Date of Award

Spring 3-29-2021

Semester of Degree


Document Type

Open Access Dissertation

Degree Name

Ph.D. in Environmental Science


Environmental Science, Division of

Major Professor

Mary Collins

Steering Committee Member

David Sonnenfeld

Steering Committee Member

Elizabeth Vidon


Government contract awardees in United States (US) industries produce toxic releases in addition to producing public goods and services. Industries that the government buys many products from and that have high environmental footprint are where procurement decisions have the most impact. This dissertation examines what factors predict variation in contractor facility environmental performance indicated by reported pollution control technology and environmental releases to inform guidance for revising procurement policy. The research question is: How are government contractors that mange toxic materials different in their environmental management behavior than noncontractors? Data from the US Environmental Protection Agency’s Toxics Release Inventory, a public dataset of chemical management information for industry, are combined with US Federal Government spending data on agency contracts and procurement to see how pollution by awardees has changed over time. Three factors explaining variation in environmental performance are explored here. The first is payment method to identify if facilities with contracts incentivizing cost reduction have different environmental performance than other contractor facilities. The second is the Federal agency that awards the contract with interest in how defense contractors are different due to exemptions from environmental purchasing standards. The last factor is variation in competition between bidding firms, which could lead to different outcomes related to management of chemicals by the facilities. This dissertation contributes to the socio-environmental frameworks of ecological modernization, environmental disproportionality, and the Porter Hypothesis on environment and competition. Quantitative models are estimated for each of the three factors of interest using both linear mixed models and Bayesian Hierarchical models. Findings support expanded use of incentive type contracts, while also identifying that defense spending as being linked to the majority of contractor pollution. Findings also show that competition intensity leads to more pollution when the chemical is not a component in the manufacturing process.