supply and demand  is  presented  to  set
the stage for subsequent analysis.   First,  data on world primary  production
are presented, and then industrial  use of PGMs in the  United  States  is
described.
SUPPLY


Due to lags in data gathering and publication, country-specific  estimates  of
world primary PGM production have been published by private  sources  only
through 1977.  More current production data, where available,  are  discussed
in producer country profiles below.  Tables 4-8 through 4-10 present world
estimates by country of primary production for platinum,  palladium,  and
rhodium, respectively.  These tables reflect the central  fact  that over 98
percent of the world's primary production of platinum-group  metals is
                                    4-12

-------
                                                                   Charles
                                                                   River
                                                                   Associates
Table 4-6

U.S. NATIONAL STOCKPILES  AS A  PERCENT
OF 1979 AUTO INDUSTRY PLATINUM METAL GROUP CONSUMPTION
(Nominal  Data in Troy Ounces)
              1979 Auto
         Industry Consumption
    U.S.
 Stockpiles
November 30,
   1980*
   Stockpiles as a
Percent of Consumption
  (Column 2/Column 1)
Platinum
Palladium
Rhodium
803,229
222,156
26,136
453,000
1,255,000
0
56.4
565.00
0
*As of November 30,  1980.
SOURCE:  U.S. Bureau of Mines,  1979, Minerals  Yearbook, and 1981, Mineral
         Commodities Profiles,  Washington,  D.CT:U.S. Government Printing
         Office.
                                      4-13

-------
                                                                   Charles
                                                                   River
                                                                   Associates
Table 4-7

RECENT U.S. REFINER,  IMPORTER,  AND DEALER STOCKS
OF PLATINUM GROUP METALS AS A PERCENT
OF 1979 AUTO INDUSTRY CONSUMPTION
(Troy Ounces)

                                      Refiner,
                                   Importer,  and
               1979 Auto         Dealer Stockpiles
          Industry Consumption    September 30,  1980
   Stockpiles  as
   a  Percent of
   Consumption
(Column I/Column 2)
Platinum
Palladium
Rhodium
803,229
222,156
26,136
402,310
321,065
50,021
50.1
144.5
191.4
*As of August 30, 1980.
SOURCE:  Charles River Associates Incorporated,  1981.
                                       4-14

-------
Table 4-8


ESTIMATED WORLD PRODUCTION OF PRIMARY
PLATINUM BY COUNTRY,  1971-1977
(Thousands of Troy Ounces)
                                                                   Charles
                                                                   River
                                                                   Associates
                     1971


Australia


Canada                219


Colombia               26


Ethiopia                o


Japan*                  3


Philippines         ,    1


South Africa          750


USSR                  690


United States*      	10_


Total               1,701
1972
-
187
24
0
4
3
870
705
4
1,799
1973
0
163
26
0
4
2
1,416
735
6
2,354
1974
0
177
21
0
4
1
1,700
750
4
2,658
1975
0
184
23
0
5
1
1,559
795
5
2,572
 1976


    o


  198


   26


    o


    9





1,680


  840
 1977


    o


  216


   25


    o


   10





1,740


  840
2,753   2,831
*Production of refined metal, some from imported ores and crude  palladium.


SOURCE:  U.S. Bureau of Mines, Minerals Yearbook, various issues,  Washington,
         D.C.:  U.S. Government Printing Office; and Roskill  Information
         Services, Ltd., 1979, The Economics of Platinum Group Metals,
         London.
                                      4-15

-------
Table 4-9

ESTIMATED WORLD PRODUCTION OF PRIMARY
PALLADIUM BY COUNTRY,  1971-1977
(Thousands of Troy Ounces)
                                                                   Charles
                                                                   River
                                                                   Associates
Australia

Canada

Japan*

Philippines

South Africa

United States*

USSR

Total
1971
-
190
5
2
438
10
1,380
2,026
1972
-
162
6
5
508
11
1,410
2,103
1973
1
142
6
4
826
13
1,470
2,463
1974
1
154
11
2
991
9
1,500
2,669
1975
1
160
14
1
909
11
1,590
2,687
1976
2
172
18
-
980
6
1,680
2,859
1977
2
188
20
-
1,015
5
1,680
2,910
*Production of refined metal,  some  from  imported ores and crude palladium.

SOURCE:  U.S. Bureau of Mines, various issues, Minerals Yearbook, Washington,
         D.C.:  U.S. Government Printing Office; and Roskill Information
         Services, Ltd., 1979, The  Economics  of Platinum Group Metals,
         London.
                                       4-16

-------
Table 4-10


ESTIMATED WORLD PRODUCTION OF PRIMARY
RHODIUM BY COUNTRY, 1971-1977 (Thousands of Troy Ounces)
                                                                   Charles
                                                                   River
                                                                   Associates
Canada


South Africa


U.S.A.


U.S.S.R.


Total (countries
       listed)
                     1971
1972    1973
1974
1975
1976
1977
14
44
0
46
104
12
51
0
47
110
11
83
0
49
143
11
99
0
50
160
12
91
0
53
156
14
98
0
56
168
14
101
0
56
171
SOURCE:  Estimates by Roskill  Information Services,  Ltd.
                                       4-17

-------
                                                                   Charles
                                                                   River
                                                                   Associates
obtained from only three countries  --  South  Africa, the Soviet Union, and
Canada.  Canada is a junior member  of  the  "big  three"; clearly the Soviet
Union and South Africa dwarf other  PGM producing countries.  South Africa
predominates in platinum with 61.5  percent of world production in 1977, while
the Soviet Union dominates in palladium with 57.7 percent of world production
in 1977.  South Africa accounted  for 59.1  percent of 1977 world rhodium
production, as reported by Roskill; the corresponding percentages for the
Soviet Union and Canada were 32.7 and  8.2, respectively.

These tables also indicate that U.S. production of PGMs is inconsequential;
we are almost completely import-dependent.   This situation is expected to be
mitigated in the future, however, with the development of U.S. PGM deposits
discussed below.

Tables 4-11 through 4-14 indicate the  countries upon which the United States
is directly import dependent.  Countries that are important exporters to the
United States differ somewhat from  the primary  producing countries, because
ores and concentrates are often exported to  other countries for refining.  In
the long run, the distribution of output among  the primary producing
countries is by far the more important consideration.  However, in the short
run, trade patterns can be difficult to redirect due to long-term contracts
between producers, refiners, and users, the  location of processing
facilities, and similar considerations.  Thus,  existing trade patterns are of
some importance for analyzing the short-run  effects of supply disruptions.

Table 4-11 contains data for all  platinum-group metals, while Tables 4-12,
4-13, and 4-14 pertain to platinum, palladium,  and rhodium, respectively.
The importance of the United Kingdom as an exporter of PGMs to the United
States is clear from the tables.  The  United Kingdom is important because of
refining done there by companies like  Johnson-Matthey, which owns part of
Rustenburg Platinum Mines, South Africa's largest platinum mine.  Other
countries that export sizable amounts  of PGMs to the United States include
Japan and Switzerland.  Figures 4-2 through  4-5 aid in the interpretation of
the import data for platinum, palladium, and rhodium.  These graphs trace
major countries' changing shares of the U.S. import market throughout the
1970s.  Two related trends emerge clearly:  the increasing importance of
South Africa and the decreasing importance of the Soviet Union in the 1970s.

U.S. refining of platinum-group metals throughout the 1970s is outlined in
Tables 4-15 through 4-18.  Data are presented on primary and secondary toll
and nontoll refining for all platinum  group  metals.  Data are presented
separately for platinum, palladium, and rhodium.

U.S. primary refining has sharply fallen over the 1970s.  These data in part
reflect South African efforts to refine more metal within that country,
rather than exporting ores or concentrates.
                                     4-18

-------
Table 4-11


U.S. IMPORTS  FOR  CONSUMPTION OF
PLATINUM METALS BY COUNTRY,1 1970-1980

(Troy Ounces)
 Country
                      1970
              1971
1972
1973
                                                                1974
1975
1976
                                                                              1977
                                                                1978
                                                                                                                      1979
                                                                           1980
Australia
Belgium-Luxembourg
Canada
Chile
Colombia
France
Germany, Federal
Republic of
Italy
Japan
Mexico
Netherlands
Norway
Romania
South Africa,
Republic of
Sweden
Switzerland
USSR
United Kingdom
i Yugoslavia
— • Other
6,093
14,184
24,745
--
27,547
--

10,320
--
25,628
6,411
10,021
11,676
--

115,500
--
1,671
494,978
651,895
--
131,138
5,099
24,466
51,365
--
26,856
5

9,684
1,240
24,170
4,624
781
12,553
--

165,672
4,067
470
407,628
558,009
--
91.3543
8,352
15,917
39,605
--
25,535
--

19,676
--
111,875
15,669
1,428
23,361
--

237,697
2,807
2,099
736,264
589,711
-_
62,188''
4,904
21,807
33,281
1,761
27,210
953

10,957
--
164,468
12,935
49,649
38,287
--

245,411
4,253
29,149
882,267
806,423
_-
168,918s

37,056
57,972
--
20,165
2,205

12,731
--
208,119
25,980
6,206
74,974
--

1,016,458
8,077
9,381
1,012,321
734,458
--
14.5026
6,476
30,418
57,962
--
21,588
2,244

22,490
350
51,567
12,132
6,000
36,137
--

837,081
9,531
11,547
331,267
362,168
1,225
20.1017
6,280
31,416
93,648
--
18,201
11,985

36,014
5,409
19,864
20,424
6,130
28,103
--

1,241,669
7,679
13,641
652,112
455,193
4,862
14.4298
1,600
26,249
88,510
--
7,989
7,074

25,370
2,800
18,581
152,402
10,602
19,480
2,680

1,267,191
4,067
20,440
617,215
226,657
3,876
7,591
..
49,472
90,305
1,608
14,550
--

--
33,408
29,597
106,780
29,014
13,798
--

1,591,925
--
23,178
552,666
343,503
--
41,607
„
34,783
80,668
--
15,707
--

--
7,723
35,002
31,867
33,068
32,085
--

2,083,209
--
40,324
693,215
305,522
--
85,955
10,396
102,246
98,330
--
3,988
5,735

54,140
15,194
22,388
41,997
59,304
17,629
--

1,908,325
8,929
31,096
376,747
503,321
3,247
242.9309
Total
1,531,807   1,388,043  1,892,184   2,502,633  3,240,605  1,820,284  2,667,059  2,510,374  2,921,411   3,479,128  3,505,042
Table continued  on  following page.
                                                                                                                               Charles
                                                                                                                               River
                                                                                                                               Associates

-------
           Table 4-11 (Continued)
           U.S. IMPORTS FOR CONSUMPTION OF
           PLATINUM METALS BY COUNTRY1, 1970-1980
           (Troy Ounces)
                'Includes unwrought and semimanufactured platinum-group metals, unspecified combinations, platinum-group metals  from precious metal
           ores, sweeping, waste, scrap, and materials  not elsewhere specified.
                2January-December 1980.
                3Includes Argentina, Austria, Brazil, Denmark, Finland, Panama, Peru, Surinam, and New Zealand.
                "•Includes Botswana, Brazil,  Costa  Rica, Finland, Ghana, Malawi, Netherlands-Antilles, New Zealand,  Panama,  Turkey,  Venezuela.
                5Includes Brazil, Costa R1ca, El  Salvador, Finland, Ireland, Panama, New Zealand, Uruguay, and Venezuela.
                6Includes Republic of Korea.
                7Includes Costa R1ca, Finland, and Panama.
                8Includes Costa R1ca, Finland, Peru, and Portugal.
                'Includes Argentina, Costa R1ca,  Finland, Hong Kong, and Namibia.
-P»
ro         SOURCE:   U.S. Bureau of Mines, Minerals Yearbook and Mineral Industry Survey, various Issues, Washington,  D.C.:   U.S.  Government Printing
o                  Office.                                    ""~
                                                                                                                                           Charles
                                                                                                                                           River
                                                                                                                                           Associates

-------
Table 4-12


PLATINUM  IMPORTS FOR CONSUMPTION
IN THE UNITED STATES BY COUNTRY,1 1970-1980
(Troy Ounces)
 Country
1970
1971
1972
1973
1974
                                                                          1975
1976
1977
                                                                         1978
                                                                                                                   1979
                                                                                                        19802
Australia
Belgium-Luxembourg
Canada
Colombia
France
Germany, Federal
Republic of
Italy
Japan
Mexico
Netherlands
Norway
Romania
South Africa,
Republic of
Sweden
Switzerland
USSR
United Kingdom
Yugoslavia
Other
Total

1,004
2,957
22,163
--

3,140
—
12,840
37
5,309
4,707
--

97,541
—
1,017
31,078
306,705
—
1,709
490,207

150
2,061
22,027
5

9,260
1,240
18,503
--
150
6,713
--

116,818
__
12
69,241
224,176
-_
3753
470,731

220
2,647
17,220
--

9,253
—
37,316
269
—
8,237
—

107,339
—
--
169,394
260,697
__
3.7951*
616,387

--
278
23,496
953

5,150
--
49,706
112
305
7,234
--

89,294
__
11,079
86,757
399,768
__
—
674,132

1,467
2,128
12,300
—

2,296
--
45,246
15
—
13,601
--

597,569
—
5,669
108,580
313,726
—
1,034s
1,103,631

2,180
1,810
12,600
--

1,976
302
10,011
112
2,252
9,229
--

412,166
—
6,554
33,642
189,589
—
9266
683,349
240
1,181
1,620
--
7,873

14,034
5,409
8,170
—
1,199
8,369
--

693,994
550
3,510
50,430
195,336
482
7.9007
1,000,297

—
3,462
--
2,770

8,273
—
2,498
2,313
3,309
4,801
2,680

675,010
__
8,100
10,432
95,955
_-
3,277
822,880

7,187
6,328
1,252
--

—
20,178
10,019
347
3,451
4,556
—

933,411
--
9,770
20,210
158,175
—
9,331
1,174,215

4,768
1,815
--
--

--
9,231
19,090
--
1,452
10,842
--

1,199,601
--
16,912
25,640
130,369
—
14,491
1,434,211
„
12,187
33,452
119
225

17,085
13,551
15,471
--
13,754
6,611
—

1,059,512
193
24,331
15,892
219,493
643
4.7058
1,436,338
Table continued on following  page.
                                                                                                                              Charles
                                                                                                                              River
                                                                                                                              Associates

-------
            Table  4-12  (Continued)
            PLATINUM  IMPORTS FOR CONSUMPTION
            IN THE UNITED  STATES BY COUNTRY,1 1970-1980
            (Troy  Ounces)
                 1 Includes  unwrought platinum grains,  nuggets,  sponge, and semimanufactured platinum.  Excludes small  amounts of platinum  contained  1n
            sweepings,  waste, scrap, and unspecified combinations.
                 2January-December 1980.
                 3Includes  Denmark.
                 * Includes  Botswana, Brazil, Costa Rica,  Finland, Ghana, Malawi, New Zealand, Panama, and Turkey.
                 5Includes  Republic of Korea.
                 6Includes  Costa Rica and Panama.
                 7Includes  Peru and Portugal.
                 8Includes  Argentina, Costa R1ca, Finland,  Hong Kong, Namibia, Cyprus, and Israel.
_fi
i           SOURCE:  U.S. Bureau of Mines. Minerals Yearbook  and Mineral Industry Survey, various Issues, Washington,  D.C.:   U.S.  Government  Printing
ro                  Office.                      "             ~~
                                                                                                                                              Charles
                                                                                                                                              River
                                                                                                                                              Associates

-------
          Table 4-13


          PALADIUM IMPORTS FOR CONSUMPTION

          IN THE UNITED STATES BY COUNTRY,1  1970-1980

          (Troy Ounces)
i
ro
GO
Country
Austral la
Belgium-Luxembourg
Canada
Colombia
France
Germany, Federal
Republic of
Italy
Japan
Mexico
Netherlands
Norway
South Africa,
Republic of
Sweden
Switzerland
USSR
United Kingdom
Yugoslavia
Other
Total
1970

24
2,275
__
—

4,538
502

3,252
6,969

10,637
_-
28
456,206
287,457
1,929
773,817
1971

8,525
18,279
—
_-

—
__
	
631
5,840

39,163
888
458
332,909
254,643
2.0123
663,348
1972
974
13,512
--
--

6,840
34,310
— -
520
15,124

111,920
._
2,098
523,112
193,819
--
902,229
1973

5,092
—
--

5,718
5,4%
	
42,366
6,950

137,615
__
16,450
668,737
265,881
—
1,154,305
1974

665
18,955
—
1,905

4,882
57,628
8
750
7,390

319,854
_-
1,502
763,343
160,172
i;033*
1,338,087
1975
1,124
2,725
11,739
—
1,979

16,604
8,920
	
2,609
11,555

294,481
4,500
4,447
75,076
117,118
1,225
554,102
1976

2,744
16,773
--
3,363

16,305
260
	
2,436
9,304

444,119
650
8,431
427,102
187,152
3,922
7505
1,123,311
1977
„
6,592
21,840
—
1,000

6,334
15,730
_-
3,387
8,%2

486,639
--
2,330
514,249
81,416
3,198
1,151,677
1978
	
32,337
20,821
—
—

—
6,286
	
4,119
6,454

498,786
__
11,233
503,438
121,134
19,485
1,224,093
1979
	
12,573
23,800
—
—

—
2,534

18,866
14,657

690,439
~-
19,812
602,307
95,002
24,439
1,504,434
19802
__
51,053
25,736
•* "~
200

32,885
4,549
yi
37,759
5,925

648,987
__.
2,003
339,570
164,231
2,604
9956
1,316,588
          Table continued on following page.
                                                                                                                                    Charles
                                                                                                                                    River
                                                                                                                                    Associates

-------
i
r\>
            Table 4-13 (continued)
            PALADIUM IMPORTS FOR  CONSUMPTION
            IN THE UNITED STATES  BY  COUNTRY,1 1970-1980
            (Troy Ounces)
                 'Includes unwrought  and semimanufactured palladium.  Excludes small amounts of palladium contained  In  sweepings, waste, scrap, and
            unspecified combinations.
                 2 January-December  1980.
                 3 Includes Austria  and Denmark.
                 '•Includes Republic of Korea.
                 5 Includes Peru.
                 6 Includes Namibia.

