The public debate on the quality of employment in Spain has settled into a simplistic narrative that is increasingly insufficient to understand the reality of our labor market. A simplistic perception would suggest that most of the employment created in recent years would be, fundamentally, precarious and poorly paid. This view, while understandable given the legitimate concerns about temporary employment and low wages in some sectors, and the job insecurity established in our country especially when the Great Recession of 2008-2012 erupted, omits a key part of the story that is not unique to our labor market: the Spanish labor market is experiencing a process of polarization that simultaneously generates high-skilled, well-paid jobs alongside occupations at the lower end of the wage distribution.
Data from the 2022 Labor Force Survey and Wage Structure Survey reveal this profile. Although on this occasion I focus on the most recent period, many studies look back and reveal this dynamic with very specific and studied causes. Thus, between 2023 and 2025, the best-paid professions, identified by the national professional classification code (CNO), experienced notable growth: technicians, scientific and intellectual professionals, with an average gross annual salary of over 43,000 euros, increased their presence by almost 450,000 people, or an increase of 22.5%. Directors and managers, whose average salary is around 65,000 euros per year, grew by 9% and technical jobs, with salaries above 33,000 euros, by 10%. This is not marginal or anecdotal data; The five highest-wage occupations increased by 834,000 jobs in just two and a half years, 47% of the increase in total employment over this period.
But it cannot be ignored that at the same time as this increase there has also been a significant growth in lower-paid occupations. Restaurant and retail workers increased by over 329,000 units, with an average salary of around 16,500 euros per year. The four basic occupations with the lowest wages (less than 20,000 euros per year) added 615,000 new jobs. Thus, two-thirds of employment was created at opposite ends of the wage distribution, indicating, albeit with some difficulty given the short time period over which it is analysed, some continuity in the polarization of the labor market.
What is particularly revealing, and I insist more evidently when we compare years further away, is the lower growth in intermediate occupations. The number of employees increased much more slowly and, with an increase of 121,000 employees, grew by just 5.8%, while installers and operators of machinery recorded a decrease of 0.7%. This weakness in job growth at the center of the employment distribution is neither coincidental nor specifically Spanish; It responds to documented phenomena in developed economies that, while difficult to identify in just two years, have been done for earlier, broader periods.
First, technological change acts as a selective sieve on the labor market. Before the advent of artificial intelligence (AI), the automation of routine, codifiable tasks, regardless of whether they were manual or cognitive, was relatively easy to replace. An assembly worker who repeats standardized sequences or an administrative employee who processes documents according to established protocols are at high risk of technological substitution. Not because their work is useless, but because its repetitive nature allows for algorithmic programming. And it is efficient to replace it with a machine.
In contrast, occupations at the extremes of the distribution have greater resistance to automation for several reasons. At the high end, scientific, medical, engineering, or managerial professionals perform complex, non-routine tasks that require analytical skills, creativity, decision making in uncertain contexts, and sophisticated interpersonal skills. And here, although not homogeneously, AI can intensify this trend. Therefore, these skills, for the time being, remain firmly in the territory of human comparative advantage. At the lower end, occupations such as healthcare workers, housekeeping staff or hospitality workers involve non-routine manual tasks, adaptability to changing contexts and personal interaction which, although they can be technically automated, are not economically viable to robotize given the low wage costs.
This polarization process has profound implications for social structure and income distribution. The weakening of intermediate occupations traditionally associated with the middle class – skilled administrators, specialized technicians, industrial workers – already erodes a fundamental pillar of social cohesion. Historically, these jobs have provided predictable trajectories of upward mobility, economic stability, and consumption capacity that have supported aggregate demand. Their relative decline raises questions about advancement opportunities for large segments of the population.
However, recognizing polarization does not imply a fatalistic resignation in the face of an alleged universal precariousness. The nearly three million new jobs created in the analyzed period include a significant percentage of well-paid and highly skilled occupations. Ignoring this reality leads to incomplete diagnoses and, therefore, potentially inadequate public policy proposals.
The phenomenon acquires further complexity if the migratory dimension is considered. Migration flows might be thought to concentrate preferentially in occupations at the lower end of the wage distribution, acting as a relief valve for sectors with structural difficulties in attracting domestic workers. This does not automatically make these jobs precarious, although many are, but rather reflects the different trajectories and restrictions faced by workers with different origins and human capital endowments recognized in the Spanish labor market. However, despite this, many immigrant workers have obtained jobs in occupations that do not necessarily have low wages, and this massification into low-skilled jobs is not so clear.
The implications for public policy are clear. First, it is imperative to invest in ongoing training and reskilling of workers whose skills are at risk of technological obsolescence, providing pathways for professional mobility. Second, it is necessary to strengthen social protection and labor institutions to mitigate the negative effects of these structural adjustments.
The public debate would gain depth by abandoning simplifying narratives that portray the labor market exclusively as generalized precariousness. The reality is more complex. Understanding polarization as a structural phenomenon linked to technological change, rather than as the result of harmful policies, allows us to design more effective responses. Recognizing that quality employment is created does not deny precariousness in some sectors. Both realities coexist and require differentiated attention. Only by starting from this nuanced understanding can we articulate the policies that the 21st century labor market requires.
