In this note, we review three recent academic papers: (i) Global supply chain pressures, inflation, and implications for monetary policy by Guido Ascari (DNB, University of Pavia), Dennis Bonam (DNB), and Andra Smadu (DNB), (ii) Why do we dislike inflation? by Stantcheva (Harvard), and (iii) The Cost of Money is Part of the Cost of Living) by Marijn A. Bolhuis (IMF), Judd Cramer (Harvard), Karl Oskar Schulz (Harvard), and Lawrence Summers (Harvard).
The first paper identifies a large role of the Euro on core HICP inflation. The paper has implications going forward, given the divergent monetary policy across the Atlantic. The second paper studies attitudes toward inflation. People dislike inflation because they perceive it to erode their purchasing power. Inflation is mostly blamed on the Biden administration and corporate greed. The third paper studies why US consumers’ sentiment is low despite favorable economic conditions. The authors argue that the cost of borrowing is a key input into consumer sentiment but not accurately reflected in published inflation rates.
Paper #1: Ascari et al. (2024)
A recent article (Global supply chain pressures, inflation, and implications for monetary policy) by Guido Ascari (DNB/University of Pavia), Dennis Bonam (DNB) and Andra Smadu (DNB) estimates the contributions to core HICP inflation since the pandemic. The impact of supply chain pressures is an important driver, but the paper also identifies a large role of the exchange rate. The paper has important implications going forward, given the divergent monetary policy across the Atlantic.
What the paper does
The authors estimate a vector autoregression to study the dynamic effects of supply chain shocks. The VAR includes industrial production, core inflation (defined as HICP inflation excluding energy), the shadow rate, the real exchange rate, the real price of oil, and an index of supply chain pressures. Supply chain shocks are identified by a combination of identification schemes: zero, sign, and narrative restrictions. For example, the supply chain shock is restricted to be positive in April 2020, when Covid led to the first wave of global lockdowns.
Main results
Supply chain pressures contributed considerably to inflation over time but the exchange rate also plays a non-negligible role. A historical decomposition based on the estimated VAR shows the contribution of the five shocks in the model to the core inflation rate (Figure 1). Supply disruptions (green bars) contributed to inflation around the global financial crisis and most notably the Covid episode, accounting for a large portion of the inflation surge. The size of the contribution has been declining since its peak in 2022 but remains large in 2023. The other large contributor is the exchange rate, as the (broad) Euro depreciated sharply at the beginning of the inflation surge, adding pressures on consumers prices. The role of demand is estimated small.
Figure 1. Contributions to (core HICP) inflation by the different shocks in VAR model of Ascari et al. (2024).
The optimal monetary policy response is to hike interest rates, though the size of monetary tightening depends on the economy’s reliance on global supply chains. The authors use a two-country New Keynesian model to study how the monetary authority would respond to a supply chain shock for different calibrations of the reliance on foreign inputs in domestic production. The peak interest rate hike implied by a Taylor rule in this model economy is nonlinear in the share of foreign inputs: The central bank will hike more if the reliance on foreign inputs is moderate. If the reliance is low, domestic inflation responds little to global supply chain shocks. If the reliance is very high, domestic production suffers strongly and the central bank faces a trade-off between output and inflation stabilization, resulting in a more muted interest rate response.
A critique of the paper
We see three main drawbacks. First, the paper includes oil prices but not natural gas prices, which have been a primary shock in the Euro area. The suspect is that the inclusion of natural gas prices in the VAR would lower the estimated contribution of global supply chain disruptions. Second, the demand side of the economy is poorly identified (essentially with the industrial production), as labor market variables (i.e. jobs openings) do not enter into the model. Finally, the empirical analysis has a short sample. This restricts the analysis to one episode of high inflation mixed with a decade of low inflation. While the time dimension of the sample is naturally restricted, the authors could gain additional insights from a cross-country comparison.
