ECONOMICS AND PSYCHOLOGY POLICY RESEARCH
: FOR ENVIRONMENTAL MANAGEMENT
AN EVALUATION OF STRATEGIES FOR
PROMOTING EFFECTIVE RADON MITIGATION
by
James K. Doyle*
Gary-H. McClelland*
William D. Schulze*
Paul A. Locke**
Steven R. Elliott*
Glenn W. Russell*
Andrew Moyad**
University of Colorado
Center for Economic Analysis
Boulder, CO 80309
USEPA COOPERATIVE AGREEMENT #CR-813686
March 1990
PROJECT OFFICER
Dr. Alan Carlin
Office of Policy, Planning and Evaluation
U.S. Environmental Protection Agency
Washington, DC 20460
* Laboratory for Economics and Psychology
University of Colorado, Boulder, Colorado
(303) 492-5242
** Environmental Law Institute, Washington, D.C.
(202)328-5150 :
-------
-------
FOREWORD
Research reported here and elsewhere has increasingly pointed to a major
problem faced by the U.S. Environmental Protection Agency and other
environmental regulatory agencies, namely, that there is a widespread difference
between the level of risks perceived by the public and those determined by
professional risk analysts. Some risks -are perceived by the public as being much
worse than analysts believe they .are; other risks are largely ignored by the public
even though analysts believe them to be of major importance. Radon is an
example of the latter type of risk. Risk analyses conducted by the National
Academy of Sciences and EPA have concluded that the risks from radon are
very high, and it appears likely that they are as high or higher than for any
other pollutant not yet substantially controlled by EPA. Yet the research
reported here strongly argues that the public is doing comparatively little in
terms of actually remediating their homes to reduce the risks they are exposed •
to.
This study reports on the relative effectiveness of a broad range of
possible strategies for reducing radon risks to the public, with particular
emphasis on two approaches: information and awareness campaigns to
encourage testing and remediation by the general public and provision of radon
-------
information at time of home sale. The conclusions that the authors reach as to
the relative Iikehihood of success of these two approaches in actually reducing
exposure to radon raise important public policy questions. What should be the
role of government in encouraging the public to take actions that are in their
own self-interest, but which they do not perceive to be of sufficient importance
to bother doing so? Should the government make testing mandatory? If so, for
whom, and who should be informed of the results? Should remdiation be
mandatory or voluntary? Although the authors do not take positions on all
these issues, their research certainly raise all these questions. How these
questions are answered is very important not only for public policy but also for
the 20,000 or so Americans that the risk analysts now believe die prematurely
every year from exposure to radon. If this Report provides a basis for raising
these issues and provides information to allow the reader to assess the relative
merits of the principal alternatives, it will have done its job.
Alan Carlin
Office of Policy, Planning and Evaluation
-------
The information in this document has been funded wholly or in part by
the United States Environmental Protection Agency under Cooperative
Agreement No, CR813686 to the University of Colorado. It has been subjected
to the Agency's peer and administrative review and has been approved for
publication as an EPA document. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
iii
-------
ACKNOWLEDGMENTS
The research presented here was supported by the Office of
Policy, Planning, and Evaluation of the U. S. Environmental Protection
Agency under Cooperative Agreement No. CR-813686. We would like
to thank Alan Carlin, Ann Fisher, and Reed Johnson of EPA and the
social psychology group at the University of Colorado for helpful
comments and suggestions. •
Special thanks are due to Bernard Alvarez of Air Chek, Inc. for
making the field study of the Washington, D. C. area campaign possible
by providing names, addresses, and test results for participants in the
campaign and for assisting in data collection. We are also indebted to
Julie Irwin for research assistance and to Melinda Berg for secretarial
assistance.
IV
-------
TABLE OF CONTENTS
CHAPTER 1
INTRODUCTION AND EXECUTIVE SUMMARY 1
1.1. Introduction: - - . (
The Radon Risk Communication Problem.... 1
1.2. Empirical Findings.. 4
1.3. Conceptual Summary 6
1.4. Recommendation - 8
CHAPTER 2
EVALUATION OF AN INTENSIVE RADON INFORMATION
AND AWARENESS CAMPAIGN.... 12
2.0. Chapter Summary 12
2.1. Introduction '. • • 13
, 2.2. The Washington, D. C. Radon Campaign... 14
2.3. Survey Methodology , 16
2.3.1. Sample Design. '. 16
2.3.2. Survey Design and Implementation. .....18
2.3.3. Characteristics of the Population..'.... 19
2.4. Evaluation of the Washington, D. ,C. Area Campaign ......20
2.4.1. The Pathway to Mitigation .20
2.4.2. Estimation of Transition Rates 24
2.4.2.1. Number of Households Needing
Mitigation....... .-24
2.4.2.2. Percentage of Households Purchasing
Test Kits ..-26
2.4.2.3. Percentage of Households Actually
Testing , ..........28
2.4.2.4. Percentage of Households Mitigating,.........2 9
2.4.3. Campaign Evaluation ...* , 31
2.5. A Model of Radon Mitigation.... 39
2.5.1. Mitigation Dependent Variable. 39
2.5.2. Potential Predictors of Mitigation '. 40
2.5.3. Results : 43
2.5.3.1. Claimed Mitigation Model... , 43
2.5.3.2. Credible Mitigation Model 49
2.5.3.3. Confirmed Mitigation Model 51
2.5.4. Conclusions.. •• -51
2.6. Accuracy in Self-reports of Radon Level 52
2.7. Conclusion • .........55
-------
CHAPTERS
AN EVALUATION OF RADON TESTING AND MITIGATION
AT TIME OF HOME SALE .>. ...59
3.1. Introduction .";..... 5 9
3.2. Survey Design ..51
3.3. Sample Design .^!!!!!!llir.'.""."."!64
3.4. Descriptive Results . .'....'.'"67
3.4.1, Introductory Questions 67
3.4.2. Radon Testing 68
3.4.3. Radon Mitigation 69
3.4.4. Radon Transactions and Negotiation 70
3.4.5. General Transactions and Negotiation.. 71
3.4.6. Characteristics of the Respondents.... 73
3.5. Data Analysis and Discussion 74
3.6. Conclusion 85
CHAPTER 4
LEGAL STRATEGIES FOR ADDRESSING RADON
DISCLOSURE AT TIME OF REAL ESTATE TRANSFER 87
4.1. Introduction'. ..." .8 7
4.2. Current Federal Strategies Addressing Radon Issues....!.1.88
4.2.1. Federal Laws 8 9
4.2.2. Federal Programs 90
4.2.2.1. Revising EPA's Citizen's Guide to
Radon , 90
4.2.2.2. Model Building Codes and Standards 92
4.3. Current State Real Estate Transfer Strategies 92
4.3.1. Florida . 93
4.3.2. Rhode Island 94
4.3.3. New Jersey. 94
4.3.4. Maine 95
4.3.5. New York ^[95
4.4. Radon Disclosure at Time of Real Estate Transfer 96
4.4.1. Sale of a Home or Building 100
4.4.1.1. Pre-sale Period 100
4.4.1.2. Negotiation Period . 103
4.4.1.3. Contract Signing and Inspection
Period 105
4.4.1.4. Final Inspection and Closing Period ..108
4.4.1.5. Post-closing Period „ 109
4.4.2. Lease of a Residence or Building 110
4.4.3. Financing Action after the Home or Building
Sale 111
4.4.4. Issuance of a Building Permit Ill
VI
-------
4.4.5. Issuance of a Final Certificate of Occupancy............. 113
4.4.6. Mandatory Testing or Mitigation of Public
Buildings , „ 113
CHAPTERS
CONCLUSION: AN EVALUATION OF STRATEGIES FOR
PROMOTING EFFECTIVE RADON MITIGATION 117
5.1. Introduction . 117
5.2. Information and Awareness 119
5.3. Incentives .. 125
5.4. Regulation. , 12?
5.5. Recommendation 129
.133
APPENDIX I: '
Facsimile Survey for the Washington, D. C. Area Campaign........ 136
APPENDIX II: '
Facsimile Survey for the Boulder County, Colorado Study,.... 151
APPENDIX III: . ,
Federal Laws Addressing Radon Issues 160
APPENDIX IV:, ,
, Constitutional Issues Relevant to Regulatory Strategies for
Radon 174
Vll
-------
LIST OF FIGURES AND TABLES
Fig. 2.1. Percentage of the population of Air Chek test kits falling into
each radon level category.
Fig. 2.2. The pathway to radon mitigation conceptualized as a
multistage process.
Fig. 2.3. Summary evaluation of the Washington, D. C. campaign
showing population estimates and transition rates for each stage in the
mitigation pathway, by radon level category.
Fig. 2.4. Self-reporting error versus Air Chek radon level for
respondents who reported using one test kit and who reported their
radon level in picocuries per liter.
Fig. 3.1. .Radon mitigation pathway for all 303 home buyers, by radon
level category.
Fig. 3.2. Radon mitigation pathway for all non-IBM home buyers, bv
radon level category.
Fig. 3.3. Radon mitigation pathway for 35 IBM employees, by radon
level category.
Fig. 3.4. Radon mitigation pathway for 88 home buyers who employed
a realtor who gave them radon information, by radon level category.
Fig. 3.5. Radon mitigation pathway for 180 home buyers who did not
receive radon information from either a realtor or employer, by radon
level category.
Table 2.1. Distribution of households in sample by radon level.
Table 2.2. Estimated households in target population by radon level.
Table 2.3. Proportion of sample households mitigating by radon level.
Table 2.4. OLS model of radon mitigation.
vm
-------
Chapter 1
Introduction and Executive Summary
1.1. Introduction: The Radon Risk Communication Problem
The Environmental Protection Agency (1986a) has estimated
that between 5,000 and 20,000 lung cancer deaths per year in the
United States can be attributed to exposure to radon gas. Although
this level of risk is two to three orders of magnitude greater than
matny risks that generate public concern and that are regulated by
' .
federal agencies, efforts to increase public concern about radon, to
promote radon testing, and to encourage appropriate protective
responses to radon have thus far been disappointing.
The difficulty in communicating radon risk and promoting radon
mitigation may be partly because radon has many risk characteristics
that typically lead people to underestimate or to dismiss the risk. The
* " •
following list summarizes the major risk characteristics that appear to
govern radon risk perception: , > •
The objective probability of the risk (between 1 and 5% chance
of lung cancer for a lifetime exposure to 4 picocuries per liter of
radon according to the EPA (1986a)) is below the level where
people understand the risk and respond appropriately ,
(Kahneman and Tversky, 1979; Schulze, McClelland, and Coursey,
1986).
There are no perceptual cues or reminders to alert people to the
presence of the risk (McClelland, Schulze, and Hurd, 1989) since
radon is colorless, odorless, and tasteless.
-------
5.
6.
7.
8.
The risk is natural as opposed to technological and there is no
"villain" to whom one can easily assign blame or responsibility
(Baum, Fleming, and Davidson, 1983; Kunreuther, Ginsberg
Miller, Sagi, Slovic, Borkin, and Katz, 1978).
People's experience with the risk is typically benign (Schulze,
McClelland, and Coursey, 1986) in the sense that they have lived
in their homes many years without experiencing any loss due to
radon).
The effect of the risk is far removed from the initial exposure
(radon-induced lung cancer takes many years to develop and
displays no early symptoms).
Deaths due to the risk are undramatic, occur singly, and are
impossible to unequivocally relate to the risk.
Exposure to the risk is voluntary in the sense that people choose
where they want to live and which home to buy (although only
those who purchased homes since radon began receiving public
attention in 1986 can possibly have accepted the risk knowingly).
The risk is not the same for everyone but varies in complex
ways depending on several dimensions (e.g., geographic location,
soil type, house structure, occupant behavior).
Studies of radon risk perception have generally confirmed the
expectation of lack of concern for radon risk. For example, Sandman,
Weinstein, and Klotz (1987) found the most common response to
radon to be one of apathy and disinterest, and Johnson and Luken
(1987) reported that the perceived risk of their respondents tended
to understate the measured objective risk by orders of magnitude. On
a more optimistic note. Smith, Desvousges, Fisher, and Johnson
(1988) reported some success in communicating radon risk, finding
that respondents did adjust their subjective risk perception in the
right direction after receiving test results. For a review of these and
many other radon studies, see Sjoberg (1989).
-------
It has also been suggested for a wide variety of risks (Adler and
Pittle, 1984; Weinstein, 1987) and for radon in particular (Weinstein,,
Klote, and Sandman, 1989) that mass media information and
awareness campaigns may be unsuccessful in promoting appropriate
protective responses to risk. This report includes the first evaluation
of such a campaign among the general public for radon, and we will
quantify the degree of success of the campaign in promoting radon
mitigation and discuss implications for public policy concerning radon
and related risks.
This report is organized as follows: The following three sections
of this chapter provide an executive summary. The first section
provides a brief overview of the empirical findings from two studies: .
(a) an evaluation of the effectiveness (in terms of ultimate mitigation
rates) for an intensive radon information and awareness campaign
conducted in the Washington, D. C. area and (b) a survey of recent
home buyers in Boulder,Gounty, Colorado. The next section provides a
conceptual, non-quantitative summary of the findings and provides a
framework for characterizing the radon mitigation process that
explains, at least partially, our empirical results and leads naturally to
several key recommendations. The final section highlights our
recommendations for an effective radon mitigation program.
Chapter 2 provides a detailed description of our evaluation of the
intensive radon information and awareness campaign. It also presents
a model that predicts which homeowners will complete effective
radon mitigation actions. Chapter 3 provides a detailed description of
the survey of recent home buyers. At issue is the proportion of newly
• i .. '•
purchased homes that have been tested for radon at or before the time
-------
of the sale and whether effective mitigation actions were taken as a
function of radon test results and other characteristics of the real
estate transaction. Chapter 4 surveys the legal context for radon
mitigation programs consistent with our recommendations which
might be instituted or mandated by government. Finally, Chapter 5
presents our conclusions in greater detail and relates them to the
empirical findings and the legal context.
1.2. Empirical Findings
This report evaluates alternative strategies for motivating people
to test for radon gas in their homes and to mitigate if appropriate. A
review of the literature on risk communication and motivating self-
protective behavior suggests that traditional .information and
awareness programs (such as advertising campaigns and public service
announcements) are likely to fail when they are targeted at the general
population. To test this hypothesis we sent a mail survey to 920
households that had purchased radon test kits as part of an intensive
information and awareness campaign in the Washington, B.C. area (see
Chapter 2). Over 100,000 test kits were purchased as a result of this
campaign. Although we estimate that about 33,000 homes in this area
exceed the EPA action level for radon by a factor of five or more (had a
«r
radon reading of 20 picocuries per liter or higher), the survey results
indicate that only 1.2% of this group have taken convincing remedial
action as a result of the campaign. In addition, only about a third of
the homes in this 1.2% group conducted a post-mitigation retest to
confirm that mitigation had been effective. These homeowners were
4
-------
sent reprints of two EPA documents, "A Citizen's Guide to Radon"
(EPA, 1986a) and "Radon Reduction Methods: A Homeowner's Guide"
(EPA, 1986b). Unfortunately, our results suggest that these pamphlets
may have encouraged people to try their own remedial measures
rattier than seek the assistance of a professional contractor. These
home remedies (e.g., opening basement windows, sealing foundation
cracks) were generally less effective and were not followed by
re testing to verify their effectiveness in spite of clear warnings given .'"
that single limited remedial measures are likely to be ineffective.
In contrast, a telephone survey of 303 home buyers in Boulder
County, Colorado found that over 40% of recently purchased homes
were tested for radon gas at the time of home sale and that this
testing was often motivated by information provided by the realtor (see
Chapter 3). Even though no intensive information and awareness
campaign has been conducted in Colorado and there are, currently no
state laws in effect concerning radon, 54% of tested homes in our
sample that had radon levels above the EPA action level underwent '
mitigation (with 87% of those completing follow-up testing) as part of
the home sale transaction. These results suggest that a radon
information and awareness program targeted at the point of home
sale, when the transaction context provides a strong economic
incentive to repair any problems a home might have, could be highly
effective in comparison to information targeted at the general
population.
EPA is currently limited to supporting information and
awareness through such means as its "Citizen's Guide" and "Radon
Reduction" pamphlets. Our study suggests that these materials
-------
require substantial revision and that new materials specifically
targeted at realtors and homebuyers would be highly desirable,
especially in concert with mandatory disclosure.
1.3. Conceptual Summary
Successful mitigation of a radon problem by a homeowner
typically requires transition through the stages of buying a test kit,
performing the test and sending it to the testing laboratory for
analysis, receiving the test results, recognizing a potential problem,
performing longer-term follow-up testing, confirming a definite radon
problem, taking remedial action, and finally retesting to confirm
mitigation effectiveness. If a homeowner fails at any of these steps,
effective mitigation will not be achieved, this implies that almost any
general information and awareness program, no matter how effectively
it is conducted, will fail to achieve a large reduction in population
exposure to radon because there are so many opportunities for the
homeowner to drop out of the process.
The first and most important bottleneck is getting people to
purchase the test kits. Although the absolute number of test kits
purchased in-the Washington, D. C. campaign was impressive, it still
was a small proportion of the households that ought to have tested.
Those who did purchase test kits were, compared to the general
population, better educated, had higher incomes, owned more
expensive homes, and more often had children. It may be very
difficult for a general information and awareness campaign to reach
the larger population which does not have these characteristics.
6
-------
Another major bottleneck is that when residents were notified of high
r radon levels they often failed to take action or they took action that
was ineffective. It appears that the current radon brochures sent to
the residents may have inadvertently misled them to believe that
opening a window or sealing a few cracks would be effective in
reducing radon levels greater than 20 pCi/1.
It is easy to suggest a number of improvements that might
achieve appreciable increases in transition rates at each stage of the
mitigation process. For example, brochures aimed at those with high
radon levels could be rewritten with more specific mitigation
recommendations. Also, the testing company could follow the
example of survey researchers and send reminders as a follow-up to
the initial test results. If. major improvements were made at all the
steps it might be possible to improve the effective mitigation rate five-
fold so that the general campaign might at least be cost effective.
However, the initial bottleneck of getting enough people to participate
would still likely remain.
The results for presenting radon information at the time of
home sale are more encouraging. This suggests that a program
focused at time of home sale might be appreciably more effective than
a general awareness campaign. Probably the most 'effective strategy
would be to require through regulation that all homes be tested at
time of sale and the test results be disclosed to all interested parties.
If regulation is not feasible, we recommend a program aimed at
realtors, lenders, and others involved in home sale transactions. This
would require a revised brochure and an information campaign
tailored for realtors and home buyers. The effectiveness of such a
-------
program would "be higher but slow because of the time it takes for the
housing stock to change ownership. However, in the long term it
probably would end up being much more effective than an information
and awareness campaign aimed at the general population.
Our results are generally consistent with experience from
attempts to motivate people to take protective action with respect to.
other risks. For example, general awareness and information
campaigns to get people to wear seat belts have generally been
ineffective in increasing voluntary use of seat belts. As a consequence,
more and more jurisdictions, both in the United States and elsewhere,
are turning to regulated, mandatory use of seat belts. There is nothing
in the literature nor in our data that suggests any other solution for
radon will be ultimately as effective.
In summary, our recommendation for increasing .effective radon
mitigation is (a) develop information and awareness materials to be
distributed, perhaps by realtors or lending agencies, to home buyers at
time of house purchase, and/or (b) require radon testing and
disclosure of information about potential radon hazards at time of
home purchase. This recommendation is detailed in the next section.
1.4. Recommendation
This study attempts to integrate three often disparate
viewpoints - psychology, economics, and law -- within the context of
addressing radon pollution. Based on the results detailed in the
following chapters, we believe it is possible to combine these three
8
-------
disciplines to devise an effective strategy to address radon
contamination in homes.
From a psychological perspective the main policy question is
under what circumstances (if any) will people respond to warnings
about radon. This study shows that general information campaigns,
" when used alone, fail to accomplish radon reduction but that radon
information provided at a key point in time, during the home sale
transaction, gets the attention necessary to mitigate radon levels.
Study results also indicate that social diffusion of radon information to
the public through realtors, employers, mortgage bankers, and
contractors may be effective.
From an economic perspective, we have determined that
general information campaigns alone do not appear to be cost
effective. Economic incentives, on the other hand, can encourage
protective behavior but require consumers to be aware of the problem;
in order to be effective, the incentives must be founded on information
and awareness. Additionally* incentive programs must be carefully
designed to avoid inefficiency or bias. Also, the potential cost to the
federal government can be a major problem with incentives.
From a legal perspective, either incentives or regulation may
provide an avenue for addressing the radon problem. As pointed out
above, incentives can be costly and inefficient. It also may be difficult
for the federal government to "police" incentives such as tax credits to
ensure that they are put to their intended use. General regulatory
strategies can suffer the same defects as incentive strategies.
Nevertheless, our research suggests that effective regulations can be
formulated by using the results of this study to design a regulatory
-------
strategy aimed at the home sale transaction. This strategy would
require mandatory disclosure of radon level at time of home sale.
A home sale transaction strategy has certain drawbacks. First, it
cannot address all radon contamination because it does not cover all
dwelling units. For example, people who rent rather than own would
not be affected by this strategy, although the owners of their dwellings
would be. Second, it is a relatively slow approach. Because only about
5% of all homes are sold each year it might take as long as 14 years to
reach one half of all the currently.existing homes. Third, it may be
inequitable. Its costs may fall hardest on the current owner/seller of
the home who may be required to test and mitigate a condition that he
did not create nor to which he contributed. •
However, the home -sale transaction regulatory strategy exploits
a key event - the decision to purchase a home - to focus the attention
of the home sale participants (e.g., buyer, seller, mortgage banker,
realtor) on the potential health effects of radon contamination. During
the home sales transaction, buyers and sellers are focused on the
condition of the home. Buyers are anxious to learn as much as possible
about the property. Sellers are likely to commit resources to correct
any perceived defects.
The home sales transaction strategy requires that before the
closing, radon tests be conducted, and their results obtained and
disclosed, to all participants in the home sale transaction. It takes
advantage of the psychological principles outlined in this report by -
providing information about radon levels in a timely fashion such that
protective behavior is framed as part of a high profile, single decision
that covers a long time span. It also uses existing channels of social
10
-------
communication to disperse information about radon by involving
mortgage bankers and realtors in disseminating radon information.
The home sales transaction strategy also is economically
efficient. Because the burdens of testing, disclosure, and mitigation
are imposed upon the participants to the home sale transaction, the
federal government'will not be forced to provide testing services or
offset costs of remediation. Since the buyer, realtor, and mortgage
banker have a strong self-interest in learning about radon, the strategy
is to a large extent self-policing. Additionally, because the strategy
does not require mitigation, it will allow the buyer and seller to
negotiate for remediation of radon pollution, if necessary. Evidence
suggests that the result of such negotiations will almost always be to
remediate rather than compensate the buyer for accepting the risk.
Thus, free market economic forces shape the ultimate resolution of
the radon problem.
In order to implement this strategy, the Congress must enact
legislation empowering a federal agency such as the Environmental
, : S . .
Protection Agency to promulgate regulations requiring radon testing
and disclosure of test results during the home sale transaction. -Some
of the potential legal impediments to such legislation are reviewed in
Chapter 4. Traditionally, the federal government has not intruded into
home sale transactions, although it has enacted at least one law
requiring disclosure of certain closing costs in home sales financed by
"federally related mortgages."
11
-------
Chapter 2
Evaluation of an Intensive Radon Information and
Awareness Campaign
2.0. Chapter Summary
This study analyzes the effectiveness of a mass media radon
information and testing campaign conducted in the Washington, D. C.
area in the Winter of 1988. Although an impressive number of test kits
(approximately 100,000) were sold, the ultimate mitigation rates
resulting, from the campaign were extremely low. Analyses show that
low mitigation rates cannot be explained by postulating that people's
responses to radon are insensitive to the level of objective risk, but
instead are due to characteristics of the protective response required
to reduce radon risk. Radon may be thought of as one of a family of
intractable risks with risk response profiles that make them
particularly difficult for people to manage and remediate. Traditional
information campaigns for such risks are likely to fail; instead they
may require regulatory, strategies or programs that provide active
guidance and assistance throughout the remediation process.
12
-------
2.1. Introduction
The literature on radon risk perception and the particular
difficulties of communicating the risks of radon exposure are briefly
summarized in Chapter. 1. Although the importance of these risk
perception issues is recognized, the focus in this study is on radon
mitigation. The respondents in this study have at least partially • • -
overcome the perceptual obstacles detailed above to make a voluntary
decision to test for radon and have received the radon test results for
their home, The findings reported below detail their behavioral
responses to receiving these test results.
Our discussion of these responses will focus on three major
issues. First, it has been suggested for a wide variety of risks (Adler
and Pittle, 1984; Weinstein, 1987) and for radon in particular
(Weinstein, Klotz, and Sandman, 1989) that mass media information
and awareness campaigns may be unsuccessful in promoting
appropriate protective responses to risk. This study represents the
first evaluation of such a campaign among the general public for radon,
and we will quantify the degree of success of the campaign in
promoting radon mitigation and discuss implications for public policy
concerning radon and related risks.
Second, given a population of people who have tested for radon,
we will examine, what variables predict who among them will mitigate
and who will not. Of special interest is the relationship between radon
level and mitigation. Johnson and Luken (1987) reported mitigation
to be independent of exposure level, whereas Ackerman (1988) found
13
-------
probability of mitigating to increase with increasing initial radon
reading. A third study conducted by Weinstein, Klotz, and
Sandman (1989) concluded that mitigation depends not on the
precise radon level, but only on whether this level is above or below
the EPA "action level" of 4 picocuries per liter (pCi/1). Because we had
available a database of 55,380 returned test kits, the present study is
the first to be able to oversample high radon levels and should provide
greater statistical power for examining the relationship between radon
level and mitigation behavior than previous studies.
Finally, we will build on the results of the first two analyses to
analyze the mitigation process and develop a profile of characteristics
of the protective response to radon (and related risks) which we feel
governs mitigation responses just as the risk characteristics listed in
Chapter 1 appear to govern risk perception responses.
It should be emphasized that the respondents to this study were
not recruited to participate in a radon study, but were contacted only
after receiving their voluntarily requested test results and having on
average nine months to make mitigation decisions. They faced a real
risk from radon and made actual mitigation decisions concerning their
own homes.
2.2. The Washington, D. C. Radon Campaign
i
The Washington, D. C. information and testing campaign
constitutes the largest and most successful mass media campaign for
radon to date and likely represents the best result that can be
achieved from a campaign of this type. The campaign was a
14
-------
cooperative effort between WJLA-TV, a Washington, B.C. television
station, Safeway foodstores, and Air Chek, Inc., a North Carolina
company that sells and analyzes radon test kits (charcoal canisters).
During the campaign (conducted in January and February of 1988)
radon test kits could be purchased at 125 Safeway stores at a 50%
discount ($4.75 selling price) or by using a discount mail coupon that
appeared in a full-page one-shot ad in the' Washington Post. In
addition, the usual postage and processing fees required to. obtain test
results were waived by Air Chek. This promotional offer was featured
during a consumer affairs news .segment called "Radon Watch" on
WJLA-TV, which began with a three-part series on January 2 and
ended with a second three-part series the week of February 15.
According to the A. C. Nielson company, WJLA-TV reaches about 1.5
million households in Washington, D. C., Virginia, and Maryland. The
total number of test kits sold during the campaign was approximately
100,000, representing about 6.5% of the target population. This level
of participation is extremely high for a media campaign aimed at the
general population. However, as we shall describe below, purchasing a
test kit only rarely results in eventual protection against radon risk.
Along with their radon test results reported in picocuries per
liter, all those who returned test kits to Air Chek received a two-page
letter. The letter briefly explained their test results, and made
suggestions for follow-up action which generally corresponded to
those in the EPA publication "A Citizen's Guide to Radon" (EPA,
1986a). Those with readings above 4 pCi/1 received additional
\ ' - . -
information that varied depending on which of three intervals their
test results fell into, those with readings between 4 and 20 pCi/1
15
-------
were sent a reprint of the "Citizen's Guide." The Citizen's Guide is a
16-page pamphlet that gives basic information about radon and radon
detection and makes suggestions for protective action which vary
according to radon level. Those with readings between 20 and 50
pCi/1 were sent the Citizen's Guide as well as a second EPA publication
titled "Radon Reduction Methods: A Homeowner's Guide" (EPA,
1986b). The "Homeowner's Guide" is a 24-page pamphlet which
describes and compares nine basic techniques for radon reduction,
ranging from inexpensive natural.ventilation to more costly methods
such as forced ventilation and sub-slab suction. Those with readings
greater than 50 pCi/1 received a letter urging an immediate retest
accompanied by a free retest kit. The interval limits of 4 and 20 pCi/1
correspond to the limits of intervals for which the EPA recommends
different actions in their Citizen's Guide. The interval limit of 50 pCi/1
was chosen by Air Chek as part of their standard testing procedure.
2.3. Survey Methodology
2.3.1. Sample Design *
In November and December of 1988 we conducted a mail survey
of a sample of participants in the Washington, D. C. area campaign.
The sampling frame consisted of all radon test kits returned to Air
Chek for processing having serial numbers in a particular range which
identified them as coming from program participants. There were
55,830 such kits.
16
-------
o
O)
ra
c:
o
o
' l_ V
0)
Q.
80-
60-
40-
•
20-
n -
71.3
- -
| '" ,'
25 .8
' " • '
: 2.2 0.6
<4
4-20 20-50
pCi/l
>50
Fig. 2.1. Percentage of the population of Air Chek test kits
falling into each radon level category.
Figure 2.1 shows the proportion of the population of test kits
falling into each of the four radon level intervals described above. The
distribution of radon test results is of course very skewed with over
70% of the readings less than 4 pCi/l and only about 3% greater than
20 pCi/l. A simple random sample from this population of test kits
would .clearly yield few kits with high radon levels. Thus, a stratified
• '^ .
sampling design was used to ensure adequate numbers of observations
to allow separate estimates of mitigation rates at each radon level. We
randomly sampled 250 test kits from each of the four intervals defined
above. After removing duplicates (many households purchased more
than one test kit) and commercial and government addresses, we had
the distribution of households shown in Table 2.1. There was a'
greater loss of usable addresses at the higher radon levels because for
17
-------
Table 2.1. Distribution of Households
in Sample by Radon Level
Radon level
(pCi/1)
<4
4-20
20-50
>50
Total
Households
in sample
248
241
229
202
920
those levels we were sampling a much larger proportion of the
population (e.g., over 75% of the test kits with radon levels greater
than 50 pCi/1) and duplicates were therefore more frequently
encountered. Surveys were mailed to all remaining 920 good
addresses.
