Reinventing OSHA: Dangerous Reductions in Enforcement During the Clinton Administration
Peter Lurie, M.D., M.P.H.
Marti Long, J.D.
Sidney M. Wolfe, M.D.
We obtained the data for this report from the Office of Statistics, Directorate of Information Technology at the Occupational Safety and Health Administration (OSHA). The data include the number of inspections, the number of violations, the nature of these violations, and the penalty imposed, by calendar year from 1972 through 1998. All monetary data were adjusted for changes in the consumer price index and are presented in 1998 dollars.
- The total number of OSHA inspections has been steadily declining since a peak in 1975, four years after the Occupational Safety and Health Act was implemented. During every year of the Clinton administration, the number of inspections has been lower than in any year during any prior administration since the Act was in place.
- From 1994 to 1995, there was a 51% decrease in the number of Serious, Willful and Repeat (SWR) violations. There has been only a modest increase in SWR violations since then. SWR inspections during the Clinton administration are still lower than at the end of the Reagan administration.
- The percentage of SWR violations downgraded to non-SWR or dismissed entirely has been higher during the Clinton administration than during any previous administration.
- Beginning in 1987, OSHA penalties began to increase fairly steadily to a peak of $121 million in 1991. This was in large part due to Congress’s increasing the penalties for OSHA violations. In spite of these increases, between 1994 and 1995 the amount penalized dropped by 47% to $61 million. Since 1995, the penalties have risen, but they still are not as high as the pre-1995 levels.
- Between 1994 and 1995, the period when Vice-President Gore’s Reinventing Government campaign became effective, a number of crucial enforcement measures decreased significantly: the number of OSHA inspections (35%), the number of SWR violations (51%) and the amount penalized (47%).
- Some may assert that, although inspections are down, the inspections that occur are more “efficient.” We used three methods to compute “efficiency”: 1. The ratio of unprogrammed to programmed inspections (see definitions on p. 5); 2. The number of SWR violations per inspection; and 3. The average penalty per OSHA inspection. All have remained stable or have decreased during the Clinton administration. Therefore, the unfavorable trends documented in this report are not due to increased efficiency of inspections, but represent real decreases in enforcement activity.
- Only the Reagan administration enacted fewer health standards per year in office than the Clinton administration (0.25/year vs. 0.33/year).
- Depending on the enforcement measure used, compared to previous administrations, the Clinton administration’s record on protecting worker safety is:
A. The worst in the history of the OSH Act
– number of annual inspections (lowest)
– percentage of proposed SWR violations dismissed or downgraded (highest)
B. Worse than the Bush administration’s
– number of SWR violations (lower)
– total penalties ultimately assessed (lower)(1)
C. No better than the Bush administration’s
– ratio of unprogrammed to programmed inspections
– number of SWR violations per inspection
– percentage of proposed penalties ultimately assessed
– average penalty per inspection.(1)
In summary, on certain measures the Clinton administration’s OSHA enforcement record is either similar to or worse than the Bush administration. On other measures, it is the worst since the OSH Act was promulgated.
Death and disability due to unsafe or unhealthy workplaces remain America’s hidden epidemic. In 1994, there were 6.8 million job-related injuries and illnesses in the private sector alone, an average of more than 18,000 injuries and/or illnesses each and every day of the year.(2) The cost of these injuries and illnesses has been estimated at $120 billion for 1994 alone.(3) Researchers at Mt. Sinai Medical School have estimated that 50,000 to 70,000 workers die each year as a result of major occupationally acquired diseases like cancer, lung disease and coronary heart disease.(4) In 1998, the number of confirmed deaths due to occupational injuries in the U.S. was 6,026, approximately one-tenth the estimated number of deaths due to occupational illnesses.(5)
Yet, despite this epidemic of illnesses and injuries, we hear little about this enormous public health problem. A recent exception was The Washington Post’s series of articles describing the deaths of radiation-exposed workers in Paducah, Kentucky.
The Occupational Safety and Health Act (OSH Act or the Act) was signed by President Nixon on December 29, 1970 and became effective on April 28, 1971. The OSH Act created three federal agencies: the Occupational Safety and Health Administration (OSHA) within the Department of Labor; the Occupational Safety and Health Review Commission; and the National Institute for Occupational Safety and Health (NIOSH) within the Department of Health and Human Services. The OSH Act covers the private, but not the public, sector. Public sector employees may be protected under other state or federal laws.
