This is part II of a note on the “Fiscal Theory of the Price Level” (or FTPL) and related literature. Part I is here. In this part of the note, we review the remaining three papers: “Fiscal Influences on Inflation in OECD Countries, 2020-2022” by Robert J. Barro (Harvard) and Francesco Bianchi (Johns Hopkins), “The Monetary Financing of a Large Fiscal Shock” by Pedro Teles (Banco de Portgual) and Oreste Tristani (ECB), and “Do Large Fiscal Deficits Cause Inflation? The Historical Record” by Michael D. Bordo (Rutgers University) and Mickey D. Levy (Berenberg). Finally, we conclude. FRB-US simulations will follow.
Paper #1: Barro and Bianchi (2023)
The paper “Fiscal Influences on Inflation in OECD Countries, 2020-2022” by Robert J. Barro (Harvard) and Francesco Bianchi (Johns Hopkins) studies the role of fiscal expansion in explaining the inflation rate in 37 OECD countries for the 2020-2022 period. The empirical results suggest an important role for fiscal policy in explaining the recent inflation episode.
What the paper does
Given the intertemporal government budget constraint and some assumptions, the authors derive an expression for the share of the rise in the debt-to-GDP ratio that is “financed” through inflation. The theory’s centrepiece is the government’s intertemporal budget constraint, which relates a country’s inflation rate in 2020-2022 (relative to a baseline rate) to a composite government-spending variable. This variable equals the cumulative increase in the ratio of government expenditure to GDP from 2020 to 2022, divided by the ratio of public debt to GDP in 2019 and the duration of the debt in 2019.
Main results
The expression can be directly taken to the data. The authors use information on headline inflation rates and government spending for 37 OECD countries for the 2020-2022 period. Figure 1 shows the relation between inflation (vertical axis) and the measure of government spending (horizontal axis). The inflation rate per year is evaluated relative to the 2010-2019 average. Countries with higher inflation have higher government spending over the sample period. According to the authors, the slope of the line gives an indication about the share of financing of government expenditure through inflation. Overall, 40-50% of the required financing comes from the negative effect of inflation on the real value of public debt, with the rest being financed by tax increases or future spending cuts. The share is estimated to be lower for euro area countries.
Figure 1. Main result of Barro and Bianchi (2023).
A critique of the paper
The regressions do not control for many factors that could affect the results. Since the equation derived by the authors is not immune from errors, it cannot explain the rise in inflation across OECD countries using only information on government spending. Cross-country differences can arise because of variation in expectations about future government spending or taxes, differences in inflation measurement or the types of government spending, different choices of governments for the share of inflation to be inflated away, or different exposure to alternative shocks (monetary, supply, inflation expectations). Some of these shocks are measurable, and it would add credibility to the results to control for them.
The link between inflation and government spending underlying the regression analysis is derived using several simplifying assumptions. For example, “the paths of real GDP […] and the real interest rate […] are assumed to be invariant with the fiscal/monetary shocks” (p. 4). This assumption abstracts from the effects of inflation on real variables. This also means that the approach does not resolve the simultaneity bias / reverse causality arising between inflation on one side of the equation and government spending on the other side: Inflation can be affected by government spending, but government spending can also be affected by inflation. This means that cross-country differences in government spending need not be exogenous to inflation and thus violates the authors’ key condition for the identification of causal effects.
Policy implications
Governments may choose to finance some of their spending via inflation. In the setup of this paper, the authors consider a government that optimally chooses which share of its spending should be financed by rising prices. While institutional factors such as central bank independence may limit the ability of governments to directly target an inflation rate optimal for their financing decision, governments have an incentive to increase inflation to finance their past and current spending.
Smaller government deficits may add additional downward pressure on inflation. Falling levels of inflation have already been attributed to a mix of weakening supply shocks and tight monetary policy. However, if the FTPL is believed to be a driving force of inflation, then the tighter government budgets in the post-Covid world may also contribute to a decrease in inflation rates.
Paper #2: Teles and Tristani (2024)
The recent paper “The Monetary Financing of a Large Fiscal Shock” by Pedro Teles (Banco de Portgual) and Oreste Tristani (ECB) studies the optimal response to inflation when governments issue long-term debt and prices are not fully flexible because of sticky information. The recent increase in inflation is persistent and large, suggesting that the Covid inflation may have partly resulted as the optimal response to rising debt levels following a strong fiscal expansion.
