Before Thanksgiving, a business reporter contacted me for an article he was writing on discussion points for families that wanted to talk politics around the holiday table. He asked me specifically to write up some thoughts about the Trump administration’s regulatory policies, and I did. He ended up not using them, even though I thought they were pretty provocative; in essence, I argued the Trump regulatory record was much less significant than what both the administration’s supporters and critics claim.
I thought they’d make for some good EconLog discussion, so I’m sharing them here:
As a policy analyst, I judge regulations on the question of whether they improve human welfare on net, compared to alternatives. This can lead me to applaud some regulations, criticize others, and call for reform—including deregulation—of still others. (As a libertarian-leaning analyst, I tend to do more of the latter two, but all three happen.)
Republicans tend to see regulation in general as harmful to human welfare, and they mainly criticize regulation and promote deregulation. By this Republican standard, there are two ways to judge the Trump administration’s record: what new regulations have they blocked, and what deregulations are they implementing?
On blocking new regulations, the administration’s record is astounding: rulemaking has largely ground to a halt. This is historic; no previous administration has come anywhere close to this decrease in output.
However, on actual deregulation—reform or removal of regulations already on the books—the administration has done little, and the successes it claims to have had are largely meaningless. Take, for instance, the ending of the Obama administration’s Clean Power Plan targeting coal-fired power plants. Though CPP is gone, coal plants continue to disappear and be replaced in part by cleaner, gas-fired “co-generation” plants. Gas simply makes better economic sense than coal, so the Trump CPP maneuver is largely meaningless (as was the original Obama CPP).
If you believe federal regulation can be more efficient—that we can improve public welfare at a lower economic cost—the Trump record on deregulation is very disappointing but not surprising. Deregulation is hard work, time consuming, and requires political consensus-building, like we had during the major deregulations of the 1970s–1990s. The Trump administration has shown little ability to do those things, and that’s unlikely to change in the future.
That said, there is a significant Trump deregulation effort now in the pipeline: the proposed change to the Waters of the United States (WOTUS) rule, overturning an Obama administration rule on what water bodies fall under federal regulation as opposed to state regulation. This would be serious deregulation, addressing problems with the Obama rule on both constitutional and public welfare grounds, and a lot of work has gone into it. But can the Trump administration ultimately deliver on it and, if so, will it survive for long after he leaves office?
READER COMMENTS
Loquitur Veritatem
Jan 17 2020 at 11:44am
You say:
Your statement suggests a belief in something like a social-welfare function, an aggregation and summation of hundreds of millions of individual states of happiness (or unhappiness) across time. Surely you don’t believe in such a thing, for if you did you would believe that that humanity is somehow better off if A gains 10 points of happiness (whatever that is) by punching B in the nose, while B loses only 9 points of happiness by being punched. My illustration is absurd because the idea of a social-welfare function is absurd.
Jon Murphy
Jan 17 2020 at 12:05pm
One can talk of human welfare while rejecting a social welfare function by separating the technical, economic definition of “welfare” from the more general use.
For example, one can conceive (as Adam Smith did) free trade in terms of human welfare by considering: “does free trade allow human beings to flourish however it is they choose to define “flourishing”?” That is a large part of The Wealth of Nations. Such a framing requires no social welfare function.
Phil H
Jan 18 2020 at 2:26am
“My illustration is absurd because the idea of a social-welfare function is absurd.”
Absurd is a very strong word. Personally, I like to reserve it for things that no-one does, in fact believe. The idea of total social welfare is believed by many, and used implicitly in many public policy decisions. Decisions over vaccines, law enforcement, eminent domain, and (most clearly) preventative military strikes all seem to employ some variant of a total welfare function. All of which you may reasonably disagree with, but to call them “absurd” just seems to to put you too far outside bounds of everyday thinking.
Jon Murphy
Jan 18 2020 at 7:26am
The popularity of a concept does not have any bearing on whether or not the concept is absurd. Not does an absurd concept imply it has no use.
Most scientific models are absurd if taken literally (indeed, glib dismissals of the models flow from precisely these grounds). Models are simplifications of reality. They just be made absurd.
The social welfare function is literally absurd. But it can also be a helpful framing device. The absurdity of the model only becomes problematic when one tries to take it literally and build policy around it (eg., Socialism or optimal tariffs).
