I choose the title of this piece carefully because I do not want to overstate the evolution of the consensus view over the past several months. The consensus does not embrace the logic of 2-Stage Disinflation. Nor has the consensus even obviously considered that template. Rather, the consensus now accepts as a possibility that the labor market may be tight and that the so-called last mile of disinflation may accordingly require a period of below-trend growth. But, so far as I can tell, it does not use my reasoning to get there, not that it necessarily matters, practically speaking.
A key premise of the 2-Stage Disinflation hypothesis is that it is very unhelpful to think of the inflation surge during 2021-22 as having been driven by labor market tightness. It was associated with a rise of the demand for labor and with a decline in the unemployment rate. And because the decline of the unemployment rate was never reversed, the labor market is now tight.
But labor market overheating, per se, was not the main driver of the higher inflation. Rather, desired output in the business sector responded sluggishly to the fiscally driven surge of nominal demand, so that surge went into price not output, particularly during the Biden phase of the stimulus. And then when the fiscal impetus passed, and the Fed got to work, inflation abated, also largely irrespective of conditions in the labor market, although labor market conditions have probably tightened somewhat over the episode on balance. These may seem like trivial points, at least once made. *
We all agree that the inflation episode cannot be explained by the unemployment rate alone
The debate is about the add-ons — or epicycles, to be snarky
Data are actual to February.
And yet the consensus is oblivious to this template, in part because of academics’ and policy makers’ desire to shoehorn all inflation dynamics into the Phillips Curve framework. Here is how the received view described the run up and (mostly) reversal of the inflation surge. Under the impetus of fiscal stimulus and to a lesser extent easy monetary policy, desired real output surged after 2021. However, this surge of desired output ran into capacity constraints in the labor market. It is not so much that the unemployment rate was low as those labor market matching frictions — generated by a shift in the composition of aggregate demand — had made the short-term natural rate of unemployment surge. Accordingly, a satisfaction of the surge of desired real output required that the labor market overshoot (short-term) full employment, resulting in a surge of expected marginal cost, driven by higher wages and weaker productivity delivered by marginal workers. And it was this expected higher marginal cost that delivered the rise of underlying (non-shock) inflation, in accordance with prevailing New Keynesian Phillips Curve. And the output response was limited by marginal cost considerations and not, as I would have it, by the business sector believing that the excessive demand growth was itself probably transitory. In any case, according to the received view, the subsequent disinflation resulted from an easing of the labor market, driven largely by a decline of matching frictions rather than by much rise of the unemployment rate.
The novel and indeed ad hoc role of labor market frictions in the steeped inflation and disinflation in two generations may strike you as funny. It strikes me that way. The JOLTS report from which the vacancy data are taken did not even exist prior to the 2000s, although the theory suggesting they might be relevant did. And for the first decade of their existence, they were viewed mostly as something that might interest labor market specialists. This does not prove that the vacancy data and the labor market friction that they presumably measure is irrelevant. The Covid shock and the fiscal response to it were both unprecedented and may have created novel dynamics. The 2-stage Disinflation hypothesis also treats this episode as unique. But maybe we should be at least as suspicious of the received view as we are of my own speculations?
A key implication of my own speculation is that we can draw no inference about the state of the labor market from the recent disinflation even of wages. Wages chased price inflation higher, when it was being driven by developments that did not run mostly through labor market tightness. And wage growth has recently abated, as the taming of nominal demand has allowed price inflation to moderate. Or at least that is a possibility, which is given no quarter by the consensus.
However, the consensus is newly receptive to the idea that the labor market might be too tight and that the last mile of disinflation may require a period of below trend growth – to get the unemployment rate up. That conclusion is strongly suggested by the 2 Stage Disinflation hypothesis, which holds both that the labor market got tight, operating as a secondary and low-frequency influence on inflation, and that the recent deceleration of wage inflation tells us nothing about the labor market. But I am not sure the consensus has their own coherent story to tell here. Maybe they are just noticing that headline price disinflation has stalled and that the Fed is talking about a still tight labor market.
Recently, researchers at the NY Fed have produced a paper that might seem to reconcile recent developments within the standard New Keynesian Phillips Curve framework. They accept the premise that inflation must run through the labor market and expected cost. And their seemingly (to me) novel contribution is to argue that the labor market is in fact tight, but that the recent disinflation has been driven largely – not entirely — by the expectation – in the real economy- that monetary policy will deliver a rise of the unemployment rate in a timely fashion. But while that expectation can reduce inflation from its peak, it cannot return it all the way to target, say the authors.
To achieve that “last mile of disinflation,” a term the authors deploy, the Fed must now deliver on the expectation – in the real economy – that the unemployment rate will in fact rise. Please do not shoot the messenger, but this is another way to get to explaining the recent disinflation, while retaining the view that the labor market has got and stayed tight. Like me, the authors are dismissive of the signal from vacancies, although they do believe that labor market frictions have operated as an influence here.
That research from the NY Fed might provide some rationalization of the newish consensus view that the labor market may be tight and that the last mile of disinflation may be difficult, in the sense of requiring a growth and employment sacrifice. To repeat, I doubt the consensus has reasoned through this with as much precision as academic researchers – or even me. But it would fit. And, as mentioned, it preserves the – wrong to me – view that the earlier inflation surge was mostly about the state of the labor market.
So, what would be the practical implication of the consensus getting to my view that the labor market may be tight and that the last mile of disinflation may be difficult, even though the consensus is indifferent to – and unsurprisingly even innocent of – my sense of why? A while ago, I would have thought that the why mattered, because the why had a bearing on why I saw the last mile as (possibly) tougher than the consensus did. But at this point, all that probably matters is the implications. I see the labor market as possibly too tight, and so too does the consensus. We both recognize that this take is a possibility, rather than a slam dunk.
Speculating a bit beyond that to quantify, I would guess that the consensus assigns a slightly lower probability, still, to this idea than I do. That is a marginally hawkish consideration, if my sense is closer to the mark. On other hand, and related, the consensus acceptance here could be fragile, because people are not really sure why they believe what they now do. With a friendly wage number or a benign inflation print, then, people could be back to their old ways pretty quickly, unless, say, the Fed gets out the horse whip. We shall see.
* One of my more needling clients calls my take on this aspect of the story “monetarist,” because I simplistically invoke that inflation is just the difference between nominal demand and output growth. Hey, any port in a storm.
