Differential enforcement of employment protection by explicit design of the legislation, for example through exemptions for small firms, has been exploited in a growing body of research. However, little is known about the effects of differential enforcement that is not defined by the letter of the law, presumably due to the lack of data. Our study incorporates aspects of both types of differential enforcement as we combine the evaluation of a partial reform with information on the more difficult-to-observe enforcement of the same reform in collective agreements. We analyse a reform of notice periods for employer-initiated separations in Sweden, which reduced the notice periods for newly hired older workers substantially but implied minor or no changes in the notices for younger workers. The reform was initiated at different times depending on collective agreement. These circumstances provide ample opportunity for the identification of its effects. Our findings indicate heterogeneous effects across collective agreements. Despite differences in terms of dynamics and size, a positive effect on hirings is found for all agreements. In most cases, our results also show an increase in separations, indicating an increase in employment turnover. A salient feature of the results is that the estimated effects increase with the treatment dose, i.e., the size of the reduction in notice periods across different age groups.