The Grumpy Economist: Comments on Costs and Benefits

The Biden administration is proposing major changes to the cost-benefit analysis used in all regulations. preamble hereAnd full text here. It’s open to the public comments until September 20.

Economists don’t often comment on proposed rules. We must do this more often. Agencies take such comments seriously. And they might have an afterlife. I have seen comments quoted in lawsuits and judgments. Even if you doubt the Biden administration’s willingness to listen to you on a cost-benefit analysis, the comment is a marker that could very well be considered by an imminent possible Supreme Court case. Comments usually come only from stakeholders and lawyers. Ordinary economists really should comment more often. I don’t do it enough either.

The user is supposed to be able to see existing comments. https://www.regulations.gov/ but I couldn’t figure out how to see the comments on that one document only. (Write if you know how to do it, with a link).

Take a look at the comments MIT team led by Deborah Lucas here and Josh Rauch as soon as they are published. These are excellent comment models. You don’t have to review everything. Make one good point.

Cost-benefit analysis is useful, even if it is imprecise. The many brilliant ideas in Washington (and Sacramento!) are unlikely to document any net benefits. Yes, these exercises can lie, cheat, and steal, but the need to come up with quantitative lies can show how reckless many of the rules are.

Both Josh and MIT’s response focus on the use of ultra-low discount rates in the proposal design, ranging from historical TIPS returns to arguments for zero or negative “social” discount rates. Josh highlights the perfect compromise: always show the annual flow of costs and benefits. Then it is enough just to apply different discount rates. No black boxes.

Discount rates seem to be a technical issue. But they are of great importance for climate policy, or for policies with significant costs but supposedly permanent benefits due to long-term prospects. For example, climate change is thought to create a cost of 5% of GDP over 100 years. So let’s assume the discount rate is 0% – treat the future as you would the present. How much is it worth spending this year to eliminate additional climate change in 2100? Spending means real spending, a real decline in everyone’s standard of living, not just ridiculous billions of money on twitter.

If you answered “5% of GDP” (approximately $3,500 per person), this is incorrect for two important reasons. First, the economy grows over time. With a modest 2% real growth, US GDP in 100 years will be 7.4 times what it is today, or 740% more. (e^2=7.4). So 5% of GDP in 100 years, discounted at 0%, is 7.4 x 5% or 37% of today’s GDP, or $17,500 per person today. Secondly, the gain is infinite – 5% of GDP in 2123, but 5% of GDP in 2124, and so on. Discounted at zero rate, 5% off 2123 forever after that is worth… an infinite amount today. But GDP continues to grow after 2123. If you discount for anything less than the GDP growth rate – 2% in my example – 5% of (growing) GDP is forever worth an infinite amount! So if $250 billion in subsidies for huge long-range battery electric vehicles built by unions in the US from hypothetical US-made lithium mines could save a handful of car carbon overall (is that even positive?), if the benefits are endless. , act.

If you discount by a low but slightly more reasonable number, like 7%, then the dollar is worth 0.09 cents today (100 xe^-7) 100 years from now. Now you know what to do with climate scales!

You may be wondering if our great-grandchildren will be so fantastically better than us, let them handle it. Or you might be wondering what maybe other things we could do with money today that could speed up this magical process of growth and make it 5% better. Can we do anything for an infinite amount of money to raise the growth rate from 2% to 2.05%?

Last opportunity cost The question is, I think this is a good way to think about discount rates. The average real return on shares is at least 5%. The average marginal product of capital before tax is higher; pick your number, but it’s in the 10% range, not 1%. The correct “discount rate” is the rate of return on the alternative use of money. Josh and the MIT team are absolutely correct in pointing out that using the risk-free government bond rate is a completely misguided way of discounting the very risky cost of climate damage – that 5% is a very little known number – and the even riskier benefits of shifting. passions of the government in the field of climate policy. But I think framing the experiment in terms of opportunity cost rather than proper discounting of risky flows makes it more visible, despite the decades I’ve spent (and a whole book!) on the latter approach. Businesses can take $1 today and turn it into an average of $1.07 next year. Why take this money for a project that will bring in $1.00 or $1.01 next year?

The first question has a deeper consequence. Why do we have to suffer to help people, even our grandchildren, who, on average, will be 7.4 times better than we are? How much would you ask your great-grandparents to donate to make you 5% better than you are today?

