Perfect markets make a nice fairy tale, but they don’t match reality. And few strawmen have been as repeatedly slain as the idea that the case for markets depends on market perfection, and thus the inevitable failure of real-world markets to match this textbook abstraction undercuts the argument for using markets. Some of the strongest defenders of the market system, like F. A. Hayek and Israel Kirzner, reject ideas like perfect markets, perfect information, perfect competition, and so forth. Their argument for the use of markets rests not on some abstract perfection of markets, but instead on the real-world dynamism inherent to the ongoing and evolving market process. Markets are useful not because in a market-driven world there are no $20 bills on the sidewalk. Markets are useful because they create the right environment for finding those $20 bills.
One real-world friction in real-world markets is price stickiness. In the fairy tale version of perfect markets, prices adjust instantly. In the real world, prices can be sticky – they might not change, or change slowly. One reason for this is transaction costs. Sometimes, changing a price isn’t free. The textbook example of this so-called “menu costs.” Even if the costs of various foods and ingredients change, restaurant prices can be sticky and remain unchanged. In order to change their prices, restaurants would have to print out an entirely new set of menus with the updated price for each dish. This costs money and time. If the price of potatoes slightly increases, it’s often not worth the time and effort for a restaurant to print out a new set of menus with updated pricing for every dish that includes potatoes.
But just like asymmetric information, transaction costs have a half-life. In markets, there is an incentive to find ways to reduce transaction costs and thus reduce the stickiness of prices – because finding ways to reduce transaction costs is itself a money-making opportunity. Menu costs are an example of this too. One way I’ve seen restaurants get around menu costs is by simply not having a listed price for particular menu items. If a restaurant in a beach town frequently serves fresh and locally caught fish or lobster, they might face significant fluctuations in costs for those items. To accommodate this, they frequently list such dishes on the menu as “market price” rather than a set dollar amount.
More recently, I’ve seen many other restaurants put their menus on digital displays rather than having them printed out, and some have dispensed with physical menus altogether and replaced them with a QR code at each table. You scan the QR code with your smartphone, and it opens up a website with the most recent menu. This drastically reduces the transaction costs associated with menu pricing, and makes prices more flexible. Price stickiness is a real problem – but at the same time, the very existence of that problem provides a market incentive to find solutions. Hence Arnold Kling’s dictum – “Markets fail. Use markets.”
On the other hand, there’s also an issue with policy stickiness. When governments create a policy to try to solve some social problem, those policies themselves become sticky. It’s surprisingly easy for people to overlook this issue. James C. Scott’s fantastic book Seeing Like a State provides an extended look at how policy interventions go awry. Toward the end of the book, he provides a few takeaways that might help improve the situation, such as:
Favor reversibility. Prefer interventions that can be easily undone if they turn out to be mistakes. Irreversible interventions have irreversible consequences. Interventions into ecosystems require particular care in this respect, given our great ignorance of how they interact. Aldo Leopold captured the spirit of caution required: “The first rule of intelligent tinkering is to keep all the parts.”
It’s not that this is bad advice in the abstract. But the idea that interventions “can be easily undone if they turn out to be mistakes” is less compelling when one takes into account that policies, too, are sticky. In practice, it’s often extremely difficult to undo interventions no matter how mistaken they turn out to have been. Policies become sticky because, as Pierre Lemieux frequently points out, any government policy necessarily benefits some at the expense of others. This quickly turns into a public choice problem. As soon as the government implements some kind of intervention, it creates a new interest group that will be invested in keeping that intervention alive, while the benefits of ending that intervention are so dispersed that there’s nobody in particular who has a strong incentive to try to put an end to it. It’s not for nothing that Milton Friedman quipped “‘Nothing is so permanent as a temporary government program.”
This quip does overstate things – not all policies are so sticky as to become immovable objects. But it’s a real problem. One classic example is the mohair subsidy. This program was initially implemented to ensure that the United States military would always have an adequate supply of wool for their uniforms. But eventually, the military stopped using this wool in their uniforms and began using synthetic materials instead. Nonetheless, the federal government continued to spend tens of millions of dollars a year subsidizing mohair production long after the initial rationale for doing so was gone. The program was eventually (mostly) eliminated – over four decades after the switch to synthetic materials.
This report from 1993 describing the ongoing efforts to eliminate these subsidies includes a comment from Senator Charles Schumer, who mentions that he’s been spending years trying to undo this policy. If ever there was a policy that should be “easily undone,” you’d think this one should be about as easy as it gets. But policy stickiness can be such a strong force that even something as ostensibly straightforward as “stop spending tens of millions of dollars per year subsidizing something you stopped needing decades ago” requires years of intensive effort to finally achieve. Pace James C. Scott, “interventions that can be easily undone if they turn out to be mistakes” are only found in fairy tales, and not in reality.
While markets provide an incentive to find ways to offset and lessen price stickiness, politics provides incentives for the beneficiaries of policies to make those policies as sticky as they possibly can. In markets, you can make money by finding ways to reduce transaction costs. In politics, you protect your largess by ensuring transaction costs are as high as possible. In the real world, price stickiness is the proverbial speck and policy stickiness is the proverbial log.
READER COMMENTS
David Henderson
Oct 2 2024 at 10:32am
Excellent piece. It weaves together a lot of insights.
David Seltzer
Oct 2 2024 at 4:37pm
Kevin: Nice work! “When governments create a policy to try to solve some social problem, those policies themselves become sticky.” I suspect much of it is the result of regulatory capture. Environmental Working Group reported 10,000 farmers got $11 billion in subsidies over 39 years. Some received payments even though they haven’t worked a farm in years. The aphorism, “There’s is a pair of them in it” applies.
Thomas L Hutcheson
Oct 2 2024 at 5:44pm
I’d amend Kling’s dictum to
“Markets fail. Government fails
Use both carefully.”
Monte
Oct 3 2024 at 11:18pm
The problem with this dictum is that markets are dynamic and self-correcting, while policymaking is, shall we say, an uncomfortable science inherently subject to policy decay.
Markets, fail, use markets. No amendment necessary.
Robert EV
Oct 3 2024 at 11:18am
Automobiles didn’t immediately replace the horse and carriage, and throw out of business the buggy whip makers. Subsidies and other monetary interventions should be built with gradual phaseouts (not phaseout cliffs). So if an intervention is no longer needed for its original purpose it fades away slowly enough that everyone dependent on it has more than enough time to switch to something else. And if it is still needed for its original purpose, it can be re-upped with a congressional vote. Votes that wouldn’t be there just to keep it as a subsidy.
For entitlements that are expected to last forever (e.g. social security) you may not include a sunset, but on the other hand it wouldn’t hurt to do so as the votes would presumably be there to re-up.
This is the reason bills such as the TCJA automatically sunset (with a cliff, not a phaseout), right?
David Seltzer
Oct 3 2024 at 2:41pm
Robert wrote, “Subsidies and other monetary interventions should be built with gradual phaseouts (not phaseout cliffs).” Yeah well; rent seeking is a byproduct of political legislation and government funding. Politicians decide funding and subsidy allocation. Occupational licensing requirements are another example. I fail to see any gradual phaseouts or cliff phaseout.