The smugness multiplier

Paul Krugman, the flagbearer of Keynesian economics, arrives in London this week to debate the impact of fiscal austerity on economies.  And he is feeling pretty smug.  The latest IMF global report (see IMF WEO Chapter 1) on the state of the world economy appears to admit that it was wrong and Keynesians, like Krugman, were right, namely that the so-called ‘fiscal multiplier’ is much larger than previously thought.

The fiscal or Keynesian multiplier argues that for every change in government spending in an economy, there is corresponding change in consumption and national output.  So if government spending is cut , there will a reduction in GDP of some amount.  The IMF and other analyses have previously suggested that the multiplier is relatively low, say between 0.5-0.7.  So a cut of $10 in government spending will generate a much smaller fall in real GDP ($5), other things being equal.  That’s not much if private investment and consumption rises to more than compensate for that reduction.  So ‘fiscal austerity’ will not be very damaging to economic growth and governments can get on with cutting spending and raising taxes to reduce budget deficits and get public debt levels down.

Well, it now seems that the IMF reckons it got its estimates wrong and that in looking at the data for 28 countries from 2009 to 2013, the multiplier will turn out to be much higher, say between 0.9-1.7.  Thus a 1% cut in government spending will produce up to a 1.7% reduction in real GDP growth.  So fiscal austerity is not working.  Indeed it is making things worse.

As Krugman put it in his blog: “I and others have been arguing for a while that the experience of austerity in the eurozone clearly suggests pretty big Keynesian effects.” and now the IMF agrees, according to the evidence of the graph below.  It shows that the more government spending cuts made by a government (Greece -16%), the worse is the decline in real GDP (Greece -19%).

Indeed, this idea has now been latched onto by the UK’s Trade Union Congress (TUC).  The TUC calculates that if the UK fiscal multiplier is as high as 1.3, the middle of the IMF new range,then the UK government’s fiscal cuts have reduced UK real GDP by 2% since 2010 (

Krugman claims that this flows from Keynesian theory, namely that cutting government spending to reduce budget deficits will keep an economy in depression when effective demand is so low: “in the little models I and others were using, we made some very striking predictions about how the world would work post-crisis – predictions that were very much at odds with what other people were saying. We predicted that trillion-dollar deficits would not drive up interest rates; that tripling the monetary base would not be inflationary; that cuts in government spending, rather than helping the economy by increasing confidence, would hurt by depressing demand, with bigger effects than in normal, non-liquidity trap times.  And guess what: the models seem to work. It appears that I wasn’t just a successful self-marketer, that I really did and do know something. 

He concludes “So I’m feeling a bit “smuggish” about all this; well, I’m only human. ….So there really was no good reason to be surprised by large fiscal multipliers. They were a predictable consequence of the kind of crisis we’re in; and the unjustified assumption of small multipliers has helped make the crisis worse.”

But hold on.  Before we bow down before JMK and Richard Kahn’s multiplier, consider a couple of issues.  First, is the IMF evidence that conclusive? Well, not according a FT analysis ( As the paper put it: “An exercise by the Financial Times to replicate and evaluate the IMF’s work, however, showed that the results suggesting very large multipliers – the relationship between deficit reduction efforts and growth – do not easily stand up to a different choice of countries or time period.”  The FT goes on: “For the countries where the full data are available on the IMF website, the results lose statistical significance if Greece and Germany are excluded. Moreover, the results were presented as general but are limited to the specific time period chosen. The 2010 forecasts of deficits were not good predictors of errors in growth forecasts for 2010 or 2011 when the years were analysed individually. Its 2011 forecasts were not good predictors of anything.”

Indeed, two leading Keynesians were also not convinced.  Jonathan Portes, director of the UK’s National Institute of Economic and Social Research, worried that cross-country studies with small samples never prove anything, even though he strongly believes that multipliers are large.  Professor Carlos Vegh of the University of Maryland said lots of evidence suggested that multipliers in countries would differ greatly and “the whole exercise of trying to forecast growth for many different countries using essentially a single multiplier, whatever the value may be, is, in and of itself, an exercise in futility”.

And when we go back and analyse previous estimates of the Keynesian multiplier, including those by the current IMF chief economist, Olivier Blanchard, we find that the fiscal multipliers will vary widely if the time periods are altered.  Indeed, the IMF notes in its report that “earlier analysis by the IMF staff suggests that on average fiscal multiplier were near 0.5 in advanced economies during the three decades leading up to 2009.”

Krugman responded to these criticisms: “The crucial thing from a macroeconomic point of view is that leveraging and deleveraging are not symmetric in their effects. Leveraging up, other things equal, leads to high aggregate demand — but this can be and is in practice offset by the central bank, which can always raise rates. Deleveraging, on the other hand, can’t be offset equally easily; the central bank can cut rates, but only to zero, and unconventional monetary policy is both controversial and an iffy proposition (which doesn’t mean that it shouldn’t be tried). So a large leveraging/deleveraging cycle is likely to be followed by a persistent shortfall in aggregate demand that can’t be cured using ordinary monetary policy; what I consider depression economics.”

