Why low interest rates force us to revisit the scope and role of fiscal policy: 45 takeaways

Over the last decade, it has become obvious that the decline in real interest rates forced us to revisit the scope and the role of fiscal policy. This is what I have tried to do in a book that I just finished. The book, Fiscal Policy Under Low Interest Rates, is now available on an MIT Press open source site, where I encourage you to leave comments and suggestions. I shall revise the book in light of those comments early next year, and the book will come out in hard copy at the end of 2022.

I thought the best way to sell the ideas in the book, and to tempt you to read it and potentially contribute to it, was to write 45 takeaways. So here they are:

On the evolution of rates

1. Safe real interest rates have declined steadily since the mid-1980s. The decline is due neither to the global financial crisis nor to the COVID-19 crisis. It has been common to all advanced economies.
2. Taking a much longer view, safe real interest rates have actually declined since the 14th century.... But the recent decline is much more pronounced.
3. The decline in safe real interest rates reflects a decline in neutral interest rates, i.e., the rates consistent with full employment. They reflect a chronic weakness in private demand—equivalently strong saving and weak investment, together with a strong demand for safe assets. This situation is what Alvin Hansen and, more recently, Lawrence Summers have called secular stagnation.
4. As central banks try to maintain full employment, the policy rates they set reflect this decline in safe neutral rates; central banks do not cause low rates.
5. The decline in rates must be traced to deep low-frequency factors, shifts in saving, investment, risk and market risk aversion, liquidity, and preference for liquidity. Many suspects have been identified. None has been convicted. Going through the list, few of them, however, seem likely to quickly turn around. One cannot and should not be sure, but secular stagnation appears likely to last.
6. The factor I see as potentially turning around is investment. The fight against global warming and green investment will require higher public investment but may also lead to spillovers to private investment. Depending on how it is financed, and on the size of spillovers, this could lead to higher neutral rates.
7. Demographics, be it the decrease in fertility and the increase in longevity, have contributed to the decline in rates and are likely to continue to contribute to low rates in the future.
8. There is little relation between growth rates and real rates. Theory does not imply a strong relation. The empirical evidence suggests that the decline in real rates is not due to a decline in growth rates.
9. Neutral rates depend on fiscal policy. A fiscal expansion leads to higher aggregate demand, higher neutral rates, and by implication an increase in actual rates. This may indeed be what happens in the near future, especially in the United States, given the strong fiscal expansion in 2021.
10. As neutral rates have declined over the last 30 years, they have, on the way down, crossed two thresholds: first, becoming lower than growth rates (r*<g), and second, often becoming lower than the lowest safe real rate achievable by monetary policy given the combination of low expected inflation and the effective lower bound on nominal rates (r*<r_min).

