Smallest Geographic Unit:
Date of Collection:
Unit of Observation:
company by year (Part 1, Part 15),
facility (permit) by month/year (Part 2),
facility (permit) (Part 3),
pipe within a facility (Part 4),
inspection (Part 5),
compliance schedule event (Part 6),
compliance schedule violation (Part 7),
single event violation (Part 8),
measurement/effluent violation or reporting violation (Part 9),
enforcement action (Part 10),
Occupational Safety and Health Administration inspection (Part 11),
court case (Part 12, Part 14),
court case settlement (Part 13),
company by quarter (Part 16),
individual by vignette (Part 17)
All United States based, publicly traded companies operating in 1995 that had their primary business in one of four Standard Industrial Classifications (SIC) codes: 2611 (Pulp Mills), 2621 (Paper Mills), 2911 (Petroleum Refining), and 3312 (Steel Works, Blast Furnaces, and Rolling) (Part 1, Part 15, Part 16). All facilities owned by sampled companies that operated in the same industry in the United States between 1995 and 2000 (Part 2, Part 3). All pipes within sampled facilities in the United States (Part 4).
All inspections conducted on sampled facilities in the United States between 1995 and 2000 (Part 5). All compliance schedule events that occurred in sampled facilities in the United States between 1974 and 2002 (Part 6). All compliance schedule violations that occurred in sampled facilities in the United States between 1995 and 2000 (Part 7). All single event violations that occurred in sampled facilities in the United States between 1995 and 2000 (Part 8). All measurement/effluent violations or reporting violations that occurred in sampled facilities in the United States between 1995 and 2000 (Part 9). All enforcement actions taken on sampled facilities in the United States between 1995 and 2000 (Part 10). All Occupational Safety and Health Administration inspections conducted in sampled facilities in the United States between 1990 and 2000 (Part 11).
All court cases filed against sampled facilities in the United States between 1995 and 2000 (Part 12, Part 14). All court case settlements by sampled facilities in the United States between 1995 and 2000 (Part 13). All managers of firms in the United States that were followed (1995-2000) in the secondary data analysis in 2005 (Part 17).
administrative records data,
Data Collection Notes:
The PERMIT, FACILITY, SIC, and MULTIPLE variables are available in every Permit Compliance System (PCS) file
(Part 3-Part 10).
Users should be aware that the time period of 1995 to 2000 was selected for data collection. Some data were also collected outside of this time period, however the research team focused on gathering secondary data from 1995 through 2000.
Users should be aware that they cannot link the Vignette Data (Part 17) to the secondary data (Part 1-Part 16).
Mark Cohen and Mike Vandenbergh of Vanderbilt University, along with Sally Simpson, designed the Web-based environmental compliance survey.
This data collection is comprised of administrative records data and survey data. The research team also conducted qualitative interviews with environmental inspectors to learn more about the context of enforcement, inspector perceptions of successful and unsuccessful enforcement strategies, and the correspondence between inspector reports of noncompliance and official records. However, the qualitative interview data are not available as part of this data collection.
Punitive strategies for crime control emphasize the use of formal legal sanctions, especially but not exclusively those found in criminal law, to deter current and future offenders from similar acts of misconduct. Cooperative crime control strategies emphasize the use of persuasion to facilitate and enhance legal compliance. As such, cooperative models integrate social norms, informal sanction threats (such as reputational damage), and intra-organizational control along with formal legal interventions to motivate firm compliance. The purpose of the study was to evaluate the extent to which deterrence or cooperative strategies motivated firms and their facilities to comply with environmental regulations. Additionally, the study sought to provide researchers and policy-makers with systematic information about the origins of company non/over-compliance, whether specific regulatory interventions and informal crime prevention strategies affected the behavior of managerial decision-makers and, in the aggregate, the firms in which they operated.
The project collected administrative data (secondary data) for a sample of publicly owned, United States companies in the pulp and paper, steel, and oil refining industries from 1995 to 2000 to track each firm's economic, environmental, and enforcement compliance history. Company Economic and Size Data (Part 1) from 1993 to 2000 were gathered from the Standard and Poor's Industrial Compustat, Mergent Online, and Securities and Exchange Commission, resulting in 512 company/year observations. The research team used qualitative company histories, company websites, and annual 10K reports to verify that the company was operating in 1995. Researchers also used these sources to investigate contradictions across data sources regarding primary industry, company status, and company name. Final coding reflected the majority consensus across sources. Next, the research team used the Directory of Corporate Affiliations, the Environmental Protection Agency's (EPA) Toxic Release Inventory (TRI), and the EPA's Permit Compliance System (PCS) to identify all facilities owned by the sample of firms between 1995 and 2000. Researchers then gathered Facility Ownership Data (Part 2), resulting in 15,408 facility/year observations.