What the paper does
The model features sticky information to model price rigidities. With fully flexible prices, inflation can be used to maintain stable real debt such that taxes do not need to be adjusted. In principle, this mechanism can lead to very volatile inflation rates. Under standard Calvo frictions to price adjustment, volatile inflation is undesirable because it leads to deviations from optimal prices, which have negative output effects. Under Calvo pricing, firms face a constant probability of adjusting prices in any given period and set a new price to maximize the present discount value of future profits. In Teles and Tristani (2024), firms select a sequence of future prices instead of a constant price (as in Calvo) to maximize expected profits. In turn, this different pricing mechanism affects the output cost from inflation.
Main results
The optimal inflation response to a large increase in public debt is a gradual, large, and persistent increase in the inflation rate. In the model, the cost of debt financing from inflation is lower than from a tax increase. This result rests on the combination of sticky information and long-term debt. Sticky information makes an inflation episode less costly the further in advance the episode has been anticipated because many firms will be able to adjust their prices given the new information about (future aggregate) inflation. If public debt has a short maturity, then an inflationary episode in the future is unable to devalue it. Instead, long-term debt can be inflated away this way. The point of the paper is that in a model where firms are restricted in setting price plans as in the sticky information model of Mankiw and Reis (2002), a pandemic-size increase in government debt ought to be financed in part by a large and highly-persistent increase in inflation. Such an increase would be costly, but less so than the other form of financing of the government debt available in the model, i.e. a permanent increase in taxes.
The model yields peak inflation above 7% in response to a pandemic-sized debt expansion. Figure 2 shows the inflation response following a 20% increase in the euro area debt-to-GDP ratio under different assumptions for price setting. The dotted line indicates the flexible price case, in which inflation jumps to over 7% on impact and declines slowly, remaining above 3% six years after impact. Under Calvo pricing (dashed line), there is no inflation response under the optimal policy because of the large costs of inflation. Sticky information (solid line) results in a hump-shaped inflation response: inflation rises gradually, peaks above 7%, and declines very slowly.
Figure 2. Optimal inflation response to a Covid-sized increase in the debt-to-GDP ratio under different assumptions for price setting, based on the model of Teles and Tristani (2024).
A critique of the paper
The quantitative evidence yields a very large and (unrealistic) persistent inflation response. It would be interesting to see how this result changes once additional frictions are introduced to make the model more realistic. If we believe that Covid inflation was largely driven by the desire to inflate away government debt, the impulse response under the assumption of flexible prices appears to provide a more accurate description of observed inflation dynamics than the highly persistent response under the sticky information assumption. In any case, under both assumptions, the inflation process is unrealistically persistent; a discussion of why and how to make it more realistic would improve the paper.
Policy implications
Governments may have a stronger incentive to inflate away debt than previously thought. Teles and Tristani (2024) supports the results from the literature on the Fiscal Theory of the Price Level, which suggests that unfinanced fiscal expansions are a key driver of inflation (for example Bianchi et al. 2023). It also provides additional motivation for borrowing with long maturities, since this can not only limit rollover risk and allow governments to lock in financing rates while rates are low, but also facilitates inflating away the real value of debt when necessary.
However, it is questionable whether the suggested policy is actually optimal in a real-world context. The model may underestimate the cost of inflation that households experience (see our discussion of Stantcheva (2024)) and does not consider the political pressure that would go along with it. Inflating away debt may also result in serious reputational damage with higher financing costs going forward. These considerations limit the policy implications of the paper.
Paper #3: Bordo and Levy (2021)
The paper “Do Large Fiscal Deficits Cause Inflation? The Historical Record” by Michael D. Bordo (Rutgers University) and Mickey D. Levy (Berenberg) provides a historical account of several fiscal expansions and discusses the relation between government deficits and inflation.
What the paper does
The authors provide historical examples of fiscal-monetary interactions in support of different theories linking fiscal policy and inflation:
- Standard Keynesian models: increases in aggregate demand increase real output until full employment is reached, then lead to rising prices once supply is constrained.
