Compared to the Bernanke-Yellin fiasco – a decade of missing inflation targets on the downside, something that should in principle never happen -- the current Fed has done spectacularly well. It did wait a bit too long to start disinflation in September 2021 and then in reducing disinflation in December 2023. Notwithstanding supply chain shocks and large changes in fiscal deficits, however, the US economy has experienced a remarkable recovery from the COVID recession, better than any other large market economy and with no more inflation.
Nevertheless, it is time to pull together some of the ways the Fed could further improve its performance. In no particular order:
1. The Fed should reaffirm its commitment to Flexible Average Inflation targeting, clarifying that
a. the “average” is a forward-looking average of expectation, not an arithmetic mean of inflation rates over some arbitrary past period.
b. both the target and the flexibility are in the service of maximizing real income
2. There should be much more frequent movements by much smaller amounts of the vector of policy instruments the Fed uses to inflate and disinflate.
3. Movement in the instrument vector should depend on information that arrives between movements. As such, movement in the instruments cannot be unconditionally predicted. Predictions of future movement depend on prediction of the new information.
4. A corollary of #3, like any other price in an efficient market that incorporates all available information, movement in the instrument vector is as likely to be toward inflation as toward disinflation. No “forward guidance” about movement in the policy vector should, nor indeed, CAN be given.
5. The Fed should carry out periodic public retrospective evaluations of its performance (perhaps as part of its semi-annual reports to Congress) both in relation to the information it had available at the time of past decisions and how/if its use of information has changed as a result of the outcomes of those decisions. In other words, did it make the right decision based on the model of the economy it was using at the time of the decision and how/if it has changed the model?[1]
6. A corollary of #5, the Fed should also publicly and periodically re-examine its inflation target, the degree of and reasons for the price stickiness it is based on and the size and frequency of expected shocks.
7. The Fed and market participants need additional data with which to make decisions
a. Bureau of Labor Statistics should begin to collect and publish disaggregated wage indices in addition to or in place of the data published on disaggregated average remuneration.
b. Treasury should begin issuing “Trillionth” securities that pay a fixed proportion of future GDP. Price fluctuations of these securities would give market participants, but especially the Fed, a real time reading of market expectations of growth over various tenors. [If these securities should be deemed not to be “debt” for purposes of the “debt ceiling,” that would be an additional benefit.]
c. Treasury should issue more shorter tenor Treasury Inflation Protected Securities (TIPS). TIPS of 5- and 10-year tenors are available, but 1-, 2-, and 3-year TIPS would be helpful.
Image Prompt: Group of well-dressed people around a table evidently intent on making a decision. Around them, perhaps in bleachers, there are a multitude of spectators.
[Standard bleg: Although my style is know-it-all-ism, I do sometime entertain the thought that, here and there, I might be mistaken on some minor detail. I would welcome comments on these views.]
[1] “Model” meaning everything used to process the information on which decisions were made, not just mathematical models
Fantastic post, and I basically agree with all of these. That being said, could you expand on "the “average” is a forward-looking average of expectation" - How exactly would you structure this?
Thomas - you indicated earlier today PCE is down to 2% already, can you point us to a source ?