What determines government spending multipliers?

Abstract

This paper studies how the effects of government spending vary with the economic environment. Using a panel of OECD countries, we identify fiscal shocks as residuals from an estimated spending rule and trace their macroeconomic impact under different conditions regarding the exchange rate regime, public indebtedness, and health of the financial system. The unconditional responses to a positive spending shock broadly confirm earlier findings. However, conditional responses differ systematically across exchange rate regimes, as real appreciation and external deficits occur mainly under currency pegs. We also find output and consumption multipliers to be unusually high during times of financial crisis.

Fiscal Policy and the Current Account

“…How restrictive is the assumption that government consumption does not react to output within a year? Corsetti et al (2010) discuss this issue in detail and, while it might be that during the 2008-09 crisis governments reacted quickly to economic conditions (maybe as fast as 5 to 8 months), this is more of the exception than the norm. Indeed, budgets are done on an annual basis and changes during the fiscal year are more cumbersome.…”

“…In this section, we follow this approach and restrict our analysis to shocks in government consumption even if this just provides a partial view on potential changes in fiscal policy. This approach has been followed by many of the recent empirical papers in the literature, including those looking at effects on the exchange rate and the current account (Corsetti et al, 2010, Beetsma et al, 2007or Monacelli and Perotti, 2007). …”

“…In fact, the evidence from VARs that use quarterly data show that in response to output shocks the response of government consumption is small and insignificant over the first quarters (in most cases it remains insignificant at any horizon). In addition, Corsetti et al (2010) also justify the use of annual data on the grounds that spending shocks might be foreseeable.…”

“…However, we are not the first looking at annual data. Recent papers in the literature such as Corsetti et al (2010) or Beetsma et al (2007) have made use of annual data. The motivation for using annual data is to look at a larger sample of countries over a longer time span.…”

Das leere Versprechen der aktiven Konjunktursteuerung

“…Eine Vielzahl von unterschiedlichen Nachfrageund Angebotsschocks kann diese konjunkturellen Schwan- 18 Diese Verzö-gerungen erhöhen das Risiko, dass sie nicht anti-, sondern sogar prozyklisch wirken. 19 Eine aktive Fiskalpolitik, die auf diese Weise selbst zu höheren Konjunkturschwankungen beiträgt, dürfte die Unsicherheit steigern und so die Investitionstätigkeit beeinträchtigen. 20 Ferner besteht das Risiko, dass die beschlossenen temporären Maßnahmen am Ende nicht zurückgenommen werden.…”