The Measurement of Social and Labour Market Policies - A Natural-Language-Processing-Based Approach

My research will make both empirical contributions to the literature on the consequences as well as determinants of social and labour market policies and theoretical innovations to the methodology of policy measurement. I want to close this empirical gap by researching how such policy measurements should look like and how various disciplines and their data sources can be brought together in order to solve this highly policy-relevant issue. These new measures will capture timing, intensity and direction of social and labour market policy changes more accurately than existing ones, which will enable researchers from various disciplines to use the new dataset in panel, cross-section and time series settings. To make a starting point of such exercises, I want to test the quality of the new measures in standard regressions as dependent and independent variable.
I am aware of the technical challenges of this project, which is why I would start to build the entire measurement pipeline for one country and one policy domain (for instance labour market regulation in the United Kingdom). I believe that my research experience and the interdisciplinary environment at the Department of Social Policy and Intervention will enable me to successfully conduct my intended research project. Because of the multi-disciplinary nature of the PhD thesis, I am happy that Professor Tim Vlandas (Department of Social Policy and Intervention) and Professor Taha Yasseri (Oxford Internet Institute; expert in Natural Language Processing) have expressed their interest to supervise the project.