3.2. Survey Design and Implementation
The survey was prepared according to the Total Design Method
described by Dillman (1978) and was mailed and collected by the
teoting company, Air Chek. Recent research has indicated that
prepaying respondents increases response rates (Berk, Mathiowetz,
Ward, and White, 1988). So, in addition to the standard reminders
commonly used to increase response rates, we included a two dollar
18
-------
bill in the initial mailing. The cover letter explained that the money
was being given to help compensate participants for their time and
efforts in filling out the survey. This approach yielded a response rate
of just over 77%. The response rate did not vary by radon level.
The survey questionnaire was a twelve-page booklet containing
59 questions, although only 36 questions applied to all respondents.
The questionnaire assessed respondent's general experience with
radon and asked questions about their radon test, their test results,
and their reactions to the test results. In addition, respondents who
indicated they had taken action to" reduce their radon levels were
asked detailed questions about their mitigation experience, including
which specific reduction methods they used, who actually did the
mitigation work, how much money they spent on mitigation, and
whether they had performed a retest to confirm mitigation success.
The questionnaire also included the standard sociodemdgraphic
questions as well as a few questions about characteristics of their
home and their behavior toward other risks (smoking and wearing
seatbelts). A facsimile of the survey instrument used in this study is _
presented in Appendix I, including results for each question by radon
level category.
2.3.3. Characteristics of the Population . - ' • •
Before proceeding to our analyses of the survey, it is worth
describing here several characteristics of the p'opulation of people who
returned test kits to Air Chek. The results, below are estimates based
on sample data that have been weighted to correct for the
19
-------
oversampling of higher radon levels described above. Results showed
the population to be highly educated (62% achieving at least a four-
year college degree) and relatively wealthy (76% having an annual
household income before taxes of more than $40,000). The
overwhelming majority of people (94%) own their own homes, and
s
82.8% of all residences are single-family detached homes. The
average age of the population is 47, and 62% are male.
Like virtually all research on radon to date, the population of
testers is clearly more educated and wealthier than average and almost
exclusively own their own homes. "Also, test kits are more likely to be
sent in by a male than a female. This survey does not address how
people with low incomes, people with limited education, or people
who rent rather than own respond to radon.
2.4. Evaluation of the Washington, D. C. Area Campaign
2-4.1. The Pathway to Mitigation
The ultimate goal of any radon risk communication program is to
get those households in need of mitigation to mitigate. Therefore our
evaluation of the Washington, D. C. campaign will be based on
estimating what percentage of households in the target population
with maximum radon readings above the EPA action level of 4 pCi/1
mitigated as a result of the campaign.1 To estimate the ultimate
1 There is quite a bit of controversy over what level 'of radon
indicates an unequivocal "need for mitigation." The Environmental
Defense Fund, for example, claims the EPA action level is too high and
recommends indoor radon levels be reduced to the level of outdoor
background radiation (0.2 pCi/1) (Yuhnke, Silbergeld, and Caswell,
1987). Others claim the EPA has greatly overestimated the risk and
20
-------
mitigation rate we will estimate the percentage of the target
population surviving at various identifiable stages along the pathway to
successfurmitigation. This will also allow us to compare transition
rates between stages to identify those stages that may be especially
troublesome for a radon campaign.
Figure 2.2 illustrates the uncertain, multistage pathway to
mitigation. Clearly households in the target population that need
mitigation can drop out of the pathway short of success at several
different stages. Stage 1 is the purchase of one or more radon test
kits at participating Safeway stores. If a household does not purchase
a test kit, then it will not mitigate as a result of this campaign,
f ' . '
although mitigation may occur through other routes not associated
with this particular risk communication effort.
Stage 2 is actually conducting the radon test. Many people may
purchase a test kit but not actually conduct the test. Or, they may
recommend no action at any level until further epidemiological studies
have been completed (Cole, 1990). In our campaign evaluation we
have chosen to define those households with readings greater than the
EPA action level as "needing mitigation" because of the widespread
adoption of this level as a standard by state agencies, testing
companies, mitigation contractors, realtors, and homeowners.
(Although EPA did not originally intend the action level to be
interpreted as a standard, it has become one in practice, largely
because no other is available. It may not in general be possible for risk
communicators to .communicate precise risk levels without them
being interpreted as standards by the public.) In any case, as shown in
Fig. 2.3, ultimate mitigation rates are very low at Ml radon levels. Our
evaluation of the effectiveness of the campaign is therefore relatively
insensitive to the precise definition of "need for mitigation."
In addition, we have made the simplifying assumption that the
level indicating a need for mitigation is the same for all households,
which may not be the case. For example, people who spend little time
in their basements might have less of a need to mitigate a given
basement level reading than people .who spend more time in their
basements.
21
-------
Population of Households
Needing Mitigation
| Buy
1
Test Kit |
1
1 Perform Test 1
1
Claimed
Mitigation
*
1 Credible
Mitigation
Confirmed
Mitigation |
Stage
stage 2
Stage 3
Stage 4
Stage 5
Place the kit in the basement but forget to send it to the testing
company for measurement of the radon level. Our operational
definition of "performing the test" is returning a test kit to Air Chek
for evaluation.
The final three stages encompass the process of carrying out
some mitigating action or actions once a high radon test result has
22
-------
been received from the testing company. In our survey there were
several possible operational definitions of mitigation. The simplest
definition is to use responses to the question, "Have you taken action
to reduce radon levels in the home which you tested?" An affirmative
response to this question constitutes Stage 3 (claimed mitigation) in
our analysis.
However, there are problems with accepting self-reports of
mitigation. First, there may be a social desirability effect. People may
have been embarrassed to admit to the testing company who sent
them their results that they had taken no action to reduce their radon
level. Second, people may have taken actions that are legitimate but
ineffective efforts to reduce their radon level.- For example, many
respondents in our survey reported attempting to reduce radon levels
by leaving their doors and windows open more frequently. We
i •
therefore added another stage to the,mitigation pathway. If someone
mitigated effectively, it is reasonable to expect them to respond with a
i
dollair amount greater than zero to the question, "About how much
have you spent on radon reduction?," and to be able to identify the
specific category of reduction method that was used (e.g., sealing,
ventilation, filtering, pressurization) from a list of possible options.
This is our operational definition of Stage 4 (credible mitigation).
Finally, attempting mitigation in a credible manner is not the
final stage in ensuring protection against radon risk. To verify that
mitigation has been successful, a post-mitigation retest should be
performed. An affirmative answer to the question, "Have you had your
home retested for radon since completing your radon reduction
23
-------
efforts?," is therefore our operational definition of Stage 5 (confirmed
mitigation).
The next section details how we arrived at estimates of the
transition rates between the stages on the pathway to mitigation for
each radon level category. However, it should first be noted that for a
multistage process such as the mitigation pathway we have described
only a relatively small proportion of the population will survive to the
final stage. For example, even if the transition rates for each of the
five stages were about 50%, then the final proportion reaching Stage 5
would be only about 3%. If the transition rates were about 85%, an
unreasonably high expectation, the ultimate confirmed mitigation rate
would still be only about 44%. Because these'observations led us to
expect a very low ultimate confirmed mitigation rate, we tried to make
assumptions, when necessary, which would err on the side of favoring
the effectiveness of the program.
2.4.2. Estimation of Transition Rates and Population Proportions
2.4.2.1. Number of Households Needing Mitigation
First we need to estimate the total number of households in the
population targeted by the Washington, D. C. area campaign.
According to the A. C. Nielson Co., the viewing area of television
station WJLA contains approximately 1.5 million households. Although
census data for the Washington, D. C. and Baltimore metropolitan areas
indicates the total number of households to be approximately 1.9
24
-------
million, we will use the conservative estimate of 1.5 million
households that could potentially have participated in the program.
Our sample of testers consisted primarily of residents of single-
family, detached homes (90.9%). Although radon is not necessarily
just a problem of single-family homes, it is more likely to be a problem
for such residences than it is for multi-unit condominiums or
apartment buildings, and residents of single-family homes are more
likely to have the authority to implement effective mitigating actions.
It therefore seems reasonable to restrict our analysis to only the
population of single-family homes in the target area. From census data
the average proportion of single-family homes in the Washington,
D. C./Baltimore area is approximately 62.2%. We therefore reduce our
original estimate of 1.5 million households to 933,630 single-family
homes.
Finally, in order to determine how many of these homes need
mitigation, we must estimate the proportion that fall into each radon
level category. We know the proportion of radon test kits returned to
Air Chek which fell into each category. However, these proportions
might underestimate the actual proportions because many households
may have purchased more than one test kit and placed them in
different areas of their homes (for example, one in the basement to
get a maximum reading and one in an upstairs bedroom to get a
minimum reading). We also have available a self-report of radon level
from respondents to the survey. However, simply using self-reports
might also underestimate the actual proportions in each category due
to people's tendency to underreport their true radon reading (see
Section 2.6). In the following analysis we chose, as the best available
25
-------
indicator of whether a household needs to mitigate, the maximum
radon reading obtained from any test kit. 2 We attempted to estimate
this from our survey by using the maximum radon level for the
household reported by either the respondent (for any test results they
received) or by Air Chek (for the test kit selected in the sample). 3
Weighted estimates 4 of the population proportions for each radon
level are shown in Table 2.2. 5 Tne estimated number of households
needing mitigation (having a maximum radon level above 4 pCi/1) is
381,714 (40.88%). -,
2.4.2.2. Percentage of Households Purchasing Test Kits
Next we need to estimate the percentage of households needing
mitigation who purchased test kits from Air Chek. The test serial
levels
2°3
<*™PU"g <* *gher radon
26
-------
Table 2.2. Estimated Households in Target
Population by Radon Level
Maximum radon
level (pCi/1)
Percentage
Estimated households
in population
<4
4-20
20-50
>50
Total
59.12%
37.38%
2.77%
0.73%
100.00%
551,916
348,958
25,894
6,862
933,630
numbers indicate that approximately 100,000 test kits were .sold as
part of this particular campaign. We must first reduce this estimate of
total households purchasing to 90,900 single-family homes,
representing the 90.9% of our sample.6 However, survey responses
indicate that on average each household purchased 1.5 test kits. Thus
we need to divide the 90,900 homes by 1.5 to obtain the estimate of
approximately 60,600 single-family homes purchasing test kits.7
6 There was a slight tendency for tests with lower radon levels to
be more likely to come from non-single-family homes. Thus the
90.9% estimate of the percentage of single-family homes from our
sample is reduced to 82.8% when weighted to account for
oversampling of higher radon levels. Using the 90.9% estimate might
overestimate the number of testers and therefore make the program
appear slightly more effective.
7 It is possible that people who returned test kits were more
likely to have purchased more than one test kit, in which case the
estimate of 60,600 homes would underestimate the true number of
27
-------
Dividing 60,600 by the number of single family homes in the target
population yields an estimated transition rate of 6.5% of households
purchasing test kits. Only 24,811 of the 381,714 houses needing
mitigation reached this stage.
2.4.2.3. Percentage of Households Actually Testing
We must now estimate the percentage of households that
actually conducted the radon test given they purchased a test kit. We
begin with the 55,830 test kits returned to Air Chek for evaluation.
Multiplying this number by 90.9% reduces the estimate of single-
family homes actually testing to 50,750. The weighted estimate from
the survey responses of 1.5 test kits per household does apply to the
population of returned test kits.8 Thus the estimate of the transition
rate from purchasing (Stage 1) to testing (Stage 2) is 33,833/60,600 =
55.8%. 9 Only 13,845 of the 381,714 homes needing mitigation
tested.
households purchasing test kits. This is not a serious problem,
however, because we do have a firmer estimate of the number of
households testing. Thus, changing the estimate of households
purchasing (Stage 1) would not affect the estimate of the overall
transition rate from the population to Stage 2.
8 There was a slight trend for households with higher radon
levels to report having purchased more test kits. If this is indeed the
case, then using a common estimate of number of kits purchased
would slightly overestimate the number of households with high radon
levels who purchased kits and actually tested. Again, this error would
be on the side of making the program appear slightly more effective.
9 Air Chek reports that they typically receive for analysis a much
higher proportion of the test kits which they sell (over 90%).
Apparently many of the people responding to the campaign are
purchasing radon test kits on impulse and these people are much less
likely to actually perform the test than people who purchase on their
own initiative.
28
-------
2.4.2.4. Percentage of Households Mitigating .
Finally, we can estimate from our survey responses, for each
radon level category, the proportion of testing households that went
on to mitigate. Table 2.3 shows these proportions for each of the
i - -
three mitigation stages we have defined: claimed, credible, and
confirmed. The proportion of households claiming mitigation
increases dramatically with radon level, going from 11.9% for the 4-20
pCi/1 category to 52.5% for the >50 pCi/1 category. We test the
reliability of this relationship in Section 2.5.
The proportion of testers who mitigated in our study, even by
the very liberal standards we used to define the "claimed mitigation"
category, is quite low compared to the result's of two previous studies.
Weinstein, Sandman, and Roberts (1988), for example, reported the
mitigation rate for a population of 123 New Jersey homeowners to be
62% even for radon readings as low as 4-8 pCi/1 (these were living
room readings, as opposed to basement readings, and therefore may
not represent maximum radon readings). Also, Ackerman (1988)
reported mitigation rates as high as 38% at 12.5 pCi/1 and 62% at
50 pCi/1 for a population of testers in a suburb of Stockholm, Sweden.
The major difference between these two studies and the present one
appears to be the amount of outside help received by testers from a
governmental agency. In the New Jersey study, participants received
free testing and advice from the New Jersey Department of
Environmental Protection's confirmatory radon monitoring program.
In the Stockholm study, tests were conducted by the local health
department for a fee of $65, and those with a high radon reading
29
-------
Table 2.3. Proportion of Sample Households Mitigating
by Radon Level
Maximum radon
level (pCl/1)
<4
4-20
20-50
>50
Proportion
Claimed
4.7%
11.9%
43.4%
52.5%
of population
Credible
2.7%
5.5%
32.0%
40.4%
mitigating
Confirmed
0.0%
1.8%
10.3%
19.6%
received free retesting and mitigation advice from the department. In
our study, participants received printed information produced by Air
Chek and the Environmental Protection Agency, but had no personal
contact with EPA or any other government agency or mitigation firm
unless they had initiated the contact. Apparently mitigation rates for
homeowners who test on their own initiative and receive only minimal
assistance from outside agencies during the testing and mitigation
process are especially low.
The decrease in the proportion of households mitigating in a
credible way and confirming mitigation success illustrate the
importance of defining "mitigation" carefully. For example, for the 4-
20 pCi/1 category, only 46% of those claiming mitigation reported
spending money on mitigation and could report the specific mitigation
method that was used, and only 15.3% of those claiming mitigation
reported retesting after mitigating.
30
-------
Although the three mitigation categories we have defined are not
necessarily (but often may be) separate stages for each individual, they
can be analyzed as separate stages in the mitigation process for the
population as a whole. Of the 13,845 homes in the target population
which needed mitigation and actually tested, only 2043 (14.8%)
claimed they had mitigated, only 1093 (7.9%) mitigated "credibly,"
and only 376 (2.7%) retested to confirm mitigation success.
2.4.3. Campaign Evaluation
Figure 2.3 summarizes our evaluation of the Washington, D. C.
campaign and includes absolute population estimates and transition
rates for each stage -in the mitigation pathway and for each radon level
category. As expected, the ultimate confirmed mitigation rates are
very low. This is an inevitable consequence of any multistage process
in which one or more transition rates may be low. Of the 381,714
single-family homes in the target population which needed mitigation,
only 376 remain at the final stage of the process. This yields an
overall confirmed, mitigation rate of just under .1%. The overall rate of
.l%i improves only to .29% If the "credible" definition of mitigation is
considered sufficient and only to .54% if the "claimed" definition of
mitigation is considered sufficient. Even for the highest radon level
category, >50 pCi/1, the ultimate confirmed mitigation rate is only :
,7%, and this improves only to about 2% for the liberal "claimed"
definition of mitigation.
The transition rate going from the population of households to
purchasing a test kit was 6.5% for this program, which is a very
31
-------
933,630 Single-Family Homes
6,862
1
6 . 5 °/
r
55.8%
pCi/l
249
i
52.5
r
%
48.4%
Returned
test kit
mitigation
Credible
mitigation
Confirmed
Ultimate confirmed
mitigation rate
376 Single-Family Homes
Fig. 2.3. Summary evaluation of the Washington, D. C. campaign
showing population estimates and transition rates for each stage in the
mitigation pathway, by radon level category.
32
-------
impressive participation rate for a mass media campaign aimed at the
general public. However, this means that, for example, out of the
> - . '•''-_ r - •"
6,862 homes with extreme radon levels (>50 pCi/1), only 445
purchased a test kit. The vast majority of homes with extreme levels
therefore remained undetected. Even so, if this 6.5% of the
population were now protected against radon risk as a result of the
campaign, the Washington, D. C. program might be considered a
success. The subsequent transition rates, however, demonstrate that
this is not the case. Only just over half of those who buy a test kit
actually return it for processing. Only about half of those who get test
results in the highest categories take even simple mitigation actions
such as opening windows more often. Overall, even for the highest
radon level categories, from 20 to 60% of the population drops out of
the mitigation process at each of several stages. Clearly, protecting
oneself from radon risk is a long and difficult process for many people,
and only a very few stick with it to the very end.
Of particular concern are the transition rates after purchasing a
test, kit for the 4-20 pCi/1 group. Only 11.9% of those receiving test
results in this range even claim to mitigate, only 46% of those who
claim to mitigate do so in a credible fashion, and only a third of those
who mitigate credibly do a confirmatory retest. Considering that over
90% of all households which need mitigation fall into this category,
these rates are alarmingly low. ;
Even given these low transition rates, however, the highest
absolute estimated number of households mitigating fall into the 4-20
pCi/1 category. This is because the base population to which these
transition rates apply is so much larger than the base populations for
33
-------
tiie higher categories. Taken a step further, the category with the
largest base rate is <4 pCi/1. We estimated that 940 households in this
category claimed to have mitigated, yielding the rather paradoxical
result that fully one-third of all those who claim to mitigate do not
even need to mitigate at all according to EPA guidelines.10 It is an
inevitable consequence of screening for low-probability events (in this
case, very high radon levels) that the absolute numbers taking action
in the rare categories, where action is most needed, will be relatively
low, and that the absolute numbers taking action in more frequent
categories, where action is less critical, will be relatively high.
In constructing the estimated transition rates and proportions
above we tried to make assumptions that would be favorable to the
evaluation of information and testing programs for radon. However,
this was not always possible. It is, in fact, probable that, although the
campaign we studied represents a state-of-the-art mass media public
information effort for radon, more localized public information
campaigns would be somewhat more effective. In fact, a more
localized and effective radon information campaign has been
* • •
conducted in Frederick, Maryland by Desvousges, Smith, and Rink
(1988). This effort integrated media messages with a community
outreach program which included presentations to community
10 It would be a mistake to conclude that all those who mitigate
after receiving test results less than 4 pCi/1 are acting irrationally,
although some may be. First, a reading of 3.9 pCi/1, for example, does
not represent substantially lower risk than a reading of 4.1 pCi/1. Also,
the EPA action level may still represent a substantial risk of lung
cancer (between 1 and 5%) over a 70-year lifetime. Finally, these
people may feel they are taking preventive measures against the
development of a radon problem in the future.
34
-------
organizations, placement of posters at various community-locations.
and a "radon awareness week," and achieved a testing rate in
Frederick of about 15%. However, if we replace the 3.6% testing rate
from our study with the 15% testing rate from the Frederick
campaign in our multistage analysis, our ultimate confirmed mitigation
rate increases only from .1% to .42%.
It is also possible that more mitigation may occur in the future as
a result of the Washington, D. C. campaign, especially given the advice
in the EPA Citizen's Guide to the effect that levels between 4 and 20
pCi/1 require action only "within a few years."H Our respondents
typically had about 9 months between the time they tested and the
time we contacted them to fill out the survey. However, on the survey
we asked respondents the specific month and year in which they first
tested their home for radon. There was, in fact, substantial variability
in the answers to this question: a few people evidently had conducted
a non^Air Chek test long before the campaign, and others clearly
waited several months after buying their Air Chek test kit to perform
the test and send it in. If in fact the proportion of those mitigating is
increasing over time, the length of time passed since their first test
should be a reliable predictor of whether or not they mitigated. We
tested this predictor in the context of the full mitigation model
Although it is true that for test results in this range the risk
does not increase substantially over a wait of a few years, our results
suggest that advising people to take action only within a few years is a
o^the1™6?-^ communication message. Given the multistage nature
of the mitigation process, if people don't take action to protect
themselves from radon immediately, they are unlikely to do so at all
35
-------
developed in Section 2.5 and found it to be nonsignificant (t = -.04,
n.s.). ,
Although we have no evidence that mitigation is increasing over
time, it is possible that people are waiting for the results of
confirmatory retests, as is recommended in the Citizen's Guide, before
proceeding with mitigation. However, only 19% of those in the 4-20
pCi/1 category (which comprises 90% of households which need to
mitigate) who did not claim to have mitigated stated that they had
performed or were conducting a follow-up test. Also, only 46% of
those in the >50 pCi/1 category who did not claim to mitigate had
conducted a foUow-up test, even though all of these people had
/
received a free retest kit from Air Chek. Even if all of these people
went on to eventually claim mitigation, in the 4-20 pCi/1 category, for
example, the ultimate confirmed mitigation rate would increase only
from .07% to about .2%.
Still another possibility is that there may be many people who
wanted to mitigate but were unable to find qualified contractors to do
the work. However, in our sample, those who mitigated and used a^
contractor gave a mean rating of 3.8 on a 1 to 7 scale to the question,
How difficult was it to find a trustworthy radon reduction contractor?
Finding a radon contractor is clearly riot extremely easy, but neither is
it extremely difficult. Although at least some future mitigation will
occur, given the ease with which people tend to forget about radon,
we view extensive future mitigation attributable to this campaign as
highly unlikely.
A final concern to our analysis is that our estimate of the
transition rate between the population needing mitigation and those
36
-------
buying test kits may be too low. For example, our estimate of the
number of households needing mitigation undoubtedly includes some
households that tested and mitigated on their own or through other
sources. Also, many of the Safeway stores participating in the
campaign sold all of their test kits. Although some of this excess
demand was met through a newspaper ad, it is possible that there
were many people who wanted to buy a test kit but couldn't. However,
even if our estimate of the transition rate to buying a test kit were too
low by an order of magnitude, the ultimate confirmed mitigation rate
for those homes needing mitigation would only be 1%. Our conclusion
concerning the ineffectiveness of the campaign is relatively insensitive
to title precise estimates we have described.
The extremely low ultimate mitigation rates resulting from the
Washington, D. C. area campaign provide convincing evidence that not
only this particular campaign was ineffective but that any radon
campaign targeted at the general population which relies only on
information, awareness, and voluntary testing is likely to fail. At the
very least, the likely credible mitigation resulting from this program
has been so small as to suggest that such programs may be a very
expensive way for society to achieve radon mitigation. ,
Although many factors likely contribute to this result, the major
one appears to be the multistage nature of the radon mitigation
process, which requires many varied actions be performed over a long
period of time and provides opportunities to drop out at any of several
stages. The majority of people simply do not make it all the way to the
end (or even to the middle) of this complex process by themselves.
37
-------
This result also raises serious questions for public policy
concerning radon. If information and awareness campaigns including
economic incentives do not achieve reasonable levels of mitigation, the
options for public policy are limited. One solution would be to adopt a
mandatory radon testing and mitigation law similar to many recent
state laws which require the use of seatbelts and child-restraint
devices in automobiles. A second solution would be to require radon
disclosure at a particular point in time during which radon testing and
mitigation might be easier than usual, during which there may be
strong preexisting incentives, for making a home risk-free, and during
which people would be available to provide homeowners active
guidance throughout the mitigation process, as part of their
professional responsibilities. The time of home sale fits this
description quite well. In a survey study conducted in Boulder.
Colorado (see Chapter 3) we found that the ultimate credible
mitigation rate for a subpopulation of people who had recently gone
through the process of buying a home was over 21%. This high
ultimate mitigation rate was found in an area in which radon has
received little media attention and there exist no state laws
concerning radon disclosure, and in a sample with maximum radon
readings below 20 pCi/1. In contrast, the ultimate credible mitigation
rate found in the target population for the Washington, D.-C. campaign
(for the comparable 4 to 20 pCi/1 category) was only .07%. Results
also showed that realtors and employers had played a significant role
in helping people get through the testing process. Although
mitigation rate estimates for the general population and a subset of
that population are not strictly comparable, these results at least
38 .
-------
suggest that the time-of-home-sale strategy might prove quite
successful in the long term as homes are bought and sold or
refinanced.
2.5. A Model of Radon Mitigation
2.5.1. Mitigation Dependent Variable
S ' : , ' '
Given the very low ultimate mitigation rates we have described
for the Washington, D. C. radon campaign, it is of prime importance
for the development of future risk communication programs for radon
to identify what factors those who do mitigate have in common. Using
various experiential, psychological, and sociodemographic variables
derived from the survey as predictors, we therefore present below a
model which predicts whether or not a person who tested and
responded to our survey subsequently mitigated.
However, as described in Section 2.4.1, there are several
possible ways of defining whether a given household has mitigated or
not. We therefore tested +hree separate models, corresponding to the
"claimed," "credible," and "confirmed" categories of mitigation we
described above, in an hierarchical fashion. That is, our first model
predicts, for all 714 respondents, who among them claimed to take
some mitigating action or actions and who did not. Our second model
predicts, for only'those 192 respondents who claimed mitigation, who
mitigated in a credible fashion and who did not. And, our third model
predicts, for only those 136 respondents who mitigated credibly, who
confirmed mitigation success with a retest and who did not. It should
39
-------
be noted that, although the ultimate mitigation rates projected for the
target population of the campaign were quite low, the proportion of
people in our sample who mitigated was quite high since we
oversampled the higher radon level categories (see Section 2.3.1).
The mitigation rates in our sample were 27% claiming mitigation,
19% mitigating credibly, and 7.4% confirming mitigation.
Since our dependent variable is categorical, there are potential
problems with the error structure using an ordinary least squares
(OLS) analysis. We therefore performed both OLS and logit analyses
for all three models. However, 'since the results as to which
predictors were important were identical from the two analyses (e.g.,
see Table 2.4) and OLS parameters are easier to interpret, we will
focus on the results of the'OLS analysis in our discussion.
2-5.2. Potential Predictors of Mitigation
We had several variables available from our survey which were
candidates for predictors of mitigation, and we included fourteen of
them in the model of, claimed mitigation and added one other in the
models of credible and confirmed mitigation. These predictor
variables are described below in the following conceptual groups:
objective risk, subjective perception of risk, characteristics of the
individual, characteristics of the home, individual's behavior for other
risks, and experience with the risk.
Of course, the objective risk, the radon level measured by a
household's Air Chek test, is expected to be an important predictor of
mitigation. In our model we used the maximum radon level for the
40
-------
household reported by either the respondent (for any test results they
received) or by Air Chek (for the test kit selected in the sample), as
we did for the mitigation pathway analysis in Section 2.4.2. l.liz Also,
since radon readings varied over three orders of magnitude, we used
the logarithm of this maximum radon level in our model to correct for
positive skewness (Judd and McClelland, 1989).
A common finding in the risk perception literature is that
people's actions are often based more on what they perceive the level
of risk to be than on the true level of objective risk. Two radon
studies, Johnson and Luken (1987) and Weinstein, Sandman, and
Roberts (1988), in fact found "perceived seriousness of the risk" to be
-v
a significant predictor of mitigation. In our study we did not attempt
to obtain a pre-mitigation measure of subjective risk since we did not
contact our respondents until after their mitigation decision had been
made. However, we did ask them to state their degree of confidence
in the accuracy of their test result. We included their responses to
this question in the model, although it certainly represents only one of
many components that people likely use in developing subjective risk
estimates.
12 There are potential problems with accepting self-reports of
radon test results in determining the maximum radon level, since
people may be subject to reporting errors, especially given the typical
nine-month lag between receiving the test result and responding to
the survey. However, as shown in Section 2.6, if people do err they
tend to underestimate, not overestimate, the true test result. And, in
any case, we would expect their actions to be based on what they
thought the true radon reading was, if this in fact differed from the
true reading. Also, the substantive conclusions of the model are the
same whether the Air Chek level or the self-reported level or the
maximum is used as the dependent measure.
41
-------
Mitigation choices may vary as a function of several
sociodemographic characteristics of the individual. For example,
older respondents may be less likely to mitigate because their risk is
much more heavily determined by their prior rather than future
exposure to radon. Or, people with higher incomes might be more
likely to mitigate simply because they are in a better position to pay
the expenses incurred by mitigation. Although we do not have specific
hypotheses about the effects of these variables on mitigation, we
examine gender,-age, number of children, education, and income as
predictors in our model. '
It is also possible that characteristics of the home, not just of the
individual, may have an effect on mitigation behavior. We therefore
included a categorical variable indicating type of home (single-family
detached versus any other type) and the value of the home as
predictors in the model. Because home values are unrestricted on the
upper end of the scale, reaching into the millions of dollars, we used
the logarithm of the reported market value of the home in the model
to correct for positive skewness.
There may also be general personality factors in people's risk
responses that play a part in radon mitigation: some people may
simply be more risk averse than others for many different kinds of
risks. We therefore included two behavioral measures of responses to
other risks in our model which we thought might represent proxy
variables for risk aversion: whether or not the respondent generally
wears a seat belt and whether or not the respondent smokes.
