The Danish flexicurity model and the crisis. Torben M. Andersen University of Aarhus CEPR, CESifo and IZA

Transcription

1 The Danish flexicurity model and the crisis Torben M. Andersen University of Aarhus CEPR, CESifo and IZA

2 The Danish Flexicurity Model Particular triple: Lax EPL Flexible hiring/firing rules Generous UIB Strong focus on ALMP Unemployment Insurance Active labour market policy

3 Flexicurity and the Great Recession More exposed to shocks Large employment response to be expected Higher risk of persistence higher unemployment response and generous unemployment insurance Can the active labour market policy cope with an increase in unemployment? Eroding financial balance of the model

4 The Great Recession and the role of institutions/policies Notion of flexibility: EPL UIB ALMP Adaptation of labour input to output - +? Speed of adjustment (actual unemployment Avoiding path dependence (change in structural unemployment) Aggregate structural unemployment) Disaggregate (groups/types of labour) -? - +/? - - -/?

5 Q1-199 Q Q Q Q Q Q Q Q Q Q1-2 Q1-21 Q1-22 Q1-23 Q1-24 Q1-25 Q1-26 Q1-27 Q1-28 Q1-29 % Unemployment Denmark OECD 6 4 2

6 1991K4 1992K4 1993K4 1994K4 1995K4 1996K4 1997K4 1998K4 1999K4 2K4 21K4 22K4 23K4 24K4 25K4 26K4 27K4 28K4 29K4 1. Persons Coping with the crisis Boom-bust pattern Prior to crisis: Very low level of unemployment Clear signs of overheating Predicted Actual Post Crisis Adjustment to lover level of activity 24 23

7 Employment adjustment relative to output hours vs employees 3,5 3 2,5 2 Employment Hours 1,5 1,5 SVK NZL DNK ITA DEU NLD KOR BEL JPN FRA SWE HUN GBR PRT CZE FIN AUT IRL USA CAN ESP NOR -,5

8 Share of working hours in labour adjusment, % US EPL and share of hours adjustment CAN UK AUS IRE NZL DEN FIN HUN GER AUT BEL ITA SWE NET CZE NOR FRA POR SPA EPL index

9 Puzzle Substantial variation in the employment to output adjustment Not clearly related to labour market policies/institutions Not even for the relative share of employment and hours All variables in logs

10 Hirings % High turnover in the labour market TUR ESP DNK IRL FIN POL USA SWE ITA SVK BEL CHE FRA PRT GBR HUN CZE GRC AUT NOR SVN DEU Separation %

11 % Job finding rates remain high but are pro-cyclical Exit from unemployment 28 q2 29 q3

12 High incidence of short-term unemployment Short term long term CAN NZL NOR AUS SWE DEN ICE USA FIN AUT ITA UK POL NET CH SLV JPN IRE SPA GRE CZE FRA BEL GER POR HUN

13 % Burden of unemployment falling on youth

14 Share of total unemployment Share of total unemployemnt (a) Short term unemployment (b) Long term unemployment CAN NZL USA AUS ICE SWE FIN DEN AUT UK ITA NET POL IRE JPN CZE CH BEL SLV HUN GER NOR GRE FRA SPA POR HUN GER POR GRE CZE FRA IRE BEL CH JPN NET SPA UK ITA POL SLV AUT USA AUS FIN SWE ICE DEN NOR CAN NZL EPL EPL

15

16 Autocorrelation Persistence in the labour market Difficult to measure How to separate Exogenous sources (shock) Endogenous sources (policy, institutions etc.) 1,,95,9,85 DEN BEL AUT NOR AUS POR CAN GER NZL IRE NET SPA ITA FIN UK CH GRE FRA JPN LUX SWE TUR Different measures indicate relatively strong persistence,8 ICE,75,75,8,85,9,95 1, Sign metric

17 Persistence Trade-off: Volatility vs. persistence Less volatile markets = more persistence? Very weak evidence also for the opposite 1,,95,9,85,8 Unemployment persistence and gross flows PRT GER AUT USA NLD CHE BEL IRL ITA SPA FRA SWE FIN UK GRC NOR y = -,18x +,9871 R² =,654 CAN ISL TUR,75, 2, 4, 6, 8, Gross job flows

18 Autocorrelation unemployment Social insurance a source of persistence? Insurance= lack of flexibility= strong persistence Incentives Reservation wages Norms 1,95,9,85,8 FRA ITA JPN GER NZL BEL AUT IRE UK FIN NOR AUS POR SPA CAN SWE NET DEN No clear evidence supporting this BUT persistence is crucial,75,7,65 USA,2,4,6,8 1 Automatic stabilizer

19

20 % High level of activation Share of unemployed in activation Insured Social assistance 5 1. kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt 29

21 Benefits and activation requirements:

22 Jan-4 Maj 24 Sep-4 Jan-5 Maj 25 Sep-5 Jan-6 Maj 26 Sep-6 Jan-7 Maj 27 Sep-7 Jan-8 Maj 28 Sep-8 Jan-9 Maj 29 Sep-9 Jan-1 Maj 21 Sep-1 % Dureation weeks Maintaining efficiency of active labour market policies? Larger inflow 1,8 4 1,6 3,5 Different composition (more core workers deadweightloss) Programmes less effective? 1,4 1,2 1,8,6,4,2 Activated in % of labour force (left scale) Average duration (right scale) 3 2,5 2 1,5 1,5 More difficult to screen participants

23 1. kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt kvt 29 Main challenge preventing long-term unemployment Overload in ALMP? 6 5 Post activation employment effect Larger risk of longterm unemployment months Can the system 1 cope with this?? 6 months 12 months

24 % Strong perceived job security Eurobarometer 21 Perceived likelihood of respondens being able to find a job (after being laid off) LV ITIE LT PTES EE HUMT BG LUCY FR PL EU27 SKSI CZ ROAT DE UKSE FI NLBE DK

25 Conclusion Large change in (un)employment but not relative to output change High gross flows remain despite crisis Most unemployment spells are short Low risk of long-term unemployment Relatively easy entry for youth So far: weak signs of marginalization and increase in long-term unemployment compared to other countries Active labour market policy can it also work in a deep recession? Financial viability of the model crucial to maintain a high employment level!