«Der Open-Access-Publikationsserver der ZBW – Leibniz-Informationszentrum Wirtschaft The Open Access Publication Server of the ZBW – Leibniz ...»
A.1.1 Income elasticity of labour supply Originating from the homothetic CES function (2), the demand functions are homogeneous of degree one in disposable extended income. We thus have26
From this we can derive the income elasticity of labour supply. To be precise, we add a small amount of non-labour income, Y0 to the extended income in (3)
A.1.3 Elasticity of participation The distribution of the U0 ’s over the population must be calibrated. We have the actual participation rate and the elasticity of labour supply at the extensive margin as our empirical basis. This is suﬃcient to calibrate the distribution of the ﬁxed costs locally (at the point of actual participation), but not globally. The rest of the distribution must be ﬁxed by some functional assumption. We assume that − + ﬁxed costs are uniformly distributed between U0 and U0. For ﬁxing the values of these bounds, we ﬁrst have to calculate the change in Ul produced by an exogenous variation in the wage. We consider the case of an isolated change in the wage of the respective individual in the case of employment. In this case, the unemployment rate and the utility in case of unemployment may be considered constant. This would not be the case for a general change in the wage, which applies to all individuals. In This is also what you have in Rutherford (1998), if you leave out the upper nest with the consumption-savings decision (assuming that the savings ratio is zero).
I follow Sørensen (1999) and assume a value of 0.1 for ηLw. The meta study of Evers et al.
(2005) suggests a somewhat higher elasticity, but it is diﬃcult to distil a “core” value from this study.
terms of elasticities, we then have
This is evaluated at the initial point, with ηN w set to 0.2, following Kleven and Kreiner (2006b).29 h is then treated as a constant in the counterfactual simulations.
This means that the elasticity at the extensive margin is precisely reproduced only for the initial point; oﬀ the initial situation, it is endogenous.
which includes both the utilitarian union (µ = 0.5) and the insider model of (8) with µ = 1 (see Graaﬂand et al. 2001, ch. 7). This, however, would leave us with two parameters, µ and λ, which cannot be calibrated in a single ﬁrst-order condition without further information.
A.2 OECD data sources
The entries in Table 1 have been generated in the following way:
• “sL ”: share of labour in value added. From OECD “Annual National Accounts
of OECD countries, Vol. 2”, Issue 2005, Table 2: Gross domestic product:
income approach. “1. Compensation of employees” / (“1. Compensation of employees” + “31. Gross operating surplus and gross mixed income” • “tC ”, “ta ”, “tπ ”: eﬀective average tax rates on consumption, labour and capital L income. From OECD “Annual National Accounts of OECD countries, Vol. 2”, Issue 2005, and OECD “Revenue Statistics”, Issue 2004. Calculated as proposed in Mendoza et al. (1994) and further developed by Gurgel et al. (2007). In order to better ﬁt the tax bases identiﬁed in the model of this paper, I have used the gross instead of the net capital income as basis for the capital (proﬁt) tax. This gives substantially lower capital tax rates than those reported in the papers cited.
• “u”: standardised unemployment rate. From OECD “Labour Force Statistics”, Issue 2005, entry “(ALFS) Total labour force, All persons, Unemployment, % total labor force”.
• “c”: replacement rate. From OECD “Beneﬁts and Wages”, Issue 2004, Table
3.3a. (p. 102) “Average of Net Replacement Rates over 60 months of unemployment 2001, for four family types and two earnings levels, in per cent”, entry “without social assistance, no children, single person”.