Chapter 5
Haunted Housing: An Epidemic of Anxiety
What to Believe?
According to epidemiologists interviewed for an article in Science magazine, convincing studies have to show "a very strong association between disease and risk factor and a highly plausible biological mechanism." A number of epidemiologists, some of whom had published erroneous results in the past, "say it is so easy to be fooled that it is almost impossible to believe less-than-stunning results."
There are two general methods of investigating the strength of an association, both of which figure prominently in the Science article. The first is "statistical significance," which refers to statistical analyses designed to estimate the possibility that the connection between exposure and effect could have occurred by chance. The epidemiologists interviewed for the article worried that the tests could be deceiving because it is literally impossible to be aware of and to guard against all the factors in a study that could contribute to a false-positive finding. Simply stated, they were not certain that the criteria for statistical significance are sufficiently rigorous. Against that background, the EPA decision to eliminate statistical significance as a criterion in its evaluation of epidemiologic studies about environmental causes of cancer appears to be a further attempt to allow the agency to pick and choose results that it likes.
The second method for judging the strength of an association in an epidemiologic study is to examine the magnitude of the relative risk that exists between the exposure and the effect. Many of the scientists interviewed for the Science article insisted that effects observed in exposed populations need to be three or four times higher than those observed in unexposed populations (that is, the relative risks must be 3 or 4 or more) to be convincing. Few of the studies that the EPA relies on to point to risks from radon, asbestos, EMF, or lead achieve that criterion. Application of that criterion would reveal EPA estimates of the risks from residential exposure to be a house of cards.
Given those areas of dispute, the lay reader is well advised to read critically and skeptically the many popular accounts of the latest scientific findings. There should at least be a very strong association and a highly plausible biological mechanism. Applying the first criterion, if it includes requiring a relative risk of 3 or 4, would eliminate any connection between the risks discussed in this book and human health. Some scientists, however, object to the use of a relative risk as high as 3 or 4 as a criterion.
Laying aside the criterion that relative risks should be 3 or 4 or more and applying the general criteria of strong association and plausible mechanism to the four risks discussed in this book separates them into two groups. EMF is in a category by itself. There is no strong association between exposure to EMF and disease even in workplaces where exposure is far greater than in homes. Moreover, as discussed in chapter 4, there is no plausible biological mechanism that can explain how EMF could cause cancer.
Asbestos and radon have certainly caused cancer in occupational groups, and decades ago lead caused terrible neurological diseases in highly exposed workers in highly polluted workplaces. The effects from those high exposures can be considered settled, but questions remain about whether the mechanisms that cause disease at high levels of exposure operate to cause disease at the far lower levels encountered in housing. Studies that have examined the effects of asbestos in buildings and of indoor radon on health provide no convincing or consistent evidence of harm, but they are not powerful enough to rule out the possibility that some small risk is present.
The most important studies warning of risks from low-level lead exposure have focused on reduced IQ scores. A handful of studies, essentially those carried out by Dr. Herbert Needleman, have produced evidence of an effect on IQ at relatively low lead levels, but Needleman's studies have been attacked because of problems in methodology and reporting. Most telling, however, is the failure of Needleman's latest paper to find the effect anticipated (see chapter 2). That failure casts renewed doubt on the link between low-level lead exposure and decreased IQ.
The science is of fundamental importance, but it has limits. It cannot tell us definitely that no risks exist. It is impossible to prove a negative, and that is what would be required to be definite. Of equal importance is the fact that it is possible to go through information that is available about every imaginable connection between exposure and health effect and construct a case for its existence. The EPA often prepares a risk assessment and puts it out for public and scientific review. Members of the public, including scientists, may object to the assessment and its conclusions, but there is no formal method by which to force the agency to reconsider its risk estimates.
Can the Science Be Improved?
The difficulties that arise when science cannot answer the questions that are asked of it have been recognized for years, and there has been no shortage of suggestions for methods to "improve the science." So far, none of the suggestions has proved sufficient.
