Note: A new (March 2021) CTC page, Carbon Pricing and Environmental Justice, summarizes this post and its Dec. 2020 follow-on, Dogmatism on Carbon Pricing Mustn’t Derail Climate Progress, and embeds them in a larger narrative about the environmental justice movement’s increasing turn against carbon pricing.
Could a new paper by two economists at the University of California at Santa Barbara upend the belief that carbon pricing cannot further environmental justice?
Drawing on millions of simulations of pollution trails from industrial smokestacks across California, UCSB PhD candidate Danae Hernandez-Cortes and Associate Prof. Kyle C. Meng have amassed convincing evidence that the state’s carbon cap-and-trade program has lessened the disproportionate dumping of pollution on disadvantaged communities.
So ingrained is this inequity that a decade ago, ground-level concentrations of pathogenic carbon “co-pollutants” — toxic particles and gaseous oxides — emanating from those smokestacks were three to four times higher in disadvantaged communities than elsewhere, according to the UCSB analysis.
The persistence of this and other stark instances of environmental injustice has nourished a conviction among many climate advocates that carbon pricing, whether rendered indirectly via emission permits or directly through carbon taxes, cannot mitigate pollution disparities affecting historically-burdened environmental-justice communities. As this belief has proliferated, some advocates have turned against carbon pricing measures as a way to reduce fossil fuel use and greenhouse gas emissions.
Now comes the UCSB economists’ finding that during the first five years of California’s carbon cap-and-trade program, what they call the state’s “EJ (environmental justice) gap” shrank considerably. From 2012 to 2017, the pollution disparity between disadvantaged and other communities fell an estimated 30 percent for particulates, 21 percent for nitrogen oxides and 24 percent for sulfur oxides.
Importantly, Hernandez-Cortes and Meng computed these shrinkages relative to a control group: industrial facilities whose smaller sizes exempted them from the cap-and-trade program. In this way, the researchers were able to conclude that “while the EJ gap was widening prior to 2013,” when cap-and-trade began, “it has since fallen by 21-30% across pollutants due to the policy” (emphasis added).
Scope of the UCSB paper
Hernandez-Cortes and Meng posted their paper, “Do Environmental Market Cause Environmental Injustice? Evidence from California’s Carbon Market,” in May on the web site of the prestigious National Bureau of Economic Research as a working paper — a preliminary form intended to make cutting-edge research available ahead of final publication. (Although the NBER posting is behind a paywall, it is available gratis at Ms. Hernandez-Cortes’ and Prof. Meng’s web sites, here and here.)
Because the EJ gap worsened considerably from 2008 to 2012, the post-2012 narrowing of the gap that the authors ascribe to the cap-and-trade program has only yielded a modest net improvement from the 2008 status quo. Nevertheless, the paper’s linkage of the 2012-2017 shrinkage of the EJ gap in California to a statewide program pricing carbon emissions is an apparent first.
Surprisingly, the Hernandez-Cortes – Meng paper has not attracted notice among climate advocates or researchers. It is not a casual read. It employs the techniques, and language, of mathematical modeling and statistics. That, along with its reliance on a “difference-in-difference research design” to sift the impacts of the cap-and-trade program from other ongoing phenomena, made it impracticable for the authors to express some of their findings as simple averages.
Moreover, the sheer size of their research sample — more than 300 power generators, refineries, cement plants, incinerators and factories that are California’s largest “stationary” carbon emitters — prevented the authors from tracing the specific pathways by which the cap-and-trade program elicited changes in individual plant designs and operations that brought about the emission reductions.
Nevertheless, their findings are unequivocal: the gaps in pollution burdens inflicted upon EJ communities vis-a-vis other locales narrowed conclusively — “statistically significantly” in academic parlance — for all four examined pollutants. The stigma against carbon-pricing policies as tools of environmental injustice is ripe for re-examination.
Breaking the mold
Emissions pricing has never gained much of a foothold among environmental justice campaigners. While mainstream economists traditionally view pricing of pollution as a necessary and benign market corrective, many activists recoil from its implicit acquiescence to capitalist means of exchange. To some, proposals to commodify pollution may serve as an unwelcome reminder of the central role of stolen labor and stolen land in the historical amassing of wealth by white planters and merchants. Emissions pricing can also appear antithetical to the Indigenous ethos that to monetize Nature is to desecrate it.
Added to this are the scars visited by an earlier California experiment with cap-and-trade — the RECLAIM (Regional Clean Air Incentives Market) program aimed at cutting emissions of smog-causing nitrogen oxides. By some accounts, the program’s softening of emission caps following the Enron-instigated power shortages in 2000-2001 greenlighted Chevron Corp. to bulk up on emission permits and expand operations at its sprawling Richmond petroleum refinery north of San Francisco rather than invest in costlier clean-up technology. The resulting concentration of pollutants in surrounding Black and brown neighborhoods erupted into a cause célèbre of environmental injustice, particularly in 2012 when a distillation unit at the refinery exploded, inundating fence-line neighborhoods in noxious plumes and reportedly sending 15,000 area residents to hospitals.
