If Middlebury College economist David Colander and theoretical physicist Roland Kupers wanted to get pro‐​market, government minimalists like me to read their new book, then they did a good job of picking its title, Complexity and the Art of Public Policy. The idea of solving problems “from the bottom up” is an appealing one because most public policy operates in the exact opposite direction, and their addition of “complexity” is intriguing.

Wondering if Colander and Kupers would shed new light on how to reform the bad policies we currently endure and avoid adopting more of them, I dove into the book and soon encountered this passage:

The standard way of doing policy considers our social system as a suburban garden. It tills, plants, and cultivates as if the parts are not interrelated. For example, it accepts that people have the tastes they have, and works within that framework. The complexity way of doing policy sees everything as interrelated; tastes are endogenous, and one must consider how tastes are affected by policy, whereas in the standard frame one does not.

So if we “do policy” according to their “complexity frame,” can we actually devise laws and regulations that are superior to the old “suburban garden” approach? Apparently they think so; I wanted to see exactly how.

Revising “I, Pencil” / The authors proceed to tell us that humans could better solve their problems if policymakers would stop thinking in either “market fundamentalist” ways or in top‐​down, command‐​and‐​control ways. Instead, they say, policymakers should adopt the more reasoned, mathematically based “complexity frame” of analysis. Hearing that “market fundamentalist” animadversion, I started to suspect that the authors’ ideal of “complexity analysis” is more amenable to state power than to laissez‐​faire.

That suspicion was confirmed in their chapter, “I Pencil Revisited.” Colander and Kupers look at Leonard Read’s famous 1958 essay and maintain that while it makes a useful point about the way prices and market competition work to coordinate the production of an item nobody could produce on his own, it shortchanges the importance of government. They go so far as to offer this addition to Read’s original text: “So, to tell the complete story of my production, I need to include government, and the many other collective groups, such as the Pencils Producers Association to which my family belongs, that assist government in its coordination role.”

Colander and Kupers suggest that the reason why Read didn’t mention government in his essay was that he feared that including it “would lead some people to expand the role of government.” In this, they’re mistaken. Read simply figured that most people understood fairly well that we need government to protect property rights and settle disputes, but that very few had any clue at all about the amazing coordinating power of the free market. That’s why his “ode” extols uncoerced human cooperation and leaves government out. Read was not an opponent of government, but argued that it has to stick to its rights‐​protecting, order‐​keeping functions. If it does so, then the market works to produce pencils and all sorts of other goods and services.

Complexity analysis / The authors, on the other hand, believe that government needs to do more so that people can better solve their problems. Complexity science, based on cutting‐​edge advances in math and computation, supposedly improves upon our ability to comprehend social systems by seeing how everything affects everything else. Supposedly it can give us better policy than the minimal state of the “market fundamentalists.”

That idea fascinates the authors and they envision a future in which “individuals can still have significant freedom of action, while achieving collective social goals.” Expecting that they would offer some concrete examples of how complexity analysis would improve upon the minimal “night watchman” state advocated by thinkers in the Smith/​Bastiat/​Read tradition, I read on.

Remarkably, there were none.

Could complexity analysis help us better adjust our welfare system, reduce teen unemployment, or improve upon our dismal educational results? The authors never venture into any such specific questions.

They do inform us that the Dutch used complexity thinking to improve traffic flow, adopting roundabouts at intersections. But it’s unclear why they think that advancement requires expanded government intervention in markets; if roads were privately owned, the owners would seem to have plenty of incentive to adopt the most efficient traffic controls.

The authors suggest that complexity analysis could be useful in banking regulation—but then admit that it might be more useful to firms in banking and finance, better enabling them to manage their operations. Again, this doesn’t give much support to expanded government intervention. One does wonder, though, if we didn’t have all the moral hazard–creating government rules for such firms, would there be any need for regulating their complex operations?

Colander and Kupers also say that complexity analysis could be important in environmental regulation. Perhaps, but with so much disagreement among experts about how, and how much, to control pollution, there would seem to be too much complexity for good policy.