The Clean Water Act prohibited the discharging of pollutants through point sources into waters of the United States without a National Pollutant Discharge Elimination System (NPDES) permit. The NPDES permit translated general requirements of the Clean Water Act into specific requirements for each facility. The Permit Compliance System (PCS) contained the NPDES permit program data on facilities permitted to discharge pollutants into United States waterways. The research team gathered various types of PCS data from the EPA for facilities in the sample. Permit Compliance System Facility Data (Part 3) were gathered on the 214 unique major NPDES permits issued to facilities in the sample. Although permits were given to facilities, facilities could have one or more discharge points (e.g., pipes) that released polluted water directly into surface waters. Thus, Permit Compliance System Discharge Points (Pipe Layout) Data (Part 4) were also collected on 1,995 pipes.
The EPA determined compliance using two methods: inspections and evaluations/assessments. NPDES permits contained effluent limitations on what and how much a facility could discharge as well as monitoring and reporting requirements. Facilities were required to take various measures of discharges into the water and submit reports of both permitted levels and actual effluent discharges to the EPA; these reports were called discharge monitoring reports (DMR). The PCS contained these data and allowed the EPA to calculate whether the facility was in violation of permitted levels (effluent or measurement violations) or whether the DMR report was late (reporting violations). The PCS contained information on inspections and on other types of violations, including compliance schedule and single event violations, and also contained data on enforcement actions against facilities that have violations of any type.
Essentially, inspections involved some sort of site visit. EPA inspections included reviewing DMR reports, interviewing knowledgeable facility personnel, inspecting the processes that generated and treated wastewater, sampling wastewater discharges, and reviewing how samples were collected and analyzed by the laboratory. Inspections that involved sampling and reviewing laboratory procedures were not regularly scheduled. They were targeted at facilities that showed evidence of permit violations or unusual trends/patterns in DMRs or labs that showed evidence of poor performance. Permit Compliance System Inspections Data (Part 5) were collected on a total of 1,943 inspections.
Compliance schedules were negotiated agreements between a pollution source and the EPA that specified dates and procedures by which a source would reduce emissions and, thereby, comply with a regulation. Permit Compliance System Compliance Schedule Data (Part 6) were collected on a total of 3,336 compliance schedule events. When facilities did not fulfill the agreement by the specified date they could be found in violation. Permit Compliance System Compliance Schedule Violation Data (Part 7) were obtained for a total of 246 compliance schedule violations. Single event violations were violations that could not be otherwise classified (as compliance schedule or effluent violations). Permit Compliance System Single Event Violations Data (Part 8) were collected on 75 single event violations.
The Permit Compliance System Measurement/Effluent and Reporting Violations Data (Part 9) contain the monthly DMR reports. Thus, it provides information on two kinds of violations: (1) discharges that exceeded permitted levels and (2) reporting violations (late or incomplete DMR reports). Permit Compliance System Measurement/Effluent and Reporting Violations Data (Part 9) were collected for 396,479 violations.
The PCS contained information regarding state and federal enforcement actions against facilities that had violations of any type, including cooperative intervention strategies (e.g., phone calls, warning letters) and more formal actions by the EPA (e.g., administrative order, enforcement conference, and emergency order) and other legal authorities. Permit Compliance System Enforcement Actions Data (Part 10) were collected on 1,730 enforcement actions.
Occupational Safety and Health Administration Data (Part 11) were collected on a total of 2,243 inspections. The OSHA data were collected by company name and include multiple facilities owned by each company and were not limited to facilities in the Permit Compliance System.
Additional information about firm noncompliance was drawn from EPA Docket and CrimDoc systems. These sources provided case-specific details regarding administrative, civil, and criminal (closed) court cases in which companies were defendants. Administrative and Judicial Docket Case Data (Part 12) were collected on 40 administrative and civil cases. Administrative and Judicial Docket Case Settlement Data (Part 13) were collected on 36 administrative and civil cases. Criminal Case Data (Part 14) were collected on three criminal cases. The information in these files overlapped to some degree.
For secondary data analysis purposes, the research team created the Yearly Final Report Data (Part 15) and the Quarterly Final Report Data (Part 16). The yearly data contain a total of 378 company/year observations; the quarterly data contain a total of 1,486 company/quarter observations.
The research team also conducted a vignette survey of the same set of companies that are in the secondary data to measure compliance and managerial decision-making. Concerning the Vignette Data (Part 17), a factorial survey was developed and administered to company managers tapping into perceptions of the costs and benefits of pro-social (e.g., over-compliance, responsiveness to counter-terrorism initiatives) and anti-social (noncompliance) conduct for themselves and their companies. A total of 114 respondents from 2 of the sampled corporations read and responded to a total of 384 vignettes (out of 456 possible) representing 4 scenario types: technical noncompliance, significant noncompliance, over-compliance, and response to counter-terrorism. Specifically, the final sample consisted of 86 respondents from a pulp and paper company and 28 participants from a steel company.
