The maid, the clerk, the doctor & their computer: automatization

This article is written with Eduardo Giménez, Professor of Economics at the University of Vigo. It is the first part out of four, which will be published in this blog. The full article will be published in the book "Automatization" , which will be published by the European Liberal Forum at the end of the year.

Rarely, technical change is neutral. The introduction of new technologies makes obsolete the old ones, monopoly rents associated with the possession of old technologies dissipate, demand for certain occupations enlarges, while demand for other falls, a new elite of millionaires arises, the leadership among nations changes,... Technological change brings allocative and distributive changes that have an impact on all spheres of our lives.

   The rise and development of the information and communications technologies has represented different opportunities for the maid, the clerk and the doctor. The computer has been a complement to the tasks of the doctor. The clerk has been witness that she has been gradually replaced by computers in many of her tasks. The computer has not significantly altered the tasks of the maid. Obviously, their wages and employment opportunities have reacted to these changes. But, in order to understand these changes, let's go bit by bit. 

The race between education and technology

The path of technical change caused by the rise and development of the information and communications technologies has been skill-biased (see Katz and Murphy (1992)). This means that the computers and computer-controlled equipment have significantly increased productivity of the high-skill workers, while their impact on the productivity of the low-skill workers has been less prominent. Consequently, relative demand of high-skill workers increased and, thus, their relative wages rose. Some empirical research reports a statistical correlation between the use of information and communications technologies and either the employment share of skilled workers or their wage share (see Bartel and Lichtenberg (1987) and Autor, Katz and Krueger (1998))

During the eighties, the skill premium –measured as the ratio of the wage earned by a worker with college education to the wage earned by a worker with high-school education- has undergone a rapid increase in most of the developed countries (see Figure 1). The increase in the skill premium happened despite the long-standing uptrend in the relative supply of college workers, which put downward pressure on the skill premium (see Figure 2). Therefore, the rate of skill-biased technical change had to be high enough to offset this downward pressure. Consequently, wage inequality was the outcome of a race between technology and education, using the famous expression of Tinbergen (1974) which gives the title to the book by Goldin and Katz (2008). Moreover, it may well have happened that the increase in the supply of skilled workers has encouraged skill-biased innovation (see Acemoglu (1998)).

However, the rate of increase of the skill premium decreased after the beginning of the nineties. Moreover, after the end of the eighties, as pointed in our first installment, the relationship between changes in the wage distribution and skill is no longer monotonic. Therefore, the relationship between technical change, wages and skills must be more complex than the one established by the explanation based on a path of skill-biased technical change.

Figure 1

Source: Acemoglu and Autor (2011). Data for the United States.

Figure 2

Source. Acemoglu and Autor (2011). Data for the United States.

The race against the machine

The routinization hypothesis (Autor, Levy and Murnane (2003)) holds that computer and computer-controlled equipment have substituted for workers in the routine tasks because computers and computer-controlled equipment are highly productive and reliable at performing the tasks that programmers can script, which are procedural, rule-based activities. This means that the development of information and communications technology has led to a race between machines and workers. The issue is extensively addressed by Brynjolfsson and McAfee in two books (Brynjolfsson and McAfee (2011, 2014)). We have taken the liberty of using title of the former book to head this section. An excellent press article by Autor and Dorn (2013) also addresses this issue.

The routinization hypothesis can be seen as a sophisticated re-elaboration of the skill-biased technical change hypothesis. According to this explanation, technical progress is biased, but not in so simple way as asserted by the skill-biased technical change hypothesis. In particular, technical change is not skill-biased, but task-biased.

Acemoglu and Autor (2011) classify occupations in four categories according to the kind of performed tasks. (1) Non-routine cognitive task-intensive occupations (managerial, professional and technical occupations), (2) routine cognitive task-intensive occupations (sales, clerical and administrative support occupations), (3) routine manual task-intensive occupations (production, craft, repair, and operative occupations), and (4) non-routine manual task-intensive occupations (service occupations between which food preparation and serving, cleaning and janitorial work, grounds cleaning and maintenance, in-person health assistance by home aides, and numerous jobs in security and protective services can be mentioned).