The model delivers a clear prediction for the best conduct of monetary policy, though alternative model setups may yield different results. The key result from the model is that the size of interest rate hikes by the central bank is hump-shaped in the reliance on foreign inputs: Increasing up to a certain point before falling as the reliance on global supply chains increases further. This result ultimately rests on the strength of the inflation and output response to global supply shocks, which need to be estimated carefully. Depending on the relative size of the output response and the central bank’s preference for output vs inflation stabilization, it could be the case that the optimal response to a supply chain shock is to cut interest rates to mitigate a drop in real activity.
Policy implications
Supply chain pressures and the exchange rate played an important role during the euro area’s inflation surge in 2022. A key question going forward not only concerns the persistence of the effect of the initial supply chain shock, but also the extent to which endogenous adjustment in the economy’s supply side will lead to a higher long-run inflation rate. A re-wiring of global supply chains takes time but appears inevitable (at least in some dimensions) given the euro area’s ambitions for a green transition and its need for greater geoeconomic independence given security concerns in Europe and beyond. Not only but with divergent monetary policy across the Atlantic, the role of the exchange rate going forward should not be dismissed. It remains unclear whether the ECB is considering the risks from the potential depreciation of the Euro.
Paper #2: Stantcheva (2024)
The paper Why do we dislike inflation? by Stefanie Stantcheva (Harvard) conducts a survey among a representative sample of US households to study people’s attitudes toward inflation. People dislike inflation mainly because they perceive it to erode their purchasing power. Many low-income households report struggling with paying their regular bills. Inflation is mostly blamed on the Biden administration and corporate greed.
Main results
People generally do not think of inflation as increasing all prices, and half of all respondents give an incorrect definition of inflation. Most other respondents give a somewhat (but not necessarily fully) correct definition. Wrong definitions usually focus on price increases being inherently excessive or “unfair”. In addition, survey respondents estimate inflation in 2023 at 5% (median) to 7.1% (mean), while expected inflation for 2024 is 5% (median) to 6.3% (median). These forecasts are considerably higher than forecasts by professional forecasters.
Respondents dislike inflation because they perceive it as diminishing their buying power since wage increases do not seem to match the pace of price increases. Many respondents report that inflation forces them to make budget adjustments, see the left panel of Figure 2. 71% of low-income respondents report difficulty with paying their regular bills, resulting in less saving (60% of low-income respondents), slower loan repayment (38%), and higher borrowing (31%). These shares are consistently lower for high-income households.
Respondents typically do not recognize positive associations with inflation such as low unemployment or decreasing real values of debt. In fact, only 25% of respondents believe that inflation and the unemployment rate are negatively related and 70% believe that high inflation indicates a poor state of the economy. 57% of respondents state that inflation makes repaying their debt harder, with 44% believing that inflation has increased the real value of their debt.
Respondents blame government and businesses for inflation. As shown in the right panel of Figure 2, “Biden and the administration” is the most common answer for the causes of inflation, followed by “Greed”. Democrats are more likely to refer to greed, while Republicans are more likely to refer to the government as the source of inflation. More generally, inflation ranks highly among various economic and social issues: 41% rank inflation as the top social issue and 33% as the top economic priority.
Figure 2. Survey responses from Stantcheva (2024). Left: Personal reactions to inflation. Right: Perceived reasons of high inflation.
A critique of the paper
The paper answers a highly important question with an often-neglected methodology: asking people what they think. Economic policy ultimately aims to increase consumer welfare, which is not only a function of actual welfare but also of consumers’ perceptions of their welfare. While most macroeconomic models focus on actual welfare, the survey results suggest that perceptions may vastly differ from economic reality: For example, most respondents are unaware that inflation decreases their debt burden and that higher inflation tends to arise in times of low unemployment.
