This column is a lead commentary in the VoxEU Debate "Secular Stagnation"
Coen Teulings and Richard Baldwin start their introduction to VoxEU’s eBook on secular stagnation (Teulings and Baldwin 2014) with the following two sentences:
“Six years after the Crisis and the recovery is still anaemic despite years of zero interest rates. Is ‘secular stagnation’ to blame?”
I think those two sentences embody a confusion that seems to arise when secular stagnation is discussed. For some (Eichengreen 2014) secular stagnation is about a downward tendency of real interest rates since 1980, but for others it seems to be about why recovery from recession has been weak. However I would argue that we know why the recovery has been so slow. It involves just two words: fiscal policy.
As IMF economists Kose et al. (2013) write, “The current and projected paths of government expenditures in the advanced economies are quite different than during past recoveries. During the past recoveries, fiscal policy was decisively expansionary, with increases in real primary government expenditures. This time is different.”
To illustrate the force of this point, I did the following back-of-the-envelope calculation. Suppose real spending on government consumption and investment (hereafter G) had grown by 2% per annum in 2010, and in each of the following years up until 2013, in the US, UK, and Eurozone. This would seem to be close to a neutral path for government spending. (Growth in G in the US – which includes expenditure at the state level – was a little above trend in 2008 and 2009 at around 3%, compared to an average of about 2.3% growth over the previous decade.)
A counterfactual without austerity
If G had followed this trend path, the level of G by 2013 would have been around 15% higher in the US, a bit less than this in the UK, and about 10% higher in the Eurozone. This indicates the extent of austerity that occurred from 2010 onwards. The question then becomes what sort of multiplier to apply to this number. I have used a multiplier of 1.5, which I think is reasonable for three reasons:
- First, when nominal interest rates are at the zero lower bound there are good reasons for expecting multipliers to be large.
- Second, in many countries fiscal consolidation included large cuts in public investment, which may have additional negative supply side effects.
- Third, the fact that G is higher in all three areas in this experiment should reduce the extent of import leakages.
- I assume the multiplier is instantaneous just for simplicity. Remember as well that I am ignoring changes to taxes and transfers, which from 2010 also tended to be contractionary. I am also assuming that none of the positive effects on GDP that I describe below would have been offset by a tighter monetary policy, but the point of the exercise is partly to argue that at some point they might well have been.
With this multiplier, higher G would have by 2013 raised the level of GDP by over 4% in the US, over 4.5% in the UK, and nearly 4% in the Eurozone. To put this in terms of changes, US GDP growth would have averaged 3.2% over those four years, rather than the actual average of 2.2%. We would probably have described that recovery as robust rather than anaemic.
A more comprehensive measure of fiscal stance is the underlying or cyclically adjusted primary deficit, as calculated by the OECD or IMF. Using this measure to gauge the impact on GDP suffers from the drawback that tax or transfer changes probably have smaller multipliers than government spending changes. However even if we make the quite conservative assumption that every 1% reduction in the underlying primary deficit reduces GDP by 1%, then we get changes in GDP by 2013 of the same order of magnitude as those already calculated. (Jordà and Taylor 2013 estimate larger GDP effects in recessionary periods, but with significant lags.)
We can compare this simple calculation with more sophisticated evidence for the Eurozone. Using the Commission’s QUEST model, in ’t Veld (2013) calculates that GDP would have been around 4.5% higher in the Eurozone in 2013 in the absence of fiscal consolidation from 2011 to 2013. Holland and Portes (2012) using the NIGEM model come up with similar numbers.
Figure 1. US GDP, logged, compared to a no-austerity counterfactual and potential
The impact of no austerity on the level of GDP by 2013 exceeds most current estimates of the output gap in all three areas. Figure 1 above plots the log of actual GDP in the US against the no-austerity counterfactual and the CBO’s estimate of potential GDP. Current output gaps may be underestimated, but the key point is that with GDP this much higher, it seems unlikely that interest rates would still be at the zero lower bound in all three areas. As a result, not only would we not be talking about an anaemic recovery, but we would no longer be wondering if interest rates would ever rise above the zero lower bound.
