Expansionary austerity: some shoddy scholarship

by John Q on October 24, 2011

I’ve just read ‘Tales of Fiscal Adjustment’ by Alesina and Ardagna, which appears to be the founding text for the idea of expansionary austerity. The level of scholarship, at least as it applies to Australia (which is their first illustration) is exceptionally poor, to the extent that it requires a rescuscitation of the ancient Internet tradition of Fisking. I’m going to quote excerpts from their text (about 50 per cent of the total), and intersperse them with my comments.

In 1985, a single-party left-wing government took office and launched a stabilization plan to correct the internal and external imbalances (the current account deficit was 4.13% of GDP and the total deficit/GDP ratio was above 3% in 1984).

The Labor government was elected in 1983, not 1985, in the early stages of recovery from a deep recession. Their stabilization plan, introduced in 1984, and called the Trilogy, pledged to hold tax/GDP and expenditure/GDP ratios at or below their current levels while reducing the budget deficit. The current account deficit, a subject of continuous concern, remained high throughout the period in question

The government wage bill and transfer programmes accounted for the biggest share of the adjustment … The cuts in transfer programmes were mainly concentrated on unemployment insurance.

There were no cuts in unemployment benefits. Expenditure fell because unemployment was falling. Haven’t Alesina and Ardagna,heard of automatic stabilizers?

Capital taxation was rationalized.

OK, I guess, except that ‘rationalized’ in this context, typically means ‘reduced’. In fact, the government introduced a capital gains tax which more than offset the end of double taxation of dividends

From 1983 to 1986, wages were bargained at a centralized level. The system was based on full indexation with twice-yearly adjustment, but there was a departure from full indexation in 1984 and 1986. In the negotiation process, government used tax reductions previously described to induce the union movement to accept reductions and delays in wages increments.

Absolutely opposite to the story told here, the trade-off in 1984 was in return for the (re)introduction of a single-payer health insurance scheme, a major expansion in the role of government and one that has endured to this day. In subsequent rounds, tax cuts were sometimes part of the deal, but the big trade-off was the introduction of compulsory employer contributions to retirement income funds. These aren’t counted in measures of tax revenue and expenditure, but in functional terms they are the equivalent of a social security scheme (though a regressive and badly designed one, with lots of historical inequities and complexities locked in).

Between 1985 and 1986, the nominal effective exchange rate decreased by about 19%.

This is presented as if it were a goal of government policy. In fact, this depreciation, and the current account deficits that drove it led to Treasurer Paul Keating’s famous observation that Australia was in danger of becoming a “banana republic”

Australia is a clear case of an‘expansionary fiscal contraction’. GDP grew faster during and in the aftermath of the adjustment, both in absolute terms and relative to the G7 countries. A private investment boom was associated with profits and easier access to credit following the financial deregulation process that took place in 1985–6.

This is like the story of the guy who jumps off a tall building and says, as he passes the 25th floor “All good so far”. Writing in 1998, Alesina and Ardagna must surely have been aware that, almost immediately after their story ends, Australia entered the worst recession in its postwar history. The recession was triggered by contractionary monetary policy, but its severity was largely due to the collapse of speculative investment projects undertaken by so-called ‘entrepreneurs’ who took advantage of easy access to credit to build conglomerate empires that failed in the crisis, almost taking down the banking system with them. Unemployment reached double digits in the early 1990s, and didn’t fall below the pre-adjustment level of 8 per cent (itself disastrous) for nearly a decade.

In July 1987, the same government and the same prime minister in office were re-elected by popular vote. In the April 1990 elections, neither the winning government nor the prime minister changed.

This is true, though there was a huge amount of luck and ham-fisted opposition involved. When these factors ran out in 1996, the government suffered a thumping defeat, based primarily on the recession of the early 1990s. Labor was out of office for another decade.

Overall, the description of Australian macroeconomic experience given here is unrecognisable to someone who lived through the period. The government did lots of things that gained the approval of neoliberals (global sense) but these were almost entirely microeconomic in nature.

Although this piece is full of silly errors and spurious claims, the central problem (which starts with the dating error) is that the direction of causality is reversed. The strong expansion that began in 1983 drove much of the fiscal consolidation directly, and created the political-economic environment in which tight fiscal discipline was feasible without economic contraction, and politically salable. The severe recession that began just after the triumphant return to budget surplus (when Paul Keating went from bragging that “this is the one that brings home the bacon” to observing that “this is the recession we had to have”) wiped out all of the fiscal consolidation of the previous decade – balance wasn’t restored until years into the expansion with a consolidation that produced an increase (admittedly temporary) in unemployment, as Keynesian theory would predict.

As a final observation, Alesina and Ardagna would have had a much better picture of the events they described if they had taken a list of Keating’s most famous sayings and checked back to discover the context.