The sample began as the universe of all United States based, publicly traded companies operating primarily in one of four Standard Industrial Classifications (SIC) (Pulp Mills; Paper Mills; Petroleum Refining; Steel Works, Blast Furnaces, and Rolling) in 1995 linked to facilities (the EPA tracks compliance at the facility level) that were regulated by the EPA. Facilities were limited to those operating in the same SIC codes in order to ensure a similar culture between parent company and facility. Companies were retained for the study if they owned at least one facility operating in the same SIC code in 1995 that was categorized as a major discharger in the National Pollutant Discharge Elimination System (NPDES). Firms/facilities were tracked from 1995 to 2000. Therefore, any changes in either the company (mergers, bankruptcy, etc.) or the facility (closings, changes in ownership, etc.) were recorded through the year 2000. The final sample contains 67 companies as of 1995 (30 Pulp and Paper; 18 Steel; 19 Oil) and drops to 55 (24 Pulp and Paper; 15 Steel; 16 Oil) by 2000 (mainly due to mergers). However, given the company and facility changes, a total of 73 companies were included in the sample at some point in the six-year period.
The sampling frame for the vignette survey (Part 17) was the 55 firms that were followed in the secondary data analysis from 1995 to 2000. Only 48 of the 55 companies were still operating in the SIC codes of interest in 2004-2005 (this number was reduced to 47 once firm contact was established). These 48 companies became the universe of firms for follow-up purposes. A total of 16 companies were then identified for follow-up telephone calls. Of this group, nine ultimately declined participation and one firm reported that it was no longer in the pulp and paper business and therefore was ineligible to participate. A total of 6 firms (evenly split between pulp and paper and the steel industry) out of the 16 originally forwarded to the project leader agreed to participate in the survey, however only 3 of the 6 companies participated in the survey and 1 of the 3 withdrew from participation due to several technical glitches that occurred during administration. Therefore, Part 17 is made up of a total of 384 vignettes (out of 456 possible) which were completed by 114 individuals in 2 corporations. Specifically, the final sample consisted of 86 respondents from a pulp and paper company and 28 participants from a steel company.
Mode of Data Collection:
Environmental Protection Agency (EPA)
EPA's Toxic Release Inventory (TRI)
EPA's Permit Compliance System (PCS)
EPA's Facility Registry System (FRS)
Ward's Business Directory
Standard and Poor's Industrial Compustat
Business and Company Resource Center
Directory of Corporate Affiliations
Securities and Exchange Commission
National Pollutant Discharge Elimination System (NPDES)
Occupational Safety and Health Administration (OSHA)
Administrative, civil, and criminal case files and records
Vignette surveys of corporate managers
Description of Variables:
Part 1 contains 19 economic and size variables including subsidiary status, holding company status, standard industrial classifications, total assets, total liabilities, total stock-holders, sales, income, number of employees, and the number of facilities owned by the parent company.
Part 2 contains a total of eight variables relating to ownership including permitted facility id, facility, standard industrial classification code, a flag for facilities with multiple permits, date, generic time measure, and company.
Part 3 contains 67 variables with regard to facility characteristics.
Part 4 contains 31 variables relating to discharge points and pipe layout information.
Part 5 contains 13 inspections characteristics variables.
Part 6 contains 13 compliance schedule event characteristics variables.
Part 7 contains 11 compliance schedule violation characteristics variables.
Part 8 contains 10 single event violation characteristics variables.
Part 9 contains 79 variables including variables for matching limits and discharge monitoring reports, actual limits (permitted levels) variables, standardized limits variables, statistical base codes variables, reported units on limits variables, units for standardized limits variables, sampling information variables, additional limits information, actual DMR reports for each limit, effluent violations, and variables relating to technical aspects of reporting.
Part 10 contains 26 enforcement actions variables.
Part 11 contains 24 Occupational Safety and Health Administration inspection variables.
Part 12 contains 39 administrative and judicial court case characteristics variables.
Part 13 contains 21 court case settlement characteristics variables.
Part 14 contains 9 criminal case characteristics variables.
Part 15 contains 95 variables created for final report analyses by year.
Part 16 contains 46 variables created for final report analyses by quarter.
Part 17 contains 157 variables including pro-social variables with security/over-compliance intentions, noncompliance variables with technical/significant noncompliance intentions, vignette characteristics variables, other variables derived from survey questions, environmental norms variables, and demographic characteristics variables. Specifically, Part 17 contains variables relating to the specific dimensions that made up each hypothetical scenario including corporate culture, environmental constraints, EPA discharge classification, facility condition, facility location, facility ownership, firm competitive position, firm environmental marketing, firm environmental record, firm EPA volunteer status, firm economic status/subsidiary economic status, firm ownership, internal compliance operation, internal compliance structure, locus of control, management location, managerial ethics, public awareness, and subsidiary status.
Not applicable (Part 1-Part 16). For the Vignette Data (Part 17), it was impossible to calculate a response rate because the research team did not know how many potential respondents were sent the survey information. Informally, however, the principal investigators were told that participation was high at one firm. According to a contact at the participating pulp and paper company, almost all potential respondents who received the survey participated in the study.
Presence of Common Scales:
Several Likert-type scales were used in the Vignette Data (Part 17).
Extent of Processing: ICPSR data undergo a confidentiality review and are altered when necessary to limit the risk of
disclosure. ICPSR also routinely creates ready-to-go data files along with setups in the major
statistical software formats as well as standard codebooks to accompany the data. In addition to
these procedures, ICPSR performed the following processing steps for this data collection:
Created variable labels and/or value labels.
Standardized missing values.
Checked for undocumented or out-of-range codes.