Institutional reforms that really matter: OECD institutional indicators vs. Dutch reform history.

1
Institutional reforms that really matter OECD
institutional indicators vs. Dutch reform history.

• Ruud Gerards, Manuel Müllers and Joan Muysken
• CofFEE-Europe
• Maastricht University, Department of Economics

Presented by Ruud Gerards at AIAS 6 November
2008 http//www.ruudgerards.nl
2
Outline

• Methodology
• Dutch reform history
• Econometric evidence
• Comparison with OECD indicators
• Conclusions
• Further research/refinement of research

3
Introduction

• Objectives
• Which reforms have had considerable impact?
• Do OECD institutional indicators correctly pick
  up these reforms?

4
Methodology

• Our methodology builds on Stegeman 2005
  (Netherlands bureau for economic policy analysis,
  CPB)
• In 1980-2003 there were too many reforms to test
  econometrically
• Qualitative analysis (literature research) gives
  us pre-selection of reforms
• Econometrics used to test the selection of
  reforms

5
Methodology

• Qualitative analysis
• Fondazione Rodolfo Debenedetti Social Reforms
  Database
• Brandt et al
• LABREF Labor Market Reforms Database
• OECD economics surveys
• Literature review
• Ranking

6
Dutch reform history
7
Dutch reform history Areas and instruments

• Reform areas
• Disability
• Sickness
• Unemployment
• ALMP
• EPL
• Reform instruments
• Level
• Responsibility
• Access
• Duration
• Stimuli

8
Dutch reform history Instruments
9
Dutch reform history Impacts
Figure 3. Reform impact ratios by area and
instrument
10
Econometric evidence Model

• Two stage error correction model
• Estimation period 1980-2003
• Reforms are represented by dummies
• First stage (long run) equation
• Second stage (short run) equation

11
Econometric evidence Dummies

• Reforms represented by Dummies
• Effects of reforms implemented in subsequent
  years difficult to measure
• Hard to distinguish between the impacts of
  dummies that are close to each other in time
• To reduce density of dummies some dummies
  represent multiple reforms

12
Econometric evidence Method
13
Econometric evidence Conclusions

• Econometric analysis confirms results from
  qualitative analysis
• Strong empirical evidence for the role of the
  business cycle
• Unemployment countercyclical
• Sickness and disability pro-cyclical

14
Do OECD indicators correctly reflect our findings?
15
Do OECD indicators correctly reflect our findings?
16
Do OECD indicators correctly reflect our findings?
17
Do OECD indicators correctly reflect our findings?

• Disability and sickness hardly discussed in the
  OECD literature
• No OECD indicators for disability and sickness
• Indicators on EPL and unemployment benefit
  replacement rates perform not so well
• ALMP indicator performs reasonably well

18
Conclusions

• Policy conclusions
• EPL and ALMP policy relatively unsuccessful
• Sickness, disability and unemployment reforms
  relatively successful
• Political business cycle influences reform timing
• OECD indicator conclusions
• OECD indicators do not perform that well, except
  ALMP indicator
• Sickness and disability not covered by OECD
  indicators

19
Further research/refinement of research

• We did not take into account tax-based reforms
• Regress all reform efforts at total number of
  claimants
• There is a certain amount of subjectivity in the
  ranking of reforms in the qualitative analysis
• Ideas for improvement??
• Maybe count and analyse newspaper citations on
  these reforms?
• Possible selection bias?
• Yes maybe, but econometrics used only to confirm
  results of qualitative analysis and this it does.
• Covariates would have become significant instead
  of the reforms
• Repeat this analysis for more countries