Non-routine cognitive tasks are highly complementary to information and communications technologies, while non-routine manual tasks, even being less complementary to information and communications technologies, cannot be replaced by computers and computer-controlled equipment because they require situational adaptability, visual and language recognition, and in-person interactions. However, routine tasks –both manual and cognitive- can be easily replaced for computers and computer-controlled equipment.

As reported by Acemoglu and Autor (2011), since the late eighties, the U.S. labor market has undergone (1) an increase of employment in non-routine cognitive task-intensive occupations; (2) an increase of employment in non-routine manual task-intensive service occupations; and (3) a decline of employment in middle-skill, routine task-intensive occupations (see Figure 3). Figure 4 and Figure 5 display changes in young-male and young-female employment for different occupations in several European countries.

Figure 3

Source: Acemoglu and Autor (2011). Data for the United States.

Figure 4

Source: Acemoglu and Autor (2011). 

Figure 5

Source: Acemoglu and Autor (2011).

Therefore, the maid, the clerk and the doctor are affected by computers in different ways. The relative demand of the clerks falls because they are replaced by computers in many tasks and her wage falls relatively to the wages of the doctor and the maid. Moreover, the doctor can significantly increase her earnings taking advantage of the information and communications technologies.

According to the routinization hypothesis, there still remains a link between technical change and skill. The reason is that non routine cognitive task-intensive occupations are mostly performed by highly-skill workers, routine task-intensive occupations are mostly performed by middle-skill workers, and non-routine manual task-intensive occupations are mostly performed by low-skill workers. However, differently to the skill-biased technical change, the relationship between technical change and skill is not monotonic, which means that technical change favored the high- and low-skill worker relatively to the middle-skill workers. Moreover, the routinization hypothesis emphasizes the tasks associated to a job as a main determinant of the wages of the workers (see Autor and Handel (2013)).

Michaels, Natraj and Van Reenen (2010) shows that demand for high-skill workers between 1980 and 2004 increased more in the industries which increased relatively more their use of the information and communications technologies, while the demand for middle-skill workers fell rapidly. However, demand for low-skill workers were broadly unaffected by the investment the industries made in information and communications technologies.

In the next installment, we address the question of how globalization is affecting the labor market.

References

Acemoglu, D. (1998). Why do new technologies complement skills? Directed technical change and wage inequality. The Quarterly Journal of Economics, 113(4), 1055-1089.

Acemoglu, D., & Autor, D. H. (2011). Skills, tasks and technologies: Implications for employment and earnings. Handbook of Labor Economics, 4, 1043-1171.

Autor, D. H., & Dorn, D. (2013). How technology wrecks the middle class. The New York Times, 24.

Autor, D. H. & Handel, M. J. (2013). Putting tasks to the test: Human capital, job tasks, and wages. Journal of Labor Economics, 31(2 Part 2), S59-S96.

Autor, D., Katz, L. F., & Krueger, A. B. (1998). Computing inequality: Have computers changed the labor market?. The Quarterly Journal of Economics, 113(4), 1169-1213.

Autor, D., Levy, F., & Murnane, R. J. (2003). The skill content of recent technological change: An empirical exploration. The Quarterly Journal of Economics, 118(4), 1279-1333.

Bartel, A. P., & Lichtenberg, F. R. (1987). The comparative advantage of educated workers in implementing new technology. The Review of Economics and Statistics, 1-11.

Brynjolfsson, E., & McAfee, A. (2011). Race against the machine: How the digital revolution is accelerating innovation, driving productivity, and irreversibly transforming employment and the economy. Digital Frontier Press, Lexington, MA.

Brynjolfsson, E., & McAfee, A. (2014). The second machine age: work, progress, and prosperity in a time of brilliant technologies. WW Norton & Company.

Goldin, C., & Katz, L. (2008). The race between technology and education. Cambridge, MA: Harvard.

Katz, L. F., & Murphy, K. M. (1992). Changes in Relative Wages, 1963–1987: Supply and Demand Factors. The Quarterly Journal of Economics, 107(1), 35-78.

Michaels, G., Natraj, A., & Van Reenen, J.  (2010). Has ICT Polarized Skill Demand? Evidence from eleven Countries over 25 Years. NBER working paper 16138.

Tinbergen, J. (1974). Substitution of graduate by other labor. Kyklos, 27, 217-26.