Initially, EJ antipathy to emissions pricing focused on cap-and-trade systems, not only because of the Chevron disaster but also because “pollution markets” were enabling blatant Wall Street profiteering. Moreover, early emissions trading schemes were riven by “offsets” that let polluters substitute offshore cuts for local action. Of late, distrust of pollution pricing has come to afflict even the straightforward taxing of carbon emissions.
“Carbon taxes will always be low, will always be evaded, do not cut pollution to the degree needed, and are greenwash.” So declared Carbon Pricing: A Critical Perspective for Community Resistance, a 2017 manifesto of the Indigenous Environmental Network and the Environmental Justice Alliance. A year later, as retiring California Governor Jerry Brown was convening a blue-chip “Global Climate Action Summit” in San Francisco, Indigenous and EJ demonstrators protesting Brown’s refusal to rein in petroleum fracking marched under a banner proclaiming that “Carbon pricing is colonialism.”
These harsh expressions could be viewed as an outgrowth of the failure of laissez-faire capitalism to contain carbon pollution, or, at a minimum, to alleviate its inequitable burdens on disadvantaged communities and households. Some of the stiffening line against carbon pricing may also be traced to the general radicalization of resistance movements during the Trump presidency, which has seen communities of color increasingly besieged by environmental assaults, systemic economic inequality, violent policing and Covid-19.
Concurrently, antipathy to carbon pricing was gaining further traction from an emerging body of research into the incidence of pollution from California’s carbon cap-and-trade law known as AB 32.
In 2015, a team of academics led by Prof. Lara Cushing, a well-known health-sciences scholar formerly at San Francisco State University and U-C Berkeley who is now at UCLA, began posting findings from an ambitious research project on AB 32. Their investigations culminated in a widely cited paper, Carbon trading, co-pollutants, and environmental equity: Evidence from California’s cap-and-trade program (2011–2015), which was published in the journal PLOS Medicine in 2018.
The Cushing team concluded that a majority (52 percent) of California “regulated facilities” — essentially, the same 300 or more factories and power plants that Hernandez-Cortes and Meng would tackle in their paper — increased rather than curbed their emissions of greenhouse gases following implementation of the cap-and-trade program. Moreover, the resulting increases in co-pollutants were disproportionately concentrated in communities with “higher proportions of people of color and poor, less educated, and linguistically isolated residents,” according to the PLOS paper.
These findings left a strong imprint on discourse about carbon pricing and environmental justice. Indeed, Prof. Hernandez-Cortes of UCSB mentioned in an email that she and Prof. Meng launched their research project after reading the original Cushing et al. working paper in 2015. Because of the salience of the issue as well as the “natural experiment” afforded by California’s carbon cap-and-trade program — the world’s second largest, after the European Union’s Emissions Trading System (ETS) — the UCSB scholars devised a research methodology to build on the Cushing team’s work in three respects.
How the UCSB paper improves on the Cushing analysis of AB-32
First, Hernandez-Cortes and Meng established a control group of 440 lesser California emitters that were not regulated by the cap-and-trade program; this “difference-in-difference research design” enabled them to weed out macroeconomic and other factors to discern the cap-and-trade program’s specific effects on emissions from the 306 larger, “covered” facilities. (Strong increases in economic activity across the state — California’s economy grew more than 17 percent during the five-year 2010-2015 period covered in the Cushing paper — also made it valuable to include a control group.)
Second, Hernandez-Cortes and Meng sought to track where the carbon co-pollutants deposit after they have been emitted. The work by Cushing et al. implicitly assumed that only households within a 2.5-mile ring of a source are exposed to its emissions. Yet pollutants can and do travel scores and even hundreds of kilometers, especially from very large emitters, whose plumes typically exhaust through tall smokestacks and at high velocity.
To “explicitly model where pollution goes,” Hernandez-Cortes and Meng fed estimates of each facility’s emissions into an atmospheric transport model that traces pollutants’ paths once they have gone up and out the stack. “This was a highly computationally intensive process,” they noted in an email, “involving modeling over 11 million trajectories [and] taking over a week’s worth of high-performance cluster computing time.”
Third, whereas the Cushing findings emphasized whether each facility’s emissions rose or fell from 2012 to 2017 — an approach that flattened broad ranges of data into simple yes-or-no form — the UCSB economists calculated facilities’ emission quantities along with their deposition. This approach steered clear of potential distortions from assigning equal weight to the 300-plus facilities.
California’s Shrinking EJ Gap, Quantified
The table below summarizes the key findings of from the Hernandez-Cortes – Meng working paper.
