A new paper by Bank of Italy Fabrizio Colonna, Roberto Torrini, and Eliana Viviano (“The profit share and firm mark-up: how to interpret them?”) explains why the ECB staff, ECB Schnabel, and ECB Panetta are wrong when arguing that rising profits have contributed to domestic price pressures. The widespread idea that “profits are pushing up prices” (typically based on the behavior of unit profits) is theoretically wrong and not supported by empirical evidence, excluding for a few sectors. Therefore, hoping that “core inflation will rapidly fall because firms will lower their margins” (as the argument goes), is most likely wrong. We assess the implications. (Slightly different conclusions apply to the US)
The issue
A few weeks ago, three ECB staff economists (Óscar Arce, Elke Hahn, and Gerrit Koester) have published an analysis “How tit-for-tat inflation can make everyone poorer” on the role of profit margins in consumer price dynamics. The analysis contains a chart -reported below- that has quickly become popular (and misinterpreted by many, including the ECB authors). The chart shows the contributions of unit profits and unit labor costs to domestic price pressures calculated by decomposing the GDP price deflator. Specifically, the chart shows that unit profits have increased in 2022 well above their pre-Covid average.
Based on this evidence, the ECB staff members argue that firms in 2022 were able to expand their profits/markups (“Many companies are apparently able to expand their profit margins”, “we are seeing rising profits”), and that the increase in profits contributed largely to the runup of consumers prices (“the effect of profits on domestic price pressures has been exceptional from a historical perspective”).
The ECB staff analysis led many to similar conclusions. For instance, ECB Isabel Schnabel said (here and here): “Rise in profits has played an important role for the rise in domestic inflation”. Along the same lines, ECB Panetta (here) said that: “[…] unit profits contributed to more than half of domestic price pressures in the last quarter of 2022. In some industries, profits are increasing strongly and retail prices are rising rapidly, in spite of the fact that wholesale prices have been decreasing for some time.”
Unfortunately, the ECB staff chart is misleading and the conclusion based on that chart is wrong.
Why it is wrong: Colonna et al. (2023)
Colonna et al. (2023) explains a basic micro concept: a profit share increase in value added is neither indicative of a higher markup nor of higher profits. The reason is simple: the ECB staff chart shows the share of unit profits in value added (emphasis on “share” and “value added”). Indeed, while the profit share is forced to increase (all else equal) when the markup increases, the opposite is not necessarily true: the profit share can increase even if the markup remains constant, for instance when total costs increase more than the wage bill (which is precisely what happened in 2022 in the Eurozone because of the energy shock). The intuition is simple: when facing a large negative supply shock, value added (which already accounts for the rise in intermediary input costs) shrinks. Therefore, the profit share rises even if the markup remains constant. Put it is simply: the size of the cake is smaller and the share that goes to profit is higher even if the markup is unchanged.
Colonna et al. (2023) derive formally for different production functions what we have described by words. The main message, under general assumptions, remains the same: the profit share of value added is not informative on the relationship between markups and price dynamics. And it is not even informative about the level of total profits. In other words, and despite the confusion in the debate, there is a difference between profit shares in value added, profit margins (markups), and level of aggregate profits; and the three of them do not necessarily move in the same direction.
(Please also note that even in case profits were elevated, it would be an indication of a strong demand, which is what we have been arguing (here and here) for the Euroarea against consensus and against the ECB doves, including Panetta)
The evidence of the markup
Colonna et al. (2023) provide an estimate of sectoral markups using national account data. The evidence points to no increase of the markup, except in a few sectors not exposed to international competition. Unfortunately, national accounts published by Eurostat do not allow to recover good proxies for markups, because the agency does not release neither estimates of the value of production nor the consumption of intermediate goods. However, the national institutes of statistics in Germany and Italy provide the necessary data to estimate the markups. The results of Colonna et al. (2023) are reported in Figure 1. The takeaway is clear: there is no visible increase in markups in 2022 across sectors, especially in Italy. The only exceptions are two sectors in Germany (“Constructions” and “Retail”) which are not exposed to international competition.
(For the record: Colonna et al. (2023) provide estimates also for the US. In this case, the evidence points to some increase in the markups in some sectors, although the BEA data have limitations).
Figure 1. Estimates of markup for Germany (2018=100) and Italy (q4.2018=100)
Implications for monetary policy
If the markup is not elevated core inflation can remain sticky, and the Governing Council can be surprised up again by the data. The dovish argument based on the GDP deflator decomposition is that if the markup is elevated, it can be a source of rapid disinflation going forward. However, Colonna et al. (2023) show that this is not the case in the Euroarea and that the widespread interpretation of the ECB staff chart is wrong (indeed, in some sectors there seems to be room for pushing the markups higher, which could be an additional inflationary source going forward). Therefore, if we cannot count on lower markups in the near-future and if wages will accelerate (as we expect), then the baseline is for core inflation to stay above the ECB target for longer. Under these conditions, it is very hard to see how the ECB can pause before reaching (at least) 4% terminal rate.