Using very detailed register data on cognitive abilities and productive personality traits for nearly all Swedish males at age 18, we show that employment in the recent past has shifted towards skill-intensive occupations. Employment growth is monotonically skill biased in relation to this set of general-purpose transferable skills, despite the well-known U-shaped (”polarizing”) relationship to occupational wage ranks. The patterns coexist because growing low-wage occupations tend to employ workers who are comparably skilled in these dimensions, whereas workers in declining mid-wage occupations instead have less of these general non-manual skills than suggested by their wages. Employment has primarily increased in occupations where workers have larger-than-average endowments of verbal and technical abilities and social maturity. Projections of future occupational decline and automation risks are even more skill-biased, but show similar associations to most of our specifc skill-measures. The most pronounced difference is that occupations relying on tolerance to stress are projected to decline in the coming decades.
In this paper, we present a new way of characterizing the skill requirements within occupations and relate these requirements to occupation-specifc employment trends, in the past and projected future. Overall, the results provides, what we believe to be, an important addition to the stock of knowledge regarding the relationship between occupational employment shifts, wage polarization, and skill demand.
Our results show that occupational employment shifts are skill-biased towards a composite measure of general-purpose transferable intellectual skills, despite the non-linear relationship to wages. The reason is that growing low-wage occupations are more intensive in these skills than their wage ranks suggest and the converse is true for the declining mid-wage occupations. Focusing on the lower half of the wage distribution, the results thus suggest that labor demand has moved away from average-paying jobs with a low need for general intellectual skills towards low-paying jobs with a high need for such skills. This process may explain why workers in declining occupations in the middle of the wage distribution appear to suffer from long-term adverse employment effects as the transition into low-wage jobs may demand more in terms of general skills than these workers possess, despite of the fact that their pre-displacement jobs were relatively well paid.
The difference between the skill-rank and wage-rank results arises because our skill measures are broad in the intellectual dimension but leave out a set of residual unobserved wage- related attributes such as, e.g., manual strength, the ability to cope with hazardous work environments, pure rent seeking abilities and knowledge that is specifc enough to not be captured by any of our general skill measures. For natural reasons, we need to leave an exploration of the relative importance of these unobserved earnings-related factors aside for the purpose of this article; our results strongly suggest that future research on the granularity of these residual components and their role in the decline of middling-wage occupations is of first-order importance.
Our second key insight is that the underlying patterns are far from uniform across skill types, even within the broader cognitive vs. non-cognitive aggregates that have been emphasized in the related literature on the changing worker-level returns to skills. In particular, we note that growing occupations are relatively dense in Verbal comprehension and Social Maturity (i.e. extroversion), both of which are related to human communication (and thus perhaps could be labeled ”soft”). Occupations that are dense in technical abilities have also grown. In contrast, we see and a reduction in employment within occupations where workers are relatively well-endowed in terms of the ability to focus (measured as Psychological Energy) and Inductive reasoning, i.e. problem-solving skills. Notably, and somewhat on the positive side from a policy perspective, both of the cognitive abilities that have seen an increased demand are in the set of (crystallized) abilities that previous research have identifed as being more malleable since they measure the ability to utilize acquired knowledge and skills.
We further show that existing projections, drawn from two very different sources, suggest that the patterns of the recent past may be reasonably representative of the near future. This is true even though, as is well known, the same projections suggest that future technology will affect a very different set of occupations. The relative growth of occupations that (currently) relies on more skilled workers is projected to continue, perhaps even more distinctly than in the past. The main consistent projected change is a decline in occupations that employ workers with higher than average tolerance to stress and a projected growth in jobs that employ workers with the ability to activate without external pressure. However, since three out of four of the attributes that defined winners and losers in the recent past will continue to do so in the projected future, the overall impression is that the same types of workers that gained in the recent past will be the winners in the near future. On the positive side, this suggest that policy makers striving to design educational systems to favor the acquisition of skills that are useful at the future labor market may draw guidance from the evolution in the recent past.
hosen excerpts by Job Market Monitor. Read the whole story @ The skill-specific impact of past and projected occupational decline – IFAU