Considering Environmental Justice in Natural Resource Damage Assessment
By Theodore D. Tomasi, Ph.D., Managing Principal, Business Director, Natural Resources and Enviromental Economics
January 27, 2022 was the one-year anniversary of President Biden’s announcing his Justice40 program as part of his approach to building environmental justice (EJ) broadly into decision-making. In December EPA announced it would spend $1 billion to fast track cleanups of contaminated sites in EJ communities. Looking back over the past year, activists bemoaned that “…we are still waiting to see results on the ground”, understandable, since, as CEQ Chair Brenda Mallory recently put it “…through the bureaucratic machinery of guidance documents and program reviews, we are turning that ship on a new course.” Complicating matters further, States have developed EJ policies for state-level enforcement.
Launching some smaller speedboats while the ship of state turns might help turn goals into actualities. I focus here on natural resource damages (NRDs), an element in the lifecycle of some contaminated sites where the potential to consider EJ is ripe but has not been much discussed. NRDs are monies collected from responsible parties by government trustee agencies to be used to restore natural resources harmed by unpermitted releases of contamination and to compensate the public for injury to such resources. The amount of NRDs equates to the cost of restoration to redress the harms. This article examines how EJ might be considered in NRD assessments (NRDAs) for contaminated sites and for oil spills.
Alternatives for Incorporating EJ in NRDA
Bringing EJ into NRDA could involve procedural changes. One approach could include targeted outreach to disadvantaged communities so their interests can be reflected in the NRDA process. This may or may not lead to different restoration outcomes, but almost certainly will alter expectations that they will. Another procedural change may be to fast track resolution for sites affecting overburdened communities, perhaps implying expedited NRDAs at such sites. Expedited approaches typically involve fewer studies and more uncertainty. As a result, more conservative assumptions often are embodied in the analysis to address these uncertainties, leading to higher damage claims than otherwise. The timely resolution hoped for may not ensue if the shakier basis for “excess” restoration leads to challenges by potentially responsible parties and protracted resolution of disputes.
EJ might be more substantively considered in restoration decisions. For example, selection of restoration projects that benefit disadvantaged communities may be artfully woven into the fabric of decision-making, without changing NRDA methods. This seems a likely path, but comes with some difficulties. One is a lack of transparency. Another is that a mismatch is created between the technical methods used to determine restoration scale and the factors that actually drive restoration decisions. This mismatch leads to biased estimation of the amount of restoration needed. To avoid this bias, EJ would need to be included directly into the technical analysis, as discussed below.
However done, bringing EJ into NRDA raises policy-level issues. EJ creates tensions with other criteria typically used in NRDAs to choose restoration actions. Consider compensating the public for an injury to a resource used for recreation. A foundational criterion is that the scale of restoration provides benefits that compensate the public for lost recreation trips. This requires that the economic value to the public of trips created by restoration equals the value of trips lost due to injury. This is called, sensibly enough, value-to-value scaling of restoration. In current analyses, public values are the simple sums of individuals’ dollar losses and gains, giving equal weight to those dollars irrespective of how the persons to whom those dollars attach may differ in ethically relevant ways. A dollar from someone in a community deprived of recreation opportunities is treated the same as a dollar from a person flush with nearby recreation resources. If those who are rich generally will pay more for given change in recreation opportunities than those poor, the simple sum implicitly weights more highly those who are well-off. Thus, restoration deemed to compensate the public by the traditional analysis can be decidedly unfair.
Whether to consider incomes begs the question: what dimensions of advantage or disadvantage should be brought into decision-making? Should NRDA focus narrowly on the resource issue at hand, such as recreation resources, or a broader set of considerations? This relates to the nexus criterion, which limits the parties from moving “outside the NRDA lane” to achieve goals such as EJ. The nexus criterion would locate the restoration near the injury (people nexus) and restrict restored resources to those injured (resource nexus). If those harmed are relatively (resource) rich, a project satisfying the people nexus criterion may not exacerbate the baseline level of inequality, but certainly forgoes opportunities to reduce it. Moreover, foregoing strict resource nexus (e.g., effects on fishing need a fishing-related project) could open the door to other restoration actions with large social benefits (a leafy park in an urban heat island). How far to go down this road toward “out-of-kind” restoration is an open question. Another criterion, cost-effectiveness, would place the restoration where compensation is satisfied at least cost. If not, restoration is inefficient, and one could rearrange things to give more resources to everyone. This goal interacts with both EJ and nexus. If restoration to redress injury that occurred in a relatively advantaged community has low cost in areas that are resource disadvantaged, one might forego people nexus to achieve a solution that is both cost-effective and reduces inequality. On a practical level, deeming sub-basins of the same watershed as bearing nexus to one another could expand opportunities for finding cost-effective projects that benefit disadvantaged communities.
The take-home message from all this (which by no means exhausts the possible interactions among criteria) is that weaving EJ into NRDA demands high-level mastery of the art; without clear guidance, various less-than-transparent, ad-hoc practices could be the norm, frustrating stakeholders.