Experience with the risk may have an important influence on
the decision to mitigate. As a crude measure of benign experience
42
-------
with radon, we included in the model the number of years the
respondent had lived in the residence which they tested. People may
also draw upon experience gained through others in responding to
radon risk. We therefore included two composite variables in the
model to try to capture respondent's exposure to radon information
and advice. The "media influence" variable represents whether and
how often people saw, for example, TV news shows or public sen/ice
announcements about radon, radon brochures in public places, radon
information in utility bill inserts, and community presentations on
radon. The "social influence" variable represents whether and how
often people discussed radon with family members, neighbors, real
estate agents, and others. •
• For the credible and confirmed mitigation models we included
another predictor, namely, who performed the radon reduction
measures on the home. We expect that those who claim to mitigate
might be more* likely to do so credibly and to retest afterward if they
hire a professional to perform the work rather than doing it
themselves. This variable was coded -1 if mitigation was conducted by
the respondent and +1 if conducted by a private contractor.
2.5,3. Results
s
2.5,3.1. Claimed Mitigation Model
Table 2.4 gives the partial regression coefficients, their
associated t statistics, and PRE (the proportional reduction in error
achieved by adding that predictor variable last, also known as the
43
-------
Table 2.4. OLS Model of Radon Mitigation3-
Variable
Claimed mitigation
(1 =yes, -1 =no)
Constant
Log maximum radon level
Confidence in test accuracy
(1-7 scale)
Gender
(1 = female. -1 = male)
Age (years)
Number of children living
at home
Education
(1-10 scale)
Income
(1-11 scale)
Type of home
(1 = single fam.. -1 = other)
Log home value
Respondent is smoker
(1 « no, -1 = yes)
Respondent wears seatbelt
(1 =no. -1 =yes)
Years lived in home
Media influence
(1-16 scale)
Personal Influence
(1-12 scale)
Sample size
f
Mean Std. Dev.
Dependent Variable
-.45 .89
Predictor Variables
2.51 1.48
4.62 1.6O
-.32 .95
46.12 12.99
•69" 1.01
731 1.72
6.32 2.37
•82 .57
12.06 .47
^
.74 .67
-.85 .53
9.60 8.68
5.28 3.51
2.51 2.00
-2.245
.235
(.406)
.073
(.102)
-.099
(-.195)
-.004
- (-.005)
.013
(.012)
.015
(.026)
-.012
(-.014)
-.015
(.063)
.067
(-177)
.022
(.028)
.0006
(-.034)
-.001
(-.004)
-.034
(-.046)
.113
(.167)
t
-2.37
9.35—"
(7.96)
3.44—
. (2.80)
-2.72"
(-2.90)
-1.03
(-.75)
.35
(-20)
.67
(.66)
-.72
, (-.47)
-.24
(-44)
.79
(1.21)
.43
(.30)
.009
(-.26)
-.21
(-.46)
-2.31*
(-1-79)
5.1O— «
(4.45)
,
PRE
.139
(.094)
.021
(.009)
.013
(.01)
.002
(...)
.0002
.0008
(...)
.001
(...)
.0001
(...)
.001
(...)
.0003
(...)
(...)
(...)
.01
(.002)
.we
(.027)
556
.27
a Logit results appear below the OLS results in parentheses
,*E.<.05 "fi<.01 —n<.001 — *D< 0001'
-------
coefficient of partial determination; see Judd and McClelland (1989))
for both the OLS and logit13 analyses of claimed mitigation.14 In
interpreting these values it should be remembered that the statistical,
tests are for partial regression coefficients. That is, the test asks
whether the given variable reliably explains a portion of the variation
in mitigation behavior after controlling for all the other variables
included in the model. With covariation among the predictor variables
this can produce conservative conclusions about the importance of a
variable. Note that approximately 27% of the variation in the coded
claimed mitigation variable is accounted for by the model. This is
substantial for a model of this type, especially given that the
dependent variable is binary. • : . -
As expected, log maximum radon level is a highly significant
predictor of mitigation (t = 9.35, rj < .0001), This is consistent with
the findings of Ackerman (1988). Moreover, the effect is
comparatively large, reducing 13.9% of the variance even after
controlling for all the other predictors in the model. For a variable
that predicts actual behavior on a binary dimension several months in
the future, this is an impressive result. The strength of the linear
13 The logit results were obtained using the LOGIST procedure
in SAS (see "SUGI Supplemental Library User's Guide," Version 5
edition, 1986, SAS Institute, Inc., Gary, North Carolina, pp. 269-294).
14 The number of respondents represented in this model is 556,
not the total sample of 714, because 158 respondents were deleted
due to missing data on one or more of the 14 independent variables.
Deleting the three variables with the highest number of missing values
(income, home value, and education) resulted in 87 more respondents
being included in the model, but this did not change the substantive
conclusions concerning any of the predictors. We therefore feel
justified .in presenting the full model with 14 predictors.
45
-------
relationship between mitigation and log radon level in this study
suggests that the failure of Johnson and Luken (1987) to find such a
result in their study was simply due to their lack of opportunity to
oversample high radon levels. .
Our result is harder to reconcile, however, with the conclusions
of Weinstein, Klotz, and Sandman (1989). In their study of New
Jersey homeowners, they reported a significant but modest-sized
linear relationship between "future remediation plans" and radon level.
However, when they dichotomized the radon level variable into a
crude measure of whether or not the test result was above or below
the EPA action level of 4 pCi/1, they found the correlation between
mitigation and radon level to increase substantially. They concluded
that there was a modest relationship between objective risk and
remediation plans, but that this correlation was "entirely due to the
distinction of being above or below the guidance level of 4 pCi/1." We
regard this result as puzzling, since dichotomizing a continuous
predictor usually produces a loss in statistical power if the standard
distributional asssumption of normality holds true. We retested our
full model of claimed mitigation replacing log radon level with the
same dichotomized predictor used by Weinstein, Klotz, and Sandman,
and found the dichotomized predictor to still be highly significant but
to be less significant than the continuous predictor and to explain a
much lower proportion of the variance (£ = 4.9, p. < .0001, PRE =
.042). Because Weinstein, Klotz, and Sandman removed people who
had already mitigated from their analysis of "future remediation plans,"
they likely had an abnormally large proportion of people in their
remaining sample with very high radon levels who did not intend to
46
-------
mitigate. These "outliers" may have weakened their ability to detect a
stronger linear relationship between mitigation and radon level in
their data. In short, we believe we have clear evidence from our
sample that the proportion of those mitigating systematically increases
as radon level increases above 4 pCi/1.
We turn now to a consideration of which predictors of mitigation
explain variation Over and above the contribution of the objective risk
level. Confidence in the accuracy of radon tests was a reliable
predictor, with those having more confidence being more likely to
mitigate. This verifies the contribution of subjective interpretations of
risk in predicting behavior and also suggests that some people may be
using lack of confidence in the accuracy of radon tests as an excuse to
"> ' ' :
drop out of the mitigation process.
Among the sociodemographic variables, age, number of children
living at home, education, and income had inconsequential effects in
our model. This, however, illustrates an interesting point, since
models which predict perceived risk from radon,typically find, for
example, that perceived risk decreases with age and increases for
households with children (Sjoberg, 1989). Clearly mitigation behavior
displays a different pattern of important predictors than perceived
risk.
Gender, however, was a significant predictor in our model, with
women being less likely to mitigate than men. This result is in direct
contrast with both previous surveys on radon (Sjoberg, 1989) and on
other risks such as nuclear energy (Kasperson, Berk, Pijawka, Sharaf,
and Wood, 1980) and industrial hazards (Stallen and Tomas, 1988),
which have found women to be more concerned than men about risk.
. 47 . :
-------
However, since our model predicts mitigation behavior and not
concern, it is possible that women in our study were still more
concerned even though they were less likely to take protective action.
In fact, this result may not indicate a gender difference at all, but
simply reflect the fact that many women have less money to spend on
radon mitigation than men or have less experience doing the kinds of
household repairs that radon mitigation requires. In any case,
although gender was a reliable predictor in our model, it's impact was
relatively small (PRE = .013).
Type of house and log of home value were not significant
predictors in our model, although there may not have been sufficient
variation in house type in our sample to test its effect adequately. The
two variables measuring behavior for other risks (wearing seatbelts and
smoking) were also nonsignificant, although again there may not have
been sufficient variation on these dimensions in our sample. In their
study of Maine households, which essentially oversampled smokers,
Johnson and Luken (1987) found smokers to be significantly less
likely to mitigate.
Years lived in home was not a significant predictor of mitigation
in our study, although it was in Johnson and Luken's study (those who
had been in their homes longer were less likely to mitigate). The
composite media influence variable was significant, with people having
more media contact about radon being less likely to mitigate. Also, the
composite personal influence variable was significant, with people
having more discussions concerning radon with others being more
likely to mitigate. Of course, these latter two variables are difficult to
interpret due to the causal ambiguity inherent in a cross-sectional
48
-------
survey of this type. For example, the counterintuitive result for media
influence might be explained by postulating that people who have
mitigated no longer need to attend to or seek out radon information
from media sources or that people who mitigate for some reason are
also people who spend less time watching television. And, the process
of mitigating is just as likely to promote discussions with others about
radon as such discussions are to promote mitigation.
2.5.3.2. Credible Mitigation Model
The credible mitigation model predicts, for those who claimed
to mitigate, who among them mitigated credibly (reported spending
money and could identify the specific reduction method used) and
who did not. The modeling procedure was identical to that described
for the claimed mitigation model. Only '152 'of the 192 respondents
who claimed mitigation were included in the model due to missing
data on one or more of the 15 predictors. The full model accounted
for 13.7% of the variance in the coded credible mitigation variable.
Log radon level was .not significant in this model (£ = 1.14, o.=
.26, n. s., PRE = .01), and in fact only two of the 15 predictors were
significant. The gender difference found in the claimed mitigation
model was also significant in this model in the same direction (bj =
-.23, t = -2.5, p. < .05) but was again a relatively small effect (PRE =
.04). In our study women were less likely to claim mitigation, and, if
they did claim to mitigate, were less likely to do so credibly. Smokers
were also less likely to mitigate credibly in this model (b| = .29, i =
2.07, p. < .05). In our study smokers, although not significantly less
49
-------
likely to claim mitigation, were less likely to mitigate credibly, given
that they had claimed to mitigate.
Surprisingly, those who hired a private contractor to do the
mitigation work were not reliably more likely to have mitigated
-..
credibly in this model (t = -0.44, p. = .66, n. s.). However, our
definition of credible mitigation was quite liberal and included all
those who reported spending any money on mitigation, even as little
as $5 or $10. It is possible that those who mitigated by themselves
tended to choose low-cost, uncertain mitigation methods such as
painting and sealing, whereas those who employed a contractor
tended to choose higher-cost, more reliable mitigation methods such
as forced ventilation, pressurization, or soil suction. To investigate
this idea we retested our credible mitigation model using a stricter
definition of credible mitigation: in this revised model only those who
reported spending more than $50 were considered to have mitigated
credibly. Given this revision those who hired a contractor were
significantly more likely to have mitigated credibly than those who did
not (bj = 0.28,1 = 2.7, D. < .01, PRE = .052). Although this argument is
somewhat circular, it illustrates an important point: people either hire
a contractor and spend a lot of money on a reliable reduction method
or they do not hire a contractor and spend very little money on a less
reliable reduction method. People generally do not implement the
most effective mitigation techniques on their own.
50
-------
2.5.3.3. Confirmed Mitigation Model
The confirmed mitigation model predicts, for those who
mitigated credibly, who among them retested to confirm mitigation
, success and who did not. The modeling procedure was identical to
those described above. Only 111 of the 136 respondents who
mitigated credibly were included in the model due to missing data on
one or more of the 15 predictors. The full model accounted for 42%
of the variance in the coded confirmed mitigation variable.
Log radon level was not significant in this model (t = 1.67, p_ =
.098, n. s.), although it did account for 3% of the variance after
controlling for the other predictors. In fact, only a single predictor in
this model, who performed the mitigation work, was significant at the
.05 level. Those who hired a -private contractor were more likely to
retest after mitigation than those who did not (bj = .45, t = 4.27, n. <
.0001). This was a strong effect, reducing the variance by 15% even
after controlling for the effects of the other predictors. Evidently
people simply do not generally retest after they have mitigated unless
the mitigation was performed by a professional contractor. This
indicates a failure of EPA's message about retesting.
2.5.4. Conclusions
Clearly radon level was the most important predictor of claimed
mitigation in our model, and the chance that a household will mitigate
increases systematically as radon level increases above 4 pCi/L The
strength of this relationship between mitigation and radon level also
suggests that low mitigation rates cannot be explained simply by
51
-------
postulating that people's protective responses to radon are insensitive
to the level of objective risk. However, radon level did not predict
who among those who initially did something to mitigate would do so
most effectively and completely.
Who performed the mitigation work was clearly the most
important predictor of who would mitigate credibly and who would
confirm mitigation with a retest, for those who had claimed to
mitigate. People who mitigated by themselves were more likely to
choose low-cost, less-effective mitigation methods and were less likely
to retest after mitigating than those who employed a professional
contractor. Apparently very few people make it all the way through
the mitigation pathway without help. . -
After controlling for the "effects of radon level, belief in test
accuracy, and who did the mitigation work, none of the other
predictor categories, including characteristics of individuals,
characteristics of homes, proxies for general risky behavior, arid
experience with the risk, were able to account for a substantial portion
of the variation in mitigation behavior (although certain specific
predictors such as gender and smoking sometimes had reliable,
though small, effects). This was especially true for the final two stages
in the mitigation pathway. One is left with the impression that
protecting oneself from radon is simply a very difficult task.
i
2.6. Accuracy in Self-reports of Radon Level
Thus far we have characterized the radon mitigation process as a
long and difficult one. We have also been able to document several
52
-------
major steps along the pathway to successful mitigation at which people
drop out: failure to return a test kit once it has been purchased, failure
to choose effective mitigation methods, and failure to retest after
mitigating. However, each of these stages likely encompasses many
(perhaps dozens of) smaller stages which our survey was unable to
document. For example, simply "returning a test kit for evaluation"
involves following instructions to set up the test correctly, getting the
proper exposure, resealing the .container properly several days later,
mailing the test back to Air Ghek, and attending to and interpreting
the results. "Mitigation" might involve seeking information from
several sources by phone or mail, asking many different people for
advice, going through the entire testing process all over again once or
twice, doing some banking and accounting, and making several trips
to the hardware store.
One of these "microstages" we do have information on is people's
ability to remember their radon level correctly. If people do not
remember or if they misremember their radon level, they might drop
out of the mitigation process because they mistakenly think their level
is safe, or may perhaps -mitigate unnecessarily if they mistakenly think
their radon level is high. To investigate this question we tested
whether the difference between people's self-reports of their initial
radon level and the radon level reported to them by the testing
company was significantly different from zero. Because we only had
access to one test kit reading from Air Chek for each respondent, and
many respondents reported using more than one test kit, we included
in the model only those 499 respondents who reported using one test
53
-------
kit. Of these 499 respondents, 125 (25.5%) were unable to state their
radon test results in picocuries per liter on the survey.
For those 374 respondents who did report a radon level, the
mean difference score was significantly less than zero (mean = -2.39
pCi/1, 1 = -3.5, rj < .001, PRE = .03), indicating a reliable tendency for
self-reports to underrepresent true radon readings. However, this
simple interpretation is misleading, which can be seen from the
pattern of the errors associated with the model. Figure 2.4 displays a
plot of self-reporting error versus Air Chek radon level for the 374
respondents in the model. Clearly, the typical person did not
underreport his or her radon level by 2.39 pCi/1. Instead, many
people knew their radon level exactly, or nearly so. However, if they
did make an error, it was1 likely to be an underreporting error, and the
pattern of errors is markedly heteroskedastic. That is, the size of the
errors increases as the true radon level increases.
Of course, we don't know if remembering an incorrect radon
level makes people less likely to mitigate or if those who do not
mitigate are less likely to ^remember their radon level correctly.
However, this analysis at least raises the possibility that some people
may not be mitigating simply because they misremember their radon
level. In fact, of the 21 respondents in our survey who
underrepresented their true radon reading by 25 pCi/1 or more, not a
single one mitigated credibly. This is a good illustration of how an
error at one small, simple stage along the mitigation pathway can lead
people to drop out of the process.
54
-------
o
LU
0) S
c O
r &
o
Q.
0
i
"o
(/)
100-
75 -
50-
, 25-
•".••'•'. . . •
.
•
" * •* • * *
o P^yy\F^ m ^"* L'*' • • "••• •
'»K -
"
-50-
-75-
_irin -
• • • v •
• •
• • •
*«••* .*
• •
- • ' ••
020 40 60 8 0 1 001 20 1 401 60
Air Chek Radon Level
Fig. 2.4. Self-reporting error versus Air Chek radon level for
respondents who reported using one test kit and who reported their
radon level in picocuries per liter.
2.7. Conclusion .
We have presented several kinds of evidence that radon is a,
particularly intractable risk in the sense of being difficult to manage
and manipulate and being resistant to well-meaning efforts at
reduction. A radon testing and information campaign aimed at the
general public was shown to result in very low ultimate mitigation
rates. Many of those who claimed to mitigate did not do so in an
effective way. Those who received professional help during the
mitigation process, however, were much more likely to mitigate
credibly and to confirm that their mitigation efforts had succeeded.
Only just over half of those who purchased radon test kits even
returned them for analysis, and many of those who did test could not
recall their radon reading or recalled it incorrectly.
55
-------
Although characteristics of radon risk such as benign experience
and lack of perceptual cues (see Chapter 1) likely play a significant
role, the major determinants of low mitigation rates appear to be
instead characteristics of the protective responses required to reduce
the risk from radon, which in general are complicated and difficult.
It has been suggested that people would be more concerned about
radon if it had an identifiable odor such as that which is added to
natural gas. Similarly, many more people would likely act to reduce
radon levels if mitigation were as simple as putting on a seatbelt.
It is also interesting to consider two home risks with solutions
that have very high adoption rates. These are crime and fire. It has
been estimated that about 40% of households have purchased security
devices such as bars, special locks, and burglar alarms to prevent
crime (Dubow, McCabe, and Kaplan, 1979) and about 60% of
households (in areas where they are not required by law) have
installed smoke detectors (Jensen, Tome, and Darby, 1989).
Admittedly, these risks are very different from radon in some ways, for
example, they are often highlighted on the evening news, they are
dramatic and perhaps "dread" risks, and many people might receive
economic incentives (lower insurance premiums) to take protective
action. Still, it is likely that one contributing factor to these high
adoption rates is that the protective behaviors required are relatively
simple, known, one-time behaviors. In the case of fire, for example,
you need only buy a smoke detector from the hardware store, install.it
like you did your ceiling lights, and put in batteries like you did for
your tape recorder. Although this task may be more difficult for some
people than for others, years of protection can be gained from a single
56
-------
afternoon's effort. In contrast, there is simply no way to reliably
reduce one's risk from radon that takes .less than several weeks.
Protecting oneself from radon is a long, difficult, multistage,
multifaceted, technical, multiperson, multiskill, recursive process.
The following list details several characteristics of the protective
reponse to radon which we feel govern radon mitigation behavior just
as the risk characteristics we presented in Chapter 1 appear to govern
radon risk perception:
1.
2.
3.
4.
5.
8.
months or
iS a multiftage Process, with successful completion of
one stage necessary for completion of subsequent stages; people
r° aitWa thr°U the Pr00ess ^ any of several P
IS a ^tiposon activity, requiring contact and
testing comPany employees and
govePnt h"?S See!dng information from neighbors,
government employees, employers, or realtors.
Radon mitigation requires a wide variety of skills and qualities (for
sSLbil^P'T nal f^ll Patience' Pe"istence, home repair
skills, ability to interpret technical information) which may not
often occur together in a single household.
AU^of the effort and expense for a lifetime of protection occurs up
be rfrn S recursi1Te' ^quiring the same procedures
be performed as, many as three or four different times.
57
-------
It is no surprise that many people do not make it all the way through
the mitigation pathway when the necessary protective responses are
characterized in this way.
Other risks which share many of these protective response
characteristics may also prove to be extremely difficult for people to
deal with and protect themselves from without help. Such a list of
intractable risks might include, for example, asbestos fibers from
building materials, lead in drinking water, contamination of wells by
toxic, chemicals in groundwater, vapors from urea-formaldehyde foam
used for insulation, and a host of other indoor air pollutants.
Traditional information and awareness campaigns are unlikely to be an
effective public policy for such risks. They may instead require
regulation or the recruitment of professionals who can provide people
with active guidance and personal assistance.
58
-------
Chapter 3
An Evaluation of Radon Testing and Mitigation at Time
of Home Sale
3.1. Introduction
Given the low overall effectiveness of an information and
awareness program targeted at' the general population, as shown in the
previous chapter, this chapter presents a study which explores the
potential effectiveness of targeting radon information and/or
regulation at the point of'home sale. Since single family homes are
sold about every ten years on the average, this approach cannot yield
an immediate solution to the radon problem. However, if individual
or institutional behavior in the setting of the home sale frequently
leads to testing and mitigation, this approach might provide an •<-
effective long run solution.
Thus, the purpose of the study reported in this chapter is to
conduct a telephone survey of recent home buyers in Boulder County,
Colorado to determine if testing at the time of home sale has become
common practice and if such testing leads to mitigation. It should be
noted that no extensive information and awareness campaign has been
conducted in the State of Colorado, so any testing that occurs is
motivated by generally available radon information. For example, news
stories indicating that Colorado has the highest percentage df homes
59
-------
across the nation violating the EPA guideline for radon have appeared
in both-newspapers and on television. If substantial mitigation is
already occurring in Boulder County, and we can determine the
underlying causes, that result would suggest that the point of home
sale is an effective time to attempt to encourage further testing and
mitigation in other areas.
The research presented here is based on a sample of 303 homes
for which telephone interviews were completed. Because the
frequencies of testing and mitigation by home buyers and sellers were
unknown, it was our view that it would be unwise to expend a large
amount of money sampling for what might be low frequency events
(i.e., testing and mitigation). For example, had we sampled the
general population rather than people who had tested for radon in the
study reported in the previous chapter, less than 10% of the sample
would have been of interest (i.e., purchased test kits and possibly
mitigated as a result of the program). Fortunately, in our study we
found that 44.6% of a sample of recent home buyers in Boulder County
had obtained radon test results prior to closing on their new home.
Home sales in and of themselves appear to motivate testing.
The rest of the chapter is organized as follows. Section 3.2
describes the design of the telephone survey, while Section 3.3
presents the sampling plan. Sections 3.4 and 3.5 give descriptive
results and present our data analysis, respectively, while Section 3.6
contains our conclusions and recommendations.
60
-------
3.2. Survey Design
The survey instrument used in this field study is reproduced in
Appendix II, along with summary results for each question. Where
appropriate, results are provided for both those who tested and'those
who did not. The survey was designed to be administered over the
telephone by interviewers who were not knowledgeable about radon
and to take no more than 10 or 15 minutes to complete.
The interviewers were instructed to speak only to the property
owner whose name we had obtained or to a co-owner of the property.
Respondents were then screened to confirm that they had recently
purchased property in Boulder County and that this property was their
primary place of residence. Respondents were read a short statement
that stated that the survey concerned radon gas, that it would take
about 15 minutes to complete, and that all responses would be
completely confidential. The interviewers were provided with three
additional statements to read if the respondent resisted or wanted to
know how they were chosen. Respondents with questions about the
survey or problems were given the name and telephone number of One
of the investigators.
All respondents were asked 23 questions, and 34 additional
questions were asked of some respondents to collect more detailed
information concerning those whose homes had been tested for radon.
The questions were selected to provide information on three
interrelated areas: ' • .
61
-------
1.
2.
3.
What percentage of recently sold homes are tested for
radon and what percentage of these homes make it through the
entire mitigation process (especially in comparison to the
corresponding percentages obtained in the information and
testing campaign detailed in Chapter 2)?
What is the story behind why radon tests at time of home sale
occur and how does radon affect the sale negotiation process?
What differences are there between people whose homes are
tested and people whose homes are not tested at time of sale?
, The survey questions were organized into six sections. The first
section was a general introductory section designed to gather some
information on respondents' experience with radon prior to finding
their new home. The second section was designed to determine if and
when a radon test occurred and to ask detailed questions about how
and why the test was done for those whose homes were tested. (Since
pretesting led us to expect that people might have difficulty reporting
the results of the radon test in picocuries per liter, respondents were
prompted with the phrase "remember that the EPA action level is 4
picocuries per liter" and were asked to rate the test result on a 5-
point scale ranging from "very low" to "very high" if they could not
report the exact result.) The third section asks if and when any
attempt was made to reduce the radon level and also asks detailed
questions about how and why the reduction effort was conducted for
those homes that underwent mitigation. The fourth section asks
specific questions concerning how radon enters into and affects the
sales transaction for those whose homes were tested and documents
the types of clauses concerning radon that are presently being added
to sales contracts in Boulder County. The fifth section asks general
62
-------
questions concerning transactions and the negotiating process in
order to assess what effect, if any, radon may have on other, more
general, aspects of the sales process such as length of time to reach an
agreement and final negotiated sales price. Also, one aspect of the
transaction process, whether a realtor was employed and, if so, what
advice the realtor gave the buyer about radon, was assessed to
determine how it might affect the likelihood of a test being conducted.
The final section covers the standard demographic information as well
as-some additional demographic information of special relevance for
radon. A few questions are designed to determine whether the
respondent has a general tendency to engage in risky behaviors such
as not wearing a seat belt when driving. Also, respondents are asked
whether they work for IBM because of IBM's radon program, which
requires testing and mitigation to below 4 pCi/1 for employees in
order for them to participate in the company's housing buyback
program in the event they are transferred away from Boulder. IBM
employees are therefore in a very different situation than non-IBM
employees with respect to the options open to them concerning
radon, and may need to be analyzed separately.
In addition, the fact that many real estate agents in the Boulder
County area are knowledgeable about radon may be partially due to
their having to deal with IBM employees as customers. The presence
!
of such a major company with a very strict radon policy may therefore
be influencing radon testing and mitigation behavior in Boulder County
even for non-IBM employees. Testing and mitigation rates at time of
home sale in Boulder County might be somewhat elevated in
comparison to communities in which no major employer with a radon
63
-------
policy exists. However, these rates are surely lower than for many
communities in Pennsylvania and New Jersey in which realtors have
been involved with radon for several years and in which realtor
associations have adopted radon policies. And, in any case, the rates
in Boulder County, which have been motivated by private market
interests, are surely minimum estimates of what could be achieved
with a government-sponsored risk communication strategy at time of
home sale.
3.3. Sample Design
The sampling plan was based on inclusive lists of individuals who
had recently bought homes in Boulder County. (It should be noted that
we attempted to contact all of the homes recently bought in Boulder
County during a specific period of time, and therefore our sample
included the entire population of interest, not a smaller randomly
chosen subset of that population.) The lists of names were taken from
a local newspaper which publishes all Boulder County property sales
on a weekly basis and provides the property buyer's name and the
address of the purchased property. Buyers of commercial property
and apartment buildings were deleted from the lists. Directory
assistance was used to obtain phone numbers for each person on the
lists. Overall one-third of the names were deleted from the lists
because they had bought commercial property rather than residential,
they did not yet have a phone number, they had an unlisted phone
number, they were not qualified to participate in the survey, or their
phone numbers turned out to be wrong or disconnected.
64
-------
An initial administration of the survey took place in December of
1988. The goal of this first administration was to obtain a pilot sample
to determine if collection of a larger sample would be useful., -• A list of
195 people with good telephone numbers who had purchased homes
from September to November of 1988 was prepared as outlined above,
and 100 of these 195 calls were successfully completed. 15
A second administration of the identical survey was conducted in
February of 1990 to increase the sample size and response rate of the
initial pilot survey. 16 All of the 95 people who did not complete the
survey in its initial administration were tried again, and 51 of these 95
people completed the survey at this time. In addition, a list of 301
new names of people with good telephone numbers who had
purchased homes between December of 1988 and early February of
1989 was prepared, and 153 of these people completed the survey.
Data from both of these administrations is combined in the
results, and analyses to be presented below. Overall, 303 surveys were
completed from lists containing 496 new home buyers for which good
is In this initial sample an attempt was made to oversample areas
of Boulder County in which we expected to find more IBM employees,
since we wanted to ensure adequate numbers of IBM employees in the
sample to investigate the impact of IBM's radon policy. However, the
percentage of IBM employees did not vary between the area which was
oversampled and the other areas of Boulder County (10% versus 13%,
respectively). We therefore did not analyze data from these areas
separately and also did not oversample any areas in the second
administration of the survey.
16 The only changes made to the survey were (1) changing the
phrase "new home" to "present home" because over a year had passed
since the first administration and (2) including a new screening
question which identified people who had moved again since.the
winter of 1988-1989.
65
-------
telephone numbers could be located, yielding an overall response rate
of 61.1%. We have no theoretical reason to expect that tho~se we were
unable to locate a telephone number for should differ from
respondents in any way. We did, however, expect that those who
refused to answer the survey when contacted might differ significantly
from respondents, perhaps having a lower testing rate. We therefore
asked these people a single simple question, Was your present home
tested for radon before the sale closed? Of the 80 people who were
contacted and refused to complete the survey, 66 did answer this one ""
question. The testing rate before closing for this group of refusals was
27.2%, significantly lower than the rate for respondents. We assess
the impact of this lower testing rate for nonrespondents in Section
3.5.
Since the sampling plan described above resulted in three
distinct samples (homes in the initial sample contacted in December
of 1988, homes in the initial sample contacted in February of 1990,
and homes in the second sample contacted in February of 1990), we
compared these samples on several key variables to determine if there
were any major differences. There was little difference between the
three samples on whether or not they employed a realtor (for all three
samples between 77 and 83% employed a realtor) but there was some
difference in the percentage of testers with test results greater than 4
pCi/1 or "above average" (between 16 and 29% for all three samples).
The testing rate before closing for the second sample (50.7%) was
higher than for the original sample (39%), indicating that testing rates
at time of home sale may be increasing over time. In addition, the
testing rate after closing for people in the original sample who were
• 66
-------
contacted in February of 1989 (19.6%) was higher than that for people
in the original sample who were contacted in December of 1988,
which likely reflects the fact that those in the former sample simply
had another year in which to conduct tests.