The results provide important guidance for models aimed at providing policy advice and for policymakers more generally. Welfare analysis may need to take into account consumers’ perceptions as an explicit argument of the utility function. Large differences in survey responses across income groups also suggest that the inflation cost for a representative agent may underestimate the average utility cost experienced across the entire income distribution. Lastly, inflation is mostly perceived as negative because of an erosion in purchasing power. This provides a motivation to closely examine if the speed of price versus wage adjustment is accurately captured by workhorse macro models.
Policy implications
Inflation may be worse than our standard models suggest. This is both because consumers may perceive inflation as worse than it actually is but also because feedback effects on real economic activity via weak sentiment may generate adverse output effects. Effective policy communication in the short run and financial education in the long run are important policy tools to guide the public opinion on inflation.
Paper #3: Bolhuis et al. (2024)
A recent NBER working paper (The Cost of Money is Part of the Cost of Living) by Marijn A. Bolhuis (IMF), Judd Cramer (Harvard), Karl Oskar Schulz (Harvard), and Lawrence Summers (Harvard) studies why consumers’ sentiment is low although the unemployment rate is low and the inflation rate has been falling. The authors argue that the cost of borrowing is a key input into consumer sentiment scores but not accurately reflected in official inflation rates. They construct a new index that closes much of the gap between economic activity and sentiment.
What the paper does and main results
Including the cost of borrowing in the price index suggests stronger price increases and still elevated inflation rates. While housing and mortgage rates used to be included in the CPI, an undesirable feature of this approach was that inflation rates artificially increased at the start of a rate hike cycle. An updated methodology was developed in 1983 that removed these components and measured housing costs using imputed rents. However, this new methodology may underestimate the impact of housing costs on consumers’ budgets. Figure 3 shows that using the old methodology results in higher inflation rates (yellow line) compared to the official CPI inflation rate (blue). Inflation peaked around 18 percent and remained above 8% in November 2023.
Figure 3. Comparison of CPI inflation using current and traditional approach.
US consumer sentiment has seemed disconnected from economic activity since the Covid recession. Unemployment is low and inflation has moderated but consumer sentiment as measured by the Michigan survey remains weak. The authors argue that this is because published inflation rates do not accurately capture the costs of borrowing, which are a key input into consumers’ cost of living. Higher interest rates have increased mortgage payments, car payments, and an array of other credit payments that are prevalent in the US economy.
The alternative measure of inflation explains changes in sentiment (during and after Covid) better than CPI inflation. While the alternative CPI index has slightly weaker explanatory power for consumer sentiment over the 1978-2019 sample, it provides a more accurate description of sentiment for 2020-2023. This is because the alternative index indicates a larger rise in inflation, not because the index has a different relationship (coefficient estimate) with consumer sentiment than official CPI inflation. The alternative index explains over 70% of the gap between sentiment and official inflation over that period.
A critique of the paper
The paper advertises for more attention to a fundamental question of economics: What is the cost of living? While this question is of great practical importance, economists usually debate the measurement procedures underlying the data that feeds their models much less than the models themselves. The authors demonstrate that a different measurement can affect the outcome and apply this insight to explain the disconnect between sentiment and economic activity. While the paper’s result provides a plausible narrative, it should be tested for robustness. An important testable implication of their hypothesis is that more financially constrained consumers (all else equal) should report weaker sentiment. Another way of testing the paper implications is to extend the analysis to countries in which the CPI includes the direct passthrough of interest rates (such as in Canada and in Sweden).
There are other possible explanations for the sentiment puzzle and the paper would benefit from testing its proposed solution against these alternatives. For example, consumers may show weak sentiment because inflation rates are still high, even though they are falling.
Policy implications
Policymakers may face a dilemma in their fight against inflation. Interest rate hikes affect different prices in the consumer basket in different directions. While some goods and services become cheaper, others (such as mortgage financing) increase in price. This not only means that the estimated effects on the aggregate cost of living may be weaker than suggested by current estimates based on CPI inflation, but also that monetary policy may have stronger distributional effects than previously thought.