            SOURCE:   U.S.  Bureau  of Mines, Minerals Yearbook  and Mineral Industry Survey, various Issues, Washington, D.C.:   U.S. Government Printing
                     Office.
                                                                                                                                          Charles
                                                                                                                                          River
                                                                                                                                          Associates

-------
Table 4-14

RHODIUM IMPORTS  FOR CONSUMPTION IN
THE UNITED STATES BY COUNTRY,1  1970-1980
(Troy Ounces)
 Country
1970
1971
                     1972
1973
                               1974
                                                    1975
                                                                                    1976
                                                              1977
                                                                                   1978
                                                               1979
Table continued on following  page.
I9602




















-p.
ro
tn
Australia
Belgium-Luxembourg
Canada
Colombia
France
Germany, Federal
Republic of
Italy
Japan
Mexico
Netherlands
Norway
South Africa,
Republic of
Sweden
Switzerland
USSR
United Kingdom
Other
Total




68
3,500
4,160
—

1,465
—
1,320
—
—
--

714
—
--
7,694
22,091
—
41,012




39
--
2,932
—

323
—
--
—
—
--

335
—
--
5,478
25,555
--
34,662




—
428
__
—

27
—
3,213
6
—
--

2,524
—
1
7,139
37,078
3883
50,804




__
	
._
__

3
	
—
113
262
—

2,045
_.
—
34,344
56,444
—
93,211




1,210
467
__
300

1,406
-_
6,567
8
96
__

8,622
__.
--
34,646
45,285
—
98,607




— —
897
__
--

3,028
	
12,831
_.-
300
	

15,810
__
100
37,977
11,025
6 11*
82,029




	
2,272

628

1,548

__
— _
	
-._

26,208
	
1,200
12,699
19,306
263
64,124




__
2,191

2,946

947

353
— —
342
132

33,690
— -
1,288
19,743
18,308
—
79,940




1,044
510



_._

106
53
2,495
77

53,041

750
23,453
20,893
516
102,938




725
1,186



_._

866

1,559
699

65,157

250
17,310
19,610
1,656
109,018




57
230



643
129
128

914
1,007

81,891
64
50
8,482
15,387
1,295s
110,277



                                                                                                                             Charles
                                                                                                                             River
                                                                                                                             Associates

-------
I
ro
en
           Table 4-14  (Continued)


           RHODIUM  IMPORTS FOR CONSUMPTION
           IN  THE UNITED STATES BY COUNTRY,1  1970-1980
           (Troy Ounces)
                Includes unwrought and semimanufactured rhodium.  Excludes small  amounts  of  rhodium contained 1n sweepings, waste, scrap,  and
          unspecified combinations.


                2January-December 1980.


                3Includes Botswana,, Brazil,  Costa  R1ca, Finland, Ghana, Malawi,  Netherlands-Antilles, Panama, Turkey, and Venezuela.


                ''Includes Finland.


                5Includes Finland and Namibia.





          SOURCE:  U.S. Bureau of Mines,  Minerals Yearbook and Mineral Industry Survey,  various  Issues,  Washington, D.C.:  U.S. Government Printing
                   Office.
                                                                                                                                        Charles
                                                                                                                                        River
                                                                                                                                        Associates

-------
ro
-g
           Table 4-15


           TOTAL PRIMARY AND SECONDARY PLATINUM-GROUP

           METALS REFINED  IN THE UNITED STATES,  1970-1979

           (Troy Ounces)
                                 1970        1971         1972        1973        1974        1975        1976         1977        1978        1979
Primary Metal :
Nontoll Refined
Toll Refined
Total

19,822
270,335
290,157

21,184
233,850
255,034

15,380
84,219
99,599

19,916
38,566
58,482

13,234
20,107
33,341

16,571
17,174
33,745

7,101
10,232
17,333

5,199
1,083
6,282

8,303
1,354
9,657

8,392
476
8,868
           Secondary Metal:


             Nontoll Refined   350,176     278,175     255,641     265,901     325,216     270,101      215,355      195,219     257,191     309,022


             Toll Refined    1.451.535   1.218.988    1.277.404   1,000.623   1.067.915   1.158.294      859,432    1.003.940    1.021,960   1.090.202


             Total           1,801,711   1,497,163    1,533,045   1,266,524   1,393,131   1,428,395    1,074,787    1,199,159    1,279,151   1,399,224
          SOURCE:  U.S. Bureau of Mines, Minerals  Yearbook, various Issues, Washington, D.C.:   U.S.  Government  Printing  Office.
                                                                                                                                         Charles
                                                                                                                                         River
                                                                                                                                         Associates

-------
            Table 4-16
            PRIMARY AND  SECONDARY  PLATINUM
            REFINED IN THE  UNITED  STATES, 1970-1979
            (Troy Ounces)
                                   1970
1971
1972
                                                                     1973
                                   1974
                                                                                             1975
                                               1976
                                                                                                                     1977
                                                                                  1978
                                                                                                                                            1979
Primary Metal :
Nontoll Refined
Toll Refined
Total

8,036
183,264
191,300

10,198
156,599
166,797

3,708
54,773
58,481

5,560
32,883
38,443

4,103
16,293
20,396

5,292
14,619
19,911

2,748
8,676
11,424

831
466
1,297

1,081
177
1,258

1,980
56
2,036
            Secondary Metal:
               Nontoll Refined   118,298     103,429
               Toll Refined      896.472     625,649
               Total            1,014,770     729,078
           75,942      94,884
          787.697     581,005
          863,639     675,889
                       95,999     103,623
                      654,156     635,148
                      750,155     738,771
 64,901      50,838       75,585      75,038
494.069     620.848     630.961     585.932
558.970     671,686     706,546     660,970
-^
CO
            SOURCE:  U.S. Bureau of Mines, Minerals Yearbook, various Issues,  Washington, D.C.:  U.S. Government Printing Office.
                                                                                                                                           Charles
                                                                                                                                           River
                                                                                                                                           Associates

-------
Table 4-17
PRIMARY AND SECONDARY  PALLADIUM
REFINED IN THE  UNITED  STATES, 1970-1979
(Troy Ounces)
                       1970
1971
1972
1973
1974
1975
1976
1977
1978
                                                                                                                                1979
Primary Metal :
Nontoll Refined
Toll Refined
Total

10,322
74,953
85,275

10,237
66,467
76,704

10,836
23,752
34,588

13,121
3,972
17,093

8,634
2,784
11,418

10,968
2,002
12,970

4,025
1,063
5,088

4,300
610
4.910

7,222
1.177
8,399

6.412
420
6,832
Secondary Metal:
   Nontoll Refined    208,555
   Toll  Refined       494,758
   Total              703,313
          162,718
          431.248
          593,966
          150,019
          373.396
          523.415
                      149,552     134,747
                      437.809     311.000
                      587,361     445,747
                                 134,086     166,371     220,639
                                 327.450     344,022     446.189
                                 461,536     510,393     666,828
SOURCE:   U.S.  Bureau  of Mines, Minerals Yearbook,  various  Issues, Washington, D.C.:   U.S.  Government  Printing Office.
                                                                                                                                  Charles
                                                                                                                                  River
                                                                                                                                  Associates

-------
I
oo
o
            Table 4-18


            PRIMARY AND SECONDARY RHODIUM
            REFINED IN THE UNITED STATES,  1970-1979
            (Troy Ounces)


                                  1970        1971        1972        1973        1974        1975        1976        1977        1978        1979


            Primary Metal:


               Nontoll Refined


               Toll Refined


               Total





            Secondary Metal:


               Nontoll Refined    13,394


               Toll Refined       47.861


               Total              61,255
64
8,885
8,949
83
8,118
8,201
62
3,354
3,416
88
381
469
38
185
223
28
164
192
35
95
130
6
3
9
8,837
43,173
52,010
11,390
44,065
55,455
11,561
36,865
48,426
11,127
36,1%
47,323
13,683
49,063
62,746
8,058
34,035
42,093
5,011
42,178
47,189
8,266
35,914
44,180
7,964
38,875
46,839
            SOURCE:  U.S. Bureau of Mines, Minerals Yearbook, various  Issues, Washington, D.C.:  U.S.  Government  Printing Office.
                                                                                                                                         Charles
                                                                                                                                         River
                                                                                                                                         Associates

-------
I
CO
                    r
                    c
                    e
                    n
                    t
                           108
                            88~
                            68-
48-
                            28-
                             8
                                  Figure 4—2

                                  Percentage of Total U.S. Platinum-Group Metal Imports from Various Countries

                                        CPrepared By:  Charles River  Associates)
                                  II      1^    I      I      I      I      I      I     I
                                 1978  1971   1972  1973  1974  1975  1976  1977  1978  1979
                               South AfrIca
                               UnI ted KIngdom
                               USSR
                                                   Other
                           years
                    SOURCE" U.S.  Bureau of Mines,  Minerals  Yearbook,
                    Mineral Industry Surveys, various  Issues,  Washington,  D.C.
                                                                                              Charles
                                                                                              River
                                                                                              Associates

-------
I
00
IN3
r
c
e
n
t
                         180
                          88-
                          60-
                          40-
                          20-
                           0
                                       Figure 4—3

                                       Percentage of Total U.S. Platinum Imports from Various Countries

                                      CPrepared By:  Charles River Associates)
                                \      I      I      I      I      I      I     I      I      I
                               1970  1971   1972  1973  1974  1975  1976  1977  1978  1979
                              South Africa
                              UnIt«d KIngdom
                              USSR
                                                           Other
                                   years
                  SOURCE" U.S. Bureau of Mines,  Mineral* Yearbook,
                  Mineral Industry Surveys, various Issues,  Washington, D.C.
                                                                                              Charles
                                                                                              River
                                                                                              Associates

-------
-Pi
I
CO
oo
                  r
                  c
                  •
                  n
                  t
                                       Figure 4—4

                                       Percentage of Total U.S. Palladium Imports from Various Countries

                                      CPrepared By  Charles River  Associates)
                         108
                          88-
                          68-
48-
                          28-
                           8
                               1978  1971   1972  1973  1974  1975  1976  1977  1978  1979
                              South AFrIca
                              UnIt«d KIngdom
                              USSR
                                                  Othere
                  SOURCEt U.S.  Bureau of Mines,  Minerals  Yearbook^
                  Mineral Industry Surveys,  various  Issues,  Washington,  D.C.
                                                                                             Charles
                                                                                             River
                                                                                             Associates

-------
I
OJ
r
c
•
n
t
                     188
                      88"
                      68-
                      48-
                      28-
                       8
                                    Figure 4—5

                                    Percentage of Total U.S. Rhodium Imports from Various Countries

                                  CPrepared  By:  Charles River  A««ocfates)
                                                                      I     I      I
                           1978  1971   1972  1973  1974  1975  1976  1977  1978  1979
                          South Africa
                          Uni ted K i ngdom
                          USSR
                                                           Other
                                   years
              SOURCE*  U.S.  Bureau of Mines,  Minerals  Yearbook,
              Mineral  Industry Surveys,  various  issues,  Washington,  D.C.
                                                                                          Charles
                                                                                          River
                                                                                          Associates

-------
                                                                   Charles
                                                                   River
                                                                   Associates


Refining of secondary metal  for all  PGMs  in  the  United States throughout the
1970s fluctuated within the  1.0 to 1.5  million ounce  range, except  for 1.8
million ounces in 1970 (Table 4-15).   Individually, platinum, palladium, and
rhodium secondary refining followed this  fluctuating  behavior throughout the
last decade, with no readily discernable  upward  or downward trend (Tables
4-16 through 4-18).

Table 4-19 identifies trends in U.S.  exports of  PGMs  throughout  the 1970s.
The data reveal that exports to Japan,  Canada, and the United Kingdom have
increased significantly, while exports  to West Germany have trended
downward.
DEMAND

Demand for PGMs fluctuates considerably,  due to the  overall  level of economic
activity, technological  changes,  and government policy  actions  (such as auto
emission and lead-free gasoline mandates).   This section  presents a brief
discussion of PGM industrial  uses.  In many applications  PGMs operate like
capital goods, the purchase of which often  can be deferred.  This fact
exacerbates fluctuations in demand due to changes in economic activity  (a
phenomenon economists refer to as the "acceleration  principle"), but also
implies consumption can be forgone during supply disruptions.


CATALYTIC USES

A catalyst is a substance that initiates  or speeds up a chemical reaction,
while not being consumed itself.   Platinum-group metals are  excellent
catalysts that are cost effective for many  uses, even compared  to other much
less expensive metals such as nickel.  A  case in point  is use of PGMs in
catalytic converters.  Currently, no alternative base-metal  catalytic
converter appears economical, as discussed  at length in a later chapter.
Catalytic uses of various types comprise  by far the  majority of U.S. PGM
demand.  Major catalytic applications are discussed  below by consuming
industry.


PETROLEUM INDUSTRY.  Platinum metals are  used as catalysts in the refining  of
petroleum products in three processes: catalytic reforming, hydrocracking,
and isomerization.  Reforming is generally  the largest  use of the three, and
it has recently grown considerably.  The  platinum catalyst eventually fails
due to high temperatures and contamination; it is then  recycled, with about a
3 percent loss rate.
                                    4-35

-------
Table 4-19


U.S.  EXPORTS OF PLATINUM-GROUP METALS,  BY COUNTRY,  1970-1980
(Troy Ounces)
                      1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
Table continued on following page.
I9601
Argentina
Australia
Belgium-Luxembourg
Brazil
Canada
Colombia
Finland
France
Germany, Federal
Republic of
Hong Kong
Italy
Japan
Mexico
Netherlands
j* Norway
co South Africa,
°> Republic of
Sweden
Switzerland
Taiwan
United Kingdom
USSR
Other
Total
673
2,154
24,818
7,644
11,299
—
—
32,842

155,222
157
16,770
70,811
4,158
12,076
--

—
—
8,415
--
64,212
—
2,515
413,766
1,055
1,932
40,204
1,490
18,506
—
__
35,084

90,616
1,015
10,934
94,265
1,926
15,831
--

1,585
--
4,437
—
80,201
—
5,529
404,610
126
2,990
45,117
5,289
10,069
--
-_
4,386

120,685
1,235
19,221
254,460
2,967
9,715
--

1,106
—
4,964
--
52.105
—
4,559
538,994
„
8,906
69,594
6,057
20.358
—
--
--

103,880
--
--
310,940
5,327
19,669
--

—
—
7,819
--
27,901
—
47.0752
627,526
„
—
49,864
--
30,666
--
-_
9,263

181,553
--
--
247,432
—
24,453
--

25,213
--
95,053
11,267
103,913
__
57.0773
835,754
„
1,780
38,812
6,351
39,607
--
—
10,284

135,754
—
20,387
168,774
--
28,389
--

55,722
--
43,238
—
97,656
—
19.1313
665,885
_ _
3,489
53.342
--
47,854
--
—
10,026

74,716
—
7,007
118,857
2,631
3,097
--

11,190
—
32,785
5,985
132,251
1,603
7, 574"
512,407
__
2,255
54,406
1,176
33,263
--
--
6,279

66,821
--
1,718
110,547
3,431
3,311
--

1,053
—
38,875
1,979
90,303
—
11.1145
426,631
1,391
1,109
36,370
10,147
51,606
11,505
6,140
11,609

73,533
2,906
1,721
225,222
58,099
7,838
3,897

2,307
15,951
26, 963
1,643
146,197
--
6,393
702,547
640
10,008
37,841
6,816
62,396
1,802
2,574
18,753

98,876
1,487
8,774
328,889
55,004
10,610
3,180

6,320
17,739
48,710
116
154,284
—
22,5996
899,598
936
799
32,283
4,634
72,399
—
4,684
li.838

111,175
984
6,393
237,963
6,144
11,786
3,967

3,252
4,249
67,194
—
173,741
—
10.5437
764,964
                                                                                                                             Charles
                                                                                                                             River
                                                                                                                             Associates

-------
I
OJ
         Table  4-19  (Continued)
         U.S. EXPORTS OF PLATINUM-GROUP METALS,  BY COUNTRY, 1970-1980
         (Troy  Ounces)

             January-December 1980.
             2Includes People's Republic of China,  Israel, and Spain.
             3Includes Israel
             *• Includes Venezuela.
             5Includes Ireland, Republic of Korea,  and Peru.
             6Inc1udes Greece, Republic of Korea, Singapore, Spain, and Venezuela.
             'Includes Greece, Ireland, Republic of Korea, Singapore, and Venezuela.

         SOURCE:  U.S. Bureau of Mines. Minerals Yearbook and Mineral Industry Survey,  various  Issues,  Washington, D.C.:  U.S. Government Printing
                 Office.
                                                                                                                                             Charles
                                                                                                                                             River
                                                                                                                                             Associates

-------
                                                                  Charles
                                                                  River
                                                                  Associates
 Catalytic  reforming  is  the  most  efficient  way  known  to  raise  the  octane
 rating  of  gasoline without  adding  lead.  The  increasing use of lead-free
 gasoline in  the United  States  insures  that reforming will  continue  to  be a
 major use  of platinum metals,  though sales to  the  petroleum industry for
 reforming  may be considerably  lower during the 1980s, after initial  loadings
 of  required  increased reforming  capacity have  been completed  (the economists'
 "accleration principle"  in  operation).  Modern technology  in  newer  plants,
 however, can reduce  the  platinum intensity of  petroleum refining.


 CHEMICAL INDUSTRY.   The  chemical industry  is a major consumer of  platinum,
 palladium, and rhodium  for  catalysis and pollution control.   Nitric  acid
 production is a major component  of chemical industry consumption  of
 platinum-group metals,  accounting  for  about 30 percent  of  industry  use of
 PGMs in an average year.  Processing losses for PGMs used  in  nitric  acid
 production are much  higher  relative to  initial  loadings than  is the  case for
 most other chemical  uses.


 AUTOMOTIVE INDUSTRY.  The automotive industry  is,  in the early 1980s,  by far
 the largest  user of  PGMs in the  United  States,  predominantly  for  use in
 catalytic converters.   New, highly sophisticated emission-control  systems
 also make  use of platinum in the coating of an  exhaust  gas oxygen sensor.
 The exact  amount of  platinum used  in each  sensor is  proprietary,  but the
 amount  is much less  than is found  in converters.

 Previously,  oxidation catalysts were used  to reduce  emissions, but with
 tighter standards for nitrogen oxide emissions, most U.S. car makers have
 switched to  the new  three-way catalyst.  The oxidation  catalyst uses
 approximately 0.05 troy ounces of platinum and  palladium in a roughly  2 to 1
 ratio.  The  three-way catalyst uses platinum,  palladium and rhodium.


