On Regulation

“You’ve Got Mail” was a hit romantic comedy movie in which a behemoth bookstore chain crowds out a popular, local, independent shop that lacks the logistical scale and marketing savvy to compete for fickle Upper West Side readers. The independent owner is renowned for her taste in children’s books and her ability to recommend titles with tailor-made suitability using only her intuition and a brief exchange of words. However, her local knowledge is no match for the commercial cognition of the colossus riding astride massive amounts of data, or so it would appear.

There was a proliferation of these large bookstores under the Barnes and Noble banner, including one of their flagships at 82^nd and Broadway.

So, it is interesting to see the latest turn of events, centered at this same beleaguered location. Here is the WSJ:

“By shifting control of the process to individual store managers across the country, Daunt is giving local booksellers permission to do things they were never able to do before. They have discretion over purchasing, placement and even pricing. He wants Barnes & Noble locations to feel welcoming but not overwhelming—a chain store should be more inviting and less intimidating than a truly independent shop—and that means he needs the people who run them to make sensible decisions for their markets.”

The new CEO is attempting a turnaround of Barnes and Noble’s brick-and-mortar locations by decentralizing significant decision-making to the manager and the team at each individual store, starting with the Upper West Side instance.

If it works, Barnes and Noble stores will vary from one location to the next, reflecting the choices made by the on-the-ground staff as they serve the preferences of the surrounding community. What works in New York may not suit Miami. The parent company becomes an umbrella brand and administrative backstop for a network of local booksellers, in which each physical store is a node in a network.

The lynchpin for success is the notion that the company can exploit massive amounts of heretofore dark local knowledge, sparking a revival of foot traffic that leads to higher sales. The alternative hypothesis in this grand experiment is that a centralized approach works best because preferences are the same everywhere, leading to the conclusion that centralized management is best. This strategy withered under the pressure of Amazon and its long tail. If Daunt is correct, people want the in-person book buying experience and local feel that transcend the ordeal of obtaining the fruits of an impersonal algorithm delivered over a cold Internet connection.

In subsequent posts, I want to explore this question of how knowledge accumulates and disseminates within and outside bureaucracy, a topic that Ian Brodie has suggested to me.

In a sense, management is regulation of the enterprise, making decisions for the optimization of the whole that we presume is superior to what we would obtain with localized control. There are rules and procedures for how to stack the books, how to lay out stores, how to submit orders, etc. that are uniform across all the stores, enforced by a central authority. Perhaps this is due to the presumption that head office functionaries are better at managing information as it flows through the organization. The Barnes and Noble shift has the potential to be interesting beyond its scope. If we can learn lessons from this experiment, we can apply them to regulation writ large.

With that context in mind, I want to take my cue from John Cochrane at the Grumpy Economist. Cochrane called out the contemporary regulatory experience in a post last month entitled “Time for a new (?) theory of regulation”:

“This doesn't fit either the econ 101, benevolent nanny, or regulatory capture view. Fundamentally, regulators have captured the industry, not the other way around. They hold arbitrary discretionary power to impose huge costs or just shut down companies. They use this power to elicit political support from the companies. There is a bit of old Chicago school capture in the deal. Companies get protected markets. But the regulators now don't just want a few three martini lunches and a cozy revolving door to "consultant" jobs. They demand, political support. The regulators are more political ideologues than gently corruptible insiders.  

“Sometimes regulators seem to attack businesses just for fun, like suing a moving company for age discrimination. But maybe here too they are showing everyone what they can do, or scoring some ideological points so people get the message.  

“The increasing arbitrariness of regulation is part of the process. I find myself nostalgic for the good old days of the Administrative Procedures Act, public comment, cost benefit analysis, and formal rule making. Now regulators just write letters or take legal action, which even if unsuccessful can bankrupt a company.  Using administrative courts, the regulators are prosecutor, judge, jury, and executioner all rolled in to one.”

The benevolent nanny addresses externalities in the marketplace:

“Econ 101 courses repeat the benevolent dictator theory of regulation: There is a "market failure," natural monopoly, externality, or asymmetric information. Benevolent regulators craft optimal restrictions to restore market order. In political life "consumer protection" is often cited, though it doesn't fit that economic structure.“

We’ve touched on regulatory capture in the past, but here is his summary:

“Then "Chicago school" scholars such as George Stigler looked at how regulations actually operated.  They found "regulatory capture." Businesses get cozy with regulators, and bit by bit regulations end up largely keeping competition down and prices up to benefit existing businesses.“

Let’s start with the benevolent nanny/dictator theory. We can see regulation as a pure public good.

What is a pure public good? It is something the consumption of which is “non-rivalrous” and “non-excludable.”

Let’s start with a definition of “non-rivalrous”:

“Non-rivalrous goods are public goods that are consumed by people but whose supply is not affected by people’s consumption. In other words, when an individual or a group of individuals use a particular good, the supply left for other people to use remains unchanged. Therefore, non-rivalrous goods can be consumed over and over again without the fear of depletion of supply.”

Moving to non-excludable goods:

“Non-excludable goods are public goods that cannot exclude a certain individual or group of individuals from using them. For this reason, it is nearly impossible to restrict access to the consumption of non-excludable goods. A public road is an example of a non-excludable good. Almost everyone has access to a public road, even if they are just walking on it (rather than driving a motorized vehicle).”

