Recently, U.S. Treasury Secretary Janet Yellen claimed confidently that “we’re not seeing the usual signs of a weakening labor market that would make you fear a recession… so it doesn’t seem at all like it’s requiring higher unemployment…” to bring inflation under control. It gives the impression that she does not believe the Phillips curve effect is out there, not even in the short term. This is somehow a bold assertion against the academic belief over half a century and against the economic projection model adopted by the Federal Reserve.
This is a very strong assertion. On the theory side, this amounts to refuting an important linkage between nominal and real variables (namely, inflation and unemployment). This also contradicts the official tone of the Federal Reserve that rate hikes would cool down real activities, including the labor market. On the reality side, it is obviously too early to claim inflation has been contained where unemployment would not worsen, given the latter is a severe lagger. In fact, layoffs have been getting more frequent while job openings data exhibit a downtrend.
Perhaps the Phillips curve is not very convenient from a modelling perspective, for it links a ratio (unemployment rate) with a growth rate over time (inflation rate). A more sensible approach would be to link the inflation rate with the wage growth rate (goods price change with labor price change), or to link up aggregate inventory with unemployment (goods excess with labor excess). Then, price change and excess (supply) of each market (goods or labor) are related by the supply-demand schedule. In theory, transmission should go via price changes.
Of course, one can skip these technical details by studying the composite of the above relationships directly if a clear picture stands out. If not, then one has to go back to the above breakdown so as to identify what blurs the picture. Here, I propose a composite, which is a variant of the Phillips curve, by replacing the unemployment rate with non-farm payrolls (NFP) YoY growth. Correlation study suggests CPI growth leads NFP growth by four quarters. The full-time series begins from 1940 to the latest, while the red one begins when inflation peaked in the last year.
As the accompanying chart shows, a proportional relationship is much more apparent under this YoY-to-YoY schedule compared with when the unemployment rate is used. More importantly, when only the recent inflation downtrend episode is concerned, the relationship is even stronger (as evidenced by a steeper slope) than the historical one. Instead of the “no trade-off” as claimed by Yellen, it is the other way around that the trade-off is even stronger in this round of tightening.
If the locus of red circles stops now, one can still argue that inflation reaches 2 percent without leading to NFP contraction. But no one can confirm this at the moment. Should the locus continue, then the U.S. will reach deflation and then job contraction. Yellen is no different from the economists she refers to. Who will eat the words is yet to be seen.
KC Law, Ka Chung
