A Study on Labor Market Transformation in the Philippines

An analysis using merged FIES-LFS microdata from 2006 to 2021

Note: This post is based on my capstone paper for the Applied Business Economics program of UA&P.

Job polarization has become a significant concern for developed economies in recent years. This phenomenon is characterized by simultaneous growth of the share of employment in high skill, high wage occupations and low skill, low wage occupations, while the middle-skill occupations experience decline or slower growth. Through the analytical lens of routine-biased technological change (RBTC) hypothesis, I examined whether the job polarization phenomenon is occurring in the Philippines by analyzing job growth patterns across dominant task classifications and major occupation groups, examining whether middle-skill jobs have been outpaced by low- and high-skill jobs, and analyzing changes in middle-class employment characteristics over time.

The RBTC hypothesis suggests that the introduction of technology, particularly those that can be described as "general purpose technology" favors jobs that are more non-routine in nature as opposed to occupations with high routine content. This, in turn, leads to job polarization where the occupations in the middle of the skill distribution experience modest growth to declining levels of employment while those at high and low end of the distribution are experiencing growth.

Key Findings

My analysis found several key insights regarding Philippine labor market transformation from 2006-2021. First, among the four dominant tasks, routine-cognitive (RC) jobs were found to have the most dramatic increase in their share of total employment.¹ On the other hand, non-routine cognitive (NRC) jobs experienced the most significant declines, not only in share of total employment but in terms of absolute levels of employment as well.² Second, "clerical support workers" and "service and sales workers", classified as medium in the International Standard Classification of Occupations (ISCO) skill level framework, were found to have the largest gains in employment, while "managers", "skilled agricultural workers", and "technicians and associate professionals" declined both in terms of share in total employment and absolute levels. Third, the occupational skill distribution analyses, using smoothed changes and decile analysis, found that the changes across the skill distribution do not exhibit the distinctive U-shaped pattern. Fourth, the middle-income class was found to have expanded within growing occupational categories, with multinomial regression analysis confirming that growing job categories like "professionals" and "clerical support workers" maintained strong associations with middle-income status.

Why the Philippines Differs

The analysis of the merged FIES-LFS microdata between 2006 to 2021 reveals that while the Philippine labor market experienced transformation, it does not have the characteristics expected under the RBTC hypothesis nor exhibits evidence of job polarization. The Philippine economic structure differs fundamentally from high-income countries that transitioned from large manufacturing sectors employing routine-intensive jobs at median wages. Instead, the Philippines developed from agriculture to services, largely bypassing manufacturing-intensive industrialization while retaining significant agricultural employment. Studies cited in the meta-analysis by Martins-Neto et al. indicate that job polarization occurs faster in manufacturing than services.³

Job polarization may eventually emerge when RC job growth plateaus, driven by firms' greater incentives to adopt labor-displacing technologies and improved technological competencies. However, the country's high concentration of non-routine manual (NRM) workers in agriculture and construction suggests mechanization and robot adoption could cause absolute declines in lowest-skill occupations. The government's pursuit of manufacturing-intensive industrialization may further minimize polarization effects, as manufacturing typically employs routine manual workers at middle-skill levels.

Recommendations

Based on these findings, I recommend that:

For Firms:

• Focus on upskilling professional and associate professional workforce with managerial competencies, based on trends observed and assuming that these trends will persist in the medium term

• Identify the optimal combination of labor input and automation technology that maximizes productivity, within their specific operational context

For Job Seekers and Workers:

• Focus on building professional, technical and sales skills

For Government:

• Establish a system enabling business and economics researchers to access anonymized data collected by the Department of Labor and Employment, for legitimate purposes

• Continue improving economic and statistical databases to ease access to labor and employment statistics and macroeconomic data in general

• Pursue continuing the World Bank's STEP Skills Measurement Program (STEP) which contains rich sources of occupational information, especially those relevant to task- and skill-based research approaches to labor and employment

• Management programs should be revisited to ensure they are versatile and adaptable to changing industry requirements

• Professional and technical programs should be re-assessed to equip students with managerial and leadership skills early in their academic development

Areas for Future Research

Future research on labor transformation can explore several directions:

• Utilize LFS microdata to build upon these findings, as opposed to the merged FIES-LFS microdata⁴

• Establish the reasons why job polarization is not evident in the Philippine context, building on the finding that initial high levels of non-routine dominant tasks may be a contributing factor

• Examine the unemployment and underemployment dimensions of labor transformation

• Analyze the role of overseas employment and emigration patterns in shaping domestic labor markets

• Conduct spatial analysis to understand regional variations in labor transformation and internal migration patterns

1. Routine cognitive jobs are occupations that generally “follow exact, straightforward, and repetitive procedures” and requires analytical skills.

2. Non-routine cognitive jobs typically consist of tasks demanding flexibility, creativity, generalized problem-solving, and complex communications.

3. Antonio Martins-Neto et al., “Is There Job Polarization in Developing Economies? A Review and Outlook,” The World Bank Research Observer 39, no. 2 (August 2024): 259–88, https://doi.org/10.1093/wbro/lkad008.

4. While an analysis specific for the middle-income class may not be feasible in LFS microdata alone, using the LFS microdata benefits from more frequent data and allows for the examination of seasonal variation.