Reassessing the NAIRUs after the Crisis

Abstract

The financial crisis has resulted in a substantial increase in unemployment in the OECD. This paper shows that this increase has reversed the reduction in structural unemployment which has been estimated to have occurred in most OECD countries since the late 1990s. Structural unemployment is defined as a time-varying NAIRU derived from the information contained in a reduced Phillips curve equation (linking inflation to the unemployment gap) by means of a Kalman filter. The overall limited revisions in historical NAIRU estimated in 2008 after such a large labour market shock support the robustness of the OECD approach. This approach is therefore extended to almost all OECD countries. Alternative specifications of the Phillips curve are proposed for some specific groups of countries. Un réexamen des NAIRUs après la crise La crise de la crise financière a entraîné une augmentation importante du chômage dans l'OCDE. Ce document montre que cette augmentation a inversé la tendance á la réduction du chômage structurel que l’on avait été estimé avoir eu lieu dans la plupart des pays de l'OCDE depuis la fin des années 1990. Le chômage structurel est défini comme à un NAIRU variant dans le temps et dérivé de l'information contenue dans une équation de courbe de Phillips réduite (reliant l'inflation à l'écart du chômage au chômage structurel) au moyen d'un filtre de Kalman. Les révisions dans l’ensemble limitées du NAIRU historique estimés en 2008 après un tel choc sur le marché du travail supportent la robustesse de l'approche de l'OCDE. Cette approche est donc étendue à presque tous les pays de l'OCDE. Des spécifications alternatives de la courbe de Phillips sont proposées pour certains groupes spécifiques du pays.unemployment, NAIRU, Phillips curves, structural unemployment, chômage structurel, chômage, courbe de Phillips, NAIRU

The OECD potential output estimation methodology

“…Unemployment (UN) and labour force (LF) series are sourced from national Labour Force Surveys and used to compute the unemployment rate (UNR = UN / LF * 100). 11 The trend unemployment rate, also called the non-accelerating inflation rate of unemployment (NAIRU), is then estimated using a Kalman filter embedded within a forward-looking Phillips curve, as described in detail in Rusticelli, Turner and Cavalleri (2015[17]) 12 :…”

“…5 With rare exceptions, the labour income share in the United States has remained between 65% and 70% since 1960 ( Figure 1, Panel A). 6 In other countries, such as Canada, the labour income share appears to have declined since the 1960s, a trend which has generated a rich literature (see for example Schwellnus et al (2018[8]), Pak and Schwellnus (2019 [9]), Karabarbounis and Neiman (2013 [10]), Growiec, McAdam and Mućk (2018 [11]), Schneider (2011 [12]), Krämer (2010 [13]) or Bentolila and Saint-Paul (2003[14])). For the purposes of potential output estimation, the value of α does not have a large impact on either the level or growth rate of potential output, although it impacts the breakdown of potential growth between different components.…”

“…The latest Eurostat population projections are normally used for European Countries and the latest United Nations population projections for other countries. 11 Labour-force based series for Argentina and India are not available from the OECD Economic Outlook database and are sourced instead from the IMF World Economic Outlook or the International Labour Organisation's databases. 12 It has been argued that the unemployment gap ( − ) may be higher than indicated by the unemployment rate itself because of a rise in part-time employment (and more specifically involuntary part-time employment) in many OECD countries.…”

Rescuing the Phillips curve

“…In addition, inflation expectations which are well anchored to central banks' inflation targets have weakened the relationship between inflation and labour market imbalances. However, previous attempts to include central bank inflation targets in the Phillips curve estimation did not produce satisfactory results (Guichard and Rusticelli, 2011). Finally, greater economic integration may have shifted the drivers of the inflation process away from domestic pressure towards more global drivers.…”

Elemzések időben változó paraméterű modellekkel

“…Hozzá kell tennünk, hogy csak akkor pontosabb, amikor egyáltalán sikerül szignifikáns kapcsolatot találni, ami viszont a konstrukció miatt nehezebb, mintGuichard-Rusticelli [2011] esetében. Mint az összes TV-NAIRU folyamatot becslő szerző, nekünk is hozzá kell tennünk, hogy maga a NAIRU pontbecslés inkább csak indikatív, a tendencia következtetése azonban robusztusabb.…”