# Fiscal Expansions in the Era of Low Real Interest Rates

When the natural real interest rate is low, the power of monetary policy to revive the economy is limited due to the zero lower bound on the nominal interest rate. In most models, increasing government debt raises the natural real interest rate, which in turn restores the power of monetary policy to stimulate the economy. However, this paper shows that even in a low real interest rate environment, higher debt doesn’t necessarily raise the natural real interest rate, not if the current period of low real interest rates is expected to be short-lived.

The author notes that if the government accumulates substantial debt and the low real interest rate period is brief, it will have to finance this debt at positive real interest rates when the rates normalize. This financing will impose costs on the economy by triggering distortionary taxes or other forces that reduce output. The anticipation of these costs may induce further savings today and reduce the positive effect of the debt. If these costs are severe enough, the precautionary savings generated by the debt could be so large that the natural real interest rate falls.

In formalizing this argument, the author shows that the effect of increasing debt depends on the probability of reverting from a low real interest rate state to a business-as-usual state—one in which the real interest rate rise. If the probability of reversion is low, increasing debt will raise the natural real interest rate; but if the probability is high, increasing debt will lower the rate. Also, when the initial debt-to-GDP ratio is low, increasing debt raises the natural real interest rate; but it decreases the rate when the ratio is high. The threshold level of the debt-to-GDP ratio—the point at which the effect of increasing debt becomes perverse—is a function of the probability of reverting from a low real interest rate state to a business-as-usual state.

#### Key Findings

- Increasing debt raises the natural real interest rate at low levels of debt, whereas at high levels it lowers the rate. This threshold level of debt, beyond which the effect becomes perverse, is a function of the expected duration of the low interest rate state.
- If the expected duration of the low real interest rate state is 1.16 years or shorter, then increasing debt beyond the current US level of approximately 110 percent of GDP will reduce the natural real interest rate.
- If the expected duration is longer than 1.16 years, then there is room to use debt policy to address the prevailing low natural real interest rate.
- The amount of room depends on the expected duration of the low real interest rate state. If the expected duration is two years, for example, then increasing debt to as much as 250 percent of GDP will increase the natural real interest rate, but further increases in debt will have a perverse effect. If the expected duration is 10 years, then there is room to increase debt to about 500 percent of GDP.

#### Implications

It is likely that due the aftermath of the Great Recession, when the economy experienced low real interest rates for as long as seven years, US households expect the current low interest rate state to last longer than 1.16 years. This assumption suggests that raising government debt from its current levels will raise the natural real interest rate. However, in an environment with uncertainty about the path of future real interest rates, there are limits to the usefulness of raising government debt. Raising government debt may not be an effective means of addressing the problem of low real interest rates indefinitely.

These findings are particularly relevant now, given the situation that the COVID-19 pandemic has created in the United States, where the nominal interest is again close to 0 percent and uncertainty surrounds the future path of the real interest rate. The government has taken on additional debt to provide households with stimulus checks, but the effectiveness of these actions in stabilizing the natural real interest rate may depend on households’ expectations of how long the pandemic, and accompanying low real interest rates, will last, and how adversely it will ultimately affect the economy.

#### Abstract

Low natural real interest rates limit the power of monetary policy to revive the economy due to the zero lower bound (ZLB) on the nominal interest rate. Fiscal stabilization via higher government debt is frequently recommended as a policy to raise the natural real interest rate. This paper builds a non-Ricardian framework to study the tradeoffs associated with a debt-financed fiscal expansion and show that even in a low real interest rate environment, higher debt doesn’t necessarily raise the real interest rate. The effect of the expansion is non-monotonic: Increasing debt raises the natural real interest rate at low levels of debt, while at high levels it perversely lowers the natural real interest rate. This threshold level of debt, beyond which the effect becomes perverse, is a function of the expected duration of the low interest rate state. In a calibrated 60-period quantitative life cycle model with aggregate uncertainty, if the low state is expected to last two years, this threshold level of debt is 250 percent of the GDP. The insights from this paper are directly applicable to a ZLB episode in a model with nominal frictions.