| ||||||||||||||||||||||||||||||||||||||||||||||||||||
This article appears in the following Oxford Review of Economic Policy issue: CAPITALISM AND INEQUALITY [View the issue table of contents]
Unemployment, institutions, and reform complementarities: re-assessing the aggregate evidence for OECD countries
* OECD and ERMES, University of Paris 2, e-mail: andrea.bassanini{at}oecd.org
** OECD, e-mail: romain.duval{at}oecd.org
 |
Abstract |
|---|
There is no or limited consensus on the quantitative impact of institutions on unemployment, which has led some to question the case for structural reforms. Recent studies suggest also that institutions interact with each other and cannot be analysed in isolation. In this paper, we estimate a standard reduced-form model to explore the institutional determinants of unemployment and assess its robustness using a large battery of robustness checks. We show that, although the impact of each individual policy varies across countries owing to policy interactions, the simple linear model can be used to draw inferences for countries with an average mix of institutions. The model is then extended to encompass systemic interactions, in which individual policies interact with the overall institutional framework. We find relatively robust evidence of broad reform complementarities.
Key Words: institutions • aggregate unemployment • reform complementarities
Comments from Sven Blöndal, Wendy Carlin, Dave Coe, Jean-Philippe Cotis, Martine Durand, Jorgen Elmeskov, David Howell, Etienne Lehmann, Edmond Malinvaud, John Martin, Steve Nickell, Giuseppe Nicoletti, Stefano Scarpetta, Paul Swaim, Raymond Torres, and audiences in Paris, Toronto, Amsterdam, and Oxford are gratefully acknowledged. Catherine Chapuis, Sebastien Martin, and Rebecca Oyomopito provided excellent research assistance. Any errors are the responsibilities of the authors alone. The views expressed in this paper are those of the authors, and do not necessarily reflect those of the OECD or of its member countries.
1 Countries are Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, Norway, New Zealand, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States. In contrast with some of the papers in the literature, we prefer to estimate a static rather than a dynamic specification. The reason is that it might be very difficult to specify the correct error structure of a model with many institutional variables (such as regulation and bargaining variables) that are subject to large, infrequent changes, whose effect might take an unknown number of years to materialize thoroughly. It is, in fact, straightforward to show that the error term in this case will hardly follow an AR process. Also, within a static framework, serial correlation in the residual can be seen essentially as a problem of efficiency and not of consistency of the estimates. Nevertheless, results from some dynamic specifications are discussed in Appendix 1, when we deal with possible reverse causality issues.
2 In a companion paper (Bassanini and Duval, 2006a), we explore the effect of other policies often considered in aggregate analyses, such as active programmes and minimum wages. However, owing to smaller samples and endogeneity issues, their analysis requires specific treatment, and they are therefore excluded from the analysis in the present paper. What matters is that controlling for these variables does not affect the estimated effects of the other institutions studied here. The companion paper also considers a number of specifications in which the output gap is replaced by a set of macroeconomic variables that capture more directly the unemployment impact of aggregate shocks and are less subject to potential endogeneity concerns. These specifications are not reported here, as they have no impact on the estimated coefficients of institutional variables.
3 Available at http://www.oecd.org/dataoecd/25/25/37431112.zip. Descriptive statistics and details on variable construction and sources are reported in Bassanini and Duval (2006b).
4 In principle, the baseline specification includes dummy variables for both high and intermediate corporatism. However, given that no country moved in or out of the intermediate level of corporatism over the sample period, the effect of this variable cannot be identified—even if controlled for—in most of the specifications. For this reason, it does not appear in Table 1.
5 Indeed, once separate indicators for duration and levels are included, both appear to be significant (see Bassanini and Duval, 2006a).
6 A negative sign implies that the detrimental effect of each policy indicator on unemployment is smaller the higher the other policy indicator, so that reforms diminishing the levels of these institutions should be undertaken together to maximize their impact. More formally, in equation (2) the partial derivative of unemployment with respect to the institutional indicator 
is: 
. If 
is negative, the marginal unemployment effect of institution 
will be larger, the lower the value of
, i.e. the more employment-friendly is the other institution, 
. In other words, the lower 
, the greater the potential employment gain from reforms reducing the level of Xk.
7 Following the ‘rule of thumb’ of Staiger and Stock (1997), the instrument is considered to be acceptable when the F test on the significance of the instrument is greater than 10.
8 More formally, equation (3) below is based upon the assumption that that: (i) labour demand is close to being iso-elastic; and (ii) policy reforms are such that they do not excessively modify the slope of the WS curve but rather entail a parallel shift of it.
9 This implies that the specification actually estimated is slightly more complex than (3), and is:
|
|
|
|
is a country dummy variable—which takes value 1 in country h and 0 otherwise—and 
is a parameter to be estimated. This approach mirrors the one considered in column 3 of Table 3. IV approaches such as those implemented in column 2 of Table 3 have not been attempted here for computational problems associated with the maximization of the joint likelihood function. It might also be argued that country fixed effects contribute to the determination of structural unemployment and should, therefore, be added to the sum of direct effects in the interaction term. Yet, this route is not followed here, both for parsimony and because of lack of convergence of the related algorithm. However, specifications where fixed effects are added to the sum of direct effects in the interaction term, while the term 
is dropped, yield qualitatively similar—albeit somewhat less significant—results. Likewise, time dummies are not included in the estimated equation. 10 As a robustness check, the specification of column 2 was re-estimated, excluding the high corporatism dummy variable in the sum of direct effects of institutions that is included in the interaction. This exercise aims at checking that the results do not hinge on the statistical treatment of corporatism which, as a dummy variable with little variation over time, has a somewhat particular status in the regressions. The results obtained are similar to those in column 2 of Table 4, except that the direct impact of high corporatism becomes insignificant.
11 As the gain is larger, the larger the extent of the reforms, we simulate the complementarity effect for large reforms in historical perspective, in order to show that its magnitude is small.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. R. Howell and M. Rehm Unemployment compensation and high European unemployment: a reassessment with new benefit indicators Oxf. Rev. Econ. Policy, March 1, 2009; 25(1): 60 - 93. [Abstract] [Full Text] [PDF]
|
|