Grantee Research Project Results
Final Report: Behavioral Reactions to Ozone Alerts: What Do They Tell Us About Willingness-to-Pay for Children's Health?
EPA Grant Number: R829544Title: Behavioral Reactions to Ozone Alerts: What Do They Tell Us About Willingness-to-Pay for Children's Health?
Investigators: Mansfield, Carol , Van Houtven, George L. , Johnson, F. Reed , Crawford-Brown, Douglas
Institution: RTI International , University of North Carolina at Chapel Hill
EPA Project Officer: Hahn, Intaek
Project Period: February 18, 2002 through February 17, 2005
Project Amount: $310,000
RFA: Valuation of Environmental Impacts on Children's Health (2001) RFA Text | Recipients Lists
Research Category: Environmental Justice , Children's Health , Human Health
Objective:
The Air Quality Index (AQI) was developed by the U.S. Environmental Protection Agency (EPA) in part to convey information to the public about the level of ozone and other pollutants that might affect their health. Currently, local governments use the AQI to issue air quality alerts during the summer on days when pollution is forecasted to be high. Children and those with respiratory diseases are considered to be especially susceptible to ozone-related health concerns. However, very little is known about public awareness of the AQI and responses to warnings about high ozone, particularly changes in behavior. This project is designed to investigate the averting behavioral responses to high ozone concentrations by parents of young children, comparing children who have asthma and those who do not. Information from these parents about the daily activities of their children under high and low ozone conditions will be used to estimate lost activity time spent outdoors for the children using either standard regression techniques and/or a random utility model (RUM). We will then derive monetary values for the lost outdoor playtime using a stated preference (SP) conjoint survey. The data will also be used to estimate averted health endpoints using an exposure-response model. The risk assessment will help evaluate whether the defensive activities taken by parents are effective in actually reducing the asthma symptoms experienced by children.
Summary/Accomplishments (Outputs/Outcomes):
Background
Even at relatively low levels, ground-level ozone is known to cause a number of acute respiratory health effects. According to EPA, active children and asthmatics are among the most highly susceptible subpopulations. Physically active children are vulnerable, because they tend to spend more time outdoors and to breathe faster and deeper. Asthma sufferers—particularly children with asthma—are at high risk, because their lungs are generally more sensitive to environmental conditions, and ozone is known to trigger asthma-related symptoms.
To protect children from ozone-related health effects, EPA and organizations such as the American Lung Association recommend that parents have their children spend more time indoors and engage in less strenuous activities on relatively high-ozone days. These same recommendations apply to active adults and adults with asthma or other conditions that make them particularly susceptible to ozone.
In part to assist parents and other individuals concerned about ozone conditions, EPA developed the AQI. This index combines information about ozone levels (and other pollutants) to produce five categories of air quality, ranging from good to very unhealthy. To more easily and effectively communicate these conditions to the general public, the five categories are also color coded, ranging from green to purple. Using some versions of this air quality categorization scheme, forecasted and actual conditions are typically reported to the public on a daily basis (particularly during the summer) through local media outlets.
The combination of widely publicized air quality conditions and explicit recommendations for avoiding adverse health effects suggests that averting behaviors, such as spending more time indoors and less time in strenuous activities, should be relatively common on days with poor air quality. Thus far, however, evidence of these types of averting behaviors is somewhat limited, particularly regarding parental behaviors to protect their children’s health.
Evidence of averting behaviors is not only useful for understanding how individuals protect themselves from potentially harmful exposures and reduce their health risks, but it also reveals potentially important information regarding individuals’ preferences and willingness-to-pay (WTP) for improved environmental conditions. Even though the averting behavior method is often cited as one of the primary revealed preference approaches for nonmarket valuation, the number of empirical applications for valuing air quality improvements is also relatively limited. To our knowledge, averting behavior approaches have not been previously used to assess parents’ values for protecting their children from ozone-related health effects.