On debt sustainability

11. The fact that r<g has important implications for debt dynamics. Put simply, it gives countries more fiscal space. They can run (some) primary deficits and keep their debt ratios (the ratio of debt to GDP) constant, or even decrease them.
12. Debt sustainability is fundamentally a probabilistic concept. Debt can be said to be sustainable if the probability that the debt ratio explodes (or at least increases steadily) is very low.
13. Debt sustainability must be assessed in two steps. Given current policies, is debt sustainable? If not, will the government be willing and able to implement policies to make it sustainable?
14. The answer cannot be reduced to a simple universal debt or/and deficit number. The answer to the first question clearly depends on the first and the second moments of current and future primary balances, real interest rates, and growth rates. The answer to the second question depends on the nature and credibility of the government, the nature of political institutions, the initial level of taxation, etc.
15. Because of the complexity of the answer, simple rules, such as the Maastricht Treaty's 60 percent debt ratio and 3 percent deficit ratio numbers (both in ratios to GDP), or the German black zero rule, will not work well. They may ensure debt sustainability but at the cost of standing in the way of the appropriate fiscal policy, sometimes at very high output cost.
16. If, nevertheless, a rule is adopted, debt dynamics suggest that it should make the required minimum primary balance a function of debt service—defined as (r-g) times the debt ratio—rather than of the debt ratio itself and allow for deviations from this bound if the policy rate is constrained by the effective lower bound.
17. Public investment spending, to the extent that it generates future increases in fiscal revenues, can be partly financed by debt without threatening debt sustainability, something that any rule should reflect. Too often, the application of simple rules has led to inefficient cuts in public investment.
18. This does not imply, however, that all public investment should be financed by debt. Even if the investment generates large social returns, if it does not generate sufficient fiscal revenues, directly in the form of fees or indirectly from higher revenues from higher future output, it may still potentially threaten debt sustainability.
19. The approach to public investment should separate the decision about the level of public investment from its financing. Public investment, e.g., green investment, should proceed so long as the risk-adjusted social rate of return exceeds the corresponding borrowing rate. Whether it is financed by debt or by taxes should depend, however, both on how much it increases future fiscal revenues and on macro stabilization objectives, discussed below.
20. Fundamentals suggest that rates are likely to remain low for a long time. Sovereign bond markets are, however, subject to multiple equilibria/sunspots/sudden stops, in which the interest rate may increase quickly and substantially. Central banks can, by playing the role of stable investors and by being willing to intervene if needed, stop pure sunspot equilibria.
21. It is less clear that central banks can keep rates low when rates reflect worse fundamentals and a higher risk of debt unsustainability. In this case, the purchase of long maturity bonds financed by interest-paying central bank reserves is just a change in the composition of the liabilities of the consolidated government (the central government plus the central bank) and does not change, by itself, default risk. Indeed, if bank reserves are perceived as safer, long bonds will be seen as riskier and, as a result of quantitative easing, investors will require a higher spread. Long rates will go up, not down.
22. Longer maturity debt allows governments to reduce the effects of temporary increases in real rates, and to have more time to adjust to permanent increases. In this respect, the purchase by central banks of long maturity bonds financed by zero maturity interest-paying bank reserves decreases maturity and goes in the wrong direction.
23. The purchase of long bonds financed by interest-paying central bank reserves does not generate inflation any more than the purchase of long maturity bonds against short maturity assets by an investment fund. Nor is such purchase a bailout of governments by central banks.
24. The cancellation of the government debt on the balance sheet of the central bank has no effect on the liabilities of the consolidated government, and thus does not increase its fiscal space. It is at best useless and at worst counterproductive by decreasing the perceived independence of the central bank.

