Why Would Anyone Want Austerity?
By Michael Munger
- A review of Austerity: When It Works and When It Doesn’t, by Alberto Alesina, Carlo Favero, and Francesco Giavazzi. Princeton, NJ: Princeton University Press. 2019.1
It is easy to see why the diversity of the Italian public finance tradition appealed to Buchanan. In fact, Buchanan would often point out that the continental thinkers, by which he meant Knut Wicksell, Bruno Leoni, and Maffeo Pantaleoni, among others, were “far ahead” of English-speaking theorists. Committed to an historical and evolutionary theory of the Roman Jus Civile and the ancient common law, this tradition concluded that our political world had collapsed into a system that restrains individual liberties. The more extreme version of this view (as later summarized by Carlo Lottieri) went so far as to question whether liberty and democracy are ultimately compatible. The question—and this will sound familiar to students of Buchanan—is under which conditions it is legitimate to coerce citizens. The gold standard for such legitimacy is consent, of course: if someone signs a contract voluntarily, they can be sanctioned for violating their promise. The question was when, or if, it is possible for citizens to consent to being coerced by majority rule.
But there is quite a bit more to the Italian tradition of public finance. Having 2,500 years of history with budgets and with the effects of different taxing and governing arrangements creates a fecund setting for research. This raises one of the central paradoxes of Italy as a nation. Italy belongs in every conversation over food, fashion, cars, art, even engineering and high quality manufacturing. It is also a notorious pit of political dysfunction and economic nonsense. Yet the view of public finance in Italian scholarship is generally sensible, informed by theory, and empirically and technically excellent.
The lead author of Austerity, Alberto Alesina, has long lived in the United States; other than a brief stint at Carnegie Mellon University from 1986 to 1988, all of his time has been spent at Harvard, with a few visiting gigs interspersed. He has been enormously productive, and some of his work is among the most-cited in all of political economy, with more than 50,000 references according to Google Scholar.
And it is, make no mistake, political economics that interests Alesina. Austerity is an attempt, working with Carlo Favero and Francesco Giavazzi as coauthors, to define austerity, determine whether it works, and explain the political reactions to it. They (plausibly) define austerity as a sizable reduction of government deficits (spending in excess of tax revenues) and the stabilization and a commitment to eventual control or reduction of debt (the accumulation of deficits financed by borrowing across years) using some combination of spending cuts and tax increases. There is no hint of the sense often given the term in some of the European press, where “excessive”, “overly restrictive”, or “damaging” are deployed in the description.
Still, why would anyone want austerity? Isn’t it a bad idea? Actually, yes, it is a bad idea. So is cutting open someone’s abdomen with a razor sharp blade. But if that person’s appendix is about to burst, the knife cut is the best of bad options. Alesina, et al begin with a sensible observation: “If governments followed adequate fiscal policies most of the time, we would almost never need austerity.” They do break out the Keynesian dogma, saying
- Economic theory and good practice suggest that a government should run deficits during recessions—when tax revenues are low and government spending is high as a result of the working of fiscal stabilizers such as unemployment subsidies… (page 1)
But they note that even this kind of deficit spending could be reduced with foresight:
- … forward-looking governments might want to accumulate funds for “rainy days” to be used when spending needs are temporarily and exceptionally high. If government followed these prescriptions, austerity would never be needed. (page 1)
So, austerity is a second best policy, contingent on a particular kind of government failure. Alesina et al note that the effectiveness of austerity is controversial, with the discussion in the press “often taking a very ideological, harsh, and unproductive tone.” (page 3) The reasons given for why austerity should be selected by a government, or should be imposed as a condition of extension of loans by creditors, are obvious: (a) the ratio of debt to GDP has grown perilously large, raising questions about whether even the existing debt can be repaid, and (b) crises, fiscal needs arising from wars or major economic downturns in the business cycle or in currency exchange markets.
The critics claim that austerity is a moralistic, punitive policy designed to cause pain for excess deficits, one that fails even on its own logic because it actually pushes debt to GDP ratios higher rather than lower. The argument is that GDP shrinks sharply as government spending is cut, and even if debt falls, GDP falls by more which worsens the problem and causes an economic storm.
