Qingyan Peng: Technology, Nonstandard Jobs, Occupations, and Firm Performance

Abstract

Essay I: This study is the first to examine the relationship between ICT adoption and nonstandard employment using country-industry-level data on ICT adoption combined with individual-level contract type data for 15 European countries. Using a multi fixed-effect model, I find that workers in industries at the 90th percentile of ICT capital per worker are about 3 percentage points less likely to hold temporary contracts and 5 percentage points more likely to hold part-time contracts compared to those in the 10th percentile. Both effects are driven by changes in non-routine occupations. The decreased incidence of temporary employment is mainly attributed to changes among involuntary temporary workers, while the increased incidence of part-time employment is most pronounced among those working part time for family, personal reasons, or care responsibilities. Overall, ICT adoption appears to improve contractual conditions by reducing the incidence of involuntary temporary employment and expanding opportunities for voluntary part-time work.

Essay II: This paper examines the impact of technology on temporary and part-time jobs using the case of surgical robots. By exploiting the staggered adoption of surgical robots across Swedish hospitals, I study the causal effects on nonstandard nursing jobs through a staggered difference-in-differences strategy. Linking hospital surgery data with job advertisements, I find that robot adoption reduces part-time job incidence among operating room nurses, which may be due to hospitals’ investment in training programs for this advanced robotic technology. Conversely, the incidence of temporary jobs increases post-adoption, which may be attributed to a screening mechanism: as the new skills required by robotic surgery are difficult to assess beforehand, hospitals use temporary contracts to evaluate applicants' suitability.

Essay III: (with Oskar Nordström Skans): We analyze the firm-level consequences of automation when firms have heterogeneous factor shares. Innovations that lower the cost of replacing specific tasks within firms should benefit firms with high employment shares in the associated occupations. Conversely, falling costs of machine services that complement other types of labor at the firm level should benefit firms with lower factor shares in declining occupations. While these two scenarios may have similar impacts on aggregate labor demand, they predict opposite outcomes at the firm level. Using Swedish register data, we correlate firms' initial occupational employment shares with subsequent growth in firm sales. We find a clear positive association between initial labor intensity in declining occupations and future market share growth, suggesting that tasks were primarily replaced through substitution within firms.