There are widespread concerns in many developed economies about the potentially damaging effects of new technologies on employment. But what if the risk in developing economies is that technology has too little impact on the job market? This column argues that in the countries of the Global South, public policy should be focused more on ensuring that technological change does penetrate the economy, generating increases in productivity.
A substantial body of recent economic research has explored the effects of technological change on the labor market. These studies not only highlight the possibility that new technologies could displace a large share of current occupations, but also the risks in terms of their distributive impact. In particular, it is suggested that the automation of certain tasks could modify the structure of employment, generating ‘labor polarization‘.
At one extreme, the employment of highly qualified workers in occupations that are intensive in non-routine cognitive tasks, with high productivity and high incomes, would tend to strengthen. At the other extreme, there would still be demand for low-skilled workers in occupations intensive in non-routine manual tasks, with low productivity and low incomes. But skilled workers at middle-income levels, generally working on routine tasks (manual and cognitive) that can be replaced by technology, would face lower demand.
A limitation of this research from the perspective of policy-makers in Latin America and other regions in the Global South is that most of it has been carried out in more developed countries. This is important given that emerging countries have different economic structures from developed countries, which could lead to different trends in technological progress.
A fundamental difference is that the process of technological acceleration linked to automation, digitization, and penetration of artificial intelligence will probably not affect Latin America with the same intensity or speed as in developed or emerging Asian countries. For example, there is evidence that the break from routine to non-routine jobs in low- and middle-income countries between 2000 and 2017 would have been significantly more modest than that seen in developed countries.
This would not be a new experience. In previous times of technological acceleration at the global level, Latin America has shown much less dynamism than the regions that led the processes. This, in turn, leads to significantly more modest increases in per capita output and a lag in productivity.
For example, in the second half of the 20th century, the US and Western European economies experienced a significant acceleration in the growth of per capita output, leading to a significant divergence with Latin American growth dynamics (see Figure 1).
More recently, over the last 40 years, there has been a strong acceleration of the emerging economies of East Asia, led by China, while the more developed economies have had a moderate slowdown and Latin America has continued with no significant break in its growth dynamics. Figure 2 shows the contrasting paths of convergence of East Asia and Latin America towards the developed economies.
Figure 1. GDP per capita by region 1900 to 2018
Source: Maddison Project Database 2020.
Figure 2. GDP per capita by region 1980–2025
(as a percentage of developed economies’ GDP per capita, at 2017 PPP prices)
Sources: WEO and IMF.
This suggests that technological change typically reaches Latin American countries with much less intensity. It also implies that the challenges associated with technological unemployment and labor polarization identified by research may be less relevant than in developed countries.
Indeed, the latest episode of technological acceleration may be another one in which Latin America is unable to capitalize to accelerate its growth and reduce productivity gaps with more developed economies. A central concern for policy-makers in the region should therefore be related to avoiding the productivity failures associated with previous technological revolutions.
The incorporation of new technologies that make it possible to avoid a growing productivity gap with the more developed countries will not happen naturally. There must be the right conditions and the appropriate economic incentives.
In terms of conditions for the incorporation of new technologies, the region is clearly at a disadvantage in terms of infrastructure, human capital, and investment in research and development (R&D). Automation and digitization require digital infrastructure, but also adequate human capital to be able to work with new technologies. Problems of educational coverage and quality can be significant limitations on capitalizing on the ‘fourth industrial revolution’.
In turn, the lower degree of local research determines that much of the time, Latin America seeks to adapt technological innovations based on motivations or contexts from elsewhere. For example, labor-saving technological developments are clearly consistent with the challenge of economies that are in advanced processes of demographic transition and which expect reductions in the size of their workforces over the coming decades.
But the objective of saving labor is less relevant in Latin America or other parts of the Global South where, despite certain heterogeneities between countries, significant growth of the working-age population is projected at least until the middle of the century. This mismatch between the objectives of technological innovations and the regional context translates into incentives less aligned with the incorporation of technology.
First, decisions to replace human labor with capital are strictly related to the relative price of both factors. Wages in Latin America are lower than in more developed economies, which could be delaying substitution decisions.
These prices are affected by structural features such as relative differences in factor endowments. As mentioned above, unlike more developed countries, a significant increase in the labor force is expected in Latin American countries, which may limit the relative increase in the cost of this factor.
In this sense, the main public policy challenges in the region should focus on reducing infrastructure gaps, continuing to increase educational coverage and quality, and aiming for greater investment in R&D, innovation, and development.
Second, in terms of labor institutions, it is essential to avoid a narrative of technological anxiety that leads to a pro-deregulation agenda. The goal of preventing the automation of specific jobs in order to preserve employment can lead to proposals for deregulation and depression of wages or employment benefits, which in turn can effectively reduce the incentives to automate.
It is crucial to avoid high costs in a transition that involves significant changes in the profile of workers’ tasks. Still, the response to this new technological impulse can in no way try to reduce incentives for the adoption of new technologies.
The downward trend in the prices of different technological options will determine that at some point, the technologies will end up being incorporated into production. The problem is that this could happen when they cease to be cutting-edge technologies and the developed world is already in a fifth technological revolution.
In that case, the distributional and job reconversion challenges for Latin America will be manageable, but a large part of the region’s workers will continue to be immersed in precarious jobs with low productivity, low incomes, and little social protection.
In summary, although it is less discussed in research on the future of employment, the greatest risk for Latin America is that the fourth global technological revolution will have too little impact on the labor markets of the region.