Disproportionality in Design and Impact: Ensuring Equity When Developing and Implementing Environmental Policies

PROBLEM

Environmental policy is focused on protecting the environment, whether that is in the form of proactive protection, or as is often the case, as a reaction to environmental harm. Any evaluation on the effectiveness of an environmental policy needs to be concerned with how the goals achieved under the policy are distributed. For example, if a policy is ameliorating harm, then it should be concerned with both the outcome of that amelioration (has the ultimate goal been met) and the process of achieving that outcome (what occurs as the policy is implemented in terms of altering environmental harms). Does everyone equally participate in the benefits achieved from improved environmental quality, or are those benefits skewed toward some recipients more than others? And is that “participation in benefits” measured not only by the outcomes achieved, but also by impacts that occur while the outcome is being achieved. This question is not only at the heart of environmental policy (whether government intervention on behalf of environmental quality is necessary), but it is also at the heart of environmental justice (whether environmental policy objectives are being equitably achieved). When there is variance in how environmental goods, or environmental harm, is distributed, society should strive to understand the causes and effects.

The purpose of this brief is to highlight the importance of looking at the distributional effects of environmental policies as they are being implemented to achieve their ultimate goal(s). It focuses on recent study that attempts to measure disproportionate effects during the implementation of a policy, rather than looking solely at the effects after the policy has been fully implemented. Much of the evaluative literature on environmental policy tends to focus on effects and impacts after policy implementation. ¹ But if we are to fully consider the importance of equity as we think about large-scale policy interventions for issues such as climate change, which carry embedded disproportionate impacts, then we likely need to more fully measure policy impacts at all phases: planning, design, implementation, and evaluation. ² One way of doing this is by examining disproportionality impacts of discrete environmental policies at a smaller scale. What follows is an overview of recent research into sulfur distributions during the implementation of the Federal Clean Air Act Amendments of 1990 (sulfur cap-and-trade program), and linking that work to issues of equity and resilience for climate change policy development.

Begun in 1990, the sulfur cap-and-trade program under the Federal Clean Air Act relied on quasi-market forces, rather than traditional command-and-control regulation, to lower sulfur emission from commercial electric utility generators. The theory was that, as implemented, incremental reductions in total sulfur output allowance, along with the ability to trade sulfur allowances between and by individual emitters, the program would result in a more efficient process of reducing overall sulfur emissions. It would happen faster than command and control, and it would be cheaper in terms of both compliance and enforcement costs. And, in fact, most studies show the program has led to substantial reductions in sulfur emissions at a rate faster than anticipated and with an economic efficiency estimated to be between 15% and 90% cheaper than the costs of traditional command-and-control regulation. ³

Although the evidence is clear the Clean Air Act Amendments of 1990 successfully reduced overall sulfur emissions, only recently has work been done to determine distributional effects during its implementation. Many large-scale environmental policies look at ultimate outcomes, for example, the overall reductions in sulfur emission as a means of achieving air quality goals under a large-scale policy regime such as the Clean Air Act. But more recent study attempts to distinguish between overall objectives and findings (reductions in total sulfur emissions), and seeks to understand if the process of the policy implementation results in new distribution patterns that are disproportionate in their effect. Much of this study has focused on how already-allowed pollution is distributed: how existing policy results in differential exposure after implementation. ⁴ For example, how industrial polluters in Saint Louis, Missouri are located in marginalized communities. Or, when examining the largest corporate polluters in the United States postregulation, finding the top 10% expose marginalized communities to disproportionate impacts. The sulfur reduction cap-and-trade program allows us to examine how a new policy intervention, successful in the aggregate on multiple measures, might itself result in disproportionate outcomes. Rather than measuring differential pollution exposure, this study focuses on measuring changes in pollution output as the policy is being implemented.