BC
Jan 20 2020 at 1:38am
Suppose B agreed to allow A to punch B in the nose in exchange for x dollars and A was willing to pay x dollars to do so, revealing that A obtained at least x dollars worth of utility from punching B and B lost less than x dollars of utility by getting punched. Would you still disagree that humanity is better off by allowing A to punch B? After all, B was willing to take the deal voluntarily, i.e., he believed he would be better off receiving x dollars than not getting punched. One could argue that professional boxing exists as a sport because some boxers are willing to get punched in exchange for compensation.
I’m not sure how “social welfare function” is defined, but there certainly is a meaningful notion that society can be made better off by giving more benefit to A than B loses. What’s questionable is whether any centralized entity, such as a committee evaluating a proposed regulation, can learn the distributed and localized information necessary to determine whether a particular regulation will improve human welfare on net. In our example, we only learn that A values punching B more than B values not getting punched because both parties agree to the transaction. It’s not clear to me how, in the general case, we can learn the necessary information from markets to evaluate the net impact of a particular regulation on human welfare.
nobody.really
Jan 17 2020 at 1:49pm
I concur that the end of the CPP has not been the primary driver of changes in the electric generation market.
That said, the developer of a coal plant must operate for a long time to be able to recover his investment. The CPP—even though it never took effect—signaled to developers that they’d better not make any investments that generate a lot of greenhouse gasses unless those investments will prove profitable before the next Democratic administration.
The CPP is like the Taliban whispering in an Afghani’s ear: “You can choose to cooperate with Trump. But someday he’ll be gone. And what do you suppose will happen then…?”
robc
Jan 18 2020 at 9:10am
There was the Reason Mag interview with Coase in which he said he couldnt find a good regulation(new) when he worked for Regulation.
So I am interested in your examples of one’s you applaud.
My theory, based on the Coase interview, is that we are well past the point where diminishing returns have gone negative, so any new reg is bad and eliminating any reg is good, no matter how good or bad they are in isolation.
Tom
Jan 18 2020 at 10:08am
You don’t believe that any of the environmental regulations that have reduced lead were a good idea?
robc
Jan 18 2020 at 3:35pm
I think it is close, but those may have preceded Coase’s time at Regulation.
That is probably a positive one, but having to verify I didnt use lead based paint when I sold a house built in 2016 is a cost. So, at this point, the benefit is probably very slowly decreasing ( as it was a minimal cost).
Mark Z
Jan 18 2020 at 3:20pm
Well, the number of regulations on the books doesn’t necessarily determine whether a nonexistent regulation will be beneficial; I think it’s more probabilistic: the more regulations currently on the books, the lower the probability that the next regulation will past the cost-benefit test, after all the ‘low hanging fruit’ have been picked. Unless, of course, there is no correlation between how good (or bad) a regulation is, and how likely it is to be implemented, in which case the millionth proposed regulation may be as likely to be beneficial as the first one, though if this is the case, it also suggests regulators and legislators are so thoroughly incompetent at assessing the usefulness of regulations that they probably shouldn’t be trusted to regulate in general.
robc
Jan 20 2020 at 8:41am
You are partially right, but I think just the number of regulations is part of the problem. Let me give an example with very simplified assumptions.
We have two regulations, A and B, each with a benefit of 1000 units.
Passing both should then be +2000 units, right? Well, no, because the crossterm AB has a small negative, lets say -1.
So A+B+AB = 1999. Still a good deal.
Then we add on C, also +1000. But, it has crossterms of AC, BC, and ABC, lets just assume -1 each. So that only gets a net benefit of +997, still a good deal.
Add on D, +1000, plus AC, BC, CD, ABD, ACD, BCD, and ABCD. That nets +993.
It won’t take long at all until the crossterms negatives outweight the main terms positive.
Like I said, a very simple model. In reality, not all main benefits are equal, they probably descend with time as you suggest. Some of the crossterms may be positive too. And maybe the larger terms, like ABCDE, get really, really tiny and don’t matter, but I think every now and then, one of those will jump up and be surprisingly big, taking out most, if not all of the benefit of the added regulation.
I really think the large morass of interlocking regulations is the problem. Add to that the fact, that as you suggest, the low hanging fruit is long gone, and it is hard for a new regulation to have positive value.
Comments are closed.