Government debt and the safe asset shortage
We can make the same point another way, by looking at the ‘safe asset’ explanation for low real interest rates which a number of economists have investigated, and which is put forward in the same eBook by Caballero and Farhi (2014). The argument is that low real interest rates primarily reflect a shortage of safe assets. Government debt is – Eurozone problems aside – a safe asset. The counterfactual I talked about above would have resulted in additional government debt, thereby reducing the shortage of safe assets.
As Caballero and Farhi stress, government debt is only a safe asset if markets believe governments have the fiscal capacity to at least sustain those debt levels once the recovery is complete. (At that point net debt could be reduced, but the level of safe assets maintained, by the government investing in a sovereign wealth fund.) This gets to the heart of the austerity debate. Those who argue that the no-austerity counterfactual was not feasible suggest that, because we entered the recession with unusually high levels of debt to GDP, allowing government spending to grow at trend would have made future levels of debt unsustainable. The safe assets argument suggests there was no question that the market would reject higher levels of debt immediately – instead they would welcome it. As a recent IMF evaluation concluded (IMF 2014), the Eurozone panic was an unfortunate red herring that led some policymakers astray. So the argument that we had to start reducing the level of government debt immediately must depend on the proposition that higher levels of debt would have become problematic at some point.
This argument seems very weak. The UK has past experience of much higher debt levels than today. US levels of tax are relatively low and nowhere near any Laffer maximum. A good deal of the austerity enacted involved cutting back on investment, whereas an efficient policy would have been to bring forward investment when interest rates were so low. As a result, it seems reasonable to conclude that the anaemic recovery from the financial crisis among the advanced economies reflects a policy choice to embark on early austerity, a choice that could have been avoided.
Austerity and low real interest rates
This is not meant to detract from what I regard as the central question in the secular stagnation debate, which is why the level of the ‘natural’ real interest rate seems to have fallen steadily since the 1980s. To the extent that a low natural real rate meant that the zero lower bound was hit earlier, it makes the questions posed by Teulings and Baldwin quoted at the start relevant. Low real interest rates allowed austerity to produce a weak recovery, because monetary policy was less able to offset its effects. However making clear the role that austerity has played in producing a weak recovery clarifies the secular stagnation debate in two respects:
- First, if without austerity interest rates were no longer at their zero lower bound, there would be less need to produce models, like that set out by Eggertsson and Mehrotra (2014), which produce negative natural real natural rates.
- Second, it would allow a clearer distinction between the secular stagnation issue and questions about future supply side growth, and whether the two are connected or not.
Caballero, R J and E Farhi (2014), “On the role of safe asset shortages in secular stagnation”, in Teulings and Baldwin (2014).
Eggertsson, G B and N Mehrotra (2014), “A model of secular stagnation”, in Teulings and Baldwin (2014).
Eichengreen, B (2014), “Secular Stagnation: The long view”, NBER Working Paper 20836.
Holland, D and J Portes (2012), “Self Defeating Austerity?”, National Institute Economic Review 222.
IMF, Independent Evaluation Office (2014), “IMF Response to the Financial and Economic Crisis”, Washington: IMF.
in ’t Veld, J (2013), “Fiscal consolidations and spillovers in the Euro area periphery and core”, Economic Papers 506, European Commission Directorate-General for Economic and Financial Affairs.
Jordà, O and A Taylor (2013), “The Time for Austerity: Estimating the Average Treatment Effect of Fiscal Policy”, NBER Working Paper 19414.
Kose, M A, P Loungani, and M E Terrones (2013), “Why is this global recovery different?” VoxEU.org, 18 April.
Teulings, C and R Baldwin (2014), Secular stagnation: Facts, causes, and cures, VoxEU eBook, August.