Fiscal Multipliers: The Olivier Blanchard-Daniel Leigh Nostra-Culpa "Box"
A "box" that is three pages of small print long, mind you…
The authors of this box are Olivier Blanchard and Daniel Leigh:
Box 1.1. are We Underestimating Short-term Fiscal Multipliers? With many economies in fiscal consolidation mode, a debate has been raging about the size of fiscal multipliers. The smaller the multipliers, the less costly the fiscal consolidation. At the same time, activity has disappointed in a number of economies undertaking fiscal consolidation. So a natural question is whether the negative short-term effects of fiscal cutbacks have been larger than expected because fiscal multipliers were underestimated.
This box sheds light on these issues using international evidence. The main finding, based on data for 28 economies, is that the multipliers used in generating [IMF] growth forecasts have been systematically too low since the start of the Great Recession, by 0.4 to 1.2, depending on the forecast source and the specifics of the estimation approach. Informal evidence suggests that the multipliers implicitly used to generate these forecasts are about 0.5. So actual multipliers may be higher, in the range of 0.9 to 1.7.
Forecast Errors and Fiscal Multipliers Our basic approach is the following: focusing on the recent episode of widespread fiscal consolidation, we regress the forecast error for real GDP growth dur- ing 2010–11 on forecasts of fiscal consolidation for 2010–11 that were made in early 2010. Under rational expectations, and assuming that the correct forecast model has been used, the coefficient on planned fiscal consolidation should be zero. The equation estimated is:
forecast error of growth = α + β forecast of fiscal consolidation + ε. (1.1.1)
The forecast error of growth is equal to actual cumulative real GDP growth during 2010–11 minus the forecast of growth in the April 2010 World Economic Outlook. The forecast of fiscal consolidation is the forecast of the change in the structural fiscal balance as a percentage of potential GDP during 2010–11 as of the April 2010 WEO. We also investigate forecasts other than the WEO. If the fiscal multipliers used for forecasting are accurate, the slope coefficient, β, should be zero. Our baseline sample consists of 28 economies: the major advanced economies included in the G20 and the member countries of the EU for which forecasts are available.
What Do the Data Show? We find the coefficient on planned fiscal consolidation to be large, negative, and significant.
The baseline estimate suggests that a planned fiscal consolidation of 1 percent of GDP is associated with a growth forecast error of about 1 percentage point (Table 1.1.1 and Figure 1.1.1, panel 1). This result indicates that the multipliers underlying growth projections have been too low by about 1. The systematic relationship between fiscal consolidation and growth holds up to a battery of robustness tests. Overall, depending on the forecast source and the specification, our estimation results for the unexpected output loss associated with a 1 percent of GDP fiscal consolidation are in the range of 0.4 to 1.2 percentage points.
First, we establish that the baseline result is not driven by crisis economies––those that had IMF programs—or other outliers (Table 1.1.1).
Next, we check whether the results are robust to controlling for additional variables that could plausibly have triggered both planned fiscal consolidation and lower-than-expected growth. The omission of such variables could bias the analysis toward finding that fiscal multipliers were larger than assumed.
We consider two groups of variables: those that were known when the growth forecasts were made and those that were not (Table 1.1.1).
Variables known at the time the forecasts were made: We start by considering the role of sovereign debt problems. Are the baseline results picking up greater-than-expected effects of sovereign debt problems rather than the effects of fiscal consolidation? Reassuringly, the results are robust to controlling for the initial (end-2009) government-debt-to-GDP ratio and for initial sovereign credit default swap (CDS) spreads. Controlling for the possible role of banking crises—based on the data set of systemic banking crises of Laeven and Valencia (2012)—yields similar results.
The baseline finding also holds up to controlling for the fiscal consolidation of trading partners. To the extent that fiscal consolidations were synchronized, fiscal consolidation by others may be driving the results. However, when we control for trade-weighted fiscal consolidation of other countries (scaled by the share of exports in GDP), the results are virtually unchanged. Finally, to investigate the role of precrisis external imbalances that may have triggered both fiscal consolidation and larger-than-expected headwinds to growth, we try controlling for the precrisis (2007) current-account-deficit-to-GDP ratio and find similar results.
Variables not known at the time the forecasts were made: We consider the role of the sharp increase in sovereign and financial market stress during 2010–11, measured by the change in the sovereign CDS spreads. Controlling for these developments again yields similar results. We also address the possibility that, even if the assumed multipliers were correct, countries with more ambitious consolidation programs may have implemented more fiscal consolidation than originally planned.
As Table 1.1.1 reports, including unexpected fiscal consolidation does not significantly affect the results, suggesting that the baseline specification is appropriate. In line with this result, we find that there was no systematic tendency for economies with larger initial fiscal consolidation plans to implement larger additional consolidation.
GDP Components, Unemployment, and Different Forecasters: When we decompose GDP, we find the largest coefficient for forecasts of investment and the most statistically significant coefficient for forecasts of consumption (Figure 1.1.1, panel 2). The coefficient associated with forecasts of the unemployment rate is also large and significant.
We also consider four different sets of forecasts: those of the WEO, the European Commission (EC), the Organization for Economic Cooperation and Development (OECD), and the Economist Intelligence Unit (EIU—Figure 1.1.1, panel 3). The largest estimated coefficient is associated with the WEO forecasts and the smallest with the OECD forecasts. The coefficient is statistically significant in all cases.
What Does This Say about Actual Fiscal Multipliers? These results suggest that actual fiscal multipliers were larger than forecasters assumed. But what did forecasters assume about fiscal multipliers? Answer- ing this question is complicated by the fact that not all forecasters make these assumptions explicit. Nevertheless, a number of policy documents, including IMF staff reports, suggest that fiscal multipliers used in the forecasting process are about 0.5. In line with these assumptions, earlier analysis by the IMF staff suggests that, on average, fiscal multipliers were near 0.5 in advanced economies during the three decades leading up to 2009.
If the multipliers underlying the growth forecasts were about 0.5, as this informal evidence suggests, our results indicate that multipliers have actually been in the 0.9 to 1.7 range since the Great Recession. This finding is consistent with research suggesting that in today’s environment of substantial economic slack, monetary policy constrained by the zero lower bound, and synchronized fiscal adjust- ment across numerous economies, multipliers may be well above 1 (Auerbach and Gorodnichenko, 2012; Batini, Callegari, and Melina, 2012; IMF, 2012b; Woodford, 2011; and others). More work on how fiscal multipliers depend on time and economic conditions is warranted.