FINDINGS

I was part of a recent study that attempted to disaggregate sulfur emissions in the Commonwealth of Massachusetts at the county level, and using the trend of sulfur emission reductions between counties, identify differences in the trends between counties. ⁵ Counties were distinguished between marginalized and nonmarginalized, defining marginalized communities as those with higher proportions of minorities, lower educational attainment, and lower socioeconomic status. ⁶ Our findings showed that although the cap-and-trade program ultimately lowered sulfur emissions across the Commonwealth between 1990 and 2014 (years of available data under analysis), it was clear that counties meeting the operational definition of marginalized carried a disproportionate load of sulfur emission as total emissions dropped. ⁷ In essence, although sulfur emissions were lowering in nonmarginalized counties, they were increasing in marginalized counties. Overall emissions were lowering over time, but they did so more rapidly in nonmarginalized counties, whereas lowering (and sometimes increasing slightly) in marginalized counties. By 2014, all counties experienced significant reductions in sulfur emissions, but the rate of reductions differed significantly between marginalized and nonmarginalized counties.

This finding is in-line with larger-scale research showing similar national programs resulting in hotspots. ⁸ In essence, aspects of the program, particularly the trading part of cap and trade, can incentivize the purchase of existing pollution credits over pollution-reducing technology upgrades. Older utility generators tend to reside in marginalized communities, face higher costs to upgrade technology, and generally receive less public demand for change than other communities. All of these factors, and others, can lead to inequitable pollution distribution even when the policy ultimately results in lower pollution levels across communities.

STAKEHOLDERS OF CONCERN

In coining the term disproportionality, Freudenburg was attempting to describe society's uneven impact on the environment, both within and between groups. Some societies pollute more than others, as do groups within society, and individuals within groups. ⁹ As noted by Collins, Munoz, and JaJa, disproportionality can be viewed from two dimensions: through the production of environmental harm and through the effects (exposure) of that production. ¹⁰ There are stakeholders of concern in both dimensions, and likely often a good deal of crossover where specific stakeholders are impacted from production disproportionality and exposure disproportionality. For example, both the intensive pollution producing industrial activities and their effects often occur in the same marginalized communities. ¹¹ But sometimes the effects of environmental harm can be decoupled from the activity. Climate change is a prime example at least at the international level, where carbon-intensive activities occur in technologically advanced and economically wealthy nations, but the effects of the activity are most acutely felt in nations that do not engage in carbon-intensive activities and have limited economic resources to respond to the harm. ¹²

Much of the environmental justice literature has focused on the effects of environmental harm. ¹³ But government should also be interested in those impacted by the design of policy initiatives, particularly when they have the capacity to cause environmental harm even as they are intended to ameliorate that harm. Focusing on the design and implementation of environmental policies, such as the 1990 Clean Air Act Amendments, is helpful when dissecting between overall goals and what happens in the process of achieving those goals. If disproportionality occurs during the mitigation of environmental harm, it should be identified and understood. This way not only the benefits, but also the costs, can be fully accounted for and understood.

As Collins, Munoz, and JaJa identified in their study of pollution production, their review of >16,000 industrial polluters across the United States showed what was expected: they existed in or near “poor and nonwhite,” that is, marginalized communities. But they also found that observed disproportionalities were more pronounced when they looked at a subset—fewer than 10%—of the 16,000 industrial polluters who generated the majority of the exposure risk. ¹⁴ It is these individuals and groups that are of concern because they are often the recipients of environmental harm that goes unnoticed and misunderstood. As noted in our study of sulfur distributions during the cap-and-trade implementation years of 1990 thru 2014, we see that although emissions levels dropped significantly everywhere across the state by 2014, the distribution of sulfur emissions was uneven between wealthy and poor communities. Sulfur emissions dropped most quickly in wealthy communities but lagged, and even incrementally increased, in marginalized communities. ¹⁵ This suggests the burden of not only environmental harm production (levels of pollution before regulation) is borne by marginalized communities, but so too is the burden of pollution during environmental regulation. It is this double burden that needs to be more readily understood by policymakers, especially when polluter choice is part of the policy intervention design.

