The Fault Line Behind Indonesia’s Soshum and STEM Divide

Sumber: dokpri

Indonesia’s long-standing debate between Soshum (social sciences and humanities) and STEM (science, technology, engineering, and mathematics) has too often been reduced to cultural stereotypes, between the “practical” and the “philosophical”, the “builders” and the “talkers”. Yet, beneath those superficial labels, lies an economic fault line: a structural disequilibrium between the supply of knowledge and the economy’s capacity to transform that knowledge to productivity.

This imbalance is beyond classroom problems. This country’s education system produces graduates at an accelerating pace, but its economy is out of sync with its own ambitions. Universities just keep producing degrees faster than industries can create meaningful work. The issue thus isn’t just about what Indonesians learn, but whether the nation can make that learning matter.

The Disequilibrium

The government has placed STEM at the center of its development narrative through several policies, from prioritizing STEM fields for Indonesia Endowment Fund for Education (LPDP), to allocating IDR 500 billion to train foundational AI and STEM skills in some 10 million students at about 500 schools across Indonesia (The Jakarta Post, 2025). Ministries and universities voice the need for more engineers, data scientists, and innovators to drive the nation’s industrial modernization. According to Bappenas (2023), only 18.47% of Indonesian university graduates come from STEM fields, far lower than in Malaysia (37.19%), Singapore (34.30%), or India (31.41%). Data from the Ministry of Education, Culture, Research, and Technology (2024) also shows that the number of students who choose education, economics, religion, and art programs can reach more than 3.7 million students while STEM-related majors such as engineering, mathematics, natural sciences, health, and agriculture only have around 1.7 million students. At first glance, this seems to justify the “STEM push” agenda. STEM graduates are expected to embody higher potential productivity due to their skill intensity and technical transferability. But the equation is not that simple. Globally, many economies have failed to align educational output with labor demand. In the book Wasted Education: How We Fail STEM Graduates, University of California San Diego sociologist, John D. Skrentny, reveals striking data: between 30% and 60% of STEM graduates end up working outside their fields. A 2021 U.S. Census Bureau study pushes that number even higher, estimating that up to 72% of STEM degree holders are employed in non-STEM fields. Indonesia mirrors this pattern. Hasibuan & Handayani (2019) estimate that 68.4% of Indonesian workers experience field-of-study mismatch, reflecting widespread misallocation of human capital. This imbalance between educational output and market absorption produces a hidden macroeconomic cost. Empirical evidence from OECD countries shows that mismatch between fields of study and occupational demand generates productivity losses. Productivity costs associated with field-of-study mismatch may amount to over 1% of GDP in England, Northern Ireland, and Estonia, and more than 0.5% in Korea, Ireland, Canada, Germany, Poland, Spain, and the United States (OECD, 2015). Beyond aggregate level, field-of-study mismatches also carry significant individual costs: workers employed outside their field of study face a wage penalty of 6.37% to 7.36% on average (Hasibuan & Handayani, 2019). This reveals how large the inefficiency can be when education systems expand without corresponding transformation in the industrial base that is supposed to absorb and reward those skills.

How STEM Contributes to the Economy

The economic case for STEM is not merely about education, but about how technological capability can shape long-term productivity. Historical evidence from Maloney and Caicedo (2022) demonstrates this with empirical evidence: a one-standard-deviation increase in the share of engineers in 1880 predicts a 10% higher GDP per capita in the United States today, while patent capacity contributes an additional 10%. Notably, they observe that countries with comparable income levels in 1900, such as Argentina, Chile, Denmark, Sweden, and the southern United States, diverged dramatically in subsequent decades depending on their engineering density (Figure 1).

Figure 1. Plot of engineering density versus Log of GDP per capita in 1900

Source: Maloney & Caicedo, 2022

Across the Americas, approximately 25% of the income gap between North and South America can be explained by differences in engineer density, underscoring how deeply technological human capital determines the trajectory of prosperity. Moreover, STEM graduates, whether domestically trained or foreign-born, have significant influence on patent intensity with coefficients of 34.29 and 28.54, respectively, while non-STEM graduates exhibit an insignificant coefficient of 1.41 (Winters, 2014). Patent protection plays a critical role in sustaining innovation incentives. Strong intellectual property rights create innovation rents which motivate scientists, engineers, and firms to invest in research and new technologies. Empirical evidence by Aghion, Howitt, and Prantl (NBER, 2013) shows that industries operating under stronger patent regimes respond to market reforms with significantly higher innovation output, measured through increased R&D spending and patent filings. In other words, when innovators can securely capture the returns on their ideas, they are more likely to take the risks required to develop them. This mechanism reinforces the Schumpeterian process of creative destruction, where protected innovation rents finance the birth of new technologies that replace outdated ones, a cycle essential to productivity growth. For developing economies, in a sub sample of 16 countries, the patent reforms also led to an increase in industry-level value added due to technology transfer by multinational companies (Branstetter et al, 2011).

