The Luddites, a group of workers during the Industrial Revolution who destroyed textile machinery, are often portrayed as a short-sighted, violent mob who were wrong about technology. Yet their core fear—that machines would replace human workers—wasn't entirely unfounded.
Today, the advancement of Artificial Intelligence has revived similar concerns about job displacement. These fears are valid, and economists have long studied how technology reshapes the nature of work.
One key finding challenges common assumptions: instead of technology simply eliminating low-skill jobs while favouring high-skill ones, what actually happened is more nuanced. In industrialized countries, middle-paying jobs declined, while both low- and high-paying jobs increased. This phenomenon is called “job polarization.”
Building on the routine-biased technological change (RBTC) hypothesis developed by David Autor, Frank Levy, and Richard Murnane, a subsequent study by Autor, Lawrence Katz, and Melissa Kearney provided one of the first empirical findings on job polarization in the US.
This study adopted the “routinization” hypothesis in exploring whether job polarization is already taking place in the Philippines.
Routine tasks are defined as a limited and well-defined set of cognitive and manual activities that can be accomplished by following explicit rules. Examples of routine-intensive occupations include library clerks and bank tellers (routine cognitive), as well as power plant operators and hand packers (routine manual).
There are at least three causes of job polarization: technology, trade and wage inequality.
The earlier theories in technology displacing jobs predicts an imbalanced growth of employment in a way that favours high-skill (high-paying) employment. However, that is not what has been observed in many developed countries.
The prevailing explanation on why is this is the case is the “routinization” hypothesis developed by David Autor, Frank Levy, and Richard Murnane (ALM). In their hypothesis, technology is not displacing jobs but is actually replacing routine tasks while augmenting non-routine tasks.
This is related to the the “routinization” hypothesis but placing emphasis on whether the service can be delivered “impersonally” (that is, through the internet or over the phone). All things being equal, routine-intensive jobs are easier to offshore but there are many non-routine, analytical jobs that are offshorable (e.g., data analyst, game developer, and software engineers).
A third explanation focuses on how rising income inequality itself may drive job polarization. As the share of income going to the rich rises in the developed countries, demand for low-skill workers increases, where they are mainly employed in jobs that provide services to the rich. Compared to earlier mentioned hypotheses, this has been less extensively explored.
Instead, most discussions focus on the effects of job polarization on wage inequality—particularly the growing gap between middle- and high-wage occupations, even as the gap between low- and middle-wage jobs appears to narrow.
If we plot the changes in the share in employment between 2006 and 2023 against the average wage in 2006, the chart will not show a distinctive U-shape pattern characteristic of a “job polarization” phenomenon.
The data used in this study are the merged microdata from the Labor Force Survey and Family Income and Expenditure Survey available at the Philippine Statistics Authority Data Archive (PSADA) from 2006 to 2023.
The 2006 average wages is being used here as a proxy for skills. Using 2018 wages also yield comparable results.
While the smoothed trends may mask actual shifts in the distribution, even the decile plot shows no evidence of “job polarization”.
Although the Philippine labour market isn’t experiencing job polarization, it is undergoing significant transformation.
This suggest a supply and demand dynamics where there is a surging firms' demand for routine cognitive employees is met with an increase in supply that is highly wage elastic. A high wage elasticity means that workers are willing to supply more work corresponding less than the increase in wage offered.
The legislated minimum wage also influence this dynamics where it acts as the binding price floor. This means that employers, particularly in the formal economy, are hiring at or near minimum wage levels.
Notice that managers experienced a decline both in their share of total employment and in absolute numbers. Professionals, on the other hand, are steadily rising.
In the advanced economies, there is a phenomenon called the “great flatenning” wherein professionals are displacing managers in the workplace. The observed pattern here suggest that this organizational resturcturing may be already occuring in the Philippines.
The findings presented demonstrate that while the Philippines experienced significant labour market transformation, the country has not experienced the characteristics expected under RBTC hypotheses, nor the job polarization observed in advanced economies.
These broad patterns tell one part of the story. To see how specific occupations are changing, explore the data yourself.
Explore trends by occupation.Aries Eroles, "Is the Philippines losing its middle-skill jobs?," EconStat, December 05, 2025, /data/special-reports/job-polar/.
@misc{Eroles2025IsthePhili,
author = {Aries Eroles},
title = {Is the Philippines losing its middle-skill jobs?},
howpublished = {EconStat},
year = {2025},
note = {Accessed: /data/special-reports/job-polar/}
}