How Tyler Cowen Learned to Stop Worrying and Love the Blob

Tyler Cowen is high variance. He can be very insightful, or annoying AF. His recent piece in Bloomberg on downsizing the Federal bureaucracy is in the AF category. (In the interest of balance I will cite to an insightful piece below).

The impetus for the article is Elon Musk’s statement that he would be willing to serve on a Department of Government Efficiency that would be tasked with cutting down the bureaucracy’s head count. Cowen refers to the fact that at Twitter Musk cut head count by 80 percent with no loss in output (and thus causing a 5-fold increase in productivity) and says it can’t be done with the bureaucracy.

Why? Sayeth Cowen:

Businesses need to make plans, and they frequently consult with regulatory agencies as to what might be permissible. The Food and Drug Administration needs to approve new drug offerings. The Federal Aviation Administration needs to approve new airline routes. The Federal Communications Commission needs to approve new versions of mobile phones. The Federal Trade Commission and Department of Justice need to give green lights for significant mergers. The Federal Deposit Insurance Corp. needs to approve plans for winding down failed banks. And so on.

There are so many problems with this argument. Where to begin?

Well, for one thing Musk’s Twitter experience demonstrates that even private business can have many zero marginal product (and indeed, zero average product) workers. But somehow the Federal bureaucracy doesn’t? It couldn’t be made more efficient by cutting the dead wood? Of which there is obviously a lot? Given that Cowen often writes about zero marginal product workers, it is rather remarkable that he implicitly argues that government workers are overwhelmingly positive marginal/average producers, especially given the extremely low power incentives and lack of accountability for poor performance that they are subject to.

In other words, identify the people who are actually performing the “customer service” to which Cowen refers, and get rid of the rest. Of whom there are likely a lot.

Moreover, these “customer service” functions represent only a fraction of what bureaucrats do. A good deal of it is administrative overhead which could be reduced if head counts and budgets are cut.

Another big part of it is writing rules and interpreting rules that were promulgated previously.

The number of pages of new rules added to the Federal Register is a common metric of increased regulatory burden. It is these rules (and their implementation, which I discuss below) that generate the deadweight costs of regulation. Fewer bureaucrats, fewer rules, greater economic efficiency. Without hitting “customer service,” and indeed, reducing the need/demand for customer service, since most of what Cowen describes as the essential function of the bureaucracy is helping business navigate a labyrinth of rules. Less bureaucrats, fewer rules, simpler labyrinth, less need for customer service (and the bodies necessary to perform it).

I would also argue that this customer service is chimerical, and is in fact a form of rent extraction that reduces wealth. That is, in supposedly “helping” business, bureaucrats’ produce bads, not goods.

I view regulation through the lends of property rights economics. Regulation is essentially an incomplete contract. Congress creates a really incomplete contract, bureaucrats craft regulations that make them somewhat more complete, but still wildly incomplete.

The allocation of residual control rights (rights not specified in the incomplete contract, or specified but non-verifiable) is vital in incomplete contracts. It is this allocation that affects incentives by affecting bargaining power. In regulation, it is the regulators who have the residual control rights. The “customer service” that Cowen describes involves negotiations with regulators who have residual control rights. They can say no to what you are asking for, or can add conditions, etc., etc., etc.

Property rights economics states that residual control rights (which in the literature basically involves control rights of access to assets) affect investment incentives because those with the rights have bargaining power. It is efficient to put bargaining power (and hence residual control rights) in the hands of those who can make non-contractible, value enhancing investments.

This is one reason why regulation, when viewed as an incomplete contract with a particular allocation of residual rights, is so likely to be highly inefficient: businesses, not government bureaucrats, are the ones who can make non-contractible, value enhancing investments but they don’t have the bargaining power. This is ass-backwards.

How do regulators use their bargaining power? To extract rents, which is what anyone with bargaining power does. Rents that do not incentivize wealth producing investments. They can’t receive these rents in monetary form (you can’t even take a regulator to lunch and pay for it). So they collect them in the form of justifying their existence, getting bigger budgets and staffs, empire building, implementing ideological objectives, ego and envy (feeling powerful by sticking it to those who make a lot more money), and creating obstacles so that they can sell themselves to the private sector as expert in working around.

(Again note the synergy between rule writing and bargaining with the regulated over implementation of the rules).

My experience with CFTC regulation, and in particular my short-lived experience as a consultant to the CFTC Inspector General, gives me an understanding of how this process works. One of the things I’ve written about over decades is the thicket of interpretations, no action letters, advisories, etc., that grew up in the aftermath of the passage of each major revision to the Commodity Exchange Act. The project I was working on with the IG was to document how extensive this practice has been, and how much the CFTC has relied on it.

The process by which this thicket is created is exactly the process that Cowen describes. “Consulting with regulatory agencies to see what might be permissible.” In essence, negotiation between the regulators and the regulated, with the bargaining power–and the ability to extract rents–resting in the hands of the regulators. (I do understand that sometimes information asymmetries give bargaining power to the regulated. My experience decades ago looking into the 1988 S&L bailouts showed me how this can work).

So, what would happen if the number of regulatory personnel was reduced? Cowen thinks it would reduce the output of successful negotiations, and that businesses would be stymied in the absence of such.

But is that really the case? For one thing, the zero MP/AP issue discussed above means that a reduced head count does not imply a reduction in output (the number of successful negotiations). But even overlooking that, if one views a major effect of these negotiations as rent extraction, reallocating bargaining power away from bureaucrats can improve efficiency.

Regulators with less capacity to negotiate have less bargaining power: the ability to string things along (which is easier when you have more bandwidth) enhances that power. Furthermore, more regulators typically means more fiefdoms within the bureaucracy that have a place at the bargaining table, hence more negotiations, more complicated negotiations (since more bureaucrats have effective veto power), and greater power to extract rents. The potential for multiple marginalization problems grows. (Think of when every local lord could extract tolls from those traveling a road that ran across their lands).

The real difference between Cowen’s view and mine seems to be that he believes that increasing the number of bureaucrats increases the output of wealth enhancing negotiations, whereas i believe that it likely increases the amount of wealth reducing rent extraction that occurs due to bureaucrats’ bargaining power.

A customer service model vs. an extortion racket model, if you will. Help with navigating through complex and confusing regulations (complex and confusing due to their incompleteness) vs. using that complexity and confusion to hold up wealth producers.

Starkly different views with starkly different implications. Cowen may not be all wrong, but his rosy “I’m from the government and here to help you” view clashes with the economics of bureaucracy, and certainly my personal experience with it.

So I say hack away. That’s the only way to find out which parts of the bureaucracy enhance wealth and which don’t. Hire back the ones that do and force those that don’t find an honest living.

Which is pretty much what Musk did at Twitter.

As promised, in the interest of balance, I recommend Cowen’s takedown of the insane idea to tax unrealized capital gains (and the companion piece by his co-blogger Alex Tabarrok).