3.4. Descriptive Results
A facsimile survey is presented in Appendix II which includes
the number of responses and the overall descriptive results, either
means or frequencies as appropriate, for each survey question. When
the comparison is useful, separate results are also reported for those
respondents whose homes were tested (either before or after the sale
closed) versus those whose homes were not tested. We shall briefly
discuss some of the highlights of these results below. Those
respondents whose homes were tested will be referred to as "testers"
and those whose homes were not tested will be referred to as
"nontesters" in the following discussion.
3.4.1. Introductory Questjrons
Responses indicate that 98% of those surveyed had at least
heard of the radon problem. Overall, over half of the respondents first
heard about radon more than 1 year ago. Testers were slightly more
likely to have first heard about radon more than one year ago and
slightly more likely to have first heard about radon before they began
their most recent home search, indicating that knowledge about radon
prior to buying a home may help people in making appropriate choices
concerning radon at time of home purchase. Question 3 documents
67
-------
the large variance and bimodality in people's attitudes toward radon:
one-fifth of the respondents indicated that it was "not at all important"
to them to buy a home with a safe radon level, whereas another one-
fourth of the respondents indicated that it was "very important."
Responses to question 6 indicate that there is little difference in the
generl home buying experience of testers versus nontesters, with
about 60% of each group having purchased a home once before. Not
surprisingly, testers rated the importance of buying a home with a safe
radon level higher than nontesters (means 5.3 vs. 3.1) and were much
more likely (47.3% vs. 22%) to have had their previous home tested
for radon. Evidently those who have experience with radon testing in
one real estate transaction do tend to make- testing part of their next
real estate transaction.
3.4.2. Radon Testing
Overall 154 of 303 respondents (50.8%) indicated that their
new home had been tested for radon. Twenty six (8.6%) of these tests
took place before the buyer first looked at the home, 109 (36%) of the
tests took place before the sale closed (and typically after negotiations
had begun), and 19 of the tests (6.3%) took place after the closing.
The buyer was the most likely person to have first requested the test
(67.5% of the time) and to have paid for the test (51.3% of the time),
but many other categories of people, including sellers, realtors,
contractors, employers, and bankers are represented in the responses
to these questions. The radon test was conducted by a professional
(either a realtor, building contractor, or radon testing firm) 48% of
68
-------
the time. Of the 154 testers, 73 of them were able to state their test
results precisely in picocuries per liter, and almost all testers could at
least describe the test results in such terms as "low," "about average,"
or "high," Of the 73 test results stated in picocuries per liter, the
mean result was 4.2 pCi/1, the lowest result was 0.2 pCi/1, and the
highest result was 18.0 pCi/1. Of the 81 test results not stated in
't ' ' •
picocuries per liter, 64.2% were described as "low" or "very low,"
14.8% were described as "average," and 7.4% were described as,
"high." In the analysis that follows, both numeric and descriptive test
results are combined into a single measure, whether or not the test
result was below or above the EPA action guideline of 4 picocuries per
liter. Descriptive results were fit into this framework by assuming
"very low," "low," .and "average" results to be less than or equal to 4
pCi/1 and "high" and "very high" results to be greater than 4 pCi/1. .
Testers seemed to have a relatively high degree of confidence that
their test results were accurate, rating their confidence, on average, at
5.5 on a 1 to 7 scale.
3.4.3. Radon Mitigation
Twenty four respondents of the 154 who tested indicated that
some attempt at mitigation had been made (answered yes to the
question, Was any attempt made to reduce the radon level in your new
home before the sale closed?), and 14 of them were able to state
exactly how much money had been spent on the mitigation effort.
Overall, 16 of the 31 testers with test results greater than 4 pCi/1 or a
"high" result claimed to mitigate, and for 14 of these 16 the success Of
69
-------
the mitigation effort had been evaluated by a retest. Again, over half of
the time the work involving radon was performed by professionals.
The maximum post-mitigation test result reported was 5.0 pCi/1 and
the mean was 2.8 pCi/1, indicating that the mitigation efforts were
generally effective.
3.4.4. Radon Transactions and Negotiation
Only 2 of the 154 testers (1.3%) reported that they had
negotiated a reduction in the sale price of their new home "specifically
because of concern about radon or because of radon test results." One
of the major concerns, from a public health perspective, about a risk
communication at time of home sale strategy without enforced
mitigation is the potential for new homeowners to accept monetary
compensation in place of mitigating the radon hazard, which would
simply leave seUers poorer and result in no reduction in radon risk for
the buyer. Based on our survey results, it appears that people simply
do not, or at least only extremely rarely, resolve a radon problem with
a seller by accepting a price reduction in place of mitigation hpfnrp
closing.
About 40% of the testers reported that there was a clause
concerning radon in the final sales contract for their home. Of these
contracts with a radon clause, 90.2% of the time the clause stated that
the sale was contingent on a radon test, 52.5% of the time the clause
stated that the seUer must reduce the radon level if found to be
unsatisfactory, and 52.5% of the time the clause stated that the home
70
-------
must meet a specific radon level (16 contracts used the EPA action
level of 4 pCi/1 and one used 20 pCi/1).
Another concern about the time of home sale strategy is that
sellers, who are often still living in a home during negotiations, might
be motivated to leave their windows open more often or in some other
way sabotage the radon test results in order to save money or to speed
negotiations along. In our survey, however, buyers indicated a high
level of confidence that their test results were accurate. In addition,
there was little difference in the percentage of homes falling into the
>4 pCi/1 category depending on whether or not the home was
occupied by someone else while the test was being conducted (25.3%
versus 18,6%) and this difference was not significant in a simple
model which used whether or not the home was occupied during
testing to predict radon level category (t = 0.9, n.s.). Although our
study does not entirely rule out the possibility that some sabotaging of
radon tests by owners is going on, it does indicate that the size of such
an effect must be rather small.
3.4.5. General Transactions and Negotiation
Overall the mean number of weeks it took to close the home sale
was 6.4 and the mean number of offers submitted by the buyers was
1.7. The overall mean sale price of the homes was $118,118. The
• f • . •
mean sale price was higher for tested ($127,180) versus nontested
($108,912) homes, indicating a tendency for more expensive homes
to be more likely to be tested, but the existence of a radon test result
had no measurable effect on aspects of the sales transaction such as
71
-------
time to close, number of offers and counteroffers, and final negotiated
sale price. Any effect radon may be having on these variables is likely
overshadowed by other effects—time of home sale therefore appears to
be a context in which the trouble and expense associated with radon
may be able to be framed as very small.
Overall 79.3% of the buyers employed the services of a realtor,
and 43.7% of these realtors gave some kind of information or advice
about radon to the buyer. Testers were more likely to have employed a
realtor (84.2% vs. 74.1%) and were substantially more likely to have
employed a realtor who gave them information on radon (57.5% vs.
26.9%) than nontesters. In general, many realtors in Boulder County
are doing a good job of informing clients about radon and of motivating
radon testing at time of home sale. They are not, however, effectively
promoting radon mitigation. Out of 31 homes in this study which
were tested and received a test result greater than 4 pCi/1, 15 did not
claim to mitigate. All 15 had employed the services of a realtor. It is
possible that realtors may be taking responsibility for testing since it is
relatively inexpensive but do not want to take responsibility for
promoting mitigation because it is rather expensive, and they do not
want to be held liable for imposing costs on a seller without a legal
basis for their action. Another possibility is that realtors are simply
subject to the same lack of concern over radon levels below 20 pCi/1 as
other people are, and see no need to ensure that mitigation occurs.
Most likely of all, realtors probably do not know exactly what they
should be doing about radon mitigation in order to best serve their
clients. In any case, it is clear that there is a great need for EPA
informational materials directed at realtors which encourage them to
72
-------
take an active role in promoting testing and seeing that radon
mitigation occurs, when appropriate, at the time of home sale.
3.4.6. Characteristics of the Respondents
Overall 63.8% of the respondents were male. The mean age of
the respondents was 37.4, and the "typical" household had 1 child, 2
adults, and only very rarely an adult over the age of 65. As a group the
respondents were very highly educated, with the median educational
level being "completed college."- The median household annual
income was in the $50,000 to $59,999 range. None of these variables
differed substantially between the testing/nqntesting groups. Also,
overall 11% of the respondents indicated they were smokers and 9%
indicated that they did not generally wear a seat belt when traveling by
automobile. The percentage of smokers did not vary between
testing/nontesting group, but nontesters were slightly more likely to
not wear a seat belt than testers. In general, time of home sale
appears to be a context in which individual differences in
demographic characteristics (e.g., age) and tendencies toward risky
behavior (e.g., smoking) that are commonly correlated with attitudes
and behaviors toward radon may often be overridden by other factors.
The one demographic variable with a marked difference between
testing/nontesting group was whether or not the respondent was
employed by IBM: 17% of testers but only 6.2% of nontesters stated
that they were currently IBM employees. The effect of the high
testing and mitigation rates of the subpopulation of IBM employees on
our overall analysis are discussed in the next section.
73
-------
3.5. Data Analysis and Discussion
Throughout this report we have emphasized that the process of
protecting oneself from the risk associated with radon gas is a long
and difficult one, with many opportunities to drop out at different
stages along the way. The effectiveness of a risk communication
program or strategy for radon is perhaps best measured by the number
of people the program is able to keep involved in the process as the
stages progress from initial interest and concern to testing to
interpretation of test results to mitigation and finally to confirmatory
retesting. . In order to interpret the data from this survey (just as we
did for the survey data in Chapter 2) we will therefore concentrate on
how many people are retained at each of the various stages, given the
time-of-home-sale context.
Although it is difficult to compare the overall relative success of
the Washington, D. C. information and awareness campaign (which was
aimed at the general public) and the Boulder County, Colorado study
(which focused on the behavior of a subpopulation of people who had
recently purchased homes and which measured preexisting behavior,
not responses to a risk communication campaign), we can document
the testing rate already occurring in absence of regulation by
government at time of home sale and we can directly compare the
relative effectiveness of each context in promoting mitigation and
confirmatory retesting. That is, given that a test has been conducted,
either within the home sale context (Boulder study) or outside of the
home sale context (Washington, D. C. study), we can directly compare
74
-------
the subsequent mitigation and retesting rates. From Chapter 2, the
transition rate in the population exposed to the information and
testing campaign, for all radon levels greater than 4 pG.i/I, from testing
through mitigation to credible mitigation to confirmatory retesting
was 14.8% X 53.5% X 34.4% = 2.7%. For radon levels between 4 and
20 pCi/1, the transition rate from testing to confirmatory retesting was
even lower: 1.1.9% X 46% X 33.3% = 1.8%. We now turn our
attention to presenting the comparable transition rates from the ,
Boulder, Colorado study.
A chart of the pathway to mitigation for all 303 respondents in~
the Boulder, Colorado study is presented in Fig. 3.1. Only those homes
which were tested before closing are included in the testing stage of
this chart and the following charts (although 19 homes tested for
radon after closing, these tests cannot be unequivocally related to the
home sale transaction). Those 26 homes which were tested before
the eventual buyer first looked at them are included in the charts,
since, although they cannot be related to characteristics of the buyer,
they are still a result of the home sale transaction. Overall, 135 out of
the 303 original homes were tested for radon before closing. The
testing rate of 44.6% for this subpopulation of home buyers is about 12
times higher than that for the general population in the Washington,
D. C. study. Twenty-eight of the test results were higher than the EPA
action level of 4 pCi/1. Of these 28 homes, 15 (53.6%) claimed
mitigation. Although we don't have a measure of "credible" mitigation
as we did for the Washington, D. C. study, we do have the same
measure of "confirmed mitigation," whether or not a confirmatory
retest was performed after the mitigation effort, so we can compute
75
-------
303 Homes
I
Tested before closing
135 (44.6%)
4 pci/l
Claimed Mitigation
Confirmed Mitigation
28 (20.7%)
15 (56.3%)
13 (86.7%)
<= 4 pci/l
Claimed Mitigation
Confirmed Mitigation
97 (71.9%)
5 (5.2%)
3 (60%)
Fig. 3.1. Radon mitigation pathway for all 303 home buyers, by
radon level category. (The percentages for partitioning by radon level
category do not sum to 100 because 10 respondents could not report
or estimate their radon level.)
the same transition rate from testing to confirmed mitigation. Of the
15 homes which claimed mitigation, 13 (86.7%) were retested to
confirm that the mitigation effort had been successful. The overall
transition rate from testing to confirmed mitigation for the Boulder
sample of new home buyers was therefore 53.6% X 86.7% = 46.5%.
Compared to the comparable 2.7% transition rate calculated above for
the Washington, D. C. study, the 46.5% transition rate is a very
impressive result, even if it does apply only to a subpopulation of new
home buyers (which typically makes up about 5% of the general
population each year). It is more impressive when one considers that
76
-------
the radon test results from the Boulder study all fell into the 4-20
pCi/1 category, since we were unable^ to oversample high radon levels,
and the 46.5% transition rate might therefore be more comparable to
the even lower transition rate of 1.8% from the 4-20 pCi/1 category in
the Washington, D. C. study. These participation rates are very high in
comparison with those commonly found in all types of risk
communication campaigns directed at the general public: we have
clearly succeeded in identifying a subpopulation in which radon
testing rates are high and in which people more often remain in the
process until protection against the risk posed by radon is assured.
It is possible that these impressive testing and mitigation rates
are being driven solely by the behavior of IBM employees, who, as we
have described, face a strong economic incentive to test and mitigate
for radon. To address this issue we present in Fig. 3.2 a mitigation
pathway identical to that in Fig. 3.1, except that all 35 IBM employees
have been removed. The testing rate decreases only from 44.6% to
40.7% when all IBM employees are removed. The transition from
testing to claimed mitigation for test results greater than 4 pCi/1
* , •
decreases slightly more, from 53.6% to 40.9%, but remains quite high.
The transition from claimed mitigation to confirmed mitigation is
identical. .It is true that testing and mitigation rates for employees
77
-------
268 Homes
I
Tested before closing
109 (40.7%)
> 4
pci/l
i
Claimed Mitigation
Confirmed Mitigation
22 (20.2%)
9 (40.9%)
8 (88.9%)
<= 4 pci/l
Claimed Mitigation
Confirmed Mitigation
77 (70.6%)
2 (2.6%)
2 (100%)
Fig. 3.2. Radon mitigation pathway for all non-IBM home buyers,
by radon level category. (The percentages for partitioning by radon
level category do not sum to 100 because 10 respondents could not
report or estimate their radon level.)
facing a strong economic incentive from their employers to test and
mitigate their new homes for radon are uncommonly high ~ but the
testing and mitigation rates for anyone purchasing a new home are
also quite high. However, IBM's radon program may be having an
indirect effect on the radon testing behavior of home buyers and
sellers in the Boulder County area at large, due to its impact on
realtors who must often sell to IBM employees. Realtors in Boulder
County may therefore be somewhat more likely to disseminate radon
information than those in communities that have no employer with a
radon program — but there are still many realtors in Boulder County
78
-------
who don't generally inform their clients about radon. .The high testing
and mitigation rates currently being achieved in Boulder County surely
are great underestimates of the potential that could be achieved from a
radon disclosure at time of home sale strategy implemented by
government.
Another concern to our estimates of testing and mitigation rates
for the Boulder study is the possibility that testing rates for
nonrespondents could be significantly lower than testing rates for
those who completed the survey. To investigate this possibility,
.during the second administration of the survey in February of 1989,
we asked those people who were contacted but refused to participate
in the survey to answer one simple question, namely, Was your present
home tested for radon before the sale closed? Sixty-six of 80 people
answered this question, and the testing rate for these people was
t '
27.2%. Although half of our nonrespondents were not refusals, but
simply people we were unable to contact, we will assume that this
lower testing rate applies to all 189 nonrespondents. We can then
construct a worst-case testing rate for all 492 homes on our survey
lists, representing all homes sold in Boulder County from September
of 1988 to early February 1989 for which a good telephone number
was available. The estimated testing rate for all 492 homes is 37.9%,
only 6.7% lower than the testing rate for the 303 respondents and
still uncommonly high, even for a subpopulation of the general
population. Although we were unable to assess mitigation rates for
nonrespondents, we have no reason to expect them to be substantially
lower than the rates for respondents.
79
-------
The obvious disadvantage of the time-of-home-sale approach to
promoting radon testing and effective radon mitigation is that it is a
long-term solution and would take years to reach a lot of people, as
homes are slowly, but inevitably, bought and sold at the rate of about
5% per year. In fact, if one takes the testing rate of 44.6% for
respondents to .the Boulder, Colorado study and multiplies it by 5%,
the result is a testing rate of 2.2% of the general population of Boulder
County testing in a given year, roughly equivalent to the testing rate
found for the general population which was the target of the
Wahington, D. C. information and testing campaign.
However, there are several reasons to expect a time-of-home-
sale strategy to be more successful in promoting effective radon
mitigation in the long term than an information and awareness
strategy. First of all, the testing and mitigation rates achieved by the
Washington, D. C. campaign were the result of an intensive, well-
organized mass media risk communication campaign. In contrast, the
testing and mitigation rates found for the Boulder study have occurred
in the absence of any radon risk .communication campaign of any sort
in the Boulder County area - surely it represents the bare minimum
that could be achieved by a policy which aims risk communication
materials at realtors and lenders and promotes or mandates disclosure
of radon test results at time of home sale. A second advantage to the
time-of-home-sale strategy is that it targets homes rather than people
-- even people who are highly resistant to radon testing or show a
great lack of concern for the risk may eventually face a situation in
which they become protected from radon (perhaps by buying a house
that has already been tested and mitigated or by facing an economic
80
-------
incentive from an employer or lender). Third, there is every reason to
expect participation rates.for an information and awareness strategy to
decrease over the years, since each year the people most likely to
paritcipate are recruited, leaving a more resistant population for the
next year. In contrast, the rate at which homes are sold remains
relatively constant year in and year out, and those who test and
mitigate are not different from those who don't. Fourth, there may be
a ceiling on the testing rate that can be achieved in the general
population. Areas in New Jersey and Pennsylvania which have elevated
radon levels and which have been in the news for years and in which
there are helpful state programs which promote testing are having
trouble achieving testing rates much higher than 40%. The time-of-
horne-sale strategy would not be subject to such a ceiling effect. Fifth,
radon testing and mitigation in the home sale context is more likely to
be conducted by professionals, probably because the cost in this
context can be framed as small, and is therefore more likely to be
complete and effective. Finally, the transition rates after testing .are
markedly low in the general population (Washington, D. C. study): only
14.8% of those with test results above 4 pCi/1 even claim to mitigate,
only about half of those who claim to mitigate do so credibly, and only
about a third of those who mitigate credibly conduct a confirmatory
retest. High transition rates at these stages may simply be
unachievable in the general population — but they are achievable given
the assistance provided by professionals and the pressures to resolve a
radon problem which occur in the time-of-home-sale context.
We shall now turn to a closer examination of two factors which
appear to be related to high testing and mitigation rates in this study:
81
-------
35 Homes
I
Tested before closing
25 (71.4%)
> 4 pci/l
I
Claimed Mitigation
1
Confirmed Mitigation
6
6
5
(24%)
(100%)
(83.3%)
<= 4 pci/l
Claimed Mitigation
Confirmed Mitigajtion
19 (76%)
3 (15.8%)
1 (33%)
Fig. 3.3. Radon mitigation pathway for 35 IBM employees, by
radon level category.
employment by IBM and hiring the services of a realtor who is
knowledgeable about radon. In order to gain more insight into these
factors we have separated the 303 respondents into three different
groups and constructed three separate charts of the pathway to
mitigation. Figure 3.3 displays a chart for the 35 respondents who
work for IBM. The homes of 25 of the 35 were tested, 6 of which
tested above 4 pCi/1, all 6 of which mitigated. The IBM policy of
requiring tests and mitigation if above 4 pCi/1 for employees to
82
-------
88 Homes
Tested before closing
60 (68.2%)
> 4
pci/l
I
Claimed Mitigation
Confirmed Mitigation
15 (26.8%)
6 (40%)
6 (100%)
<= 4 pci/l
Claimed Mitigation
Confirmed Mitigation
41 (73.2%)
1 (2.4%)
(100%)
Fig. 3.4. Radon mitigation pathway for 88 home buyers who
employed a realtor who gave them radon information, by radon level
category. (The percentages for partitioning by radon level category do
not sum to 100 because 4 respondents could not report or estimate
their radon level.)
participate in their housing buyback program is clearly very effective.
Figure 3.4 charts the pathway for 88 non-IBM homes whose
buyers employed a realtor who gave them some information about
radon. The testing rate (68.2%) is quite high, indicating that realtors
are helping to motivate radon testing. The mitigation rate for those
with tests above 4 pCi/1 (40%) is good compared to rates among the
^ ' , '
general population but clearly there is room for improvement --
several people whose realtors gave them advice about radon did not go
on to mitigate levels above 4 pCi/1.
83
-------
180 Homes
I
Tested before closing
50 (27.8%)
> 4
pci/l
i
Claimed Mitigation
Confirmed Mitigation
(14%)
3 (42.9%)
2 (66.7.%)
<= 4 pci/l
Claimed Mitigation
Confirmed Mitigation
37 (74%)
(2.7%)
1 (100%)
Fig. 3.5. Radon mitigation pathway for 180 home buyers who did
not receive radon information from either a realtor or employer, by
radon level category. (The percentages for partitioning by radon level
category do not sum to 100 because 6 respondents could not report or
estimate their radon level.)
Figure 3.5 charts the pathway for the remaining 180 homes for
which the respondents did not receive information from either IBM or
a realtor. The testing rate for these people (27.8%) is much lower
than for the first two groups, but still much higher than that of the
general population. These 180 people seem to have not received
much of the advantage which the other 123 people gained by being in
the time-of-home-sale context.
In fact, what 12 of the 15 people in this study whose homes
tested above 4 pCi/1 before the sale closed and were subsequently
mitigated have in common is personal contact with a professional,
84
-------
who is knowledgeable about radon, who knows what to do if the house
has a radon problem, and who in many instances may be taking some
of the responsibility for making decisions about radon off the
shoulders of the buyer. This, more than any other characteristic of the
time-of-home-sale context, is most likely responsible for the high
participation rates we have described.
3.6. Conclusion
. In contrast to our conclusions from the previous chapter,
pro-vision of radon information at time of sale to home buyers appears
to be very effective in inducing both testing- and mitigation. In fact, an
information and awareness program utilizing realtors as the primary
suppliers of information might well make an information !only
approach attractive for single family homes if one is willing to accept
the lengthy delay in resolving the radon problem associated with the
slow turnover rate of the housing market. This approach would
certainly be cost effective, and the development of better information
by EPA for use by realtors would be highly desirable. However, it
should be remembered that, in the group motivated by realtor
information, 68.2% tested their new homes and of those with "high"
radon levels (>4 pCi/1) 40% mitigated. Thus, the overall rate of
mitigation was about 27%. This overall rate of mitigation, though far
better than that obtained from a general information and awareness .
program, clearly leaves room for improvement.
Two approaches may work to improve mitigation rates at the
time of home sale. First, radon testing and/or mitigation could be
required at time of home sale, which would likely yield mitigation
85
-------
rates similar to those for IBM employees. Second, better information
could be made available to realtors either for voluntary use or for use as
part of local mandatory disclosure at time of home sale.
86
-------
Chapter 4
Legal Strategies for Addressing Radon Disclosure at
Time of Real Estate Transfer
4.1. Introduction
Radon is a unique environmental contaminant that is difficult to
control through traditional pollution control laws and regulations. .
Because it is naturally occurring and ubiquitous, it cannot be
controlled effectively at a "point source" like many air and water
pollutants. Since radon contamination generally is not the direct
result of any harmful, negligent, or other behavior on the part of a
person or organization, it cannot be minimized through a permit
mechanism or an enforcement program. Even though the level of
radon contamination may be indirectly affected by human activity (e.g.,
housing design and construction), the natural origins of the gas
prevent shifting the cost of radon contamination to a class of persons
who caused the harm, because no such class of persons exists.
Nonetheless, certain legal strategies, combined with an effective
information and awareness program, should be useful in encouraging
individuals to take protective measures to minimize their exposure to
radon contamination.
Many different radon legal strategies are currently being
practiced by both the federal and many state governments, including
information and awareness programs, incentive programs, and
87
-------
regulatory,programs (see, e.g., EPA, 1987). Our focus in this chapter,
however, will be to address legal issues concerning the strategy for
radon disclosure at time of real estate transfer, which the research we
have detailed in Chapters 2 and 3 suggests will prove to be the most
X
effective strategy for promoting radon mitigation in the long term.
The remainder of the chapter is organized as follows. Section
7.2 will provide an overview of current federal radon strategies in
order to develop the context in which future strategies must be
considered. Section 7.3 summarizes those state programs which are
relevant to the time-of-real-estate-transfer strategy and may therefore
serve as models for similar federal strategies. Section 7.4 then
discusses in detail the legal issues surrounding implementation of
such a strategy at the federal level.
4.2. Current Federal Strategies Addressing Radon issues
Congress has not at this time provided the United States
Environmental Protection Agency with regulatory power over naturally
occurring radon, nor has it adopted incentives, such as tax credits or
deductions, to encourage radon mitigation and testing. Congress,
however, has required EPA to undertake radon studies and provide
information and technical aid to states in the Radon Gas and Indoor
Air Quality Research Act of 1986, Pub. L. No. 99-499, §§ 401-405, 100
Stat. 1758-59 (1986) ("Radon/Indoor Air Research Act"), and the
Radon Pollution Control Act of 1988, Pub. L. No. 100-551, 102 Stat.
i
2755 (1988) ("RPCA"). Based on these two laws, the current federal
strategies for controlling radon can be divided into two groups: (a)
88
-------
studying the radon problem and (b) providing information, technology,
and other aid to states to assist them in developing radon programs.
4.2.1. Federal Laws
As its name suggests, the Radon/Indoor Air Research Act creates
a program to study and assess the indoor radon problem. The Act
requires EPA to establish a research program on radon gas and indoor
air quality designed to gather data and information to contribute to the
understanding of health problems associated with the existence of air
pollutants in the indoor environment, to coordinate state, federal,
local, and private research, and to assess appropriate federal
government actions to mitigate the environmental and health risks
associated with all indoor air pollution, including radon. The.
Radon/Indoor Air Research Act specifies the requirements of this
research program, which include research efforts to identify,
characterize, and monitor sources and levels of indoor air pollution
including radon, control'technologies and other mitigation measures,
demonstration programs, and information dissemination. This Act
specifically states that Congress has not yet authorized EPA to carry
out a regulatory program.
Like the Radon/Indoor Air Research Act, the Radon Pollution
Control Act does not authorize or create a regulatory program. It is
more comprehensive than the Radon /Indoor Air Research Act and
contains several interesting provisions directly aimed at state
programs. For example, it provides grant assistance to the states to
set up radon programs and requires EPA to develop and implement
89
-------
activities to assist state radon programs. Additionally, this law sets a
national goal for the level of acceptable radon in buildings, requires
the EPA to update and republish its "Citizen's Guide to Radon" (EPA,
1986a), and orders EPA to develop model construction techniques
and standards for controlling radon. A more complete description of .
the Radon Pollution Control Act and other federal laws appears in
Appendix III.
4.2.2. Federal Programs
4.2.2.1. Revising EPA's Citizen's Guide to Radon
The RPCA requires EPA to revise its Citizen's Guide to Radon to
include information regarding a. series of "action levels" indicating the
health risk associated with different levels of radon exposure. Certain
"other information" must be. provided as well, including a discussion of
the increased health risk associated with the exposure of potentially
sensitive populations to different levels of radon, the increased health
risk associated with radon exposure and risk taking behavior (e.g.,
smoking), the cost and technological feasibility of reducing radon
concentrations, the relationship between short-term and long-term
testing techniques and measurements and action levels, and data on
outdoor radon levels around the nation. Through the RPCA, Congress
also set a national long-term goal to lower indoor radon levels to the
same level as ambient air.
Obviously, Congress is very concerned about the health effects of
long-term exposure to radon gas. Apparently it feels that information
about these health effects was not clearly conveyed in EPA's current
90
-------
Citizen's Guide. In fact, the House Report explains why Congress
mandated the additional information in the Guide:
The legislative requirement to update the Citizen's Guide is
principally based upon the Committee's concern that the public
is interpreting radon levels as safe if they fall below EPA's action
level of 4 picocuries per liter. The EPA guidance document
currently advises that "follow-up measures are probably not
required" if screening measurements are less than 4 picocuries
per liter In addition EPA states that "[e]xposures in this range
are considered above average for residential structures." Jhe
Committee believes that many people have misinterpreted EPAs
designated action level and the statements in the current
citizen's guide as meaning that there is little or no risk from
radon levels below 4 picocuries per liter (H.R. 1047, 100th
Cong., 2d Sess. 12, 1988).
Congress1 required changes and rationale are significant and may
impact other strategies for at least two reasons. First, they suggest
that EPA's current "action level" of 4 picocuries per liter will be
adjusted downward, i.e., made more stringent; Although Congress did
not establish a mandatory radon level, its mandate for a continuously
decreasing long-term national goal is likely to be adopted by states,
local governments/and others involved in radon mitigation. Thus, if a
radon contamination problem is discovered in the future and the
decision is made to correct it, more complete, and probably more
complex and expensive, mitigation and remediation methods may
likely be required to reach this continually decreasing level. Second.
the House Report indicates that Congress favors a health-based radon
standard over a technology-based one. Accordingly, mitigation
methods and techniques will be forced to progress toward this ever
decreasing long-term radon goal. This type of "technology forcing"
may mean that costly and complex remediation and mitigation
91
-------
techniques must be employed, at least until -reasonably priced
mitigation technology can achieve the lowered radon levels.
4.2.2.2. Model Building Codes and Standards
The RPCA also requires that EPA develop model construction
standards and techniques for controlling radon levels in new buildings
by June 1, 1990. To the maximum extent possible, EPA is required to
consult with organizations involved in establishing national building
standards and techniques in order to develop these standards. The
Act does not require states to adopt these model building codes and
standards, but it does instruct EPA to give preference for grant
assistance to states that have made reasonable efforts to ensure the
adoption of these standards.
While not requiring that state and local governments adopt these
building standards, by withholding funding from states that do not
make "reasonable efforts" to adopt them, Congress has sent a clear
signal that it expects the standards to be adopted. Thus, Congress has
sent the message to the sjates that making new buildings radon-proof
should be a high priority.