- Fiscal dominance of monetary policy: Expansionary fiscal policy can force the monetary authority to inflate away debt through expansionary monetary policy, assuming that debt levels are forced to remain sustainable.
- The fiscal theory of the price level.
Main results
Wars are periods of distinctly large government expenditure which often coincide with high inflation. In many cases, governments financed their war efforts by printing money. Early examples are the Seven Years War in Sweden, the American Revolutionary War, and the French Revolutionary War. All of these episodes were characterized by high inflation or even hyperinflation. The World Wars of the 20th century saw large increases in debt-to-GDP ratios among the participating nations. For example, France’s debt-to-GDP ratio peaked at almost 200% in WWI. The fiscal expansions were mostly debt-financed, and borrowing was facilitated by low interest rates, leading to high inflation rates. These examples support the idea of fiscal dominance of monetary policy.
Differences in fiscal discipline may explain some inflationary episodes. The authors consider the experience of France and the UK after World War I (Figure 3). Coming out of the war, both countries faced a quickly rising price level, large fiscal deficits, and a devalued exchange rate. In contrast to the UK, France did not resume the gold standard immediately after the war, instead delaying its policy stabilization until 1926. During this time, France’s price level rose dramatically. The authors consider this evidence consistent with the Fiscal Theory. However, it should be noted that even after the government deficit was contained, France’s price level kept rising strongly while it remained stable in the UK.
Figure 3. Left panel: Price levels around WWI in UK, USA, France (1910 normalized to 100). Right panel: Government budget deficit in percent of GDP for the UK and France. Bordo and Levy (2021).
Fiscal expansions are not always followed by inflationary episodes, even if monetary policy is also accommodative. The Great Recession is a leading example. The US government’s economic stimulus programmes were the largest peacetime expenditure since the Great Depression. Simultaneously, zero interest rate policy and quantitative easing provided strong monetary stimulus. Nonetheless, inflation did not rise. The authors argue this is because the monetary and fiscal stimulus failed to spark a sustained acceleration in aggregate demand. This supports the Keynesian idea of aggregate demand being the cause of inflation, not fiscal deficits per se.
A critique of the paper
The paper provides a useful set of case studies that can be used to put current inflation episodes in context. Some historical examples may be close to natural experiments and therefore insightful for the understanding of causal links, at least given the economic structure prevalent at the time. Despite its title, the paper does not provide clear evidence on the causal relationship between fiscal deficits and inflation. There is only one run of history, and each war or recession are subject to many simultaneous forces. As usual, correlation is not causality, and a counterfactual is always necessary.
It is unclear what the size of the relation between fiscal spending and inflation is. The fiscal expansions considered by the authors varied in size, and there is no sense of the magnitude of the effects on inflation. This magnitude is likely to depend on surrounding economic conditions, making it hard to estimate. Nonetheless, there would be value in providing an approximate estimate of the causal impact of a marginal fiscal expansion on inflation.
Policy implications
As debt levels rise, the risks of fiscal dominance rise as well. The burden from interest payments on government debt is a function of the value of outstanding debt and the coupon payments due on this debt. This can limit the government’s ability in fighting inflation effectively: High inflation means the real value of government debt is inflated away plus keeping rates low means the debt burden can be easily refinanced. Even with central bank independence officially in place, political and public pressures on central bankers can be large in practice, leading to dangerous policy outcomes. Rising debt levels imply fiscal dominance is not only a concern of the developed economies’ past, or of today’s developing economies, but may also be a feature of the future.
General conclusion
Fiscal matters, period. Fiscal issues will likely dominate going forward. In the last 15 years or so, the (inter)national sport has become blaming the central banks, no matter what happens in the economy. Central banks have their own responsibilities, of course. But the literature is clear: monetary policy operates for a given fiscal policy. The aim of our note is to remind that there are different channels of the interaction between monetary and fiscal policies. From an empirical point of view, it remains hard to identify fiscal shocks properly. Nevertheless, under reasonable assumptions, it seems clear that fiscal policy has been part of the problem post-Covid. The result is larger public debts, in some cases (US) the debt/GDP ratio is on an unstable path. The FRB-US simulations that we will circulate in the next weeks will show that stabilizing the debt/GDP ratio is not an easy task at this point and that it requires a protracted period of fiscal consolidation.