In a classical way of thinking the economic term natural rate of unemployment has as underlying assumption that the rate of unemployment have a “constant” equilibrium value that will oscillate much like the output gap. The natural rate of unemployment is then the equilibrium rate, which would be observable when the economy is at full employment output.
In the early 60’s when this concept was forged, disagreements on the validity of this concepts arise, and then start growing, when with the first oil chock in many OECD countries almost tripled unemployment rate and didn’t managed to decrease it considerably – A equilibrium unemployment rate was not to be found.
In 1986 Blanchard and summers approach an idea of hysteresis first introduced by Phelps in 72, with their insiders – outsiders model. The hysteresis idea contrasts with the natural rate of unemployment – while this second assumes stability over the time, (assuming therefore that the rate of unemployment is a stationary one), the hysteresis hypothesis states that this is a non stationary variable, specifically a unit root, which imply that a momentarily shock will have permanent effects in the unemployment rate, meaning that the unemployment rate is explained by its past values.
As it was said before this hypothesis came with the increasing unemployment rate in most of the European countries since the 70’s. As an example the following chart shows the moving average to 5 years, of the Portuguese unemployment rate since 1960’s until 2011, with the inclusion of perspectives until 2014:
As it is observable there is an increasing tendency on the unemployment rate growth. Also it is observable that the shocks suffered by Portugal in the mid 70’s are quite visible (the end of the war, with the end of the regime and the oil chock all together), but even when they come down in the 80’s the never retracted as much as the increased, and with the crisis a new shock, even greater is observed.
The extended literature that appeared after the Blanchard and Summers paper backed so far the hysteresis hypothesis empirically, although some of the reason for it is maybe due to the lack of strong tests in this part of econometrics. To resolve this, stronger tests that have been developed, have been applied to this particular problem, and showed the same results. Although there is still some “place” for flaws in studies, the hysteresis hypothesis, that came to deny the natural rate of unemployment, is strongly backed up by empirical evidence.
Given this it is of strong importance to adopt contra-cyclical policies, as the unemployment rate, otherwise, may rise forever. This hypothesis highlight the need of careful policymaking – a mistake today has much more impact with hysteresis that without.
Knut Roed – Unemployment hysteresis- Macro evidence from 16 OECD Countries
Ameco data base
Maria Roque Martins nº540