The New York Fed has published a post titled “What Are Consumers’ Inflation Expectations Telling Us Today?” by Olivier Armantier, Leo Goldman, Gizem Koşar, Giorgio Topa, Wilbert van der Klaauw, and John C. Williams.
To keep in mind
Research shows that (i) longer-term US consumers inflation expectations since the start of the pandemic have become less sensitive to inflation surprises and (ii) that revisions to short-term expectations translate to smaller revisions to longer-term expectations compared to the past. Taken together, these findings show that longer-term inflation expectations appear to be very well anchored (and even more than before the pandemic). In other words, research shows that households are still treating the current period as “exceptional” and the current high inflation is not leaving a mark on their expectations.
What the paper does
The theory (see for instance Orphanides and Williams (2005)) suggests that the sensitivity of expectations to inflation surprises is a useful metric to think about how shocks to inflation may get embedded into persistently high inflation and inflation expectations. For this reason, the authors use data from the New York Fed’s Survey of Consumer Expectations (SCE) and from the Michigan Survey of Consumers to examine how much consumers revise their inflation expectations in response to inflation surprises and to look at the pass-through of revisions in short-term inflation expectations to revisions in longer-term inflation expectations.
Results
The authors analysis contains two main results.
First, the authors claim that over the past two years consumers’ shorter-horizon expectations have been highly attuned to current inflation news: one-year-ahead inflation expectations are very responsive to inflation surprises, in a pattern similar to what we witnessed before the pandemic. In contrast, medium-term (three-year-ahead) inflation expectations are now far less responsive to inflation surprises than they were before the pandemic, indicating that consumers are taking less signal from recent movements than they did before Covid.
The table below shows the change of inflation expectations to inflation surprises before and since the pandemic. Before the pandemic a 1 percentage point surprise in inflation on average translated into a 0.7 percentage point revision in one-year-ahead expectations, but only a 0.45 percentage point revision in three-year-ahead expectations. However, the three-year-ahead expectations are now much less sensitive to surprises than before Covid: a 1 percentage point surprise translates only in 0.19 percentage point revision.
This result suggests that while consumers are highly attuned to current inflation news in updating their short-term inflation expectations, they are taking less signal than before the pandemic from the recent sharp movements in realized inflation when revising their medium-term expectations.
The second result has to do with the interaction between short-term and longer-term expectations. In this regard, the authors show that the pass-through from short-term to longer-term inflation expectations has been lower, on average, during the pandemic than it was before. The chart below plots the authors’ estimated pass-through coefficients from a series of rolling regressions taken over a six-month window. The evidence suggests that when consumers revise their short-term expectations, they do not revise their long-term expectations, or at least much less than before the pandemic.
The authors conclude that consumers are taking less signal than before the pandemic from inflation news in updating their longer-term expectations, and that households do not view the current elevated inflation as very long-lasting.
Comment
The NY Fed post comes at a time in which measures of long-term inflation expectations are moving sideways (Michigan, TIPS) or ticking down (SPF, NY-Fed SCE 3y-ahead) while observed inflation keeps trending higher. The NY Fed research is a useful analysis in our view because it suggests that US consumers are still treating the current period as “exceptional”. As such, at the moment high inflation is not leaving any mark on households’ expectations.
The observed behavior of longer-term expectations is good news for the Fed which is relying on this anchor as a guide for monetary policy. Having said so, despite their central role, there is a longer list of things we don’t understand about inflation expectations than of what we are confident we know (see various papers by Coibion and co-authors, and by Carola Binder). For instance, we don’t know whose expectations matters most for future inflation, over what horizon, and through what mechanism. We also don’t have robust, representative measures of the inflation expectations of different groups and we know very little about firms’ expectations. Finally, the literature is still debating whether long-term inflation expectations can be influenced (and how) by targeted communication (see for instance this newly released paper by Drager et al (2022)). While the theoretical mechanism linking expectations to actual inflation is clear, the Fed is forced to operate on an empirical minefield. In our view, relying too much on these indicators would probably neglect the complexity of the inflation process. The empirical evidence, so far, is in favor of the Fed and hopefully will remain as such. Nevertheless, in our view it shouldn’t taken for granted that expectations will remain so unresponsive, should inflation persist more than expected or the price level of some salient items remain elevated. As we discussed in the past, in our view focusing on a combination of actual inflation and long-term inflation expectation is probably a wise approach for the Fed going forward.