Researchers in industrial economics have devoted particular attention to the estimation of markups, i.e. the ratio of price over marginal costs. Given the fact that markups measure the deviation from perfect competition, it has important welfare implications, what makes its measurement extremely relevant for policy makers. Markup dynamics also plays an important role in neo-keynesian macroeconomic models. However, the current methodologies do not allow researchers to compute markups at very disaggregated levels. In this project, we propose to develop a novel methodology adapted to the use of rich micro data that deals with the presence of multi product firms. As such, it will go back to an old literature in production economics that suggested to exploit joint information from the primal and dual Solow residuals. These papers were not always rigorous though in their use of econometric methods and data availability was still scarce. Computing capabilities were also quite limited at the time. Our analysis will exploit the more recent econometric advancement in the estimation of production and cost functions to precisely estimate marginal costs at the firm-product level, and aggregate the micro level estimate at the macro level. As such, it will permit various analyses regarding the determinants of markups as well as an understanding of the drivers of markup dynamics and its macroeconomic implications.