A new working paper from the IMF, written by Luc Eyraud and Anke Weber, has brought more attention to austerity in the current economic climate. Specifically, Eyraud & Anke explored what effect fiscal austerity would have on the debt to GDP ratio when the fiscal multiplier is higher than it otherwise would be. Via both simulations and econometric techniques, the authors show that austerity can actually increase the debt ratio in the short term, rather than reduce it as intended.
Eyraud & Anke articulate the following fiscal equation to demonstrate the importance of the fiscal multiplier:
The multiplier appears twice on the right hand side as a mitigating influence via its effect on the debt ratio and the revenue ratio. The impact of the multiplier on the debt ratio captures the effect of lower GDP on the debt ratio, while the impact on the revenue ratio captures the automatic stabilisers that also work to offset the direct effect of the consolidation.
By way of example, they cite a country with a debt ratio of 80%, a revenue ratio of 40% and a first year multiplier of 0.6. In this scenario, a 1% of GDP fiscal consolidation reduces the debt ratio by 0.3%, as the 1% reduction is offset by the mitigating factors. The debt ratio will increase, however, in the current economic environment. As other work from the IMF has discussed, with interest rates already low (and unable therefore to offset fiscal consolidation) and a weak external environment, fiscal multipliers can be over 1. Using a multiplier of 1 in the example above changes the outcome: the result being a debt ratio increase of 0.2% rather than a debt ratio decrease.
How they did it
While Eyraud & Anke conduct simulations to show the debt ratio increasing from a fiscal consolidation, I would like to focus on the empirical work that they used to support these results. The authors employ a structural VAR approach (SVAR) to account for the government’s intertemporal budget constraint. This framework allows them to conduct fiscal shocks on the system and to keep track of the debt dynamics. They use Japan in their example, as it has a high debt to GDP ratio and the reasonable assumption of a high fiscal multiplier.
On a technical note, the SVAR approach was quite intriguing as they had to use an approach by Blanchard & Perotti (2002) to correctly identify the fiscal shocks. This is due to the complication that the residuals from the VAR contain both the errors from a fiscal shock as well as the normal errors from the automatic response of the fiscal balance to macro variables. With the correct identification of the truly exogenous shift in fiscal policy, Eyraud & Anke are then able to shock the SVAR with a fiscal tightening of 1% and compare this to the baseline with no fiscal tightening.
What they found
The chart below shows the results from the SVAR. Similar to the simulations they conducted the fiscal tightening leads to an increase in the debt to GDP ratio for 6 quarters. These results are due to the responses of the endogenous variables. Real output and inflation fall immediately, which offsets the decrease in the primary deficit and reduces the denominator in the debt to GDP ratio.
Why is this important?
As Eyraud & Anke note, misunderstanding fiscal multipliers can be problematic for policymakers. In the current environment of a depressed economy and monetary policy that is at the zero lower bound, an incorrect assumption on the multiplier can lead authorities to set unrealistic debt and deficit targets, leading to missed targets and uncertainty about future policy. The problem may be exacerbated if authorities then embark on repeated rounds of fiscal tightening to achieve these unrealistic targets, all the while harming the economy in the process.
Possibly the most important point from this study is that the timing of fiscal consolidation is crucial. Eyraud & Anke state that a more gradual consolidation is usually preferable as the multiplier will be lower when more normal economic conditions prevail. Crucially, delaying the fiscal consolidation would lessen the size and duration of the policy’s impact on growth. Eyraud & Anke, therefore, have provided a highly useful and timely piece of research for the current macroeconomic debate.