Fears are being expressed that in 2013 the American economy will plunge over a so-called fiscal cliff. On unchanged policies the budget deficit (cyclically- adjusted) is due to fall sharply. Keynesian textbook orthodoxy says that a large decline of that sort represents a marked tightening of fiscal policy which will ‘withdraw spending power from the economy’, and so reduce demand, output and employment.
But will it? The thinking behind modern fiscal policy were first developed in Keynes’ 1936 General Theory, particularly in its chapter 10, and his 1939 essay How to Pay for the War. The ideas were attractive in theory, not least because they accorded the government a large role in ‘managing the economy’. That appealed, and continues to appeal, to ‘the socialists in all parties’. (The phrase ‘socialists in all parties’ comes from Friedrich Hayek, The Road to Serfdom.) But does Keynesian thinking on fiscal policy work in practice? In the last few years the International Monetary Fund has published a database which includes numbers for both the output gap and the cyclically-adjusted (or ‘structural’) budget balance for all its important member nations. Analysts can therefore check the evidence on the relationship between changes in the structural budget balance and growth relative to trend. The Keynesian worldview would be confirmed if above-trend growth were associated with (or ‘caused by’) increases in the structural budget deficit. In my 2011 book Money in a Free Society I looked at the US evidence and established, at least to my own satisfaction, that the data up to 2008 flatly contradicted Keynesianism. In today’s note I extend the analysis to 2012 and also allow for some changes to old data. My conclusion in Money in a Free Society is not only confirmed, but reinforced. Naïve fiscalist Keynesianism does not work in the USA. The further implication is that the fiscal cliff does not mean that the American economy will suffer from another recession, or even demand weakness, in 2013.
American macroeconomic policy is at an awkward turning point. The budget deficit cannot be allowed to remain at or near its present level because the ratio of public debt to GDP would explode upwards. On the general government definition (i.e., Federal and state governments combined) and using IMF data, the deficit peaked at $1,864b., or over 13% of gross domestic product, in 2009. It is estimated to be down to $1,358b.in 2012, still almost 9% of GDP. A fall is expected in 2013 and later on unchanged policies, but these ‘unchanged policies’ incorporate a big rise in taxation and consequent fall in the deficit. Keynesian textbook orthodoxy says that the projected large decline in the deficit next year, the so-called ‘fiscal cliff’, represents a marked tightening of fiscal policy, which will ‘withdraw spending power from the economy’, and so reduce demand, output and employment.
But will it? The thinking behind modern fiscal policy was first developed in Keynes’ 1936 General Theory, particularly in its chapter 10, and his 1939 essay How to Pay for the War. The ideas were plausible and attractive in theory, not least because they accorded the government a large role in ‘managing the economy’. That appealed, and continues to appeal, to ‘the socialists in all parties’. (The phrase ‘socialists in all parties’ comes from Friedrich Hayek, The Road to Serfdom.) But does Keynesian thinking on fiscal policy work in practice? In the last few years the International Monetary Fund has published a database which includes numbers for both the output gap and the cyclically- adjusted (or ‘structural’) budget balance for all its important member nations. Analysts can therefore check the evidence on the relationship between changes in the structural budget balance and growth relative to trend.
The Keynesian worldview would be confirmed if above-trend growth were associated with (or ‘caused by’) increases in the structural budget deficit. In my 2011 book Money in a Free Society I looked at the US evidence and established, at least to my own satisfaction, that the data up to 2008 flatly contradicted Keynesianism. In today’s note I extend the analysis to 2012 and also allow for some changes to old data. My conclusion in Money in a Free Society is not only confirmed, but reinforced. Naïve fiscalist Keynesianism does not work in the USA. The further implication is that the fiscal cliff does not mean that the American economy will suffer from another recession, or even demand weakness, in 2013. In the next section I set out the second round of evidence on fiscal policy’s ineffectiveness. In the section after that I consider why fiscal policy doesn’t work or, at any rate, doesn’t seem to work in the fashion understood by the textbooks.
Fiscal Cliff: The evidence: an update for the data to 2012
Essay 8 of my November 2011 book Money in a Free Society discussed Milton Friedman’s attitude towards fiscal policy (Question: ‘what role do you see for fiscal policy?’, Friedman’s answer: ‘none’, p.191 of Money in a Free Society), and set out some evidence on the relationship between changes in the structural budget balance and the output gap (i.e., growth relative to trend) for the 1980 – 2008 period. As explained above, I used date from the IMF website, as it was in 2009 and 2010. The 1980 – 2008 period of course included 29 years, but it contained the data on ‘changes’ in the two relevant variables for 28 years. I showed that, out of these 28 years, there were only five in which the output gap and the budget balance changed, by significant amounts (i.e., more than ½% of GDP), in opposite directions. In other words, the American economy behaved in a clearly Keynesian manner less than a fifth of the time. There were twice as many years (i.e., ten years) in which the output gap and the structural budget balance moved in the same direction, which contradicts Keynesian theorizing. (If Keynesian fiscalism were correct, output ought to have risen in those years when the deficit increased, i.e., in those years when the budget balance went more negative.) Friedman seems to have thought about pursuing rigorous intellectual research on the topic, but never in fact published a careful piece of statistical work on changes in the structural budget balance, perhaps because the preparation of the data would be too time-consuming. At any rate, the IMF has a large research staff and hence the resources to estimate some numbers. The IMF data I put together for the USA had a pretty clear-cut message, that Friedman was right. With the US economy responding weakly and erratically to changes in the budget balance, fiscal policy did not appear to be very effective.
