A natural alternative normalization would set the weights such that k∈K ωikt = 1,
where K is the set of our 20 OECD countries. One could also normalize weights by
maxj ωijt . In a series of robustness checks, we will show that the choice of normalization
has little qualitative effect on our results.
4.2 Data
Bassanini and Duval (2006) have assembled the most comprehensive data set on labor
market variables. It reflects intensive efforts at the OECD to come up with harmonized
measures. Unfortunately, it covers only 20 countries36 for the years 1982 to 2003. How-
ever, it should be mentioned that virtually all cross-country unemployment regressions
in the literature make use of exactly this data set (or its numerous precursors), see for
example Nickell, Nunziata, and Ochel (2005). The key problem with unemployment rates
from a wider spectrum of countries is their lack of comparability across time and space.
Moreover, detailed data on labor market institutions does not exist, except for a cross-
section (Botero, Djankov, La Porta, De Silanes, and Shleifer, 2004). Data on the degree
of product market regulation (PMR), the output gap, and the array of exogenous shocks
also come from Bassanini and Duval (2006).
Data on geographical distance come from CEPII.37 Population data is from the Penn
World Tables mark 6.2. We compute capital stocks using the perpetual inventory method;
see Benhabib and Spiegel (2005). Summary statistics are in the Appendix.
4.3 Conditional cross-country correlation of unemployment rates
As a first step, we show that our data replicates the typical results found in the empirical
literature. Column (1) in Table 1 shows the results of estimating (28) using OLS. The
coefficient on the tax wedge bit , our key labor market variable of interest, implies that a
20 percentage point increase (approximately equal to one standard deviation of b in the
data) increases the equilibrium rate of unemployment by about 1.8 percentage points.
Union density and employment protection legislation (EPL) do not have any measurable
influence on equilibrium unemployment. This is a standard finding; see Bassanini and
Duval (2006) or Baker, Glyn, Howell, and Schmitt (2004). Countries featuring a high
degree of corporatism (such as the Scandinavian countries or Austria) have a rate of
unemployment that is by about 1.7 percentage point lower. Finally, the output gap (gap)
is an important determinant of the unemployment rate. Note that country fixed effects
alone explain about 78 percent of the total variation (adjusted R2) of unemployment
rates in our sample (not shown). Accounting for the common business cycle by including
year fixed effects adds five percentage points of explanatory power; adding country-specific
36Australia (AUS), Austria (AUT), Belgium (BEL), Canada (CAN), Switzerland (CHE), Germany
(DEU), Denmark (DNK), Spain (ESP), Finland (FIN), France (FRA), Great Britain (GBR), Ireland
(IRL), Italy (ITA), Japan (JPN), Netherlands (NLD), Norway (NOR), New Zealand (NZL), Portugal
(PRT), Sweden (SWE), and the United States of America (USA).
37www.cepii.fr/anglaisgraph/bdd/distances.htm.
34