OSHA’s duties include promulgating standards, inspecting workplaces for violations of standards, and prosecuting violations. The Commission is responsible for resolving disputes between OSHA and violators (usually employers) of the OSH Act. NIOSH conducts research on occupational hazards and makes recommendations for standards.
The Secretary of Labor is charged with inspecting workplaces for violations. An inspection may be the result of an employee complaint, or it may occur at OSHA’s own initiative. An inspector can either issue an order to comply with the requirements of the Act or he or she can issue a penalty. There is only one criminal penalty available under the Act, and that is if a worker actually dies from a willful violation.
OSHA is also charged with promulgating exposure standards for particularly hazardous chemicals or working conditions. Interim standards, which were permitted during the first two years of the Act, are in effect until revoked. Some of these standards (e.g., ionizing radiation and the “400 air contaminants”) are still in effect today. Permanent standards are enacted after an elaborate process, involving many opportunities for review by both the public and the courts. This lengthy review process has stretched the process of enacting a new permanent standard to as much as 5 to 10 years from proposal to enactment.
In 1984, Public Citizen’s Health Research Group released “Decreased Law Enforcement at the Food and Drug Administration and the Occupational Safety and Health Administration FY 1981-1984.” The report compared the Reagan administration’s enforcement record from 1981 through 1984 with the four years of the Carter administration. The results showed reduced numbers of inspections, fewer Serious, Willful and Repeat (SWR) violations and lower penalties for SWR and non-SWR violations. As this report will document, the Reagan administration made some improvements in its second term. We now compare the Clinton administration’s record on occupational safety and health to previous administrations’ records on many of these same measures.
Changes in OSH Act enforcement should be viewed in their overall policy context. Beginning in 1993, the Clinton administration, spearheaded by Vice-President Al Gore, embarked on a program called Reinventing Government. In this initiative, the government shifted its priorities away from process evaluations (e.g., number of inspections made) and punishments (e.g., number of violations found). Agencies were directed instead to cut “obsolete” regulations, and to negotiate with regulated industries instead of dictating to them.(6) Examples of the policy shift are:
- OSHA was instructed to build “partnerships among regulators and business”(7)
- employers who correct workplace safety and health hazards within 24 hours of an OSHA inspection can now receive a 15-percent penalty reduction(8)
- federal agencies should promulgate only such regulations “as are required by law, are necessary to interpret the law, or are made necessary by compelling public need.”(9)
This report documents the disastrous consequences of this shift in policy. Rather than increased efficiency, the result has been to decrease worker protections in the workplace. On this Labor Day, as we honor the contributions of U.S. workers, it is time to admit that this policy has been a failure that has left workers unprotected in their workplaces.
On August 27, 1999, we obtained the data for this report from the Office of Statistics, Directorate of Information Technology at OSHA. The data were arranged by calendar year from 1972 through 1998 and comprised the following seven categories:
- Number of Inspections
- Number of Proposed SWR Violations
- Number of Current SWR Violations
- Number of Proposed non-SWR Violations
- Number of Current non-SWR Violations
- Proposed Penalties (in dollars)
- Current Penalties (in dollars).
Violations are in two main categories: those that are “Serious, Willful and Repeat” (SWR) and those that are non-SWR. Ninety-eight percent of SWR violations are “serious.”(10) SWR violations are sometimes downgraded to non-SWR violations during the appeals process. It is extremely rare for the opposite to occur.
Inspections were further categorized as “programmed” and “unprogrammed.” Programmed inspections may either be entirely random (as is the case for construction) or they may be based on high reported accident rates at a particular facility or knowledge that a particular industry is hazardous. Unprogrammed inspections follow the report of an accident in which a worker is killed or 3 or more workers are hospitalized. While these categories are not strictly the equivalents of random and for-cause, they are rough approximations, with programmed being closer to random and unprogrammed being closer to for-cause. A shift from one category to another is a reasonable measure of regulatory priorities.