 NONCATALYTIC USES
ELECTRICAL END USES.  Platinum, palladium, and rhodium have high electrical
conductivity and hence are used in the electrical equipment industry.
Platinum is used in thermocouples, electrical contacts, and electroplating of
printed circuits, while palladium is used for relays and metal contacts  in
telecommunications equipment.

The Bell System uses large amounts of palladium in its switching equipment.
Most contacts are now a silver-palladium alloy rather than all palladium.
Conversion from mechanical to electronic switching by the Bell System is
decreasing its demand for palladium.  By the mid-1980s it is anticipated that
the Bell System will be able to meet all its internal needs for palladium
from its own recycling of obsolete equipment.
                                   4-38

-------
                                                                   Charles
                                                                   River
                                                                   Associates


GLASS PRODUCTION.   Platinum group  metals  are  used in the manufacture of glass
and glass fiber because of high  corrosion resistance, ability to withstand
high temperatures,  and compatible  expansion coefficients.  The glass industry
consumes relatively minor quantities  of platinum and palladium, but in 1979
it had the third largest reported  sales figures for rhodium of any  industry
in the United States.


MEDICAL, DENTAL, AND JEWELRY USES.  Very  little rhodium is consumed in dental
and medical uses,  but substantial  amounts of  platinum and palladium are used.
In 1979, reported sales of palladium  were second highest in the dental and
medical categories.  Application of PGMs  in dental work include dental
crowns, caps, and bridges.  Medical uses  for  PGMs are quantitatively minor.
They are used in cancer chemotherapy, heart pacemakers, and hypodermic needle
tubing.

On the whole, U.S.  consumers have  not been as much enamored with platinum
jewelry as with gold and silver  jewelry,  and  U.S. reported sales to this
industrial category are relatively low.  Japanese jewelry consumers have
historically favored platinum jewelry, but the relative demand for  platinum
in jewelry apparently may be fading there.


CONSUMPTION TRENDS IN THE 1970S

Table 4-20 gives a percentage breakdown of world platinum-group metal
consumption by end use.  The figures  are  estimated by private publications
through 1977, and are probably not as reliable as similar figures for the
United States alone.  Two readily  discernible worldwide trends from
Table 4-20 are that electrical use has declined relatively and that
automotive use began and climbed rapidly  in the mid-1970s.  Relative world
shares for petroleum,  dental and medical, and glass have remained fairly
constant, while chemical and jewelry  shares were erratic from 1972 to 1977.

Tables 4-21, 4-22,  and 4-23 and  Figures 4-6,  4-7, and 4-8 provide an overview
of U.S. consumption of platinum, palladium, and rhodium.  As discussed in the
previous section,  the U.S. Bureau  of  Mines PGM consumption data are not
comprehensive, but are the sum of  sales reported to consuming industries.
Total "apparent consumption" figures, based on net import and production
data, are consistently above reported sales,  but are not reported on an
industry-specific basis.  Reported sales, despite being less than
comprehensive, still can be used reliably to  examine changing patterns of
interindustry PGM consumption.

As mentioned in the previous section, sales data reflect not only usage of
PGMs directly or indirectly in production, but also possible changes in shelf
stocks.  For example,  if the chemical industry embarked on a plan to increase
its PGM shelf stocks in a particular  year, reported sales data might look
                                     4-39

-------
Table 4-20

WORLD ESTIMATED CONSUMPTION OF PLATINUM
GROUP METALS BY END USE,  1972-1977
(Percent)
                                                                   Charles
                                                                   River
                                                                   Associates
End Use
1972
1973
1974
1975
1976
1977
Electrical
Chemical
Jewelry
Automotive
Petroleum
Dental and Medical
Glass
Others
Total
33.8
27.5
18.4
-
5.7
6.4
3.8
4.4
100.0
33.5
25.2
19.8
-
5.8
6.7
3.6
5.4
100.0
26.5
23.0
19.4
11.2
6.4
6.7
4.0
2.8
100.0
19.1
22.4
27.4
11.3
5.9
6.5
3.0
4.4
100.0
20.8
21.4
20.8
16.7
4.3
8.4
2.7
4.9
100.0
22.6
23.4
20.5
12.7
5.1
6.3
3.7
5.5
100.0
SOURCE:  J. Aron and Roskill estimates from Roskill  Information  Services,
         Ltd., 1979, The Economics of Platinum Group Metals.  London.
                                       4-40

-------
Table  4-21


PLATINUM SOLD TO CONSUMING INDUSTRIES IN
THE UNITED STATES,  BY END USE,1  1970-1980
(Troy  Ounces)
Automotive
Chemical
Dental and Medical
Electrical
Glass
Jewelry and
Decorative
Petroleum
•£» Miscellaneous
i
^ Total
1970
—
148,289
18,302
103,318
46,687
29,203
202,015
18,555
566,369
1971
—
135,112
23,097
51,940
40,703
18,577
141,800
19,859
431,088
'Includes primary and nontoll -refined
2Excludes companies
SOURCE: U.S. Bureau
reporting
of Mines
annually
1972

225,895
30,462
92,381
26,970
20,655
98,847
50,089
545,299
1973
—
238,974
27,887
117,352
72,543
22.433
123,649
55,695
658,533
1974
350,000
215,663
25,513
98,608
74,398
22,968
139,519
17,020
943,689
1975
273,000
148,813
17,097
73,624
33,813
22,900
107,988
21,318
698,553
1976
480,965
83,560
26,858
89,319
41,683
23,371
59,103
46,246
851,105
1977
354,338
84,414
27,083
90,217
59,995
34,650
74,772
64,350
789,819
1978
597 , 538
149,696
44,139
106,422
98,094
25,751
108,365
66,336
1,196,341
1979
803,229
98,600
27,053
115,775
88,594
27,712
170,013
77,949
1,408,925 1
19802
517,143
116,609
26,191
142,442
51,843
38,360
141,197
56,372
,090,157
secondary metals.
•
, Minerals Yearbook and

Mineral


Industry Surveys, various

issues,

Washington

, D.C.: U.

S. Government

Prlntlm
        Office.
                                                                                                                          Charles
                                                                                                                          River
                                                                                                                          Associates

-------
Table 4-22


PALLADIUM SOLD TO CONSUMING INDUSTRIES IN
THE UNITED STATES,  BY END USE,1  1970-1980
(Troy Ounces)
Automotive
Chemical
Dental and Medical
Electrical
Glass
Jewelry and
Decorative
.p. Petroleum
i
PO Miscellaneous
Total
1970
—
184,618
47,583
429,032
21,147
17,329
15,494
24,140
739,343
1971
--
218,651
61,594
431,505
237
18,752
2,916
26,451
760,106
•includes primary and nontoll -refined
2Excludes companies
SOURCE: U.S. Bureau
reporting
of Mines,
annually
1972
—
292,710
94,274
425,081
2,250
19,375
14,499
27,835
876,024 1
1973
—
259,959
135,060
524,056
1,439
23,052
3,761
65,157
,012,484
1974
150,000
163,205
124,074
390,237
9,549
21,701
14,877
12,420
886,063
1975
97,000
142,975
114,970
132,247
17,633
23,026
1,755
11,942
541,548
1976
194,496
128,229
139,279
152,312
2,989
5,700
7,291
26,766
657,062
1977
125,010
161,234
112,473
223,748
907
15,567
8,507
53,023
700,469
1978
198,809
146,352
206,312
286,574
2,757
12,570
18,909
45,645
917,928
1979
222,156
199,743
243,627
392,372
1,729
11,766
24,588
36,640
1,132,621
19802
176,518
116,515
174,8321
289,797
1,121
12,874
21,391
21,320
814,368
secondary metals.
•
Minerals Yearbook and

Mineral

Industry Surve

ys, various

Issues,

Washington,

D.C.: U.

S. Government

Printing
        Office.
                                                                                                                        Charles
                                                                                                                        River
                                                                                                                        Associates

-------
Table  4-23


RHODIUM SOLD TO CONSUMING INDUSTRIES IN
THE UNITED STATES, BY END USE,1 1970-1980

(Troy  Ounces)
Automotive
Chemical
Dental and Medical
Electrical
Glass
Jewelry and
Decorative
Petroleum
-P»
4=» Miscellaneous
to
Total
1970
--
26,445
51
9,056
7,138
5,343
59
805
48,897
1971
--
14,910
31
9,084
3,362
5,419
176
1,384
34,366
'Includes primary and nontoll -refined
2Excludes companies
SOURCE: U.S. Bureau
reporting
of Mines
annually
1972
--
15,358
48
7,867
13,923
6,593
149
2,157
46,095
1973
—
23,772
297
13,187
16,689
12,526
3,057
3,987
73,515
1974
—
23,328
373
15,538
7,464
10,460
1,239
3,200
61,602
1975
--
15,440
41
8,252
4,471
4,932
114
3,598
36,848
1976
391
19,225
75
9,062
3,828
5,170
1
3,123
40,875
1977
871
20,245
275
10,758
13,986
5,011
4,070
55,216
1978
2,939
19,397
232
14,329
16,605
9,950
281
5,907
69,640
1979
26,136
11,684
45
16,923
15,376
7,458
1,223
4,625
83,470
19802
37,012
5,174
42
6,646
8,420
9,588
650
1,665
63,197
secondary metals.
•
, Minerals Yearbook and

Mineral

Industry Survej

/s, various

Issues,

Washington,

D.C.: U.S.

Government

Printing
        Office.
                                                                                                                      Charles
                                                                                                                       River
                                                                                                                      Associates

-------
t
r
o
o
u
n
o
   1508008
   1200000-
    900000 ~
    600000 ~
    300000 ~
         0
                     Figure 4—6

                     Platinum Sold to U.S. Consuming Industries

                    CPrepared  By:  Charlie River Associates)
                                           I     I      I      I      I
             1970  1971  1972  1973  1974  1975  1976  1977  1978  1979
            Automot i ve
            Chemi caI
            Dental  & Medical
yean
Electri cal
 Petroleum
     Other
                               IIIIIHMII
SOURCE' U.S.  Bureau of Mines,  Mineral* Yearbook,
Mineral Industry Surveys, various issues, Washington, D.C.
                                                                           Charles
                                                                           River
                                                                           Associates

-------
-pa
I
                                   Figure 4—7

                                   Palladium Sold to U.S. Consuming Industries

                                   CPrepared By:  Charles River Associates}

1000000-
t
r
o
V 750808 ~
0
u
0 500000 ~
e
8
250000 -







1








mil

















u








1








111








11

















[








1








1

















1








1

















U








1








11








1


















1J








1








1








11
















1








1








11

















i








i








i








i

















i
i







i
i















1








i








m

















                            1970  1971   1972  1973  1974  1975  1976  1977  1978  1979

MINI

MM

                           AutomotIve
                           ChemIcaI
                           Dental  & Medical
year i
Electr i cal
 Petroleum
     Other
               SOURCE•  U.S.  Bureau of Mines,  Mineral* Yearbook,
               Mineral  Industry Surveys, various Issues, Wash Inoton, D.C.
                                                                                         Charles
                                                                                         River
                                                                                         Associates

-------
I
-C.
CTl
               t
               r
               o
               y

               o
               u
               n
               c
                                    Figure 4—8

                                    Rhodium Sold to U.S. Consuming Industries

                                   (Prepared By: Charles River Associates}
67288 ~
58488 -
33688 ~
1 6888 ~


LU


••


III
uuu






mi
1
Uil T1



1 I
" ' T


\



u









y
H


•


•ev
•
•H



1
•


T"
MH




1
                            1978  1971   1972  1973  1974  1975  1976  1977  1978  1979

MMI

linn

                           AutomotIve
                           Chem i caI
                           Dental  & Medical
years
Electrical
 Petroleum
     Other
               SOURCEi  U.S. Bureau of Mines,  Minerals Yearbook^
               Mineral  Industry Surveys,  various  Issues,  Washington, D.C.
                                                                                           Charles
                                                                                           River
                                                                                           Associates

-------
                                                                   Charles
                                                                   River
                                                                   Associates
abnormally large relative to changes  in  the  production of  the chemical
industry.

Figure 4-6 shows the upward trend in  U.S.  reported  platinum  consumption
throughout the 1970s.  Figures 4-7 and 4-8 indicate that U.S. reported
palladium and rhodium consumption has been more  erratic; stock  adjustments
may account for part of this.

From Table 4-21, it can be seen that  the industries to whom  the largest  sales
of platinum were reported in 1979 are automotive, petroleum, electrical, and
chemical.  For palladium, reported sales were  highest in 1978 to the
electrical, dental  and medical, automotive,  and  chemical industries  (Table
4-22).  For rhodium, reported sales were highest in 1979 to  the automotive,
electrical, glass,  and chemical industries (Table 4-23).


PRICES

The market for PGMs has a two-tier pricing structure.  In  broad terms, there
is one price which  primary producers  charge their regular  contractual
customers, called the producer price, and there  is  another price that others
must pay on the spot market, called the  dealer price.

As Figures 4-9, 4-10, and 4-11 demonstrate,  producer and dealer prices for
platinum, palladium, and rhodium have closely  tracked one  another throughout
the 1970s.  This is because the economic forces  of  supply  and demand, which
typically first influence the spot price,  are  also  taken into account during
producer negotiations with the contractual consumers.  Occasionally,
short-term forces cause the spot price to deviate substantially from the
producer price.

Figures 4-12, 4-13, and 4-14 plot 1980 monthly producer and  dealer prices for
platinum, palladium, and rhodium.  Dealer prices are more  variable,  reacting
constantly to market events that will only slowly influence  contractual
arrangements.  Especially noticeable  is  the large excess of  dealer prices
over producer prices for platinum in  early 1980, which many  industry
observers attributed to speculative activity.   The  producers of platinum-
group metals have kept prices to traditional industrial consumers at more
stable levels so that long-run consumption and profitability for the platinum
producers are not damaged for short-run  gains.
                                     4-47

-------
-pi
I
-pi
00
t
r
o

y

o
u
n
c
                    Figure 4—9


                    Averooe Annual Platinum Producer & Dealer Price*,  1970-1980


                                  CPrepored By: Char lee River Associates}
                    1000
                     800-
                     600-
400-
                     200-
                       0
                            \     I     I      I     I     I     I     I     I      I     I
                           1970       1972      1974       1976       1978      1980
                                1971       1973       1975       1977      1979
                          Producer PrIce
                          Dealer Price
                                  years
              SOURCE'  U.S.  Bureau of Mine*, Minerals  Yearbook.

              Mineral  Industry Surveys, various  Issues,  Washington, D.C.
                                                                                     Charles
                                                                                     River
                                                                                     Associates

-------
t
r
o
y

o
u
n
c
     Figure 4—10

     Average Annual  Palladium Producer & Dealer Prices, 1970-1988

                    (Prepared By:  Charles River Associates)
       250
       200
150-
100-
        50-
         0
              I     I     I     I     I     I      I     I     I     I     I
            1970       1972       1974      1976      1978       I960
                  1971       1973       1975      1977       1979
           Producer PrIce
           Dealer Price
                           years
SOURCE* U.S. Bureau of Mines,  Mineral* Yearbook,
Mineral Industry Surveys,  various Issues, Washlnaton, D.C.
                                                                        Charles
                                                                        River
                                                                        Assonintes

-------
seppossv
    ' jsAia
   S3|JDl|O
                      "O'Q
              8861
                                         SZ.61
                  9Z61
               I     I
                                                      I     I
 •oijj ~i»onpojj

      U61
ZL&\        0£61
       I	I
                                                   :Ag
                                                                          8
                                                                         -882
                                                                         -009
                                                                         L008
                                                                          0001
0881-8^61  '
                                                               |onuuy
                                                                                   o
                                                                                   u
                                                                                   n
                                                                                   o

                                                                                   A
                                                                                   o
                                                                 j
                                                                 e
                                                                 d

                                                                 $
                                                                                o

                                                                                ^t-

-------
en
$

p
•
r

t
r
o
y

o
u
n
c
                   1000
                        Figure 4-12

                        Monthly Plallnum Producer and Dealer  Price*,  1880

                                 CPrepared Bys Charles River  Associates)
                    800-
                    600-
                    400-
                    200-^
                      0
                           I     I    I     I     I     I    I
                         Jan  Feb  Mar Apr  May  Jun  Jul
                                               I     1    I     I     I
                                             Aug  Sep  Oct  Nov Dec
                         Producer PrIce
                         Dealer Price
                                   1980
             SOURCE'  U.S.  Bureau of Mines, Mineral* Yearbook,
             Mineral  Industry Surveys, various  Issues,  Wash I noton, D.C.
                                                                                       Charles
                                                                                       River
                                                                                       Associates

-------
en
ro
t

r
o

Y


o

u
n

c
                        Figure 4—13


                        Monthly Pal I adturn Producer and Dealer prices,  1980

                                 CPrepored By: Chart** River  Associates)
                    580
                    400-
                    300-
                    200-
                    100-
                      0
                          n     i    i    i     i     i    r    \     r    i    i     r
                         Jan  Feb  Mar  Apr  May  Jun  Jul   Aua S*p  Oct  Nov  Dec
                         Producer PrIce

                         DeaIer Pr i ce
                                   1980
              SOURCE'  U.S. Bureau of Mines, Minerals Yearbook.,

              Mineral  Industry Surveys, various  Issues,  Washington, D.C.
                                                                                        Charles
                                                                                        River
                                                                                        Associates

-------
01
CO
t
r
o
y

o
u
n
c
                        Figure 4—14

                        Monthly Rhodium Producer and Dealer  Prices,  1980

                                CPrepared By:  Charles  River  Associates}
                   1088
                   900-
                   800-
700-
                   600-
                   508
                          \     I     I    I     F    \     I    I     I     I     I     I
                         Jan Feb  Mar  Apr  May  Jun  Jul  Aug  Sep Oct  Nov  Dec
                        Producer PrIce
                        DeaIer PrIce
                                   1980
             SOURCEi  U.S. Bureau of Mines,  Minerals Yearbook,
             Mineral  Industry Surveys, various Issues, Washington,  D.C.
                                                                                      Charles
                                                                                      River
                                                                                      Associates

-------
                                                                   Charles
                                                                   River
                                                                   Associates
                              PRIMARY PRODUCERS
This section describes capacity,  production,  and  supply elasticities of
countries mining platinum-group metals.   Current  conditions are analyzed,
focusing on how each country could respond to increased PGM demand, whether
due to increased consumption for  vehicular emissions control or some other
cause.
SOUTH AFRICA
CURRENT PRODUCTION

As Tables 4-8, 4-9, and 4-10 indicate,  South Africa  produces most  of  the
world's primary platinum and rhodium, and a  significant  amount of  palladium.
More recent 1979 platinum production figures for the "big  three" South
African producers are:  Rustenburg Platinum  Mines, 1.3 million troy ounces;
Impala Platinum Mines, 900,000 troy ounces;  and  Western  Platinum,  80,000 troy
ounces. Total  South African 1979 production  of palladium was 906,000  troy
ounces; 1979 rhodium,production was 108,000  troy ounces.

