That’s the title of a new book (Palgrave MacMillan) by Clifford Winston, Senior Nonresident Fellow in Economic Studies at Brookings.
From the Introduction:
Economists have long argued that government should complement markets by implementing a policy intervention if markets fail. For example, if a firm monopolizes a market by using anticompetitive strategies and sets supracompetitive prices, they hold that the antitrust authorities should bring a monopolization case against the firm and seek
a remedy that eliminates the cost to consumers from elevated prices.However, economists generally have not considered that markets could complement government by self-correcting when government does not intervene or when a government intervention fails. For example, if the government loses a monopolization case or decides not to bring one, a new entrant could reduce consumer prices by using a new technology
to develop a product that competes effectively against the monopolist’s product.…
…the pendulum has not moved when it comes to government failure. Economists have long acknowledged government failures but tend to associate them narrowly within their field of specialization.
For example, government failure results from inefficient regulations and rules according to industrial organization economists; from price controls according to health economists; from rent controls according to urban economists; from tariffs and quotas according to trade economists; and so on. And in most cases when a market failure occurs, more economists would contend that imperfect government action to correct an alleged
market failure is preferable to no government action, instead of no government action being preferable to harmful government action.Because economists generally believe that market failure is a greater social problem than government failure, they are unequivocal about recommending a government policy intervention to address an economic goal, such as correcting an alleged market failure due to imperfect information about a product, or a social goal, such as reducing poverty.
My view, however, is that economists should be cautious before recommending any policy intervention because government policies often fail in practice, even if they should work in theory, and they are rarely reformed in the long run. In contrast, markets may fail in the short run but could self-correct in the long run, so economists should be more cognizant
of markets’ capabilities.
Here’s the table of contents:
1 Introduction 1
Part I Government Failure
2 Defining and Measuring Government Failure 9
3 Government Policies, Institutions, and Benchmarks for Assessing Government Performance 15
3.1 Government Institutions to Formulate and Implement Public Policies 18
3.2 Benchmarks for Assessing Government Policies 19
4 Empirical Evidence of Government Failures 25
4.1 Government Policies that Reduce Social Welfare 26
4.2 Government Policies that Have a Negligible Effect on Social Welfare 36
4.3 Government Policies that Increase Social Welfare at Excessive Cost 39
4.4 Common Features of Government Failures 42
4.5 Pushback From Policymakers and Their Supporters 45
4.6 Potential Policy Improvements to Reduce Government Failures 475 The Failure to Explain Government Failure 53
5.1 Theories of Government Successes 54
5.2 Theories of Government Failure 55
5.3 Empirical Evidence on Explanations of Government Failure 57
Part II Market Corrections
6 Defining and Illustrating Market Corrections 71
6.1 Why are Market Corrections Possible in Theory? 72
6.2 Illustrating Market Corrections 74
7 Empirical Evidence of Market Corrections to Address Economic and Social Goals 81
7.1 Market Corrections that are Independent of Government Policy 82
7.2 Market Corrections Facilitated by Government Policy Reforms 91
7.3 Preliminary Market Corrections That Have the Potential to Produce Large Benefits 94
8 An Assessment of and Ongoing Challenges for Market Corrections 101
8.1 A Comparison of Government Failures and Market Corrections 101
8.2 Challenges for Markets 108
Part III Implications for Economists and Policymakers
9 Implications for Economists 119
9.1 Distinguishing Between Market Failure and Government Failure 120
9.2 Comparing Market and Government Long-Run Performance 124
9.3 Identifying a Market Failure but Ignoring Government’s Failure to Address It Efficiently 128
9.4 New Empirical and Theoretical Contributions by Economists to Improve Policy Debates 133
9.5 Include Extensive Discussion of Government and Market Performance in Course Content 13510 Implications for Public and Private Sector Policymakers
10.1 Government Failures in Multiple Branches of Government
10.2 Could Journalists Help Bridge the Gap Between Economists and Policymakers?
10.3 Recommendations for the Private Sector
Epilogue: The Challenge of Changing Attitudes Toward Markets and Government in a Divided Nation
Given the timing of the publication, one might wonder how policy-making in the Trump 2.0 era fits in with the prescriptions in the book. Readers interested in that question could well turn to the Epilogue.