Why can't Congress take the lead and judge the validity of the agency's risk assessments? Everyone knows the answer. Congressional committees and subcommittees are composed of lawyers, not scientists. Congress simply lacks the technical expertise either to voice doubts or to search out countervailing views, and it fears appearing insensitive on issues that supposedly affect public health.
Almost 20 years ago, the "science court" was advanced as a method for bringing better science to bear. The "court" had a variety of forms; basically, it was to be a group of qualified scientists who would sit as judge or as judge and jury, hear scientific evidence, and decide whether the data were convincing. Usually the idea of a court included provisions for attorneys and cross-examination.
There is little enthusiasm for the court idea now, one reason for its demise being the realization that many of the tricky "scientific" questions or uncertainties are not scientific. Instead, they are "trans-scientific," meaning that they are asked in scientific terms but impossible to answer with current scientific knowledge. For example, "Does indoor radon cause cancer?" is a scientific question. When, however, analysis of the studies that can be used to investigate the question reveals that the studies are not sufficiently powerful to answer the question, it becomes trans-scientific. That does not mean that a trans-scientific issue will always remain trans-scientific; new techniques or new insights may make it possible to address the issue scientifically. Here and now, however, many questions are trans-scientific and likely to remain so for some time. With that realization, the science court, which would consider only the science that could be brought to bear on an issue, lost much of its appeal.
Short of a science court that would make scientific decisions, there are many suggestions for and a few examples of committees of scientists who advise the government. As the controversies described in this book illustrate, none of the existing committees has distinguished itself as a source of reliable advice. Cynical Washington observers view many of the advisory panels as covers for congressional or regulatory indecision. While the committees deliberate, investigate, review, and write, the politicians and regulators can delay decisionmaking, claiming that they are awaiting more information.
Congress established the Scientific Advisory Board (SAB) of the Environmental Protection Agency to provide scientific oversight to the agency. The SAB argued against the EPA's proposed regulation of radon in water, and its deliberations may have influenced the decision of Congress to place a moratorium on that regulation in the early 1990s (see chapter 1). It is difficult to determine which carried more weight with Congress--the SAB's judgment or the cries of protest from the water utilities, which would have been forced to foot the bill. Certain it is that the SAB did take a stand on the issue and that Congress did place a moratorium on the regulation.
Nevertheless, there is little reason to expect great things of the SAB. The mechanisms by which its members are appointed are a general mystery. The chairman and committee chairmen are clearly political appointees, serving at the pleasure of the president and the administration, a procedure that underlines the political implications of the work of the board. Moreover, all the staff support comes from the EPA. No matter how independent some of the staff may be from the EPA's drive to regulate, the overwhelming dependence on EPA experts, reports, and staff support must inevitably bias the SAB.
There are alternatives to placing an advisory body in the agency it is to advise. For example, Congress established a committee in the Department of Health and Human Services to review and advise on the study by the U.S. Air Force of the health of Air Force personnel who sprayed Agent Orange. That might be a useful model, but it is extremely clumsy to consider setting up a new advisory body for every issue. Even if it were attempted, it can be expected that the number of critical, qualified scientists who would be willing to serve on such boards would be far fewer than the number of seats to be filled.
The best known source of scientific advice for the government has been around for well over a century. In 1863, during the Civil War, Congress chartered the National Academy of Sciences as an adviser to the federal government. The academy selects its own members, and membership is a mark of scientific achievement. Like many scientists, academy members are not eager to participate in activities that take them away from their laboratories, field sites, clinics, and offices, and very few academy members serve on the "academy panels" that investigate the issues on which the government seeks advice. Those issues are generally identified by Congress, and Congress appropriates a sum of money to an executive branch agency that then contracts with the academy for a study.
