Inflation is back in the headlines. The decade following the Global Financial Crisis and up to the onset of the COVID-19 pandemic was characterised by low inflation almost everywhere (Jordà and Nechio 2020), even running precariously low in many advanced economies (primarily in the euro area, but also in the US). Speculation that the Phillips curve (Phillips 1958) was no longer a useful barometer for central banks was fuelled by an extended period in which inflation in advanced economies had been running consistently below target (over a decade in the case of the US). Researchers provided several explanations for the flattening of the Phillips curve (e.g. Mishkin et al. 2019, Tenreyro and MacLeay 2018) The conversation has since dramatically shifted, and central banks are now struggling to bring inflation down. Has the Phillips mechanism returned?
The pandemic generated all sorts of disruptions to supply chains as well as shifts in consumer demand away from services to goods. For example, di Giovanni et al. (2023) discuss how supply disruptions amplified the effects on inflation from demand stimulus. The invasion of Ukraine further aggravated the situation in commodity and oil markets. So, it is not difficult to come up with explanations for why inflation took off. Some of these explanations are behind us, however, so it is becoming harder to see how they would justify the persistence of high inflation.
Are demand factors equally relevant to understand the recent inflation resurgence? We think that they are. In many countries, pandemic fiscal support reached the highest levels on record. If the Phillips mechanism is not dead, could it explain why inflation has remained stubbornly high? The goal of our recent research (Jordà and Nechio 2023) has been to uncover the quantitative importance of these demand factors and to understand their contribution to recent inflation and wage dynamics.
The pandemic: A natural experiment
The COVID-19 pandemic was a global health crisis, but the manner in which governments responded to the crisis varied widely. Moreover, many governments instituted a variety of programmes rather than pursuing a single policy. The CARES Act (2020) and the ARP Act (2021) in the US included direct cash transfers to citizens, in addition to other loan assistance programmes and several other measures. Other countries, such as Germany, relied more heavily on short-time work programmes (Kurzarbeit) to preserve jobs and support demand, for example. Heterogeneity in the level of pandemic support is challenging, but it provides unique empirical possibilities.
In order to come up with a uniform measure of the degree of support experienced in each country, we focused on measuring how real disposable income (RDI) fared relative to its pre-COVID trend – what we will call the ‘RDI gap’. This variable is easy to construct and is available at a quarterly frequency for the countries in our sample. Moreover, it directly relates to consumers’ spending capacity, an important demand factor. How well does this variable correlate with the magnitude of the COVID fiscal effort? Figure 1 shows the correlation between our RDI gap measure and the IMF’s calculations of the COVID fiscal response relative to GDP (which is available only annually). The figure shows that this correlation is fairly strong among the countries in our sample, and that the RDI gap is likely a good proxy for the size of different fiscal packages.
Figure 1 The RDI gap mirrors the COVID fiscal effort
Notes: Sample: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Netherlands, Norway, Portugal, Spain, Sweden, the United Kingdom, and the United States. Real disposable income in deviations from trend.
Next, trying to disentangle the drivers of the inflation surge would be difficult. There is no shortage of potential explanations on the supply side. Instead, we will focus on demand factors as captured by the degree of pandemic fiscal support. A country’s choice of such support probably depends on political, social, and economic factors. Our working assumption will be that these factors (in the statistical parlance, characteristics that determine selection into treatment) remained more or less constant relative to the onset of the pandemic. As a result, one way to tease out the effect of these support programmes on inflation is to bin countries into two categories, according to how generous the support programmes were, and then compare the evolution of inflation before and after the pandemic – a strategy known as differences-in-differences. The hope is that the double-differencing (across time and across countries) will largely eliminate other confounding factors (mainly, supply-side factors that affected countries in more or less equal measure).
Figure 2 shows the variation in our RDI gap measure as of 2021Q1, approximately a year after the onset of the pandemic. It shows that there were considerable differences across countries along this dimension. Based on these differences, we bin countries into two groups, what we call an ‘active’ group that includes countries where the RDI gap was positive, and a ‘passive’ group, where the RDI gap was negative.
Figure 2 Ranking countries by their RDI gap
Notes: The figure reports RDI gaps by country as of 2021Q2. Countries with a positive gap implemented ‘active’ fiscal policies, while countries with a negative gap implemented ‘passive’ policies.
Evaluating the Philips mechanism: Local projections difference-in-differences
As Mishkin et al. (2019) and Tenreyro and McLeay (2018) discuss, empirically tracing the slope of the Phillips curve is difficult – central banks will work against the very mechanisms that give rise to the relationship between economic slack and inflation. One way to exploit the features of the data presented in the previous section and gain some traction in this identification quest is to implement a difference-in-difference estimation approach that in addition incorporates two components. First, a flexible dynamic component that allows demand factors to affect inflation not just on impact, but over a period of time. Second, the specification builds on a traditional hybrid Phillips curve that incorporates inflation expectations, previous inflation and the RDI gap (more details are available in Jordà and Nechio 2023). The estimation strategy builds on a recent paper on local projections difference-in-differences (LPDiD) by Dube et al. (2023).