We now have four more years of data, although 2012 is not quite over yet. Some changes have been made by IMF staff to the earlier numbers for the output gap. (The IMF has also stopped publishing the structural budget deficit data for the USA for the years from 1980 to 2000. The rationale for this omission is a puzzle. To arrive at the figures for the level of the structural budget deficit below, I used the IMF’s latest figure for the level of the deficit in 2001 and then deducted the change in the deficit in 2000, 1999 and so on, to produce a level series for those years. The figures for the change in the deficit come from an earlier IMF database.)
Just to make sure that my conclusion is based on fact, not phoney figures, it makes sense to check the year-by-year history told by the table above. Let us look at the budget balance first. We start in the early 1980s with a very ‘restrictive’ fiscal policy, which seems to be the legacy of the Carter administration. By contrast, in every one of the five years 1982 – 86 fiscal policy is eased as Reagan cut taxes as part of his supply-side agenda. A few indeterminate years follow until the Clinton Presidency, when ‘the peace dividend’ allows large cuts in defence expenditure. Every one of the eight years 1993 – 2000 sees a reduction in the structural budget deficit, converting the deficit of 6.8% of GDP in 1992 to a surplus – a tiny one – in 2000. The first term of the second Bush Presidency sees a widening of the deficit, partly justified at the time on Keynesian lines (i.e., to defeat the downturn associated with the bursting of the dotcom bubble). Finally, the Great Recession is met by deliberate Keynesian stimulus, as first Bush and then Obama listen to their (mostly) Keynesian advisers. The net effect is an increase in the structural deficit of 5% – 6% of GDP, equivalent to over $700b. at an annual rate. In short, the numbers make sense from what is generally understood of the intentions and actions of the various US Presidents and their advisory teams. (The President can be fairly seen as the conductor of the fiscal policy orchestra in the American context.) If we also cross-check the series for the output gap, that too is consistent with the economic history narrative of these years. Bad recessions occur in 1981 and late 2008 (plus early 2009), and we notice that the most negative values of the output gap are in the years 1982 and 2009. Output is well above its trend level in 2000, with all the euphoria of the dotcom bubble. In short, the facts in the table look like the ‘right’ facts. The message from the table is not one which can be attributed to any doctoring of the data. (And, if readers do not believe that message, they can go to the IMF website and do their own reconnaissance. I am not making anything up.)
Anyhow, the table above summarizes the results of adding four more years to the previous exercise and adjusting for new estimates of the earlier values of the two variables. It turns out that – in this 32- year period, as in the 28-year sample of the previous analysis – there were only five years in which the output gap and the budget balance moved in opposite directions. In other words, the American economy behaved in the way understood by the Keynesian textbooks less than a sixth of the time! Indeed, given that there were 14 years in which the output gap and the budget balance moved in the same direction, the American economy behaved in an anti-Keynesian way almost three times as often as it behaved according to the textbooks.
My conclusion is that the textbook theory is wrong. I have drawn this conclusion several times before, for anyone who has bothered to read my work. (And there could be more of those!) Last summer I showed in an exchange (some of it in letters on the Financial Times correspondence page) with Martin Wolf, the chief economic commentator on the Financial Times, that ‘expansionary fiscal contraction’ had been the norm in the UK since the 1980s. (See the International Monetary Research weekly e-mails of 11th May 2011 and 11th June 2011.) In a regression exercise, admittedly a primitive one, the regression coefficient on an estimated relationship between the output gap and the budget balance took a positive value. In other words, an increase in the budget surplus (or a reduction in the deficit, i.e., a fiscal tightening) was associated with above-trend growth. Indeed, when five-year moving averages of the two variables were used, the regression coefficient easily met the usual significance test (i.e., the t statistic was well over 2). I have little doubt that something similar could be found in the USA over the 1981 – 2012 period, but haven’t yet done the calculation. This finding is pretty weird and surprised even me, and I am a long-term sceptic about fiscal policy. (I would have expected a negative coefficient of little significance in a relationship of very low quality.) But statistics are statistics, and anyone can draw their own more impressionistic conclusions – as far as the USA is concerned – from the two tables above. (Needless to say, Wolf continues to prattle in his column about the recession-causing dangers of fiscal restriction, including the USA’s ‘fiscal cliff’. He does so regardless of evidence and without any meaningful reply to the points I made. What does one say?)