4.3. Current State Real Estate Transfer Strategies
Two complementary real estate transfer strategies have been
considered or adopted by some states: (a) contract notification, which
can be a general warning of the radon threat, a required disclosure of
previous test results to prospective buyers, or the requirement of
radon testing prior to the final execution of a contract; and (b)
92
-------
contract nullification or new owner reimbursement for remedial costs
when prior notification was not given. Florida, Maine, New York, New
Jersey, and Rhode Island have considered legislation that would
require warnings or disclosures to prospective real estate buyers,
renters, or lessees. To date, only New Jersey, Florida, and Rhode
Island have passed such legislation. The state of Maine has considered
legislation that would allow the buyer to nullify a real estate contract if
notification of a radon problem is not provided. The New York state
legislature debated the question of reimbursement to new owners for
radon remediation if previous owners or realtors failed to disclose a
radon problem. No state has yet adopted such legislation. The next
few sections provide detailed information for some of these state
programs.
4.3.1. Florida
As of January 1, 1989, Florida requires notification of the
potential radon threat in at least one real estate document. The
notification must read as follows:
RADON GAS: Radon is a naturally occurring radioactive gas-that,
when it has accumulated in a building in sufficient quantities,
may present health risks to persons who are exposed to it over
time. Levels of radon that exceed federal and state guidelines
have been found in buildings in Florida. Additional information
regarding radon,and radon testing may be obtained from your
local county public health unit.
93
-------
4.3.2. Rhode Island
Rhode Island passed a law in June 1988 requiring that, among
other things, by January 1, 1989, all real estate contracts must contain
the following language:
Radon has been determined to exist in the state of Rhode Island
Testing for the presence of radon in residential real estate prior
to purchase is advisable.
4.3.3. New Jersey
New Jersey law requires that "[i]n the case of a prospective sale
of a building which has been tested for radon gas and radon progeny,
the seller shall provide the buyer, at the time of the contract of sale is
entered into, with a copy of the results of that test and evidence of any
subsequent mitigation or treatment, and any prospective buyer who
contracts for the testing shall have the right to receive the results of
that testing." In reaction to this law, New Jersey realtors have
incorporated testing provisions into the real estate transaction. They
require the seller to test before the sale, and the buyer to test
immediately after the sale. If remediation work is needed, it is funded
out of an escrow account set up by the seller and buyer before closing.
A pending state bill (S. B. 2964) would require the completion of
radon tests as a condition of real estate transfer.
94
-------
4.3.4. Maine
The state of Maine proposed but did not pass notification
legislation for certain real estate documents. The proposed Maine
notice (L. D. 552) was written as follows:
Naturally occurring radon gas may pose a health hazard in
residential dwellings. Testing services are available from the
Department of Human Services, Public Health Laboratory,
Statehouse Station #11. Augusta, Maine 04333.
This proposed legislation was designed to warn Maine residents of
possible radon health risks and to establish penalties for real estate
agents and home sellers neglecting to inform buyers of the notice.
A second bill, the Radon Gas Liability Act (L. D. 965), was
introduced to the Maine State House of Representatives in 1987 but
also did not pass. One proposal in this bill would have rendered any
real estate contract null if radon tests revealed concentrations greater
than 4 pCi/1 in the building being considered for sale and if the
purchaser wished to nullify the contract as a result.
4.3.5. New York
A major piece of legislation proposed in 1988 is still pending in
the state of New York. Assembly Bill 10293 would amend Article 20
and Sections 600-602 of the New York real property law to include a
radon disclosure notice. The notice would appear on all residential
property contracts and would read as follows:
95
-------
RADON DISCLOSURE-NOTICE
Radon is an odorless and colorless radioactive gas identified as a
leading cause of lung cancer. It may be released naturally in the
ground in locations throughout New York State. Radon can seep
into homes and accumulate to dangerous levels. The buyer
should be aware that indoor radon levels can be determined by
proper testing and, where necessary, remedial steps can be
taken to correct the problem. These steps may include sealing
the foundation or ventilating the basement. The parties to this
contract may negotiate concerning the allocation of the costs of
detecting and correcting any identified hazardous radon gas
problem. For further information on radon contamination,
contact federal, state or local health or environmental agencies.
In cases in which such notice is not provided, New York's proposed
legislation would make sellers liable to buyers for the reasonable costs
of radon detection and remediation occurring within one year of the
date of property transfer. Moreover, another proposal in New York
would aUow the nullification of any new real estate contracts when a
home tests above 4 pCi/1.
4.4. Radon Disclosure at Time of Real Estate Transfer
In this section we set forth the radon control strategy we believe
offers the most promise for promoting effective mitigation: mandatory
radon testing and disclosure during real estate transactions. We will
discuss the advantages and disadvantages of this strategy based on the
following six criteria:
1.
2.
effectiveness ~ whether the strategy can achieve the goal of
lowering exposure to radon;
cost -- whether governmental resources are associated with
implementation of the strategy;
96
-------
3. equity ~ whether the strategy distributes the costs fairly among
individuals affected by it;
4. legality ~ whether the strategy conflicts with other principles of
law, such as constitutional law (see Appendix IV for a detailed
analysis of federal constitutional issues);
•x.
5. federalism — whether the strategy affects the relationship
between the states and the federal government; and
6. enforceability - whether the strategy can be policed to ensure its
efficacy. .
. The time-of-real-estate-transaction regulatory approach is a
f
flexible, multifaceted, yet powerful strategy under which the
government can require or recommend varying degrees of disclosure,
testing, or mitigation including (a) general disclosure of the potential
of radon gas and its effects, (b) specific disclosure regarding known
levels of radon gas, (c) recommended or required testing for radon
contamination, (d) recommended or required mitigation, if ._
appropriate, or (e) voiding of the purchase and sale.
To analyze this strategy, we divide real estate transactions into
five categories and subdivide one category, home sales, into
•
chronological subcategories. The five categories are as follows:
1. a sale of a residence or building;
2. a lease of a residence or building;
3. a financing action after the rental, or home or building sale, such
as a home equity loan, improvement loan, or second mortgage;
4. issuance of a building permit for modification of existing or new
structures; and "
5. issuance of a final "certificate of occupancy" or a final inspection
for new structures.
97
-------
The sale situation, will be discussed at length. Because the steps
involved in leasing, financing, and issuance of a certificate of
occupancy are analogous to steps in the sale process, these categories
will be discussed only briefly.
Before examining these specific strategies, however, some
general observations can be made about intervention during the real
estate transfer process. Assuming that the information disclosed at
the stages suggested below is accurate and that the parties involved in
the real estate transaction comprehend the information, such
strategies can be very effective in ultimately reducing nationwide
exposure to harmful radon levels. The strategies could have a broad
impact because the vast majority of building and homes are eventually
bought and sold through the services of real estate agents, inspectors,
and mortgage institutions. The strategies may also be considered
optimal from an economic perspective, because affected parties are
allowed to bargain freely about the economic and health risks of radon
exposure. Although the strategies have the advantage of low costs for
the regulatory agency, certain nontrivial costs for testing and
paperwork would be imposed on sellers, realtors, and mortgage
bankers.
The notable advantage of these real estate transaction strategies
is that during the sale process, buyer, seller, and the supporting
financial institutions (e.g., mortgage companies) have a heightened
awareness of risks to property and will be motivated to obtain and
disseminate accurate information and/or to undertake remediation.
Also, real estate transaction strategies provide an appropriate vehicle
98
-------
for translating health risks into economic terms in a way that may
encourage mitigation that would not otherwise occur. .
The effectiveness of real estate transaction strategies might,
however, be hampered by certain factors. For example, success would
depend on the availability of reliable test methods. Radon has become
a national issue only recently, and testing and mitigation methods have
not been fully developed. The fast pace of most real estate
transactions requires accurate short-term test methods that may not
yet be available. Perhaps, more importantly, a safe level for exposure
to radon has not yet been established. Congress only recently declared
it a national goal to reduce indoor radon levels to the same levels as
ambient air (see Appendix III). Without an agreed upon level of safety,
it is extremely difficult to determine an appropriate level for
mitigation or to determine the real devaluation properly attributable to
a radon problem. This uncertainty might disrupt real estate
transactions, with or without required disclosure, testing, or
mitigation, although the evidence presented in Chapter 3 suggests this
is unlikely. The success of these strategies also may depend on the
development of certification criteria for radon testing and mitigation
companies. The assurance of minimum qualifications of those involved
in the radon service industry is crucial to provide confidence about the
information exchanged in the real estate negotiation process.
The place at which the federal government might most
appropriately intervene in the real estate transaction process is in the
second mortgage market, when mortgages are resold to financial
institutions which are often located in a different state than the
original lender and therefore naturally fall under federal jurisdiction,
99
-------
although intervention at other points may also be feasible. However,
any federal intervention in the real estate transaction process must be
measured against the potential constitutional concerns (e.g., tenth .
amendment) that might arise in the event these strategies are
federally implemented (see Appendix IV). With these general
observations in mind, the potential strategies and their merits can be
examined.
4.4.1. Sale of a Home or Building
The sale of a home or a building is perhaps the most opportune
time to implement a real estate transaction strategy, For purposes of
analyzing this approach in greater detail, we have divided the sale into
five periods: (a) the pre-sale period, during which the buyer generally
reviews homes with the realtor; (b) the negotiation period, during
which the buyer and seller discuss, and agree upon, the terms of a
sale; (c) the contract signing and inspection period, during which the
buyer signs the purchase contract, seeks financing, inspects the
seller's premises, and generally gathers information about the
property; (d) the final inspection and the closing, when the buyer
inspects the premises immediately before closing, and closes on the
premises; and (e) the post-closing period, at which time the buyer
purchases the property and takes possession.
4.4.1.1. Pre-sale Period
During the pre-sale period, a buyer generally "goes to the
market" by contacting a real estate agent. The buyer and agent
100
-------
preview buildings or homes so the buyer can select potential premises
for purchase. While it is probably inappropriate to require realtors to
conduct testing or mitigation, requiring the real estate agent to
provide information with respect to radon levels, testing, and/or
mitigation at this point might be an effective disclosure approach. (As
described in Chapter 3, many realtors are already taking it upon
themselves to disclose radon information to their clients.) The real
estate agent would bear the burden of determining if the property had
ever been tested, and if so, the results of the test. A realtor might be
subject to suit upon breaching the obligation to inform or disclose.
While imposing such a condition upon real estate agents may
seem burdensome, it is not unreasonable. Realtors act as agents for
the seller and generally are familiar with the community. They occupy
a position of trust with respect to both the seller and buyer, who rely
upon their expertise. Furthermore, realtors have created a database
containing detailed information about sale properties called the
Multiple Listing System (MLS). Thus, an appropriate notation in the
MLS system could inform all realtors that a parcel of property contains
radon or has not yet been tested. Disclosure at this early stage allows
the buyer to learn about radon and to consider whether to accept the
risks associated with long-term exposure.
Disclosure by realtors has other advantages. It is cost-effective
for the government, because the real estate agents are shouldering the
expenses related to the disclosure requirements. This strategy also
puts disclosure in a context that the seller and buyer can understand;
radon becomes associated directly with the sale and purchase of a
home or building. It is not difficult to enforce, because disclosure can
101
-------
be determined by reviewing any written materials given to the buyer
by the real estate agent, or by asking about oral warnings provided by
the realtor. This approach is relatively comprehensive because the
majority of property sales involve a realtor, allowing a wide audience to
be reached. Realtors may add a disclosure clause to the standard
purchase contract for homes in order to fulfill this disclosure
requirement, they may prepare a separate information booklet on
radon, or they may use "safe harbor" forms.17
The major disadvantage of this trategy is the burden it places
upon real estate agents. Depending on the type and extent of
disclosure required, it may force realtors to require sellers to conduct
a radon test, or mitigate, before a building or home is listed for sale.
More importantly, if the buyer later discovers high radon levels, this
strategy could create liability for realtors who, in good faith, relied
upon assertions by the seller that radon was not a problem at the
premises. Some of these inequities can be cured by requiring or
recommending that the realtor provide only a generic or standardized
written notice to the buyer (although evidence suggests it may be
»
important that a test be required). Such a notice could state simply
that radon is a colorless, odorless gas that may cause lung cancer, and
that it may be appropriate to test property for radon contamination.
Two states (Rhode Island and Florida) have enacted such general
radon notices for real estate transactions (see Sections 7.3.1 and
17By using safe harbor forms, which set out a checklist of
potential hazards for which real estate agents must inspect, realtors
can be more confident that they have fulfilled their disclosure
obligations and protect themselves from liability.
102
-------
7.3.2). Such a notice might lead the buyer to engage in the necessary
self-protective behavior. The seller, knowing that the buyer is to
receive such a general disclosure, is more likely to test the home or
building and conduct remedial action before attempting to sell it.
Such a pre-sale disclosure strategy could be very effective in
encouraging widespread mitigation because most real estate
transactions involve a real estate agent. This strategy would
significantly increase the distribution of information to buyers, who
have the biggest incentive to evaluate it carefully. Delivery of this
information personally and directly by a real estate agent also would
increase seller and buyer awareness. Also, real estate agents would
themselves have an incentive to learn about radon so that they can
advise their clients and so that the radon issue does not greatly
interfere with the sales process. And, disclosure at this early stage in
the sale process still allows, substantial time for further inquiry,
testing, evaluation, or mitigation.
4.4.1.2. Negotiation Period
The negotiation period generally involves three parties: the
seller, the buyer, and a realtor. During this period, the buyer has
tentatively decided to purchase the premises and approaches the
seller with an offer to buy. Generally, negotiations ensue over the
terms of the sale, including the purchase price, time of closing, and
other conditions and contingencies. At the end of the negotiation
period, seller and buyer agree on the terms of purchase.
103
-------
The negotiation period is the most obvious period in which the
radon disclosure/testing strategy may be utilized, because it is during
this period that seller and buyer are evaluating all of the factors
relevant to the sale. Seller and buyer are assigning a monetary value to
such factors and making decisions about them. During the negotiation
period, the radon contamination would likely be quantified by the
seller and buyer with respect to the purchase price of the premises.
Thus, the final agreement would reflect the diminution in the value of
the premises because of the radon problem.
There may be one potential drawback to disclosure without
required mitigation at any point in the sale process. Assuming the
buyer understands the risks associated with exposure to radon,
allowing the seller and buyer to negotiate about the radon levels means
that the buyer may accept a trade-off by willingly purchasing a radon-
infested home for a lower cost despite the potential health effects of
long-term exposure to radon. If the aim of any radon strategy is to
encourage self-protective behavior "and correct radon contamination,
then disclosure without required mitigation may not reach this goal.
However, the evidence presented in Chapter 2 suggests that
mitigation is nearly always the option chosen. Alternatively, forced
mitigation at the time of sale is problematic. First, it requires that
**
some level of mitigation be agreed upon by seller and buyer or set by
the government. Second, if the government sets a required mitigation
level, it could reduce the market value of the premises to such an
extent that the seller can argue that the property was
unconstitutionally taken without just compensation (see Appendix IV
for a discussion of the takings issue).
104
-------
The real estate negotiations process is one of the most effective
times during the home sale process for encouraging or requiring
disclosure, testing, or mitigation, because the parties are focusing on
the economic aspects of the transaction, are highly motivated to obtain
detailed information about the building or home, yet still have time, if
necessary, to undertake additional testing or evaluation or mitigation.
During this stage, information about any radon contamination can be
readily translated into concrete economic terms and can become
another factor in negotiations over the price. And, in the event that
testing is incomplete or uncertain, the parties can add contingency -
clauses or allocate future potential liabilities.
4.4.1.3. Contract Signing .and Inspection Period
The contract signing and inspection period begins when the
contract for sale is signed and the buyer begins the search for
financing the sale, and it ends after the buyer has gathered
information about the premises and conducted a building or home
inspection before closing. Thus, this period mainly involves the buyer,
.the financial institution providing the mortgage, and the individuals
who gather facts and conduct various inspection services for the buyer
(e.g., home or building inspector).
The mortgage process also is one of the most effective avenues
for required or recommended testing or disclosure. Because financial
institutions have a strong interest in protecting the market value of,
collateral, as a condition of the mortgage, mortgagees may provide
105
-------
radon information to the mortgagor!* or require mortgagors to certify
that radon levels are within a certain range and/or to test for radon
and take remedial action, if appropriate. Because most home purchase
contracts contain a contingency clause permitting the buyer to void
the contract without penalty in the event that a mortgage cannot be
procured, a home buyer discovering radon contamination would not be
penalized for failing to procure a mortgage because of a radon problem.
A seller who lost a home or building sale because of radon would be
encouraged to mitigate the problem in order to increase the sales
potential of the building or home.
Involving financial institutions in a strategy to increase
disclosure, testing, and mitigation may be the most powerful method
of intervening in the vast number of real estate transactions. Because
obtaining the mortgage is typically the "make or break" point of a
building or home sale, both seller and buyer have strong incentives to
comply with whatever the mortgagee requires. More importantly, the
purely economic interest of the mortgagee in properly evaluating the
value of the property and in preventing subsequent devaluations (such
as the latent discovery of high radon levels) would motivate financial
institutions to ensure that comprehensive and accurate information
was disclosed. An additional benefit also may be that fee collective
106
-------
economic weight of mortgage institutions creates larger incentives for
accurate testing, innovations in mitigation methods, more thorough
information dissemination, and better assessments of health risks.
Generic or standardized disclosure, by the persons who conduct
home or building inspection services may have a positive effect.
.Arguably, it is somewhat late in the sale process to begin testing or
mitigation because the purchase contract has been signed and the sale
process is almost complete. For example, while conducting the home
inspection, the inspector can ask the buyer whether there has been
any attempt to assess radon contamination. Inspectors can also inform
buyers of structural and/or other characteristics of the premises that
indicate a potential radon problem, such as a completely underground
cellar that creates a negative pressure likely to draw radon into the
home, or granite or other geological formations upon which the
premises are built. The major drawbacks of this strategy are analogous
to the drawbacks of using realtors as disclosure agents, i.e., it may be
unduly burdensome and provide the basis for potential liability.
Additionally, home inspecv>rs may not be trained to detect and assess
potential radon risk factors and thus may provide incomplete,
inaccurate, or false information.
The effectiveness of requiring radon consultation to be included
in the inspection process would depend largely on the training and
qualifications of the inspectors. Nonetheless, because inspectors are
generally hired by-the buyer, the buyer would have an incentive to
conduct a thorough examination, which the buyer could then carefully
evaluate. This strategy could provide an effective double-check on any
information provided by the -seller. And, because inspections are
107
-------
routine for virtually all home and building sales, this strategy could
reach the vast number of homes and buildings involved in real estate
transactions.
4.4.1.4. Final Inspection and Closing Period
The final inspection and closing period includes the final "walk-
through" immediately before closing and the closing ceremony itself,
in which the appropriate documents are executed and, if necessary,
attested to. The major actors during this period are the seUer, the
buyer, and their attorneys.
Again, as discussed with respect to the contract signing and
inspection period, the seller's disclosure to the buyer is probably not
an appropriate strategy at this stage. Mandatory or recommended
testing or mitigation probably is not feasible. One potential strategy,
which is a type of disclosure, could be to place a warranty or warning
into the deed. The warranty, given by the seller to the buyer, could
state that, to the best of the seller's knowledge, the premises are not
contaminated with radon. It could state that the seUer has tested the
premises and list the date of the test, the level of radon discovered,
and the company that conducted the test. The warning, placed in the
.deed, could contain the same statements as the warranty or it could
contain a generic disclosure statement about the risks of radon. These
warnings or warranties could provide the buyer with a strong cause of
action against the seller if they proved false. Moreover, a warning in
the deed would notify any future purchaser of the radon-related
conditions. Thus, the next buyer's lawyer, examining a deed, would
108
-------
alert the buyer about radon. While the use of warnings or warranties
would not be an effective strategy for the present parties to the
negotiations (because bargaining at this point is essentially finished),
such provisions could be an effective tool for informing future buyers of
the property's radon risk. With this long-term view in mind, this
strategy could be uniquely effective.
4.4.1.5. Post-closing Period
The post-closing period begins when the closing is completed
and continues until the premises is sold again. The buyer, now the
property's owner, is the only individual directly connected with the
premises.
Disclosure, mandatory or recommended testing, and other \
strategies would not be effective in altering behavior that has occurred
in the past. Nonetheless, empowering the buyer/owner to recover ~
costs of mitigation and/or testing from the seller if the buyer/owner
discovers high radon levels after taking possession may be a strategy to
employ during this period.
This strategy has many potential disadvantages. It probably
depends on litigation, which is expensive, uncertain, and time
consuming; it may be difficult to prove that the seller falsified test
results or conducted an inadequate radon survey, because radon levels
fluctuate dramatically over time. There is also no guarantee that any
amount recovered by the buyer/owner would be spent on radon
mitigation. Nonetheless, the major advantage of this approach centers
on its potential deterrent effect. Because it severely punishes the
109
-------
seller's failure to determine and/or disclose radon levels, sellers will
be more likely to carry out required or recommended radon testing
and disclose the,results.
4.4.2. Lease of a Residence or Building
The same approaches presented above involving the sale of a
home or building could.be applied in the context of negotiating a lease.
For example, the government could require or recommend disclosure
by a realtor, insertion of a "radon clause" in the lease agreement, or
condition financing on testing and/or mitigation. Particularly because
leasing of commercial buildings is pervasive, this avenue for
addressing the radon threat should be considered an important
supplemental strategy. '
_ The main distinction between the lease situation and the sale
situation is that (depending of course on the term of the lease) the
parties may have less of an interest in testing, disclosure, and/or
mitigation, because they may be only thinking of the short-term
economic consequences of the transaction. On the other hand,
because the lessor and lessee have an ongoing relationship through the
end of the term (and, by the exercise of options, perhaps beyond), the
time pressures inherent in the sale negotiations context may be
partially removed. For example, compared to a buyer, a lessee could
more easily agree to condition the amount of the rent on the results of
long-term or periodic testing for radon contamination. Lessees also
could obtain a clause providing that expenses for mitigation are
deducted from the rent. As with sales, the success of this strategy
110
-------
would depend in great part on the quality of information available to
the parties and their own economic evaluation of the health risks
involved.
4.4.3. Financing Action after the Home or Building Sale
Another point at which a radon strategy could be implemented
is during a financing action after the home or building has already
been sold, e.g., when the buyer applies for a home equity loan, a home
improvement loan, or a second mortgage. The analysis outlined above
with, regard to mortgages would apply in these situations as well. At
this stage, a strategy of recommended or required testing, disclosure,
and/or mitigation could be effective in encouraging owners to reduce
their exposure to radon. The unique advantage of this strategy, of
course, is that it reaches buildings and homes not involved in sale or
transfer. Even though only a small percentage of buildings or homes •
may be affected by this strategy, it is a logical complement to the
transaction-based approach discussed above.
4.4.4. Issuance of a Building Permit
State and local laws generally require that building permits be
obtained for new structures and for certain modifications of existing
buildings or homes. It is at this point that building code and other
structural requirements minimizing radon exposure could be imposed.
For example, building plans could be evaluated for ventilation systems,
structure of underground areas, floor construction, and weatherization
techniques. The soon to be drafted model building code (see
111
-------
Appendix III), which states are encouraged but not required to adopt,
provides national leadership for this strategy.
Because construction cannot legally proceed without a permit,
the owner may face "red tagging" delay, and possible fines for
attempting to circumvent the permit system. Thus, restrictions on
building permits would be an effective technique for ensuring that at
least new and modified structures are radon safe if not radon proof.
The primary advantage to such a strategy is that it is preventive
in nature. It would be relatively easy to enforce, because radon
mitigation measures would simply become another design
requirement for architects and contractors. The social costs of such a
policy also would be low for those houses which need mitigation .
because, at least for new "structures, no radon problem yet exists and
no costly post-construction remediation would be required. However,
considering that the large majority of houses do not require
mitigation, the social costs of the policy may be high in the sense that
costs must be imposed on many homes which have low radon levels in
order to find the homes which have high radon levels.
The main disadvantage to permit requirements is that it may be
difficult to achieve concensus concerning the appropriate building
code requirements. And, because structural design, though important,
is only one element that determines potential radon exposure levels,
this strategy cannot address or anticipate the largely natural causes of
contamination (i.e., the geologic composition of the area).
112
-------
4.4.5. Issuance of a Final Certificate of Occupancy
Requirements for disclosure or for testing or mitigating radon
contamination levels could also be imposed during the issuance of a
final certificate of occupancy or a final inspection for new structures.
As discussed above with respect to the final inspection during the sale
process, this stage is probably not an appropriate time to impose a
primary radon strategy. Particularly with regard to new structures,
because the certificate of occupancy and final inspection occur after
the completion of design and construction, the flexibility of an owner
to respond to new requirements is limited and the costs may be high.
On the other hand, actual radon exposure levels may not be
ascertainable until after a structure is completed. If such a strategy
accounts for these latent problems, then requiring testing and/or
mitigation prior to occupancy could be a very effective secondary
strategy for enforcing exposure goals.
4.4.6. Mandatory Testing or Mitigation of Pnhllc Buildings
The radon disclosure during real estate transaction strategies
outlined above, if properly implemented, will address the majority of
radon problems in homes, and a large number of radon contamination
problems in buildings. Because these real estate strategies hinge upon
property transfer, however, they will not be successful in addressing
premises which are not generally sold, leased or otherwise
transferred. Also, the real estate strategy may be inappropriate for
transfers involving state or federal governmental entities because
applying the approach to a governmental body may be inappropriate as
113
-------
a matter of law. Thus, there is a category of "public buildings" which
would require application of a different strategy as a supplement to a
real estate strategy. This section focuses on one category of public
buildings: governmental or quasi-governmental structures.19
A potential federal strategy applicable to governmental or quasi-
governmental facilities could involve the mandatory testing and
disclosure of results to the occupants of these buildings. Mandatory
mitigation is a second option aligned with this strategy.
Mandatory,testing of governmental and quasi-governmental
facilities would depend largely upon intra- and intergovernmental
cooperation. Mandatory testing of these premises is easier to
administer than mandatory testing of privately owned, operated, and
occupied property because the government can easily carry out the
testing on its own property, or exert its will to require testing. For
example, because there are no private parties involved, a facility which
is owned or operated by a state government can be tested easily if the
state legislature requires testing of public buildings. For a structure
leased or occupied by a governmental entity, the government can
public building includes the following three categories of real
property: (a) governmental or quasi-governmental structures, the
majority of which are buildings owned, operated, or occupied by the
federal or state government and containing federal or state
government-facilities especially offices; (b) public accommodation
iacihties, which are facilities used by members of the public, or open
to members of the public, such as a museum, library, sports arena or
concert hall; and (c) "special population" facilities, such as hospitals,
mental institutions, schools, day care centers, or prisons
This section of the report does not address public accommodation
iacihties or special population facilities. Although strategies applicable
to these two types of public buildings should be analyzed, they are
outside the scope of this report.
114
-------
exert its bargaining power to include in its lease a clause regarding
testing.
By restricting testing, to governmental or quasi-governmental
facilities, the federal government would not impose financial, legal, or
other inequities or burdens on private citizens. A mandatory testing
program would demonstrate the government's commitment to the
seriousness of the radon problem, and would serve as a model for
citizens to follow. Thus, testing by the government could be a valuable
public relations tool. ]/' .
But testing alone may not be sufficient. Testing without
mitigation uncovers radon contamination without resolving it. .
However, disclosure of the test results may motivate the building
occupants to demand mitigation and follow-up testing, and may
encourage others to demand testing and disclosure.
The potential cost of requiring testing is the major disadvantage
of this strategy. But required public testing may be a necessary cost to
increase citizen confidence in government declarations of the radon
threat, and to demonstrate that the radon problem requires prompt
action and the government is prepared to take such action. In fact,
without an investment in public testing, the federal government may
undermine the value of their other radon investments by sending an
. ambivalent message to citizens. In addition, mandatory public testing
guarantees that the federal government will learn the full extent of its
portion of the radon problem.
Required mitigation of radon contamination can be combined
with the mandatory testing requirement. Though the costs of
mandatory mitigation are high, the potential benefits are great. First,
115
-------
mandatory mitigation would establish the federal government as the
leader on radon remediation. Second, the occupational health threat
would be reduced or eliminated, thus protecting workers. Third,
required mitigation could inspire private mitigation. Fourth,
mandatory mitigation allows the government to test new remediation
techniques.
Aside from cost, the disruption caused by mitigation
construction is the only other serious disadvantage of this strategy.
Required mitigation also is sure to attract publicity, increasing the
likelihood that mitigation target levels adopted for governmental and
quasi-governmental buildings will then be applied to private buildings.
Thus, public mitigation efforts may eventually compel the government
to establish defined levels of acceptable radon exposure.
116
-------
Chapters
Conclusion: An Evaluation of Strategies for Promoting
Effective Radon Mitigation
5.1. Introduction
! .* -
The purpose of this study has been to evaluate the effectiveness
of alternative strategies for motivating people to test for radon gas in
their homes and to mitigate if necessary, and we have conducted,
surveys which evaluate two of these strategies: (a) a traditional
information and awareness strategy aimed at the general public and (b)
a strategy which discloses radon information (and perhaps requires
testing) at the time of home sale. A review of the literature on risk
communication and motivating self-protective behavior suggests that
traditional information and awareness programs will likely fail when
they are targeted at the general population. To test this conclusion ,
•
from the literature we sent a mail survey to 920 households which had
purchased radon test kits as part of an intensive information and
awareness campaign in the Washington, D.C. area (see Chapter 2).
Although we estimate that about 33,000 homes in this area exceed the
federal guideline by a factor of five or more (had a radon reading of 20
pCi/1 or higher), the survey results indicate that only 1.2% of this
group have taken convincing remedial action as a result of the
campaign. Those, few homeowners who tested Were sent reprints of
two EPA documents, "A Citizen's Guide to Radon" (EPA, 1986a) and
117
-------
"Radon Reduction Methods: A Homeowner's Guide" (EPA, 1986b).
Unfortunately, our results suggest that these pamphlets may have
encouraged people to try their own remedial measures rather than
employ a professional contractor. These home remedies (e.g., opening
basement windows more often) were not followed by retesting to
verify their effectiveness in spite of clear warnings given that single
limited remedial measures are likely to be ineffective.
