It is well known that technological change is not only one of the key drivers of overall economic growth, but also has implications for inequality.
Developed economies have seen a large reallocation from goods to the service sector. In the US for example, while the goods sector accounted for about 44% of total hours worked in 1960, by 2010 this was down to just 21%. The economic literature on structural transformation typically explains these changes by pointing to differences in productivity growth across sectors. As goods and services tend to be complements in consumption and labor productivity grows faster in goods than in other sectors, supply outgrows demand for goods, leading to a reduction of employment in the goods sector and a rise in service employment.
A different strand of literature has documented the polarization of occupational employment. The demand for labor in traditionally middle-earning occupations has fallen, resulting in a reduction in their employment share and their wages compared to low and high-earning jobs. Again the main explanation put forward in the literature is one based on biased technological change, in this case at the level of tasks. The argument is that the advent of computers and information and communication technologies adversely affected workers in occupations more intensive in routine tasks (routine occupations), which are in the middle of the wage distribution, and complemented workers performing abstract tasks, at the top of the wage distribution.
A closer look at the data shows that the contraction in goods sector employment happened exclusively through a reduction of routine employment, and vice versa almost all of the decline in routine employment occurred in the goods sector. These patterns not only highlight that there is a close connection between the sectoral and occupational employment reallocations, but also are the reason for why it is difficult to identify the true nature of technological change. In this paper we take a new approach to disentangle to what degree technological change is truly biased across sectors and to what degree across occupations.
In our analysis we focus on the production side of the economy without imposing a priori restrictions on how technological change occurs. Specifying a flexible model that is consistent with structural change and labor market polarization, we start from the assumption that a job’s productivity is specific both to the occupation and to the sector of the worker. Making some assumptions about the interactions between different occupations (i.e. the sectoral production function), we use US data over 1960 to 2010 on income shares of various occupations within a sector, as well as relative prices of sectoral outputs and overall growth in GDP per worker, to infer how productivities have evolved over time for each sector-occupation cell.
We then look for common factors in technological change to gauge the extent to which productivity growth is specific to certain sectors (after controlling for the role of occupations) or to certain occupations (controlling for sectors). Our results show that most of productivity changes are not neutral, but biased across occupations (around 70%) and to a lesser extent across sectors (around 5%), and that a significant part of technology is specific to the sector-occupation cell (around 25%). Moreover, our model highlights that by far most of the observed employment reallocations are due to occupational differences in technological change.
We view our analysis as a first step towards evaluating policies in the rapidly changing labor markets. Recently much of the political debate has focused on active labor market policies (such as training programs), and on protectionist policies aiming at maintaining certain industries of the home economy. While in our model there are no frictions or externalities which would justify these policies, our results shed light on the technology side of the economy and attribute a large role to occupation-specific changes. An implication of these findings is that policies targeting workers’ occupational choice might be better at improving labor market outcomes than industrial policies.
You can download the complete paper here.