Technological change is one of the main drivers of economic growth and improved living standards. But its impact on the labour market and wage inequality continues to be a subject of intense debate. Concerns about the effects of technological advances on wage distribution contribute to social anxiety and reinforce the perception that innovation can leave a significant proportion of the population behind.

Against this backdrop, recent research (Palomino et al., 2025) examines the distributional consequences of technological change in the Spanish economy over the last two decades, thereby offering a new perspective in order to understand this highly relevant yet complex phenomenon. The study quantifies the impact of technology on inequality by using an innovative methodology that accurately captures the dynamic nature of automation. The results are clear: technological change has been a key factor in increasing wage inequality in Spain.
 

The research’s main finding is that automation decisively contributed to greater wage inequality in Spain between 2000 and 2019. This result was obtained using an empirical strategy that combines the creation of a dynamic automation index and regression analysis with a thorough counterfactual simulation that involved an estimate of what wages would have been if workers had not been exposed to automation technologies.

The comparison between the real and simulated scenarios yields clear results. Without automation, the Gini index – a standard measure of inequality – would have been 21.7% lower in 2019, reflecting the significance of the effect. A distributional analysis of the wage bill also reveals a clear pattern of redistribution. Without technological change, the top 10% of workers with the highest wages would have accumulated a 4% smaller share of wages. In contrast, the middle and bottom segments of the distribution would have seen their share increase. These data confirm that automation has led to a transfer of wage income from the bottom and middle segments to the top of the pyramid.

Another of the most significant effects of automation is the widening of the wage gap associated with educational differences. Analysis shows that workers with a higher level of education have been less exposed to the automation of their tasks and in many cases have often benefited from technologies that complement their skills.

The study estimates that the education wage premium – the average wage gap between higher- and lower-educated workers – would have been 76.6% lower in a scenario without automation between 2000 and 2019. This finding lends weight to the theory that automation replaces tasks previously performed by medium- and low-skilled workers, while complementing those of workers with more training (Acemoglu and Autor, 2011). This not only increases the relative demand for the latter workers, but also strengthens their bargaining position in the labour market, resulting in higher wages.

As can be seen in the following figure, the adverse impact of automation is systematically greater for workers with lower levels of education (dark blue line), for both men and women and across all age groups. In contrast, those with higher education (yellow line) are the least affected, experiencing an impact far below the average (dotted line).

One of the most notable findings of the research is that automation has helped to reduce the gender pay gap in Spain. Without automation, this gap would have been approximately 68% larger between 2000 and 2019.

The explanation lies in the structure of the Spanish labour market. Automation has disproportionately affected traditionally male-dominated jobs, particularly in the industrial sector. As tasks within male-dominated occupations have been replaced, the negative impact on men’s wages was greater in relative terms, thereby leading to a convergence in the average wages of men and women. Figure 1 above illustrates this point: for each level of education, the lines corresponding to women tend to be below those for men, indicating a smaller adverse impact.

However, the effect of technology is not consistent across all age groups. Analysis reveals that the impact tends to be more negative for younger workers of both sexes. As can also be seen in Figure 1, the lines slope downward in most groups, suggesting that older workers have suffered to a lesser extent from the wage consequences of automation than their younger colleagues.

The study also examines the impact of the most disruptive technology of recent years: artificial intelligence (AI). For the period 2015-2019, during which time this technology began to be implemented in certain industries, it was found that exposure to AI also led to an increase in wage inequality, albeit with a different mechanism to that of traditional automation.

While automation tends to depress the wages of medium- and low-skilled workers, the results suggest that AI has the potential of increasing the wages of more qualified workers. However, the net effect on inequality is similar. In a counterfactual scenario without the influence of AI, as measured by the Felten index, the Gini index in 2019 would have been 11.3% lower. 

The graph below shows how the positive wage changes observed (blue bubbles) are concentrated in high wages with high levels of AI (top graph) and low levels of automation (bottom graph). Nevertheless, negative wage changes (brown bubbles) in both cases are concentrated in the middle segment of wage distribution for both automation and the AI index, so that wage inequality increases in both cases.

In addition to technological change, the research examines the impact of various factors, including trade openness, regional employment structures and the market power of companies (profit margins).

Greater trade openness and higher regional employment rates are found to be positively associated with wage growth. Conversely, the ability to outsource and relocate economic activities (offshorability) and the profit margins of companies do not appear to have had a significant or consistent effect throughout the entire period, although they have been observed in specific subperiods. In any case, the impact of these factors is considerably smaller than that of technological change.
 

The results of this research confirm that fears about technology in terms of its impact on inequality are well founded. Technological change over the last two decades has altered Spain’s wage structure, negatively impacting the middle segment of wage distribution and less-skilled workers in particular, thereby increasing wage inequality. Surmising that the gains from technological change will eventually ‘trickle down’ to society as a whole therefore fails to capture the complexity of the phenomenon. In fact, the findings highlight the need for public policies that ensure the benefits of innovation are shared more equitably.

In order to address this challenge, various authors have proposed updating tax systems by developing new protection schemes aimed at the most vulnerable groups, as well as redistributing the burden between capital (whose returns would increase given the nature of technological progress) and labour. While these measures would alleviate the harmful effects of technological change to some extent, they would not solve the problem: the replacement of human labour by new technologies and the power relations arising from this interaction. It is therefore urgent to facilitate the workforce’s adaptation to the technological challenges that our society must face in the coming years. This would improve the employability and position of workers in bargaining processes. In order to achieve this, substantial investments in human capital development and reskilling programmes are required to ensure that the benefits of innovation can be fully realised.

ACEMOGLU, D.; D. AUTOR (2011): «Skills, tasks and technologies: implications for employment earnings», en O. Ashenfelter y D. Card (eds.), Handbook of Labor Economics, vol. 4B, 1043-1171, Elsevier: Ámsterdam.

PALOMINO, J. C. et al. (2025): New Technologies and the rise of wage inequality, ICAE Working Paper, en prensa.