In contrast, a telephone survey of 303 home buyers in Boulder
County, Colorado found that over 40% of recently purchased homes
were tested for radon gas at the time of home sale and that this
testing was often motivated by information provided by the realtor (see
Chapter 3). Even though no intensive information and awareness
campaign had been conducted in Colorado and there are currently no
state laws in effect concerning radon, 54% of tested homes in our
sample which had radon levels above the EPA action level underwent
mitigation (with 87% of those completing follow-up testing) as part of
the home sale transaction. These results suggest that a radon
information and awareness program targeted at the point of home
sale, when the transaction context provides a strong economic
incentive to repair any problems a home might have, could be highly
effective in comparison to information targeted at the general
population. Since several approaches for mandating disclosure of
radon levels at the time of home sale appear to be available to the
federal government (see Chapter 4), a regulatory approach may be the
most effective available strategy.
EPA is currently limited to supporting information and
awareness through such means as its "Citizen's Guide" and "Radon
118
-------
Reduction" pamphlets. Our study suggests that these materials
require substantial revision and that new materials specifically
targeted at realtors and home buyers would be highly desirable,
especially in concert with mandatory disclosure.
The remainder of this chapter discusses three potential radon
strategies:
1 Information and Awareness: Programs or strategies whose
primary purpose is to inform property owners or occupants ot
the potential problems of radon gas in buildings.
2 Economic Incentives: Programs or strategies using financial
inducements to motivate owners or occupants to take personal
action against radon.
3 Regulation: Programs or strategies which involve setting rules,
standards, and procedural guidelines to make responsible
parties take action against radon.
5.2. Information and Awareness
A large amount of research has been conducted on risk
communication and on motivating self-protective behavior. Although
there are still large gaps in our understanding of many issues, there
are some substantive conclusions that yield recommendations for
radon risk communication.
First, it must be recognized that different people define risk in
different ways. For example, experts commonly judge risk in terms of
probabilities and losses, whereas laypeople commonly have a much
broader definition of risk which includes such factors as whether the
risk is controllable or uncontrollable, voluntary or involuntary, natural
or technological, or known to science or unknown. Different people
119
-------
also have different goals concerning risks and are interested in
different information. For example, risk managers are usually
concerned with the aggregate risk to a large population, whereas
laypeople want to know what they should do, individually, to protect
themselves and their family.
Second, people commonly have great difficulty when judging
probabilities, making predictions, coping with uncertainty and, in
general, thinking intuitively about risk. They typically rely on
judgmental heuristics which can lead to systematic biases and errors
in judgment. For example, people often judge the frequency of an
event by its availability, that is, the ease with which examples of the
event can be imagined or recalled. For a risk which lacks perceptual
reminders, for which prior experience is benign, and for which deaths
occur singly and in isolation, such as radon, availability is low and the
risk is commonly underestimated. People also generally fail to
understand the limits of their knowledge. For example, people often
erroneously believe they can exert control over events that are in fact
random and they often display too much confidence in their ability to
estimate uncertain quantities.
Third, naive beliefs concerning risky events tend to be very
unstable, especially when the risk is new and unfamiliar. Such beliefs
can often be easily manipulated by seemingly subtle differences in the
way in which risk components such as outcomes and probabilities are
framed. For example, people dislike suffering losses more than they
like receiving gains and may take different actions depending on
which perspective they are encouraged to adopt.
120
-------
Fourth, people have the most difficulty when judging very low
probability risks (for example, less than 1%). The distribution of
people's decisions for low probability risks is commonly bimodal, with
one mode at or near a level indicating no concern for the risk and a
second mode at a level indicating overconcern for the risk.
A radon risk communication program must take into account
what is known about how people define risks, how they judge risks,
how they respond to alternate framings of risk, and how they interpret
probabilistic information. However, it should be noted that radon has
a profile of characteristics that would be expected to lead people to
underestimate its associated risk or even to dismiss it entirely, which
will make an information and awareness approach very difficult. The
risk is objectively below the level at which people commonly respond
accurately, the consequences of the risk are far removed from the
exposure, and deaths related to the risk occur in isolation and are
impossible to relate directly to the hazard. Also, radon is a naturally
occurring risk for which no one can be blamed and people's prior
'. experience with the risk is overwhelmingly benign. Since radon is
colorless and odorless, there are no perceptual reminders to alert
people to the presence of the risk. Finally, the risk varies widely
depending on time of year, geographical location, behavior patterns,
and other factors, making it difficult, if not impossible, for people to
assess their risk very accurately.
Although the test to determine the risk from radon is relatively
quick and inexpensive, several other characteristics of the radon
testing and mitigation process are likely to discourage appropriate
protective responses. For example, there is little time urgency for
' ;
121
-------
conducting the test, and the results of the test can only bring bad
news compared to the status quo. Also, interpretation of the test
results is subjective and open to question, and people are likely to
view remediation as costly and difficult. Due to the technical nature of
radon testing and mitigation, most people will be forced to rely on the
opinions and advice of others. In fact, to ensure protection against
radon people must engage in a long, drawn out, complicated series of
behaviors with the opportunity to drop out at each step along the way.
Finally, several characteristics of the current social context have
implications for the radon problem. People are now being bombarded
by information concerning a large number of low probability risks, and
the radon issue may not be a very high priority for many people. Also,
the same factors that lead, to apathy on the part of individuals make
the radon problem uninteresting for the mass media and unlikely to be
consistently in the news. In addition, unlike most environmental
risks, radon is largely under the control of individual homeowners,
and risk communication must reach down to this level.
A large amount of research in such domains as health, natural
hazards, crime prevention, injury prevention, and energy conservation
has been conducted on ways of encouraging and motivating self-
protective behavior. Although researchers are only beginning to
explore commonalities among these different domains, experience
with real risk communication programs has yielded a variety of
recommendations applicable to radon risk communication.
The major result of this experience is that, despite an
overwhelming general interest in self-protection on the part of both
professionals and the public, it is enormously difficult to get specific
122
-------
people to perform specific behaviors in specific situations. Well-
intentioned, common sense suggestions are often ineffective, and
(
simply dispersing information and increasing knowledge is usually
insufficient to motivate people to act. Also, people's behavior seems to
be largely governed by short-term consequences, and they are very
reluctant to accept definite costs iii the present to prevent indefinite
harm in the future. Finally, self-efficacy, that is, a person's beliefs
concerning his or her ability to perform an action and its chances for
success, appears to be a very important determinant of protective
behavior — attention must be paid not only to generating concern
about a risk but to providing easy solutions that individuals can be
confident of handling themselves.
Traditional information and awareness campaigns and fear-
arousing appeals have proven to be generally ineffective, in and of
themselves, in motivating people to act. There are, however, two
approaches to communicating risk that appear to be relatively
promising. The first approach uses knowledge from decision theory
to communicate or frame risks in an effective manner. For example,
we have experimented with communicating a low-probability risk as
an integration over time where people can act to protect themselves
from the risk for a long period of time. The results of the-experiment
show that framing a protective behavior as a single decision that
covers a long span of time results in more accurate, more consistent,
and less variable responses to risk. The second approach, social
diffusion, applies knowledge gained from the study of how new
information is commonly diffused through a social system. A typical
program first identifies social referent groups and preexisting
123
-------
channels of communication and then identifies and activates
innovators and opinion leaders and seeks their help in dispersing
information to other members of the system. Such social diffusion
programs have already been shown to be more successful than
traditional information campaigns in several domains.
Risk communication for radon might therefore most profitably
focus on framing simpler, easier, longer-lasting protective responses
and communicating this information through preexisting social
networks and institutions. However, there are severe limitations to a
communication-only approach.' First, there are practical limits on any
communication effort: not everyone will be reached and not everyone
will be persuaded to act appropriately. Second, there are cognitive
limits on people's ability to judge low-probability risks, and highly
accurate judgments and precisely appropriate responses would be
difficult to achieve without regulation. Third, there are attentional
limits: it may prove difficult to encourage busy people who have many
other concerns and who face many hazards in addition to radon to
maintain interest and concern during the long process of testing and
mitigation. Finally, there are limits to our ability to frame problems
and solutions in real-life contexts. In conclusion, consistent both with
the available literature and with the new research presented here,
such purely voluntary programs alone are likely to be ineffective for
promoting effective radon mitigation. However, risk information and
awareness remains a necessary component in the application of either
incentives or regulation.
124
-------
5.3. Incentives
Incentive programs have the general advantages of giving policy
makers a more direct influence on the radon problem than an
information and awareness campaign. One program might involve
government subsidized testing and/or mitigation services. This
strategy can defer or eliminate costs for low income homeowners or
owners in areas with exceptionally high radon levels. The
disadvantage of this strategy is its narrow focus and potential cost.
Moreover, it raises questions, from a legal standpoint, of government
liability whenever it results in damage to a home or building, or fails to
reduce radon levels.
Another possible strategy would be a system of tax credits such
as those used to encourage the use of solar energy in the late 1970's.
These might include income tax credits for radon testing and
mitigation. Radon mitigation work might also be exempted from
property tax assessment, or sales tax rebates might be given on radon
remediation related purchases. Also included in this category could be
government sponsored low-interest or interest-free loans for
mitigation.
There are advantages and disadvantages to these incentive
strategies. Tax credits or tax deductions are primarily options offering
equitable financial relief to home purchasers who were probably not
aware whether the building or home they bought had a radon problem.
On the other hand, tax relief limits collectable government revenues,
and it distributes some of the cost of mitigation to people who do not
have a radon problem.
125
-------
An example of a tax credit approach is the Energy Tax Act of
1978 which-amended the U.S. tax code to give tax credits for energy
conservation and renewable energy source expenditures. For energy
conservation this allowed a deduction, of 15% per year (on
expenditures up to $2000) and for renewable energy a deduction of
SOo/o on the first $2000, and 20% on purchases between $2000 and
$10,000 per year. At the same time 17 states passed legislation to
exempt solar energy installations from property tax assessment. While
there is little data on the effectiveness of individual state efforts,
results on the federal tax credit program are available (Carpenter et
fll. 1981). It was found that while 90% of the homeowners were
familiar with the tax credits, only 30% took advantage of them, and
only 1% would not have made the improvements without tax credits.
This includes both the conservation and the renewable resource
credits. If one disaggregates these statistics it appears that the
conservation tax credit has had little effect while the renewable
energy credit has had a positive effect on the demand for solar space
and water heating. One explanation for the weak showing of energy
conservation credits relative to the solar energy credit might be that
solar energy equipment appears to be very high profile and "high
tech," while energy conservation measures (which are often more cost
effective) are not. That is, many people were drawn to solar
equipment because of its novelty. Conservation measures, on the other
hand, are somewhat "run of the mill" (the neighbors are more
impressed with solar panels than new window caulking). Therefore,
we would expect this type of credit to increase the use of high-tech,
126
-------
high-visibility mitigation techniques at the expense of more simple,
inexpensive methods.
Low-interest and interest-free loans for mitigation also distribute
/-
costs to society, but they are potentially costly for the government.
They may also encourage "gold plated" mitigation efforts and it may ^
also be difficult to ensure that loan money is expended on actual radon
mitigation. Further, these loans may not be accessible to low income
homeowners. Economic incentives alone may result in very high
direct costs to government with no guarantee that those costs will
necessarily produce the intended result. For this reason incentives by
themselves are not, from an economic standpoint, well suited to the
present radon problem. Finally, we must seriously question whether
or not these incentives are worth the cost unless they are used in
concert with other programs (i.e., information or regulation).
5.4. Regulation
In some respects the radon problem, in a traditional sense, is
difficult to regulate. If standards for acceptable radon levels were
mandated it would be up to the government to police these standards,
which would be a very costly process. Further, there is an equity issue
in imposing standards which make individuals bear mitigation costs,
for a substance which occurs naturally but unevenly.
Sweden has enacted the most ambitious regulatory approach to
radon so far. Based on health studies, the national government set
radon exposure standards in 1980 for radon concentrations within the
home and required all homes above a certain level to mitigate. New
127
-------
buildings were also required to meet standards in specified building
codes. The national government had the legal power to condemn
property or withhold necessary permits to those who did not comply.
Despite such measures, Swedens' programs have met with little
success.
Most of the responsibilities for promoting and monitoring
testing and mitigation practices were decentralized to the municipal
governments, which often had limited resources to devote to the
problem. Little in the way of information was provided to homeowners
who often received only a single-page list of mitigation alternatives as
guidance. The Swedish government relied heavily upon individual
voluntary testing and did little to motivate homeowners to test and
mitigate.
With the establishment of national standards in 1980, Sweden
had hoped to have mitigated most of the homes with high radon
concentrations by now. However, their standards have met with little
more compliance than EPA's voluntary action levels. As a result,
Sweden's regulatory solution to the radon problem (which did not
commit the necessary resources to enforcement) cannot be viewed as
a success.
In our view, a better regulatory strategy would be to focus on
action at the time of home or property sales. Such policies include
mandatory disclosure of radon based on testing at time of sale or a
standard applied at time of home sale. In addition to sales, other real
estate regulatory strategies may be applied during leasing, home or
building financing, building permit issuance, and inspection. Also,
mandatory testing and mitigation of public buildings would, though
128
-------
costly, reduce risk to public employees, and increase citizen
confidence in government concern about radon, which in turn might
increase private action.
\
A final suggested regulatory strategy is the development of
model building codes for radon. Such codes are usually quickly
adopted nationwide and would focus 'attention on the radon issue.
Among these options, the most appealing from our perspective
is mandatory testing and disclosure at time of home sale. However,
this approach must be combined with an information and awareness ,
strategy also applied at time of home sale. We discuss this approach
below.
5.5. Recommendation
This study has attempted to integrate three often disparate
viewpoints - psychology, economics, and law — within the context of
addressing radon pollution. Based on the results we have obtained, we
believe it is possible to combine these three disciplines to devise an
effective strategy to address radon contamination in homes.
From a psychological perspective the main policy question is
under what circumstances (if any) will people respond to warnings
about radon. This study has shown that general information
campaigns, when used alone, fail to accomplish radon reduction but
that radon information provided at a key point in time, during the
home sale transaction, gets the attention necessary to mitigate radon
levels. Study results have also indicated that social diffusion of radon
129
-------
information through, for example, realtors, contractors, and lenders,
may be effective.
From an economic perspective, we have determined that
general information campaigns alone do not appear to be cost
effective. Economic incentives, on the other hand, have worked to
encourage self-protective behavior but require consumers to be aware
of the problem; in order to be effective, the incentives must be
founded on information and awareness. Additionally, incentive
programs must be carefully designed to avoid inefficiency or bias. Of
course, the potential cost to the federal government is a major
problem with incentives. Regulations can be very effective in
motivating self-protective behavior, but must be carefully designed to
avoid inefficiency and excessive cost.
From a legal perspective, incentives or regulation may provide
an avenue to address the radon problem. As pointed out above,
incentives can be costly and inefficient. It also may be difficult for the
federal government to "police" incentives such as tax credits to ensure
that they are put to their intended use. General regulatory strategies
can suffer the same defects as incentive strategies. Nevertheless, our
research suggests that effective regulations can be formulated by using
the results of this study to design a regulatory strategy aimed at the
home sale transaction. This strategy would require mandatory
disclosure of radon level at time of home sale.
A home sales transaction strategy has certain drawbacks. First, *
it cannot address all radon contamination because it does not cover all
dwelling units. For example, people who rent rather than own would
not be affected by this strategy, although the owners of their dwellings
130
-------
could be. Second, it is a relatively slow approach. Because only 'about
5% of all homes are sold each year it might take as long as 14 years to
reach one half of all the currently existing homes. Third, it may be
inequitable, although this is mostly due to the nature of the radon
hazard. For example, its costs may fall hardest on the current
owner/seller of the home who may be required to test and mitigate a
condition that he did not create nor to which he contributed.
The home sale transaction regulatory strategy exploits a key
event - the decision to purchase a home - to focus the attention of the
home sale participants (e.g., buyer, seller, mortgage banker, realtor)
on the potential health effects of radon contamination. During the
home sales transactions, buyers and sellers are focused on the
condition of the home. Buyers are anxious to learn as much as possible
about the property. Sellers are likely to commit resources to correct
any perceived defects.
The home sales transaction strategy requires that before the
*
closing, radon tests be conducted, and their results obtained and
disclosed, to all participants in the home sale transaction. It takes
advantage of the psychological principles outlined above by providing
information about radon levels in a timely fashion so that protective
behavior is framed as part of a high profile, single decision that covers
a long time span. It also uses existing channels of social
communication to disperse information about radon by involving
mortgage bankers and realtors in disseminating radon information.
The home sales transaction strategy also is economically
efficient. Because the burdens of testing, disclosure, and mitigation
are imposed upon the participants to the home sale transaction, the
131
-------
federal government will not be forced to provide testing services or
offset costs of remediation. Since the buyer, realtor, and mortgage
banker have a strong self-interest in learning about radon, the strategy
is to a large extent self-policing. Additionally, because the strategy
does not require mitigation, it will aUow the buyer and seller to
negotiate for remediation of radon pollution, if necessary. Evidence
suggests that the result of such negotiations will almost always be to
remediate rather than compensate the buyer for accepting the risk.
Thus, free market economic forces shape the ultimate resolution of
the radon problem.
In order to implement this strategy, the Congress must enact
legislation empowering a federal agency such as the Environmental
Protection Agency to promulgate regulations requiring radon.testing
and disclosure of test results during the home sale transaction. Some
of the legal impediments to such potential legislation are reviewed in
Chapter 4 and Appendix IV. Traditionally, the federal government has
not .ntruded into home sale transactions, although it has enacted at
least one law requiring disclosure of certain closing costs in home
sales financed by "federally related mortgages."
132
-------
REFERENCES
Akerman J (1988). Economic valuation of risk reduction: The case of
indoor radon. EFI "Research Report 91-7258-266-9. Stockholm,
Sweden: Stockholm School of Economics.
Adler, R. S., and Pittle. R. D. (1984). Cajolery or command: Are
education campaigns an adequate substitute for regulation? Yale_
Journal on Regulation. I, 159-193,
Baum, A., Fleming, R., and Davidson, L. M. (1983). Natural disaster
and technological catastrophe. Environment and Behavior. 15,
333-354.
Berk, M. L., Mathiowetz, N. A., Ward, E. P., and White, A. A. (1988).
The effect of prepaid and promised incentives: Results of a
controlled experiment. .Imirnal of Official Statistics. 3(4), 1988.
Cole, L. A. (1990). Much ado about radon. The Sciences. 30(1), 18-23.
Carpenter, E.H., and Chester, ST., Jr. (1981). Are federal energy tax
credits effective? A Western United States survey. The Energy
Journal. 5, 139-147.
Desvousges, W. H., Smith, V. K., and Rink, H. H., Ill (1988).
Communicating Radon Risk Effectively: Radon Testing in
Maryland, EPA 230/03-89-048. U. S. Environmental Protection
Agency, Office of Policy, Planning, and Evaluation, Washington,
B.C.
Dillman, D. (1978). Mail and Telephone Surreys: The Total Design
Method. New York: John Wiley & Sons.
Dubow, F., McCabe, E., and Kaplan, G. (1979). Reactions to Crime:
A Critical Review of the Literature. Washington, D. C.: U. S.
Department of Justice, National Institute of Justice.
Jensen, D. D., Tome, A. E., and Darby, W. P. (1989). Applying decision
analysis to determine the effect of smoke detector laws on fire
loss in the United States. Risk Analysis. & 79-90.
Johnson, F. R., and Luken, R. A. (1987). Radon risk information and
voluntary protection: Evidence from a natural experiment. Kisj^
Analysis. 7(1), 97-107. .
Judd, C. M., and McClelland, G. H. (1989). Data Analysis: A Model
Comparison Approach. San Diego: Harcourt Brace Jovanovich.
133
-------
Kahneman, D. and Tversky, A. (1979). Prospect theory- An
analysis of decision under risk. Econometric 47, 263-291.
Kasperson, R, Berk, G., Pijawka, D., Sharaf, A., and Wood J (1980)
Public opposition to nuclear energy: Retrospects a^d prospecte
Science, Technology. * wuman vT&iPs 5, f 1-23 ProsPects.
ic P Borkin
'
Kunreuther, H., Ginsberg, R., Miller, L., Sagi. P.,
B. and Katz. N. (1978). Disaster In.snrJn.. '
.
Lessons. New York: Wiley.
McClelland G H., Schulze, W. D., and Kurd, B. (1989). The effect
of nsk behefs on property values: A case study of a hazardous
waste site. Risk Analysis submitted. Hazardous
Sandman, P. M., Welnstein. N. D., and Klotz, M. L. (1987) Public
response to the risk from geological radon. Journal of
CommunipaHnn 37(3), 93-108. --
Schulze, W. D., McClelland. G. H., and Coursey, D. L (1986)
Valuing risk: A comparison of expected utility with models from
cognitive psycho ogy. Technical Report, Laboratory for EconoiScs
and Psychology, Univ. of Colorado, Boulder, CO 80309. nomics
U'
- Per^ptions, and Actions,
Offlr- «p 1- - . Environmental Protection Agency,
Office of Policy, Planning, and Evaluation, Washington, D7c
" FiSher' A"
risk-
Johnson, F. R.
concem
(1986a). "A Citizen's Guide
Fn • t Wnat to Do A130111 It." OPA-86-004. U S
Environmental Protection Agency, Washington, D. C.
Meft Protect^on Agency (1986b). "Radon Reduction
Methods: A Homeowner's Guide," OPA-86-005 US
Environmental Protection Agency, Washington', D'. C.
Protection Agency, Washington, D. C.
134
-------
Weinstein, N. D. (Ed.). (1987). Taking Care: Understanding and
Encouraging Self-Protective Behavior. Cambridge: Cambridge
University Press.
Weinstein, N. D., Sandman, P. M., and Roberts, N. E. (1988).
"Homeowner Radon Mitigation," Report to the Division of...
Environmental Quality, New Jersey Department of Environmental
Protection, Trenton, New Jersey.
Weinstein, N. D., Klotz, M. L., and Sandman, P. M. (1989). Promoting
remedial response to the risk of radon: Are information
campaigns enough? Science. Technology. & Human Values. .14(4),
360-379.
Yuhnke, R. E., Silbergeld, E. K., and Caswell, J. E. (1987). "Radon:
The Citizen's Guide," Environmental Defense Fund, Inc., New
York, New York. •
135
-------
Appendix I
Facsimile Survey for the Washington, D. C. Area
. Campaign
136
-------
I. THE ISSUES
We would like to find out if your Air Chek radon test has been helpful to you and your family.
Please answer all questions for the home you tested.
Q-l Why did you decide to test your home for radon? (Circle all that apply)
<4 " ~~ -- -
1. PRICE OF TEST KIT
2. NEIGHBORS WERE TESTING-THEIR HOME
3. CONCERN ABOUT FAMILY'S HEALTH
4. WANT TO SELL HOME SOON
5. RADIO, TV, MAGAZINE, OR NEWSPAPER STORY 69.90
6. WANTED TO CHECK HOUSE BEFORE BUYING IT 1,02
4<20 2CX50 >50
38.27 43.55 38.64 40.65
4.59 5.38 10.80 14.19
76.88 77.27 78.71
2.69 2.27 2.58
70.43 65.91 63.87
0.54 0.00 0.65
80.10
1.53
Q-2
Have you ever: (circle answer for all that
1 . Received an insert in your utility
. hill rnnrernincf radon?.. .-.....••..: I
2.
3.
4.
5.
6.
7.
Called a radon 800 number
hotline?
Heard a public service
announcement about
radon on the radio?
Seen a public service
announcement about
radon on television?
Read an article about radon
in a newspaper?
Attended a community
group presentation or
meeting on radon?
Seen a TV news story or
documentary program
on radon?.
4<20
20<40
>50
<4
4<20
20<50
>50
4<20
20<50
_>5Q
4<20
20<50
>50
4<20
20<50
>50
<4
4<20
20<50
>50
<4
4<20
20<50
>5Q
apply)
NEVER
72.47
70.30
73.08
73.29
...NEVER
94.65
89.33
80.95
83.01
..NEVER
22.92
26.82
27.49
31.58
..NEVER
12.57
9.78
13.37
15.69
..NEVER
4.10
3.89
5.26
2,63
.NEVER
96.28
94.38
91.18
94.08
....NEVER
11.34
11.96
7.51
10.46 '
ONCE OR TWICE
27.53
27.88
26.92
26.03
ONCE OR TWICE
4.81
10.67 .
17.86
16.34
ONCE OR TWICE
38.54
34.08
32.16
39.47
ONCE OR TWICE
28.80
32.61
33.72
28.10
ONCE OR TWICE-
43.59
.38.33
38.60
42.11
ONCE OR TWICE
3.72
5.06
7.65
3.95
ONCE OR TWICE
52.58
52.17
54.91
52.94
MANY TIMES
0.00
1.82
0.00
0.68
MANY TIMES
0.53
0.00
1.19
0.65
MANY TIMES
38.54
39.11
40.35
28.95
MANY TIMES
58.64
57.61
52.91
56.21 '
MANY TIMES
52.31
57.78
56.14
55.26
MANY TIMES
0.00
0.56
1.18
1.97
MANY TIMES
36.08
35.87
37.57
36.60
137
-------
8. Seen a radon poster or
brochure In supermarkets
or other public places? NEVER
_<4 17.62
4<20 12.50
20<50 23.67
>50 20.13
ONCE OR TWICE
41.97
46.20
43.20
, 54.55
MANY TIMES
40.41
41.30
33.14
25.32
Q-3 When did you first hear about the radon problem? (circle number)
1. LESS THAN 6 MONTHS AGO
2. 6 MONTHS TO 1 YEAR AGO
3. 1TO3YEARSAGO
4. MORE THAN 3 YEARS AGO
0.00
39.58
52.60
7.81
4<20
1.09
30.43
61.41
7.07
20<50
0.57
34.09
60.23
4.55
>50
1.29
29.03
65.16
4.52
Q-4 How often have you discussed radon with: (circle answer for all that apply)
1. A family member? NEVER
<4 5.29
4<20 2.19
20<40 1.14
>50 0.67
2. A friend or neighbor? NEVER
. <4 12.09
4<20 9.47
20<40 6.59
>50 10.88
3. A government employee? NEVER
<4 77.78
4<20 75.00
20<40 71.33
>50 69.01
4. A doctor? NEVER
<4 96.82
. 4<20 94.16
2O<40 94.67
>50 91.30
5. A real estate agent? NEVER
<4 85.00
4<20 82.28
20<40 82.67
>50 80.58
6. A building contractor? NEVER
' <4 94.38
4<20 87.26
20<40 87.42
>50 82.73
7. Other (Please specify) loo% at <4
ONCE OR TWICE
48.15
42.62
37.14
28.00
ONCE OR TWICE
64.84
66.27
59.28
56.46
ONCE OR TWICE
18.52
19.23
20.00
25.35
ONCE OR TWICE
1.91
4.55
4.00
870
ONCE OR TWICE
11.88
16.46
16.67
15.83
ONCE OR TWICE
4.38
10.83
11.26
14.39
MANY TIMES
46.56
55.19
61.71
71.33
MANY TIMES
23.08
24.26
34.13
32.63
MANY TIMES
3.70
5.77
8.67
5.63
MANY TIMES
1.27
1.30
1.33
0.00
MANY TIMES
3.13
1.27
0.67
3.60
MANY TIMES
1.25
1.91
1.32
2.88
138
-------
Q-5 How useful were your Air Chek test results In your evaluation of radon levels
in your home? (circle number)
NOT AT ALL
EXTREMELY
mean
<4 5.81
4<20 5.10
20<5Q 5.19
< >50 5.43
H. ABOUT YOUR RADON TEST
Q-6 About when did you first test your home for radon?
M3MIH
mode
1988
4<20 1988
2CX50 1988
>50 1988
'Q-7 How many radon test kits did you use in your initial testing for radon?
KITS mean std.de\
<4 1.26 0.60
4<20 1.34 0.62
20<50 1.51 0.90
>50 1.55 0.74
Q-8 In which area of your home did you have the highest radon level?
<4 4<20 2CX50 >50
basement/lower level 86.43 89.53 96.49 - 97.87
bedroom 3.57 2.91 0.00 0.00
living room 5.71 3.49 2.34 1.42
kitchen/dining room 1.43 1.16 O.OO " 0.00
hallway 1.43 1.16 0.00 0.00
other , 1.43 1.74 1.17 ..0.71
Q-9 What were the test results for this area?
PICOCURIES PER LITER
139
-------
Q-10 Do you think your home has a radon problem? (circle number)
1. NO •»• Please skin tn OiimzHnn 19 96~05
-»- Please skip to Question 12.
4<20
60.00
20<50 >50
25.66 22.79
^
Q-ii w
number)
1.
2.
3.
4.
5.
\
X
nen do you plan to take action
NO ACTION REQUIRED
ACTION HAS BEEN TAKEN
SEVERAL WEEKS
SEVERAL MONTHS
SEVERAL YEARS
'
to fix. the
57.89
10.53
0.00
21.05
10.53
radon problem?
4<20
17.19
17.19
12.50
34.38
18.75
20<5Q
3.20
46.40
16.00
28.80
5.60
3.95 40.00 74.34 77.21
(circle
5.41
56.76
6.31
22.52
9.01
-
Q-12 How serious a health risk do you feel radon is to you and your family?
number) J J
(circle
LOW RISK
1
HIGH RISK
<4
4<20
20<50
>50
mean
2.76
2.74
3.87
4.28
Q-13 How confident are you that the test results sent to you accurately measure actual
radOn Pxnnsiirp in \rr\fir h/ii-noO r^f^^l^. n.._i \ J
radon exposure in your home? (circle number)
NOT AT ALL
CONFIDENT
VERY
CONFIDENT
Q-14
4<20
20<50
>50
mean
5.09
4.44
4.31
4.58
What information did you receive with your test results? (circle number for all that
<4 4<20 20<50 >50
1. LETTER FROM AIR CHEK AND A SECOND TEST KIT
2. AIR CHEK'S CITIZENS1 GUIDE TO RADON PAMPHLET 65.31
3. U.S. ENVIRONMENTAL PROTECTION AGENCYS
RADON REDUCTION METHODS PAMPHLET
4. A SECOND TEST KIT
5. OTHER (Please specify)
8.67
65.31
17.86
1.53
8.67
9.68
73.66
21.51
1.61
5.38
7.34
69.49
54.80
2.26
2.26
58.06
56.77
47.10
18.71
3.90
140
-------
Q-15 About how much of this information did you read? (circle number)
READ
NONE
READ
ALL
i -
99
mean std.dev.