South African PGMs come predominantly from the Merensky  Reef in the Bushveld
Complex.  About two-thirds of the PGMs  in the Merensky Reef by weight is
platinum.  The ore grade is approximately 8.1 grams  of PGMs per metric ton of
ore.  The Merensky Reef is one of the few deposits in the  world, and  by far
the largest, where PGMs are the principle product.   In the Soviet  Union and
Canada, PGMs are byproducts from the mining  of other metals, mostly nickel.
This fact is important, because the supply of PGMs from  byproduct  operations
responds more strongly to changes in production  of the parent metal than to
changes in PGM prices.  Byproduct supply is  discussed further below.

Current production technologies in South Africa  and  elsewhere involve
significant time lags  between mining and delivery of refined metal.   At
present it requires about seven months  after mining  to deliver refined
platinum or palladium, and about 16 months for other PGMs.  In part,  these
lags are due to ores,  concentrates, and slimes being shipped to other
countries for refining.  In 1978, the Bureau of  Mines estimated the cost of
refining platinum and  palladium at $7 to $8  per  troy ounce, and two to three
times this for other PGMs (Mineral  Commodity Profiles, 1978).

New production technologies are being developed  which should reduce these
production lags.  Texasgulf, Inc. is testing a new plasma  smelting process to
obtain PGMs from concentrates containing high amounts of chromite.  (See
Engineering and Mining Journal, 1979.)   The  South African  National Institute
Tor Metallurgy has already developed a  new PGM extraction  process  from high
chromite ores which Western Platinum will  use to mine the  UG2 seam, a high
chromite deposit below the Merensky Reef.
                                     4-54

-------
                                                                   Charles
                                                                   River
                                                                   Associates


The UG2 seam's rhodium concentration is almost three times that of the
Merensky Reef.  One source (Buchanan, 1980) puts the rhodium proportion
(percent of PGMs accounted for by rhodium) at 8 percent for UG2 versus
3 percent for Merensky.  Consequently, both the platinum/rhodium and
palladium/rhodium ratios are lower for UG2 than for Merensky.  The
platinum/rhodium ratio is 5.25 for UG2 and 19.66 for Merensky,  while  the
palladium/rhodium ratio is 4.375 for UG2 and 8.33 for Merensky.

The high rhodium concentration in UG2 is important because three-way
catalytic converters in 1980 used platinum and rhodium in roughly a 10 to 1
ratio, and there is concern about this ratio being much higher  than the
current mining ratios from Merensky of 19 to 1.  However, Ford's three-way
catalyst reportedly utilizes rhodium, palladium, and platinum in proportions
"not very far off" from those in which they are currently mined.

As the UG2 deposit comes on stream, platinum/rhodium mining ratios could
approach the 10 to 1 usage ratio.  CRA communications with mining companies
in South Africa indicate that Impala and Rustenburg Mines, the  two largest
producers, are not planning to mine the UG2 seam in the near future.
However, Western Platinum, due to its smaller holdings in the Merensky Reef,
is planning to recover about 110,000 troy ounces of PGMs annually from UG2.
This quantity, however, is small when compared to total 1979 South African
PGM production of 2.28 million ounces by Rustenburg, Impala, and Western.

Despite the fact that the UG2 deposit lies only 100 to 350 meters under the
Merensky Reef, and the two could be simultaneously mined from a vertical
shaft, Rustenburg and Impala have indicated little interest in  developing
UG2.  Impala personnel have indicated, however, that should auto emission
standards become tighter and drive up the relative demand for rhodium, they
would consider mining UG2.

It is likely that higher rhodium production would have to come  from UG2 due
to rhodium's relative scarcity in Merensky deposits.  Working with October
1979 producer prices, Buchanan calculated that platinum accounts for  61.7
percent of PGM dollar revenues from the Rustenburg area of Merensky,  compared
to 6.6 percent for rhodium (Buchanan, 1979).  The current platinum/rhodium
producer price ratio is higher than in 1979 (0.679 versus 0.475), so  revenue
from platinum would now account for an even higher percentage relative to
rhodium.

The economics of increased rhodium production are thus clear.  Platinum
prices and not rhodium prices currently control PGM production  decisions in
Merensky Reef deposits.  To significantly increase the rhodium/platinum
production ratio, the UG2 seam will have to be exploited by more than the
planned 110,000 ounce annual  production by Western.  To entice  development of
UG2 by Rustenburg and Impala, relative rhodium prices probably  will have to
                                     4-55

-------
                                                                   Charles
                                                                   River
                                                                   Associates


rise; long-term contracts would maximize the incentive effects of any given
price increase.  Because UG2 lies directly below Merensky and could be
simultaneously mined from vertical  shafts, additional  incentives  would not
have to be great to encourage development of UG2,  but  development could still
require several years.

In 1978 the U.S. Bureau of Mines published data on South  African  PGM  refinery
(as opposed to mining) capacity:  2.1  million troy ounces for Rustenburg;  1.5
million troy ounces for Impala; and 135,000 ounces for Western.   Rustenburg's
output is refined by Matthey Rustenburg refiners in South Africa  and  England.
Impala refines all  of its material  in  South Africa, while Western platinum
ships its concentrate to Falconbridge  Nickel's refinery in Norway for removal
of nickel and copper, and then back to South Africa for removal to PGMs.
According to recent news reports, Impala undertook a 10 to 20 percent
increase in refinery capacity from 1978 to 1980, and Western  hopes soon to
increase its refinery capacity to 245,000 troy ounces  at a cost of
$32.7 million.

T. P. Mohide has published estimated 1978 mining (as opposed  to refining)
capacity for South African producers (in troy ounces)  (Metal  Bulletin
Monthly, 1980) :  Rustenburg, 1,516,781 for platinum and 564,383  for
palladium; Impala, 1,093,493 for platinum and 423,288  for palladium;  Western,
105,822 for platinum and 56,438 for palladium.  While  Mohide  does not
estimate rhodium capacities, they can  be estimated by  dividing the platinum
figures by 19.66, the platinum/rhodium production ratio in the Merensky Reef.
Using this method,  we arrive at 1978 mining capacity figures  for  rhodium  (in
troy ounces) of 77,151 for Rustenburg, 55,620 for Impala, and 5,383 for
Western.
SUPPLY ELASTICITY

Long-term purchase contracts can play a very important role in  inducing  South
African producers to increase capacity and production.  Increasing  mining
capacity is a lengthy and expensive process, and inherently risky  from a
business viewpoint since demand may suddenly fall  after the expansion.   The
price incentives required to induce expansion may  not be nearly so  large if
long-term contracts allow consumers to share risks with producers.

For PGM producers in South Africa and elsewhere, two facts  make mining
expansion particularly risky.  First, PGM demand,  whether for catalytic
converter, petroleum, chemical, or electronic uses,  is largely  dependent on
macroeconomic activity in the United States, Japan,  and Western Europe.  A
second variable in PGM markets is Soviet behavior.  While the data  discussed
above showed that Soviet PGM sales to the West are declining, South African
producers are still reluctant to expand capacity when large Soviet  PGM
supplies could begin flowing to the West again.  In  addition, the  Soviets are
thought to have large stockpiles of platinum and palladium  (but no
substantial rhodium stockpile).  Mohide estimated  that in 1979  the  Soviets
                                     4-56

-------
                                                                   Charles
                                                                   River
                                                                   Associates


had 20 metric tons, or 705,479 troy ounces  of platinum,  and  50 metric  tons,
or 1,763,698 troy ounces of palladium stockpiled.  These stockpiles  loom
large as platinum and palladium sold to U.S.  consuming  industries  in 1979
were 1,408,925 and 1,132,621 troy ounces, respectively.

Currently, most major U.S.  PGM users, including  General  Motors,  Ford,  and
Chrysler, have long-term contracts with South African producers.   General
Motors has a ten-year contract with Impala, while  Rustenburg supplies  Ford
under long-term contract.  The GM contract  reportedly includes annual
delivery of 300,000 troy ounces of platinum.

While details of these contracts are proprietary,  they  generally call  for an
annual amount of PGMs to be purchased at a  fixed price,  usually  the  producer
price or lower.  Often there is an inflation clause  in  the contract.  These
long-term contracts for fixed annual deliveries  shift the burden of
fluctuating demand to the auto companies.   When  auto sales and hence PGM
demand are down, the auto companies incur inventory  storage  and  holding
costs, while the South African producers maintain  their  revenue  flows  to
cover production and amortized expansion costs.  One implication of  this
practice is that auto company stocks of PGMs are presumably  inversely  related
to sales, allowing for lags involved in routing  PGMs through catalytic
converter manufacturers.

According to industry sources, large chemical and  petroleum  refining firms
often have long-term contracts as well with the  South African producers.
Long-term contracts are also common for the electronic  industry.   This leaves
the jewelry, medical and dental, and miscellaneous other industries  to buy
mainly on the spot market,  although there also may be some contracting in
these industries.  Overall, it appears that approximately 70 percent of
platinum, palladium, and rhodium consumption (not  counting reuse of
toll-refined material) is typically sold under long-term contract in the
United States.  One implication of this situation  for the Environmental
Protection Agency is that it will generally be quite costly  to implement new
regulations so quickly that the regulated industry is forced onto  the  spot
market, without allowing time for long-term contracts and capacity expansion
by South African producers.

The degree to which capacity and production expansion decisions  are  dictated
by long-term contracts and not short-term PGM prices is  illustrated  by the
historical behavior of South African producers.   Two instances are
particularly illustrative.   The first occurred during the 1979 slump in  auto
sales.  Despite the slump,  Rustenburg, with the  security of  long-term
contracts, undertook to increase output from 1 to  1.2 million troy ounces,
and to aim for 1.4 million  ounces for 1981.  A company  executive stated  that
the expansion was undertaken to meet perceived future increased  demand from
the auto industry (see American Metal Market. 1979a).
                                     4-57

-------
                                                                   Charles
                                                                   River
                                                                   Associates
A second example illustrates the extent to  which  production and producer
stocks of PGMs are tied to long-term contracts  and  thus  unavailable on the
spot market.  Recently, a major U.S computer  manufacturer  attempted to
contract for a two-year supply of platinum  at producer prices and was unable
to do so, allegedly because all producer inventory  was committed in long-term
contracts.  One industry observer commented that  South African producers
would not increase production for a two-year  contract, because this was not
considered a long-term commitment compared  to more  common  ten-year contracts
(see American Metals Market, 1979b).   The cost  of capacity expansion could
not be amortized over so short a time span.

Under normal circumstances, substantial  demands for PGMs are best met from
primary production by the major producers in  South  Africa, through long-term
contracts of more than two years.  Producers  are  fond of emphasizing this
consideration in press reports or negotiations  with potential new customers.
Furthermore, since the above example indicates  that producer stocks are
likely to be committed in long-term contracts,  a  potential purchaser on the
spot market in the United States would have to  rely largely on refiner,
importer, and dealer stocks.  Earlier analysis  of these  stocks in the United
States revealed that while physically they  could  meet a  substantial increase
in PGM demand, past behavior suggests that  these  inventories would be
released only for large price increases.

A further advantage of long-term contracts  is that  they  insulate platinum
producers somewhat from competition with their  own  material, when it is
recycled.  This could be a particularly relevant  concern during the 1980s,
should large amounts of PGMs be recycled from vehicular  emissions control
devices.

Given the key role of long-term contracts,  the  question  then becomes the
extent and speed with which South African PGM production could be expanded
with such contracts.  Discussions with industry sources  ,and CRA analyses of
South African production technology, indicate that  increases in auto industry
platinum demand on the order of 20 to 40 percent  (160,045  to 321,291 troy
ounces over 1979 demand) could be readily met in  two to  three years without
substantial platinum price increases.  Substantially larger production
increases could be implemented over longer  periods  of time without
substantial increases in the (deflated or "real") price  of platinum.

The Merensky Reef is situated such that expansion of capacity and production
is not very difficult.  In the past,  when platinum  demand  has increased, the
length and shallowness of the reef have made  it possible to increase
production quickly.  To extract the ore, surface  adits,  or openings sloping
down along the reef, are used.  In the Rustenburg area,  the slope is
generally only 10 degrees, in which case simple incline  haulage tracks are
used to remove the ore.
                                     4-58

-------
                                                                   Charles
                                                                   River
                                                                   Associates


Deeper deposits in the reef are opened by sinking vertical shafts from 500 to
3,000 feet down.  One difficulty with the vertical shafts and deeper
operations is that the temperature gradient in the Merensky Reef is quite
steep, about one degree Fahrenheit for each 90 feet (compared, for example,
to one degree for each 200 feet in South African gold mines).

One factor that would tend to mitigate price increases in response to higher
PGM demand is the threat of Soviet production and stockpiles entering the
supply stream.  Soviet behavior as a PGM supplier is analyzed below.

It was noted above that platinum, palladium, and rhodium accounted for 61.7,
8.2, and 6.6 percent, respectively, of the 1979 revenue generated from
Rustenburg area ores (Buchanan, 1980.)  In 1980, palladium's share was up a
bit and rhodium's down due to higher and lower current producer prices, but
the point is the same:  increased platinum (not just palladium or rhodium)
demand and prices is usually necessary to encourage increased South African
production from the Merensky Reef.

For South African PGM production ratios to move toward rhodium (and
palladium) and away from platinum, the UG2 seam will have to be mined in
conjunction with the Merensky Reef.  On the demand side, a solution is to
produce three-way catalytic converters that utilize platinum, palladium, and
rhodium in a manner more consistent with their mining ratios.  GM converter
technology apparently utilizes rhodium more intensively than its current
mining ratio, while Ford has indicated that their three-way catalyst  PGM
ratios are not very far from current mining ratios.  Given this apparent
flexibility in the ratio in which rhodium and platinum are used, it may not
take large changes in relative prices to induce equilibrium in the platinum,
rhodium, and palladium markets, even if more extensive use of three-way
catalytic converters is required.

The key to increasing South African production of PGMs for emissions  control
without undue price disruptions is to announce regulatory changes two or
three years in advance,  so the auto companies can negotiate higher output
long-term contracts,  and producers can implement expansion plans in an
orderly fashion.


THE SOVIET UNION

Until  1971,  the Soviet Union was the world's leading supplier of PGMs, but
conditions have changed throughout the 1970s.  Soviet sales to the West of
platinum alone steadily increased from 225,000 troy ounces in 1970 to 631,000
ounces in 1976,  then  steadily decreased to 300,000 ounces in 1979.  PGM
proportions in Soviet production are platinum,  25 percent; palladium,  71
percent;  and other PGMs,  4 percent.   Soviet PGM production is thus heavily
skewed toward palladium.
                                    4-59

-------
                                                                   Charles
                                                                   River
                                                                   Associates


PGM production in the Soviet Union  is  largely a byproduct of nickel
operations, and hence responds  to decisions concerning nickel production.
Soviet production of PGMs  in 1979 can  be  estimated by applying ore grades and
PGM proportions to estimated Soviet nickel production.  Buchanan arrives at
estimates of 757,000 troy  ounces of platinum and 2,130,000 troy ounces of
palladium.  Extending this methodology, estimated 1979 Soviet production of
rhodium is 90,000 troy ounces.

PGMs are found in three major deposits in the Soviet Union:  Norilsk,
Petsama, and the Ural Mountains.  Norilsk, located in the north-central part
of the Soviet Union in the Krasnoyarsk Territory, is by far the most
important deposit, accounting for 85 percent of Soviet PGM production.

U.S. imports for consumption of platinum,  palladium, and rhodium from the
Soviet Union have displayed a more  erratic trend over the 1970s than total
Soviet shipments to the West.  From Table 4-15 it can be seen that 1979 U.S.
imports of PGMs from the Soviet Union  totaled 693,215 troy ounces, down from
a 1974 high of 1,012,321 ounces.  The  1979 figure of 693,215 troy ounces was
20 percent of the 3,479,128 total ounces  imported from all countries for
consumption that year in the United States.

Tables 4-12, 4-13, and 4-14 focus respectively on platinum, palladium, and
rhodium imports by the United States.   Platinum imports from the Soviet Union
have fallen from the mid-1970s, and in 1979 they accounted for only 1.8
percent of total U.S. platinum  imports.   Palladium imports from the the
Soviet Union have been erratic, but in 1979 they accounted for a significant
40 percent of total U.S. palladium  imports for consumption.  Likewise,
rhodium imports from the Soviet Union  have been erratic, but accounted for
15.9 percent of 1979 U.S.  rhodium imports for consumption.

From these calculations it can  be seen that palladium is the only PGM for
which Soviet imports are of much importance to the United States.  Late in
1980, the Soviets virtually halted  palladium exports to the West for reasons
which are unclear.  Despite the cutoff, analysts quoted in press reports were
sanguine about adequate compensatory supplies from South Africa, albeit at a
higher price.  The Soviet palladium embargo in 1980 demonstrated that, while
the dealer price of palladium is significantly influenced by Soviet sales,
supplies from South Africa can  take up much slack and large increases in the
price of palladium are not required to equilibrate the market.  Figure 4-13
shows that the dealer price rose from  around $165 an ounce in early June, to
around $220 in mid-September.


SUPPLY ELASTICITY

Having seen that U.S. dependence on imports of platinum, palladium, and
rhodium froi.i the Soviet Union is now quite low, we still must consider the
                                     4-60

-------
                                                                   Charles
                                                                   River
                                                                   Associates


elasticity of Soviet supplies of these  three  metals  should U.S. demand
increase.  Soviet actions in PGM markets  are  difficult  to predict.  One
theory with some historical  merit holds that  the  Soviets sell metal to the
West mainly to earn foreign  exchange funds, so  sales are inversely  related to
Soviet grain harvests and other determinants  of net  exports.  Soviet export
behavior has followed this pattern less clearly in the  1970s than in earlier
years.  In any case, Soviet  production  of PGMs, and  hence potential exports,
are bound by decisions about nickel  production.  A rough consensus  of
industry sources is that Soviet PGM  flows to  the  United States will remain
quite insignificant during the 1980s and  will not approach higher historic
levels, unless there is a drastic improvement in  East-West relations.