Using a series of web-based surveys with parents of young children in relatively high-ozone metropolitan areas, this study is intended to address these research gaps. Some of the key research questions to be addressed by the study are:
- To what extent are parents aware of ozone alert systems?
- To what extent do parents change their and their children’s behavior in response to ozone alert systems?
- How does the presence of asthmatic children in the household affect awareness and behaviors?
- What costs (direct and indirect) are incurred by parents and children as a result of behaviors to avert ozone exposures?
- To what extent are children’s risks from exposure to high ozone levels offset by defensive/averting behaviors?
- How much do parents value reductions in potentially harmful ozone exposures to their children?
Results to Date
Please note that work on the data collected through this grant continues, and we are preparing additional publications. Thus, the following results are preliminary.
Sample and Survey Design. During the summer of 2002, a series of web-based activity diaries was administered to a panel of parents who are members of the Harris Interactive online panel (a self-selected panel) living in the 35 U.S. cities with the highest ozone levels. Their children were ages 2–12; half had asthma and half were free from asthma. We selected households where one parent was home with the child during the day, and the children were not away from home for more than 1 month during the summer. Diaries were administered over the internet on sunny weekdays representing a mix of high and low ozone days within a target temperature range. Panel members were not told the purpose of the study or alerted to ozone conditions. By using a web-based survey and e-mail contact, we were able to target specific days and to confine responses to those collected within 48 hours to improve data quality. Parents were asked to complete the diary from the time the children woke up until they went to bed or 8 pm. A panel of 768 parents completed a total of 2,876 activity diaries. A final debriefing survey (completed by 469 parents) collected information on knowledge of ozone alerts and self-reported changes to their children’s schedules on high-ozone days. Table 1 presents some information about the sample and the diaries.
Table 1. Full Diary Sample Characteristics
|
Mean |
Std Dev |
Min |
Max |
Household Level |
|
|
|
|
Average summer temperature last 2 years |
85.20 |
7.01 |
74.17 |
102.22 |
Average # red days last 2 years |
7.90 |
12.43 |
0 |
46 |
Asthmatic child |
0.46 |
0.50 |
0 |
1 |
Child age |
6.71 |
3.37 |
2 |
12 |
Child male |
0.58 |
0.49 |
0 |
1 |
Parent white |
0.88 |
0.32 |
0 |
1 |
Parent education in years |
14.77 |
2.32 |
7 |
20 |
HH Income |
6.61 |
3.80 |
0.5 |
25 |
Number of obs |
768 |
|
|
|
Diary Level |
|
|
|
|
Spent time totally outdoors |
0.48 |
0.50 |
0 |
1 |
Number of totally outdoors hours if spent time totally outdoors |
1.52 |
1.49 |
0.25 |
11.25 |
% diaries on Code Red days |
0.27 |
0.44 |
0 |
1 |
Number of diaries |
2,876 |
|
|
|
Number of diaries with time totally outdoors |
1,380 |
|
|
|
Results. Starting with the final survey that included questions about awareness and use of AQI-based warnings, we found high levels of awareness of the ozone alert system across our panel. About 91% of the 535 households who completed the final survey in the sequence have heard of at least one type of air pollution warning (i.e., Code Purple/Red/Orange Days, Air Pollution Watch, Ozone Action Days, and Smog Alert). Seventy percent of the households can recall hearing about the level of air pollution or checking on the level of air pollution in Summer 2002. However, 20% of the parents who completed the final debriefing survey indicated that they never checked the daily ozone forecast, while almost 50% indicated that they did not check the forecast, but they paid attention if they saw that high ozone levels were predicted. About two-thirds of the sample who completed the final debriefing survey said that they kept their children indoors more on high ozone days.
Comparing parents of asthmatic and non-asthmatic children:
- Parents of asthmatic children were more likely to say they checked the AQI frequently.
- More parents of children with asthma said they check the AQI forecast daily (difference significant with a p < 0.01).
- Fewer parents of children with asthma said they rarely or never paid attention to the AQI (difference is significant with a p < 0.01).
