Abstract

Rural labor markets in developing countries remain highly vulnerable to structural inequalities, where informality, agricultural dependence, and limited access to education and technology combine to suppress productivity and wage outcomes. Indonesia provides a compelling case, as the majority of its rural workforce is engaged in low-paying, insecure jobs with limited opportunities for upward mobility. Understanding the dynamics of rural employment, particularly the intersection of informality, gender disparities, and digital transformation, is essential for shaping effective labor and development policies. This study seeks to illuminate these dynamics and explore pathways toward reducing earning gaps and unlocking rural productivity.
Using the August 2024 National Labor Force Survey (Sakernas), with a sample of 221,264 rural workers aged 15 years and above who earn income, we applied linear multilevel models across all districts and municipalities in Indonesia to assess determinants of rural earning. We further conducted Ordinary Least Squares (OLS) analyses focusing on three disadvantaged groups: women, informal workers, and young workers. To examine the potential of digital technologies, especially generative AI, in shaping rural employment prospects, we employed multinomial logit models to capture patterns of occupational exposure to AI-driven tasks.
The findings reveal several structural challenges. Informal status and employment in agriculture substantially depress wages, while sharp gender disparities persist, with women systematically earning less than men. Wages rise with age (after adolescence) and education, although the strongest returns are associated with non-STEM tertiary education. Regional variation across districts, while relatively modest, remains relevant for wage-setting and local training policies.
For rural women, higher education—particularly in non-STEM fields—internet access for work, and marital status significantly raise earnings, with formal training providing additional though limited benefits. Agricultural attachment, however, continues to lower wages. Among informal workers, tertiary STEM education and internet use for work yield substantial earning premiums, while marital status provides an additional advantage. Formal training shows no consistent positive effect, suggesting current training schemes may be misaligned with informal sector realities. For young workers, secondary education, formal training, and internet use significantly enhance earning, though the effectiveness of tertiary education remains questionable given the non-significant and negative coefficients.
A critical insight of this study lies in the transformative potential of digitalization. Across all three disadvantaged groups, internet use for work consistently improves earning prospects, signaling its role as a key enabler of rural productivity. Extending the analysis to generative AI exposure highlights both opportunities and barriers. Rural agricultural workers remain nearly excluded from AI-related tasks, with less than five percent probability of achieving even minimal exposure compared to non-agricultural workers. Within the informal sector, exposure follows a layered pattern: wage employees dominate access to advanced exposure (“Gradient 4”), self-employed workers are concentrated in intermediate levels, while casual workers remain almost untouched. Education strongly mediates AI exposure—secondary schooling raises basic exposure odds by 40–50 percent, yet only STEM degrees consistently unlock higher levels of AI task integration, with odds ratios reaching double-digit magnitudes. Interestingly, formal training programs reduce initial AI exposure odds but provide a marginal boost at higher levels, suggesting a mismatch between training curricula and digital entry-level needs.
These results underscore the urgency of targeted interventions. Strengthening digital inclusion through affordable internet access and rural digital hubs can narrow income gaps and expand economic opportunities. Aligning vocational training with AI-relevant foundational skills, particularly for women and young workers, would bridge the entry-level digital divide. Encouraging STEM education and ensuring its accessibility to marginalized groups remains critical for long-term competitiveness. Additionally, diversifying rural employment beyond agriculture and integrating AI-enabled technologies into agricultural value chains can anchor inclusive rural transformation.