Environmental regulations and business decisions

Regulatory differences and looking under the lamppost

One under-appreciated aspect of environmental regulation in developed countries is that regulatory approaches vary considerably for different pollution media. In the US, air pollution follows a top-down approach, with the Environmental Protection Agency setting overall national ambient air quality standards for each pollutant that must be met in every county in every state. State agencies must then develop state implementation plans that impose stricter regulations in “non-attainment” counties—counties whose pollution levels exceed the air quality standards. A similar approach is taken in the EU, where the European Commission sets air quality standards and countries must develop plans to reduce pollution in any region that fails to meet the standards.

In contrast, water pollution regulation follows a bottom-up approach in the US but a top-down approach in the EU. In the US, each plant that discharges into navigable waters requires a permit for those discharges. State regulators write most water permits, which require monthly reporting of discharges. The permits tend to follow the Environmental Protection Agency’s industry-specific guidelines but can be stricter when water conditions warrant. EU water regulation is more like air regulation, with countries required to develop a management plan for each river basin district. Concern about toxic materials in the US led to a requirement that manufacturing plants provide annual public reports in the Toxic Release Inventory.

Market-based regulatory approaches such as tradable permits have long been favored by economists for providing an efficient allocation of pollution reduction by equalizing the marginal cost of reductions across polluters. Market-based approaches also provide a measure of regulatory intensity, since permit prices reflect the marginal cost of abating pollution. In the US, sulfur allowance trading under the 1990 Clean Air Act Amendments was associated with lower than expected compliance costs and widely considered a success. In recent years, tradable permits have been proposed for reducing greenhouse gas emissions, most notably in the EU Emissions Trading System.

Empirical research on the economic impact of environmental regulation has followed the tradition of “looking under the lamppost”—focusing on the most readily available data. Data on US environmental regulation far exceed what are available for other countries, helping explain why so much of this research has focused on the US, despite several limitations.

County location is easily connected to economic data, and the non-attainment (“dirty air”) counties that should face stricter regulation are easily identified. Thus, there are many studies on US air regulations at the county level. Regulatory stringency for water pollution is set on a plant-by-plant basis, which makes data collection and analysis much more difficult, so there are far fewer studies on water pollution. The Toxic Release Inventory provides easily accessible annual data for tens of thousands of manufacturing plants—but there is often no clear connection between the reported releases and any particular regulatory program.

This paper focuses on environmental regulations that affect firms’ production processes. Regulations that mandate changes in product characteristics, such as increased fuel economy requirements for motor vehicles, reduced volatility in paints, or bans on specific pesticides, certainly affect firms’ decisions, but they are not addressed here.

Pollution abatement costs and productivity impacts

Three basic approaches are used to measure the costs of environmental regulation: surveys, engineering studies, and econometric analyses. For the US manufacturing sector, the Pollution Abatement Costs and Expenditures (PACE) survey, conducted annually by the Census Bureau between 1973 and 1994 and only sporadically after 1994 (in 1999 and 2005), provides detailed information on air, water, and solid waste pollution abatement spending for capital investment and operating costs. Engineering studies forecast the cost of new environmental standards based on the cost of purchasing and operating the equipment needed for compliance. Econometric analysis can measure the costs of existing environmental regulation, given data on production costs and (crucially) measures of the differences across observations or over time in the extent of regulation faced by the plants, firms, industries, or economies being studied. Each approach has its disadvantages. Surveys rely on respondents to allocate some costs between abatement and production (for example, in the case of investment in new capital that is both cleaner and more productive). Engineering analyses rely on predictive methods and do not allow for improvements in abatement technology. And econometric analyses are subject to biases if there are errors in measurement or omitted key variables. This review focuses on econometric approaches, but it can be important to benchmark those results against the other approaches where possible.

Many of the earliest analyses of the impact of environmental regulations on business focused on their impact on productivity, defined as output per unit of input. Many different productivity measures are available, varying in the definition of inputs. Total factor productivity measures include all of a plant’s inputs (labor, capital, and materials); other measures such as labor productivity (output per worker hour) use a single input. Productivity growth happens when output increases faster than inputs. Spending money on pollution abatement increases measured inputs, but the resulting improvement in environmental quality is not counted as an output. Thus, a plant’s measured productivity declines. The reduction in productivity should be roughly equivalent to the share of pollution abatement costs in total costs. A firm whose pollution abatement costs are 2% of its total costs should have a 2% lower level of measured productivity, all else being equal. Econometric analyses can test whether the relationship between abatement costs and productivity is larger or smaller than this, implicitly testing whether reported abatement costs overstate or understate the true abatement costs.

