From 2010 onwards, European Union countries embarked on a prolonged phase of fiscal consolidation to restore fiscal sustainability in the aftermath of the financial crisis. By 2012, many EU countries were back into recession. Following these events, policymakers and economists began to question the size and timing of fiscal consolidation, and a new debate began concerning the size of short-term fiscal multipliers. Blanchard and Leigh (2013), for instance, argue that policymakers considerably understated the impact of fiscal policy on output at the start of the crisis – they argue that multipliers assumed prior to the crisis were around 0.5, while according to their estimates actual multipliers were substantially larger.
In our paper (Cugnasca and Rother 2015) we estimate the short-term output cost of fiscal consolidation for a panel of EU countries between 2004 and 2013. We identify consolidation episodes by looking at countries which are under the EU's Excessive Deficit Procedure, using as our exogenous fiscal instrument the structural adjustment required in Excessive Deficit Procedure recommendations.
Before the 2007-08 financial crisis, most of the economic literature focused on the estimation of fiscal multipliers that were independent of the state of the economy, which could be interpreted as representing the average response of output to fiscal policy. The actual impact of fiscal policy, however, is known to depend on the prevailing economic conditions, such as the state of the business cycle (see Auerbach and Gorodnichenko 2012), or the composition of the fiscal adjustment (see Gechert and Will 2012 for a survey of studies). We therefore start by estimating an average (or state-independent) multiplier and then consider how the impact of consolidation varies depending on whether the economy is in recession or not, whether it is open to trade or closed and whether the fiscal adjustment is achieved via cuts to direct government spending (i.e. consumption, investment and wages) or increases in taxes net of transfer payments. Finally, there is evidence that the output cost of consolidation can be lower in times of weak public finances (see ECB 2014) – the sovereign debt crisis that hit several EU Member States propagated to their respective banking systems and led to an overall increase in lending rates to the private sector. We therefore consider also whether credit stress with high lending rates affects the multiplier.
A reliable estimate of the fiscal multiplier requires finding a sound strategy to identify the growth impact of exogenous fiscal policy changes. We adopt the following strategy – under the so-called Maastricht criteria, an EU Member State with a deficit above 3% of GDP receives a recommendation from the ECOFIN Council to correct this excessive deficit within a given period. This recommendation includes a quantification of the necessary structural adjustment. The concerned government then implements its policies in reaction to the Excessive Deficit Procedure recommendation. The recommended structural adjustment is therefore set by the Council based only on economic developments realised up to the moment when the Excessive Deficit Procedure is launched and is not affected by developments realised at the time of the actual consolidation. Controlling for past economic developments, we use the recommended structural adjustment as an exogenous instrument for discretionary fiscal policy actions, in a dynamic panel model for 27 EU countries.1
Our results in Table 1 show that the average (or state-independent) output cost of a fiscal adjustment equal to 1% of GDP is 0.5% of GDP for the EU as a whole – this result is in line with the size of multipliers assumed before the crisis according to Blanchard and Leigh (2012), despite the fact that approximately three quarters of the consolidation episodes that we consider occur after 2009. The multiplier is somewhat larger at 0.76 for Eurozone countries and EU countries with a currency pegged to the euro, where monetary policy is less able to compensate for country-specific adjustments.
Next, we estimate state-dependent multipliers in the following way – first we estimate a baseline multiplier, which refers to an economy which is not in recession, open to trade,2 and where more than half of the adjustment takes place on the revenue side and where there is no stress in the credit market. Second, we estimate the additional impact on growth that occurs when each of these four states does occur.
In our baseline case, the impact of fiscal consolidation is not statistically significant. If consolidation is, however, implemented during a recession (the output cost of an adjustment of 1% of GDP increases by 0.9% of GDP with respect to the baseline case – or 1.3% for Eurozone countries and EU countries with a currency pegged to the euro). If the economy is closed to trade (the multiplier is again significantly larger, by approximately 0.7, while shifting the composition of the adjustment toward direct spending) raises the output cost by 0.3 to 0.4 percent of GDP. Finally, when fiscal consolidation is implemented during times of credit stress, the fiscal multiplier decreases significantly, particularly for euro area and euro-pegged countries, signalling the presence of confidence effects (see Table 1).
Table 1. Output effect of a 1% of GDP improvement in the structural primary balance
In Figure 1 we show a cumulative distribution of the multiplier values under the various states that actually occurred in our sample. The multiplier for each state of the economy is recovered by aggregation of the additional impacts to the baseline. The distribution appears to be quite asymmetric – a large majority of multipliers are smaller than the average level of 0.5 that we found in the state-independent model and for a substantial part of the sample they are non-significant – in a limited number of cases they are even significantly positive. In approximately one third of the cases, however, the negative multiplier is larger than average and consolidation can be over twice as costly.
Figure 1. Cumulative distribution of state-dependent multipliers
Authors' note: The views expressed in this article are the sole responsibility of the authors and should not be attributed to the European Commission.
Auerbach, A J, Y Gorodnichenko (2012), "Fiscal Multipliers in Recession and Expansion", NBER Chapters in National Bureau of Economic research Inc., Fiscal Policy after the Financial Crisis, pp. 63-98.
Blanchard, O J, D Leigh (2013), "Growth Forecast Errors and Fiscal Multipliers", American Economic Review, American Economic Association 103(3): 117-20, May.
Cugnasca, A, P Rother (2015), "Fiscal multipliers during consolidation: evidence from the European Union", Working Paper Series 1863, European Central Bank.
ECB (2014), "Fiscal Multipliers and the Timing of Consolidation", Monthly Bulletin, European Central Bank, April.
Gechert, S and W Henner (2012), "Fiscal Multipliers: A Meta Regression Analysis", IMK Working Paper 97-2012, IMK at the Hans Boeckler Foundation, Macroeconomic Policy Institute.
1 Croatia, which entered the European Union only in 2013, is excluded.
2 We consider open to trade the countries with the largest average sum of import and export share relative to GDP.