On Wednesday, I attended an event hosted by the APPG on ‘Economics, Banking and Finance’ on the ‘Future of political economy’ in the UK. The discussion was discursive in topic nature but highly impressive in intellectual content. The extent of differences in opinion perhaps showed why economics can never be regarded as a science. The three panellists: Richard Murphy, Jesse Norman MP and Dr. Mark Pennington, were all well-versed in economic theory and literature, but had come to vastly different conclusions as to the lessons for government policy.
Among the topics discussed included the issue of fiscal stimuli: i.e. have the US and UK governments been right in their responses to the recession with their tax and spending policies. Richard Murphy was very much on the Keynesian side of the argument (well, Keynesian to the extent that he believed a substantial government spending stimulus was necessary) whilst Dr. Mark Pennington was more defensive of the path that the UK government has chosen to take. “You speak as if the state is somehow small at the moment,” he said to Mr. Murphy. “But in truth, the state already accounts for 47.3% of GDP and will add around £400 billion to the national debt over this Parliament.”
Quite. As we showed in ‘Five Fiscal Fallacies,’ the last government's response to the recession saw a combined monetary and fiscal stimulus of up to 40% of GDP over three years. Yet growth last year was still extremely sluggish. In that paper, Dr Tim Morgan argued that this was because of the nature of the recession. He claimed that the stimulus was not as effective as it would be after an ordinary destocking recession, because we are in a deleveraging one:
“Businesses and individuals have looked over the debt abyss and have stepped back in fear. Their aim now is to reduce their leverage. Stimulus pumped in by government will be used to reduce debt, not to boost consumption. This is why stimulus has not, this time, worked according to the Keynesian calculus. It simply transfers debt from private balance sheets to that of the government.”
The key debate around whether fiscal stimuli are effective is the disputed size of the fiscal multiplier, i.e. the implied change in national income brought by a change in government spending. Much economic literature in the past has presumed that this multiplier is above 1, as well as the Obama administration whose stimulus plan used an assumed multiplier of 1.5. This implies that if the government spends an extra one per cent of GDP, national income will rise by one and a half per cent of GDP due to the increase in consumption spending resulting though the creation of more jobs. In short, the government, through fiscal policy, can increase GDP faster than the increase in spending relative to the taxes it collects.
But there are reasons to suspect that at the moment the multiplier will be much lower. First, the high indebtedness of businesses and individuals, added to the fear of the economic outlook, are likely to have reduced the marginal propensity to consume – making the second stage consumption spending effects much smaller. In addition, observed high levels of government debt mean that increases in government expenditures act as a signal that larger fiscal tightening will be required in the future, as now evidenced in Greece. If the public anticipate these adjustments, there could be a contractionary effect on consumption levels which will offset any short-term expansionary impact that an increase of government consumption might have.
Some recent research by the IMF (How Big (Small?) are Fiscal Multipliers?) seems to back this up. It found that:
“During episodes where the outstanding debt of the central government was high (exceeding 60 percent of GDP) the fiscal multiplier was not statistically significant from zero on impact and was negative in the long run”
The UK’s official debt today is 66.1% of GDP, and is predicted to peak at 70.9% in 2013/14.
In addition, the research found that economies with flexible exchange rates essentially have zero multipliers: a fiscal expansion has the initial effect of raising output, which in turn raises interest rates through an increase in money demand. This induces a flow of foreign capital and creates an appreciation pressure on the domestic currency, offsetting the initial government spending effect through lowering next exports. And these results are robust to government investment spending.
But let’s suppose the UK does have a positive multiplier. This does not guarantee that in the long-term a stimulus would necessarily have a beneficial effect on the economy. Work by Robert J. Barro, a Harvard economist who recently gave the Hayek lecture, estimated that government spending has a multiplier of around 0.8. But since spending must be repaid in the future, most likely through tax increases (which he calculates to have a multiplier of -1.1), the overall decision can sometimes have a negative effect on GDP. That is, the government spending actually has more cost than benefit.
It is worth noting that the OBR uses an assumed government consumption multiplier of 0.6 for welfare and departmental expenditure, and 1.0 for government capital expenditure, in its impact forecasts. These impacts diminish over time. This recent IMF evidence, however, now suggests that the multiplier might be even less significant still. Of course, the overall multiplier will depend on where money is spent, but given the nature of the last recession and the structure of the UK economy (open, highly indebted, and with a flexible exchange rate), there is not much evidence to suggest that a large fiscal stimulus would have been particularly effective.