CANADA

Although Canada ranks third  in world PGM  production, it is dwarfed  by South
Africa and the Soviet Union.  In 1979,  Canada produced  327,000 troy ounces of
PGMs, 44 percent of which was platinum, 46 percent palladium, and 10 percent
other PGMs.  By comparison,  in 1979 the Soviet  Union produced an estimated 3
million troy ounces of PGMs, and South  Africa produced  almost 3.5 million
troy ounces.

Canada has not played an important direct role  in U.S.  PGM imports; Tables
4-11 through 4-14 show that  throughout  the 1970s, Canada supplied only a very
small fraction of U.S. imports of PGMs  for consumption.

Canadian production of PGMs, like that  of the Soviet Union, comes as a
byproduct of nickel operations, principally by  International Nickel (INCO) in
Sudbury, Ontario.  Other mining operations are  found in Quebec, British
Columbia, and Pickle Lake, Ontario.   The  PGM  concentration in the nickel ore
at Sudbury is quite low, averaging only 0.8 to  0.9 grams per metric ton of
ore.

A deposit that could come on line at much higher  PGM prices is Lac  des lies,
88 kilometers north of Thunder Bay,  Ontario.  The deposit contains  3 million
tons of ore with 4.450 grams of PGMs per  metric ton, and a platinum/palladium
ratio of 1:8.  In close proximity to Lac  des  lies is the Roly zone,
containing 7 million tons of ore with a PGM concentration of 5.5 grams per
metric ton.  However, for these deposits  to be  profitable, PGM prices would
have to increase by several  times current levels. At the Roly zone, PGM
prices would have to increase to 5 to 10  times  their current level  to make
production profitable.

Because PGMs are produced in Canada  as  a  nickel byproduct, output is
predominantly dependent on events in the  nickel market.
                                     4-61

-------
                                                                   Charles
                                                                   River
                                                                   Associates
UNITED STATES
Heretofore, primary production  of PGMs  in  the  United States has been quite
insignificant (Tables 4-8 to 4-10).   However,  PGM deposits have been
discovered in Montana, Minnesota, and Alaska,  and it is possible that in 5 to
10 years the majority of U.S. palladium requirements, and a significant
portion of platinum requirements, could be met domestically.

The most promising U.S. deposit,  in  Still water, Montana, is currently being
explored jointly by the Jonns-Manville  and Chevron corporations.  Based on
public announcements by Johns-Manville  and Chevron executives, the deposit
can be profitable at 1980 PGM prices, but  a  delay of three to five years to
set up operation is required.   Potential annual production has been estimated
at 643,000 ounces of palladium  and 225,057 ounces of platinum.

Total PGM reserves at Stillwater  are estimated to be around 35 million
ounces.  There are three zones  in the Stillwater complex:  the Basil,
Ultramafic, and Banded.  According to a 1978 press release by Johns-Manville,
the Banded zone has a strike length  of  18,000  feet and an average grade of 2
grams of PGMs per metric ton of ore, with  minor amounts of nickel and copper
across a seam width of 7 feet.  There is another section in the Banded zone
with an average grade of 24.7 grams  of  PGMs  per metric ton of ore.  In both
sections, the platinum/palladium  ratio  is  1  to 3.5.  In the Ultramafic zone,
PGM occurs from traces up to 8.9  grams  platinum and 2 grams palladium per
metric ton.

The Minnamax project, about 60  miles north of  Duluth, Minnesota, is being
explored by Amax, Inc. and is believed  to  contain about 4.4 billion tons of
copper-nickel mineralized rock.  Tests  on  a  120-ton bulk sample by the U.S.
Bureau of Mines found 0.7 to 1.2  grams  of  platinum  , 4.1 to 4.4 grams of
palladium, and small concentrations  of  other PGMs per metric ton of ore.
Using values from a sample tested by International Nickel Co. (0.037 grams
platinum and 0.1 grams palladium  per metric  ton of ore), estimated total PGM
resources at Minnamax are 18 million troy  ounces.  The National Materials
Advisory Board estimated PGM reserves there  to be 50 million troy ounces.

In Alaska, the Crillion-La Perouse Complex contains an estimated 100 million
tons of copper-nickel ore with  a  PGM grade of  0.17 grams per metric ton.  The
platinum/palladium ratio is 0.8.   Goodnews Bay, Alaska, produced over 500,000
troy ounces of PGMs between 1927  and 1976, but is currently not actively
mined.

In the United States, only the  Stillwater  and  Goodnews Bay deposits could be
mined independently of nickel and copper values.  Only Stillwater appears
potentially profitable at 1980  PGM prices.   Plans now call for an annual
output at Stillwater of 55,000  troy  ounces of  platinum, 190,000 of palladium,
22,000 of rhodium, and 2C.OOO of  other  PGMs.   It is economically viable to
                                    4-62

-------
                                                                   Charles
                                                                   River
                                                                   Associates


quadruple these figures at current prices,  given  enough  development  time.
Goodnews Bay could produce around 10,000 troy ounces  of  platinum a year, but
only if prices were to increase tenfold.

If test reports are accurate, the United States for the  first  time could meet
significant portions of its PGM demand from domestic  primary production.
Quadrupling the annual output plans stated  above  would yield 220,000 troy
ounces of platinum, 760,000 ounces of palladium,  and  88,000 ounces of
rhodium.  Considering total PGMs sold to consuming  industries  in the United
States in 1979 (Tables 4-21 through 4-23),  the above  figures represent
15.6 percent of the platinum demand, 67.1 percent of  the palladium demand,
and 105.4 percent of the rhodium demand. Of course,  the estimated quadrupled
output should be compared with higher future consumption in most end uses.

Output of the magnitudes discussed above would put  the United  States squarely
in competition with established South African producers.  Whether Still water
is developed to the extent discussed above  depends  not only on the accuracy
of U.S. production cost estimates, but also on the  price and production
decisions in South Africa.  South African PGM production costs are believed
to be somewhat lower than those projected at Still water.


COLOMBIA

Production of PGMs in Colombia is a byproduct of  gold mining.   Currently
25,000 ounces of platinum are produced yearly from  placer deposits in
Colombia.  Reserves are estimated to be under 1 million  troy ounces  of
platinum.  There is virtually no palladium  production in Colombia.   Due  to
the byproduct nature of PGM productiuon, Colombia would  not be a significant
factor in meeting increased PGM demand, short of  a  price rise  to many times
current levels.


OTHER COUNTRIES

Over 98 percent of the world's PGM primary  production currently comes from
South Africa, the Soviet Union, and Canada.   There  is only byproduct PGM
production from nickel-copper ores in Australia,  Indonesia, the Philippines,
Zimbabwe, and Finland.  At much higher prices, China  could produce small
amounts of platinum as a byproduct of copper-nickel production in the Garsu
province.


WORLD RESERVES

Table 4-23 summarizes estimated world PGM reserves  by country,  with  data on
ore grades and proportions for each PGM. The huge  reserves in  South Africa
indicates that long-run supply response should be very price elastic, as long
as this source of supply remains available.
                                    4-63

-------
                                                                   Charles
                                                                   River
                                                                   Associates
The importance of the South African UG2 seam in  terms  of world  reserves  is
clear.  Moreover, the table facilitates analysis of  the  proportions of
various platinum group metals in each  deposit.   For  instance, rhodium is seen
to be in high proportion at UG2 and Stillwater.

It is also important to remember that  PGMs  are primary products  in South
Africa.  Unike Canada and the Soviet Union,  where PGMs are  nickel-copper
byproducts, production in South Africa should respond  primarily  to events in
PGM markets.


                             SUPPLY RELIABILITY


Any discussion of the reliability of U.S.  platinum and palladium supplies
must, of course, begin with South Africa.   Rustenburg  and Western Platinum
Mines operate in South Africa, whereas Impala's  mines, concentrator, and
smelter are in Bophuthatswana, which became a quasi-independent state in
1977, under the South African Homelands policy.   Impala's head  office and
refinery are in South Africa proper, however. The fact  that much of Impala's
operations are in Bophuthatswana is not necessarily  a  cause for concern,
since Impala is the country's principal  industry and relations  with the
homeland government generally have been cordial.  A  portion of  Rustenburg's
ore bodies are also in Bophuthatswana.

Labor conditions in the Republic of South  Africa actually may be somewhat
more uncertain than those in Bophuthatswana, at  least  in the short run.  The
most recent labor incident in South Africa  affecting mining occurred in  March
1979 when the South African Mine Workers Union called  on all of its 18,000
members to walk out one week after one-third of  its  miners  struck.  The  major
issue was removal of job reservations  for  whites only.  The strike ended
after six days, after reasonably organized  negotiations  among mining
companies, union leadership, and the union  rank  and  file.  The  South African
Chamber of Mines, which can act as an  intermediary in  labor disputes, did not
have to intervene in the 1979 incident.

The South African mining industry naturally has  been sensitive  to concern
about its reliability as a long-run source  of supply,  and it can present a
persuasive case for confidence.  On a  recent tour of South  Africa, U.S.
Representative James D. Santini (D-Nev.),  an advocate  of an expanded U.S.
national minerals stockpile policy, found  little cause for supply concern in
South Africa (American Metal Market, 1980).   Santini serves as  chairman  of
the House Mines and Mining Subcommittee.  He stated  after his tour that
"South Africa's Achille's Heel is resolving their racial  problems," and  he
found that the mining companies themselves  were  on the cutting  edge of social
change.
                                      4-64

-------
                                                                   Charles
                                                                   River
                                                                   Associates


It is certainly outside the scope of the present project to  forecast  the
likelihood of civil unrest in South Africa comparable  to that  which has
occurred in outlying areas such as Rhodesia.   However,  it is worth  noting
that in most cases where such unrest has taken place  in Africa,  mining
operations have not been direct targets for catastrophic sabotage,  presumably
because any of the parties vying for power strongly  prefer to  have  the mines'
revenues, should they be successful.  However, such  historical  precedents  are
hardly conclusive, and there would be many imponderables even  if we had  a
great deal more time to give to the issue.  We have  no immediate reason  for
great concern about supplies of PGMs from South Africa.

Canada has experienced rather severe labor problems  at the Sudbury  nickel
operations.  Most recently, in 1979 INCO (Canada's largest producer of PGMs),
reduced customer deliveries to 40 percent of 1978 levels due to  an
eight-month strike.  Regular INCO customers had to turn to the spot market
and pay premium prices.

The Soviet Union cannot of course be considered a reliable source of  supply.
It has cut off supplies of metals to the United States in the  past, both for
political reasons (e.g., prior to the Korean War) and  for possible  economic
gain (e.g., to support much higher export prices for palladium).  Even
without such identifiable motivations, Soviet actions  on export  markets  can
seem capricious to Western consumers.  The moderate  market response to the
cessation of Soviet exports of palladium in 1980, in  part because of
compensating supplies from South Africa, is encouraging.


               CONSUMPTION ELASTICITY AND SECONDARY  RECOVERY


In addition to increased primary production,  secondary recovery  could help
meet increased demand for PGMs due to tighter emissions standards.
Furthermore, other industries could reduce their consumption if  PGM prices
increase sufficently.  The central issues then become  by how much would
demand by other industries fall  and secondary recovery increase,  after a
given increase in PGM prices.

CRA has recently designed an engineering-based model  of demand elasticity  and
secondary recovery for PGMs in each major industrial  end use.  This analysis
for the U.S. Department of Interior is for the purpose of specifying  a
complete model  of the platinum and palladium industries.  The  results of this
work are utilized for an analysis of industry-specific demand  elasticities
and secondary recovery below.  Analyses of these issues for  PGM  consumption
by the U.S. automotive industry are treated in some  detail  in  another chapter
of this report.
                                    4-65

-------
                                                                   Charles
                                                                   River
                                                                   Associates
PETROLEUM REFORMING
The demand for PGMs in petroleum processing exhibits low demand elasticity
(and then only after substantial lags)  due to lack of substitutes.   Our
estimates of the long-run price response in petroleum reforming is  tabulated
below:
Relative Price Increase
Long-Run Relative
Consumption Decrease
1
1
2
0.967
5
0.867
10
0.700
As the numbers indicate, a doubling of PGM prices would result in  less  than  a
4 percent decrease in quantity demanded even in the long run.   Moderate
additional reductions in PGM demand could be obtained from petroleum
reforming only with a 5- or 10-fold price increase.  The recycling recovery
rate for petroleum reforming is quite high, perhaps 97 percent.


PETROLEUM CRACKING

Most platinum purchased by the petroleum industry is used for  petroleum
reforming, but moderate amounts used for cracking (a recent development,
about which the industry is still  quite secretive)  are unusual  because  there
is no secondary recovery and consumption is more responsive to price
increases than is typical  for PGMs.  Price responsiveness of demand is
estimated in the table below.
Relative Price Increase
Long-Run Rel
Consumption
lative
Decrease
1
1

0.
2
70

0.
5
44

0.
10
31
NITRIC ACID PRODUCTION
Nitric acid production, which uses platinum in  a  catalytic  process,  currently
accounts on average for about 30 percent of U.S.  platinum purchases  for
chemical processing.   (Palladium is used in very  small  quantities  relative to
platinum in nitric acid production; the  elasticity  figures  below refer to
platinum.)

Higher platinum prices can reduce PGM  demand in nitric  acid production by
encouraging better recycling and usage of the so-called "random  pack"
technology, which uses platinum less intensively.

Relative Price Increase	1	2	5	10	
Long-Run Relative
Consumption Decrease                1      0.74      0.49      0.37
                                    4-66

-------
                                                                   Charles
                                                                   River
                                                                   Associates


The table above indicates that a doubling of platinum prices will  decrease
platinum consumption in nitric acid production by approximately 26 percent in
the long run.

CHEMICAL PROCESSES OTHER THAN NITRIC ACID

In most chemical applications other than nitric acid production there is no
significant loss of platinum or palladium, hence little room to improve PGM
recovery rates.  The following very low demand elasticity estimates apply to
both platinum and palladium.
Relative Price Increase
Long-Run Re
Consumption
lative
Decrease
1
1

0.
2
997

0.
5
992
10
0.
989
TELEPHONE SWITCHING EQUIPMENT

Our estimates for price elasticity of demand for palladium in telephone
switching equipment is given below.
Relative Price Increase
Long-Run ReT
Consumption
lative
Decrease
1
1

0.
2
92

0
5
.67
10
0.
25
For a price increase of 100 percent, demand for palladium would decrease by
only 8 percent, but would diminish by 75 percent for a 10-fold price
increase.  This is possible because electronic switching, which will be used
for most new switching by the 1990s, could be used immediately in  place of
palladium (electromechanical) switches if price incentives were sufficient.

Between 1955 and 1974, the Bell  System was the largest purchaser of palladium
in the United States, accounting for around 60 percent of palladium sales
reported by the U.S. Bureau of Mines for the electric  category.  In 1975 a
silver-palladium alloy was substituted for pure palladium contacts.  By 1980
the installation of electronic switching instead of electromechanical  relays
was common.  By 1980, Bell System purchases of nontoll-refined palladium were
down to around 100,000 ounces per year.

Our estimates are that the amount of recycled palladium from scrapped Bell
equipment throughout the rest of the century will  be roughly 100,000 troy
ounces annually.  Gross consumption of palladium by the Bell System is
projected to decline from 200,000 ounces in 1980 to 100,000 ounces  in 1985,
and 50,000 ounces in 1990.  Thus, after 1985,  the Bell  System will  be a net
supplier of palladium.
                                     4-67

-------
                                                                   Charles
                                                                   River
                                                                   Associates
DENTAL AND MEDICAL USES
Gold is the most- generally competitive substitute for platinum and  palladium
in dental work.  If the price of gold were to remain constant, substitution
away from palladium or platinum in response to increases  in their prices
would be rapid and extensive:
Relative Price Increase
Long-Run Re"
Consumption
lative
Decrease
1
1
2
0.15
5
0.01
10
0
(If increased prices for platinum and palladium were  caused  by  a  countrywide
disruption in South Africa, it is plausible that gold prices would  also
increase, in which case substitution away from platinum  and  palladium  would
be much less extensive.)
ELECTRICAL — OTHER THAN TELEPHONE SWITCHES

Platinum and palladium are used in "electronic  inks"  which  dry  into
conductive paths in miniature electronic components.   Our estimates  for  price
elasticity of demand for these two platinum metals are given  below:
Relative Price Increase
Long-Run Relative
Consumption Decrease
1
1
2
0.78
5
0.59
10
0.50
Even if PGM prices were to increase 10-fold,  PGM  consumption  in electrical
uses (other than telephone switching)  would decrease  only  50  percent.   It  is
not anticipated that recovery of PGMs  from scrapped electronic gear  will be
profitable unless prices increase greatly.


GLASS
Platinum and rhodium are used extensively in the  glass  industry,  in  glass
handling and forming equipment,  because of its tendency not to react with
other agents.  There are no close substitutes for PGMs  in  glass production,
and as a consequence price responsiveness is low:

Relative Price Increase	1	2	5	10	
Long-Run Relative
Consumption Decrease               1      0.97    0.93      0.90
                                     4-68

-------
                                                                   Charles
                                                                   River
                                                                   Associates


             PRICE ELASTICITIES FOR CALCULATING THE CRITICALITY
                     OF PLATINUM, PALLADIUM, AND RHODIUM


The above estimates of the responsiveness of platinum and palladium
consumption to large price increases were specifically developed for  analysis
of severe supply disruptions.  Thus, they are designed for the type of
analysis of materials criticality that we carried out in the preceding
chapter.

The only category of platinum and palladium consumption not discussed above
is vehicular emissions control.  The concluding chapter of this study
discusses the limited possibilities for reducing consumption of platinum and
palladium, assuming currently promulgated emissions standards are maintained.
For purposes of estimating the criticality of platinum, palladium, and
rhodium, we suppose conservatively that a five-fold increase in the prices of
platinum, palladium, and rhodium would make efficient only a 12 percent
decrease in quantities consumed (beyond decreases that would occur anyway,
due to technological advances, decreases in engine sizes and so on).   We
assume the same price responsiveness for rhodium consumption.  We do  not yet
regard this price responsiveness to be a definitive estimate, but we  use it
as a placeholder until further work can be done.
Finally, weighting the price responsiveness of all  the categories  of platinum
and palladium consumption by their expected relative importance in the
mid-1980s (not discussed here), we obtain the following price elasticities  of
U.S. demand that were used in the national  criticality measurement in the
second chapter:

•    Platinum  0.03

•    Palladium 0.045

Without similarly detailed analysis,  we specify  the price  elasticity  of  U.S.
demand for rhodium to be the same as  that for platinum.


