Pollution and GDP

Some of the earliest work in green accounting was done in the early 1970s by two Nobel Prize–winning economists, James Tobin and Bill Nordhaus. Their paper really laid out the research agenda; it provided estimates of the value of these different components, of what was missing. Although very insightful, their paper used the primitive empirical techniques that were available back in 1973.

As you mentioned, some of the most accomplished economists have realized and advocated for a long time that incorporating some of these measures is useful and would help us get a better sense of the economy at large, and accounting for things that are outside of markets. But to my knowledge, green accounting hasn't really been adopted widely across the United States or other nations. Why do you think we haven't seen these measures incorporated at large scales?

Taking the long view, by thinking back to that initial work by Nordhaus and Tobin in the ’70s, there were empirical obstacles. On the one hand, there are actually the approaches that economists use to derive monetary values for things like recreation experiences, or aesthetics, or existence values for different species. The measurement of some of these values, which are key ingredients in an extended set of green accounts, is frankly very hard.

On the other end of the spectrum, in terms of thinking about steps to incorporate green accounting: we have pollution damage measurements that, just like the valuation metrics, have matured between 1973 and today, such as some of the environmental modeling methods. For instance, if we're thinking about the impacts of pollution emitted by a power plant, we need to know something about where that pollution goes, what it might turn into along the way, some of the chemistry involved, who or what is exposed to it, their response in terms of elevated health risks, and then, ultimately, what is the value of those impacts. There are empirical challenges all along the way in that modeling chain.

It's nice to be able to say in 2019 that we've made great progress in some of those empirical steps that allow us to now have rigorous estimates of the damage for some of these pollutants. Of course, uncertainties remain, as they always will with any modeling exercise. But we've made a lot of progress. We're now in a position where we can credibly report to policymakers what these values might be, and have a serious discussion about extending the existing accounts to include the green accounts.

There's another side to this, which is vested interests. Firms in certain industries may not want environmental accounts to be officially on the books. For example, think about firms that own or consume large quantities of pollution-intensive fossil fuels. They may not want to have their value-added be the net of the pollution damage that their production activities cause. This is not a statement about their production activities not having value—just that the existing accounts turn out to mismeasure that net value, when one takes into account pollution damage.

Really, the obstacles are twofold. There's an historic, practical, empirical set of obstacles; and there's the obstacle of the status-quo way of doing things.

Let's get into some of the research. The paper we're going to talk about, from the journal Proceedings of the National Academy of Sciences, is called, “Fine particulate matter damages and value added in the US economy.” It's a paper you've authored with your colleagues Peter Tschofen and Inês M.L. Azevedo. You measure something called “gross external damages” from PM2.5 (fine particulate matter that measures 2.5 microns or less). You assess the health damages from this kind of pollution across different sectors of the economy.

Writ large, how do these gross external damages from PM2.5 compare with the traditional measure of GDP that we've been talking about, and how has that changed over time?

In the most recent year for which we have the data necessary to do the modeling exercise that was conducted for the paper—which is 2014—we find that the gross external damages amount to about 4 or 5 percent of GDP, which might not sound like much, but it's really a pretty big number. That 5 percent number is, of course, subject to assumptions made in the models about various key parameters, such as the value of damages and how sensitive humans are to PM2.5 exposure. We certainly go to the literature to find appropriate choices for those parameters, but I’ll just note that there are other ways to do this that can make that 5 percent number go up or down.

Generally, as a long-term trend, PM2.5 and associated damages have fallen precipitously in the United States since the 1970s. In some of my additional work, I've found that the share of GDP contributed by PM2.5 damage was much, much higher back in the ’70s. Not coincidentally, the Clean Air Act—our primary set of regulatory tools in the United States—was passed in 1970, and it was implemented meaningfully throughout the ’70s. I think those two things are certainly related.

When I have looked at the pollution monitoring data for years more recent than 2014 (just looking at the air pollution monitoring network in a collection of cities in the United States, and not the full modeling exercise that Tschofen and Azevedo and I conducted) I find a disturbing trend that the PM2.5 levels have started to go up in both 2017 and 2018, after a decade of continuous decline.

Just to put some numbers on the trends that you identify in the paper, in terms of the gross external damages from PM2.5: you find about 6 percent in 2008, about 4.6 percent in 2011, and then declining to 4.2 percent in 2014. But, as far as you know, 2014 is the most recent year for which the full set of data is available.

In the paper, you note that a relatively small number of economic sectors contribute a large share of these PM2.5 emissions and associated damages. Can you talk about those economic sectors, and which of them might offer some of the best opportunities for near-term emissions reductions?

We find this result is associated with damages from the agriculture sector, the utility sector, the manufacturing sector, and transportation. As you stated, those sectors together contribute just 20 percent of GDP, and they contribute three quarters of the total air pollution damages that we track.