- More parents of asthmatic children said they reduce their child’s outdoor hours on high pollution days than parents of children without asthma (difference is significant with a p-value < 0.01).
Looking at the activity diary data, the preliminary analysis shows a complicated pattern of results. For each activity their child did, parents were asked if the child was totally indoors, mostly indoors, half indoors/half outdoors, mostly outdoors or totally outdoors. We found that:
- Children spent less time totally outdoors and totally or mostly outdoors in cities that experience more red days on average during the summer, but the effect is small. There is no reduction in hours outdoors on Code Red days.
- Among respondents who took the debriefing survey, asthmatic children spent less time totally outdoors on Code Red days, while the non-asthmatic children did not change time outdoors.
However, we are currently addressing alternative specifications, so these results are preliminary.
The final element of the survey was an SP conjoint survey. In the attached paper (Mansfield, et al., 2006), we estimate parents’ WTP to prevent restrictions on a child’s outdoor time from an SP conjoint survey. As far as we know, this is the first attempt to estimate the shadow price of outdoor time for children. We combine this WTP measure with an estimate of reductions in time spent outdoors on high-ozone days from the activity-diary study to estimate this averting behavior component of WTP for reductions in ozone pollution. In order to elicit the value of lifting restrictions on children’s time outdoors, we needed to design a scenario in which time outdoors was independent of air pollution levels. Figure 1 presents the text for the question.
Assume that at the beginning of the summer, your family doctor tells you that [child’s name] needs to take a medicine during the summer as a preventive measure. In other words, [child’s name] is not sick, but [he/she] needs to take medicine to prevent an illness from developing.… |
[Child’s name] would have to limit the time spent outdoors on the days [he/she] takes [his/her] medicine. Even on cloudy days or when [he/she] is wearing sunscreen, extended exposure to the sun will make the medicine less effective. |
Figure 1. Definition of the Outdoor Time Attribute
The data were estimated using a hierarchal Bayes procedure. As discussed in the paper, the lowest WTP is about $20 to avoid 3 days during which outdoor time would be restricted to 90 minutes and the highest is about $200 to avoid 20 days during which outdoor time would be restricted to 10 minutes. Based on preliminary results from the activity diary data, a subset of the asthmatic children reduced the amount of time spent totally outdoors by 30 minutes on a Code Red day relative to a Code Green, Yellow, or Orange day, and parents are willing to pay $35.18 (90% Confidence Interval [CI] 33.96–36.46) for an additional 30 minutes of time outdoors on a given day. The paper reports regressions that examine covariates significantly correlated with an individual’s WTP. The average number of orange and red days, average temperature, whether the child is a boy, outside play preferences, and household income all influence the size of WTP.
Significance to the Field
The results should make a significant contribution to the field. Very few economic studies have attempted to quantify the degree to which people alter the amount of time spent outdoors on high ozone days. As far as we know, our survey is the most detailed. Furthermore, because very little data exist on how people actually change their outdoor activities in response to pollution, little is known about how well people’s stated changes in behavior match their actual changes in behavior. Furthermore, this study focused on two important subpopulations: young children and young children with asthma. The few studies in the economic literature that have examined the question of whether people limit outdoor time on high pollution days have mostly focused on adults. Finally, the use of the internet to collect activity diary data represents an innovation that has not been widely used in the field of economics. The internet provides significant advantages in the collection of daily time and activity diary information. While there are concerns about the representativeness of the sample, as internet access increases, this mode of survey administration will become more important and viable.
Relationship to the Goals of the Grant
Overall, we met all the original goals of the grant except one. The survey faithfully reflects the original goals of the grant. Data were collected through an internet survey, which was a primary goal of the grant. We analyzed the data to provide evidence of behavioral change on poor air quality days and of awareness of AQI-based warnings. We also used the data from the SP conjoint question to calculate WTP for the absence of restrictions on the time children can spend outdoors during the day. We combined the WTP data and estimates of the reductions in outdoor time on high ozone days to provide an estimate of the costs associated with one type of behavioral adjustment to poor air quality. The only area of the original grant where we have made less progress is the risk-exposure model, primarily because of the loss of Dr. Pekar to the team (discussed below).