John Quiggin 10.24.11 at 7:30 pm

Last para should be my comments in itals


SamChevre 10.24.11 at 8:00 pm

At the line, “A private investment boom was associated with
profits and easier access to credit following the financial deregulation process that took place in 1985–6″everything prior to that repeats before continuing. The formatting gets weird a couple sentences before, at “aftermath of the adjustment”.


William Timberman 10.24.11 at 8:03 pm

Something went terribly wrong with the cut-and-paste, it seems. Time for a reload.


Satan Mayo 10.24.11 at 8:34 pm

This post formatted by guest editor Stéphane Mallarmé.


John Quiggin 10.24.11 at 9:14 pm

Indeed, reformatting is on its way


P O'Neill 10.24.11 at 9:46 pm

It’s worth noting that the literature has moved on substantially since this paper, and in particular Roberto Perotti, who became the more recent preferred cite of the austerity crowd, has recanted.


Walt 10.24.11 at 10:15 pm

I think it would read better if the quotes from the paper were in italics (or maybe block quotes), and your comments were in the normal style.


sg 10.25.11 at 1:17 am

Is this reflective of economists’ penchant for working papers rather than good old peer review? Getting the date of the election wrong (by two years!) and ignoring the reintroduction of medicare is just classic. As a consequence all their tables are wrong (“bef.(85-86)” indeed!)

I guess they’ve never heard of Paul Keating or read anything he ever wrote or said. Their loss, really…


js. 10.25.11 at 1:27 am

WRT #8:

Actually wanted to ask: was this a peer-reviewed publication? Not that shoddy scholarship (or analysis, etc.) doesn’t get published in other fields, but getting the basic chronology wrong seems pretty egregious.


Mr Art 10.25.11 at 6:37 am

“the introduction of compulsory employer contributions to retirement income funds. These aren’t counted in measures of tax revenue and expenditure, but in functional terms they are the equivalent of a social security scheme.”

Do you mean functional as in ‘the functional macro-economic effects’? I was under the impression that these contributions were defined contribution and therefore very different from a social security scheme.


Kevin Donoghue 10.25.11 at 7:52 am

“…was this a peer-reviewed publication?”

Alesina lists it as such (see link below). But my layman’s impression is that economic journals don’t worry too much about getting mere facts right. I’d be happy to hear that I’m wrong though.



Jack Strocchi 10.25.11 at 8:22 am

What planet are these guys on?

The post-Whitlam [1]vAUS macro-eco story is a boringly conventional Keynsian morality play. Fraser-LNP and Keating-ALP both ran expansionary deficits to counter-act the recession. Hawke-ALP and Howard-LNP both ran contractionary surpluses to counter-act the recovery.

Rudd-ALP ran a deficit stimulus in the aftermath of the GFC, as per the Keynsian rule book. This helped somewhat, although most of the good work had been done by the RBA’s interest rate cuts and Costello’s prudent financial regulation.

This is exactly how the Keynsian text books prescribe it and as it should be. To the extent that AUS’s post-Whitlam macro-eco performance is creditable it is a vindication of Keynsianism.

The economic recovery under the Hawke-ALP was almost totally cyclical in nature, driven by the breaking of the drought, the general post-OPEC II global boom, some recessionary belt tightening by unionized wage earners and a somewhat reckless loosening of credit standards (“the entrepreneurs” plus housing boom).

Its true that Hawke-ALP ran a contractionary fiscal policy during the second half of the eighties, a series of surpluses and amortization of public debt (though this was shoddily funded by fire sale privatisations). But this was more or less an effect, not cause, of the economic recovery in train from the early eighties on-wards.

The AUS macro-eco story does vindicate one liberal policy, namely floating exchange rates which have greatly stabilized national income in the face of wild flunctuations in global trade and investment flows.

[1] Whitlam-ALP, under the “stewardship” of Treasurer Jim Cairns, completely bungled macro-eco policy by prescribing Keynsian deficit pump-priming to counteract structural unemployment (major factoral supply shocks) in the midst of a wage-price inflation spiral. This was not a refutation of Keynes, merely the misapplication of Keynsianism to what was basically an institutional governance problem.


sg 10.25.11 at 8:28 am

It’s not peer reviewed by any kind of standard definition, as far as I can tell. The articles are discussed by a “panel of experts.” Alesina is also a research fellow at the Centre for Economic Policy Research, which publishes Economic Policy. I wonder if Alesina is also a panel member?

In any case, their panel would probably benefit from inviting Paul Keating to join it.


Alex 10.25.11 at 9:52 am

It really is time to dissolve economics and elect a new one.


