In this work we explore how the international outsourcing of production impacts the skill composition of employment within Italian manufacturing firms. In particular, our aim is to assess whether the choice to offshore production activities to cheap‐labour countries implies a bias in the employment of skilled workers relative to unskilled ones.
Using a balanced panel of firms covering the period 1995–2003, we set up a counterfactual analysis in which, by using a difference‐in‐differences propensity score matching estimator, we compare the dynamics of skill demand for treated and control firms while addressing the possible problem of selection bias.
Our results identify a ‘potential’ skill bias effect of production offshoring. In particular, we find that treated firms tend to show an upward shift in the skill ratio with respect to the counterfactual sample, but coefficients are not significantly different from zero. When we look at the elements of the skill ratio separately, we find that the skill bias is driven by a fall in the employment of production workers (blue collars), rather than by the increase in the employment of non‐production workers (white collars), thus providing further evidence on the unskilled labour‐saving nature of international outsourcing. 相似文献
Voting referenda provide direct evidence of the demand for public goods. A number of previous studies have used referenda to analyze the support for public environmental goods. These studies have used aggregate data from large jurisdictional units (usually counties) and summary income measures such as the mean or median, and have usually found that higher income areas offer greater support for environmental propositions. We examine environmental referenda voting in California using census block group data, spatial dependence controls, and detailed income distribution data. We find that household income has a negative marginal effect on environmental referenda voting for most of the income range when using census block data. In addition, controls for spatial dependence significantly reduce the magnitude of most coefficients. This suggests that OLS estimates of referenda determinants are biased. We also show that county level data may be subject to severe aggregation bias and might not be appropriate for referenda studies. 相似文献
Cross-cultural research (CCR) is extremely important for advancing understanding of the range and limits of theoretical constructs and frameworks across different cultural contexts. CCR also provides an important understanding of the differences between countries and cultures. This commentary essay examines the review conducted by Engelen and Brettel (2010) and comments on shortcomings, as well as identifying areas where more needs to be done. A critical first step is to properly define the unit of analysis. This step is particularly critical as the concept of “national culture” is becoming increasingly less relevant and the appropriate culture unit for examination is often a smaller, more homogeneous grouping within a given geographic location. This view allows more meaningful inferences to be made and controls for possible contextual confounds, as well as providing a richer understanding of the limits of national culture in an increasingly global economy. 相似文献