POLICY ALTERNATIVES AND RECOMMENDATIONS

Cap and trade has proven to be an effective mechanism to achieve environmental goals in a variety of contexts, ¹⁶ and certainly with respect to sulfur emissions in the United States. But success needs to be operationally defined from numerous vantage points. Cost savings in enforcement and providing leeway in how private firms choose to meet overall goals (buy-in of the regulated) are certainly important considerations from a policy design standpoint. But impacts in design and implementation must also be considered. It is not enough to see whether a particular goal has been met, but we must also strive to understand how the goal has been met: the dynamics of policy in action. For example, what groups of people are taking part in the design of environmental policy solutions? And what standards are being used, if any, to judge equity issues in the carrying out of environmental goals? Who is at the table, and how are the metrics of success being operationally defined and measured? The idea is to consider all attendant aspects of policy design and implementation, making policies “resilient” through a lens of equity.

Today, much of climate policy is driven not only by the environmental harm and human costs observed today, but also by the anticipated toll from cumulative impacts years and decades from today. That anticipation requires a precautionary approach to policy design and analysis. This means that goals have to be assessed incrementally as well as holistically, and from multiple vantage points. As Collins, Munoz, and Jaja point out, highlighting disproportionality as a metric to examine in both the design and implementation of environmental policies is a key to focusing on potential inequities that arise in both the unregulated and regulated environments. The analysis of sulfur emissions in Massachusetts during the implementation of the cap-and-trade program would undoubtedly have shown evidence of disproportionality: substantial reductions in certain areas, whereas maintaining and even increasing emissions in other areas. A closer inspection of the disproportionality would reveal communities with higher concentrations of sulfur emissions presented characteristics of marginalization. Such evidence would allow for a more nuanced approach to the program to ensure environmental harm was not being shifted in inequitable ways while the policy was progressing to its ultimate goal of overall lower sulfur emissions.

Comparing global climate change policy with statewide sulfur emissions patterns under a national cap-and-trade program is not an “apples-to-apples” comparison. For example, there are significant disparities between a relatively straightforward national sulfur reduction program and a dynamic and multifaceted multinational program to lower carbon emissions. The point attempting to be made here is that disparities can occur in relatively straightforward policy goals undertaken on a relatively small scale. We can assume (as evidence has shown) that proposed global solutions to climate change will have significant disparities in how they are carried out: the ultimate goal will be clear, but the road to achieving that goal will be fraught with potential and realized inequities. They already exist. ¹⁷

In many ways our desires to develop meaningful climate change policy today exemplifies the need to place disproportionality at the forefront of our policymaking. There are at least two reasons that warrant mention. First, climate change impacts are aggregated over time; the full effects and impacts are not felt with each incremental discharge of carbon, but only after emissions have culminated in a consistently warmer planet. Billions of individual actions today culminate in a disproportionate impact tomorrow. Without considering that temporal disproportionality, climate change policy cannot occur proactively. Second, both the causes and effects of climate change are, currently, disproportionate. Developed nations produce the majority of the greenhouse gas emissions responsible for climate change. But the effects are not limited to these developed nations. ¹⁸ The effects are dispersed over the entire globe, where areas such as the Arctic or small island states are disproportionately impacted.

It is clear that environmental policy, in general, can benefit from a more nuanced approach. Not only should ultimate policy goals be considered (lowering overall sulfur or carbon emissions), but universal goals of equity and fairness should also imbue each new policy directive. Adding a disproportionality component to environmental policies will enhance them at both the creation and implementation stage. When being designed, proposed policies will ensure that methods include a consideration of impact: whether, as designed, the policy might lead to unequal outcomes. And when implemented, the policy will be evaluated during implementation to ensure that it is not shifting burdens and obligations of environmental harm in the near term. Such a requirement will help to ensure that new policies are equitable at all phases of policy evolution.