EPA/400-11/76-002
WOULD DECLARING FEDERAL SMOG
DAYS REDUCE AUTOMOBILE TRAVEL
IN THE WASHINGTON. D.C.AREA?
THE RESULTS OF A SURVEY OF TRAVEL BY EPA
EMPLOYEES ON APRIL 19,1976
OFFICE OF POLICY ANALYSIS
U.S.ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C.
NOVEMBER 1976
-------
EPA/400-11/76-002
WOULD DECLARING FEDERAL SMOG
DAYS REDUCE AUTOMOBILE TRAVEL
THE WASHINGTON, D.C. AREA?
THE RESULTS OF A SURVEY OF TRAVEL BY EPA
EMPLOYEES ON APRIL 19,1976
by
JOEL HOROWITZ
OFFICE OF POLICY ANALYSIS
U.S.ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C.
NOVEMBER 1976
-------
TABLE OF CONTENTS
Page
Abstract iii
Would Declaring Federal Smog Days Reduce Automobile
Travel in the Washington, D. C. Area? 1
Table 1 - Estimates of Average Automobile
Driver Trips and VMT Per Dismissed
Employee 5
Table 2 - Estimated Reductions in Daily Trips
and VMT on a Smog Day 7
Acknowledgements 9
Appendix - The Survey Letter and Questionnaire 10
-------
iii
ABSTRACT
It is sometimes suggested that the intensity of smog episodes in
the Washington, D. C. area could be reduced by asking Federal employees
in the area not to go to work on days forecast to be smoggy, thereby
reducing the volumes of automobile travel and emissions during smog
episodes. Estimates of the reduction in automobile travel that might be
achieved by declaring a Federal "smog day" have been difficult to develop
owing to a lack of travel data for situations analogous to smog days. A
somewhat analogous situation occurred at EPA on April 19, 1976. Employees
of the West Tower were dismissed from work at 9:30 AM because of a breakdown
of the building's air conditioning. A survey of travel by employees dismissed
on April 19 was conducted, and the results were used to estimate the trip
and VMT reductions that might occur on a smog day. The survey results
suggest that declaring a Federal smog day might reduce Federal employee
automobile trips by 18 to 42 percent and Federal employee VMT by 37 to 50
percent. It is probably safest to interpret these results as being indicative
of travel reductions on an isolated smog day or on the first of a series of
smog days.
-------
WOULD DECLARING FEDERAL SMOG DAYS REDUCE
AUTOMOBILE TRAVEL IN THE WASHINGTON, D. C. AREA?
by
Joel Horowitz
It is sometimes suggested that the intensity of oxidant episodes
in the Washington, D. C. area could be reduced by asking Federal employees
in the area not to go to work on days forecast to be smoggy. Proponents
of this approach claim that the decrease in hydrocarbon emissions
associated with the elimination home-to-work automobile travel by Federal
employees would be beneficial to air quality. However, opponents have
argued that employees who do not have to go to work might take more shopping,
social and recreational trips than they normally do on working days, thereby
negating the beneficial effects of decreased home-to-work travel.
The extent to which declaring a Federal "smog day" might decrease
automobile travel could be determined empirically if appropriate travel
data were available for previous smog days or analogous events. However,
there exist no travel data sets for smog days, and reasonable analogies
are difficult to find. A smog day would be an unanticipated holiday
from work during hot weather. Normal non-working days, such as weekends,
legal holidays, and vacation days, are not analogous because they are
anticipated. Snow days are not anticipated, but the weather on snow days
makes driving difficult or impossible.
An event that is somewhat analogous to the declaration of a smog
day occurred at EPA on April 19, 1976. Employees in the West Tower were
-------
-2-
dismissed from work at 9:30 AM due to a breakdown of the building's air
conditioning. The dismissal was analogous to a smog day in that it was
unanticipated, the weather was hot, and it occurred early enough in the
day to allow employees time to engage in non-work travel. Thus, it is
possible that employees' non-work travel activity on April 19 was similar
to the activity that would take place on a smog day. If so, employee's
non-work travel on April 19 approximates their total travel on a smog day,
and the reduction in employee travel that declaring a smog day would
produce can be estimated.
The analogy between the April 19 dismissal and a smog day is not
perfect. Employees did have to go to work on April 19, and nobody was
advised to avoid driving due to atmospheric or other conditions. Nonethe-
less, the dismissal is more closely analogous to a smog day than are other
non-work days for which travel data are available. Using the dismissal as
an analogy, estimates have been developed of the reductions in employee
automobile driver trips and vehicle miles traveled (VMT) that might be
achieved by declaring a smog day. This paper describes the estimation
methodology and results.