In a way, that’s the end of the story. Because Alesina, et al. are able to give a decisive resolution to the controversy: not all austerities are the same. In fact, there are two very different types: a focus on raising taxes, and a focus on cutting spending. When it comes to imposing an austerity of “sharply increased taxes” variety, the anti-austerity activists probably have the best of it. But when the austerity takes the form of large-scale cuts, not just in budgets but in entire programs, the larger weight of evidence by far falls on the “austerity works” side of the scale.
This may sound obvious, but that’s because Alesina and company have explained it clearly. That distinction is not generally made, certainly not by critics of austerity and often not even by its supporters. It is perfectly clear why tax-focused austerity would fail, because it chokes off any hope of expanded growth starting to shrink the debt. Conversely, a cut in government spending represents a credible commitment, at least in the near term, to reduce the debt. Yes, it’s painful, but Alesina et al. note that the problem for rescheduling debt is made much worse by uncertainty about when fiscal stabilization will take place. Just the act of cutting government spending, and possibly even just cutting the rate of growth of government spending, is a costly signal that the administration has some political will, and enough electoral power to implement an unpopular policy.
One of the central questions, and in fact insights, in the book is about whether austerity can be expansionary. The narrow conventional wisdom is that austerity is always contractionary, and the only question is by how much. And then the debate would turn on whether austerity is “worth it,” given the pain it causes. Alesina, et al. show straightforwardly that this is not true empirically, and explain why it would not be true theoretically.
The simplest way to describe the logic of an “expansionary austerity” is that it is the opposite of Frederic Bastiat’s “broken window” fallacy. The naive Keynesian [see John Maynard Keynes] prescription is that we can get growth at no cost as long as there are unemployed resources, leading to Paul Krugman’s famous observation that an alien invasion would stimulate the U.S. economy, because it would spur a large “investment” in defensive infrastructure. The problem is that this spending has an opportunity cost; you don’t get something for nothing.
Just as an increase in government spending has hidden costs, Alesina et al. note that cutting government spending can have hidden benefits. Remember, for Keynesians, Y=C+I+G. If you cut G, it must be true that Y falls. Unless… unless C and I expand by more than G falls. As Alesina et al. put it:
- The possibility of expansionary austerity does not mean that every time a government reduces public spending the economy expands. The term instead implies that in certain cases the direct output costs of spending cuts is more than compensated for by increases in other components of aggregate demand. (page 5)
The authors note that one possibility might be that the government is responsible enough to impose the “reduced government spending” flavor of austerity during an expansion. But this might simply mean that austerity hurts, just not enough to change economic growth from positive to negative. Alesina et al. choose a more restrictive, and sensible, definition: expansionary austerity occurs when growth following an austerity regime increases, or is maintained, at or near the top of the distributions of growth paths that particular economy is capable of generating. This is a rather nuanced problem, of course, because it requires the estimation of the country’s capacity for growth, and an estimation of the counterfactual: what would have happened if no austerity had been imposed? Expansionary austerity means austerity increases growth over what would have happened without austerity, which in the standard naive Keynesian model would be impossible.
I don’t mean to say that Alesina et al. reject Keynesianism; they don’t. Instead, they adopt the sensible extensions of the “New Keynesian” perspective on expectations, and also “supply side” considerations about availability of credit. The connection between expectations of creditors and the availability of investment is obvious: business will be more willing to invest, and creditors to lend to both the government and to private businesses, if austerity has ended the rapid expansion of debt load.
The book is mostly trying to make an empirical contribution relevant to policy makers. The question is not whether austerity always hurts, and how much, but rather which countries might be candidates for an austerity treatment and under what circumstances.
The simple answer is that countries should generally consider austerity when the cost is lowest, which will almost always be during a period of rapid growth. To be fair, this is also the standard “old school” Keynesian answer, where spending is cut back during expansions and accelerated during recessions. But the mechanism is different here, because austerity is a signal of a permanent and widely spread out reduction in government spending, rather than countercyclical “leaning into the wind.”
The problem with the sensible, simple answer is that austerity is politically disastrous. Alesina et al. quote Jean-Claude Juncker, then President of the European Commission, as saying “We all know what are the policies which we should follow, but we do not know how to introduce them and then be re-elected.” (page 8). This idea, which the authors note is held as “vastly obvious” among both academics and politicians, is that voters always punish leaders who even propose, much less implement, austerity policies.