                    SPECULATION AND INCREASED DEMAND FOR
                           PLATINUM GROUP METALS


In recent years speculating in platinum-group metal  spot and futures  markets
has become more pronounced.  Among the platinum-group metals,  only platinum
and palladium are traded in futures contracts on the New York  Mercantile
Exchange.
                                     4-69

-------
                                                                   Charles
                                                                   River
                                                                   Associates


Speculating in platinum can be simply defined as  buying  and selling  by  those
who otherwise have no use for the metal  in anticipation  of future  price
changes.  While speculation more immediately affects  the spot  price  of
platinum, it can also affect the producer price,  since underlying  forces  of
supply and demand are eventually reflected in the provisions of long-term
contracts as well as in spot prices.   One industry observer has ventured  that
the speculative factor is on average  equal to about 25 percent of  the
activity in the spot market (see American Metals  Market, 1980).

The reasons for speculation in platinum, just as  in other metals,  vary  with
the speculator.  Some investors may reason that future demand  or supply
events which the current price does not reflect may increase platinum prices,
in which case they can purchase on the spot market and hold the metal,  or buy
a futures contract for platinum delivery at a later date at a  specified
price.  Engaging in a futures contract to buy is  referred to as "going  long."
On the other hand, if the speculator  believes that future market events are
likely to decrease platinum's price relative to the currently  quoted futures
trading price, he can "sell short", i.e., engage  in a contract to  deliver
platinum at a future date at a fixed  price.  If the speculator has guessed
correctly, he can purchase the platinum to deliver at the future date when
the spot price is lower than the previously contracted price,  hence  earning a
profit.  Only two to three percent of the platinum futures contracts at the
New York Mercantile Exchange are ever consummated with actual  physical
delivery of the metal.  In most cases, traders cancel their positions on  a
daily basis by entering futures contracts on both the buying and selling
sides, incurring either a profit or loss on the difference.

In addition to speculators who engage in the above-described activity,  there
are those who desire to hold platinum as security for future severe  economic
or political  instability, where "hard" currency like  gold and  silver would be
a more viable medium of exchange.  It appears however that currently most
speculators in platinum are there purely for the  possibility of short-run
economic gains.

At times, platinum prices follow the  speculative  trends  in gold and  silver
prices, allegedly because of concern  over the value of paper currency in
inflationary times.  Often, however,  platinum does not move in tandem with
gold or silver, in part because platinum demand,  unlike  gold,  is
predominantly determined by industrial consumption.

Speculation is often responsible for  discrepancies between dealer  and
producer prices.  As Figures 4-9 through 4-14 indicate,  the discrepancies are
usually short-lived but sometimes sharp.  The speculative sector in  the
platinum market today is credited by  some observers with causing the clear
divergence of platinum dealer prices  over producer prices during the late
1970s and early 1980s (Figure 4-9).  Primary producers of PGMs have  kept
prices to their industrial customers  significantly below those on  the spot
market so as to stabilize demand (and thereby presumably maximize  profits in
the long run, if not the short run).
                                     4-70

-------
                                                                 Charles
                                                                 River
                                                                 Associates
The reaction of speculators can be a significant consideration for evaluating
any proposed regulatory policy that would be implemented so quickly as to
force auto manufacturers or their suppliers onto the the spot market  to
purchase platinum-grown metals.  Even in the absence of speculative activity,
demand and supply for platinum-group metals tend to be insensitive to price
changes, so that relatively large price changes are required to equilibrate
the market when a new demand for PGMs suddenly appears.  With speculative
activity, an unanticipated demand can destabilize the market and greatly
exacerbate the resulting price variability.

In the long run, much more platinum and palladium can be supplied from South
Africa and even deposits in the United States without greatly increasing the
metals'  prices.  If at all possible, regulatory changes requiring sizable new
amounts of PGMs should be planned and promulgated more than two years in
advance, so that long term contracts can be signed with producers and the
required new PGM capacity can be created.
                                    4-71

-------
                                                                   Charles
                                                                   River
                                                                   Associates
                                  Appendix


                ANNOTATED BIBLIOGRAPHY AND GUIDE TO SOURCES
                  OF INFORMATION ON PLATINUM-GROUP METALS
To keep abreast of the PGM market at a general  level,  or to analyze  policy
impacts, there is a nucleus of publications which serve quite well.   Below
are listed organizations that publish PGM information,  followed by a
discussion of which aspects of the market are adequately covered and which
aspects are poorly covered.  Finally, an annotated bibiliography is
provided.

U.S. government publications are readily available and very useful.   The U.S.
Bureau of Mines (US BOM) regularly publishes PGM data  and market analyses;
these publications are often cited in other PGM reports.  The Canada Ministry
of Natural Resources occasionally publishes reports on PGM world markets.
Also, the Mineral Bureau of South Africa publishes regular information.

There are also many private sources of PGM information.  Roskill Information
Services in London provides regular newsletters as well as periodic
publications that bring together data from around the  world.   J. Aron and
Co., New York, is also a source of information  on PGM  markets.
                                      4-72

-------
                                                                   Charles
                                                                   River
                                                                   Associates


A variety of journals and newsletters publish  regular  PGM  information.   Some,
like American Metal  Market,  provide daily price  statistics,  while  others,
like Engineering and Mining  Journal,  provide information of  a  more technical
nature.

Other organizations  provide  irregular but sometimes  very useful  publications
on PGMs.  The International  Precious  Metals  Institute  seminars generate
publications often containing PGM information.   Also,  the  National  Material
Advisory Board has published monographs  on PGM supply  and  use  patterns.

The increased importance of  PGMs has  led to  publication of increasingly
comprehensive data.   Primary production  data are generally good, except, of
course, for the Soviet Union.  Reasonably reliable data exist  for  consumption
in industralized countries.   In the United States, the Bureau  of Mines
estimates "apparent consumption" to compensate  for the underreporting of
actual consumption.   Trade of PGMs among industrialized countries  is  fairly
well captured; data on trade with less developed countries are not as good.

The most obvious deficiency  is data on PGM stocks.  The U.S. Bureau of Mines
publishes data on refiner, dealer, and importer  stocks, but  these  do  not
include consumer shelf and in-use stocks, or private investor  stocks. Only
unofficial estimates.are available for other countries and other kinds of
private stockpiles.


                           ANNOTATED  BIBLIOGRAPHY


U.S. BUREAU OF MINES

1.   Mineral Commodity Profiles, Platinum Group  Metals, No.  22,  September
     1978, 23 pages.  An excellent general reference for the structure of PGM
     markets, and discussion of major supply and demand issues.  U.S. and
     world coverage is provided.  Price, world production  and  trade,  and
     consumption data are included.

2.   Mineral Commodity Summaries, annual, Platinum Group Metals  chapter.  A
     brief (usually two page) update  of important events in  PGM  markets.
     Data provided include percent breakdowns  of U.S.  consumption, U.S.
     production and recycling, U.S. government  stockpiles, and world  mine
     production and reserves.  In addition to  these  data,  a  general
     discussion of trends and issues  is provided.  Publishing  is prompt.
                                     4-73

-------
                                                                   Charles
                                                                   River
                                                                   Associates
3.    Minerals Yearbook,  annual,  Platinum Group Metals chapter.  This is the
     most comprehensive  yearly US  BOM publication on PGMs.  Data are provided
     on producer and dealer  prices,  U.S. consumption by end-use and type of
     production, foreign trade by  country, and world primary production.
     Also included are general discussions on recent trends in many aspects
     of the PGM market,  including  technological changes in the use
     and production of PGMs. Publication lags are  significant, however.

4.    Mineral Industry Surveys, quarterly, Platinum  Group Metals.  A pamphlet
     with quarterly U.S. data on PGMs.   Included are data on PGMs recovered
     by refiners; refiner,  importer  and  dealer stocks; consumption by end
     use; imports and exports by country; and producer and dealer prices.

5.    Mineral Trade Notes, monthly.  This publication compiles PGM trade data
     obtained in part from  the State Department.

6.    Minerals and Materials, monthly edition on Platinum Group metals.  Two
     or three "pages of charts and tables are provided.  Coverage includes
     domestic consumption,  primary and  secondary production, trade, and
     prices.


ONTARIO, CANADA MINISTRY OF NATURAL RESOURCES

1.    Platinum Group Metals  --Ontario and the World  by Thomas P. Mohide,
     Minerals Resources Branch,  March 1979,  162 pages.  Description of each
     of the PGMs, with emphasis  on how they  are used.  For each PGM, a
     complete analysis of where  and how primary and secondary production
     occurs is included.  A good general  reference  or "textbook."


MINERALS BUREAU OF SOUTH AFRICA

1.    The Bureau periodically publishes  memoranda and reports dealing with
     PGMs.  Some relevant publications  are  Internal Memoranda No. 8,
     "Platinum Group Metals in  Canada,"  and  No. 25, "Platinum Group Metals  in
     the People's Republic of China."


ROSKILL  INFORMATION SERVICES, LTD. (LONDON)
1.   Roskill's Letters from Japan, monthly.  Periodic information on PGM
     demand by category in Japan, and dealer and producer price is reported.
     Also, information is provided on ore content and deposits of leading
     world PGM producers.
                                     4-74

-------
                                                                   Charles
                                                                   River
                                                                   Associates
     The Economics of Platinum Group Metals,  Second  Edition,  1979.   Useful
     compendium of worldwide PGM data"Among the  hundreds  of tables are
     end-use for the world and major countries,  voluminous  primary  and
     secondary production data by country,  and international  trade  data.   One
     difficulty, however, is that the 1979  (second)  edition reports data  only
     through 1978.
J. ARON COMMODITIES CORPORATION

1.   J. Aron's Precious Metals Research Department periodically  publishes
     reports on supply and demand characteristics  of PGM markets.
INTERNATIONAL PRECIOUS METALS INSTITUTE

1.   The IPMI conducts regular seminars and publishes the papers from these
     seminars.  Often papers are presented on the economics and metallurgy
     of PGMs.
NATIONAL MATERIALS ADVISORY BOARD

1.   While the NMAB does not publish regular PGM reports, "Supply and Use
     Patterns for the Platinum Group Metals" (1980)  discusses the criticality
     of PGMs and recommends stockpiling objectives.


JOURNALS

1.   American Metals Market (daily) and Metals Week (weekly) provide PGM
     price and market information.  Several other journals periodically
     contain PGM information of a more technical nature:  Engineering and
     Mining Journal, Mining Journal (London), and World Mining.
                                      4-75

-------
                                                                   Charles
                                                                   River
                                                                   Associates
                            CHAPTER 4 REFERENCES
1.   American Metal  Market.   1979a.   Statement by  Albert Robinson, Chairman,
     Rustenberg Platinum Holdings,  Ltd.   Sept.  28,  1979:  12.

2.   	.   1979b.   December 6.

3.   	.   1980.  Feb.  15:  1.

4.   Buchanan, D.L.   1980.  "Platinum:  Great Importance of Bushveld  Complex."
     World Mining.   (August.)

5.               .   1979.  "Platinum-Group Metals  Production  from the
     Bushveld Complex and Its Relationship to World Markets."   University  of
     Witwaterstrand, Bureau of Mineral  Studies, Johannesburg.

6.   Charles River  Associates Incorporated.   1976.   "Policy Implications of
     Producer Country Supply Restrictions: The World Platinum  and Palladium
     Markets."  National Bureau  of Standards, U.S.  Department  of Commerce.

7.   "Cobalt, Platinum Scene Stockpile  Focus."   1981.   American Metal Market
     Feb. 5: 1.

8.   Engineering and Mining Journal.  1979.   September: 38.

9.   Mohide, T.P. 1980.  Metal Bulletin Monthly.  February:  55 (Converted
     from metric tons to troy ounces by Charles River Associates).

10.  National Materials Advisory Board.   1980.   Supply and Use Patterns  for
     the Platinum Group Metals.   Washington, D.C.:   109.

11.  U.S. Bureau of Mines.   1978.   "Platinum."   Mineral Commodity Profiles.
     No. 22.
                                    4-76

-------
                                                                 Charles
                                                                 River
                                                                 Associates
                     RECYCLING OF PLATINUM-GROUP METALS
                     	hRUM CAIALYliC OJNVLHIURS


Since 1975, catalytic converters have been fitted onto the exhaust  systems  of
American automobiles to reduce the level of hydrocarbon and carbon  monoxide
emissions.  The converters contain small quantities of platinum-group metals
(PGMs) deposited on a substrate that facilitate the chemical breakdown  of
exhaust pollutants into harmless components such as carbon dioxide  and  water
vapor.  The PGM content of each catalytic converter is quite small,
approximately 0.05 troy ounces worth about $20 at 1981 prices.   1981 PGM
prices are relatively high by historical standards, and there  is  some
question whether market forces will be  sufficient to  induce recycling of
these metals from spent converters.

Consumption of platinum by the automobile industry beginning in  1974 is shown
in Table 4-21, taken from U.S. Bureau of Mines data.  Catalytic  converters
purchased by the auto industry in 1980  accounted for  approximately  40 to
45 percent of total  U.S. consumption of platinum and  palladium.   By way of
contrast, the chemical  and electrical industries each consumed approximately
15 percent, of these precious metals.  No major reductions in the  use of
platinum-group metals have occurred through 1980.  The 1981 model year
employs a "three-way" catalyst nationwide (instituted in 1980  in  California)
which contains rhodium in addition to platinum and reduced quantities of
palladium.
                                    5-1

-------
                                                                 Charles
                                                                 River
                                                                 Associates
The recovery and recycling of PGMs from the chemical, petroleum,  electrical,
glass and other industries have well-establ ished technologies that can
recover more than 95 percent of the precious metals in scrap residues.  On
the other hand, recovery technologies for catalytic converters are still
being developed.  We will analyze these issues further after a preliminary
discussion of the functions of catalytic converters in automobiles.


           BACKGROUND ON THE FUNCTION OF THE CATALYTIC CONVERTOR


Exhaust gases from the internal combustion engine contain, in the absence
of any controls, four environmentally hazardous components:  unburned
hydrocarbons, carbon monoxide, oxides of nitrogen, and lead.  The first three
can be controlled using the catalytic converter and the fourth by changing  to
unleaded gasolines.  There is a clear distinction between catalytic  systems
for the oxidation of carbon monoxide and unburned hydrocarbons and
systems for the reduction of the oxides of nitrogen.  Problems in developing
exhaust catalyst systems lie not only with the catalyst itself, but  also with
support systems and reactors capable of withstanding engine exhaust
conditions as well as thermal shock, vibration, and general misuse.
Emissions are worst at start-up, when the catalyst is cold and below its
effective operating temperature.


                  THE CHARACTERISTICS OF CATALYST MATERIAL
PGMs are the crucial catalysts being employed in converters.  The PGMs are
deposited in a very thin layer onto one of two forms of inert substrate,
either a monolithic honeycomb or pellets.  To obtain effective performance as
rapidly as possible after engine start-up, the density of the support
material  is kept as low as is practical.  Recently, a stainless steel
honeycomb has been developed.  The entire catalyst is contained in a
stainless steel casing which is placed into the exhaust system between the
exhaust manifold and the muffler.  The casing directs the exhaust flow
through the catalyst bed and protects the catalyst from mechanical damage.
The harmful components in the exhaust gases are converted to carbon dioxide
and water vapor in the convertor.

The PGM loading in each convertor manufactured by GM from 1975 to 1979 was
about 0.05 troy ounces of platinum and palladium, combined in a 5:2 ratio,
amounting to 0.036 troy ounces platinum and 0.014 troy ounces palladium.
Prior to 1980, the total  catalyst (including substrate) weighed between 4 and
6.4 pounds.  In 1980, the density of the substrate was reduced, lowering the
range of weichts to between 2.8 and 4.4 pounds.  In 1981, GM began using the
three-way catalyst with a loading of about 0.05 troy ounces platinum, 0.02
troy ounces palladium, and 0.005 troy ounces rhodium per convertor, which
                                    5-2

-------
                                                                  Charles
                                                                  River
                                                                  Associates
amounts to about $30 worth of PGMs at  1981 prices:   ($475  per  troy  ounce
platinum, $140 per troy ounce palladium, and $700 per  troy ounce  rhodium).
The  total weight of the catalyst  system  averages 5  pounds  in the  latest
converters.  We present below a breakdown of the costs of  a catalytic
converter when it is new and another breakdown of the  scrap value of the  same
converter after 50,000 miles of use.
COST BREAKDOWN FOR NEW AND USED CONVERTORS
The breakdown of the components of a typical new 1980-1981 three-way  pellet
converter are given below:
     Part
     Converter Assembly
     Outer Wrap
     Shell
     Input/Output pipes
     Bed Support
     Insulation
     Pellets
     TOTAL
    Material

409 Stainless Steel
409 Stainless Steel
409 Stainless Steel
409 Stainless Steel
Fiberglass
Alumina + PGMs
Manufactured Costs
    $  1.92
       5.42
       3.00
       1.68
       2.86
       1.85
      30.05
     Total  "aftermarket" selling cost,
     including markup by auto parts dealer
                           46.75
                         $204.00
SOURCE:   Rath & Strong, 1980.
                                   5-3

-------
                                                                  Charles
                                                                  River
                                                                  Associates
The  value of materials  in  a  spent  three-way converter  after  50,000  miles  of
operation (not counting  recovery costs)  is approximately  as  follows:


                                                             Value of
                                                             Material
Part                    Material         Convertor   After Processing  Losses

Converter can and   409  Stainless  Steel  $ 1.25*              $  1.25
 support structures

Pellets                 PGMs              $30.05               $24.04**

                                         $31.30               $25.29
SOURCE:  Charles River Associates, 1980.


From the above table, it can be seen that what remains of  value  from  a  spent
converter is the can and the PGMs.  The can is 409 stainless  steel  which,
because of contamination, will only be worth about $0.05 per  pound.   This  409
stainless is not high-quality scrap, especially after contamination by
exhaust gases.