So in periods when the capitalist economy is deep in recession, the multiplier is likely to be higher because monetary policy cannot help and excess capacity is huge in the economy.  So boosting government spending may produce a better kick to growth than in normal times.  In the short term (2009-13?), more government spending could alleviate the hit to economic growth, while more austerity makes it worse (see this Goldman Sachs paper, GS Fiscal multipliers at times of economic slack and when monetary policy is at zero bound 041912.

But then there is another issue: causation.  All these studies do not tell you what causes what.  Did the recession cause deficits to rise and debt to increase and thus force governments into austerity, not the other way round?  It was not fiscal austerity that caused the Great Recession, but the Great Recession led to fiscal austerity.   Krugman’s reply is rather unconvincing, namely that as the IMF got it wrong and the impact of government spending cuts has been worse that expected, it shows that austerity must be the cause and slowing growth or contracting economies is the result.

Well, there has been a heap of studies that argue that it is large budget deficits and high debt that will cause GDP growth to falter.  The most famous one is that Reinhart and Rogoff (Reinhart and Rogoff Growth in a Time of Debt) that shows if public debt levels get to over 85-90% of GDP (as they now are in most advanced capitalist economies, then it will take years (5-7 years or more) to restore economic growth.  The implication is that the quicker public debt ratios are reduced, the quicker the sustained growth can resume.

But the causation is not clear: 1) a recession causes high debt, so the only way to get debt down is to boost growth (Keynesian) or 2) high debt causes recessions, so the only way to restore growth is to cut debt.  The evidence one or way or another from all these studies is not there.  As John Cochrane put it in his blog (The grumpy economist, Two views of debt and stagnation, 20 September 2012,, “when I read the review of the ‘studies’, they are the usual sort of growth regressions or instruments, hardly decisive of causality.”  In other words, the studies show a correlation between high debt, big budget deficits and recessions, but not the causal direction.

And there is the issue of: what sort of government spending and what sort of tax cuts would help restore the capitalist economy?  When you ask that question, it shows that word ‘austerity’ is really a political, not an economic term, steeped in ideology. This brings me to the distinction that has been made previously in this blog (see Keynes, the profits equation and the Marxist multiplier, 13 June, 2012) and has been highlighted in particular by the work of G Carchedi recently (Carchedi and the Marxist multiplier). There is an important difference between the Keynesian multiplier and what we could call the Marxist multiplier.

The Keynesian analysis denies or ignores the class nature of the capitalist economy and the law of value under which it operates by creating profits from the exploitation of labour.  As a result, Keynesian macro identities start from consumption and investment (“effective demand”) and go onto incomes and employment.  So the Keynesian multiplier measures the impact of more or less spending (demand) on income (GDP).  The Marxist multiplier starts from profit generated from the class struggle and the law of value.  So the causation is from profits to capitalist investment and then from investment to employment, wages and consumption.  Spending and growth in GDP are dependent variables on profitability of investment, not the other way round.  The Marxist multiplier measures changes in profitability and their impact on investment and growth.  Then, as Carchedi puts it:  “in the Marxist multiplier, profitability is central…. The question is whether n rounds of subsequent investments generate a rate of profit higher than, lower than, or equal to the original average rate of profit”.

If the Marxist multiplier is the right way to view the modern economy, then what follows is that government spending and tax increases or cuts must be viewed from whether they boost or reduce profitability. If they do not, then any short-term boost to GDP from more government spending will only be at the expense of  a lengthier period of low growth and an eventual return to recession.

If government spending goes into social transfers and welfare, that will cut profitability as it is a cost to the capitalist sector and adds no new value to the economy.  If it goes into public services like education and health (human capital), it may help to raise the productivity of labour over time, but it won’t help profitability.  If it goes into government investment in infrastructure that may boost profitability for those capitalist sectors getting the contracts, but if it is paid for by higher taxes on profits, there is no gain overall.  If it is financed by borrowing, profitability will be constrained by a rising cost of capital.  There is no assurance that more spending means more profits, on the contrary.  Carchedi and I are working on a joint paper that details the arguments around the Keynesian and Marxist multipliers.  So I’ll return to these issues in future posts.

Krugman’s triumphal tone is based on just the IMF’s new measure of the multiplier based on the last two years plus estimates and forecasts of this year and next. The question is whether Keynesian promotion of more government spending through budget deficits would deliver growth over a sustained period (and not just after a huge slump, with zero-bound interest rates).

I did a little piece of statistical research on this issue comparing the average budget deficit to GDP for Japan, the US and the Euro area against real GDP growth since 1998.  1998 is the date that most economists argue was the point when the Japanese authorities went for broke with Keynesian-type government spending policies  designed to restore economic growth.  Did it work?