The data are categorized as either “proposed” or “current.” The proposed number is that proposed by OSHA at the time of the inspection, while the current number is the classification of these same violations as it is now. The proposed fine may be reduced or dismissed during the appeals process. Because many years have elapsed since most of the inspections occurred, “current” often represents the fine actually collected. A reduced or dismissed penalty appears with the data for the year in which the inspection occurred. For example, if a penalty was originally imposed in 1980, and was decreased in 1985, OSHA will retroactively correct the 1980 figure, not the 1985 figure.
Maximum penalties for OSHA violations were significantly increased by Congress in 1990. Before 1990, the penalty for a non-SWR violation was $0 to $1,000; for a serious violation it was $0 to $1,000; and the maximums for willful and repeat violations were only $10,000. The penalty for a non-SWR violation now is from $0 to $7,000; for a serious violation it is from $1 to $7,000; for a repeat violation it is from $0 to $70,000; and for a willful violation it is from $5,000 to $70,000. All penalties are adjusted to 1998 dollars. We used the annual Consumer Price Index-Urban (CPI-U),(13) published by the U.S. Department of Commerce, for this adjustment.
We also constructed a list of health standards promulgated by OSHA since its inception by reviewing the Code of Federal Regulations Title 29 §1910 and cross-referencing this list with a similar list prepared by the AFL-CIO in 1995.(14) Interim standards, such as ionizing radiation and the 400 air contaminants, were excluded. Regulations not covering chemical exposures, such as medical records and noise, were also excluded. Standards are characterized according to the date they were finalized in the Federal Register.
Trends in the Number and Types of OSHA Inspections
The total number of OSHA inspections reached its peak in 1975, shortly after the enactment of the OSH Act (see Figure 1). Inspections have since declined steadily, with a sharp drop of 35% in 1995 compared to 1994. While the number of inspections rose somewhat in the period after 1995, the number of inspections in 1998 was still 22% below the number in 1992, the year before the Clinton administration took office. Every year the Clinton administration has been in office, inspections have been lower than in any year during any prior administration since the Act was in place.
Figure 1 also demonstrates trends in programmed and unprogrammed inspections. Since 1983, programmed inspections have continued to fall. The number of unprogrammed inspections (those closer to “for-cause”) increased during the Reagan administration, before starting to decline in 1989 during the Bush administration. This downward trend has continued during the Clinton administration, with the number of unprogrammed inspections in 1995 and 1996 lower than in any year since 1972 except for 1982 and 1983.
In addition to the seven primary enforcement indicators listed in the Methods section, we also developed three indices of inspection efficiency. The first is the ratio of unprogrammed to programmed inspections. While we believe that both unprogrammed and programmed inspection rates are substantially lower than they should be, the higher the unprogrammed to programmed ratio, the more cost-effective the inspection program is likely to be because the inspection is based on a higher suspicion of danger. The ratio of unprogrammed to programmed inspections reached its peak in 1978, during the Carter administration, before falling rapidly to its all-time low in 1982 (see Figure 2). After 1982, the ratio increased steadily until 1988, when the ratio began to level off at between 1 and 1.5. There has been very little change in this ratio during the Clinton administration.
In the remainder of this report we do not present programmed and unprogrammed inspections graphically as the trends for these inspections for the remaining indicators are generally very similar to the trends for total inspections.
Trends in the Number and Types of OSHA Violations
The total number of current SWR violations in OSHA inspections rose fairly consistently until 1989, and remained stable until 1994. In 1995, a sharp 51% decline occurred (see Figure 3). Since 1996, the numbers have risen somewhat but are still lower than at the end of the Reagan administration. The number of proposed SWR violations shows a similar pattern to the current SWR and so is not depicted graphically.
The percentage of proposed SWR violations (for a given year) dismissed or downgraded to non-SWR has steadily increased (see Figure 4). Until 1983, less than 1% of proposed violations were dismissed or downgraded. Since then, the percentage of SWR violations downgraded to non-SWR violations or dismissed entirely has risen to about 5% from 1993 onward, more than during any previous administration. This percentage is more than four times the average percentage during the Reagan administration. In 1998 alone, this represented 4,224 SWR violations dismissed or downgraded.
The total number of current non-SWR violations in OSHA inspections rose to its peak in 1975 and then dropped dramatically (see Figure 5). This trend was driven primarily by changes in programmed inspections. Since 1978, there has been a slow decline in these violations. Non-SWR violations during the Clinton administration are lower than during any previous presidency. Proposed non-SWR violations show the same patterns as do the current non-SWR violations and so are not depicted graphically. The ratio of current non-SWR violations to proposed non-SWR violations is also not presented as it represents both downgraded SWR violations and dismissed non-SWR violations, and so is difficult to interpret.
Our second index of inspection efficiency is the number of violations detected per inspection. If inspections were becoming more efficient, the number of violations detected per inspection should increase. This was the case for current SWR violations until 1990, when the numbers began to level off (see Figure 6). The rate of SWR violations per inspection in 1995 showed a drop of 24% compared to 1994 and has remained stable ever since. The number of proposed SWR violations per inspection shows a very similar pattern to the current SWR data and so is not presented graphically.
Current non-SWR violations per inspection showed a peak in the mid-1970s, and have generally declined ever since, reaching their lowest points ever in the Clinton administration (see Figure 7). The number of proposed non-SWR violations per inspection shows a similar pattern to the current non-SWR data and so is not presented graphically.
Trends in OSHA Penalties
After adjusting to 1998 dollars, the data show that until 1987, the current penalties remained at a fairly stable level of under $40 million per year (see Figure 8). In 1988, the amounts began to increase fairly steadily to a peak of $121 million in 1991. The increase coincides in part with the increases in permissible OSHA penalties that were passed in 1990 and went into effect in 1991, the year of the greatest increase in penalties. Between 1994 and 1995, the amount penalized dropped by 47% from $116 million in 1994 to $61 million in 1995. Since 1995, the current penalties have risen, but they still are not as high as the levels prior to the Clinton administration’s Reinventing Government initiatives. The average current penalty during the Clinton administration is $26 million less than during the last two years of the Bush administration, the first years when the increased fines were in effect. In most years, unprogrammed inspections generated considerably higher total penalties than programmed ones.
The proposed penalties show a similar pattern to the current penalties but are substantially higher than the current penalties. From 1972 to 1980 the ratio of current to proposed penalties decreased consistently from about 100% to about 60%, where it has remained for the majority of the period since (see Figure 9). This means that about 40% of the total number of dollars initially fined were eliminated by OSHA or the courts. The Clinton administration is therefore no better than the Reagan and Bush administrations and worse than the administrations of Nixon, Ford and Carter on this index.
Our third index of inspection efficiency is the average current penalty per OSHA inspection. The data show that this index was constant at $1,000 or less for the years 1972 to 1988 (see Figure 10). Starting in the late 1980s, the index began to climb rapidly, reaching a 1991-1992 average of $2,644. Again, 1991 and 1992 were used for comparison purposes, because they are the first two years after the new OSHA penalty scale went into effect, The average current penalty per OSHA inspection during the Clinton administration is $2,501, 5% lower than during the last two years of the Bush administration. Unprogrammed inspections tend to be somewhat more efficient on this index. The average proposed and current penalties per OSHA inspection showed similar trends. Consequently, the proposed penalties are not shown graphically. The current penalties per inspection are substantially lower than the proposed ones; the comparison between current and proposed penalties has already been presented in Figure 9.
We also evaluated the number of final health standards that have been enacted since OSHA’s creation. Since 1972, there have been 17 health standards enacted under the “final standards” (as opposed to interim standards) process defined in the OSH Act (see Table). The rates of standards promulgated per year by presidents serving at least one full term in office are: Jimmy Carter: 1.5/year; Ronald Reagan: 0.25/year; George Bush: 0.75/year; and Bill Clinton: 0.33/year. President Clinton’s rate is therefore comparable to the low-water mark set by Ronald Reagan.