If we were to think about the value that those sectors contribute to economic performance, it's important to note that we as authors of the paper are not arguing for the elimination of the agriculture sector, or utilities, or anything of the sort. What we're doing is merely saying, “Here's a way that you can characterize the value-added that those sectors contribute to the US economy.” And when you build in some of these extra-market, or non-market, impacts (the costs that they confer on the population) their value-added really changes appreciably—especially agriculture and utilities.

From the point of view of policies and additional abatement opportunities, I think the first thing I want to say is that the United States has made a lot of progress in improving our air quality over the 50 years that the Clean Air Act has been in place. And I would argue that it is a very bad idea to relax or not enforce the current standards, which have been very hard won. I'm thinking about things like fuel economy and the vehicle fleet, or relaxing or not enforcing some of the ambient standards for PM2.5. I would suggest that we work (at least initially) on maintaining our current goals as stated—statutorily and administratively.

I would also note that the agriculture sector is a really interesting place to think about additional abatement, because we traditionally think about air pollution control from smokestacks and tailpipes. (At least, that’s the way I think about it.) Agriculture offers different opportunities: We might consider changes in the composition or intensity of fertilizer, which contributes to emissions of ammonia, which contributes greatly to the damages that we're measuring in that sector. Additionally, animal wastes that are produced in the course of producing livestock are also important contributors. If we can manage that waste in a way that's more cognizant of the air pollution impacts from those production activities, that would be great.

In a more broad sense, it also speaks to how we as consumers think about the composition of our menus or the composition of our diet—right? If we're more aware of these upstream costs associated with the production of livestock for food, then we may decide to change our habits, or we may need nudges in the form of public policy to help us do that.

The most dramatic trend is a steep reduction in sulfur dioxide (SO₂) emissions from the utility sector. And I imagine that's mostly the decline of coal-fired electricity generation in the United States—is that right?

Yes. It's not only the decline in how much electricity we're producing by burning coal, which is largely about the switch to natural gas. But it's also the fact that many of the remaining coal-fired power generators are using flue-gas desulfurization (scrubbers), which removes SO₂ from the waste stream quite a bit—to the tune of 80 percent or more. Both of those market and regulatory forces are associated with or causing that steep decline in SO₂ emissions.

And that, of course, brings us back to the Clean Air Act and the importance of policy measures to deal with some of these issues.

Another really interesting feature is Table 1 of the paper. This table shows the ratio of damages to value-added across different economic sectors—how much health damage is associated with a given sector, and how much value-added is associated with that sector. Let's compare this simple metric across different parts of the economy. Can you talk a little bit about which economic sectors are most damaging, relative to their economic contributions, and which are the least damaging? How has that changed over time?

One interesting question is: How do we assess “most damaging”?

One perspective might be: You total up all the impacts, you get a gross external damage (GED) number, and then you rank the sectors and say the one at the top is the most damaging. As an economist, my view is that's not quite right, because we need to remember that these sectors are there for a reason—they're producing something of value (at least in principle) to the economy. So, we need to compare both the external damages (the air pollution impacts from these sectors) to the monetary value of the products that those sectors are producing. That’s what really leads us to this GED-to-value-added ratio that we report in the table.

From that perspective, it appears that livestock production is the most damaging relative to value-added in 2014. We estimate that the gross external damages from that sector are greater than their market value-added.

A cautionary note is that the fact that damages from that sector exceed its value-added should not suggest that we shut that sector down or ban its production. What it does say is that the regulatory apparatus, insofar as it's targeting air pollution and air pollution damages, is probably not stringent enough for that sector. We need to bring that ratio down. This is evidence that the regulatory stringency applied to that sector is apparently far too lax.

We also see evidence that sectors, or subsectors like waste management, are generating lots of damage relative to value-added—these are things like incinerators. But I would note that we need to think carefully about what value-added is for the waste management sector. There are probably non-market health sanitation benefits associated with waste management that may not be included in the value-added figure, which might inflate the ratio of damages from air pollution to value-added for that sector.

So, the measures that we're using here in this work may not account for the health values of having clean homes, clean streets, clean parks, and so on.

Absolutely.

When you look across these different economic sectors at their health damages and their value-added, do you see any particularly low-hanging fruit for either the private sector or public policies, to address emissions at low cost?

As a microeconomist, I don't typically think about businesses pursuing emission reductions as a primary objective. What we need to think carefully about is whether or not there are complementarities between profit maximization as the objective and emission reductions.

I’ll note something about low-hanging fruit, for continuing to reduce emissions and improve the environment: when EPA and other scholars have historically looked back at the Clean Air Act and assessed the benefits and costs associated with pollution improvements, they typically find ratios of benefits to costs in excess of five to one, or even ten to one. So, every dollar that's invested in pollution control, according to those studies, is leading to an additional $5 or $10 in human health and environmental improvement. Continuing to maintain or even strengthen standards, and to enforce existing standards, is not generally acting as a drag on economic performance—provided your measure of economic performance is inclusive of both the benefits (which may extend beyond measures like GDP) and the costs. If a five-to-one or a ten-to-one ratio is not low hanging fruit, I frankly don't know what is.