However, the economic payoff of STEM education depends on whether industries exist to absorb the skills being produced. STEM fields have a narrower and more industry-specific employment space, meaning their labor market outcomes are far more sensitive to shifts in industrial demand. Engineers amplify productivity only when economies possess sufficient industrial capacity to absorb their skills. Theodora (2023) stated that in 2013, every IDR 1 trillion of manufacturing investment in Indonesia could employ around 4,594 workers. By 2016, that number had fallen to 2,271, and by 2021, to only 1,340 workers. This trend stems from two underlying factors: first, most new investments are concentrated in capital and technology-intensive industries; second, formal sector employment has been steadily declining, forcing much of the surplus workforce to seek livelihoods in the informal economy, where productivity and security are far lower. This reflects a pattern of capital-deepening without corresponding labor absorption capacity, indicating much of Indonesia’s surplus labor, including STEM-trained workers, ends up absorbed into the informal or non-technical occupations, where their skills are underutilized and productivity gains remain limited.

The Social Architecture of Growth

Innovation alone does not guarantee prosperity. Public policy and institutional design determine whether new ideas translate into broad-based economic transformation. This principle lies at the heart of Abramovitz’s (1986) Social Capability Theory, which defines a nation’s social capability as its ability to absorb, adapt, and benefit from modern technology through effective institutions and governance. Without this capability, even cutting-edge innovations remain isolated, benefiting only a small elite. Technology diffusion, the process by which knowledge from research and innovation spreads across sectors, depends critically on social infrastructure. While STEM disciplines drive innovation, it is the social sciences and humanities that enable transformation. However, there is also a trade-off between innovation and diffusion. As patents lengthen, firms are more likely to innovate because they can secure exclusive profits, but this comes at the cost of slower knowledge sharing. Once patents expire, the social benefits rise sharply as innovations become freely accessible and diffuse throughout the economy. The optimal point lies where creativity is rewarded but knowledge remains accessible. This is where the social sciences and humanities take part through laws and public policies that mediate between private profit and social benefit, ensuring that innovation serves not only inventors but society as a whole (Figure 2).

Figure 2. Optimal Patent Duration

Source: CORE, 2017

Public policy has always been central to innovation. For innovators to take risks, they must trust that their innovation rents will be protected by strong property rights and reliable legal systems, as seen in early industrial leaders like the UK and the Netherlands. Modern innovation hubs such as Silicon Valley and Germany’s industrial ecosystem thrive because governments provide complementary inputs such as infrastructure, education, basic research, and temporary monopolies that later yield to competition. This balance between private incentives and supportive public policy makes capitalism dynamic, raising living standards and driving technological renewal through what Joseph Schumpeter called creative destruction (CORE, 2017).

For Indonesia, where R&D intensity remains at only around 0.3% of GDP and patenting activity limited, the aim should not just to produce more STEM graduates, but to create a system that can turn its growing base talent into tangible innovation outcomes. Innovation without diffusion causes inequality, diffusion without innovation causes stagnation. Indonesia’s problem lies in this missing balance. The government’s push to expand STEM education is commendable, but without efforts to strengthen economic and institutional capacity to absorb and apply that knowledge, it risks worsening the inefficiencies it hopes to fix.

References

Ayyubi, M. S. A., & Pratomo, D. S. (2025). Graduating at the wrong time: Labor market outcomes of colleges graduates during economic contraction. Jurnal Ekonomi Indonesia, 14(1), 38–53. https://doi.org/10.52813/jei.v14i1.397

Hasibuan, E. A. (2019). Qualification mismatch dan pendapatan tenaga kerja di Indonesia = Qualification mismatch and labor earnings in Indonesia. Universitas Indonesia Library. https://lib.ui.ac.id/detail?id=20501921&lokasi=lokal

Maloney, W. F., & Caicedo, F. V. (2022). Engineering growth. Journal of the European Economic Association, 20(4), 1554–1594. https://doi.org/10.1093/jeea/jvac014

Montt, G. (2015). The causes and consequences of field-of-study mismatch: An analysis using PIAAC. In OECD Social, Employment and Migration Working Papers (No. 167). https://www.oecd.org/content/dam/oecd/en/publications/reports/2015/07/the-causes-and-consequences-of-field-of-study-mismatch_g17a26a5/5jrxm4dhv9r2-en.pdf

Nababan, T. S., & Purba, E. F. (2023, November 3). Labour absorption in manufacturing industry in Indonesia: anomalous and regressive phenomena. arXiv.org. https://arxiv.org/abs/2311.01787

Paunov, C., Planes-Satorra, S., Moriguchi, T., OECD, & Japan Patent Office. (2017). WHAT ROLE FOR SOCIAL SCIENCES IN INNOVATION? RE-ASSESSING HOW SCIENTIFIC DISCIPLINES CONTRIBUTE TO DIFFERENT INDUSTRIES. OECD SCIENCE, TECHNOLOGY AND INDUSTRY POLICY PAPERS. https://www.oecd.org/content/dam/oecd/en/publications/reports/2017/11/what-role-for-social-sciences-in-innovation_92aa744c/8a306011-en.pdf

Post, J. (2025, June 1). National AI, STEM initiative started by ministries, foundation. The Jakarta Post. https://www.thejakartapost.com/indonesia/2025/06/01/national-ai-stem-initiative-started-by-ministries-foundation.html

Winters, J. V. (2014). Foreign and Native-Born STEM graduates and innovation intensity in the United States. In IZA, IZA Discussion Paper Series (No. 8575). IZA. https://docs.iza.org/dp8575.pdf