<4 6.24 1.39
4<20 6.19 1.51
2CX50 6.22 1.34
>50 6.45 1.03
DIDNT RECEIVE
INFORMATION
Q-16 Do you still have any of this information? (circle number)
<4 4<20 20<50 >50
1. NO 31.55 23.56 19.54 20.39
2. YES 68.45 76.44 80.46 79.61
Q-17 How well did the information sent with your test results answer your questions concerning
radon? (circle number) • .
VERY POORLY
1 2
6
VERY WELL
mean
<4 5.48
4<20 5.00
20<50 5.18
>50 5.05
Q-18 Please circle all numbers that describe anything you have done to find out
about radon.
1. CALLED OR WROTE AIR CHEK
2. CALLED OR WROTE THE U.S. ENVIRONMENTAL
PROTECTION AGENCY
3. CALLED OR WROTE J.OCAL OR STATE PUBLIC
HEALTH AGENCIES
4. CONTACTED A LOCAL RADON REDUCTION
CONTRACTOR
5. CONTACTED A TESTING CONTRACTOR OTHER THAN
ABRCHEK
6. OTHER" (Please specify) ;
more
<4 4<2O 20<50 >50
6.04 3.95 5.29 12.00
9.89 12.43 17.65 22.00
6.59 11.86 19.41 30.00
1.10 3.39 17.65 20.67
3.30 3.95 16.47 17.33
14.29 18.08 12.94 9.33
141
-------
9-19 How difficult has It been to get information on radon reduction? (circle number)
DIFFICULT
1 2
EASY
mean
50 4.79
9-20 Have you performed a second or follow-up radon test since receiving
the results of your first test? (circle number)
PLEASE EXPLAIN YOUR REASON IN THIS BOX
1. NO
92.31
7.10
2. YES
Please skip to Question 25.
Q-21 About when did you retest your residence for radon?
.MCNIH
.YEAR Mode (all levels): 1988
Q-22 How many test kits did you use?
KITS
mean
<4 1.92
4<20 1.73
2Ck50 1.75
>50 1.68
9-23 In which area of your home did you have the highest radon
retest? •
<4 4^20 20<50 >50
basement/lower level 92.86 82.50 84.13 94.32
bedroom o.OO 5.00 4.76 0.00
living room 7.14 7.50 6.35 1.14
kitchen/dining room 0.00 2.50 0.00 0.00
hallway o.OO 0.00 1.59 0.00
other 0.00 2.50 3.17 4,55
9-24 What were the test results for this area?
PICOCURIES PER LITER
<4 4<20 20<50 >50
<4 57.14 15.38 13.79 13.33
new 4<20 42.86 - 82.05 36.21 21.33
levels: 20<50 0.00 2.56 44.83 26.67
>50 0.00 0.00 5.17 38.67
level in the
9-25 Have you moved from the home to which this survey was originally sent?
142
-------
(circle number)
1. NO
2. YES
97.41 97.28 98.29 96.73
2.59 2.72
1.71
3.27
Q-26 Were the results of your radon test part of the reason for your
move? (circle number)
<4 4<20 20<50 >50
1. NO 93.33 100.0 100.0 " 100.0
2. YES 6.67 0.00 0.00 0.00
Q-27 Have you taken action to reduce radon levels in the home which you tested?
(circle number) '
l.YES +*
Please continue with Section m below.
2. NO • —
<4 4<20 20<5O >50
2.73 14.29 41.52 48.63
97.27 85.71 58.48 51.37
Q-28 What are your future plans with respect to radon in this home?
(Circle all that apply) > •
<4 4<20 20<50 >50
1. NO FUTURE PLANS 65.82 28.49 14.12 14.84
2. FURTHER TESTING »> WHEN? 25.64 39.78 31.61 25.81
3. SELL HOME *> WHEN? 2.05 1.08 1.14 2.58
4. TAKE ACTION TO REDUCE RADON AT LATER DATE
-»> WHEN?
PLAN TO SPEND ABOUT $
2.05 11.83 15.43 16.13
TO REDUCE RADON.
5. OTHER (Please, specify)
1.53 2.70 3.41 1.29
Please skip the next section and continue with Section IV on page 9.
143
-------
m. RADON REDUCTION IN YOUR HOME
9-29 Why did you attempt to reduce radon levels In your home? (circle all that apply)
1. CONCERN ABOUT FAMILY'S HEALTH
2. NEIGHBORS WERE UNDERTAKING RADON
REDUCTION
3. REDUCTION WAS SUGGESTED WITH MY TEST
RESULTS
4. REDUCTION WAS SUGGESTED AS PART OF THE
SALE OF THE HOUSE
5. OTHER (Please specify)
—<4 4<20 20<50 >50
11.73 23.66 44.07 52.90
0.00 0.00 2.26 4.52
1.02 6.45 25.42 30.32
0.00 0.00 . 0.00 1.29
1.02 1.62 2.26 1.29
Q-30
9-31
9-32
About when did you begin to take action to reduce radon levels in your home?
iv/*~ivmj- mode
, MUV1H ^ TEtfL YEAR .MONTH
_<_4. 1988 2
4<20 1988 5
2CX50 1988 5
>50 1988 6
About when did you or will you complete your solution to your radon problem?
,_._^ ' mode
MJVffl _ -<£fR YEAR MONTH
<4 1988 3
4<20 1988 12
20<50 1988 12
>50 1988 12
Which of the following have you done to reduce radon levels in your home? (circle
all that apply)
1. Opened windows and doorssmore often
(natural ventilation)
2. Installed fans (active ventilation)
3. Sealed cracks and or/foundation Joints
4. Sealed sump holes and/or entry points in
plumbing system
Sealed or paved dirt crawl spaces
Painted walls and/or floors
Installed air cleaner/filter
Installed sub-slab ventilation system
Installed basement wall suction system
Installed heat recovery ventilation system
11. Installed basement pressurization system
12. Installed block-wall pressurization system
Ventilated crawl space
Contractor did something.
but I'm not sure what
Other ». (Please specify)
<4 4<20 20<50 ->50
10.71 21,15 28.25 35.48
4.59 4.30 12.43 12.26
3.06 8.60 31.64 30.32
5.
6.
7.
8.
9.
10.
13.
14.
15.
0.51
0.51
2.55
1.02
0.00
0.00
0.00
0.00
0.51
0.51
6.99
1.61
3.23
2.15
1.08
0.00
0.00
0.00
0.00
2.15
24.86
5.06
10.17
1.69
9.O4
3.39
0.00
0.00
0.00
5.08
28.39
6.45
9.68
4.52
16.77
2.58
0.00
0.00
0.00
3.23
0.00
0.00
0.00
0.00
0.00
1.13
0.00
1.29
144
-------
Q-33 Who performed these radon reduction measures on your home? (circle all that
apply)
<4 4<20 20<50 >50
1. DID THE WORK OURSELVES
2. LANDLORD
3. PRIVATE CONTRACTOR
4. OTHER
Q-34 How difficult was it to find a trustworthy radon reduction contractor?
(circle number) .
NOT
DIFFICULT
VERY
DIFFICULT
123,4567
mean std. dev.
<4 5.5 2.12
4<20 4.24 . 2.71
20<50 3.92 2.30
' >50 3.30 2.41
Q-35 Did you work with a radon reduction contractor who was certified or endorsed by the U.S.
Environmental Protection Agency or a state or local government agency? (circle number)
1. NO ,
2. YES .
3. DONTKNOW .
Q-36 How important was it that your radon reduction contractor was certified? (circle
number) ,
NOT
IMPORTANT
VERY
IMPORTANT
567
mean std. dev.
<4 5.00 0.00
4<20 5.50 225
20<50 5.59 2,06
>50 6.60 0.78
145
-------
9-37
How did you determine which radon reduction methods to use? (circle all that
. „., «.,^.v,w,_.li,_,1., \^t_/i\ iiwl^iUK
4. NEWSPAPER AND MAGAZINE ARTICLES
5. RADIO OR TV FEATURES
6. REAL ESTATE AGENT
7. FRIENDS OR NEIGHBORS
8. OTHER (Please specify)
6.67
3.06
0.51
4.62
4.10
0.00
0.00
(local EPA) 0.00
4<20
11.83
20<5O >50
28.25 27.74
9.14
2.15
10.22
8.06
0.00
1.08
1.63
24.86
12.43
19.21
16.38
2.26
6.78
2.26
27.10
16.77
21.29
11.61
0.00
3.87
1.96
9-38 About how much have you spent on radon reduction?
INSTALLATION COST
$
mean std. dev.
115.55 227.98
319.21 1029.02
294.43 381.81
344.83 395.43
OPERATING AND MAINTENANCE
mean std.dev.
<4 43.85 78.06
4<20 22.14 44.07
20<50 17.28 29.57
>50 23.88 25.67
PER YEAR
How much do you
number)
. NOTATALL
1
thCSe SCtl0nS have reduced y^ ** *** radon? (circle
VERY MUCH
6 7
mean sjd.dg
<4 4.00 1.73
45Q 5.07 1.87
146
-------
Q-4O . Have you had your home retested for radon since completing your radon
reduction efforts?
, <4 4<20 2CX50 >50
1. NO
Please skip to Question 44 on page 9.
2. YES
92.31 84.21 68.35 59.76
7.69 15.79 31.65 40.24
Q-41 About when did you again test this residence for radon?
MONTH _ YEAR
MXE
year month
<4 1988 6
4<20 1988 4
20<50 1988 4
>50 1988 5
Q-42 In which area of your home did you have the highest radon level?
<4 4<20 2CX50 >50
basement/lower level 100.0 83.33 95.65 90.63
bedroom 0.00 0.00 0.00 3.13
living room 0.00 0.00 4.35 3.13
other ' 0.00 16.67 0.00 3.13
Q-43 What is the new level in this area of your home?
PICOCURIES PER LITER
<4 4<20 20<50 >50
<4 50.00 28.57 30.43 67.74
new 4<20 50.00 71.43 56.52 25.81
• levels: 20<50 0.00 0.00 13.04 6.43
>50 0.00 0.00 0.00 0.00
Q-44 What are your future plans for action with respect to radon?
<4 4<20 20<50 >50
30.61 22.58 18.64 16.13
11.22 24.73 32.20 38.06
0.00 1.61 7.34 11.61
1. NO FUTURE PLANS'
+* Please skip to Question 46.
2. FOLLOW UP TESTING
—»• WHEN?
3. BEGIN FURTHER REDUCTION
147
-------
9-45
How much do you plan to spend on further radon reduction? (circle number)
<4 4<2O or^f^n -os/1
$0-$99
2. $100-$199
3. $200-$499
4. $500-$999
5. $1,000 - $1,999
6. $2,000 - $2,999
7. $3.000 - $4.999
8. MORE THAN $5,000
<4
85.71
0.00
9.52
4.76
0.00
0.00
0.00
0.00
4<20
75.00
15.00
5.00
5.00
0.00
0.00
0.00
0.00
20<5Q
49.09
20.00
12.73
10.91
7.27
0.00
0.00
Q.OO
>50
53.73
8.96
20.90
13.43
2.99
0.00
0.00
0.00
IV. ABOUT YOU
9-46 What is your sex?
1. FEMALE
2. MALE
_<4. 4<30 20<50 _
39.06 35.56 29.14 30.92
60.94 64.44 70.86 69.80
Q-47 What is your age:
YEARS
9-48
mean
47.00
48.44
45.13
43.24
yourself, how many members in your household are in
age group
(MEANS)
UNDER 18 YEARS OF AGE ~^ -*3tt MsSfl -ZfiQ
18-64
65 AND OVER
9-49 Are you presently:
1. EMPLOYED
2. UNEMPLOYED
3. RETIRED
4. FULL TIME HOMEMAKER
5. STUDENT
6. EMPLOYED PART-TIME
7. OTHER (Please specify)
9-50 Do you own this residence?
l.NO-
2. YES
1.63
2.07
1.28
-5Q
73.68
1.97
10.53
7.89
0.00
3.95
1.97
_>§Q
2.63
97.37
148
-------
Q-51
Q-52
Q-53
9-54
What type of residence is this? (circle number)
1. SINGLE FAMILY DETACHED HOME
2. MOBILE HOME OR HOUSE TRAILER
3. DUPLEX, TRIPLEX OR FOURPLEX
4. TOWNHOUSE CONDOMINIUM OR ROWHOUSE
5. APARTMENT.
(Building with five or more residences)
What is the zip code of the home you tested?
How long have you lived in this residence?
<4 4<20 20<50 >50
78.24 94.54 95.45 97.40
0.00 O.OO 0.00 0.00
0.55 0.57 0.65
3.22
16.58 ..4.92.
YEARS
mfean
<4 10.88
4<20 9.81
2CX50 8.74
>50 8.65
'How much longer do, you expect to live here?
YEARS
4<20
>50
mean
9.05
11.06
12.14
14.02
3.98
1.95
2.07 ..0.00 0.00 0.00
std.dev.
9.18
9.27
a 19
7.63
Q-55 What do you think the current market value of your home is?
Q-56
•
*
How much formal education have you
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
NO FORMAL EDUCATION
SOME GRADE SCHOOL
COMPLETED GRADE SCHOOL
SOME HIGH SCHOOL
COMPLETED HIGH SCHOOL
TRADE SCHOOL
SOME COLLEGE
COMPLETED COLLEGE
SOME GRADUATE WORK
. ADVANCED COLLEGE DEGREE
<4
4<20
20<50
>50
178,178.28
222,164.84
196,674.69
184.696.55
completed? (circle number)
<4
"o.oo
0.00
0.53
1.06
7.94
3.17
25.40
19.05
17.99
24.87
4<20
0.00
0.56
0.56
0.56
11.30
2.82
22.03
20.34
14.12
27.68
2CX50
0.00
0.00
0.00
3.49
8.14
2.91
30.23
19.19
8.14
27.91
>50
0.00
0.65
1.31
1.96
15.03
3.27
16.99
26.80
16.34
17.65
(IN DOLLARS)
mean std.dev.
110.226.89
124.802.47
135.958.55
93.863.32'
149
-------
Q-57 Do you smoke?
1. NO
2. YES
<4 4<20 20<50 >50
88.42 86.52 82.76 90.85
11.58 13.48 17.24 9.15
Q-58 Do you generally wear a seat belt when you travel by automobile?
l.NO
2. YES
4<20 20<50 >50
9.39 4.57 6.54
90.61 95.43 93.46
Q-59 What is your total annual household Income before taxes and
other deductions? (circle one)
1. UNDER $9,999
2. $10,000 - 19,999
3. $20,000-29,999
4. $30,000-39,999
5. $40,000-49,999
6. $50,000-59,999
7. $60,000-69,999
8. $70.000-79,999
9. $80.000-89,999
10. $90.000 - 99.999
11. OVER $100.000
** •; Is,there anything we may have overlooked? Please use this space for any
additional comments you would like to make concerning radon, your radon test or
radon reduction.
<4
2.37
2.96
9.47
11.83
17.16
17.16
13.61
8.88
3.55
2.96
10.06
4<20
0.00
3.23
5.81
9.68
11.61
15.48
18.06
12.26
10.97
3.87
9.03
20<50
0.67
2.68
6.04
8.72
22.15
18.79
14.77
8.05
4.70
5.37
8.05
>50
0.71
5.71
2.14
15.00
17.86
22.14
12.86
10.00
6.43
2.14
5.00
e, ™ Your contribution to this effort is very greatly appreciated. If you would like a
summary of the results, please print your name and address on the back of the return
envelope (NOT on this questionnaire). We will see that you receive it
150
-------
Appendix II
Facsimile Survey for the Boulder County,
Colorado Study
151
-------
I. INTRODUCTION
First we'd like to ask you a few questions about your experience with radon.
9-1 Did you first hear about the radon problem less than 6 months ago,
6 months to 1 year ago, 1 to 3 years ago, or more than 3 years ago?
Overall Test
1. LESS THAN 6 MONTHS AGO 5.3% 39%
2. 6 MONTHS TO 1 YEAR AGO 136 97
3. 1TO3YEARSAGO 53.0 558
4. MORE THAN 3 YEARS AGO 25.5 305
5. DONTKNOW 07 o'
9-2
9-3
297
9-4
9-5
296
9-6
297
6. NEVER —> (SKIP TO Q-38)
2.0
0
Was this before or after you began looking for a home in Boulder County?
Overall Test No test
1. BEFORE 82.5% 85% 799%
2. AFTER 17.5 15 20.1
On a scale from 1 to 7, where 1 indicates not at all Important and 7 indicates
very important, how important was it to you to buy a home with a safe
radon level?
19.5% 9.1% ILP/o 9.8% 14.50,6 9.40^ 266%
NOTATALL 1 2 34 5 6 7 VERV
IMPORTANT IMPORTANT
Overall mean = 4.25, Test mean = 5.3, No test meaij. = 3.1
When looking for a new home in Boulder County, did you always, usuallv
sometimes, rarely, or never ask the seller or realtor if the home had been'
tested for radon?
1. ALWAYS
2. USUALLY
3. SOMETIMES
4. RARELY
5. NEVER
29.0%
5.1
7.4
7.4
51.2
When looking for a new home in Boulder County, did you ever stop pursuing
your interest in a home specifically because of radon?
1. YES
2. NO
5.4%
94.6%
Did you own and sell a home before buying your present home?
Overall Test No test
1. YES 58.9% 60.4% 57.3%
2. NO 41.1 39.6 42.7
9-7 Was your previous home tested for radon?
Overall Test
N=173 1. YES 35.3% 47.3%
2. NO 64.7% 52.7
152
-------
II. RADON TESTING
Now we'd like to ask you- some specific questions about radon and your new home.
Q-8
N = 303
Q-10
N=136
Q-ll
N=154
Q-12
N=154
Q-13
N=152
Was your new home tested for radon before the sale closed?
1. YES
2. NO
3. DONTKNOW
9-9
,- 44.6%
48.8
6.6
Was your new home tested for radon after the
- sale closed?
N=163 1. YES——> (GO TO Q-ll) 11.7%
2. NO > (SKIP TO Q-37) 88.3
Did the first radon test take place before you first looked at the house, before
sale negotiations began, before sale negotiations completed, or before the
sale closed?
1. BEFORE LOOKING • 19.1%
.2. BEFORE NEGOTIATIONS BEGAN 13.2
3. BEFORE NEGOTIATIONS COMPLETED 29.4
4. BEFORE SALE CLOSED 38.2
Who requested that the first radon test be performed?
1. SELF OR SPOUSE
2. SELLER
4. REALTOR
5. BUILDING CONTRACTOR
6. EMPLOYER
7. BANKER
8. DONT KNOW
9. OTHER (
67.5%
7.8
7.8
3.9
2.6
3.2
3.9
3.2
Who conducted the first radon test?
1. SELF OR SPOUSE 18.2%
2. SELLER 5.2
3. REALTOR 3.9
4. BUILDING CONTRACTOR 7.1
5. TESTING FIRM 37.0
6. DONT KNOW 16.2
7. OTHER( 12.3
Was someone else still living in the home when the first radon
test was conducted?
1. YES
2. NO
55.9%
44.1%.
153
-------
Q-14
N=154
9-15
Who payed for, the first radon test?
1. SELF OR SPOUSE 51 3%
2. SELLER 22 7
3. REALTOR 2 6
4. BUILDING CONTRACTOR 65
5. DONT KNOW 9 i
6. OTHER( '78
mSsyuredetame What thC highest radon level found to y°ur new home was,
. pe?TitS! 6S PCr ltGT? Remember t*131 the EPA ac«on
PICOCURIES PER LITER
DONT KNOW/NOT SURE
Mean = 4.2
Min = 0.2
Max = 18.0
Q-17
Q-18
Q-19
N=151
9-16
Would you say the highest radon level
found in your new home
was very low, low, about average,
high, or very high?
1. VERYLOW
2. LOW
3. ABOUT AVERAGE
4. HIGH
5. VERY HIGH
6. DONT KNOW
32.1%
32.1
14.8
7.4
, 12.3
Based on these test results, did you think the home had a radon problem?
1. YES
2. NO
14.4%
86.6%
vrdi, 10 7' r!fre l tadicates not at ^ confident and 7 indicates
very confident, how confident are you that the test results you were told
were an accurate measure of the actual radon level in the home?
MOT AT ATT L3% 2'6% 4<6% 13'2% 19'7% 26.3% 31.6%
NOT AT ALL 1 2 3 4 5 67
"hen
1. YES
2. NO
78.1%
21.9
154
-------
III. RADON MITIGATION
Q-20
N=155
Q-22
N = 18.
Q-23
N = '25
Q-24
N = 25
Q-25
N=14
Was any attempt made to reduce the radon level in your new home
before the sale closed?
1. YES
2. NO
3. DONT KNOW
11.6%
87.1
1.3 ,
/ i '
Q-21 Was any attempt made to reduce the radon level in
your new home after the sale-closed?
N=137 1. YES-—-T > (GO TO Q-23) 4.4%
2. NO .- > (SKIP TO Q-30) 95.6
3. DONT KNOW—> (SKIP TO Q-30)
">
Did the first reduction effort take place before you first looked at the home.
before sale negotiations began, before sale negotiations completed, or
before the sale closed?
1. BEFORE LOOKING " 22.2%
2, BEFORE NEGOTIATIONS BEGAN 5.6
3. BEFORE NEGOTIATIONS COMPLETED 27.8
4. BEFORE SALE CLOSED . 44.4
Who first requested that an attempt, to reduce the radon level be made?
1. SELF OR SPOUSE 76%
2. SELLER '8 .
4. REALTOR
5. BUILDING CONTRACTOR 4
6. EMPLOYER
7. BANKER 4
8. DONT KNOW. 8
9. OTHER ( " )
Who did the work involved in trying to reduce the radon level?
1. SELF OR SPOUSE 12%
2. BUILDING CONTRACTOR 36
3. RADON MITIGATION FIRM 16
4. SELLER • 12
5. DONT KNOW , 20
6. OTHER( -
J
About how much did the reduction effort cost?
mean = $522, low = $20, high = $1500
$
155
-------
Q-26
Who payed for it?
9-28
N-13
1. SELF OR SPOUSE
2. SELLER
3. REALTOR
4. BUILDING CONTRACTOR
5. DONT KNOW
6. OTHERL
26.1%
47.8
8.7
13.0
. 4.3
Q-27 was a radon test conducted after the reduction effort?
1. YES
2. NO > (SKIP TO Q-30)
3. DONT KNOW > (SKIP TO Q-30)
70.8%
20.8
8.3
Mean = 2.8
PICOCURIES PER LITER
DON'T KNOW/NOT SURE
Q-29 Would you say the radon level
: found after the reduction effort
N = 4 was very low. low, about average
high, or very high?
,. ' 1. VERYLOW
2. LOW
3. ABOUT AVERAGE
4. HIGH
5. VERY HIGH
6. DONT KNOW
IV. RADON TRANSACTIONS AND NEGOTIATIONS
Now we'd like to know a little bit about radon and your home-buying transactions.
about^don^^b ri°e °f ^ home reduced specifically because of concern
N=154
1. YES
2. NO
1.3%
98.7
Q-31 By how much was the price reduced because of radon?
N=l *
156
-------
Q-32
N=155
Were any clauses concerning radon written Into the final sales contract
for the home? .
1. YES 39.4% '
2; NO 56.8
3. DONTKNOW 33 -
Q-33 Was the contract contingent on a radon test?
N = 61 1. YES 90.2%
2. NO 9.8%
3. DONTKNOW
Q-34 Did the contract state that the seller must reduce the radon
'• level in the home if unsatisfactory?
N = 61 1. YES 52.5%
2. NO 37.7
3. DONTKNOW 9.8
Q-35 Did the contract specify that the home must meet a specific
radon level, in picocuries per liter?
N=61 1. YES 52.5%
2. NO 36.1
3. DONTKNOW 11.5
IF YES: " Q-36 What was this level?
16 = 4.0, 1=20
N=17 PICOCURIES PER LITER
DON'T KNOW
Q-37 On a scale from 1 to 7, where 1 indicates not at all serious and 7 indicates
very serious, how serious a health risk do you feel radon is to you and
N = 295 your family right now?
54.9% - 22.7% 12.9% 8.1% 9.8% 3.7% 7.8%
NOT AT ALL 1 2 3 4 56 7
SERIOUS
SERIOUS Overall mean = 2.8, Test mean = 2.8, No test mean = 2.7
V. GENERAL TRANSACTIONS AND NEGOTIATIONS ' .
Now we'd like to ask a few general questions about your new home purchase.
Q-38 Are you the first owner of your new home?
Overall Test No test
N = 303 1. YES 20.8% . 18.8% 22.8%
2. NO 79.2 81.2 77.2
Q-39 How many weeks went by from the time you gave the seller your initial
offer until the sale closed?
Overall mean = 6.4, Test mean = 6.3, No test mean = 6.5
N = 295 _WEEKS
Q-40 How many offers did you make on the home before one was accepted?
Overall mean =1.7
N = 291 Testmean = 1.8 ; OFFERS
No test mean = 1.6
VERY
157
-------
9-41
N = 2
9-42
N = 2
9-43
N = 2
9-44
What was the Initial asking price of the seller of your new home?
$ _ Mean = 123,244 ___
What was your initial offer on the home?
$, _ Mean = 124,364 __
What was the final selling price of the home?
$ _ Overall mean = 118,118
P.,, ,,,,,,, Test mean =127, 180 No test mean = 108,9 12
Did you deal directly with the seller or employ the services of a realtor?
1. DIRECTLY WITH SELLER
2. EMPLOYED REALTOR
79.3
842
9-45 Did your realtor give you any information or
advice about radon?
Overall Test
N = 231 l.YES 43.7% 575%
2. NO 56.3 42.5
9-46 Did your realtor agree to reduce any fees or
commissions in order to close the deal?
No test
26.9%
73.1
N = 235
1. YES
2. NO
IF YES:
9^7
8.9%
91.1%
How much was the reduction?
mean = 707
VI. ABOUT THE RESPONDENT
We're almost finished. I'd now like to ask you a few final questions about yourself.
9-48 GENDER (DO NOT ASK):
36.2
38.3 34
§r^ What is your age? YEARS
Overall mean = 37.4, Test mean = 37.6, No test mean = 37.2
158
-------
Q-50
N = 300
Q-51
N = 299
Q-52 .
N = 299
Q-53
N = 300
Q-54
N = 298
9-55
N = 300
Q-56
N = 299
1.
2.
3.
4.
5.
Q-57
N.= 232
Including yourself, how many members of your household are
under 18 years of age? 0.88
between age 18 and age 65? 1.93
older than age 65? 0.04
Are you or a member of your household employed by IBM?
Overall Test No test
1. YES 11.7% • 17% 6.2%
2. NO 88.3 83 93.8
Are you or a member o'f your household employed by U. S. West?
1. YES
2. NO
Do you smoke?
1. YES
2. NO
2%
98%
Overall
11%
89
Test
11%
89
No test
11%
89
Does any other member of your household smoke?
Overall Test No test
1. YES 11.7% 13.1% 10.3%
2. NO 88.3 86.9 89.7
Do you generally wear a seat belt when traveling by automobile?
Overall Test No test
1- .YES 91% 94.2% 87.7%
2. NO 9 5.8 12.3
How much formal education have you completed?
NO FORMAL EDUCATION
SOME GRADE SCHOOL
COMPLETED GRADE SCHOOL
SOME HIGH SCHOOL
COMPLETED HIGH SCHOOL
6. TRADE SCHOOL
7. SOME COLLEGE
8. COMPLETED COLLEGE
9. SOME GRADUATE WORK
0.3 6. TRADE SCHOOL , 1.3%
0.3 7. SOME COLLEGE .19.7
32.8
13.0
9.4% 10. ADVANCED COLLEGE DEGREE 23.1
• Median = completed college
What is your household's total annual income before taxes and other
deductions?
1. UNDER $9,999 OS
2. $10,000 - 19,999 3%
3. $20,000 - 29,999 11.2
4. $30,000 - 39,999 16.8
5. $4O,000 - 49,999 15.9
6. $5O,000 - 59.999 16.4
7. $60.000 - 69,999
8. $70,000 - 79,999
9. $80.000 - 89,999
10. $90,000 - 99.999
11. OVER $100,000
9.9%
9.1
4.7
2.6
9.5
Median = $50,000 - 59,999
That's it, we're finished. Thank you very much for your time. Your responses will be
of great value to us in evaluating public health management policies for radon.
159
-------
Appendix III
Federal Laws Addressing Radon Issues
The Radon Pollution Control Act
The Radon Pollution Control Act of 1988, Pub. L. No. 100-
551,102 Stat. 2755 (1988) ("RPCA" or the "Act"), to be codified as
Title III of the Toxic Substances Control Act, 15 U.S.C. §§ 2601 -
2629 (1983), was passed by Congress and signed into law by the
president on October 28, 1988.
The Act.sets a national goal for radon in buildings; requires the
United States Environmental Protection Agency to update and
republish its "Citizen's Guide to Radon" (EPA, 1986a); orders EPA to
develop model construction techniques and standards for controlling
radon; requires EPA to develop and implement activities to assist state
radon programs; and provides a grant assistance program for state
radon programs. It does not provide for direct federal regulation of
radon, nor does it establish national radon standards. This appendix
reviews the provisions of the RPCA in detail.
Sections 301. 3O2, and 303
Section 301 of the RPCA declares that the long-term goal of the
United States is to lower the radon levels in buildings to the same
level as ambient air. Section 302 defines certain terms used in the'
160
-------
Act, and Section 303 requires EPA to revise and republish its Citizen's
Guide to Radon. EPA must include the following information in its
updated guide:
1. a series of "action levels" indicating the health risk associated
with different levels of radon exposure; and
2 certain "other information," including a discussion of the, •
increased health risk associated with the exposure of potentially
sensitive populations to different levels of radon, the increased
health risk associated with radon exposure and risk taking
behavior, the cost and technological feasibility of reducing radon
concentrations, the relationship between short-term and long-
term testing techniques and measurements and action levels,
and outdoor radon levels around the nation.