When requests for studies, with their accompanying hefty sums of money, arrive at the academy, the National Research Council, the Institute of Medicine, or the National Academy of Engineering, the operating arms of the academy, organize "academy studies" and set up "academy committees" that have few academy members. Whatever conclusion a committee may reach about the merits of an issue, it can be counted on to recommend more research. That is far from unexpected. More research might provide more information. If the information is incontrovertible, the decision to be made will be apparent to everyone and noncontroversial. What more can anyone in public life desire? Another good reason lies behind calls for more research: most of the committee members will benefit from the flow of more dollars into the research effort.
In addition to calls for more research, most academy reports contain ``on the one hand, on the other hand'' conclusions. Unsatisfying as such conclusions can be, they are expected. After all, if the answer were clear-cut, Congress would not have requested the study.
The Office of Management and Budget, which is located in the Executive Office of the President, is now seldom mentioned as a brake on regulatory drive, probably because of politics. During the "Reagan years," the OMB was criticized because it was trying to block the programs initiated by the Democratic Congress. During the "Clinton years," the executive branch has been in the different position of trying to maintain and expand programs in the face of congressional opposition.
It is reasonably easy to propose constituting a new group of scientists to try to moderate the drive of Congress and the regulatory agencies to legislate regulatory solutions to every identifiable environmental concern. In practice, it would be exceedingly difficult. Who would fund such a group? Even if the executive branch or the Congress were of such a mind, funding by the federal government would increase the chances that one or another executive branch agency would capture the group.
Ideally, voluntary contributions from citizens, including scientists, would provide funding. There is no shortage of private-sector groups working to make regulations more rational, so such funding could in all probability be organized. But the idea will make sense only after thought is given to how the group would work. Such a group has a chance of success only if pledges of cooperation are secured from scientists with prestige sufficient to guarantee that they will be heard. The location of the group also raises problems. All universities are dependent on the federal government for part of their funding, so a group set up within a university department could be subject to political pressures. A free-standing organization is another possibility, but it might be subject to attacks because of its sources of funding (whatever they might be). The possibility of such attacks reinforces the idea that such an organization will succeed only if its scientific advisers or staff have excellent credentials and demonstrated integrity.
The science that underlies the furor over residential risks is uncertain and mechanisms to improve the science are far from clear, but there are few doubts about the costs imposed by government risk assessments and regulations. The costs are real, regardless of the reality or unreality of the risks that are used to justify the expenditures.
Cost
Scientists at the Harvard Center for Risk Analysis collated the costs of 500 ``life-saving interventions'' and published the results in 1995. One of their analyses compared the median "cost per life-year saved" for interventions in medicine, injury reduction, and toxin control. The median cost per life-year saved ranged from a low of $19,000 for medicine, through $48,000 for injury reduction, to a high of $2,782,000 for toxin control. The comparison of estimated costs at various regulatory agencies reveals that the Federal Aviation Administration saves one life-year for every $23,000, whereas the EPA requires the expenditure of $7,269,000, about 300 times as much, for the same (EPA-estimated) benefit.
Indeed, when one judges by the costs of programs, the EPA leads the pack of regulatory agencies. The number of dollars spent to reduce the EPA-identified risks is a great deal more certain than the risks themselves. If the number of cases of lung cancer that would be eliminated by controlling radon, for instance, is a great exaggeration, as seems likely, the unit costs are higher than estimated. As the preceding chapters of this book demonstrate, scientific considerations apart, the regulations promulgated by the EPA constitute an enormous drain on the economy.
The Homebuyer's Dilemma
Epidemiology and the risk assessments based on it are all-important for the threats addressed in this book. The checks and balances that normally apply to regulations--that give regulated parties opportunities to object and to go to court to petition for redress--do not apply to risk assessments. Neither indoor radon nor residential asbestos nor EMF is regulated, so "getting the science (or, more correctly, the risk assessment) right" or, at least, "better," is the only check and balance on EPA efforts directed at those risks.
