Fiscal Multipliers: Post-Crisis Controversies and their Implications

Iyanatul IslamThere has been a surge of scholarly work on estimating fiscal multipliers and assessing their implications in the wake of the global financial and economic crises of 2007-2009. A large number of countries enacted fiscal stimulus packages in response to the twin crises, but an evaluation of their efficacy remains mired in controversy. Such controversy has now evolved into disagreement over the contractionary consequences of fiscal consolidation exercises, most notably in the EU.

The empirical dissension that one observes reflects alternative values of fiscal multipliers that can range from zero to above 1. In traditional Keynesian models fiscal multipliers can be zero (meaning fiscal policy has no impact on output and employment) if the economy is operating at full capacity and if there is full crowding out through the interest rate and exchange rate channel. The notion that fiscal multipliers can be negative – which validates the idea of ‘expansionary’ fiscal contractions – depends upon strong assumptions with respect to Ricardian equivalence, crowding out through the interest channel and ‘confidence effects’ that can outweigh the contractionary consequences of fiscal policy.

The ‘New Keynesian’ framework that dominates academic research on macroeconomics through DSCGE models and seems to have influenced policy-makers in central banks of advanced economies is as malleable as its conventional Keynesian counterpart. A new Keynesian DSCGE model can become Keynesian or new classical if assumptions about key parameter values are changed. For example, it is possible to show that if there are a significant proportion of liquidity-constrained households in an economy, or supply of work-hours is demand-determined, or there is a so-called zero lower bound on monetary policy, then one can generate fiscal multipliers well above unity – which is the prediction of a simple Keynesian model.

There is, as noted, a growing and sizeable literature embodying estimates of fiscal multipliers both in aggregate and by type of measures. They are largely based on high-income countries and focus on discretionary interventions, not on automatic stabilizers. Figure 1 provides some updated estimates for USA and EU – two major G20 members – based on seven structural models. They corroborate the empirical regularity that spending multipliers are usually above 1 and higher than tax multipliers.

Figure 1: Fiscal multipliers by type from seven ‘structural’ models


Source: Derived from Coenen, Gunter et al (2012). “Effects of Fiscal Stimulus in Structural Models.” American Economic Journal: Macroeconomics, 4(1): 22-68.

Note: Government consumption multipliers assume 2-years accommodation of monetary policy.

The empirical literature on fiscal multipliers has concentrated on OECD G20 member states, but this is beginning to change. There has, in recent years, been a spate of studies that seek to assess fiscal multipliers for a number of non-OECD G20 member states.

In the cases of Brazil, China, India, GCC countries (of which Saudi Arabia is a typical example) and South Africa, the size of the fiscal multipliers, especially for capital expenditure, appears significant (that is, close to 1 or above). These findings, if taken seriously, challenge the pessimistic view that fiscal policy has no role to play in developing countries, because estimated fiscal multipliers are low and even negative. This might be true for pooled data for a group of countries and in some countries for the G20, but such a claim cannot be substantiated for other large and populous non-OECD G20 countries.

The policy message embedded in this assessment is that governments enacting fiscal stimulus packages should concentrate on spending measures rather than tax cuts. Furthermore, within the category of spending measures, governments should focus on capital expenditure, such as investments in employment-intensive infrastructure. This will yield higher dividends in terms of the impact on output and jobs for any given size of fiscal expansion. One can also interpret the results to mean that if fiscal consolidation programmes rely heavily on spending cuts, they are likely to be more contractionary than revenue raising measures (assuming there are no compensating changes through other policy instruments).


[1] Iyanatul Islam, Chief, Country Employment Policy Unit, ILO, Geneva.

[2] Fiscal multipliers (in an ex-ante sense) are defined as a percentage change of GDP (the numerator) given the same percentage change in discretionary fiscal policy (the denominator) entailing either expenditure or tax adjustments or both. Ramey, Valerie A. 2011. “Can Government Purchases Stimulate the Economy?” Journal of Economic Literature, 49(3): 673-85 offers a thorough review based on US data.

[3] IMF (2012) World Economic Outlook: coping with high debt and sluggish growth, October, Box1.1, p.41;European Commission (2012)  European Economic Forecast Autumn 2012, Brussels, Box 1.5, p.41

[4] Ilzetzki, Ethan et al (2010) ‘How Big (Small) are Fiscal Multipliers’? October, NBER Working Paper No.16479

[5] Central Bank of Brazil (2011) Inflation Report, March, pp.99-106; Cova, Petri et al (2010)

[6] ‘Macroeconomic Effects of China’s Fiscal Stimulus’, IDB Working Paper Series No.211; Guimaraes, Roberto (2010) ‘What are the Effects of Fiscal Policy Shocks in India?, March, mimeo, IMF; Espinoza, Raphael and Senhadji, Adbelhak (2011) ‘How Strong are Fiscal Multipliers in the GCC? An Empirical Investigation`? IMF Working Paper No.61; Jooste, Charl et al (2012) ‘Analysing the Effects of Fiscal Policy Shocks in the South African Economy’, February, Department of Economics Working Paper Series, University of Pretoria, South Africa;