New Evidence on the Effect of Technology on Employment and Skill Demand
(with Johannes Hirvonen and Aapo Stenhammar)
under revision for the Quarterly Journal of Economics
Abstract: We present novel evidence on the effects of advanced technologies on employment, skill demand, and firm performance. The main finding is that advanced technologies led to increases in employment and no change in skill composition. Our main research design focuses on a technology subsidy program in Finland that induced sharp increases in technology investment in manufacturing firms. Our data directly measure multiple technologies and skills and track firms and workers over time. We demonstrate novel text analysis and machine learning methods to perform matching and to measure specific technological changes. To understand our findings, we outline a theoretical framework that contrasts two types of technological change: process versus product. We document that the firms used new technologies to produce new types of output rather than replace workers with technologies within the same type of production. The results contrast with the ideas that technologies necessarily replace workers or are skill biased.
The Economist: "Economists are revising their views on robots and jobs."
The Economist: "Leader: The world should welcome the rise of the robots."
Wired: "Automation Isn't the Biggest Threat to US Factory Jobs"
Psychological Traits and Adaptation in the Labor Market
(with Ramin Izadi)
Abstract: Labor markets are in constant change. Which personality traits and skills help workers to deal with a changing environment? This paper documents how responses to labor-market shocks vary by individuals’ psychological traits. We construct measures of cognitive ability, extraversion, and conscientiousness using standardized personality and cognitive tests administered during military service to 79% of Finnish men born 1962–1979. We analyze establishment closures and mass layoffs between 1995–2010 and document heterogeneous responses to the shock. Extraversion is the strongest predictor of adaptation: the negative effect of a mass layoff on earnings is 20% smaller for those with one standard deviation higher scores of extraversion. Conscientiousness appears to have no differential impact conditional on other traits. Cognitive ability and education predict a significantly smaller initial drop in earnings but have no long-term advantage. Our findings appear to be driven directly by smaller dis-employment effects: extraverted and high cognitive-ability individuals find re-employment faster in a similar occupation and industry they worked in before. Extraversion’s adaptive value is robust to controlling for pre-shock education, occupation, and industry, which rules out selection into different careers as the driving mechanism. Extraverts are slightly more likely to retain employment in their current establishment during a mass layoff event, but the retention effect is not large enough to explain the smaller earnings drop.
Featured in Marginal Revolution.
Abstract: This paper analyzes the impact of manufacturing decline on children. To do so, it considers local employment structure—characterizing lost manufacturing jobs and left-behind places—high-school dropout rates, and college access in the US over 1990–2010. To establish a basis for causal inference, the paper uses variations in trade exposure from China, following its entry to the WTO, as an instrument for manufacturing decline in the US. While the literature on job loss has emphasized negative effects on children, the main conclusion of this research is that the rapid US manufacturing decline decreased high-school dropout rates and possibly increased college access. The magnitudes of the estimates suggest that for every 3-percentage-point decline in manufacturing as a share of total employment, the high-school dropout rate declined by 1 percentage point. The effects are largest in the areas with high racial and socioeconomic segregation and in those with larger African American populations. The results are consistent with the idea that the manufacturing decline increased returns and decreased opportunity costs of education, and with sociological accounts linking working-class environment and children’s education.
School vs. Action-Oriented Personalities in the Labor Market
(with Ramin Izadi)
Abstract: How do different dimensions of personality predict school vs. labor-market performance? How has the value of these traits changed over time? We answer these questions using data that includes multidimensional personality and cognitive test scores from mandatory military conscription for approximately 80% of Finnish men. We document that some dimensions of noncognitive skills are productive at school, and some dimensions are counterproductive at school but still valued in the labor market. Action-oriented traits predict low school performance but high labor market performance. School-oriented traits, such as dutifulness, deliberation, and achievement striving, predict high school performance but are not independently valued in the labor market after controlling for school achievement. We further document that the labor-market premium to action-oriented personality traits has rapidly increased over the past two decades. To interpret the empirical results, we outline a model of multidimensional skill specialization. The model and evidence highlight two paths to labor-market success: one through school-oriented traits and formal skills, and one through action-oriented traits and informal skills.
Featured in Marginal Revolution.
Research in Progress
Scarcity vs. Surplus: New Evidence on Labor Supply and Industrialization
(with Jonas Mueller-Gastell)
Abstract: Does shortage of labor or abundance of labor encourage technology adoption? Are machines and men substitutes so that labor scarcity induces investment in technology, or are they complements so that availability of workers facilitates technology adoption? The project uses local labor supply shocks in Finland at the verge of industrialization to study how technology and labor supply interact. These shocks come from two sources: combat deaths and evacuations from invaded areas into designated towns during the Second World War, 1939–45. The project uses newly digitized local and plant-level data on technology use by type, employment, and organization. We find a positive effect of labor abundance on manufacturing development. Evidence on horsepower per person shows that additional labor does not crowd out capital but complements capital investment. Manufacturing employment share and gross value-added per person increase substantially across all identification strategies, including strategies based on military and evacuation plans.