The Act states that the guide should be revised "li]n order to make
continuous progress toward the long term goal established in Section
301." A key piece of the RPCA's legislative history explains the intent
of this section in greater detail:
The legislative requirement to update the Citizen's Guide is
principally based upon the Committee's concern that the
• public is interpreting radon levels as safe if they fall below
EPA's action level of 4 picocuries per liter. The EPA guidance
document currently advises that "follow-up measures are
probably not required" if screening measurements are less
than 4 picocuries per liter. In addition EPA states that
"[ejxposures in this range are considered above average for
residential structures." The Committee believes that many
people have misinterpreted EPA's designated action level
and the statements in the current Citizen's Guide as meaning
that there is little or no risk from radon levels below .
4 picocuries per liter. . . . [AJlthough EPA's current Citizen's
Guide includes a radon risk evaluation chart and other very
useful information, the Committee believes the public is
relying upon EPA's 4 picocuries per liter as a health-based
standard. . . . The Committee wants to encourage the public
to make efforts to bring radon levels in existing homes
161
-------
and buildings down as low as practicable. (H.R. 1047 100 Cong
2d Sess., 12-13 reprinted in 1988 U.S. Code Cong & Admin
News 3617-18 (hereinafter House Report 1047))20
This passage makes it clear that Congress felt that the current
Citizen's Guide and the 4 picocuries per liter standard was misleading.
Congress mandated that EPA revise its guide to clarify that EPA's
present "action level" entails some health risks, and to add other
action levels and their related risks. In fact, House Report 1047 states
that "EPA should not designate a single particular radon level as an
'action level1 or 'guidance level'. . . . Rather, the Citizen's Guide should
contain a series of action or guidance levels including levels below 4
picocuries per liter so that homeowners and other members of the
public can evaluate the health risk at each radon level." (House Report
1047, reprinted in 1988 U.S. Code Cong. & Admin. News 3618-19
(emphasis in original))
f . ,
The House Report also shows concern about "the reliability and
accuracy of currently available radon testing techniques." It notes that
"information [is available] suggesting that results from instantaneous or
short-term radon tests may not provide a reliable and accurate
indication of long-term radon levels. The Committee is concerned
about people making decisions not to mitigate based on low readings
from short-term radon tests....[T]he Committee expects EPA to
20According to the Act's legislative history, the Senate bill was
passed in lieu of the House bill after its language was amended to
contain the text of the House biU. Thus, the House report, prepared by
the Committee on Energy and Commerce, is essential in construing
the meaning of the Act (see Allied Towing vs. Great Eastern Petroleum
Corp.. 642 F. Supp. 1339, 1351-52 (E.D. Vs. 1986).
162
-------
consider whether the Agency should recommend that only results
from long-term tests be used." (House Report 1047 at 16, reprinted
in 1988 U.S. Code Cong. & Admin. News 3620)
Section 304
This section requires that EPA develop model construction
standards and techniques for controlling radon levels in new buildings
by June 1, 1990. To the "maximum extent possible," EPA is directed
to consult with organizations involved in establishing national building
standards and techniques. In developing the standards and
techniques, EPA must take into account the geographic differences in
construction types and materials, geology, weather and any other
•i , , ' i ' .
factors that may affect radon levels. A draft copy of these standards
and techniques shall be made available to the public for review and
comment.
The RPCA does not empower the Agency to issue regulations
requiring adherence to the building standards, nor does it otherwise
make these standards effective as a matter of law. It states merely that
EPA "shall work to ensure that organizations responsible for
developing national model building codes, and authorities which
regulate building construction ... adopt the Agency's model standards
and techniques."21 (RPCA §304)
2iHowever, see the discussion of RPCA Section 306(d) below.
This section provides that EPA shall give preference for grant
assistance to states that have made "reasonable efforts" to ensure the
adoption of these building standards.
163
-------
In its explanation of this section, the Committee noted that EPA
has collaborated with the National Association of Home Builders
(NAHB) to develop a guidance document identifying construction
techniques that can significantly reduce radon levels in new home
construction. EPA is encouraged to continue its efforts to promote the
nationwide use of construction techniques that reduce radon levels.
The Committee desires that these national building standards be
incorporated into model national building codes that will be adopted
by state and local communities. (House Report 1047 at 16-17,
reprinted in 1988 U.S. Code Cong. & Admin. News 3621)
Section 304 is not intended to require EPA to establish a
performance-based standard at outdoor radon levels. The initial
model construction standards and techniques and subsequent
revisions are meant to "assist the public in making progress toward
the national long-term goal." (House Report 1047 at 16-17, reprinted
in 1988 U.S. Code Cong. & Admin. News 3621)
Section 305 .
Section 305, entitled 'Technical Assistance to States for Radon
Programs," requires EPA to develop and implement certain activities
designed to assist state radon programs, including a clearinghouse for
radon information, a voluntary proficiency program for rating radon
measurement devices and firms and individuals that offer radon-
related services, training seminars for public and private firms dealing
with radon, publication of public information materials, operation of
state/federal cooperative projects, demonstration of radon mitigation
164
-------
methods and establishment of a national data base regarding the
amounts and location of radon. It also allows EPA to provide certain
discretionary assistance to states if states (or a state) request such
assistance, including designing and implementing a radon survey, a
public information and education program, and a program to control
radon in existing and new structures.
This section also requires EPA to provide information regarding
technology and methods of radon assessment and mitigation to private
professional organizations. By July 1, 1989, and annually thereafter,
the Agency must prepare a plan- to implement'this section, and submit
it to Congress. _
Congress authorized the appropriation of $1.5 million22 so that
EPA could establish a "proficiency rating program, and training
seminars" for any person desiring such services, including private
firms and organizations and state and local governments. The Act
requires EPA to charge for attendance at the training seminars and for
participation in the proficiency rating program to "cover the operating
costs of such proficiency rating program and training seminars."
(RPCA §305(e)(2)) No charge will be imposed upon state or local
governments. During the first three years of the rating program and
seminars, the charges imposed are to be in excess of the operation
costs. The excess amount collected is to be used to reimburse the
General Fund of the United States Treasury for the $1.5 million
22RPCA Section 305(f)rstates that for the purposes of carrying
out Sections 303, 304, and 305, "[tjhere is authorized to be
appropriated an amount not to exceed [$3.0 million]."
165
-------
appropriated to establish the training seminars and rating program.
(RPCA §305(e)(4))
House Report 1047 states that EPA should provide "seed money"
to help states initiate and establish radon programs. This financial
assistance was only designed to help get state programs "off the
ground" and was not intended to establish a permanent federal grant
program. (House Report 1047 at 17, reprinted in 1988 U.S. Code
Cong. & Admin. News at 3622)
Section
Section 306 covers grant assistance to states for radon
programs. It provides that EPA may make a grant to a state for "the
purpose of assisting the State in the development and implementation
of programs for the assessment and mitigation of radon."23 (RPCA
§306(a)) Section 306 lists some of the information that must be
provided in a grant application, such as a description of the
seriousness and extent of radon exposure in the state, the
identification of the state agency that has responsibility for radon
23The Act limits the federal share of the cost of radon '
m any fiscal year according to the following schedule (a] I TO
^ purchase radon measurement equipment
pay for costs of demonstration of radon
C°StS °f general overhead
25% of "" "mount of
166
-------
programs and which will receive the grant, a description of the radon-
related activities and programs proposed by the state, and a three-year
plan that outlines long-range program goals and objectives, desired
federal funding, and state funding. Activities eligible for funding N
include radon surveys, development of public information and
educational materials, implementation of programs to control radon,
purchase and maintenance of analytical equipment and measurement
devices, payment of general overhead and program administration
costs, and development of a data storage and management system.
Starting in 1991, preference for grant assistance will be given to
states that "have made reasonable efforts to ensure the adoption, by
the authorities which regulate building construction within that State
or political subdivisions within States, of the model construction
standards and techniques for. new buildings developed under section
304." The RPCA provides that EPA shall fuUy support eligible activities
contained in state applications with the full amount of funds. If the
state applications exceed the total funds available, EPA shall give
priority to activities or projects based upon the seriousness and extent
of the radon problem, the potential that the proposed activity or
project will reduce radon levels, the potential for development of
innovative radon assessment techniques or program management
approaches, and any other criteria designated by EPA.
An award of federal funds is conditioned upon the state
providing to the federal government all radon-related information,
including survey results and risk communication studies.24 Each state
.« 24States may use federal grant funds to assist local governments
. in carrying out their radon programs (RPCA Section 306(g))
167
-------
also must maintain a list of firms and individuals within the state that
"have received a passing rating under the [EPA] proficiency rating
program referred to in Section 305(a)(2)." The list shall include the
rating received by each firm and shall be available to the public. (RPCA
§306(h))
Congress has authorized the appropriation of $10 miUion for
grant assistance for fiscal years 1989,1990, and 1991. This money may
not be used to cover the costs of the proficiency rating program
established pursuant to section 305. (RPCA §306(j)(5))
Section 3Q7
This section requires EPA to conduct a nation-wide study of
radon in schools. Based on geological data and data concerning radon
in homes and other buildings, EPA must identify and compile a list of
high probability radon contamination areas where schools are located.
EPA is empowered to assist state agencies in carrying out this survey.
It must provide to the state agency a list of high probability areas,
other data about schools in the state, technical guidance, and
information concerning methods of reducing radon contamination.
EPA also may provide testing devices and the services of EPA's
laboratories to evaluate radon test information.
On or before October 1, 1989, EPA must submit to Congress a
status report regarding the school study. By October 1, 1990, EPA
must submit its final report setting forth the results of the study and
its recommendations. Congress has authorized up to $1 million for
the purposes of carrying out Section 307, except for the "diagnostic
168
-------
and remedial efforts" described in Section 307(a)(6), for which it
authorized up to $500,000.-
Sections 308. 309. 310. and 311
Section 308 gives EPA the discretion to enter into cooperative
agreements or provide grants to establish regional radon training
centers at colleges, universities, institutions of higher learning or a
consortia of such institutions. EPA is directed to make grants to at
least three applicants for training centers. (House Report 1047 at 19,
reprinted in 1988 U.S. Code Gong. & Admin. News 3624) The
regional centers are directed to develop information and provide
i •
training. RPCA sets forth certain criteria for the establishment of such
regional training centers. One million dollars has been authorized for
each of the fiscal years 1980, 1990, and 1991 to cany out this
program.
Section 309 requires that a study of radon in federal buildings be
conducted. The study shall include examination of radon
contamination in nonpublic water supplied to the buildings. EPA must
identify and compile a list of areas where federal buildings are located
that have a high probability of radon contamination. On or before
October 1, 1990, the Agency must submit to Congress a report
describing the results of the study.
Section 310 authorizes EPA to issue regulations to carry out the
provisions of the Act, and Section 311 states that amounts authorized
to be appropriated in RPCA are in addition to amounts authorized to be
appropriated under other laws for radon-related activities.
169
-------
The Superfund Amendments
The 1986 Superfund Amendments and Reauthorization Act '
(SARA) contains two separate sections directly addressing the subject
of radon gas - Section 118 and Title IV - and a limitation on the
federal response to radon contamination in Section 104(a)(3).
In Section 118(k), SARA calls for EPA to prepare a national
assessment of radon to identify the locations where the gas is found, to
determine the radon levels and health risks at these locations and
methods of reducing those risks, and to provide guidance and
information to the public. (42 U.S.C. §9618(k)) This national
assessment was to be submitted to Congress by October 1987. EPA
submitted a survey of seven states in September 1988 in partial
fulfillment of section 118(k). The results for tests in more states were
released in October of 1989.
Section 118(k) also requires EPA to conduct a radon mitigation
demonstration program to test methods and technologies of reducing
or eliminating the radon threat. (42 U.S.C. §9618(k)(l)) SARA
specifies that the demonstration program should be conducted in the
Reading Prong area of Pennsylvania and New Jersey and other sites
EPA considers appropriate. Annual reports on the status of the
demonstration program are to be submitted to Congress on February 1
of each year, beginning in 1987.25 (42 u.s.C. §9618(k)(2)(B))
als° added a "sense of Congress" in Section 118(m) that
n rec*uil^d to use fully demonstrated methods when carrying
bausfoPf ?3SdCnn Th at * ^^ Usted On *« National Priorities Ust
because of radon. This provision seeks to encourage innovative or
170
-------
The section addressing radon gas is found in SARA Title IV, a
free standing act entitled'The Radon Gas and Indoor Air Quality
Research Act of 1986." Title IV contains findings by Congress
regarding the serious health risk posed by radon gas and the need for
more coordination among research programs and a better information
base. Title IV also establishes a research program to gather
information on indoor air quality, coordinate federal, state, local, and
private research, and assess appropriate federal action to mitigate the
risks associated with indoor air pollution. (§403(a))
Under Title IV, Congress has required EPA to set up a research
program to (a) identify, characterize, and monitor sources and levels
of indoor air pollution, including radon, and including measurement of
various pollutant concentrations, high-risk building types, and
instruments for indoor air quality data collection; (b) study the effects
of indoor air pollution and radon on human health; (c) research
control technologies and other mitigation measures; (d) demonstrate
methods for reducing and eliminating indoor air pollution and radon;
(e) research methods of assessing the potential for radon
contamination of new construction and design measures; and (f)
disseminate information to assure the public availablilty of this
research program. (§403(b)) Congress also set up an advisory
committee composed of representatives from federal agencies, state
governments, the scientific community, industry, and public interest
organizations. (§403 (c))
alternative methods, particularly those involving the off-site transport
and disposition of radon-contaminated material.
171
-------
Title IV also specifically limits EPA's authority to conduct
research, development, and related reporting, information
dissemination, and coordination activities. ,t does not, however, Umlt
the authority of EPA or ^ other ag£ncy
autho^ of law. (§404) EPA submjtte
Congress for these research activities, as required by Me TV. in 1987.
A status report was planned for fall of 1989.
SARA Section 104WB) also adds limitations on EPA's Section
104 response authority. The limitations regarding radon state that
EPA "shall not provide for a removal or remedial action under [104] In
response to a release or threat of a release - of a naturaUy occurring
substance te its unaltered form, or altered ^ ^^ ^^
occurring processes or phenomena, from a location where it Is
naturally found."26 (§i04(a)(3)(A))
EPA's subsequent interpretation of this provision, a clear reading
of the statute, and the legislative history support a finding that the
§104(a,(3HA, limitation on EPA's response authority includes a threat
in the
of hazardous
's example regadinffr Hi o, °ngr lst Sess" at 16^ ™e
2
in
172
-------
from radon contamination. That limitation may be overcome,
however, if EPA finds that the contamination constitutes a "public
health or environmental emergency" and there is no other person
with the authority and capability to respond to the emergency.
-^
\ ' . •
(§104(a)(4)) Thus, under certain circumstances radon contamination
might be addressed under EPA's §104 authority. Whether radon
contamination could be considered a "hazardous substance" that EPA is
authorized to clean up absent an emergency with respect to §106, or
that could constitute the basis for a cost recovery action under §107,
are complex legal issues beyond the scope of this report.
173
-------
Appendix IV
Constitutional Issues Relevant to Regulatory
Strategies for Radon
This appendix addresses the potential constitutional
impediments to adopting various regulatory strategies at the federal
level for controlling naturally occurring radon pollution.
The Commerce Clause
The federal government has inherently limited authority and can
only act on the basis of specific enumerated powers. As with most
federal environmental statutes, the most probable basis for regulation
of radon pollution would be the commerce clause. Even though indoor
radon contamination is distinct in several ways from other
environmental problems regulated under commerce clause theory,
judicial interpretation has so expanded the scope of powers available
to Congress that at least one commentator has noted that "no
conceivable measure reasonably intended to protect the environment
is beyond the reach of congressional authority."27
Article I, Section 8 of the Constitution provides that Congress
has the power "to regulate Commerce . . . among the several states."
n 27PhiUp Soper, 'The Constitutional Framework of Environmental
/1 in Federal Environmental Law 22 (Environmental Law Institute
ed.) (1974). -
174
-------
(Emphasis added.) in the seminal case Gibbons v. Ogden, 22 U.S. (9
Wheat.) 1 (1824), the Supreme Court held 'that Congress may not
regulate activities that are completely confined to a particular state,
that do not affect other states, and with which it is not necessary to
interfere for the purpose of executing some of the general powers of
the government. Subsequent decisions have, however, significantly
expanded the permissible realm of activities that Congress can
regulate, to the point that activities that are essentially local in nature
have been found to have an impact on interstate commerce, however
theoretical, sufficient to support congressional regulation. For
example, in Wickard v. Filburn, 317 U.S. Ill (1942), the Court upheld
federal commodity regulations as applied to a local farmer who
produced wheat exclusively for use on his own farm on the basis that
the cumulative price effect of such home production on interstate
commerce could be significant. The Court applied a similarly
expansive analysis of "substantial economic effect on interstate
commerce" in upholding the Civil Rights Act of 1964. In Katzenbach
y. McClung, 379 U.S. 294 (1964), the Court reasoned that race
discrimination by public facilities, such as restaurants, resulted in the
sales of fewer goods, obstruction of interstate travel, and depressed
general business conditions.
The distinctions between radon regulation and these other
exercises of commerce clause powers are,primarily (a) that radon
contamination involves real property and not the traditional goods of
commerce,28 (b) radon contamination is uniquely confined to the
28As suggested above, however, the courts have quite generously
interpreted the "article of commerce" requirement (United States v.
175
-------
indoors of buildings and residences, i.e., it is not "ambient,"29 and (c)
that Congress has not traditionally "occupied the field" of indoor air
pollution, real estate transactions, or building construction. On the
other hand, federal regulation in other areas of longstanding state or
local concern (e.g., fish and wildlife management) has steadily
increased and been approved by the courts. And, the recent
enactments of federal legislation (see Appendix III) encouraging
research and providing assistance to states, though expressly non-
regulatory, may provide a sound basis for later demonstrating the
significance of the federal interest.
The most plausible theory supporting a commerce clause basis
for federal regulation of radon contamination is the "effect on
Sullivan. 332 U.S. 689, 698 (1948)) (commerce clause powers extend
to articles that have completed an interstate shipment and are being
held for future sales in purely local or intrastate commerce).
29In contrast, the commerce clause theory behind federal
regulation of ambient air pollution is that ambient air connot be
confined to one state's borders. As one court noted, "emitted particles
themselves may be seen as themselves constituting articles moving in
commerce and hence directly subject to regulation." (United States v
gishop Processing Co.. 287 F. Supp. 624, 629 (D. Md. 1968), affd, 423
F.2d 469 (4th Cir. 1970), cert, denied. 398 U.S. 904 (1970)) "Bishop
also demonstrates that an "article of commerce" need not have~a
commercial value, and it is irrelevant that movement of the article
across state lines was not intentional. On the other hand, as
mentioned above, the radon problem occurs precisely because of
radons lack of ambience, i.e., confinement to buildings. Thus it
would be difficult to argue that radon contamination itself can form a
suihcient interstate basis for federal regulation under the commerce-
clause. One could argue, however, that even though the particles
themselves are not ambient, the geological formations that create the
problem are by nature interstate (i.e., deposits of phosphate, granite
and shale are not confined neatly to state borders).
176
-------
interstate commerce" argument.30 Under this approach, the courts
would examine the competitive effects of the problem or the lack of '
federal regulation. For example, different radon testing, disclosure or
mitigation standards in different states can impact on industry
location, pricing of real estate, personnel relocation decisions, and
individual consumption patterns. The commerce clause argument to
support federal radon regulation of contracts for sales of property (i.e.,
disclosure, testing, or mitigation) would run as follows: although
concentrations vary regionally, radon contamination is a significant
nationwide problem; leaving regulation to the states could lead to
inconsistent standards among the. states, directly affecting interstate
commerce related to real estate transactions, in building materials,
and in testing and mitigation services, and indirectly affecting
interstate commerce in other areas (e.g., industry relocation); and
there is a significant need for federal coordination because states lack
the necessary financing and technology to address the problem
comprehensively.31 In short,, the commerce clause provides the
30Another interesting, though relatively unused, theory that
could support congressional regulation is Congress' power to approve
interstate compacts. (U.S. Const., art. 1, sec. 10) "Since under the
Constitution congressional approval is required for these agreements,
an indirect handle is provided for imposing minimum federal
environmental standards by conditioning approval of such agreements
on the compatibility of the compact's terms with the federal
standards." (Soper, supra note 27, at 32) Such leverage hinges, of
course, on the states' own incentives to enter into such compacts,
which despite the regional nature of some radon "hot spots," are
probably minimal.
31 For example, the New Jersey Department of Environmental
Protection ran out of money in its efforts to mitigate radon
contamination from uranium mill tailings fill in three communities in
New Jersey (Galen, "Lawyer's grapple with radon issue," Nat'l L.J.. July
.21, 1986, at 10).
177
-------
strongest "hook" for an exercise of congressional regulatory power,
and such regulation would likely be upheld despite the unusually
localized aspects of the radon contamination problem.
The Tenth Amendment
Under the tenth amendment to the Constitution, powers not
delegated to the federal government are reserved to the states. The
police powers reserved to the states include those involving public
safety and public health. (See Berman v. Parkpr 343 U.S. 26, 32
(1954).) Whether the tenth amendment acts as a limitation on federal
power depends on the constitutional underpinning for that federal
exercise. For example, if Congress can legitimately utilize its
commerce power to regulate, then the tenth amendment does not
constrain its actions. (United States v. Bally, 345 F. Supp. 410 (D.
La. 1972)) However, similar to pre-emption theory (discussed below),
when Congress legislates in a field that the states have traditionally
occupied, the historic police powers reserved to the states under the
tenth amendment are not superceded unless that was the clear and
manifest purpose of Congress. (Fouke v. MandPl 386 F. Supp. 1341
(D. Md. 1974))
To succeed, a claim that congressional commerce power
legislation is invalid under the tenth amendment must show that (a)
the challenged statute regulates "states as states," (b) the federal
regulation addresses matters that are indisputably attributes of state
sovereignty, and (c) state compliance with the federal law would
directly impair the state's ability to structure integral operations in
areas of traditional governmental function. (Hodel v. Virginia Surface
178
-------
Mining and Reclamation Ass'n, 452 U.S. 264, 287-88 (1981)) Despite
this seemingly high standard, some federal legislation has been struck
down as unconstitutional under the tenth amendment. For example,
the Federal Resettlement Administration (part of the Emergency
I ' •• , , " ' ~
Relief Appropriation Act of 1935), which attempted to regulate and
control-housing and shift destitute and low-income populations, was
found unconstitutional as invading the reserved powers of the states.
(Township of Franklin, Somerset County, N.J. v. Tugwell, 85 F.2d 208
(D.c! Cir. 1936)) On the other hand, the Federal Emergency Price
Control Act "of 1942, which authorized a federal administrator to fix
maximum rentals in a defense area was upheld even though rentals
were considered a matter for local legislation and one not delegated to
the United States. (Ritchie v. Johnson, 144 P.2d 925, 158 Kan. 103
(1944))
" A more detailed analysis of the potential tenth amendment
challenges to radon regulation depends on the specific regulatory
strategy proposed. In general, however, such regulation would
probably survive a tenth amendment challenge so long as Congress
convincingly set forth its longstanding interest in the problem.
179
-------
The Fifth . Amendment - Takings Clause
Intrusive regulation at both the state and federal level could be
subject to challenge on the basis of the takings clause of the fifth
amendment^ which provides that private property shall not "be
taken for public use without just compensation." The takings
argument most applicable to the radon regulation strategies described
above would be the "diminution in value" theory. Under this theory, a
seller could argue that the requirements of disclosure, testing, and/or
mitigation diminish the value of the property to the extent that a
taking by the government has occurred. How much diminution must
occur before a court will find a taking and order compensation is a
question that cannot be answered satisfactorily given the murky state
of takings law. particularly without an examination of the specific
32pther potential fifth amendment issues might arise under
different regulatory strategies. For example, if EPA wS to reauire
testing of privately owned buildings (with attend Enforcement
powers providing for physical access to property) Tthere might bl a
fifth amendment search and seizure concern But, such governmental
nn< *** circui»stances has been upheld in the^e of
^te Waste Management Bna.5 v
hv, w , ' APP- 1984) (electrical sensitivity testing conducted
by the Waste Management Board in connection with selection of
C"llh^d?f W?St,e diSp°SaI Sites did not^S^lfa
fuSstS^nJnV?1yfed ^temporary, minimal intrusion and
suostantially interfere with property rights or cause
^ ta ^^ ^^ SimTl^mfo^rmSion and
requirements may involve fifth amendment self-
£ h°wever' disdosure of information
c ™ thte tyPe of takin^ challenge
«*
to
180
-------
governmental action at issue and the alleged loss in value.33 And,
whether such federal regulation would be considered for "public use"
is questionable.34
In the seminal case of Pennsylvania Coal y. Mahon, 260 U.S. 393
(1922), the Supreme Court held that a drastic reduction in the • •
economic value of property triggers the need for compensation, but
declined to define the point at which a regulation "goes too far." The
lower courts have subsequently developed the general rules that: a
taking occurs if the regulation deprives the property of all potential
value or use;35 a taking has not occurred if the owner merely is not
33"[N]o single formula [for federal takings principles] is either
possible or desirable. All such formulations, for example, may prove to
be only extrapolations from, what is basically an ethical judgment about
the fairness of refusing to distribute across a broad base the costs
entailed in implementing certain public programs perceived to have
positive net benefits. As such, the takings clause, like the Due Process
Clause and other constitutional expressions of broad, social policy may
be expected to reflect changes in society in a way that allows doctrinal
development to keep pace with shifting priorities in societal values."
(Soper, supra note 27, at 61)
34Even though the general public benefits incidentally from
mitigation of radon in private buildings, if the legislation placed the
burden on sellers to mitigate, it arguably could be challenged as not
being for public use because it creates only a private benefit to the new
owner. But see Monsanto, 467 U.S. at 1014 (broadly interpreting the
public use" requirement); and National Board of Young Men's
Christian Ass'ns. v. United States. 395 U.S. 85, 92 (1969) (any
protection of private property also serves a broader public purpose).
35See, e.g., Bartlett v. Zoning Cnmm's Town of Old Lvme. 161
Conn. 24, 282 A.2d 907 (1971) (invalidating tidal wetlands
restrictions that deprive owner complete use of property); Maine v
Johnson, 265 A.2d 711 (Me. 1970) (Maine Wetlands Act deprived
plaintiff of aU commercial value and constituted a taking); Morris
County Land Improvement Co. v. Parsippany-TVny Wiiic, 40 N.J. 539,
193 A.2d 232 (1963) (invalidating meadow development zone as
applied to certain wetlands). A less than confiscatory impact may be
severe enough to constitute a taking if the governmental action is not
181
-------
realizing a speculative investment potential or is not allowed the most
profitable use of the land. As the Supreme Court recently commented,
the factors a court will consider include "the character of the
governmental action, its economic impact, and its interference with
reasonable investment backed expectations." (Pruneyard Shopping
Center v. Robbins, 447 U.S. 74, 83 (1980)) Thus, if a reasonable,
economically profitable use of the land remains, a takings claim will
probably not succeed.36 ^ one 8tate court commented, "there is no
constitutional impediment to making a landowner personally
responsible for the reasonable cost of abating hazardous conditions'.
existing on the property." (City of Patterson v. Fargo Realty, Inc., 174
N.J. Super. 178, 415 A.2d 1210 (N.J. Super. 1980))
In addition, given that the diminution in value in the typical
mitigation case would be small (relative to the value of the property),37
a takings claim would undoubtedly fail. However, in extreme cases, for
example, where mitigation costs approach the market value of the
property, a seller might theoretically have a viable argument. The
takings argument might also be of greater concern should EPA's safety
v
v.
n ! te governmental interest. See
Coastal Commisinr.gr- 107 S. Ct. 3141 (1987).
Forest
TT q i 15
substantial reduction of attractiveness of property to
to
™s hlgher
182
-------
standard be set to very low levels (resulting in increased mitigation
costs).
Pre-emption Issues
Under the Supremacy Clause, U.S. Const., art. VI, cl. 2, state laws
must yield to federal regulation in areas where the governments have
overlapping authority. Given that the tenth amendment reserves
certain powers to the states, including the power to protect public
health, and the extensive nature of current state regulation, an analysis
of the desirability of further federal regulation should be cognizant of
the political and legal issues raised by the potential for federal pre^
emption of state programs. To avoid uncertainty. Congress should
always make explicit in any new legislation whether it intends that
state laws be preempted. Otherwise, courts and litigants will later be
left the difficult and uncertain task of discerning congressional
intent.38 Principles of pre-emption may then invalidate state
regulation in part or altogether, or may limit states to enforcing the
safety standard set by the federal government.39
38See Pennsylvania v. Nelson. 350 U.S. 497 (1956) (analyzing
factors such as "pervasiveness" of federal scheme, "dominance" of
federal interest, and whether the area has traditionally been one of
state and local, rather than federal, control).
39See Northern State Power Co. v. Minnesota. 447 F.2d 1143
(8th Cir. 1971), affd. 405 U.S. 1035 (1972) (state attempt to impose
radiation emission standards on a nuclear power plant stricter than
those required under federal law held pre-empted by Atomic Energy
Act).
183
-------
With regard to radon contamination, in the absence of expressed
congressional intention, the strong state interest in health and
welfare, and the historical lack of interest in indoor pollution (and in
private property transfers) by the federal government, would strongly
support a state's argument that its laws establishing higher standards
or additional requirements should not be preempted.4<> The outcome
of such a challenge would, of course, once again turn on the specifics
of the conflicting regulatory schemes at issue.
„ * JlOS°Per sug£ests that .state regulation in the environmental
context enjoys a presumption of validity in the context of a pre-
^£™^fng%£^n°te ^'• at "-10°): "Attempts to cope with
environmental problems have led to increasing recognition of the
complexity of the solutions and of the often inadequate knowledge
base for providing such solutions. These facts strongly support an
approach to interpretation of the Commerce Clause that preserves
maximum flexibility for state experimentation in the absence of
explicit contrary directions from Congress."
184
------- |