Does remediation save lives? On the cost of cleaning up arsenic-contaminated sites in Sweden

Swedish environmental policy is based on 16 environmental quality objectives (Gov. Bill 2000/01:130 and Gov.Bill 2004/05:150).1 One of the most challenging objectives,‘A non toxic environment’, has two interim targets that concern remediation of contaminated sites. In sum, they state that the highest priority should be given to sites posing the highest risks to human health and the environment.2 By eliminating pollutants in soil, groundwater and sediment, the interim targets aim to reduce risks to human health and the environment. In Sweden, 83,000 sites are potentially contaminated due to previous industrial activities. According to the Swedish Environmental Protection Agency (EPA), the administrator of the governmental funds for remediation, approximately 1500 of these sites contain contaminant concentrations that could seriously harm human health and the environment (Swedish EPA, 2008a). To reach the interim targets, all these sites need to be remediated by 2050. Remediation of contaminated sites has so far cost more than SEK 3,000 million.3 The approximated cost to mitigate the potential risks at the most harmful sites is estimated at SEK 60,000 million.4 The Swedish government’s funding for remediation presently comes in the form of a directed grant (sakanslag). The directed grant, administrated by the Swedish EPA, subsidises remediation of contaminated sites that were contaminated prior to modern environmental legislation (in 1969) or for which no liable party can be found. The directed grant amounts to approximately 455 millions annually, which corresponds to about 10 percent of the annual national funds for environmental protection (Gov. Bill 2007/08:1). To make it possible to prioritise among contaminated sites, the Swedish EPA has developed a method for risk assessment called the ‘MIFO’ (i.e. the Method for Inventory of Contaminated Sites). The risk assessment does not take into account the actual exposure at a contaminated site. Risk is instead assessed based on divergence from guideline values for acceptable concentrations given a standardised (i.e. worst case) exposure situation on an individual level. This means that a site can be remediated without any individuals actually being exposed. The expected risk reduction is consequently not quantified. This eliminates the possibility of valuing the risk reduction, which should be weighed against the remediation cost. The purpose of this paper is to analyse how health effects, in the form of cancer risks, from sites contaminated by arsenic are valued implicitly in remediation. By using an environmental medicine approach that takes exposure into account, and without underestimating the potential health consequences of arsenic exposure, our purpose is to place arsenic risk management in the overall picture of live-saving interventions. In the case of cancer prevention, it is necessary to recognise that focus on an environmental carcinogen like arsenic may draw public attention – and funding – away from mental health risks like ambient air pollution and indoor radon. Although environmental pollution accounts for less than ten percent of all cancer cases (Harvard Centre for Cancer Prevention, 1996; Saracci and Vineis, 2007), environmental factors are important to recognize since they may be preventable. We emphasise, however, the inefficiency in becoming overly concerned about small risks while, at the same time, losing sight of the large risks. If society’s spending on lifesaving measures with small effects (i.e. a small number of lives saved) crowds out spending on lifesaving measures with large effects, then remediation can, in fact, even be said to waste lives. By using data on 23 arsenic-contaminated sites in Sweden, we estimate the sitespecific cancer risks and calculate the cost per life saved by using the sites’ remediation costs. Our results show that the cost per life saved through remediation is much higher than that associated with other primary prevention measures, indicating that the ambition level of Swedish remediation may be too high.