Depending on the enforcement measure used, the Clinton administration’s record on protecting worker safety is the worst in the history of the OSH Act (e.g., the total number of annual inspections, percentage of proposed SWR violations dismissed or downgraded, number of current non-SWR violations, number of current non-SWR violations per inspection), worse than the Bush administration’s (e.g., number of current SWR violations, total current penalties) and no better than the Bush administration’s (e.g., ratio of unprogrammed to programmed inspections, number of current SWR violations per inspection, percentage of proposed penalties made current, average current penalty per inspection). In addition to these measures, President Clinton also has a poor record of promulgating new standards. In his six years in office, only two new standards have been enacted. In the history of the OSH Act, the Clinton administration’s record on health standards is most similar to Ronald Reagan’s, who promulgated two standards in his eight years, the lowest rate in OSHA history.
The effect of the Reinventing Government program is most apparent in the dramatic enforcement changes that occurred between 1994 and 1995. Decreases in the enforcement of four of our measures occurred in 1995. This took place immediately after the Reinventing Government program was implemented.
Some may assert that although inspections, for example, have declined, this is consistent with the Reinventing Government program’s goal of increasing efficiency of regulatory agencies. The data belie this assertion. We used three different methods to measure inspection efficiency, and found that there has been no improvement in any of these during the Clinton administration. In fact, the only improvements in efficiency occurred during the final years of the Reagan administration or during the Bush administration. What has occurred is not an increase in efficiency, but a decrease in enforcement. This is the true legacy of Reinventing Government.
There is some evidence that fatal workplace injuries may be declining. The fatal occupational injury rate in 1998 was 4.5/100,000, compared to 5.3/100,000 in 1992.(15) Similarly, the rate of non-fatal injuries and illnesses among U.S. workers (excluding the self-employed and those working for the state and federal governments) has declined from 11.0/100 full-time workers in 1973 to 7.1 in 1997. However, there has been no decline in the rate of injuries and illnesses resulting in lost days from work during this period, although there has been a relatively recent decline from the high point reached at the end of the Bush administration.(16)
We do not think that these data provide persuasive evidence that occupational safety and health hazards are declining adequately in the U.S. First, there is enormous underreporting of occupational injuries (these are monitored through logbooks maintained by employers) and many occupational illnesses occur years after exposure and may not be attributed to occupational exposures. In many cases, the worker will have left the workplace or have died and these illnesses will not be reflected in the statistics. Even the illness data collected are suspect; many may not be occupationally related.
Any improvement in injury rates must be viewed in the context that the slightly over 6,000 fatal injuries that now occur is only one-tenth the 50,000-70,000 fatal occupationally related illnesses that are estimated to occur.4 More attention must be paid to the much larger problem of occupational illness, a problem the federal government monitors poorly. A 1984 Congressional report described the U.S. occupational disease surveillance system as “fragmented, unreliable and 70 years behind the times.”(17) Little has changed since then.
Finally, it is not appropriate to ascribe any decrease in injury or illness rates to a particular government intervention such as the Reinventing Government program. If the program were to have any impact, it would like not occur until years after the program was initiated. Even then it would be difficult to differentiate its impact from other changes in the workplace such as increases in employment in the service sector of the economy.
We acknowledge that attempts to regulate by one administration may not result in final regulations until years later, in some cases in the next administration. Due to these delays, some of the standards finalized during one administration were in fact proposed during a previous administration. But with six years in office, the Clinton administration has only finalized two health regulations. It has proposed no new health standards of its own, even though there is no shortage of occupational chemicals in great need of proper regulation. The administration fought Public Citizen’s lawsuit to get chromium, a well-documented lung carcinogen, properly regulated, and still has not issued even a proposed rule to regulate it. Much-needed ergonomic standards as well as standards on silica, metalworking fluids and beryllium are also proceeding at a snail’s pace. Almost all of the standards that have been promulgated by all administrations have followed suits from consumer groups or unions.
There is a real need for OSHA to be modernized. The American workplace has evolved in ways not reflected in current regulations. The few standards that exist are based on an eight-hour workday and a forty-hour workweek, but many Americans find themselves required to work far longer hours. The shifting of jobs to cheaper overseas facilities and into the home makes it less likely that workers will report hazardous conditions. This makes strong enforcement by OSHA all the more critical.
It is time to admit that “reinventing” OSHA has been a disaster for American workers. OSHA was enacted, and is still needed, precisely because employers will not voluntarily provide adequate protections. Existing law provides at least some framework for identifying and punishing those who failed to obey the law. On this Labor Day, we should be renewing the OSH Act’s original promises to American workers, rather than continuing a policy in which employers have little incentive to improve.