Martin Bento 10.25.11 at 10:43 am

Am I the only one who thinks the right-wing econosphere is ripe for a Sokal hoax? Give them something that fits their ideological priors – expansionary austerity would be a good one, or blaming the crisis on government encouragement of minority homeownership. Dress it up with lots of difficult math, with fundamental but not obvious errors. Or else, as here, botch the facts substantially. Like Sokal, but unlike here, have the author come out and loudly proclaim that he ratf*cked the editors, peer reviewers, and all who cited approvingly. Much good would come of this. Look how consequential Sokal was.


ajay 10.25.11 at 10:53 am

It wouldn’t make any difference, Martin. Something very similar has been happening, except that the authors haven’t been confessing that they were hoaxing, they’ve just been making testable predictions about (eg) inflation and bond rates that repeatedly fail to come true. But that hasn’t really had any impact at all on the field of economics.
In an ideal science, predictions not coming true would mean that a theory gets discarded. In a real-world science, what happens more often is that the predictions don’t come true and the theory gets tweaked to accommodate them. But in economics, the predictions don’t come true and… no one cares.


Kevin Donoghue 10.25.11 at 11:00 am

sg, is there a standard definition of peer review? Anyway Economic Policy ranks right up there at No. 6, according to:



Kevin Donoghue 10.25.11 at 11:08 am

In many ways the archetypal economic model is the Fibonacci rabbit-breeding model. It’s not meant to tell you how rabbits breed, it’s just a story to motivate some interesting mathematics.


Martin Bento 10.25.11 at 11:47 am

ajay, I think author confessions make all the difference. There was a lot of junk published in the heyday of literary theory, but the ball kept rolling till someone stood up and said “I was BSing you all! And you all fell for it!” I realize that some people exaggerate the significance of what Sokal did, but the effectiveness of the thing is hard to dispute.


ajay 10.25.11 at 12:12 pm

Martin, I take your point, but I think the difference is that economics is more open to being provably, unambiguously wrong in its predictions than literary theory – so the only way to have a lit-theory paper be declared wrong is for the author to actually stand up and say “it was all rubbish”.


sg 10.25.11 at 12:33 pm

Kevin, it says that the papers are invited and discussed by the panel:

All the articles are specifically commissioned from leading professional economists. Their brief is to illuminate topical policy issues by combining the insights of modern economics with best available evidence. The presentation is incisive and written in plain language accessible to a wide range of participants in the policy debate.

Prior to publication, each article is discussed by a Panel of distinguished economists. Summaries of Panelists’ comments are included in each volume in order to provide readers with alternative interpretations of the evidence and a sense of the liveliness of the current debate. The Panel, which rotates annually, represents a broad range of specialisations and nationalities.

Obviously the panel that “discussed” this article knew nothing about the Australian political economy. If only they could have sent it to a peer, working in Australian economics, who could have reviewed it… do we know anyone with those qualifications who was working in the field in 1998?

Under a proper peer review process, articles are unsolicited, the journal sends the article to a small number of peer reviewers chosen for their expertise on the matter, and ideally reviewer and victim reviewee are unaware of each others’ identities.

Is this how the economics profession understands the concept of peer review? Because it certainly wouldn’t pass muster in epidemiology or public health, as far as I’m aware.


sg 10.25.11 at 12:34 pm

of course, the third paragraph should be blockquoted too…


Walt 10.25.11 at 12:57 pm

sg, economics has the normal notion of peer review, though some journals are single-blind and some are double-blind. Economic Journal is not the #6 journal — the list that Kevin links to is pretty goofy (I assume because it’s generated by an algorithm).


Lemuel Pitkin 10.25.11 at 1:12 pm

This is great!

(Although it would be more conventional to put A&A in italics, and your stuff in normal type.)

A more systematic critique of Alesina and Ardagna’s work is here. Even their econometric work depends onto rather arbitrary ways of grouping countries into austerity and non-austerity piles, and a rather strange measure of expansion vs. contraction.


Kevin Donoghue 10.25.11 at 1:46 pm

Walt, no doubt that ranking (6th) is goofy, but it’s a pretty well-regarded journal nonetheless. See the Keele list for example:



Substance McGravitas 10.25.11 at 5:52 pm

I think it would read better if the quotes from the paper were in italics (or maybe block quotes), and your comments were in the normal style.

Yes, normal style is established at the top of the post and then immediately abandoned.


John Quiggin 10.25.11 at 5:57 pm

OK, I’ve done the reformat


GK 10.25.11 at 5:59 pm

John Quiggin is absolutely correct. I lived through all this.


Kevin Donoghue 10.25.11 at 6:14 pm

Thanks John, you might want to strike the “…my comments in italics.”

Format aside, I’d be interested in your take on the question raised by sg: what are the duties of journal referees/reviewers and editors in such cases?


John Quiggin 10.25.11 at 6:30 pm

I guess I would cut the editors some slack here. The referees aren’t like newspaper fact-checkers. They are looking for analytical errors and weak arguments, and would normally trust the authors to get things like dates right. If it were a single country study, the referee would normally be familiar with the country concerned, and the errors would jump out, but obviously the journal can’t go to a separate referee for each country in the data set.

