Implications for Fiscal Policy of Sustaining a Large Banking Sector

This paper investigates common determinants of fiscal crises using a standard Early Warning System (EWS) approach, with a particular focus on the role of the financial sector. We find that the probability of a fiscal crisis decreases with the level of domestic credit (as a share of GDP), but that at very high levels of credit it starts to increase. The critical threshold above which an increase in the level of credit signals an increase in the likelihood of a fiscal crisis, appears to be country (or group) specific, rather than an absolute level valid across all countries as previous research on this issue seemed to suggest. The paper also presents some preliminary results suggesting that, to determine a country's vulnerability to fiscal crises, it might play a role whether the credit is provided to the real economy (e.g. households, non-financial corporations) as opposed to the financial sector. In fact, after controlling for the stage of financial development of a country, the likelihood of a fiscal crisis decreases with the ratio of credit to the real economy (as a share of GDP) and increases with the ratio of credit to the financial sector (as a share of GDP). Consistent with previous findings in this literature, we find that higher levels of gross government debt, larger budget deficits, lower GDP growth and a loss of competitiveness (at least for more advanced economies) increase the likelihood of a fiscal crisis. We also find that countries with larger negative Net International Investment Positions (NIIPs) are more vulnerable to fiscal crises, especially if the level of debt liabilities (as opposed to FDIs) is large. This paper does not, however, account for other important factors that are likely to have an impact on a country's vulnerability to a fiscal crisis. These include the strength and credibility of domestic institutions, the potentially stabilising role of an independent monetary policy, progress made on structural reforms; and other political economy factors. These limitations inevitably call for some care in assessing the key policy implications of this paper