The recent finding of an increasingly polarized employment structure in the United States has sparked an intense debate whether low-skilled service occupations are also expanding in Western Europe. Our research project wishes to contribute to this debate by empirically analyzing the patterns of job expansion in Britain, Denmark, Germany and Switzerland since 1980. Based on individual-level data, we want to find out whether the employment structure has undergone a process of upgrading, downgrading or polarization: was job growth over the last 25 years concentrated among well paid professions or were low-paid services expanding as well? To answer this question, we bring in the issue of unemployment and examine whether countries with a good employment performance over the last decade (such as Denmark) created jobs in different occupations than countries with lower employment growth (such as Germany). Our research design puts heavy emphasis on the country comparison on the one hand, the comparison between two time periods on the other. We thus examine and then compare the patterns of job expansion over a first (1980-1993) and second period (1993-2006). For our empirical analysis, we select large-scale individual-level surveys for Britain and Germany (the British Labour Force Survey and the German Socio-Economic Panel). For Denmark, we rely on register data (the Integrated Database for Labour Market Research). Finally, for Switzerland, we use data from two different sources (the Federal Population Census and the Swiss Labour Force Survey). The theoretical ambition is to supplement the contributions of labour economy and work sociology with the insight gained from comparative political economy: do different institutions lead to different outcomes in terms of occupational change? For this reason, we discuss and empirically examine four hypotheses developed in parallel by these different disciplines. They include the proposal of skill-biased technological change, the argument of increasing international division of labour, the thesis of a trade-off between low-paid service job creation and unemployment as well as the argument explaining the patterns of job expansion primarily with the skill composition of labour supply.