But empirically it is far from obvious. One is reminded of the observation made by Josh Billings (often misattributed to Mark Twain or Artemus Ward, which given the content of the observation adds irony): “I honestly believe it is better to know nothing than to know what ain’t so.” Those who know that austerity always result in electoral losses are just as bad as those who know that austerity always causes economic harm, because both groups “know” something “what ain’t so.”
As Alesina et al. note:
- If one looks at the data more closely, this view is much less supported by the evidence than one may think, even outside of traditionally fiscally conservative countries like Germany. Many governments that have implemented tight fiscal policies and reduced deficits have been reelected, and the other way around, fiscally careless governments were punished by the voters…. The evidence does not support Juncker’s statement: many governments have been able to implement austerity policies and be reelected. Of course this does not mean that governments that cut spending or raise taxes are always reelected; it means that reality is more subtle and complex…. (page 8)
How Much Austerity? Four Contributions
To aid policy-makers by going beyond “never austerity” or “always austerity” to “when austerity, and how much?” Alesina et al. claim to make four contributions in four areas. The first is the availability of their data online, documenting nearly 200 austerity plans in 16 OECD countries from the late 1970s to 2014. Starting with 3500 individual fiscal measures or actions, they classified such actions into 27 categories, and then further aggregated the data into 15 measures (transfers, direct taxes, credits, indirect taxes, deductions, and so on). These data are publicly available, in quite well documented and disaggregated form.2 The analysis in the book takes place at a more highly aggregated level, which means that the publicly available data can added up differently, or analyzed in their existing components, a great benefit for researchers who want to study the nuts and bolts of austerity measures and their history.
The second contribution is method. A common understanding of the nature of policy is to imagine “shocks” to a mechanical system. But this view ignores two problems: first, austerity programs are multiyear and can be phased in in stages, often in complicated ways. Second, the strategy is not executed like a computer program or contract written in advance; rather, adjustments are made continuously, based on both economic and electoral feedback.
The third contribution has already been mentioned, but it is important. Analysts must distinguish between austerity based primarily on tax increases and austerity that takes the form of shrinking the size and scope of government. “Supply side” effects are simply ignored in the classical Keynesian model, and even in New Keynesian versions of the model are incorporated only as an afterthought. The book offers detailed summaries of the results in individual nations, but there is also an aggregated analysis that tries to reach an overall conclusion, admitting all the problems of heterogeneity of cases and endogeneity of policies. The result is quite clear: countries that implemented expenditure-based austerity plans either suffered little measurable costs in terms of growth or actually expanded, after a period of only one year. Countries that implemented tax-based austerity did suffer, in some cases sharply and for several years.
In terms of the Keynesian view, cutting government spending to effect a 1% of GDP reduction in debt (in simulation) saw increases in consumption and investment spending that mostly, and in some cases completely, offset the “loss” in GDP. An analogous simulation of a tax increase that reduced the debt by 1% of GDP was not offset, and in fact in some cases the resulting recession actually expanded the debt-to-GDP ratio within two years. Interestingly, this latter effect, “tax increases reduce GDP and actually increase debt” is the inverse of the standard supply side argument often decried by the left. Incentives and expectations matter, after all: raising taxes without showing the ability to constrain the growth of government spending does nothing to increase business or consumer confidence.
Finally, the fourth point the authors make is about prudence. It is not true that austerity always results in electoral loss for the party in power, though it can be dangerous and there are examples of voters blaming politicians. But it may be just as dangerous to do nothing, in hopes that voters will credit neglect when the economy is clearly suffering from burgeoning debt. In a way, this is a hopeful note on which to finish. When it comes to unmanageable debt, the party in power is better off implementing austerity, providing it can do so during an economic expansion and also provided that the program focuses on sharply cutting government spending.
While these conclusions are clearly contentious, the authors have given us a detailed analysis, backed by publicly available data that can be examined and further analyzed to glean details. Austerity is the new standard in the fiscal policy literature on deficits and public policy.
 The book website is: https://press.princeton.edu/titles/13244.html. The downloadable online data appendix is at: https://www.aeaweb.org/doi/10.1257/jep.33.2.141.data.
Read more of Michael Munger’s writing at the Econlib Archive.