Framing Out the Inflation Response to Tariffs
- The short-run impacts of tariffs on inflation are fairly easy to get a handle on with fairly direct pass-throughs into consumer prices, for a given tariff policy shock (the great unknown at the moment).
- Margin shifts and volume reductions (for PCE, whose weights dynamically shift) may play a role in attenuating the 1-for-1 impact, we would then some see some further downward adjustment in impacted goods’ pricing over time as trade flows reroute. Unlike in 2018 though, we are tariffing everyone everywhere so the help from diversion is more limited.
- Inflation’s medium-term path is more important for the Fed than the short-run impacts of tariffs on inflation, but you cannot assume those short-run impacts don’t matter at all given the possible magnitude of the m/m prints in Q2-Q3. The March SEP showed that the Fed sees the softer growth outlook canceling out the impacts from second round effects and expectations. A key, if for now reasonable, assumption. The uncomfortable echoes of 2021’s transitory are not lost on the Fed.
- The Fed will be very hesitant to hike into the peak inflationary months at the end of the spring and summer. If there is only a minimal near-term pass-through into wages and core services inflation, the slowing m/m pace of core PCE inflation by the fall should allow them to refocus on the labor market side of the mandate.
When thinking through the inflationary impacts of tariffs, I have fallen back on the following mental model. First there are the direct impacts, at roughly the same time supplementary and complementary goods prices respond, then more underlying inflation measures may or may not see a response, and finally, but running in parallel to the above, inflation expectations.
As I said last week after the ‘reciprocal’ tariffs were announced, assuming they or something only mildly toned down remains in place, inflation seems likely to rise into the 4-5% range, with more precision only possible if one makes overly specific assumptions about models and exact final tariff outcomes. That 4-5% range is contingent on a specific set of policies. It seems easier to make a current policy baseline and frame risks around that than try to constantly update a weighted moving average of scenarios in response to the volatile and never-ending news flow on the subject. As we learn more in the coming months, that baseline, including retaliation, will become more precise in regarding the direct impacts, although the indirect effects will be observed more slowly.
1a) Direct impacts on imported goods. These will show up fairly immediately over the course of the spring and summer. How much pass-through ultimately takes place depends on the balance between margin and volume preservation for firms. With more normal targeted tariffs there would some gradual attenuation over time as trade flows rerouted away from tariffed countries towards those with lower barriers; with current tariff rates applying to so many countries at such high levels this seems much more difficult to imagine.
1b) Pass-through to other related non-tariffed goods. This happens fairly quickly as well as relative price shifts take place (the classic example in the 2018-19 episode was the increase of prices in dyers, a complementary good, in response to tariffs on washing machines). Think about used autos responding to new car price changes as well firms managing the tariff hit across their list of SKUs but keeping some ex-tariff relative price rankings in place.
2) Wages and core services inflation see impacts on a 6-18m horizon. This is about more persistent inflationary pressures which build in response to the initial shocks and the relative price adjustments which are driven by those. Even with easier labor markets workers’ wages may see a very slow and gradual upward impulse in response, not necessarily taking it higher but making it decrease more slowly than it otherwise would. The Fed’s March SEP suggested that for the tariff policies in place then, combined with other immigration, regulatory, and fiscal policy shifts’ impacts it anticipated, this would roughly net out.
3) Longer-run inflation expectations across households and firms. Technically this could belong in the first position as we have already seen some preemptive moves higher in inflation expectations. My current belief, tentatively held, is that in order to spook the Fed these will either need to broaden out across all measures and forecast horizons, while moving notably higher still, or see moves notably above those that simple regression of spot inflation would suggest. The circumstances of this inflation shock heighten the risk of expectations deanchoring but it should not be the base case (here), precisely because the Fed continues to seem concerned about this possibility (monetary policy as an offset to tariff/fiscal policy), imparting a gradual hawkish push down on inflation over time. The real risk here is that these measures do start to broadly jump higher and the Fed’s reaction function suffers a notable hawkish lurch.
As we move further into the inflationary shock from these tariffs analytic tools to help separate out the direct and indirect impacts of tariffed goods, import price shocks, and trying to assess underlying inflation expectations moves higher (i.e. stripping out the fairly predictable movements in response to core and food and energy price developments). If the labor market starts weakening rapidly, the Fed will hope it can lean on these sorts of tools to justify looking through the inflation shock and focusing on the underlying labor market problem.