New converters of the major automakers were sold by auto parts dealers
("after-market") for approximately $200 in 1980.  (See Chi!tons,  1980.)  GM
had the following choices of pellet systems in 1980:

     Oxidizing catalysts only                       $188

     Oxidizing and reducing catalyst (Cadillac)     $204

For renewing spent converter catalysts, GM had a repair kit containing  new
pellets for $38, with a labor charge of $20, bringing the  total  cost  for
replacement to $58.

Ford had a range of honeycomb converter prices ranging from $172 to $320  (on
the Maverick and Granada models).  Ford also charged $20 for  labor  to replace
a spent converter, so the cost of putting in a whole new honeycomb  catalyst
totaled between $192 and $340.
*Convertor can weighs approximately 25 pounds, and 409 stainless steel scrap
from this source is valued at approximately $0.05 per pound.

**80 percent PGM recovery assumed (10 percent loss from abrasion during
operation and 10 percent loss during processing).  There is not yet an
industry concensus on these loss percentages; these assumptions are a
compromise among industry sources with which CRA talked.
                                    5-4

-------
                                                                 Charles
                                                                 River
                                                                 Associates
Chrysler had honeycomb converters that ranged in price from $176 to $196.
Replacement of a spent converter required about 1.2 hours and cost $24 for
labor, bringing the total for converter replacement to $200 to $220 for  an
automobile.

The above prices were typical in 1980.  The manufacturing/vendor cost is  from
one fifth to one quarter of the $200 selling price, or $40 to $50.  The
corresponding actual plant manufacturing cost would be roughly $27 to $34 per
converter.  There is, of course, a significant markup over manufacturing
costs by the time the converter reaches the customer.  However, a spent
converter, after 50,000 miles of use, will contain scrap metals worth at most
only about $31, including the value of the stainless steel can, which in  1980
was worth about $.05 per pound.  If the recovery rate for PGM processing and
recycling is 80 percent, the total  scrap value of contained metals drops from
$31 to $25 per converter.  (See the table above.)


                       RECYCLING CATALYTIC CONVERTORS


There are three phases in the lifetime of a catalytic converter when it may
be scrapped, allowing recovery of contained PGMS:  1) after failure to meet
original specifications; 2) after usage makes the performance of the
converter on an operational  vehicle unsatisfactory; and 3) when the vehicle
is scrapped.  We now consider each of these cases in turn.


PHASE I;  REJECTED CATALYSTS

This category consists of calatysts in the form of beads, pellets, monoliths,
honeycomb or biscuits which are supplied by the manufacturers to the
automotive industry, but do not meet specifications for insertion into
convertors.  For this classification of scrap, the total amount produced
through 1980 was about 4 million pounds, with roughly an additional 2 million
pounds being generated in 1981.


PHASE II;  REPLACEMENT AFTER 50,000 MILES

Catalytic convertors are guaranteed for only 50,000 miles of vehicle
operation.  The material  from failed convertors scrapped prior to scrappage
of vehicles will  be a particularly important source of supply in states with
inspection programs.

The amount of automotive catalyst which will  become available will  be solely
dependent upon mandatory replacement of spent catalysts after 50,000 miles.
Sebastian Musco of Gemini Industries, a new catalyst recycling company,
                                     5-5

-------
                                                                  Charles
                                                                  River
                                                                  Associates
 estimates  this  market  will  eventually  be  8 million  pounds per year,
 containing  approximately  50,000 troy ounces of  platinum and 21,000 troy
 ounces  of  palladium.   (See  Musco,  1979.)   (Replacement catalytic  converters
 will  likely be  taken off  of late model  wrecked  cars where the converter is
 relatively  new  and  undamaged and resold at prices  ranging from $35 to $65
 each.)


 PHASE  III:   AUTO  CATALYST FROM  SALVAGED AUTOS

 This  phase  represents  the largest  resource for  recycled PGMs.  Musco at
 Gemini  Industries has  estimated that by 1984, 4 million cars containing 18
 million pounds  of catalyst  will  become  available.   He  feels that  catalysts
 from  automobiles  in California  will begin  to surface in significant amounts
 by  1981.  He projects  by  1988 approximately 8.5 million converter equipped
 cars will be scrapped, containing  45 million pounds of catalyst,  or
 approximately 300,000  troy  ounces  of platinum and  125,000 ounces  of
 palladium;  by 1998, he maintains these  figures  will  double.

 Phase I and  II  converter  recovery  has been going on for several years,  but
 Phase III is very much in an  embryonic  state.   Phase III  presents a more
 complex problem, involving  dismantling  the converters  from salvaged auto
 wrecks.

 Discussions with several  large  auto scrap  handling  companies and  precious
 metal producers, who are  entering  the business  of PGM  recycling from
 converters, have indicated  that the removal  of  the  converters is  quite  easy
 and fast in most cases.   Either  the converters  are  torched off, which takes a
 couple of minutes, or  a hydraulic  cutter  is  used, which can  remove three
 converter cans  per minute.   One  scrap yard, when removing the gas tank  before
 shredding the auto wrecks,  uses  a  fork  lift  to  remove  the converter along
 with the whole  exhaust system in about  thirty seconds.   There was early
 concern that recovering converters is too  labor-intensive to be cost
 effective.  However, we have  uncovered  no  evidence  that this is in fact the
 case.

 One major precious metals producer has  estimated that  60  percent  of all  PGMs
 in converters going to scrap yards in 1981 are  being recycled in  the last two
 phases.

 AMC and GM converters, containing  loose beads or pellets,  are simple to open.
The beads are removed easily by  a  vacuum suction hose  and shipped to PGM
processors in drums.  The Ford monolithic  converter  can  is more difficult to
 shear open, since a torch is  not used.  The whole converter  is usually
shipped to the  processor.   Probably because of  the  difficulty with Ford
converters (and the probability  of getting some empty  cans),  Gemini
 Industries, a leader in the field,  has  stopped  processing  monolithics and now
only handles the beads.
                                    5-6

-------
                                                                  Charles
                                                                  River
                                                                  Associates
The  converters being  recovered  in  1980  came  predominantly from 1975 through
1979 model  autos,  which  used  the oxidation catalyst only.  The typical
loading was  a total of 0.05 troy ounces of PGMs  in  a platinum:  palladium
ratio  of  5:2, that is, 0.036  ounces  platinum and 0.014 ounces palladium.
(The PGM  loading  varied  with  model year and  engine  size,  but this was the
average loading level.)   At 1980 PGM prices  (producer prices) of $475 per
troy ounce  for platinum  and $140 per troy ounce  for palladium, the total
value  of  PGM in a  typical  two-way  converter  was  $19.15.   Johnson-Matthey, a
large  PGM refiner  which  is getting into converter catalyst recycling, has
reported  their refining  costs,  including PGM losses, to be $8.35 per
converter.   This  leaves  only  $10.80  per converter for the purchase and
transportation of  used catalysts from the scrap  dealers,  which,  as will  be
shown  below, is near  the  breakeven point that makes this  activity at all
profitable.

Used converters usually  pass  through several  hands  before ending up at  the
refinery.  The converters are first  recovered from  salvaged autos at scrap
yards  and auto dismantlers, of  which there are estimated  to be between  15,000
and  20,000 in this country.   (See  McKinnon,  1978.)   Auto  dismantlers
typically buy wrecked cars from individuals,  auto insurance companies,  and
municipal governments.   The converters,  as well  as  other  major auto parts,
then enter the "core-exchange"  market through a  core buyer.  The core buyer
typically drives a truck  around to various scrap yards and picks up barrels
of used converters.   He  then  may sell to a scrap broker,  who then sells  to
refiners  like Gemini, Engelhard, or  Johnson-Matthey.   If  the dismantler  is
large  enough, he may  deal directly with  the  refiner.

Cohen  quoted the following operating cost breakdown for gathering and
processing scrap and  catalytic  converters (Cohen, 1979):

                                                     Range of Cost
     Stage of Recycling	                         (dollars)

    Auto  dismantler                                    4.00-5.50

     Intermediary stages  (core buyer                    1.50-2.50
     and/or core supplier)

    Transportation to refinery                         0.05-0.12

    Refining                                           0.73-1.22

    TOTAL COST                                        $6.28-9.34

We now consider more recent information on this  issue.

Information from a 1980 ADRA* survey  of scrap dealers  in  various  states has
indicated that dismantling time, the  availability of  dry  storage  areas for
300 or more convertors,  and shipping  costs will   be  the major  costs
determining the profitability of recycling PGMs.  If  the  convertors  were
picked up, then $4 per unit would make the activity  profitable for  scrap
                                   5-7

-------
                                                                 Charles
                                                                 River
                                                                 Associates
dealers.  But if the converters had to be shipped, then $7 or $8 would be
required in the case of a Michigan dealer shipping to Gemini Industries in
California.  In general, from a survey of Massachusetts auto dismantlers, as
well as the ADRA respondents, prices ranging from $8 to $11 per converter for
bulk quantities delivered to the refiner were found prevailing.  Gemini, for
instance, is offering $7.86 for pellet converters and $5.70 for honeycomb
converters at existing platinum prices of $475 per troy ounce.  Gemini will
buy only in minimum orders of 1,000 pounds.  This is the price for the
catalyst alone.  If the can also is bought for its chromium value, another
$1.50 to $2.00 is paid per converter.  A Louisiana firm. Southern Scrap
Materials Corporation, now entering the business of refining converter
catalysts, is offering $12.00 to $13.50 per converter.

Transportation costs quoted by Engelhard** are $0.05 per pound of catalyst,
just for shipping from Tennessee to their New Jersey plant, which translates
to $0.30 per pellet converter with 6 pounds of catalyst and $0.20 per
monolithic converter.  Assuming 4 pounds for the catalyst system, and using
the refining costs of Johnson-Matthey noted earlier, a revised table of
gathering and processing costs would be as follows for 1980 prices:


Stage of Recycling   	Pellet Converter           Honeycomb  Converter


                          12                12
Auto Dismantler      $4.00-6.00     $7.36-13.50       4.00-6.00     $5.70

Intermediary Stages   1.50-2.50         —            1.50-2.50
 (core buyer or
  supplier)

Transportation to     0.20-0.30         —            0.20-0.30
 Refinery

Refining*	       8.35          8.35             8.35        8.35

Total Cost           $14.05-17.15   $16.21-21.85     14.05-17.15   14.05


1 = Selling to intermediaries from dismantler.

2 = Selling direct to refineriers from dismantler.
*ADRA - Automotive Dismantlers and Recyclers of America, Washington, D.C.

**Personal communication with Englehard Corp., New Jersey.

+See Warwick, 1980.
                                    5-8

-------
                                                                  Charles
                                                                  River
                                                                  Associates
 These  are  average  costs.   Costs  for  more  specific  situations depend on the
 origin  of  the  scrapped  convertor shipment,  the  location of the refinery,  and
 the  quality  of shipped  catalyst  material.   Moisture,  iron, lead,  and other
 contaminants can reduce the  price that refiners will  ultimately pay for spent
 convertor  catalysts.


                         SPENT  CATALYST PGM  REFINING
The  convertor  recyling  business  is  clearly  in  its  infancy,  and  details on how
the  refining companies  process the  catalyst to  recover  PGMs is  very much
proprietary information  right now.   However, there  are  two  possible
processing routes that could be  taken,  which have been  discussed  publicly in
general terms.  The first  is by  smelting  the spent  catalyst material,
slagging off impurities, and separating out the PGMs.   The  second route is  by
chemically leaching the  PGMs from their alumina substrate and applying
standard recovery methods  now being  used  for primary  production of PGMs.

The  smelting route appears quite costly in  terms of energy  consumption, since
platinum melts at 1769°C.  Rhodium melts  at 1966°C, and the substrate  melts
at over 2000°C.  It appears that smelting would be  the  less efficient
processing route.

Chemical leaching of the PGMs leaves behind the substrate.   It could be
performed with a number  of industrial acids.  The spent catalyst  would likely
be ground to finer particle size to  expose  more surface area for  more
effective leaching.  While no information was obtained  on process details,
the  outline of a plausible method is as follows:  Platinum  and palladium are
readily dissolved in aqua  regia -- a mixture of hydrochloric and  nitric
acids, and HN03, while rhodium will   not dissolve and  is  left behind.   This
separates platinum and palladium into a solution where  platinum could  be
precipitated with ammonium chloride  solution, as the  impure platinum salt.
Several more steps of redissolving and  precipitating  platinum would  be
required for purification.  Subsequently, a pure platinum chloride  salt would
be roasted in muffle furnaces at 1000°C to  give platinum sponge of  99.99
percent purity.

The  palladium remaining  in solution  after platinum  removal   might  be
precipitated in a manner similar to  that for platinum,  and  ultimately  roasted
to form palladium metal   sponge.
                                    5-9

-------
                                                                  Charles
                                                                  River
                                                                  Associates
 The  insoluble  rhodium  in  the  catalyst could be removed with molten sodium
 trisulfate  at  500°C.   From  the  rhodium sulfate solution,  rhodium hydroxide
 can  be  precipitated  and  then  dissolved in  hydrochloric acid to form rhodium
 chloride.   Several more  redissolving  and precipitating stages would be used
 to purify the  chloride before it goes to a glass  lined vessel  to be boiled
 with formic  acid,  forming "rhodium  black"  precipitate.  This is then ignited
 in a muffle  furnace  to produce  99.9 percent pure  rhodium  powder.

 This is  a hypothetical chemical  leaching process  based on the current
 techniques  used  in primary  PGM  production  by Johnson-Matthey, INCO, Engelhard
 and  others.  Recovery  rates for PGMs  from  spent catalysts upon processing are
 estimated to be  approximately 90 percent of the PGMs  that physically remain
 after use.   However, buildup  of lead  on  the catalyst  can  complicate the
 recovery process and reduce the yield.   Other losses  have been identified by
 the  PGM  refiners,  attributed  to abrasion of the beads, which removes some of
 the  PGM  surface  coating.   It  apparently  is difficult  to discern whether  loss
 is due to abrasion or  chemical  reactions during operation of the converter.
 Some spokesmen for vehicle  manufacturers still  claim  that very little loss of
 PGMs  occurs, by  either volatization or by  abrasion.   Another type of PGM loss
 can  occur, called  afterburn,  which  is initiated by  the high operating
 temperatures inside  the converter (over  1000"F).   If  the  engine is run with a
 rich  fuel mix, unburned gases in  the  exhaust can  ignite in  the converter and
 even melt down some  of the  catalyst material,  which can make PGM recovery
 very difficult or  impossible.

 The  current  leader in  the technology  of  recycling PGMs from converters is
 generally acknowledged to be  Gemini Industries  in California.   It is a small
 company  that was retained by  General  Motors to  review the problem of
 expensive existing recovery processes and  low yields.   Gemini  claims it  has
 developed a  new, less  expensive  process  to  recover  the PGMs,  and the company
 anticipates  no difficulties in  recovering  PGMs  from the three-way catalytic
 converters going into  new automobiles in 1981  (Musco,  1979).   Gemini,  as of
 February 1980, was processing 2 million  pounds  of catalyst  material  per  year,
 and  had capacity to process 3.5 million  pounds  per year.   A planned
 $2 million capital  improvement  would  expand capacity  to 10  million pounds per
year  (see Chemical  Week,  1980).

 Available data on PGM  recycling  technologies  is insufficient  to  evaluate
 independently the summary analyses  of their costs provided  by  refiners such
 as Johnson-Matthey.  Our best judgment is  that  industry claims about the cost
of refining converter material  are  plausible, and improved  technologies  such
 as those claimed by Gemini  Industries could  indeed determine  whether
 recycling PGMs in obsolete catalysts  would  be profitable  at PGM  prices lower
 than  those existing today.  We  do not forecast  such substantial  decreases in
 PGM  prices,  but they are a  possibility,  for  example,  if the Soviet Union
begins exporting much more extensively.  At  1981 PGM  prices,  recycling
appears at least marginally profitable even  with "traditional"  refining
methods.
                                    5-10

-------
                                                                  Charles
                                                                  River
                                                                  Associates
            FUTURE AVAILABILITY OF PGMS FROM  SCRAPPED CONVERTORS


During the first three years of substantial catalytic converter use,  from
1975 to 1977, approximately 1.1 million troy  ounces of PGMs were  consumed.
(See Roskill, 1978.)  By the end of 1980,  the cumulative total had  risen to
1,500,000 troy ounces of platinum and 600,000 ounces of palladium (see  Musco,
1979).  By the early 1980s, 550,000 troy ounces of this PGM content had
become available, according to Sebastian Musco of Gemini Industries.
Engelhard has characterized 1985 as the "kick off" year for PGM recycling,
because at that time most scrapped automobiles coming into auto dismantlers
will have catalytic converters.  A significant fraction of new demand for
PGMs by the automotive industry could then be supplied from recycling
catalytic converters.
                                    5-11

-------
                                                                 Charles
                                                                 River
                                                                 Associates
                            CHAPTER 5 REFERENCES
Chiltons 1980 Labor Guide and Parts Manual.  Chi 1 ton Book Co., Radnor, Pa.

Chemical Week.  1980.  February 17.

Cohen, Mark A.  1979.  "Recycling of Catalytic Converters."  Draft.  EPA,
Washington.  April.

McKinnon, R.F.  1978.  "Panel on Trends in the Use of Platinum Group Metals."
Presentation Before the National Academy of Science, Washington, D.C.:
August 30.

Musco, S.P.  1979.  "Reclaiming of Precious Metals From Automobile Catalytic
Converters."  Third International  Precious Metals Institute, Chicago.  May:
178.

Roskill.  1978.  "Platinum Group Metals."  118.

Wernick, N.  1980.  Paper presented at ADRA Annual  Convention.
Johnson-Matthey Corp., Honolulu, Hawaii, November.
                                    5-12

-------
                                                                 Charles
                                                                 River
                                                                 Associates
  SUBSTITUTES FOR PLATINUM-GROUP CATALYSTS IN VEHICULAR EMISSIONS CONTROL


                                BACKGROUND


The requirements of the  Clean Air Act of 1967, together with the amendments
in 1970 and 1977,  provide  the framework for the establishment of emissions
standards from automobile  engines.  These basically set maximum permissible
emission levels for three  pollutants, CO, hydrocarbons, and NOX, in terms of
grams per mile, as determined by a standardized test procedure that averages
the representative modes of operation of an automobile.