The expansion of government regulation in the 1970s coincided with a slowdown in productivity growth in US manufacturing industries, raising speculation that regulation had contributed to the productivity slowdown. This apparent relationship prompted a series of research projects examining the impact of environmental regulation on productivity, many of them using the US Pollution Abatement and Control Expenditures (PACE) survey.

Early studies using economy-wide data found that pollution abatement costs explained a small part (perhaps one-fifth) of the productivity slowdown. Using industry-level data, one study found a greater productivity slowdown in the 1970s in industries with higher pollution abatement costs [2]. Using plant-level data, another study found large impacts on productivity (suggesting under-reported costs) for paper mills and steel mills but not for oil refineries, though these impacts were substantially smaller when plant-specific fixed-effects were included [3]. One study found less significant impacts of abatement costs on productivity using a different fixed-effects model [4]. Another study of paper mills found large impacts on productivity but with significant variability across plants—paper mills incorporating a pulping process were significantly more sensitive to abatement costs, while mills that had recently been renovated were significantly less sensitive [5]. More recently, a study created a county-year index of abatement costs from the PACE survey data and found no statistically significant impact of higher pollution abatement costs on productivity for the average manufacturing plant [6].

Studies of the impact of pollution abatement costs on productivity in other countries seem to find smaller impacts than in the US. An analysis of ten German industries found significant differences in productivity impacts across industries, but the estimated reductions in productivity growth were relatively small [7].

Productivity analyses using measures of regulation other than reported pollution abatement costs have found varying results. An examination of oil refineries in the Los Angeles area, which face extremely stringent air pollution regulations, found no evidence of falling productivity as a result of increased regulation [8]. (Recall that the oil industry also seemed less sensitive to abatement costs in [3].) A study focusing on US counties with changes in their pollution attainment status found that polluting plants had significantly lower productivity in years when their county was in non-attainment and was thus facing stricter regulations [9].

Location and investment decisions

In addition to examining the direct costs of environmental regulation on productivity, economists have studied how firms respond to regulatory-driven cost differences. Much of this research has examined the decision on where to locate a plant. When locating a new plant, a profit-maximizing firm should consider many factors besides factor prices and availability, such as regulatory compliance costs. If more stringent environmental regulations discourage firms from locating in a particular jurisdiction, politicians might weaken regulations to attract new plants. This possibility of a “race to the bottom” in environmental stringency is especially likely when the damages from a pollutant are widely dispersed (or can be directed to a neighboring jurisdiction by locating plants near the border). If a pollutant’s damages are more locally concentrated than the benefits from the new plant, the opposite effect could occur, with politicians showing a “race to the top” or “not in my back yard” bias by imposing increasingly stricter regulations.

One of the initial justifications in the US for establishing federal environmental standards was to avoid having individual states relax their environmental standards for competitive reasons. Amendments to the Clean Air Act, such as New Source Performance Standards and Prevention of Significant Deterioration, increased the regulatory stringency for new plants in areas with clean air (attainment counties), to avoid incentives for plants to move there from dirtier locations with stricter regulations (non-attainment counties). Studies show that Congressional voting on those programs was driven less by politicians’ environmental stance and more by the economic interests of their constituents: the principal beneficiaries of the new regulations were Eastern industrial firms and coal-mining firms that would otherwise have been at a competitive disadvantage compared with new plants in attainment counties.

Considerations of both data availability and the federal nature of US regulation, which allow for measurable differences in stringency across states and counties, account for the fact that most studies on the location decisions of plants have worked with US manufacturing sector data. The magnitude of regulatory cost differences within the US may be small compared with differences across countries, but potentially confounding differences in factor prices and availability are also likely to be smaller within the US.

One study looked at the number of new plants in ozone-generating manufacturing industries that were opened in each county and found large and statistically significant declines (26–45%) in counties that had failed to meet ozone standards [10]. Large multi-plant firms responded sooner than single-plant firms, suggesting that large firms managed the regulatory process better. Existing plants survived longer in non-attainment counties, while new plants were larger initially but grew more slowly, consistent with facing stricter regulations on plant expansions. The overall impact of the regulation was to spread ozone pollution geographically from initially high-ozone counties to initially low-ozone counties. Reducing peak ozone exposures in high-population areas may be beneficial, but increasing ozone elsewhere is clearly an unintended consequence of the regulation.

A similar outcome was found for a comparison across European countries of industry location in the early 1990s [11]. The share of each industry’s production that occurred in each country was related to that country’s stringency of environmental regulation and the industry’s pollution intensity, controlling for other factors that might affect location. As expected, less stringent countries were more attractive for dirtier industries, although the effect was significant only for the dirtiest industry studied, industrial chemicals.

