We adapt the multidimensional poverty methodology to study job quality dynamics using a unique household survey panel for Chile. We use information on wages, type of contract, training and em- ployment duration to build an aggregate job quality index. Panel data allow us to properly separate individual heterogeneity and true dependence. We estimate a dynamic panel with random effects finding higher job quality among larger and unionised firms. Moreover, labor history predicts job quality con- firming the existence of persistence in job quality.