Methodology
Data on travel by dismissed employees were obtained by distributing
a self-administered, written questionnaire to each of 238 randomly selected
West Tower employees. The questionnaire is shown in the appendix. Employees
were asked to list the origin location, destination location, and travel
mode of each trip they took prior to 6:00 PM on April 19. The 6:00 PM cutoff
was established on the hypotheses that employees would be more likely to
-------
-3-
answer questions about daytime travel than evening travel and that any
unusual travel resulting from the dismissal would have occurred during
normal working hours. The questionnaire was not distributed until May 6,
roughly three weeks after the dismissal occurred. This late distribution
may have biased the survey results owing to the possibility that some
employees may have forgotten what trips they took on April 19. If it
is assumed that employees who took many trips were more likely to forget
than employees who took few trips, then the survey results will tend to
overestimate the reduction in travel that might be achieved by declaring
a smog day.
The survey results were used to compute the average numbers of auto-
mobile driver work trips and 9:30 AM - 6:00 PM non-work trips per employee,
and the average work-trip VMT and 9:30 AM - 6:00 PM non-work trip VMT
per employee. VMT were computed by assigning the origin and destination
of each trip to the appropriate Washington area transportation planning zones
and-summing the interzonal distances over trips. The interzonal distances
were obtained from the Washington Council of Governments and are based on
data developed in the 1968 Washington area transportation survey.
Results
Of the 238 questionnaires distributed, 114 were returned, making the
response rate 48 percent. The 114 returned questionnaires included 12
invalid responses (i.e., the questionnaire was not completed properly), 12
responses from employees who did not leave work on April 19, and 5 responses
from employees who were sick or on annual leave that day. Thus, the survey
produced a total of 85 useable responses.
-------
-4-
The estimates of average automobile trips and VMT per dismissed
employee derived from the survey are shown in Table 1. On the average,
employees took 1.15 automobile work trips and 0.89 automobile non-work
trips before 6:00 PM. Average VMT were 12.29 miles per employee for work
trips and 6.17 miles per employee for non-work trips before 6:00 PM.
The reductions in pre-6:00 PM trips and VMT that would occur on a
smog day were estimated from the survey results using the following
equation:
(1) R = (HW + X) - NW
HW + X
where
R = Fractional reduction in trips (VMT)
HW = Home-to-work automobile trips (VMT) per employee on a normal day
X = Pre-6:00 PM non-work automobile trips (VMT) per employee on a normal
day
NW = Pre-6:00 PM non-work automobile trips (VMT) per employee on a smog day.
It is assumed that home-to-work travel on April 19 was normal. Thus, HW has
the values observed in the survey. Using the presumed analogy between the
April 19 dismissal and a smog day, NW is set equal to non-work trips (VMT)
observed in the survey.
X was not measured in the survey, and recent estimates of X are not
available from other sources. However, a range of likely values of X is
easy to develop. The lowest possible value of X is zero. If X = 0, then
all non-work automobile trips taken on April 19 by dismissed employees
occurred because of the dismissal. A likely upper bound for X is NW. If
X=NW, then all non-work trips by dismissed employees would have occurred
-------
-5-
TABLE 1 - ESTIMATES OF AVERAGE AUTOMOBILE DRIVER
TRIPS AND VMT PER DISMISSED EMPLOYEE
Sampling Error8
0.20
0.29
0.31
3.20
2.55
3.92
Work Trips
Non Work Trips b
Total Trips
Work VMT
Non Work VMT b
Total VMT
Mean
1.15
0.89
2.04
12.29
6.17
18.46
a. Computed at 95% confidence level
b. Prior to 6:00 PM
-------
-6-
even if there had been no dismissal. Another possible value of X can be
obtained from the 1968 Washington area transportation survey. In that
survey, there were roughly 0.40 automobile trips and 2.5 VMT per government
employee prior to 6:00 PM. If X has these values, then 40 percent of
dismissed employees' non-work travel on April 19 would have occurred without
the dismissal.
The reductions in pre-6:00 PM trips and VMT that might occur on a smog
day were estimated using each of the three foregoing values of X. The pre-
6:00 PM reductions were converted into reductions in daily trips (VMT)
by assuming that the April 19 dismissal did not affect post - 6:00 PM
travel. Such travel accounts for about 25 percent of daily travel. The
results are shown in Table 2. Depending on the value of X, there is an 18 to
42 percent reduction in trips and a 37 to 50 percent reduction in VMT.
Discussion and Conclusions
The survey results suggest that declaring a Federal smog day might
achieve an 18 to 42 percent reduction in automobile driver trips and a 37
to 50 percent reduction in automobile VMT by Federal employees. This
conclusion is based on the assumption that the survey results provide a
reasonably accurate description of travel behavior under circumstances
analogous to a smog day. Some important qualifications to the conclusion
will now be discussed.