Well, between 1998 and 2007, Japan’s average budget deficit was 6.9% of GDP, while real GDP growth averaged just 1%.  In the same period, the US budget deficit was just 2% of GDP, less than one-third of that of Japan, but real GDP growth was 3% a year, or three times as fast as Japan.  In the Euro area, the budget deficit was even lower at 1.9% of GDP, but real GDP growth still averaged 2.3% a year, or more than twice that of Japan.  So the Keynesian multiplier did not seem to do its job in Japan over a ten-year period.   Again, in the credit boom period of 2002-07, Japan’s average real GDP growth was the lowest even though its budget deficit was way higher than the US or the Eurozone.

Now let’s look at the period from the point of view of the Marxist multiplier.  As I have shown in many previous posts, after 1997, the rate of profit in most of the advanced capitalist world began to decline.  The Marxist multiplier would then forecast that investment growth would start to slow, and so would GDP growth.  Well, in the four years from 1998 to the mild recession of 2001, US real investment growth averaged 6.1% a year while the government ran a small budget surplus and real GDP growth was 3.6% a year.  But after the recession of 2001, during the credit boom of 2002-07, US real investment growth slowed to 2.2% a year, but the government ran a budget deficit that averaged 3.6% of GDP and real GDP growth slowed to 2.6% a year.  It was the same story for Europe.  Even more revealing is capitalist investment in the productive sectors of the economy (real non-residential private capital formation in OECD terms).  Between 1998-01, US real investment in productive sectors rose 7.2% a year, but from 2001-07 it rose only 3.5% a year (half the rate).  Again it was the same story for Europe.

There does not seem to be any evidence that bigger government spending or wider budget deficits will lead to faster investment or economic growth over time in capitalist economies.  Indeed, the evidence is to the contrary much of time.  The Marxist multiplier of profitability and investment seems more convincing.

6 thoughts on “The smugness multiplier

  1. It’s not just the fiscal multipliers that are wrong, it’s their economic theory, whether neo-classical, Keynesian or even Austrian; they all wrong base their theory on marginalism.
    Marx, just like the classical economists, had an objective theory of value, just that his was better.
    Once again it falls to Marxists to explain to the capitalists how their system works & why it is prone to crisis.

  2. Really interesting post. Though I think that Carchedi paper you link may contain an error:

    “This chart shows that up to 1986, wages rise relative to profits and the ARP falls, conform to Marx but not to Keynesian underconsumptionism. From 1987 to 2009 wages fall relative to profits and the ARP rises, again conform to Marx but not to Keynesian underconsumptionism. Both pro-labour and pro-capital redistribution did not prevent the ARP from falling.”

    That’s the opposite of what the chart shows. I think the lines are miscolored or mislabeled. (Also some of the grid lines got splashed with color.)

    Also, what do you think of this post regarding using something like the “MMT” job guarantee idea, albeit not so much with Keynesian stimulus in mind as to shrink the domain of the economy strictly governed by the logic of capital?

    1. Hedlund

      I’ll check the Carchedi piece and see what’s wrong. The idea of a guaranteed job is something that any progressive should support and of course a left government should use any measures that can help improve the position of labour. So there is no conflict here. But the MMT idea that a country that ‘controls its own currency’ (ie the UK ot the US) as opposed to Greece has more room to play is really just a modern way of saying that one economy can devalue and gain market share at the expense of others – it’s a race to the bottom for currency values. It’s a temporary national ‘solution’ that soon poses new problems (inflation, real wage reduction etc).

  3. “So the causation is from profits to capitalist investment and then from investment to employment, wages and consumption.”

    This is of course correct – as opposed to the Keynesian approach – and profitability is the key driver in a capitalist economy. But I’m not so sure that it is straightforward, i.e. that investment growth will fall as a result of a fall in profitability.

    It would be instructive to look at investment and profitabillity since WWII – rather than starting in 1997 – as this would cover the long boom and the 1970s crisis period.

  4. “1998 is the date that most economists argue was the point when the Japanese authorities went for broke with Keynesian-type government spending policies designed to restore economic growth”

    My understanding was that they went for broke with enormous QE programs. I believe Krugman was pushing for more Keynesian style fiscal spending but it was really a watered down version in the end. So its not really a good example of a true Keynsian fiscal stimulus – in the end it was a half-way house between QE and limited stimulus.

  5. Please let me notice you that Blanchard and Leigh responded to FT’s observations on their multipliers’ estimates in this paper, January 2013:

    Click to access wp1301.pdf

    “when we remove the
    two largest policy changes (those for Germany and Greece), the estimate of β declines to –
    0.776 (t-statistic = –2.249) but remains statistically significant at the 5 percent level. Thus,
    concerns raised by some in reaction to an earlier version of this paper, that excluding the
    largest policy changes from the sample might render the results insignificant, seem

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: