There is an old story about a policeman who sees a drunk looking for something under a streetlight and asks what he is looking for. The drunk replies he has lost his car keys and the policeman joins in the search. A few minutes later the policeman asks if he is sure he lost them here and the drunk replies “No, I lost them in the park.” The policeman then asks “So why are you looking here?” to which the drunk replies “Because this is where the light is.”That story has much relevance for the economics profession’s approach to the Phillips curve.

Recently, there has been another flare-up in discussion of the Phillips curve involving Paul Krugman [Here], Chris House [Here], and Simon Wren-Lewis [Here]. It is précised by Mark Thoma [Here].

The question triggering the discussion is: Can Phillips curve (PC) theory account for inflation and the non-emergence of sustained deflation in the Great Recession? Four approaches are considered: (1) the original PC without inflation expectations; (2) the adaptive inflation expectations augmented PC; (3) the rational inflation expectations new classical vertical PC; and (4) the new Keynesian “sluggish price adjustment” PC that embeds a mix of lagged inflation and forward looking rational inflation expectations. The conclusion seems to be the original PC does best with regard to recent inflation experience but, of course, it fails with regard to past experience.

There is another obvious explanation that has been over-looked by mainstream economists for nearly forty years because they have preferred to keep looking under the “lamppost” of their conventional constructions. That alternative explanation rests on a combination of downward nominal wage rigidity plus incomplete incorporation of inflation expectations in a multi-sector economy.

The alternative has its roots in the seminal ideas of James Tobin, expressed in his 1971 presidential address to the American Economic Association [Here]. Tobin identified the critical multi-sector aspect of inflation for the Phillips curve. However, he failed to identify the issue of incomplete incorporation of inflation expectations, and nor did he present a mathematical model of the inflation process, which is a cardinal sin in today’s profession.

The mainstream profession’s focus has been, and continues to be, the formation of expectations (i.e. adaptive, rational, and most recently, near-rational). In my view, the real issue is the extent to which inflation expectations are incorporated into wage behavior. Workers may have absolutely correct expectations of inflation but not incorporate them into nominal wage demands because of job fears.

Unfortunately, this approach requires a multi-sector perspective, which makes it more complicated. However, it turns out that it is possible to construct stylized multi-sector models that aggregate nicely.

The history of debate about the PC and this proposed alternative approach are comprehensively presented in a 2012 paper titled “The Economics of the Phillips Curve: Formation of Inflation Expectations versus Incorporation of Inflation Expectations” [Published here: early working paper here].

The original version of the multi-sector model involves partial downward nominal wage rigidity and zero incorporation of inflation expectations in sectors with unemployment. That model was published in 1994 in the Scandinavian Journal of Economics [Here], before the two famous Brookings Papers articles by Akerlof, Dickens, and Perry [1996 here: 2000 here]. The latter also adopted a multi-sector (actually multi-firm) frame, but they then veered off target and fell back into the expectation formation trap by focusing on near-rational expectations.

The simplest version of the multi-sector model has complete downward nominal wage rigidity and zero incorporation of inflation expectations in sectors with unemployment. That model does very well explaining inflation during the Great Recession. As the proportion of sectors with unemployment diminishes we should expect inflation to start increasing. That is my “old Keynesian” prediction, which I hope can inoculate against the inevitable claims that will come from the Friedman – Lucas tribe that we have over-shot the natural rate of unemployment.

Lastly, there is a clear lesson for economists. If the profession would move away from the lamppost, it might actually find the keys to inflation and the Phillips curve.