Relevance to EPA and Practical Applications
The findings are highly relevant to the type of work that EPA does. Very few studies exist that measure the degree to which individuals engage in defensive behaviors (such as staying indoors) to protect themselves from air pollution. These measures are needed to properly specify dose-response relationships between pollutants and health outcomes. A complete assessment of the benefits of air quality improvements requires information on behavioral changes and the monetary value of these changes. Our survey will ultimately provide data on the changes in the behavior of children and asthmatic children, both sensitive subpopulations, as well as a lower bound on the monetary value their parents place on the reductions in outdoor time. We have shared our results and our data, and we will continue to do so, with staff in the EPA Office of Air Quality Planning and Standards to provide information that could be useful when calculating the benefits of reductions in ozone levels.
EPA’s Office of Air Quality Planning and Standards is funding development of a theoretical health production model that links benefits (WTP) and the concentration-response (risk-response) relationship and an empirical strategy for estimating the model with the Science To Achieve Results (STAR) data (or similar data).
Beyond the goals of our grant, the time and activity diary data provides EPA with more data on which to base exposure assessments and calculate baseline activity levels. The survey data have been provided to EPA for use in EPA’s Consolidated Human Activity Database and exposure analysis.
Measures of public awareness of the AQI and behavioral changes related to the information in the AQI will help measure the effectiveness of the AQI. This information will be useful for assessing the success of the AQI, designing information programs about the AQI, and plans to extend or revise the AQI. We have shared our preliminary results with EPA staff who have an interest in the design, public awareness of, and response to, the AQI.
The survey instrument and data collection process provide a model for EPA researchers interested in replicating the data collection for different populations and different pollutants. We have shared our survey instrument with several EPA staff who are interested in data collection of this type. Through a work assignment with EPA’s Office of Air Quality Planning and Standards, the STAR survey will be used as the model to design a national survey to measure awareness of the AQI and behavioral changes.
Finally, we will continue to share our results with policymakers and the research community. We have presented work based on the survey at professional conferences and at EPA-sponsored conferences including the 2005 EPA Science Forum and the 2006 National Air Quality Conference. These events provided us with the opportunity to interact with Federal, state, and local air quality regulators and people developing public education campaigns. The survey instrument and results have been provided to state and local air quality agencies.
Changes in Scope and Personnel from Original Grant Proposal
One of the co-principal investigators, Dr. Zachary Pekar, has left RTI International to take a job with EPA. He is no longer an investigator on this grant. As discussed in the 2002–2003 annual report, the survey data collection cost more than we had budgeted in the proposal. To deal with the budget cut and the loss of Dr. Pekar to the team, we decided to seek outside funding for the risk assessment/exposure analysis portion of the proposal.
Journal Articles on this Report : 1 Displayed | Download in RIS Format
Other project views: | All 9 publications | 1 publications in selected types | All 1 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Mansfield C, Johnson FR, Van Houtven G. The missing piece: valuing averting behavior for children’s ozone exposures. Resource and Energy Economics 2006;28(3):215-228. |
R829544 (Final) |
Exit Exit |
Supplemental Keywords:
ambient air, risk assessment, children, public policy, nonmarket valuation, geographic information system, GIS,, RFA, Economic, Social, & Behavioral Science Research Program, Health, Scientific Discipline, Health Risk Assessment, Economics, Environmental Monitoring, Children's Health, decision-making, Ecology and Ecosystems, Social Science, Economics & Decision Making, social psychology, surveys, contingent valuation, behavioral effects, social impact analysis, valuation, random utility model, air toxics, economic valuation, internet survey, incentives, ozone, decision analysis, behavioral assessment, preference formation, air pollution, environmental values, survey, adult valuation of children's health, willingness to pay (WTP), psychological attitudes, children's environmental health, environmental health hazard, stated preference, willingness to pay, economic objectivesProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.