TABLE: HEALTH STANDARDS ENACTED
1974 Thirteen carcinogens (1 standard), vinyl chloride, coke oven emissions
1978 1,2-dibromo-3-chloropropane (DBCP), acrylonitrile, arsenic, benzene, cotton dust, lead
1984 Ethylene oxide
1991 Bloodborne pathogens
1992 Cadmium, methylenedianiline
1997 Methylene chloride
This information is based on when the standards were finalized in the Federal Register. Subsequent changes to a standard are not included. Two other standards (ionizing radiation and air contaminants) were enacted as interim standards (see Introduction), and consequently these are not included in the Table above.
10. While none of the three subcategories of SWR violations are explicitly defined in the statute, the courts have defined “serious” as a violation that creates the possibility of an accident associated with a substantial probability of death or serious physical harm (45 A.L.R. Fed. 785). The courts have not provided a clear definition of “willful”, but according to the OSHA Field Operations Manual, an employer need not entertain a bad purpose or malicious intent in order to have a violation classified as willful (31 A.L.R. Fed. 551).(11)
11. See William J. Danne, Jr., Annotation, What Constitutes “Willful” Violation for Purposes of §§17(a) and (e) of Occupational Safety and Health Act of 1970, 31 ALR Fed 551. (12)
12. See Ann Wooster, Annotation, When has Employer “Repeatedly” Violated Occupational Safety and Health Act of 1970 within meaning of §17(a) of Act; see also, Reich v. D.M. Sabia Co., 90 F.3d. 854, 17 O.S.H. Cas. (BNA) 1680, 1996 O.S.H. Dec. (CCH) para 31117, 151 A.L.R. Fed. 669 (3rd. Cir. 1996).
13. Downloaded from Bureau of Labor Statistics web site on August 28, 1999 (ftp://ftp.bls.gov/pub/special.requests/cpi/cpiai.txt).
15. Data for number of fatal injuries are from: Occupational Safety and Health Administration. Fatal occupational injuries by event or exposure, 1992-1997. Downloaded on August 28, 1999 from http://stats.bls.gov/oshfat1.htm and from U.S. Department of Labor, Bureau of Labor Statistics. Census of Fatal Occupational Injuries, 1998. U.S. Department of Labor, August 4, 1999. Data for the number of employed persons in the U.S. is from Department of Occupational Safety and Health. Death on the job: the toll of neglect, 8th edition. AFL-CIO, April 1999 and written communication from Bureau of Labor Statistics, August 30, 1999.
16. Occupational Safety and Health Administration. Occupational injury & illness incidence rates per 100 full-time workers 1973-1997. U.S. Department of Labor. Downloaded from http://www.osha.gov/oshstats/bltable.html.
17. Committee on Government Operations. Occupational Illness Data Collection: Fragmented, Unreliable and Seventy Years Behind Communicable Disease Surveillance. Washington, D.C.; 1984. 60th Report by the Committee on Government Operations. House Report 98-1144.
The following figures may be found in the PDF of this report.
Figure 1: Number of OSHA Inspections 1972 – 1998
Figure 2: Ratio of Unprogrammed to Programmed OSHA Inspections, 1972 – 1998
Figure 3: Current Serious, Willful and Repeat Violations in OSHA Inspections 1972 – 1998
Figure 4: Percentage of Proposed Serious, Willful and Repeat Violations Dismissed or Downgraded, 1972 -1998
Figure 5: Current Non-Serious, Willful and Repeat Violations in OSHA Inspections 1972 – 1998
Figure 6: Number of Current Serious, Willful and Repeat Violations per OSHA Inspection 1972 – 1998
Figure 7: Number of Current Non-Serious, Willful and Repeat Violations per OSHA Inspection 1972 – 1998
Figure 8: Current Penalties in OSHA Inspections, 1972 – 1998
Figure 9: Percentage of Proposed Penalties in OSHA Inspections Made Current, 1972 – 1998
Figure 10: Average Current Penalty Per OSHA Inspection, 1972 – 1998