The columns in yellow display ground-level concentrations of particulate and gaseous carbon co-pollutants from California’s 306 largest carbon emitters in 2008, expressed in the standard pollution-exposure metric of micrograms of pollution suspended in air. Perhaps even more striking than the “deltas” (numerical differences) between pollution concentrations in environmental justice vs. other communities are the ratios. Pollution levels in the EJ or disadvantaged communities were three to four times as high as in other areas. While these disparities were probably attenuated by pollution from mobile sources, which tend to be evenly distributed but were not included in the cap-and-trade program or the Hernandez-Cortes – Meng analysis, the differences are stark nonetheless.
The next column, in green, shows that by 2012 each pollutant’s “EJ gap” — the excess pollution deposited on disadvantaged communities relative to other California locales — had worsened even from their shocking 2008 base. Thereafter, however, coinciding with the onset of the cap-and-trade program, the gap narrowed significantly, as shown in the final two columns.
Consider the Hernandez-Cortes – Meng findings for PM2.5, in the third row of the table. PM2.5, designating fine particulate matter that lodges deep in the lungs and is implicated in illness and death from heart and lung diseases and strokes, is considered the most deadly air pollutant associated with industrial processes that emit climate-damaging carbon dioxide.
Over the four years starting in 2008, the UCSB economists found, the gap in levels of PM2.5 between EJ and other localities widened, from 3.0 micrograms per cubic meter to 3.8. However, beginning in 2013, the EJ gap for fine particulates contracted, falling to 2.7 µg/m3 in 2017 (the last year for which pollution data was available). That was 30 percent less than the 2012 gap of 3.8, and less than the baseline figure of 3.0 µg/m3 as well.
The state’s cap-and-trade program had a similarly beneficial impact on oxides of nitrogen, or NOx. This dangerous pollutant is the key constituent (along with volatile organic compounds) of the photochemical smog that since the late 1950s has infamously blanketed skies and seared eyes and lungs across much of California, with disadvantaged communities bearing more of the impact. NOx is also a precursor of “secondary” formation of deadly particulates. As shown in the table’s top row, from 2012 to 2017 the EJ gap for NOx shrank to 4.8 µg/m3, a level 21 percent below the 2012 figure.
Over the same five years, the environmental justice gap shrank by 24 percent for sulfur oxides and 30 percent for larger particulate matter, known as PM10, matching the decline in PM2.5.
Implications of California’s narrowed EJ gap
For all of its depth and rigor — or perhaps on account of it — the Hernandez-Cortes and Meng paper does not seek to explain why California’s cap-and-trade program should have shrunk the EJ gap in carbon co-pollution. The marked reductions they reported in their paper for all four pollutants are perhaps surprising, given that California’s cap-and-trade program was designed solely to engender the greatest pollution declines from emission sources that could most easily be abated at the lowest cost, without regard to location or incidence.
If, and only if, low-abatement-cost sources already are concentrated in EJ communities, would those populations be expected to enjoy above-average percentage rates of pollution reduction as a result of carbon pricing. Or so most economists reason. (Although AB 32 contains provisions for allocating a portion of the carbon permit revenues to disadvantaged communities, these would not be expected to induce facility owners and operators to concentrate emission reductions in those communities.)
For now, the UCSB economists’ finding that California’s comprehensive statewide cap-and-trade program has narrowed the EJ gap should be considered a fortuitous result. It should not be used to argue that carbon pricing policies, whether delivered via cap-and-trade or carbon taxes, will necessarily shrink EJ gaps elsewhere.
This caveat should not be taken too far, however. Even if disadvantaged communities shouldn’t expect carbon pricing to bestow disproportionate percentage reductions in emissions, pricing will still tend to deliver greater quantity reductions to those communities — and to any communities that suffer from excessive burdens of pollutants that the pricing measures address. This follows mathematically from the fact that a given percentage reduction applied to a larger amount translates to a greater quantity reduction than the same percentage reduction applied to a smaller amount.
A schematic example
To grasp this, consider Community A that suffers 100 daily doses of pollution while Community B suffers 50. A 20 percent reduction applied to both will cut the burden to A by 20 while cutting the burden to B by just 10. If the “before” EJ gap is expressed as a percent, with A suffering a 100 percent greater burden relative to B, then the percentage gap is unaffected by the across-the-board (20 percent) reduction, since the new burdens, which are 80 for A and 40 for B, still amount to a 100 percent gap. Yet it is also true that A’s reduction of 20 is twice B’s reduction of 10 — and that the EJ gap expressed as an amount has shrunk from 50 to 40.
This schematic suggests that community-neutral pricing policies stand to additionally benefit disadvantaged communities in health terms, even if the relative bias of disproportionate burdens on those communities remains untouched.
This is not to argue for community-neutral policies in the environmental realm or any other arena. Nevertheless, it is important to point out that such policies need not replicate the felicitous outcome thus far from California’s cap-and-trade program in order to benefit disadvantaged communities in an absolute sense, and, further, to confer larger benefits on them than on relatively privileged areas.