I suspect the bigger problem is that an error like getting the election date might well go uncorrected even if a reader noticed it. It’s scarcely worth an erratum in itself, and a journal might not want to publish a critique like the one I’ve presented above, even if it points out some clear factual errors.


Colin Danby 10.25.11 at 7:14 pm

Econ also has weaker traditions of national and regional expertise than other social sciences. There are people who develop such expertise, but it’s easy to get away with the kind of thing discussed above. I remember going to the AEA as a grad student in 1995, after a year of work in Mexico, and being stunned by elementary data errors in papers that prominent people had whipped up on that country. You also discover, from both ends of the refereeing process, that editors struggle to find referees with place-specific knowledge.


Kevin Donoghue 10.25.11 at 7:58 pm

Given what John Quiggin and Colin Danby say (and it’s in line with my impressions as an economics consumer) the moral is obviously caveat emptor.

Ireland, 1987, is frequently cited as a leading example of expansionary austerity, but you’d have a hard time finding even one Irish economist who takes that claim seriously. I know of a retired professor who used to think there was something in it and maybe still does; also a finance professor who, however, seems a bit confused as to what the hypothesis actually is.


sg 10.25.11 at 11:45 pm

John, if that’s the case I’m kind of flabber-gasted at the expectations of basic quality in economics journals. This would be like doing a multi-country study of injury or obesity and getting basic prevalence statistics wrong – then working with them all through the paper. Essentially the mistaken date they have used sets the time point at which they judge the “intervention” to have happened, and they’re wrong by 2 years. To have done that in the first place is enormously sloppy; to have breezed it through peer review unnoticed is terrible, for several reasons:

1. They’re doing an intervention study, and nobody in the “panel” had the basic understanding of intervention analysis to challenge them on their choice of starting point for the intervention – this is a crucial concept in analysing interventions
2. A fundamental parameter of their analysis is unreferenced, as are all their statements about the properties of the different analysis periods
3. They have shown no country-specific knowledge – not even to the level of having done a literature review.

This level of scholarship is literally pulling opinions out of your arse, and it wouldn’t pass muster in even a low-ranked public health journal. That no one on the journal’s distinguished panel of experts noticed these aspects of the intervention analysis is bad enough; that these aspects were thoroughly wrong both before and after review is really really poor.


John Quiggin 10.26.11 at 12:07 am

@sg The result is appalling, but it seems to me that the responsibility for this kind of error (easily checkable facts like election dates) rests with the authors and no one else. The panel is certainly to blame for the generally weak analysis, as discussed above.

It’s also depressing to observe that the errors in this trainwreck of a paper have only come to attention because it has been taken up, leading me and others to scrutinise it. Presumably, the literature is full of similarly erroneous papers no one can be bothered to look at.


John Quiggin 10.26.11 at 1:10 pm

Thanks everyone for patience with my formatting disasters and some useful links.

Among them, the Perotti link was va useful, and I was very pleased to find this quote from Keynes which sums up the prescriptions of the ‘hard Keynesianism’ Henry and I have been advocating, and which is cited by Jayadev and Konczal in the critique of A&A, linked by LP at 14

“The boom, not the slump, is the right time for austerity at the Treasury.” –
‘Keynes, John Maynard. 1937/1983. Collected Writings of John Maynard Keynes, vol
21. London: Palgrave Macmillan’

The original source appears to be a newspaper article, How to Avoid a Slump’, The Times (11-14 January 1937). I wonder if that’s online somewhere


Watson Ladd 10.26.11 at 3:59 pm

So an article can be published with basic factual mistakes, and the journal will do nothing about it in economics? The authors better have a good explanation for a few of these howlers.


Gabriel 10.26.11 at 6:54 pm

For the Keynes article, the Times has a digital archive (see here: http://gale.cengage.co.uk/times.aspx/) but your library has to have purchased a subscription in order to get the article.


Norwegian Guy 10.26.11 at 11:09 pm

What’s your problem with Robert Fisk? It turned out that the warmongers that were “Fisking” him were wrong, Iraq had no weapons of mass destruction. So while I know you don’t intend it that way, the use of this term is both self defeating, and a disservice to a great journalist.


ajay 10.31.11 at 1:33 pm

Presumably, the literature is full of similarly erroneous papers no one can be bothered to look at.

I wouldn’t be at all surprised: one of my happiest memories from undergraduate days is discovering that the head of my department had got an equation the wrong way up half way through a string of them in a paper in Phil Trans R Soc, leading him to all sorts of wholly erroneous conclusions.

I wonder what would happen if you got a few students and gave them some random published papers to fact-check – for simple things like dates and places. Kind of an Innocence Project but for stupidity.

Comments on this entry are closed.