Environmental regulation could also affect investment decisions by firms. Plants facing greater regulatory stringency may be less profitable, leading multi-plant firms to shift production (and investment) to their other plants and possibly to close the plant facing high regulation if compliance costs are especially high or if the industry is in decline and some plants need to close anyway. Regulations may also require substantial capital investments in abatement equipment, reducing the financing available for investments in production capital if the firm follows a rule-of-thumb for allocating capital expenditures across plants, such as a pre-set capital budget for each plant. Investment in production capital at paper mills was crowded out by capital investments in abatement equipment, with plants in a multi-plant company that had high abatement capital costs receiving significantly less in production capital investments [12].

Environmental regulation can also affect the choice of production technology. If different technologies have different environmental consequences, firms should choose cleaner (and presumably more expensive) technologies at plants facing more stringent regulation. Multiple production techniques are available for making paper, so plants can tailor their production technology decisions to the regulatory environment. New paper mills in states with stricter regulations tend to choose a low-pollution technology, while mills in states with a different mix of air and water pollution respond to the relative regulatory stringencies on air and water pollution in their choice of technology [12]. The mere existence of a regulation may change the set of production techniques available, providing incentives to develop cleaner methods. In some cases, “technology-forcing” regulations impose impossibly strict pollution reductions (which existing techniques cannot meet), anticipating that research will produce technologies that can accomplish those reductions before the standard takes effect.

Benefits from pollution reductions

While the costs associated with environmental regulation have been substantial, the best available information indicates that the overall benefits have far exceeded the costs. Benefits come in the form of reduced illness and death, as well as ecosystem services, such as improved visibility and better recreational water quality. Expressing these benefits in monetary terms requires a variety of assumptions. For example, the “value of a statistical life” calculation comes from workers’ choices between risky and safe jobs. The value of a statistical life is the increase in wages that a group of workers would require to accept a risky job with the probability of one death on average per year. If the probability of death on the job is one in 1,000, the value of a statistical life would be the extra pay to 1,000 workers. If each worker requires, say, $5,000 to accept the higher risk, the overall cost is $5 million ($5,000 × 1,000)—the value of a statistical life. If workers require only $500 each to accept that risk, then the value of a statistical life would be only $500,000. Calculating this value allows us to express lives saved (benefits) in monetary terms for comparison with costs. It seems clear that reductions in air pollution account for the vast bulk of these benefits, with most of the benefits coming from reductions in mortality as a result of lower ambient concentrations of fine particulates.

A systematic examination of the benefits and costs of US environmental regulation of both air and water pollution finds that the largest benefits came from reducing particulate emissions from industrial point sources, with benefits many times the costs [13]. A series of epidemiological studies over time have confirmed large impacts of particulates on mortality. The US Environmental Protection Agency’s most recent (2011) assessment of the benefits and costs associated with the Clean Air Act Amendments of 1990 estimates that by 2020 the reductions in particulate emissions will result in 230,000 fewer deaths annually, providing 85% of the predicted $2 trillion in annual benefits, far exceeding the predicted $65 billion in annual costs [1]. The gap between benefits and costs is so large in this case that it is unlikely to be reversed by any plausible adjustment in the measurement methodology, although the averted deaths tend to occur among older people with other health issues, while the value of a statistical life used to calculate the benefits comes from decisions by much younger people choosing between safe and risky jobs.

While most people in developed countries have relatively low exposure to air and water pollution, people in many developing countries face much higher levels, implying that the health benefits of pollution reductions are likely to be greater. The World Health Organization estimates that air pollution exposure leads to 3.7 million premature deaths, mostly in developing countries. Estimates of annual premature deaths from air pollution in China alone range from 350,000 to 1.2 million. Similarly, the health benefits from reducing water pollution are limited in developed countries, which already have high drinking water quality and adequate sanitation, but could be much larger in developing countries, where access to clean drinking water and sanitation is poor.

While the evidence is strong that overall environmental regulations have benefits that exceed costs, this conclusion does not necessarily apply to individual regulations. As noted, the vast majority of pollution abatement benefits in the US are associated with reductions in fine particulates. A systematic examination of the benefits and costs of US environmental regulation concluded that air pollution controls on mobile sources (which generate little particulate pollution) probably have costs that exceed benefits, as do most water pollution controls, except in the case of a few highly polluted bodies of water [13]. Even a new regulation to reduce particulate pollution would not necessarily pass a benefit–cost test. The benefits must be evaluated on the margin, considering only the incremental reductions in particulates attributable to the new regulation, not the large benefits that have been achieved by earlier regulations.