(As a side note, Dr. Winston, along with Dr. Robert Crandall, was my supervisor as a RA at Brookings – right after working for Peter Navarro).
Without having read the book, I may be saying something Winston has already written, but here goes.
Regulatory capture and lobbying are among the reasons that government intervention doesn’t often correct course once it has failed. In competitive markets and in nature, there is a continuous effort to get the better of the other guy. In regulation, the greedy-guts private actor adjusts quickly, but regulators do not. Regulators, and their pokitical masters, should take a lesson from nature. Lina Khan showed that a committed regulator can move reasonably quickly, despite the built-in impediments to quick action.
Again, not having read the book, I don’t know whether this next criticism is apt, but here goes.
Letting markets work where government has failed sound like neoliberalism and like wishful thinking. The claim is that markets have already failed; now we’ll rely on them to undo their own failure, based on the logic of competition. The logic of competition is based on Econ 101 assumptions, including the assumption of perfect competition. It’s a question-begging argument in the real world. What we need is to reinforce regulators against capture and to give them rules which allow faster response when market manipulators try to skitter away from existing regulations.
But maybe I should read the book.
Meanwhile, here’s evidence that market power is alive, well and causes inflation:
https://cepr.org/voxeu/columns/granular-origins-inflation
Lead-steer firms use their power to boost prices, which then generalizes to higher prices across their markets. The contention of the authors is that the market power of leading firms slows the transmission of monetary tightening in reducing inflation.
Macroduck: On capture, see Sections 5.2-5.3 for somewhat skeptical evaluation.
Off topic – time to obsess over mid-term polls:
Way before polls are very informative about election outcomes, there’s a recent poll causing palpitations in both parties. It’s a Marist poll, which tend to be pretty good. It shows a whopping 14 ppt advantage for Democrats in a generic House ballot for next year’s elections:
https://maristpoll.marist.edu/polls/a-look-to-the-2026-midterms-november-2025/
That’s big, and among independent voters, Democrats have a huge 33 ppt lead.
Here’s the thing – the Marist poll is a big old outlier. The average among polls covered by RCP gives Democrats only a 4.8 ppt advantage:
https://www.realclearpolling.com/polls/state-of-the-union/generic-congressional-vote
In the RCP generic House polling average just prior to the 2020 election, Democrats had a 6.8 ppt advantage over Republicans. In that year’s election, Democrats won 222 of 435 House seats, just squeaking into the majority. There’s been one redistricting since then, and another partial redistricting is (maybe) underway right now, so it would unwise to extrapolate from 2020 results to next year’s. There is also still plenty of time for voters’ preferences to shift. However, based on polling so far, the race to control the House looks close. Unless, of course, one ignores the average and looks only at the Marist results.
Off topic – Miran and bank regulation :
https://www.federalreserve.gov/newsevents/speech/miran20251119a.htm
Newly installed FOMC gig worker Stephen Miran says that the Fed’s balance sheet is largely a function of bank regulation, and that once bank regulation is “right sized”, the balance sheet can be reduced. Okay…
My understanding is that the size of private banks’ balance sheets is a function of prudential regulatory requirements and that the size of the Fed’s balance sheet is a function of demand for liquidity. I’m sure Miran hasn’t confused the two, but still… Demand for Fed liquidity certainly reflects banks’ needs, but to a much greater extent it reflects Treasury’s needs and general market needs. We’ve already seen evidence that Fed asset holdings have fallen to the point that seasonal fluctuations can cause a liquidity scare. And Miran thinks banks should be deregulated to the point that the Fed can reduce its portfolio, while the rest of the Fed policy-making apparatus anticipates adding to reserves in the near future.
How much bank deregulation does Miran think we should have? It’s gotta be a lot of deregulating if it will mean a smaller Fed portfolio than today’s. I have to wonder, which bank does Miran think is going to pay him millions for this favor after he leaves government service?