The emissions standards  in effect until 1974 could be met by engine
modifications, but those for the 1975 and later model years required the use
of a catalyst for most cars.  For the period of 1975 to 1979, the standards
for NOX (3.1 to 2.0 gm/mile, except for California) could be met by engine
modifications, so the catalyst was an oxidizing catalyst only, designed to
oxidize most of the CO and hydrocarbons in the exhaust to C02 and water.

The standards for 1980 model cars required a further reduction in permissible
CO and hydrocarbon levels. The standards for the 1981 model year reduced the
allowable CO emissions further.  The reduction in allowable NOX emissions was
of great importance technically, since it could not be met on all cars by
engine modifications but required the development of a new catalyst system.
The situation with respect to CO emissions is in a state of flux.  EPA has
waived the 3.4 g/mile standard for many 1981 and 1982 model cars, allowing
the 7.0 g/mile standard  to remain for this period.
                                   6-1

-------
                                                                   Charles
                                                                   River
                                                                   Associates
                             OXIDIZING CATALYSTS
During the research and development leading to the oxidizing catalyst used
during the period 1975 to 1980, literally thousands of catalyst compositions
were developed and tested by chemical  companies, oil  companies, catalyst
manufacturers and automobile companies.  It was found that no catalyst
compositions could begin to meet the requirements of activity and durability
with the use of gasoline containing lead compounds and the accompanying
halogenated additives (such as ethylene chloride) required to prevent the
buildup of lead deposits in the engine.  Hence it was necesesary to require
the use of a lead-free fuel.

Designing catalyst systems to meet mandated emissions standards is a very
intricate matter.  The main factors involved are basically the following:

1.   Catalyst durability;

2.   Emissions before the catalyst reaches operating temperature, that is,
     the "cold start" problem; and

3.   Intrinsic catalyst activity and response to poisoning from sulfur in the
     gasoline.

These factors interact with each other in a complicated manner.  An enormous
amount of technical literature exists on various ramifications of the
problem.  For example, an excellent, critical and recent review by Kummer of
the Ford Motor Co. (1980) cites 168 references.  The key points are the
following:


1.  Catalyst Durability

The catalyst is required to meet the required emission levels specified by
the Federal  Test Procedure after completing 50,000 miles of driving over a
prescribed route in a specified manner.  The catalyst may fail  because of
(i) sintering, that reduces active area, (ii) loss of active catalyst by
attrition or spalling, or (iii) deactivation by poisons, especially sulfur.
Catalyst durability is also determined by engine durability.  A variety of
engine malfunctions can cause excessive amounts of unburned fuel  in the
exhaust that can lead in a short time to excessive catalyst temperatures that
permanently destroy its activity.


2.  "Cold Start"

Much of the CO and hydrocarbons that are emitted in the Federal  Test
Procedure escape before the catalyst reaches operating temperature, and the
problem is exacerbated by the necessity to choke the  engine for start-up,
which unfortunately increases hydrocarbon and CO emissions from the engine.
                                     6-2

-------
                                                                   Charles
                                                                   River
                                                                   Associates
Thus it is vitally important that the  catalyst  be brought  up  to operating
temperature as rapidly as possible,  a  process termed  "light-off."


3.  Intrinsic Activity

    A.  Oxidation Catalysis

The best base-metal  oxide catalysts, such as copper oxide  (CuO) or  cobalt
oxide (00304), show activities (per  unit surface area)  for CO oxidation
comparable to that of noble-metal  catalysts, but they are  less active  for
hydrocarbon oxidation, especially for  saturated hydrocarbons.  Moreover,
these oxides in an unsupported form  sinter readily.   When  supported on
alumina, which increases the available catalytic area,  they tend  to react
with the alumina at high temperatures  to form less active  species,  but
alumina is still considered the best available, reasonably economic support.

Non-leaded gasoline contains about 150-600 parts per  million  of sulfur,  which
is converted to S02 during combustion.  All  base-metal  catalysts  become
gradually deactivated by S02 in the  exhaust,  at the temperature range  of
400-600QC, as a result of adsorbed sulfate species  (Yao,  1975).   Those
catalysts containing copper or chromium are least  affected and the  situation
can be alleviated to some degree by  operating  at temperatures above about
600oc.  The presence of a small amount of noble metal on  the  surface of  a
base metal oxide can also help suppress sulfur  poisoning  (Gallagher, et  al.,
1975).  However, supported noble metal catalysts are  much  less deactivated
than base metal catalysts by S02 at  temperatures below about  500°C. Probably
the poisoning of base-metal oxide catalysts could  be  prevented if the
catalyst were occasionally heated to above 700°C  (Fishel,  1974),  but
temperatures of 1000°C or so can cause their rapid, irreversible  sintering.

In summary, U.S. car manufacturers have gone completely to the use  of
platinum or a mixture of platinum and palladium for oxidation catalysts.  No
catalysts consisting only of base metals are used.  The reasons are as
follows:

•    These noble metals are less deactivated by sulfur compounds  at
     temperatures below 500oC;

•    They are more active for hydrocarbon oxidation than  base metal oxides;
     and

•    They are more thermally resistant to sintering.

Development of suitable automobile catalysts requires enormous expense to
demonstrate durability.  (It has been estimated that  each  50,000  mile
durability test of one car may cost  between $50,000 and $100,000.)   Therefore
there are very strong incentives to  test only  those catalysts that  clearly
have the potential for the required  durability. By about 1973 it had  become
                                     6-3

-------
                                                                 Charles
                                                                 River
                                                                 Associates
evident that base metal  systems were noncompetitive, and catalyst suppliers
and the automobile companies turned completely to optimizing the performance
of noble metal  catalyst  systems.


                           NOX REMOVAL CATALYSTS


The commitment to noble  metal catalysts by car manufacturers was further
reinforced by the necessity to develop a new catalyst system to reduce NOx
levels for the 1981 model  cars.  No known catalyst of requisite activity is
available to decompose NO  to N2 and 02.  However, by operating a suitable
catalyst within a narrow gas composition range, termed the "window," at
essentially the stoichiometric value* it is possible to markedly lower the
emissions of all  three pollutants.  This is termed a "three way  catalyst
(TWO, because all three pollutants -- CO, hydrocarbons, and NOX -- react
simultaneously.  In effect, the exhaust is brought very close to a mixture of
only H20, C02,  and N2.

The three-way plus oxidation catalyst system may also be used.  It consists
of one bed operated as a three-way converter to reduce NOX, after which
secondary air is added and the mixture passed through an oxidizing converter.
Removal of CO and hydrocarbons is somewhat improved, but NOX conversion is
not as good as with the  three-way system, largely because some NH3 is formed
in the first bed and is  reoxidized to NOX in the second.  However, it can be
operated satisfactorily  over a wider range of air-fuel ratios than the TWC.
A three-way catalyst is  apparently becoming the preferred system by U.S. car
makers.  The fuel-air ratio to the engine is carefully controlled to
essentially the stoichiometric value by use of an oxygen sensor on the engine
exhaust.

The reduction of NOX to  nitrogen by CO or H2 is readily catalyzed by base
metal oxide catalysts such as NiO, CuO, or CuCr204, although some of the NOx
may be converted to NH3.  NiO is of particular interest in that, in the
presence of H2 or H20, conversion of NO to NH3 is quite low (Shelef and
Gandhi, 1971),  and the Ford three-way catalyst used in California in 1979
incorporates NiO as well as Pt and Rh (see the discussion below).  However,
these base metal  catalysts are severely poisoned by surface sulfides which
are formed under reducing  conditions, in contrast to sulfates formed under
oxidizing conditions. Deactivation by sulfur can be lessened by operating at
temperatures above 650°C,  but thermal degradation can then be severe.
*The stoichiometric ratio is that ratio of air to fuel at which the oxygen
present is just sufficient to convert  the fuel to C02 and H20 if it were
burned completely.  If the air-fuel  ratio is less than this it is termed
"fuel-rich  operation; if the air-fuel ratio is greater than this it is
termed "fuel-lean" operation.
                                    6-4

-------
                                                                   Charles
                                                                   River
                                                                   Associates


For removal of NOX by reduction, the noble metal  catalysts  are  again  superior
to the base metal  catalysts.  Rhodium is more active  than platinum  or
palladium and produces less NH3, but because of its scarcity  and  high cost,  a
mixture of rhodium with platinum, or with platinum and palladium, is  usually
used.

With three-way catalysts, the closed-loop system that controls  the  air-fuel
ratio to the engine operates essentially by an on-off method  that causes this
ratio to oscillate at a frequency that varies moderately above  and  below
about one Hertz (sec-1).  The catalyst thus operates  under  transient
conditions, the exhaust gas alternately being net oxygen-rich (fuel lean)  and
then net oxygen deficient (fuel  rich).  The performance of  noble-metal
catalysts is improved by incorporating an additional  component  such as cerium
oxide, which is usually described as having an "oxygen storage  capacity."   In
an oxidized state this component provides oxygen for  CO and hydrocarbon
oxidization during the fuel-rich portion of the cycle, when the catalyst is
simultaneously reduced.  When the cycle changes to the fuel-lean  portion,  the
catalyst component is reoxidized from 02 or by NOX; the latter  process aids
in NOX removal.  The effects are still only partly understood and the water
gas shift reaction, 1^0 + CO -»• H2 = C02, may also play a role in  CO removal
(Schlatter and Mitchell, 1980; Hegedus and Gumbleton, 1980).

The monolith three-way catalyst used by the Ford Motor Co.  in California in
1979 contained about 14 percent A1203, 1.6 percent Ni02, 0.7  percent  Ce02,
0.15 percent platinum and 0.015 percent rhodium.   The pelleted  catalyst used
by General Motors in California in 1978 contained noble metals  at about
0.05 percent of total weight, with a platinum/rhodium ratio of  about  2 in  one
type of vehicle and about 15 in another (R. Canole, et al., 1978).  The noble
metal catalysts can be further improved, at least in  principle, by  depositing
the noble metal in layers slightly displaced below the outside  surface of  the
porous support, which gives added protection from poisons such  as phosphorus
compounds that come from lubricating oil (Hegedus, et al.,  1979;  Summers and
Hegedus, 1979; and Hegedus, 1981).  The additional expense  for  development of
the three-way catalysts represents a very large sunk  cost that  in effect
commits the automobile manufacturers even more deeply to the  use  of
noble-metal catalysts.  (The Japanese car makers have followed  essentially
the same path as U.S. car makers, but with slightly different mechanical
methods of controlling the air-fuel ratio with the use of a three-way
catalyst.)


                        BASE-METAL CATALYST RESEARCH


During recent years, a modest research effort on base metal catalysts has
been maintained at General  Motors and by Professor W. Keith Hall  (University
of Wisconsin, Milwaukee, Wisconsin), working with G.M. (Hall, 1981).   This
work has focused on zeolite catalysts which were not  considered in  the early
1970s.  (A zeolite is a crystalline alumino-silicate  containing a very fine
pore structure.  A large variety are known and several are  currently  used
                                     6-5

-------
                                                                 Charles
                                                                 River
                                                                 Associates
commercially as catalysts, primarily  in  petroleum processing.  Some are quite
stable at high temperatures and in the presence  of water vapor.)  Of a large
variety of compositions considered by Hall  and co-workers, a Fe-Y zeolite was
the most attractive (Fu, et al., 1980).   It can  be reversibly oxidized and
reduced between Fe2+ and Fe3+,  and it shows reasonable  activity for reaction
of NOx with CO, although not as good  activity as that exhibited by a standard
GM oxidizing catalyst.  Nonetheless many questions remain, including actual
thermal stability and resistance to poisoning by S02.   Hall speculates that
it is possible that Sfy may be  excluded  by  its size  from reaction sites
accessible to CO, so that S02 might not  be  an effective poison in this case,
but this possibility remains to be demonstrated.

A new class of catalysts having strong interactions  between a supported metal
and the support have been developed by researchers at Exxon (Tauster, et al.,
1981) (so-called S.M.S.I. catalysts).  They consist  primarily of group VIII
metals dispersed on transition  metal  oxides.  (Group VIII includes Fe, Co,
and Ni, as well as the six noble metals.)   They  have not been studied as
automobile catalysts, but they  have been shown to have  unusual chemisorption
properties which suggests that  some members of the class might possibly
adsorb sulfur compounds less readily  and hence be more  resistant to
poisoning.  It is to be emphasized that  this possibility is speculative and
no experimental studies have been done.   However, it is another direction
that could be explored in the future.

Kummer (1980) emphasizes that one of  several requisites for development of a
solely base-metal oxide catalyst is improved stability  at high temperatures
and in the presence of water vapor.  Some work at Ford  Motor Co. was done
with Zr02 as a support, which has a high melting point.  New silicalite-type
zeolites developed by Union Carbide (Flanigan, et al.,  1978) are stable to
very high temperatures and might be interesting  in this regard.


          SUMMARY ON POSSIBLE REPLACEMENT OF PLATINUM-GROUP METALS


The above history of the development  of  automobile catalysts to date shows
the great complexities that have been  overcome in order to develop
noble-metal  catalyst systems that meet the  presently-mandated requirements
for emissions standards and durability.   With the present state of the art,
it probably would be possible to design  a solely base-metal oxidizing
catalyst system that would at least initially meet (Federal Test Procedure)
standards for emissions of carbon monoxide  and hydrocarbons.  This system
would have to be somewhat larger than  present units  utilizing noble metals
and there might be some difficulty with  achieving sufficiently fast warm-up.
However,  because of sulfur poisoning such a unit could almost certainly not
meet present durability standards and  it would not remove NOX, so it could
only meet pre-1981 NOX emissions standards.
                                   6-6

-------
                                                                   Charles
                                                                   River
                                                                   Associates


Technology exists for reducing the sulfur in gasoline to a very low level,
which could substantially prolong the life of a base-metal catalyst.   We are
not aware, however, of any analysis of how costly this desulfurizaton would
be.  It would probably vary considerably from refiner to refiner and  would
depend on the sulfur content of the crude oil.   Low sulfur crude oil  is  even
today a premium material  that commands a higher market price  than lower
quality, higher sulfur crudes.  Unfortunately,  crude oils being imported into
the United States increasingly contain higher sulfur contents.

There is no published evidence that a solely base-metal  catalyst could match
the performance of the present three-way catalyst system, even  initially.
The Ford three-way monolith catalyst incorporating NiO and Ce02 still uses
approximately the same quantity of platinum that is required  for a solely
oxidizing catalyst designed to meet 1980 CO and hydrocarbon emission
standards.  The very high prices of platinum-group metals of  course provide
car makers with a very strong incentive to reduce the cost of their catalyst
units by substituting base metals.  Most promising avenues for  such
substitutions have been extensively explored.

To reduce the use of platinum significantly further would apparently  require
the relaxation of present emissions standards,  either with respect to
allowable rates of emission or durability.  Exploratory  work  on certain
zeolite catalysts by General  Motors and Professor Hall  is interesting, but
much further effort would be required to determine if these catalysts truly
have potential.  Considering the fact that all  previous  base-metal  catalysts
have proved inadequate, the like!hi hood of new, solely base-metal  catalysts
becoming practical for meeting present standards of durability  and emissions
must be considered slight.  We have searched the literature and talked to a
number of people knowledgeable in  the field of  catalysts in general and  auto
catalysts in particular,  but there appear to be no other leads  to solely
base-metal  catalysts that would meet present auto emission standards.
                                    6-7

-------
                                                                   Charles
                                                                   River
                                                                   Associates
                            CHAPTER 6  REFERENCES


1.    Canole, R., et al.   1978.   S.A.E.  Paper 780205.

2.    Fishel, N.A., et al.   1974.   Environmental  Science and Technology,  8,
      260.

3.    Flanigan, E.M., et al.   1978.  Nature,  271.

4.    Fu,  C.M.,  M. Deeba  and W.K. Hall.   1980.   Industrial  Engineering  &
      Chemistry, Product Research  and Development,  19,  299.~

5.    Gallagher, P.K., etal.   1975.  Materials Research Bulletin  10,  623.

6.    Hall, W.K.  1981.   Personal  communication.

7.    Hegedus, L.L.  1981.   Personal communication.

8.    Hegedus, L.L., and J.J.  Gumbleton.  1980.  Chemtech,  10,  630.

9.    Hegedus, L.L.,.etal.   1979.  Journal  of Catalysis, 56, 321.

10.   Kummer, J.T.  1980.   Progress in Energy and Combustion Science,
      6, 177.

11.  Schlatter, J.C. and P.J.  Mitchell.   1980.   Industrial Engineering &
      Chemistry, Product Research  and Development!19,  288.

12.   Shelef, M. and H.S.  Gandhi.   1972.  Industrial  Engineering & Chemistry,
      Product Research and  Development,  11,  2.

13.   Summers,  J.C. and L.L.  Hegedus.  1979.
      Industrial Engineering & Chemistry,  Product Research and  Development.
      18, 318.

14.   Tauster, S.J., etal.   1981.  Science,  211,  1121.

15.   Yao, Y.Y.F.  1975.  Journal  of Catalysis, 39,  104.
                                    6-8

-------
Peer Reviewers'  Comments on Report Number EPA 460/3-82-012


Larry C. Landman, Project Officer, CTAB


Readers of this Report


Due to  the  timing of the  writing  of this final  report  and the
implementation of EPA's  peer  review process, it  was  not possi-
ble  to  incorporate   the  reviewers'  comments  into  the  final
report  entitled   "Scarcity,   Recycling  and   Substitution  of
Potentially  Critical Materials   Used   for  Vehicular  Emissions
Control."

The  following  corrections  are applicable  to  the  above  cited
report which was  prepared under EPA contract  number  68-03-2910
and dated February 1982:

                             Page 2-62

The report now reads as  follows:

    "...EPA might consider  requiring U.S.  vehicle manufacturers
    to  hold  specified  minimum levels  of platinum,  palladium,
    and rhodium."

EPA has no authority to  require manufacturers to do this.

                             Page 3-5

The report now reads as  follows:

    "Thus, their  [Rath  & Strong subproject] report  is  included
    with this study as a separately bound appendix."

At  the  recommendation of  EPA,  this subproject  report  was  not
included with this study.

                             Page 3A-2

The report now reads as  follows:

    "Grams of Platinum-Group Metals Consumed per Vehicle:"

-------
                              -3-
                             Page 4-35

The authors state:
    "Currently,  no  alternative  base-metal catalytic  converter
    appears  economical,  as  discussed  at  length  in  a  later
    chapter."

While base metal catalysts  have  a  number  of problems,  economics
(i.e., cost)  does not appear to be a major difficulty.

-------