There exists an extensive empirical literature on the macroeconomic effects of government deficits and debt. A few studies investigate the relationship between fiscal deficits and output directly, while most papers consider whether interest rates rise in response to an increase in government debt. If no increase in interest rates can be measured in response to an increase in debt, it would confirm Ricardian equivalence proposition. If a link between interest rates and debt can be established, the finding does not necessarily imply support for the conventional approach with crowding-out through the aggregate savings effect, as the increase in interest rates may be – at least partly – due to an increase in the risk premium and not necessarily from the savings channel.(105) Portfolio
effects may affect the risk premia across sovereigns in case of efficient international financial markets or between sovereign and corporate spreads.(106)
Box III.3.1 finds that the effect of increasing interest rates on government bond yields is significantly higher for high debt countries than for countries with average debt levels, which points to the importance of the risk premium-channel relative to the savings channel.
(104) See for instance Blanchard (1984)
(105) The empirical studies that explicitly aim to test the Ricardian equivalence proposition seem at best unable to reject the proposition. For instance, while Evans (1988) and Rockerbie (1997) could not reject it, Evans (1993) and Lopez et al. (2000) find evidence against the hypothesis. On the other hand, Afonso (2008) focuses on EU countries and dismisses debt neutrality, while Reitschuler (2008), who tests the theory in 11 new member states, cannot reject the null in all countries.
(106) Evidence from the US suggests that an increase in government debt tends to reduce the spread between government and corporate bonds. Krishnamurthy and Vissing-Jorgensen (2007) show that an increase in Treasury debt held by public leads to decline in the yield spread of AAA corporate debt over Treasuries.
This finding is also supported by a much cited recent study by Reinhart and Rogoff (2010), which shows evidence of a link between growth and debt when debt-to-GDP levels are high. The authors use an extensive database of forty-four countries and about 200 years of observations. They find that the growth impact of government debt is negligible for levels of debt below a threshold of 90 percent of GDP, but above that threshold median growth rates fall by one percent, and average growth falls considerably more. It is not clear, however, whether the causality is unidirectional or whether this observation partly reflects the fact that countries with low growth are more likely to have encountered debt sustainability problems. (107)
There is some empirical evidence which suggests that government debt is associated with an increase in real interest rates on government bonds, ranging from a 1 to 6 basis-point increase in interest rates on government bonds, for each 1 percentage point increase in the government debt to GDP ratio (see, for example, Laubach (2009). (108) Engen and
Hubbard (2004) obtain similar results when investigating the impact of government debt. There is no consensus on whether the estimated increase is confined to government bonds or whether it affects the general level of interest rates in the country in question. In countries that rely heavily on external financing of investment, an increase in government debt could lead to a general increase in the risk premium and raise interest rates for both government and private bonds.(109)
The extensive literature has been more recently surveyed by Gale and Orszag (2003), who also point out that most studies tend find a significant, positive link between fiscal deficits and debt and the long-term interest rate. However, they also find
(107) Another example of a study that investigates the
relationship between fiscal deficits and output directly is Sawhney and DiPietro (1994) who, after controlling for a number of variables affecting growth, find no evidence that government debt would retard growth.
(108) Laubach aims to tackle the endogeneity problems (of higher interest rates also affecting debt and deficit dynamics and higher debt affecting fiscal policies) through taking into account the anticipated path of the fiscal variables as well.
(109) This can reflect, for example, increased currency risk or increased corporate default risk if large fiscal consolidation needs risk hampering economic activity (this assumes that taxes are distortionary). In a financial crisis, it can reflect the reduced capacity of the government to support the financial sector to avoid systemic crises.
that while about half of the studies found a predominantly positive and significant impact, about one-third concluded that the relationship was not significant. The authors claim that international capital flows significantly reduce the impact of budget deficits on interest rates as part of the increase demand for funds could be met with foreign capital. Hence, only part of the impact of government borrowing affects domestic interest rates. This explanation assumes an increased domestic demand for funds to be the reason for an interest rate increase and would be consistent with the crowding out channel through national savings. Two, recent studies examined this interplay of foreign capital markets and the effect of government deficits in a European context. Claeys et al. (2008) and Faini (2006) find significant spill- over effects and explain that debt-financed expansionary fiscal policy not only increases the sovereign interest rate spread vis-à-vis the rest of the region but would also raise the level of interest rates across the entire currency union. They also points out that in countries with a high debt-to- GDP ratio this harmful spill-over is even stronger.
Note that possible negative effects of higher public debt on growth can be offset by the positive effects of debt-financed public expenditures, if this raises the economy's productive capacity. In this vein Tanzi and Zee (1997) advocate the accumulation of human capital, while Semmler et al. (2007) distinguish between different types of productive government spending and find that the government debt ratio could be stabilized as long as government investments are used in a growth- maximizing way. Greiner and Fincke (2009), by analysing an endogenous growth model, find that when real wages are sufficiently flexible in the long-run, the optimal government debt-to-GDP ratio is zero as the associated debt service distracts resources away from productive investment. However, when real wages are rigid, maintaining non-zero government debt can be beneficial, as the proceeds of the additional public investment in terms of higher employment and output can compensate for the interest burden.
Box III.3.1: The effect of deficits and debt on the sovereign risk premium
The main factors determining the risk premium on sovereign debt are: the perceived credit risk of this debt, the liquidity of the debt issuance, the degree of global risk aversion and the microstructure of the bond market. The level of debt, in turn, is also affected by these determinants, as the cost of financing debt affects the government's willingness to issue bonds and add to the stock of debt. These interactions make the relationship between the level of government debt and the sovereign interest rate complex (1).
While a deteriorating domestic outlook for fiscal deficits and debt is associated with higher interest rates, there is also evidence that countries with high debt levels are more likely to experience increases in interest rates if debt increases further. An econometric analysis using data for euro-area countries over the period 2003q1-2009q2 suggests an interaction of debt and deficits in determining government bond yields. In particular, the impact of deteriorated fiscal balances on government bond yields appears significantly higher for high debt countries than for countries with average debt levels. Figure B.1 plots the estimated linear relationship between spreads and deficit for a given level of initial debt while controlling for a number of other determinants of interest rate spreads vs. Germany (such as liquidity conditions, and global risk aversion). It shows that the impact of higher deficit on the yield spread tends to increase significantly with a higher initial level of debt.
Figure B.1: The impact of budgetary balance on 10-year government bond spread at high and average debt levels
Source: Commission services
(1) This box draws heavily on Barrios, Iversen, Lewandowska and Setzer (2009), "Determinants of
intra-euro area government bond spreads during the financial crisis" European Economy, Economic Papers 388, November 2009.
3.3. THE QUEST MODEL: OUTPUT EFFECTS OF