As explained earlier, the analogy between the April 19 dismissal and a
smog day is not perfect. It is not clear a priori how this might bias the
foregoing conclusion. Not having to go to work on a smog day might cause
employees to take more or longer non-work trips than they did on April 19.
-------
-7-
TABLE 2 - ESTIMATED REDUCTIONS IN DAILY TRIPS
AND VMT ON A SMOG DAY
Assumed Level
of Normal Non-Work Trip VMT
Travel (X)a Reduction(%) Reduction (%}
X = 0 18 37
X = NW 42 50
X = 0.40 trips 32 44
(2.5 VMT) per
employee
a. See page 4 of text for explanation.
-------
-8-
However, if exhortations to avoid driving on smoggy days are effective,
there might be less non-work travel on a smog day than on April 19.
The late distribution of the survey of April 19 travel is another
potential source of bias. As discussed earlier, this may cause the survey's
results to overestimate the reductions in travel that would occur on a smog
day.
A third potential source of bias is the duration of oxidant episodes
in Washington. These episodes typically last for several days, whereas
the April 19 dismissal affected only one day. If employees are asked to
stay away from work during a series of smog days, their driving patterns
may change as the series progresses. For example, they may get restless,
with the result that they travel more toward the end of the oxidant episode
than toward the beginning. Or they may anticipate the duration of the
episode and leave town. It is probably safest to interpret the April 19
results as being indicative of travel on an isolated smog day or on the
first of a series of smog days.
-------
ACKNOWLEDGEMENTS
The author thanks Larry R. DeBoever and William G. Laxton for
their assistance in developing the survey questionnaire, and David
Syskowski and Robert Emslie for helping to compile the survey results.
-------
-10-
APPENDIX
The Survey Letter and Questionnaire
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
OFFICE OF
AIR AND WASTE MANAGEMENT
May 6, 1976
Dear EPA Employee:
On Monday April 19, 1976 employees of the West Tower were released
from work at 9:30 A.M. when the air conditioning in the building broke
down. The Office of Air and Waste Management would like to acquire
information on the automobile trips released employees took that day. The
information will be used by OAWM to help determine the extent of emissions
reductions that might be achieved if employees were asked not to come to
work during smog episodes. The information obtained in the survey will be
used for statistical purposes only. No information about individuals will
be released. Please do not put your name on the survey form.
The attached questionnaire requests you to supply the addresses of
the locations where the trips you took between 9:30 A.M. and 6:00 P.M.
on April 19 started and ended. If you do not know an address or would
prefer not to provide it, please fill in the street intersection or major
landmark nearest the location where your trip started or ended (e.g.,
4th & M. Street, S. W., Tyson's Corner, Washington Monument). If you know
neither the address nor the intersection nor a landmark, write "don't
know." If you did not leave work early, please write "did not leave work"
on the form, and return it.
Your participation in this survey is entirely voluntary. If you choose
to participate, please return the completed survey form in the attached
envelope to Margaret Lee (AW-444) within 48 hours.
Thank you for your cooperation.
Sincerel'yj yours.
Roger Strelow
Assistant Administrator
for Air and Waste Management
Enclosures
-------
OAWM SURVEY OF AUTOMOBILE TRAVEL ON APRIL 19. 1976
Your participation 1n this survey 1s voluntary. Information obtained from
the survey will be used for statistical purposes only. No Information about
Individuals will be released. Do not put your name on the survey form. Please
return the completed form 1n the attached envelope to Margaret Lee (AW-444)
within 48 hours.
What is your home address or the street Intersection nearest your home, and the
city or county and state where you live?
Did you drive a car to work on Monday, April 19, the day the air conditioning
broke down?
Yes No
Please supply the requested information for each trip you took between 9:30 A.M.
and 6:00 P.M. on Monday April 19, the day the air conditioning broke down.
Under "Place Trip Started" and "Place Trip Ended" enter the appropriate street
address, nearby intersection, or major landmark and the city or county and
state.
Place Trip Place Trip Did you Drive a
Trip No. Started Ended Car on this Trip?
1 401 M. Street, S. W. Yes
Washington, D. C. No
Yes_
No
Yes_
No
-------
-2-
Questionnalre (cont'd)
Trip No.
4
Place Trip
Started
Place Trip
Ended
Did You Drive a
Car on this Trip?
Yes
No
Yes_
No _
Yes_
No _
Yes
No
8
Yes_
No
Yes_
No
10
Yes
No
------- |