A classic example of a pure public good (for the strength of its non-rivalry and its non-excludability) is defense. We can imagine a primitive case in which a tribe of our ancestors, sitting around the fire, negotiate how much each member will contribute in time and resources to protecting the collective. The level of protection against attack from a hostile tribe is unaffected by the number of people enjoying it. If Bob contributes nothing or a lot, he enjoys the same amount of protection. Bob can be a “free-rider” on the contributions of everyone else. This is the externality associated with public goods.

If we think about it, we might argue that regulation is non-rivalrous and non-excludable. The act of regulating industry X does not limit the ability to regulate industry Y. The benefits of regulation inure to everyone. When the SEC regulates security markets, everyone in the United States benefits from free and fair capital markets, not just the investors in a particular issue that must abide by the rules deriving from securities laws.

This is what this paper on the utilities industry argues:

“Regulation shares the attributes of public goods: non-excludability and non-rivalry. Regulation ought to be non-excludable. However, this attribute of publicness is not a fixed attribute. It depends on policy design. It is sometimes difficult to characterize regulation as a public good because more often than not, the public bad – regulatory capture – is present … Regulation should be non-rival. The existence of an additional consumer being subjected to the same regulation does not entail any disadvantage to existing users. On the contrary, when regulation is extended to additional consumers, it creates positive externalities. For instance, regulation that improves access to water and sewerage benefits not just the individual, but the entire community by lowering health risks.”

So far, so good.

Returning to Bob the free-rider, Paul Samuelson developed a model that showed the consequences of this externality. Private provision of pure public goods is sub-optimal. It’s not just Bob. Everyone around the campfire (assuming they have the same preferences) will behave the same way. In the corner solution, there is no defense provided. Generally we would imagine that there are too few resources provided for defense. We could improve the overall outcomes for everyone if we demanded that everyone put up more than they would choose to provide in a private framework. A benevolent dictator would make the optimal decision. This requires simplified assumptions about utility functions.

If we apply this to regulation, having a government decide on the optimal level of regulation leads to a greater amount of rulemaking than what we would expect if private agents negotiated amongst themselves.

Here’s Wikipedia:

“The Pareto optimal provision of a public good in a society occurs when the sum of the marginal valuations of the public good (taken across all individuals) is equal to the marginal cost of providing that public good. These marginal valuations are, formally, marginal rates of substitution relative to some reference private good, and the marginal cost is a marginal rate of transformation that describes how much of that private good it costs to produce an incremental unit of the public good. This contrasts to the Pareto optimality condition of private goods, which equates each consumer's valuation of the private good to its marginal cost of production.”

A model in which politicians, bureaucrats, and private players negotiate over how much of the public good to provide suggests public choice theory. This is closer perhaps to what we observe.

“Public choice, or public choice theory, is "the use of economic tools to deal with traditional problems of political science". Its content includes the study of political behavior. In political science, it is the subset of positive political theory that studies self-interested agents (voters, politicians, bureaucrats) and their interactions, which can be represented in a number of ways – using (for example) standard constrained utility maximization, game theory, or decision theory. It is the origin and intellectual foundation of contemporary work in political economy.^[2]

“In popular use, "public choice" is often used as a shorthand for components of modern public choice theory that focus on how elected officials, bureaucrats and other government agents can be influenced by their own perceived self-interest when making decisions in their official roles. Economist James M. Buchanan received the 1986 Nobel Memorial Prize in Economic Sciences "for his development of the contractual and constitutional bases for the theory of economic and political decision-making" in this space.^[3]

"Public choice analysis has roots in positive analysis ("what is") but is often used for normative purposes ("what ought to be") in order to identify a problem or to suggest improvements to constitutional rules (i.e., constitutional economics).^”

We can imagine a spectrum of outcomes with the benevolent dictator occupying the central spot, with regulatory capture at one end, and with regulatory discretion at the other. The level of regulation is least with regulatory capture and most with regulatory discretion. The equilibrium resting point on this spectrum is a product of negotiation and policy design. Leverage is a question of politics. In the United States, years of Congressional political retreat, expanding powers for the civil service, and diminishing regard for institutional precedent of the kind Cochrane describes mean that the executive branch has gained power. Executive branch officeholders can tip the scales one way or the other, subject to the constraint of the built-in buoyancy that favors regulatory discretion such as administrative tribunals in which regulators act as prosecutor, judge, and jury. We saw this with the massive deregulatory imperative of the Trump Administration and its unwinding by his successor.

If we attempt to model the future, we should forecast that there will be continued (if not increased) volatility as the fulcrum point on the spectrum swings back-and-forth for as long as political divides remain sharp and even. This will impose costs on all economic agents in the form of additional uncertainty.

Things become only more complicated in terms of the over-provision of regulation qua public good when we consider the disparity in preferences by state or region. The federal government makes rules on things like securities trading and antitrust with national scope, even as political and private preference differ across the geographic dimension. We might expect that some regions would prefer much less regulation than what the center dictates.

This makes Barnes and Noble’s decision to decentralize decision-making all the more remarkable. It represents a unilateral rejection of management’s negotiating leverage in its abdication of consolidated power, with the apparent belief that the decisions made by the center had become far from optimal. This is a function of the competitive dynamic that led to years of decay. If not a Zombie, we can say that Barnes and Noble teeters on the edge of the abyss.

Would the same decentralized thinking benefit government?

What would be the catalyst? Perhaps it would be a future fiscal or global financial crisis.

Time will tell.