Here, the low discount rate interestingly clashes with another part of the supply: equity and transfers.

From the preamble, page 12:

The standard assumption in economics, based on empirical evidence (as discussed below), is that an extra $100 given to a low-income person increases that person’s wealth more than an extra $100 given to a rich person. Traditional cost-benefit analysis, which applies uniform weights to willingness-to-pay measures, usually does not take into account how distributional effects may affect aggregate welfare due to differences in the marginal utility of individual income. Related to the topic of distributional analysis is whether agencies should be allowed or encouraged to develop estimates of net benefits using weights that account for these differences.26 The proposed revisions to Circular A-4 suggest that agencies may wish to consider for each income group for which regulated is equal to the average income of the group divided by the average income in the United States raised to the elasticity of marginal utility multiplied by a negative unit.

Now wait a damn minute. The “standard” doctrine in economics is that you can’t make intrapersonal comparisons of utility. Utility is ordinal, not cardinal. Here, cardinal utility utilitarianism with equal Pareto weights is about to be set in federal stone. (To determine the social benefit of taking from A and giving back to B, you construct a social welfare function \(u(c_A) + \lambda u(c_B)\). To do this, you need to use the same \(u()\) for A and B, and agree on the Pareto weight \(\lambda\) here implicitly.)

Imagine a simple rule: Take a dollar from Joe (income $100,000) and give it to Kathy (income $50,000). According to this standard, such a direct transfer is tested for profitability.

But it doesn’t apply over time. Taking a dollar from you and me and giving it at 0% to our great-grandchildren, who will be 7.4 times richer, should be a wake-up call of mass inequality. Nope.

Indeed, you can derive the discount rate from the inequality target. Net undiscounted intergenerational equity requires a discount rate that is proportional to the rate of economic growth.

(In the case of power supply, an intervention that costs A $1 to give B $\(e^{rt}\) simply passes the cost-benefit test if \[c_A^{-\gamma} = e^{rt} (c_B)^{-\gamma}.\] If B is \(e^{gt}\) times better than A, \(c_B=e^{gt} \times c_A\), then we need \(r=\gamma g\). \(\gamma\) is usually a number slightly greater than one. The discussion of \(\gamma\) values ​​in the preamble is pretty good, stopping at numbers one through two. However, they have not really heard of the finance literature:

Risk aversion data can be used to estimate the elasticity of marginal utility. In a utility specification with constant elasticity, the coefficient of relative risk aversion is the elasticity of marginal utility. There are many different estimates of risk aversion ratio (CRRA) using data from various markets, including labor supply markets,29 the stock market,30 and insurance markets.31 Estimates vary widely, although assumed CRRA values ​​between 1 and 2 are common.32

30 Robert S. Pindik, “Risk Aversion and the Determinants of Stock Market Behavior,” Review of Economics and Statistics, 70, no. 2 (1988): 183-90 uses stock market data and rates CRRA as “in the range of 3 to 4”.

Since then, of course, there has been a whole literature on stock premiums ranging from 10 to 50. Shh. It would justify the insane level of fairness.

The project also encourages all sorts of non-quantifiable non-economic “benefits”, but I’ll leave that for another day.

Read and comment.

By the way, despite my negative tone and nitpicking about these elements, most of the draft is pretty good. Here is a particularly good excerpt from page 26 of the full text

J. Note on Certain Economic Regulations

In light of both economic theory and actual experience, it is particularly difficult to demonstrate positive net benefits for any of the following types of regulation:

 price regulation in well-functioning competitive markets;

 production or sales quotas in well-functioning competitive markets;

 Mandatory uniform quality standards for goods or services, if a potential problem can be adequately addressed by voluntary standards or by disclosure of information about the hazard to buyers or users; P

 employment or production controls, except (a) as necessary to protect health and safety (for example, FAA tests for commercial pilots) or (b) to control the use of common property resources (for example, fishing , radio waves, federal states and offshore zones).

Well, the FAA tests and regulations for commercial pilots aren’t really that obvious and really need a cost-benefit test. “Commercial pilot” does not mean “airline pilot”, it means that you can do anything on the plane and get paid for it. But leave that for another day, these principles, if applied, can remove a lot of harm. Well, I believe that many of the progressive left or the emerging national conservative right would deny there is such a thing as a “well-functioning competitive market.”