Would be worth a detailed post by Menzie.
You’d get a different approach if you used a portfolio approach to ideologies, ie, balancing three different objective functions in a political context with declining marginal utility of income…
well we politicians dont care about much more that getting elected or re-elected, or just avoiding bad press if their priority is being in office (and it is)
Pay also matters, both official and ‘non-official’.
But to my point, which objective function should a politician follow to be elected or re-elected? Egalitarian (Mamdani)? Liberal (ie, fiscal conservatives like, say, Sherrill, Spanberger)? Or conservative (Mike Johnson, MTG)?
The politician has three separate, non-reconcilable objective functions to follow. How to choose? How to balance? What’s the difference between rhetoric and substance? And how are promises enforced as a result, that is, how are principal-agent issues resolved in a three-objective function world?
Once you frame the question like that, you arrive at a different model from Winston’s above. It’s not that Winston is necessarily wrong, just not complete.
Btw, note that the fiscal conservatives are on the left, as I have been saying for a very long time.
Will we ever rationalize our ag sector? We have too many farmers farming. We have ag welfare for sentimental reasons as well as good reasons. Can we use government to buy out some farms, let the land go fallow or used for solar installations or other non-ag purposes, and greatly cut the ag welfare?
we cant mention that some politician need their votes
The idea that end stage capitalism can ever function for anything else than its design (extraction and exploitation) is a bit naive – or simply evil.
very naive, not sure how you can expect those who run companies to do more than just make the company profitable as possible, by any means possible (legal or illegal means), and if there are penalties to pay for using what is unscrupulous means to improve profits as long as they arent so big that profits are badly impacted by the penalties, then they wont care (case in point a lot of penalties today can be taken off the companies tax return, and many companies dont even have to admit they violated the law either in court). this is where we have been for decades. companies get caught doing illegal things, they make the news (bad PR), but continue on as long as the matter is in court (usually years if not decades)
Or libertoonian. End stage capitalism will be a form of feudalism, with most of us being serfs.
Off topic – AI usage:
The Census Bureau has added a question about use of artificial intelligence by businesses to its Business Trends and Outlook Survey:
https://www.census.gov/newsroom/press-releases/2025/btos-nov-20.html
The press release says data from this new question will be released starting next year, but the early results are circulating in the press already. Reportedly, business use of AI in actual production has recently declined:
https://www.economist.com/finance-and-economics/2025/11/26/investors-expect-ai-use-to-soar-thats-not-happening
The same caution applies here as does in the case of all time-series data: one data point doesn’t make a trend. We have a good bit of data on uptake and use of AI, most of which shows increased use. Here’s a pretty good survey:
https://www.brookings.edu/articles/how-are-americans-using-ai-evidence-from-a-nationwide-survey/
It’s early days for AI, so the Brookings survey’s finding that use is on the rise is what one would expect, and the fact that Census has found even a brief pause in adopting AI is kinda spooky. Certainly, once knowledge of how AI can be productively used is more widespread, adoption by business will continue.
Still, the whole “AI is our future” story has picked up another small dent.
Informal poll. How many readers on this site have used AI on a semi regular interval in order to complete a task or improve efficiency? I have used it regularly, but cannot say it really completed a task or improved efficiency yet. It is still a curiosity without a payoff. And i have actually used it quite a bit.
Nope. Too lazy to learn new stuff.
Its an advanced google search that will process and write out a draft – rather than you having to read the links and synthesize the text yourself. I know someone who writes a fair amount of recommendation letters for graduate students – and he claims it cuts the work from 4 hours to 1 hour. The big problem is with the quality of the draft, it has not got much capacity to sort input quality. So whenever it gives you information you didn’t know you have to take the time to dig into sources and evaluate their trustworthiness. Few people will do that, so the quality of the output is Fox “news” level.
China will eat our lunch on this. They have systems that give you 90% of our systems but at less than half the cost. That is what will be commercially viable.