Air pollution and children's health in Sweden : An enquiry into how the economic benefit of improvements in children's health resulting from reductions in air pollution can be assessed

Abstract

lean air is one of 16 Environmental Objectives adopted by the Swedish parliament to guide action towards a sustainable environment. This project is part of the work undertaken by the Swedish Environment Protection Agency (Swedish EPA) to bring about the fulfilment of this goal. Much research has been undertaken regarding air pollution and health impacts in the adult population but much less is known about how pollutants influence children's health. The overriding purpose of this study has therefore been to see how and to what extent the economic benefit from reducing these impacts can be calculated. To answer this question we provide a brief introduction on the method commonly used to do these kinds of benefit calculations. Two crucial inputs into these calculations are estimates of the health impacts and estimates of the economic values for the health impacts. We therefore start by providing a summary of the current stateof-art regarding these inputs which is based on a survey of the literature in each area. We then perform two case studies that describe how these economic benefits can be calculated and what influences the results. The calculation is based on the findings in the literature reviews and we also describe the exposure assessment that is another crucial input into these calculations. The report ends with suggestions for future research. Regarding air pollution and health impacts, the finding is that air pollution exposure has been associated with a number of health outcomes in children, many of these partly overlapping and related to respiratory effects. Both long-term exposure and short-term fluctuations have been correlated with adverse effects. However, the involved exposure variables are often not source specific, but may in some cases act as acceptable indicators of traffic related air pollution. Only for a limited number of health effects we have found exposure-response functions that may be used to quantify health effects in children. Most of these have been described also in a previous report (Naturvårdsverket, 2010). New for this report is an estimated exposure-response function for the development of air-way disease in the 5-18 age group. For the short-term effects such as hospital admissions, it is possible to calculate baseline frequencies needed for the impact assessments from register data. It is more complicated to estimate the baseline in terms of prevalence (occurrence of disease) or onset of disease, but some types of impacts can be estimated using combinations of data and assumptions. On the relationship between traffic pollution and restricted activity days (for example school absences), effects on pregnancy outcome and in infancy as well as effects of early exposure later in life there is limited information. As for the economic valuation of health impacts, the conclusion in the literature is that the valuation of children's health risks is more challenging than that of adults. There are several reasons for this where children not being able to assess and value risk reductions by themselves is the most important one. There is however also the difference in age between children and adults which is likely to make a difference for the values. As in the case of the quantification of health impacts, little research has been done on the valuation of children's health risks. Therefore, so far mainly proxies have been used such as willingness to pay estimates derived from parents' choices and behaviour. The general conclusion is that economic values used for adults in general underestimate the benefits to children and that as high as two times these estimates can be relevant. Since almost no economic valuation studies of this kind have been undertaken in Sweden the estimates we propose are those used in other, mainly European, studies. Based on the findings in the literature surveys we have, as an example, calculated the benefit of a reduction in children's exposure of 1 µg/m³ of NO2 in Stockholm and Umeå. The difference between the cities that we could account for was the number of children that are exposed. The calculation was done for two endpoints; that children having wheeze develop asthma and that asthmatic children are admitted to hospital due to respiratory symptoms. According to our calculations this reduction in exposure in Greater Stockholm would generate a benefit to society of 168 million SEK per year because of fewer cases of asthma, and 47 000 SEK due to fewer hospital admissions (for the price levels in 2000). For Umeå the benefits are smaller, 8 million SEK and 2000 SEK per year. These benefit estimates however are based on a quite large reduction in air pollution. 1 µg/m³ NO2 is approximately the reduction in population exposure that resulted in the inner city of Stockholm from the trial with congestion charges where traffic in this area decreased by 15%. To achieve the same reduction in Greater Stockholm or Umeå would require measures that result in quite important emission reductions from transport. To determine if such measures are beneficial from a socioeconomic point of view would require a comparison of benefits and costs on the local scale of the chosen measures. In general it is found in the literature that the benefits are larger when emissions are reduced in densely populated areas. We also discuss how different assumptions influence the results and the uncertainties related to these types of calculations. There are uncertainties in every part of the calculation chain; exposure, impact assessment and economic valuation. One way to account for these uncertainties is by doing a sensitivity analysis where alternative assumptions are used for important inputs. In our calculations an influential assumption is for example the probability that children with wheeze develop asthma later in life. The largest uncertainty however is probably the cause and effect of single pollutants. In this study NO2 is used since it is a good indicator of emissions from traffic but if this is the true cause of the effects is still a matter of research and discussion. This is the first attempt to calculate the benefits for children in Sweden of reducing air pollution. Due to lack of data we have only been able to give an indication of the size of the benefits and only for endpoints related to respiratory diseases. Therefore, further research is needed in order to determine the accuracy of these estimates, the size of the benefit for other endpoints and all children in Sweden and how the benefits vary between different geographical areas. However we consider such research to be warranted since our estimates suggests that reducing children's exposure to air pollution result in important economic benefits and there is a need for policy makers to know if and when this is the case.