Wirtschaft und Gesellschaft 42. Jahrgang (2016), Heft 4 quently a change in the aggregate wage share can results from changes in the sectoral composition, referred to as the between component, or changes in the sectoral wage shares, referred to as the within component as distinguished by the first and second product in equation (2): n ⎛ VAi, t ⎞ LCi, t ⎛ LCi, t ⎞ VAi, t Eq. (2) + ∆⎜ WSC ⎟∗ ⎟∗ t = ∑ ∆⎜ C ⎝ VAi, t ⎠ VAC VA VA i, t i =1 ⎝ t ⎠ t 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 Jayadev (2007); Stockhammer (2015). Onaran (2011, 2012). Stockhammer (2015). Kristal (2012); Argitis and Pitelis (2001). Choi (2001). Nunziata (2005). Although some economists argued that stronger unions can lead to higher unemployment there is very little econometric evidence for this hypothesis (OECD [2006]; Jaumotte and Buitron [2015]). OECD (2006). E.g. Harrison (2002); Jayadev (2007); Onaran (2009); Stockhammer (2015). Stockhammer (2015). Jayadev (2007). Lazonick and O’Sullivan (2000). Langley (2007). Stockhammer (2009, 2015). Jayadev (2007). Hein and Schoder (2011); Dünhaupt (2013). The use of an international database is instructional for making the variables and estimations comparable between countries. See Guschanski and Onaran (2016a, 2016b) for further information on sector definitions and the skill taxonomy. It would be preferable to use value added at factor cost for the calculation of the wage share. Unfortunately, there are no long series on taxes minus subsidies on production in EU KLEMS. Since self-employed are not included in the measure of labour compensation in OECD STAN we impute their wages by applying the same technique as in EU KLEMS. We exclude observations where the number of self-employed suddenly falls to zero, assuming that it must be related to a measurement error. We refer to our data as “at the 2-digit level” if we use manufacturing sectors at 2-digits. Most service sectors are always used at the 1-digit level. Unfortunately, data for most countries includes re-export and re-imports as most countries do not report these series separately. Given the asset/liability principle of the measure negative FDI positions can result “when the loans from the affiliate to its parent exceed the loans and equity capital given by the parent to the affiliate” (OECD, 2016). Since data for foreign labour and population by nationality is not available for the US we use foreign labour and population differentiated by country of birth for the US only. The variable is adjusted for the possibility that some sectors or occupations are excluded from the right to bargain (removing such groups from the employment count before dividing the number of covered employees by the total number of dependent workers in employment). Visser (2015). Solt (2014). Alvaredo et al. (2015). 584