Measuring Jobs Beyond Employment: How Task-Based Data Reveals the Future of Work

The Asian Development Bank’s 2025 report shows that traditional labor statistics fail to capture how jobs, skills, and tasks are changing due to technology, automation, and new work arrangements, leaving policymakers unprepared for future risks. By piloting a Jobs and Skills Survey in Bhutan, Georgia, and the Philippines, the study demonstrates that task-based data can reveal skills mismatches, gender inequalities, and automation exposure, helping governments design better jobs and skills policies.

Produced by the Asian Development Bank (ADB) through its Economic Research and Development Impact Department, in close collaboration with the National Statistics Bureau of Bhutan, the National Statistics Office of Georgia, and the Philippine Statistics Authority, the December 2025 report reflects joint work with and insights from international institutions such as the International Labour Organization, the World Bank, the OECD, and academic partners including the Centre for Economic Policy Research. The report argues that while labor markets across Asia and the Pacific are changing rapidly, official statistics have struggled to keep up, leaving policymakers without a clear picture of how people actually work today.

Economic growth, digital technologies, globalization, and the COVID-19 pandemic have transformed jobs in ways that go far beyond simple employment counts. More people now work remotely, use digital tools, or earn income through short-term or platform-based jobs. At the same time, automation is replacing repetitive tasks while increasing demand for skills such as problem-solving, communication, and creativity. Yet most labor force surveys still focus on whether someone is employed, not on what they do at work. This gap makes it difficult to design effective skills policies or prepare workers for future risks.

The Jobs and Skills Survey: A New Tool

To address this problem, ADB developed the Jobs and Skills Survey (JSS) module, designed to be added to existing labor force surveys rather than replace them. The idea is simple: keep the familiar structure of official surveys, but add questions that capture job tasks, skills used at work, exposure to new technologies, and new work arrangements such as remote work and gig employment. This approach makes it possible to collect richer data without overburdening national statistical systems.

The JSS module was pilot-tested in Bhutan, Georgia, and the Philippines because these countries conduct labor force surveys regularly and face important employment challenges. Bhutan struggles with high youth unemployment and heavy reliance on public sector jobs, Georgia faces persistent skills mismatches, and the Philippines combines strong service-sector growth with high exposure to automation. The pilots showed that collecting detailed task-level data is both feasible and valuable.

What Workers Actually Do on the Job

The report’s findings reveal strong patterns in how work is organized. Tasks are grouped into four main types: nonroutine cognitive analytical tasks, such as problem-solving and writing, nonroutine cognitive interpersonal tasks, such as supervision and public speaking; routine cognitive tasks, like filling forms and basic calculations, and manual tasks involving physical or precision work.

Across all three countries, men are more likely to perform nonroutine tasks linked to decision-making and leadership, while women are more often concentrated in routine cognitive tasks. Manual work requiring physical strength remains dominated by men, although women are often more involved in precision-based manual activities. These patterns matter because routine tasks are more vulnerable to automation, while nonroutine tasks tend to be better paid and more secure.

Who Faces the Greatest Automation Risk

To better understand this vulnerability, the report uses a measure called routine task intensity, which shows how heavily a job relies on repetitive, rules-based tasks. The analysis confirms a clear global trend: as countries become richer, routine task intensity generally falls, especially in professional and managerial jobs. However, clerical, service, and elementary occupations remain routine-heavy even in middle-income economies.

In the pilot countries, women consistently show higher routine task intensity than men within the same occupations. This means gender inequality is not only about access to jobs, but also about job content. The Philippines stands out for especially high routine intensity in clerical and service jobs, linked to business process outsourcing, while Bhutan shows high routine intensity in agriculture and crafts. These patterns signal where automation risks may be most severe.

What This Means for Policy

The report concludes that understanding job tasks is essential for effective labor and skills policy. Without knowing what workers actually do, governments cannot design training programs that match real labor market needs or protect those most at risk of displacement. The JSS module offers a practical approach to bridging this knowledge gap by modernizing labor statistics within existing systems.

By demonstrating that task-based data can be collected reliably and at scale, the pilots make a compelling case for its wider adoption. Better data can help governments anticipate change rather than react to it, ensuring that technological progress leads to better jobs, fairer opportunities, and more inclusive growth across Asia and the Pacific.