We live in a world today whereby machines are replacing human-labor. Not only are machines driving people out of manual labor, but they are also expanding their knowledge to handle more complex tasks.
[1] Are we heading towards a jobless future? The diminishing number of jobs available in the labor market could imply not only fewer employment opportunities for job-market entrants but also more forced exits of older workers from the workforce.
[2],[3] Could workers’ retirement decisions be dependent not solely on health or financial concerns but also on people’s varying degrees of adaptability to technology and of automatability of their jobs and skill-sets? Can we assume that it is the white-collar jobs are always less automatable by technology than blue-collar jobs? Such are the question that motivated the current study. Data is drawn from the Health and Retirement Study (HRS) used in conjunction with the ONET data to categorize workers’ occupations to assess potential variations in the retirement patterns as follows: (1) white-collar versus blue-collar and (2) highly-automatable versus unautomatable jobs.
[4] Panel fixed-effects models with clustered standard errors at the individual level for a continuous outcome variable indicating the planned retirement year as well as linear probability models for the probability of working past age 62/65 have been conducted. Based on the theory of skill-biased technological growth (SBTG),
[5],[6] I conjecture the following technology-induced changes in retirement decisions: (1) In experiencing a higher wage growths compared to those of the automatable workers, the unautomatable workers can resort to retiring earlier if the income effects predominates over the substitution effects, or (2) the unautomatable workers could postpone retirement to be later than that of the highly-automatable workers if the substitution effects predominate over the income effects, and (3) in experiencing a growing sense of competence associated with the increasing productivity, the unautomatable workers could retire later than the highly-automatable counterparts. Preliminary results show that workers who are unautomatable tend to postpone retirement compared to the highly-automatable counterparts—rending support for the last two hypotheses. Further analyses will be conducted to distinguish between the substitution effect of the income-trajectory change and the psychological effect.
[1] Rotman, David. "How technology is destroying jobs." Technology Review 16, no. 4 (2013): 28-35
[2] Aaronson, Daniel, and Kenneth Housinger. "The impact of technology on displacement and reemployment." Economic Perspectives-Federal Reserve Bank of Chicago 23 (1999): 14-30
[3] Bartel, Ann P., and Nachum Sicherman. "Technological change and retirement decisions of older workers." Journal of Labor Economics 11, no. 1, Part 1 (1993): 162-183
[4] The rationale behind the second way of grouping occupations comes from the theory of skill-biased technology growth (SBTG) (Card & DiNardo 2002; Goldin 1998)—which is to be discussed in further detail in the theory section.
[5] Card, David, and John E. DiNardo. "Skill-biased technological change and rising wage inequality: some problems and puzzles." Journal of Labor Economics 20, no. 4 (2002): 733-783
[6] Goldin, C., and L. F. Katz. 1998. “The Origins of Technology-Skill Complementarity.” The Quarterly Journal of Economics 113 (3) (August 1): 693–732.
