The New Executive Order: Cost-Benefit Analysis and Market Failure

I’m late in getting to this news, but on January 18th, President Bush amended Executive Order 12866 on Regulatory Planning and Review. The new order, EO # 13422, has some interesting and potentially troubling new provisions. Here are some links: pro, pro, pro, con, con, con.

Public Citizen identifies three problems with the new order, and I’m going to add a few more:

  1. “[I]t requires agencies to get White House approval of many important kinds of guidance for the public, which would allow the White House to create a bureaucratic bottleneck that would slow down agencies’ ability to give the public information it needs.”
  2. “[T]he new order stresses the concept of ‘market failure’ in its revised command for agencies to state justifications for new regulations for public health, privacy, safety, civil rights and the environment.”
  3. “[T]he order requires agencies to develop annual plans for upcoming rulemakings that identify ‘the combined aggregate costs and benefits of all … regulations planned for that calendar year to assist with the identification of priorities.'”

In addition to these, I’d add a few of my own:

  1. The President may not have the power to micromanage agencies empowered by Congress in this way while respecting the separation of powers. (Both the old and new Executive Orders may be unconstitutional.)
  2. This appears to add a good deal of red tape to the process of producing red tape. While that may seem appealing to knee-jerk libertarians, there’s no indication that this won’t just bloat the bureaucracy further. Sadly, most regulations are needed, and the market fails all the time. Perhaps, though, this process will give us more insight into just how it fails, and shut down the most egregiously bad policies before they’re enacted. It remains to be seen whether outlier regulations like will help pay for the cost of evaluating them all (a GAO study a few years down the road seems appropriate.)

Richard Belzer at Neutral Source disagrees with the first three explicitly. (He’s spent ten years at the Office of Management and Budget trying to regulate the regulators, so he’s understandably supportive of the whole scheme of executive oversight.)

  1. He points out that guidance documents are substantially identical to regulations, since they explain likely agency actions and reasoning to the industries they regulate. The Reagan-era Executive Order only covered actual rulemaking the agency engages in, but increasingly industries seek pre-ruling guidance in order to avoid conflict. The move from oversight over present rules to foresight over proposed new rules is not, he argues, a principled distinction. (This leaves open the question of oversight for regulations… but these agencies are under the charge of the executive branch, and their heads are employed at the pleasure of the President.)
  2. The language on ‘market failure’ is not new, though it may be given greater pride of place in the new text. Before, it was the first in a parenthetical list of possible reasons for regulation. Now, it’s the first in a non-paranthetical disjunct: regulation must be justified by market failure or ‘other problems.’ Again, this doesn’t address the question of whether agencies should be forced to answer to the White House.
  3. However, according to the plain text of the order, the amendments do not give anyone veto power over proposed regulation. Rather, it requires all agency heads to go to a meeting. That’s it: “The Director may convene a meeting of agency heads and other government personnel as appropriate to seek a common understanding of priorities and to coordinate regulatory efforts to be accomplished in the upcoming year.”
  4. Belzer is silent on the constitutional question, at least in these essays. I’ll just repeat my usual argument, which is that Congress arrogated their responsibilities vis-à-vis the administrative state decades ago. They’d have to work more than three days a week if they wanted to actually keep the government going themselves, so they’re content to leave it to various scientists, policy wonks, and efficiency experts. If the White House really wants to wade in and try to make sense of all that, I say let ’em. Maybe it’ll keep them from invading Iran.
  5. Belzer acknowledges that the new order “increases agencies regulatory planning obligations by requiring them to report estimates of the aggregate benefits and costs of regulations planned for the year. How much increase in analytic burden this imposes depends on whether the task is a clerical summation of estimates for individual rules (and now guidance) or it requires additional analytic effort.” How much is demanded is really up the executive in charge: if President Bush (or his deputy, Susan Dudley) doesn’t like the report, he can threaten to fire the head. That’s the extent of his practical power over these agencies; my bet is that most agency heads won’t make waves, so they’ll write whatever report Dudley wants.

None of this speaks to the desirability of cost-benefit analyses for new regulation, which in my opinion is quite high. Many people suggest that cost-benefit analyses deprioritize the intangibles, but I think it’s the opposite. When we finally understand the real costs of environmental pollution, for instance, we’ll quit polluting. The same thing goes for the costs of seat belts v. the cost of accident fatalities, the cost of accounting rules v. the cost of accounting corruption, ad infinitum. A good cost-benefit analysis is really about challenging our presumptions, looking at alternatives, and forcing us to weigh what matters to us. Public Citizen disagrees: “”These cost/benefit analyses are notoriously biased against regulation, especially long-term goals such as preventing global warming or cancers that manifest years after exposure to toxic substances. The upshot of this whole executive order is that the White House is already working to undermine not just agencies but also the new Congress’ ability to protect the public.” Maybe… but don’t blame economics for that: blame the White House.

That’s not to say that the next Administrator of the Office of Information and Regulatory Affairs, Susan Dudley, won’t misuse her influence in some way. Here’s some background on her: bio, pro, pro, meh, con, con, CON. (The last is a 68 page report put out by Public Citizen and OMB Watch.) Basically, she’s a free market ideologist: the new emphasis on ‘market failure’ is her baby, and she’ll undoubtedly nurture it for the next couple of years if she ever gets appointed. But she also looks like a crackpot, blindly opposing safety, financial, and environmental regulation that seems eminently reasonable to most people. These things are common sense to conservatives and liberals alike: air bags, banking privacy, regulations to prevent overfishing, or arsenic-free water. We all want these things. I don’t think she should be given a position of power for the same reason I wouldn’t give a hardcore Marxist an agency headship: ideologists lack the flexibility to be good administrators and managers. It’s worse in her case, insofar as she’s continually guilty of a basic logical fallacy. She regularly reasons like this: If people demand a safe or environmentally product, they will get it. This product does not exist. They do not really want this product. That’s a fine example of modus tollens, though of course it conflates the economic sense of demand with the common sense one. I may want to drive a car that is safe, but if I cannot afford one, economists do not count my desires as ‘demand.’

Dudley’s theory of market-failure, however, is not logical. She reasons: If the market has failed, we need government regulation. The market cannot fail: markets are self-regulating. Therefore, we do not need government regulation. Sadly, this is an example of denying the antecedent: If I’m asleep, my eyes are closed. I’m not asleep. By Dudley’s logic, therefore, my eyes are not closed. However, it is equally possible that I have blinked, just as it is possible that we make laws and regulations so as to register our demands as a people. Think of the government as an enormous Costco: we’re taking advantage of all the volume pricing deals to be had on safe roads, clean environments, and honest corporations and stock traders. The market may not fail in Dudley’s definition, but it regularly fails us. Markets aren’t nice: in a good market, rich people ‘demand’ more, and poor people ‘demand’ less: the market supplies them according to their means, not according to their needs. Yet even rich people can’t afford to pay Ford to build a car with seat belts if that requires Ford to retool the whole production line just for those few rich people. Markets suffer from a collective action problem: they are not always able to take advantage of economies of scale, and actually existing markets generally fail to plan for the future: they reward near-term profits, with which they hope to buy someone else’s innovations.

Nothing about what I have written negates the value of considering the costs and benefits of regulation. Men are not markets: we can think ahead, plan for uncertainties, and take lessons from the past in a way that the senseless beast of the market cannot. We partly inhabit markets, but we also have other identities. We evaluate cost-benefit statements as parents and partners, churchgoers and gardeners, sports fans and knitters, in other words, as citizens.

Second Opinions