2. Literature review

Date

The impacts of government debt and its effects on interest rates can be explained in a number of different ways, and with differing underlying assumptions. For a closed economy, and assuming that Ricardian equivalence1 does not hold, a budget deficit reduces national saving, which implies a shortage of funds to finance investment. This would place upward pressure on interest rates as firms compete to finance their investments from the existing pool of domestic saving (Elmendorf and Mankiw 1999).

If the flow of capital from overseas is assumed to be infinitely elastic, a budget deficit may not reduce the domestic capital stock as the adjustment can occur through higher capital inflows — which may not necessarily change interest rates.

Economic theory suggests that in an open economy with imperfect capital mobility2, the decline in national saving and rise in interest rates resulting from a budget deficit will induce a decline in domestic investment and net foreign investment. Under these circumstances higher net capital inflows would bid up the exchange rate.

Given some of the theoretical ambiguities about the connection between debt and interest rates, much of the literature has followed an empirical approach. However, the empirical evidence focusing specifically on the link between fiscal deficits and interest rates is mixed.

Barth et al. (1991) survey 42 papers through to 1989, of which 17 claimed positive effects, 19 showed negative effects, and 6 found mixed effects. In an additional survey of the empirical evidence, Gale and Orszag (2003) report that of 59 papers reviewed, 29 found a significant effect of deficits on interest rates, 19 found a predominantly insignificant effect, and 11 had mixed results. Gale and Orszag conclude that an increase in the fiscal deficit by one percentage point of GDP raises interest rates by about 30 to 60 basis points. In another survey, the European Commission (2004) concludes that the evidence points to an effect of 15 to 80 basis points.

The OECD (2009) also summarises recent empirical work on the link between fiscal policy and interest rates. Overall, the OECD’s review indicates that a 1 per cent deterioration in the fiscal balance produces a 10 to 60 basis point increase in long-term interest rates. Focusing on stock variables, the OECD’s survey suggests that a 1 per cent increase in net public debt raises long term interest rates by 3 to 50 basis points.

Studies based on cross-sectional data have typically found smaller effects than analysis of individual countries. Reinhart and Sack (2000) find that the impact of a deterioration in the fiscal balance by one percentage point of GDP in the current and following year raises government bond yields by nine basis points in OECD countries, and by 12 basis points in G-7 countries (Horton et al. 2009).

Empirical studies which consider the effect of current fiscal deficits on bond yields will inevitably suffer from some degree of endogeneity or reverse causality. For example, this can occur where high levels of debt lead to rising risk premia. One approach to alleviate this problem has been to study the announcement effects of fiscal policy on interest rates. This is because, in an efficient market, if one believes that larger budget deficits will increase interest rates, an announcement of deficits larger than previously anticipated will immediately boost long-term rates. This can occur through two channels. First, the term structure of interest rates hypothesises that the current yield on a long-term bond is related to the geometric average of current and future expected short-term bond yields. If agents believe larger future deficits will raise short-term yields as the deficits are incurred, long-term rates increase as soon as the deficit expectations are formed. Second, larger expected deficits may increase uncertainty about future monetary and other economic conditions and thereby increase the term premium embedded in long-term yields.

Early studies that adopted this approach include those of Cohen and Garnier (1991) and Barro (1991) who estimate the effect of projected government debt on the current real interest rate for the US. This was done by using forecasts of US federal deficits made by the Office of Management and Budget (OMB) or by the OECD.

Canzoneri et al. (2002) use the Congressional Budget Office’s (CBO) projected budget balances and find that an increase in projected future deficits averaging one percentage point of current GDP raises the long-term interest rate relative to the short-term rate by 53 to 60 basis points. Laubach (2003) also uses projections from CBO and the OMB and finds that a one percentage point increase in the deficit-to-GDP ratio raises long-term interest rates by 25 basis points. Engen and Hubbard (2004) claim that an increase in government debt of one percentage point of GDP, regardless of whether it is expected or current debt, increases the real interest rate by three basis points.

More recently, Chinn and Frankel (2005) show that current and expected levels of debt affect long term interest rates in Europe and the United States, but the estimates are sensitive to the sample period. Ardagna et al. (2004) find that a one percentage point of GDP increase in the primary deficit leads to a 10 basis point increase in the long-term rate, while public debt has a non linear effect.

Ardagna (2009) identifies periods of large fiscal contractions and expansions in OECD countries, and then studies how large changes affect interest rates. Ardagna’s results suggest that sharp changes in the fiscal stance have the largest and most significant impact on long-term bond yields. Interest rates on 10-year government bonds decrease, on average, by 124 basis points around episodes of fiscal consolidations and increase by 162 basis points during periods of loose fiscal policy.

Thomas and Wu (2009) study the impact of fiscal policy on interest rates by using CBO forecasts of budget deficits five years into the future as well as bond yields expected to prevail five years in the future. Results suggest that bond yields increase by 30 to 60 basis points for each percentage point increase in the deficit to GDP ratio expected to prevail five years in the future.

Ardagna et al. (2004) highlight the non linear effects of public debt on interest rates. Considering a panel of 16 OECD countries, the authors find that the impact of debt on long-term bond yields depends on initial debt levels. Higher initial debt raises the perception that governments will be less able to service their liabilities — leading to increased credit risk. Further, countries with large debt accumulation tend to be more at risk of inflationary pressures. These factors affect the long end of the term structure curve and raise borrowing costs for long-term government securities non-linearly.

For Australia, Comley et al. (2002) investigated the link between government debt and the real interest margin between Australian and US 10-year government bond yields over the period 1985 Q1 to 2001 Q2.3 Their results indicate that the real interest margin increases by around 20 basis points in response to a one percentage point of GDP deterioration in the headline budget balance in the short run. A one percentage point of GDP increase in the stock of public debt was found to increase the long-run real interest margin by around 15 basis points. As the authors note, these estimates are implausibly large for a small open economy and are likely to have been affected by high public debt levels over their sample period.


1 In its strict form, Barro’s (1974) Ricardian equivalence proposition asserts that government bonds do not constitute net wealth, meaning that
financing decisions have no real effects on consumption and interest rates. This implies that reductions in public saving resulting from tax cuts are offset one for one by increased private saving leaving consumption, national saving and interest rates unchanged. The same effect also holds for a deficit financed permanent spending increase as private agents cut their consumption in anticipation of future tax increases (Röhn 2010).

2 It is likely that the Australian economy faces some degree of capital immobility in the short run — but with long-run markets being closer to perfectly competitive.

3 Further details of the Comley et al. model and its results are outlined in Appendix 1.