Longitudinal Data Analysis
In 2014, I published the dissertation "When Employees Leap to Self-Employment", which was based on 4 quantitative studies, 2 of them using advanced methods to study longitudinal data. Research on entrepreneurship entry and related work on labor market transitions has grown rapidly, and I continue my research project studying 22 years of employer-employee matched data from Statistics Sweden (SCB).
Data Mining, Inference and Prediction
As statistical problems have exploded both in size and complexity, “data mining” techniques are necessary to extract important patterns and trends from the data. I do work on how data mining can help us to make better inferences and predictions, especially in the context of big data and the population-based labor market data set by SCB.
Experimental and Quasi-Experimental Design
- Business ideas are important for firms and their employes, but also very uncertain–both when developing one's own idea and when adopting those of others. Examining when people make the right decisions about which business ideas to pursue and which to avoid, and when they make these decisions is important both for understanding the effects of Knightian uncertainty on decision making and for understanding the spread of business ideas. I have done work on both the adoption of business ideas and on their creation using economic laboratory experiments.