On optimal fiscal policy

25. On the welfare costs of debt: Other things equal, higher debt crowds out capital, and thus is widely perceived by policymakers and the general public as mortgaging the future and burdening future generations. The fact that r is less than g forces a reconsideration of this proposition.
26. A fundamental result in growth theory is that when r is less than g, an increase in debt can actually increase welfare for all generations. r<g means the net marginal product of capital is smaller than the investment needed to keep the capital growing at rate g. Thus, while lower capital indeed means lower future output, the decrease in the required investment allows higher future consumption.
27. This result, due to Edmund Phelps and Milton Friedman, and developed later by Peter Diamond, has been derived, however, under certainty, in which case interest rates are all equal, and equal to the net marginal product of capital.
28. Under uncertainty, there are many rates, from the safer rate to the (typically higher) average net marginal product of capital. Which rate is relevant in assessing whether debt increases or decreases welfare turns out to be difficult to assess. Theory suggests that the relevant rate depends on the nature of the production function, the existence of other distortions, and that it is somewhere between the safe rate and the average marginal product of capital.
29. Empirically, it is thus not clear whether the relevant rate is higher or lower than the growth rate. A pragmatic approach is to assume that, other things equal, debt is not good but is not very bad. And the lower the neutral rate, the less bad it is.
30. Turning from the costs to the benefits of debt and deficits, the second threshold is the most important. In the absence of an effective lower bound, we can think of central banks as setting the policy rate equal to the neutral rate, thus maintaining output at potential. When central banks are constrained by the effective lower bound, the policy rate cannot be set to the neutral rate and monetary policy cannot be used to maintain output at potential. This role then falls to fiscal policy.
31. The evidence on so-called multipliers —i.e., the effects of various dimensions of fiscal policy, be it spending, taxes, or the level of debt itself—on output, is that, most of the time, a fiscal expansion increases aggregate demand. And that the effect is stronger when monetary policy is at the effective lower bound.
32. In discussing optimal fiscal policy, it is useful to start with two extreme views. The first can be called the "pure public finance" view. It implicitly assumes that monetary policy can maintain output at potential and focuses on the role of debt in smoothing taxes in the face of variations in spending or in affecting the welfare of current versus future generations.
33. The second view assumes implicitly that monetary policy is not or cannot be used and that the main task of fiscal policy is thus macro stabilization. This view is known as the "functional finance" view, so baptized by Abba Lerner. In this case, fiscal policy must compensate for variations in private demand in order to maintain output at potential. If private demand is chronically low, then governments must run sustained deficits.
34. This suggests the following characterization of optimal fiscal policy, in relation to the two thresholds for the neutral rate. First, the lower the interest rate relative to the growth rate, the lower both the fiscal and welfare costs of debt. Second, the closer the interest rate is to the effective lower bound, the smaller the room for monetary policy to stabilize output, and the more important is the use of fiscal policy for macro stabilization.
35. The case for output stabilization is strongest when the effective lower bound is strictly binding. But it remains relevant when it is only potentially binding, leaving little room for monetary policy to react to decreases in private demand.
36. Put another way, the weaker private demand, and thus the lower the neutral rate, the smaller the costs and the larger the benefits of deficits and debt.
37. Put yet another way, the weaker private demand, and thus the lower the neutral rate, the larger the weight on functional finance, and the smaller the weight on pure public finance.
38. I interpret Modern Monetary Theory (MMT) as putting all the weight of functional finance, on the use of fiscal rather than monetary policy to keep output at potential. If so, this is too extreme a view.
39. Fiscal policy affects the neutral rate. One can then think of optimal fiscal policy as setting the neutral rate high enough that monetary policy has enough room to sustain output except against extremely bad demand shocks but low enough that the welfare and fiscal costs of debt remain limited.
40. Going back to debt sustainability: Chronically weak private demand may keep central banks at the effective lower bound and lead governments to run deficits so large as to lead to a steady increase in debt ratios, potentially raising issues of debt sustainability. This raises the issue of whether there are alternatives to deficits to sustain aggregate demand.
41. One approach is to relax the effective lower bound, either by increasing the target rate of inflation, and by implication average inflation and nominal interest rates, giving more room for monetary policy to decrease nominal rates if needed, or, as suggested by Kenneth Rogoff, by shifting to digital money and making cash illegal—which seems difficult to achieve, at least for the time being.
42. Another approach is to work on the factors that determine the neutral rate, especially if some of these factors represent distortions that should be removed, independently of their effect on private demand and fiscal policy. For example, offering more social insurance, such as Medicare for all, may decrease precautionary saving and increase the neutral rate without increasing deficits.

On fiscal policy in the flesh: Three applications

43. A case of too little? The shift from output stabilization to debt reduction in the wake of the global financial crisis in Europe was too strong and too costly, reflecting an excessive weight on the costs of debt and an insufficient belief in the adverse effects of contractionary fiscal policy on demand and output.
44. A case of just right? Faced with a strong case of secular stagnation, Japan has run large deficits for three decades and debt ratios have increased to very high levels, while the Bank of Japan remained at the effective lower bound. Was it the right strategy (if indeed it was a strategy)? The answer is a qualified yes, but, looking forward, the high debt ratios raise issues of debt sustainability. Alternative ways of boosting demand should be a high priority.
45. A case of too much? To boost the US recovery from the initial COVID-19 shocks, the Biden administration embarked in 2021 on a major fiscal expansion. The strategy (again, if indeed it was a strategy) was for fiscal policy to increase demand and thus increase the neutral rate, and for monetary policy to delay the adjustment of the policy rate to the neutral rate, and in the process generate temporary inflation. Inflation has turned out to be much higher than expected. Was the fiscal expansion too strong? Was the strategy a mistake?