Complexity analysis, no doubt, has important uses. But after reading the book, I’m not persuaded that government policy is among them. Despite the authors’ criticism of “market fundamentalism,” I don’t see how complexity analysis could have improved societies where that “fundamentalism” prevailed. Would Hong Kong, for example, have been more prosperous and its people happier if Sir John Cowperthwaite had tried “laissez‐​faire activism” (as the authors call their preferred policy) instead of plain old laissez‐​faire?

A big part of their activism involves governmental “nudging” of people to do what they “really want to do” but can’t discipline themselves to do. Colander and Kupers like the idea of policy that doesn’t dictate to people, but merely encourages them to do what they ought to do. They embrace the idea, most famously advanced by Cass Sunstein and Richard Thaler, that behavioral economics “creates a new role for applied policy economists, that of choice architect; government policy creates the choice architecture within which people make decisions.” Thus, public policy might be fashioned to nudge people to eat less junk food and save more money.

The authors fail to explain, however, what “nudging” has to do with complexity and, more importantly, why nudging should be a function of government rather than of society’s many voluntary organizations. Churches have been nudging people for thousands of years to behave better, probably with some success. So have a host of personal improvement groups such as Alcoholics Anonymous. Moreover, the apparent need for “nudging” would often disappear if we made the law less complex. For instance, a big reason why most Americans no longer save very much is that the tax code penalizes thrift. It seemingly would make more sense to change the tax code than to create countervailing nudges.

Another policy change Colander and Kupers advocate is the encouragement of “for‐​benefit” organizations. Such organizations “blend the social motives of a nonprofit with the financial sustainability motives of a for‐​profit” and “turn the power of the market toward social problems,” they write.

Indeed, it might prove to be the case that successful philanthropy is better accomplished through for‐​benefits than either through the “earn it and donate it” approach or through government welfare systems. What is missing here is any need for policy change. Nothing is preventing, for example, eBay founder Pierre Omidyar from engaging in philanthropy through a for‐​benefit model. The free market’s discovery process will find the optimal kinds of philanthropic organization without any governmental nudging or complexity analysis.

Bad complexity / In their enthusiasm for “the complexity frame,” Colander and Kupers overlook the possibility that it might actually lead to policies that interfere with the ability of many individuals to solve their problems from the bottom up. Nowhere do they advert to the problems that public choice economists argue are almost inevitable once the government goes beyond those neutral, rights‐​protecting functions that people like Read thought should be its limits. Rent‐​seeking factions could try using complexity analysis as a cover to get what they want, hiding their desires behind a smokescreen of abstruse math and torrents of words.

Or maybe we’re already there; consider the Affordable Care Act. MIT economist Jonathan Gruber did a great deal of highly complex analytical work purporting to optimize our health care system. After examining how numerous mandates and prohibitions would affect the various parts of the system, he then made further adjustments for those effects. His research was incorporated into the mountainous 2010 legislation that, Gruber admitted, could only be “sold” to the voters through statements of doubtful veracity. This law has led to huge distortions in the market and serious problems for many individuals.

Did Gruber not do his complexity analysis correctly? Or, is it perhaps the case that there are too many unknowns for anyone to “solve” the nation’s health policy problems, just as there were too many unknowns for Soviet economic planners to solve the problem of optimal resource allocation? In any event, I think we need to be wary of anyone who claims to have devised ideal new public policies based on his uniquely deep understanding of how complex our problems are.

Twenty years ago, Richard Epstein wrote a marvelous book entitled Simple Rules for a Complex World. Colander and Kupers never mention it, but comparing Epstein’s case for the simple rules of the common law (e.g., contract, property, tort) with the Pandora’s Box that complexity theory and nudging would open, I remain strongly inclined toward Epstein’s view that the legal system and public policy should remain simple.

In the end, the argument Colander and Kupers make for a new way of doing public policy that revolves around “complexity” analysis and governmental nudges leaves me worried. While the authors say they don’t want government top‐​down planning to rule the economy and our lives, I fear that their ideas, if taken seriously, would be useful to those who do.

Or maybe I’m worrying for nothing. After all, Colander entitled one of his earlier books Why Aren’t Economists as Important as Garbagemen?