A Method for Including EJ in NRD Calculations
In light of this discussion, it would seem desirable to have a technical approach that identifies projects that compensate the public while balancing efficiency and equity in a transparent way. Such a model would provide a framework and language for measuring tradeoffs and evaluating alternative outcomes. Moreover, it would remove the mismatch between what goes into models for restoration scaling and considerations driving actual project selection. Economists have developed approaches with these features that can be applied in a relatively straightforward way in the recreation context. Rather than value-to-value restoration scaling, with its simple sum of trips values across people, it would use a weighted sum, where the weights reflect social preferences for reducing inequality. The approach might be called social value-to-social value scaling.
Some might note that EJ is neither mentioned as an aspect of injury quantification, nor identified as a criterion for project selection in the federal NRD laws or regulations. Therefore, they may assert, we are limited to having listening sessions and artful weaving in our toolkit. I see more leeway.
The regulations suggest methods for restoration scaling but not the details of their implementation. For example, in the OPA NRD regulations value-to-value scaling is described (43 CFR §990.53(d)(3)(i)), and individual values are defined in economic terms (43 CFR §990.30), but the regulations do not say how the individual values should be incorporated into value-to-value scaling. The traditional way to do this is for recreation injuries is via unweighted sums. This embodies the specific (if implicit) ethical assumption that each person’s value gets a weight of 1. This is nowhere mandated in NRD statutory or regulatory language. Adopting an alternative ethical assumption results in alternative weights. This approach has been discussed in published economic literature since the 1950s, and so has a firm basis.
Applying this approach in NRDA, however, would be new. The social value-to-social value approach would be based on a generalized summation with ethical weights assigned to persons via a parameter of the scaling model, call it z , that reflects the trustees’ aversion to inequality in the distribution of recreation opportunities across persons. In the sum of values across people, a larger z places higher relative weight on those who are relatively more resource disadvantaged. Restoration scaling would consist of finding the low-cost projects (addressing cost-effectiveness) with a generalized sum value (reflecting the degree of attention paid to EJ) at least as great as the generalized sum loss (which assures overall compensation of the public). The ethical parameter z could be varied as a sensitivity analysis to see how EJ interacts with other project selection criteria, combined with guidance from estimates of z taken from the economics literature.
How might this change outcomes? A higher value would be placed on a given increment to recreation resources where the baseline acreage is small than would be placed on the same increment to a baseline that is large (all else equal). A stylized example plays this out. Assume injury takes place near a community with x acres of recreation opportunities at baseline, results in a loss of 10 acres for one period, and has a dollar loss equal to $20. If the trustees place overriding importance on the nexus criterion, restoration is placed where the injury occurred. Assuming the restoration project provides the same value per acre as the injured acres ($2/acre) and costs $1 per acre, 10 acres of restoration compensates those harmed and NRDs equal $10 (the discount rate is zero).
Now suppose instead of following nexus, that same restoration project is placed in a different location with y acres at baseline, and that recreation acreage is the only way affected communities will be deemed to differ in the NRDA. It can be shown that the dollar value of a unit of restoration in the alternative location is worth (x/y), raised to the power z, times the value of injury. For example, if the trustees locate restoration in a community with y=x2 , and social aversion to inequality (z) equals 1, then an acre of restoration provides twice the value as it would in the injured region. Now 5 acres will compensate society for the injury, and NRDs are cut in half. Increasing z to 1.25 cuts NRDs to $4.2, while cutting z to 0.75 raises NRDs to $5.9. Empirical evidence places z in a range between 0.5 and 3 or more.
This example demonstrates that to include EJ formally in NRDAs for recreation, the basic structure of traditional analysis needs to change very little and the incremental data needs are relatively modest. The innovation comes from the weight given to EJ, as summarized by a single variable parameter, about which there is at least indirect evidence based on observable behaviors.
The principles exemplified by the above model can readily be extended to more realistic settings. Things get considerably more complex if differences across people other than availability of resources are considered. For example, a new park in an urban heat island cools its surroundings, providing benefits that correlate with income (the ability to buy an air conditioner) and other stressors. This relates EJ in restoration scaling to cumulative risk assessment. Although multiple health stressors are not directly related to NRDA injury, they may be, with sufficient creativity, part of restoration considerations, both in terms of “knobs to turn” to achieve compensation (however, recall the nexus discussion) and the collateral social benefits of certain projects relative to others.
When recreation resources are injured, economic valuation is routinely undertaken. The above approach does not apply to other ecological services that in most NRDAs are not valued. Related methods have been developed for such services, however, that have very similar implications if not an identical formal structure.
NRDA practitioners are fond of saying that good restoration solves difficult cases. By extending that creativity towards consideration of EJ as outlined above, we may well find that good restoration can do a whole lot more.
 InsideEPA.com, December 17, 2021.
 Energy and Environment News, December 17, 2021.
 A compendium of state EJ programs is available at https://www.integral-corp.com/integral-issues-the-first-comprehensive-compendium-of-state-regulatory-activities-on-environmental-justice/
 T. Tomasi, 2021. Environmental Justice in Natural Resource Damage Assessment: Recreation Resources Integral White Paper 2021-1.
 T. Tomasi, 2021. Environmental Justice in Natural Resource Damage Assessment: Ecological Resources Integral White Paper 2021-1.