The NY Fed has posted an analysis (“High Import Prices along the Global Supply Chain Feed Through to U.S. Domestic Prices”) by Mary Amiti, Sebastian Heise, and Aidan Wang.
Please, find below our review and comments.
What the paper does. The authors estimate the passthrough from import prices to US PPIs distinguishing between two channels: (i) strategic complementarities (that is, how much a US producer in a sector can raise its price in reaction to higher imported prices in the same sector) and (ii) marginal cost (that is, how much a US producer is forced to increased its own price in reaction to higher intermediate input prices). The authors use a two-step procedure and split the results between pre- and post-Covid period.
Results. Authors’ main results are shown in the Table and Figure below (note: the authors’ results are net of labor costs, that is they identify only the passthrough from imported prices to US PPIs). According to the authors’ estimates, a 10% increase in import prices resulted in 1 percent increase in PPI pre-Covid, with most of the effect coming from the marginal cost channel. The proportions between strategic complementarities and marginal cost channel remain the same in the post-Covid period but the magnitude of the passthrough more than doubles (from 1% to 2.56%).
However, the paper also shows that the overall contribution of the passthrough from import prices to domestic PPIs can explain only a small fraction of the recent runup (see gray bars in the Figure below). In other words, as they authors specify, [it seems] “there are also other important, and atypical, forces hitting the economy during the pandemic” that remain largely unexplained at the moment.
Implications for the Fed Board staff. The Fed Board staff framework (see Detmeister et al. (2014)) uses a measure of core import prices to capture relative price movements in an estimated Phillips curve model. In an estimated model, a 10% increase in (relative) core import prices result in about 0.8% upward pressure on core PCE price inflation (after 4 quarters). Therefore, under reasonable assumption (i.e., the distributors’ margins absorb part of the shock), the authors’ estimates are a confirmation of an estimated Phillips curve model. Having said so, a puzzle remains about the recent runup in domestic PPIs (and CPIs) that cannot be explained by higher imported prices. In any case, to the extent that imported prices will moderate going forward, we continue to believe that the Fed staff will judge the currently high inflation as “transitory”.