Ebook New Evidence on the Interest Rate Effects of Budget Deficits and Debt
Much controversy surrounds the quantitative effects of government debt and deficits on long-term real interest rates. Economic theory provides different answers depending on issues such as whether deficits reflect changes in government expenditures or shifts in the timing of taxes, and on the planning horizon of households who hold government debt and pay taxes. One might hope that empirical evidence could be brought to bear on this question, but here the results are just as ambiguous.
One major obstacle in obtaining empirical estimates is the need to isolate the effects of fiscal policy from the many other factors affecting interest rates. The most obvious of these factors is the state of the business cycle. If automatic fiscal stabilizers raise deficits during recessions, while at the same time long-term interest rates fall due to monetary easing, deficits and interest rates may be negatively correlated even if the partial effect of deficits on interest rates controlling for all other influences is positive.
This paper proposes to address this identification problem by focusing on the relationship between long-horizon expectations of both interest rates and fiscal variables. Deficits, debt, and interest rates expected to prevail several years in the future are presumably little affected by the current state of the business cycle, thus greatly reducing the reverse-causality effects induced by counter cyclical monetary policy and automatic fiscal stabilizers.
Of course, there are many conceivable factors that jointly determine fiscal variables and interest rates, and it is unlikely that a reduced-form regression would ever completely overcome this endogeneity problem, but focusing on long-horizon expectations is an important step in the right direction. Deficits projected several years into the future may be informative about the longer-run fiscal position, and may therefore approximate investors’ expectations about the eventual level of government debt relative to GDP. Such measures of expectations thus hold out the prospect of uncovering any causal relationship from fiscal variables to interest rates.
Expectations of future fiscal policy are proxied in this paper by projections published by the Congressional Budget Office (CBO) for the federal government’s unified budget deficit, the stock of federal government debt held by the public, and other fiscal variables, all expressed as percentages of projected GNP or GDP. The forecast horizon is five years in the future, which is the longest horizon for which a reasonably long time series of projections is available. Consistent with the use of 5-year-ahead projections of fiscal variables by the CBO, the analysis focuses on forward rates 5 years ahead embedded in the term structure of interest rates.
The results reported below show that a percentage point increase in the projected deficit-to-GDP ratio raises the five-year-ahead 10-year forward rate by 20 to 29 basis points; a typical estimate is about 22 basis points. The estimates are precise compared to most of the literature mentioned below. Similarly, a percentage point increase in the projected debt-to-GDP ratio raises the forward rate by about 3 to 4 basis points, and these estimates are statistically significant, too. These estimates are shown to be robust along many dimensions.
This study is by no means the first to use published projections of future budget deficits. Wachtel and Young (1987) use projections by the CBO and the Office of Management and Budget (OMB) to analyze changes in long-term interest rates on the day of the release of the respective projection.1 Unlike those shown here, their results therefore depend on correctly identifying the unanticipated component of the release. They find that a $1 billion increase in the projected deficit (at that time roughly 0.025 percent of nominal GDP) raises interest rates by between 0.15 and 0.4 basis points, depending on the maturity of the interest rate series and the source of the projections. Their estimates therefore imply an increase in interest rates on the order of 6 to 16 basis points in response to a percentage point increase in the deficit-to-GDP ratio. However, many of their estimates are statistically insignificant.
Cohen and Garnier (1991) and Elmendorf (1993) present results concerning the effect of deficit projections on the change in interest rates between release dates. Like the present one, these studies are based on the weaker assumption (in comparison to Wachtel and Young’s) that the deficit projections are good proxies of private agent’s expectations of future fiscal policy at the time of the release. The projections used in these studies, as well as in Wachtel and Young, are relatively short for the current and next fiscal year in Wachtel and Young and in Cohen and Garnier; for up to eight quarters ahead in Elmendorf. Forecasts at this horizon are presumably still affected by the state of the business cycle.
Using OMB projections, Cohen and Garnier find statistically significant effects of a percentage point unexpected (relative to the previous year’s projection) increase in the deficit-to-GDP ratio on interest rates on the order of 40 to 55 basis points. Using DRI forecasts, Elmendorf finds a statistically significant increase in interest rates at maturities up to five years of about 50 basis points, but the effects on long-term interest rates are smaller and statistically insignificant. Canzoneri, Cumby, and Diba (2002) use 5-year-ahead and 10-year-ahead CBO projections of cumulative budget deficits and study their effects on the spread between 5-year or 10-year, and 3-month Treasury yields. Their estimates are of similar magnitude as those reported in Cohen and Garnier and in Elmendorf, but are considerably more precise.
The present study confirms the importance of carefully measuring long-horizon expectations of deficits and debt for identifying their effects on interest rates. It departs from the previous studies in several respects, notably by using long-horizon forward rates as the dependent variable instead of current long-term rates or the slope of the yield curve. In comparison to previous studies, it also examines the role of additional regressors suggested by economic theory.
The specifications and the data used in the empirical analysis are introduced in sections 2 and 3, respectively. Baseline empirical results are presented in section 4. Because economic models differ in their view on whether deficits or the stock of debt is what matters for interest rate determination, I present results concerning the effects of both projected deficits and projected debt on interest rates. Taking the view that what ultimately matters is the stock of debt, Feldstein (1986) argues that empirical estimates of the interest rate effects of deficits depend on how persistent these deficits are expected to be. The relative magnitudes of the estimated effects of deficits and the estimated effects of debt reported below are shown to be consistent, under this view, with the observed historical autocorrelation of actual deficits.
Section 5 examines the importance of using fiscal and interest rate projections by comparing results to those obtained using current long-term interest rates and current fiscal variables. As shown there, removing the short end of the yield curve by focusing on long-horizon forward rates improves substantially the precision of the estimates of the interest rate effects of the fiscal variables. Given the historically large forecast errors in 5-year projections for deficits and debt, section 5 also examines whether these projections can be considered as proxies for market expectations of deficits and debt. Section 6 addresses several issues concerning the robustness of the baseline results.
Excluding measures of near-term economic conditions from the basic specification is found to be consistent with the data. Stability tests reveal substantial evidence for time variation in the effects of fiscal variables. For the regressions including debt, the evidence suggests a break in the mid-1980s, but the estimate for the later subsample is close to the full-sample estimate, whereas for the early subsample it is substantially higher. By contrast, the break in the relation-ship between deficits and interest rates has most likely occurred in the late 1990s. Since then, this relationship is estimated to have been negative, likely pointing to omitted variable problems associated with the unusual behavior of long-term yields and distant forward rates over recent years.
Section 7 discusses the predictions of the neoclassical growth model the simplest general equilibrium framework for this purpose for the relationship between the stock of debt and interest rates. Under plausible assumptions, the empirical results are close to the predictions from this model. Section 8 concludes.
Download
PDF Ebook New Evidence on the Interest Rate Effects of Budget Deficits and Debt
Posted in :