As a complement to the analysis of demand-side inflation drivers, we also consider a similar Phillips curve specification for wages based on Galí (2011). The idea is to investigate whether wages are more or less sensitive to the same factors that affect inflation. A summary of the main results is presented in Figure 3. The figure displays the response of inflation to the RDI gap in the Phillips curve, over time. Two main lessons can be derived from the analysis. First, the effects on impact are relatively small (almost zero), yet they reach about 0.5 percentage points a year after impact. This means that for every 10 percentage point increase in the RDI gap during the pandemic results into prices being about 5 percentage points higher – a sizable response. Second, wages display a slightly different dynamic pattern, but essentially end up at about the same endpoint, that is, for a 10 percentage point increase in RDI, wages would be expected to be 5 percentage points higher after five quarters.
Figure 3 Pandemic support demand pushed inflation and wages up
Notes: Contribution to price and wage inflation from the effects of pandemic stimulus in the US on the RDI gap. Whisker bands are reported at the 90% confidence level.
A deeper dive into the determinants of wages
How do workers approach wage contract negotiations? When inflation is stable, looking at the previous contract is probably a good guideline. Moreover, looking forward, if inflation is stable, workers probably expect a similar rate of inflation going forward, so determining how to maintain purchasing power is easier. Finally, workers will likely demand a higher wage when demand for the goods they produce increases.
What about when inflation strays off target? We would expect to see less reliance of previous contracts as a baseline for negotiation. Workers know how to make up some of the lost purchasing power from the surge of inflation and will probably want to be more careful in factoring in that inflation may remain high in the future so that they do not suffer further erosion of purchasing power going forward. Is this what we see in the data?
Figure 4 presents a simple exercise. It asks, of the three main determinants of wages (previous wages, inflation expectations, and demand as measured by the RDI gap), which components appear to have more weight in predicting how wages will grow in the future. In order to measure any differences due to the surge in inflation after the pandemic, we consider two samples: one that ends in 2020Q1 (the start of the pandemic), and one that begins in 2020Q3 (thus removing the worst of the lockdown periods).
Figure 4 Expectations now play a bigger role in wage determination
Notes: Decomposition based on estimates of Eq. (8). 2020Q1 and Q2 omitted from the sample for clarity. Lag wages refers to the lag of wage inflation; inf. exp. refers to the Consensus Forecasts of 1-year-ahead inflation; and others collects all other regressors.
Figure 4 offers a clear picture. Before the pandemic lagged wage growth is the main component, followed by demand factors (labelled “other”, which also includes unobserved factors), and in last place, inflation expectations. However, since the pandemic, the role of inflation expectations has nearly doubled. This means that workers are now responding to fluctuations in inflation much more directly than in the past – a feature that could make bringing inflation back to heel more difficult and costly.
Inflation in many advanced economies surged (though at different times) in the middle of 2021. It peaked in many countries around the second half of 2022, though it has been difficult to bring it down back to target since and in fact, in some countries (such as the UK) inflation is only now starting to show signs of improvement. Our calculations show that an important share of this surge can be explained by the generous fiscal support packages implemented during the worst of the pandemic. Moreover, the surge of inflation has enhanced the role of inflation expectations in wage contract negotiations, which will likely make further declines in inflation more challenging.
Di Giovanni, J and S Kalemli-Özcan, A Silva, and M A Yildirim (2023), “Supply Chains Trade and Inflation", VoxEU.org, 7 January.
Dube, A, D Girardi, O Jordà, and A M Taylor (2023), “A local projections approach to difference-in-differences event studies”, NBER Working Paper No. w31184.
Galí, J (2011), “The Return of the Wage Phillips Curve”, Journal of the European Economic Association 9(3): 436—461.
Jordà, O and F Nechio (2020), “Inflation globally”, Serie Banca Central, análisis y políticas económicas No. 27.
Jordà, O and F Nechio (2023), “Inflation and wage growth since the pandemic”, European Economic Review 156, 104474.
Mishkin, F, A Sufi and P Hooper (2019), “The Phillips Curve: Dead or Alive”, VoxEU.org, 23 October.
Phillips, A W (1958), “The Relation between Unemployment and the Rate of Change of Money Wage Rates in the United Kingdom, 1861-1957”, Economica 25(100): 283—299.
Tenreyro, S and M McLeay (2018), “Optimal Inflation and the Identification of the Phillips Curve”, VoxEU.org, 3 July.