The paper considers time series evidence on the relationships, and possible trade-offs, between productivity and employment, and on the impact of taxes in this connection. First, a theoretical model is built for an open economy leading to the identification of technology, non-technology and tax shocks. Then structural VAR models are estimated for all the EU-15 and some other OECD countries to infer the above links. Our conclusion is that there is in the EU a fairly uniform and significant short-run negative impulse on employment from a positive productivity shock, while this becomes smaller and statistically insignificant over time in most, but not in some member countries. The former situation is interpreted to be an indication of nominal and the latter that of real or structural rigidity in the economy. In the US, there is no such trade-off, either in the short or long run. Tax shocks are found to have mostly a short-run negative effect which is stronger on productivity than on aggregate employment. However, if we separate the effects of labour taxes and corporate taxes, the former have in the EU-15 a strong negative effect on employment while the latter are fairly neutral. Second, we simulate an aggregative econometric labour market model and insert various types of shocks into it : a rise in productivity, achieved, e.g. by enhancing R&D, or by rationalising the use of labour, and a change in the tax/benefit system. We find that although there is no long-run trade-off between productivity and employment, over the medium run acceleration of productivity has a clear positive effect on employment.