On the face of it, American policy-makers would be well-advised to plan significant reductions in the USA’s structural budget deficit (expressed as a %age of GDP), perhaps by 1% – 2% a year, with a view to its elimination by 2016 or 2017 or so. They should do so in order to check the growth of the public debt, and the incurrence of a large and ever-rising future burden of debt interest and servicing. They should certainly not bother themselves about the risk of another recession in 2013 because of a supposed ‘fiscal cliff’.
Why is fiscal policy so useless?
So fiscal policy is useless. Policy-makers should – I suggest – concentrate on keeping public debt to the lowest possible figure, eschewing Keynesian demand management altogether. They should do that to ensure that the USA’s long-term fiscal position is solvent. But why is fiscal policy useless as a means of influencing demand? What is wrong with it?
Various potential answers come into consideration. (See essay 9 of Money in a Free Society and particularly p. 204.) But even to me, a long-term cynic/sceptic about Keynesian ‘fiscal pump- priming’, the clarity of the message from the USA’s experience over the last 30 years is surprising. After all, if the government cuts spending and lays off workers, that must mean – at least for a period – a fall in aggregate demand and employment, unless the private sector agents take action almost immediately and with a similar effect to offset the negatives from the public sector’s retrenchment. I am not going to offer a final answer here to the apparent puzzle of fiscal policy ineffectiveness in the USA. But may I suggest here that the basic flaw in Keynesian fiscalism is to over-estimate the importance of actions by the state compared with the effects of the private sector’s adjustments (to expenditure and investment portfolios) to changes in the quantity of money?
It is vital to remember that over the medium and long runs the evidence is pretty definite in the USA, as in other countries, that changes in money, nominal national income and money wealth are equi- proportional (or, at any rate, as near to equi-proportionality as to be consistent with standard monetary theory, that the demand to hold money is a stable function of a small number of variables, with income being the dominant one). I have argued that – when the quantity of money rises by, say, 5% – the equilibrium values of the variable-income forms of wealth (i.e., particularly, corporate equity and real estate, in both residential and commercial forms) also rise by 5%. (See, for example, Money and Asset Prices in Boom and Bust, my 2005 monograph for the Institute of Economic Affairs, and essays 15 and 16 in Money in a Free Society.) The equi-proportionality may not be evident in periods of a few quarters, but over longer periods – say, five years or more – the data are consistent with this view. We need then to remember that wealth is invariably a multiple of income.
In the USA the net worth of the household sector was $60,110b. at the end of last year, more than five times disposable personal income. Between the end of 2007 and the end of 2008 net worth slumped from $66,057b. to $53,457b., that is, by $12,600b. or almost 20%. Of course people do not consume all their wealth in any one year, but their propensity to spend out of income is strongly influenced by fluctuations in their net worth. If the value of someone’s equity portfolio and real estate assets drops by almost 20%, it is inconceivable that his or her spending from income will be unaffected. Suppose that the US household sector were to make the attempt in 2009 to rebuild assets to the previous level – the level that evidently was preferred at the earlier date of end-2007 – by, say, 3% of the current, end- of-2008 level. Then the cutback in spending in 2009 would be over $1,600b. (yes!) or over 10% of GDP. Even if the Fed slashed interest rates, it would be travelling against a strong headwind. And, even if President Obama were to endorse aggressive ‘fiscal expansionism’ (as of course he did), fiscal policy could not make much progress against the asset price deflation which was to be explained largely in monetary terms.
Conclusion: need to monitor growth of the quantity of money
My conclusion is that – because the rate of growth of broad money is, over the medium term, the key driving force in the determination of nominal wealth (and hence the net worth of households and businesses) – it is money growth trends that will matter critically to the macroeconomic outlook in the early years of Obama’s second term. The ‘fiscal cliff’ is far less important than implied by the media hullabaloo about it.
The latest estimates from Shadow Government Statistics are that M3 increased by 0.5% in September, while in the three months to September the annualized rate of growth was 5.0%. The USA’s stock market has had a reasonable year, while house prices are rising. On this basis, the American economy should enjoy a year of satisfactory growth in 2013 despite the fiscal cliff. Indeed, the prospects are quite good even if the politicians reach no agreement on averting the fiscal cliff. A large reduction in the budget deficit would in fact be positive for the USA’s long-run fiscal solvency, and for business and household confidence in the next year or two.Tags: Associated Press, Beijing, Bernard Tapie, Board of directors, Christine Lagarde, France, Friedrich Hayek, International Monetary Fund, Iraq, Media of France