Once the government identified, publicized, and exaggerated risks, sellers and brokers, worried about possible liability, and buyers, worried about their health and regulations that might be forthcoming, acted to defend themselves. The lengthy disclosures now routinely included in sales contracts, the increased emphasis on inspections, and the establishment of escrow accounts for possible remediation represent protection from liability. Granted, inspections for asbestos and radon are aided and abetted by EPA programs that certify people to do the work, but the government does not require them. (In fact, the speed with which the real estate industry has adopted disclosure statements casts doubt on any need for a federal requirement.)
All of the regulatory reforms, had they been enacted, would have had no effect on the problems buyers and sellers of homes confront because of radon and asbestos. Fear takes the place of regulation--fear of being sued, fear of compromising one's health or the health of one's children. The resultant costs of inspections and remediation have become part of the sales contract and are visible to everyone connected with them.
EMF costs do not appear on a contract, but EMF can decrease the value of a home. Most often the culprit is power transmission lines or power distribution lines. The seller, unable to force the utility company to bury them, pays for the assumed risk in a reduced price when he tries to sell his house. The reduction does nothing to help the home owner, nor does it protect the utility company from liability. There are more direct costs as well. Practicing "prudent avoidance," the utility can bury the lines, site new lines elsewhere, rethink plans for expansion, and pass the costs on to rate payers--and utilities are doing so. As a result, costs are likely to rise for everyone.
Lead paint is different; it is regulated. Its sale to the construction industry and thus to consumers was legally banned nearly 20 years ago. In 1996, regulations reached the private homeowner. As of September 6, 1996, an owner of four or more housing units, any of which was built before 1978, when lead paint was banned, must tell a prospective buyer about any lead hazard in the building, give the buyer a pamphlet about the hazards of lead paint, and allow him or her 10 days to have the house inspected for "unacceptable levels of lead" before becoming obligated under the sales contract. After December 6, 1996, every owner, including the private homeowner, must comply with the notification rules. The penalty for noncompliance is $10,000 and possibly imprisonment.
The regulation of lead in public housing requires all taxpayers to pay for the ``get-the-lead-out'' programs. Given $1,000 to spend on bettering the chances of a poor child for a successful life, some people at least would prefer to spend those dollars on improving the local school and on making sure the child goes to that school. The lead in bullets, it has been pointed out, is a greater threat to many children than the lead on the walls, but the Department of Housing and Urban Development (HUD), directed by Congress, goes after the lead paint and lead dust. It removes lead or, more accurately, as reported in chapter 2, it makes efforts to remove lead, efforts that have met with indifferent success when measured by reduced lead levels in the housing or in the children. Success in terms of higher IQ or in terms of school performance, the hoped-for improvement, even if realized, would be so small that evidence of it would escape detection.
Given the wide-ranging and expensive effects of EPA risk assessments, the regulatory process with its hearings, balancing of views and opinions, and opportunities for legal interventions seems almost attractive. But there is no way of knowing what would have happened if radon, asbestos, and EMF had gone through the regulatory process. Open argument and discussion might have produced more rational and less expensive policies, as it has done (so far) with EPA's attempted regulation of radon in water (see chapter 1), but that is not the only possible outcome. A $10,000 fine and imprisonment seem rather severe penalties for not telling someone of the hypothetical risk from lead paint.
If the fears engendered by EPA's risk assessments are justified--and that seems very unlikely--lives are at stake. If they are not--and that is very likely--housing costs are being driven up needlessly. The EPA apparently thinks that the scientific niceties--considering studies and analyses that provide no support for its position, acknowledging uncertainties, weighing the benefits of spending money differently--are inconvenient. It prefers to operate in a vacuum characterized by good intentions and claims that it is protecting the public. The intentions and the promotional hype, however, seem to have done little so far except to frighten homebuyers and sellers with stories based on questionable hypotheses and to drive up costs. As repeated time and again in this book, any improvement in health will be so small as to escape detection. There may well be none at all.