In Indonesia, the phenomenon of education-job mismatch among educated workers has emerged as a critical issue with profound implications for both individuals and the economy at large. Education-job mismatch occurs when there is a discrepancy between an individual’s educational qualifications and the requirements of their job ((Shahidan & Ismail, 2021); (Serikbayeva & Abdulla, 2022)). This phenomenon is typically categorized into two forms: over-education, where individuals possess higher qualifications than their job demands, and under-education, where their educational attainment is below the job’s requirements ((Battu & Bender, 2020); (Wen & Maani, 2022)). Historically, educational systems have emphasized higher education as a path to better job opportunities and economic advancement. However, rapid changes in labor market demands, technological advancements, and the evolving nature of work have increasingly led to situations where the supply of educated labor does not align with job requirements ((Mahy et al., 2015); (Esposito & Scicchitano, 2020), (Hane-Weijman, 2021)). This misalignment raises concerns about the efficacy of educational investments and the optimal use of human capital.

Job search theory provides a critical lens for understanding job-education mismatch by examining the processes through which individuals seek and accept employment. Job search theory provides a framework for understanding job-education mismatch. Stigler's early work in the 1960s emphasized how search costs and imperfect information impact labor markets (Stigler, 1962). Job seekers often lack complete knowledge of all available opportunities, while employers do not fully understand each candidate's capabilities. This information asymmetry forces individuals to weigh the costs of job searching—such as time, money, and effort—against the potential benefits of finding a job that better matches their skills and qualifications.

Further contributions by Diamond, Mortensen, and Pissarides in (Pissarides, 2011), who received the Nobel Prize in Economic Sciences in 2010, expanded on this by introducing the concept of matching frictions. These frictions refer to delays and inefficiencies both job seekers and employers face in finding optimal matches (Pissarides, 2011) (Mortensen, 2011). In the context of educational mismatch, these frictions can cause highly educated individuals to accept jobs below their qualification level due to prolonged search times or high search costs, leading to over-education. Conversely, under-education can occur when less qualified individuals take on roles that typically require higher qualifications, as employers might compromise due to an inability to find ideal candidates. Thus, as highlighted by Banerjee & Sequira (2023), job search theory emphasizes how the dynamics of the search process and labor market conditions contribute to educational mismatches, influenced by imperfect information and the costs associated with job searching.

Figure 1 displays the percentages of over-education and under-education across various countries, with each country represented by two adjacent bars. Over-education is marked by darker bars, while under-education is depicted by lighter bars. South Africa leads the chart with the highest percentage of over-education, reaching nearly 50%, and a notable level of under-education slightly below 30%. Mexico follows with a high over-education rate above 40%, while the under-education rate is less pronounced, slightly above 10%. European countries such as Portugal, Greece, and Turkey also exhibit significant levels of over-education, though their under-education levels vary. Conversely, the European Union as a whole shows a more balanced distribution with moderate levels of both over-education and under-education.

Indonesia is positioned towards the left of the chart, indicating a substantial percentage of over-education, comparable to Mexico and Portugal. Specifically, Indonesia's over-education rate is slightly over 40%, placing it among the top countries in terms of over-education. Its under-education percentage, however, is lower, aligning more closely with countries like Turkey and Greece. This highlights a significant issue of over-education in Indonesia, suggesting that a notable portion of the population may possess educational qualifications that exceed the requirements of their jobs. Compared to countries towards the right of the chart, such as Poland, Sweden, and Hungary, where the levels of over-education and under-education are relatively balanced, Indonesia exhibits a clear skew towards over-education.

More specificly, the Figure 2 compare the percentages of educational match and mismatch (over-education and under-education) across different educational levels (Elementary, Junior High, Senior High, Vocational, and University) for 2019, 2021, and 2023 in Indonesia. Over these years, a consistent pattern emerges where a significant proportion of vocational and university graduates are over-educated for their jobs. For instance, the over-education rate among vocational graduates was around 56.8% in 2019, slightly decreasing to 55.8% in 2021, and then rising to 57.1% in 2023. Similarly, for university graduates, the over-education rate hovered around 50% throughout the period. The under-education rates for vocational graduates showed an increasing trend from 23.8% in 2019 to 26.8% in 2021 but then slightly decreased to 23.9% in 2023. There was no under-education reported for university graduates in any of the years.

For Senior High School graduates, the match rates remained relatively stable, approximately around 75%, with slight fluctuations over the years. Over-education for this group declined from 13.9% in 2019 to 11.0% in 2021 but rose again to 17.5% in 2023. Under-education for Senior High School graduates showed a slight decrease from 10.8% in 2019 to 9.2% in 2021, and a minor increase to 10.4% in 2023. These trends highlight persistent over-education issues at higher educational levels while indicating some improvements and fluctuations in educational alignment for Senior High School graduates. Overall, the data points to an ongoing challenge of aligning educational outcomes with job market demands, particularly for those with vocational and university degrees.

The persistent over-education observed among vocational and university graduates has several socio-economic consequences. For individuals, it often leads to decreased job satisfaction as they perform roles that do not utilize their full skill set or education level, potentially resulting in lower productivity and career dissatisfaction (Frei & Grund, 2022). Financially, these individuals may face a return on investment dilemma, where their educational expenditure does not correspond to their income potential, leading to economic inefficiencies. For employers, this mismatch can mean an inefficient allocation of human resources, where highly educated workers occupy positions that require less skill, potentially stifling innovation and growth (Salas-Velasco, 2021). At a macroeconomic level, over-education can signal a misalignment between the education system and labor market needs, causing a skills mismatch that affects overall economic competitiveness and growth (Battu & Bender, 2020). Addressing these issues requires better coordination between educational institutions and labor market demands, with a focus on aligning curricula with industry needs and enhancing career guidance to ensure that educational attainment translates effectively into suitable employment opportunities.

Understanding the determinants of job-education mismatch is crucial for addressing labor market inefficiencies and improving educational and employment outcomes. Research shows that factors such as field of study, labor market conditions, individual skills, regional economic conditions, and institutional frameworks significantly influence educational mismatches. Graduates in fields with lower market demand are more likely to experience over-education, while those in high-demand disciplines are less prone to mismatches (Capsada-Munsech, 2024). Regional economic disparities also contribute, with job opportunities varying widely across different areas, influencing the extent of mismatches (Andersson et al., 2018). Additionally, the alignment between educational curricula and market needs, as well as the quality of career guidance, affects how well graduates' qualifications match job requirements ((Cleary et al., 2017); (Lichy & Khvatova, 2019); (Aljohani et al., 2022)). Prior studies highlight the need for tailored educational policies and labor market interventions to reduce mismatches and enhance the alignment of skills with job opportunities, leading to better utilization of human capital and improved economic outcomes.

This research aims to explore the determinants of educational mismatch among educated (Senior High School and University graduates) in Indonesia by utilizing microdata. The primary objective is to identify the socio-economic and demographic factors contributing to over-education and under-education within these groups. By analyzing individual-level data, the study seeks to uncover patterns and predictors that explain why graduates experience a misalignment between their educational qualifications and job roles. The implications of this research are significant for policy and practice: it can inform educational policy makers to align curricula more closely with labor market demands, help institutions design targeted career services to better prepare graduates for appropriate job placements, and guide employers in structuring roles and training programs to more effectively utilize the skills of their workforce. Moreover, the findings can contribute to reducing the economic inefficiencies associated with educational mismatch, ultimately leading to improved job satisfaction, higher productivity, and better alignment of human capital with industry needs.

This study investigates the factors contributing to educational mismatches among Senior High School and University graduates in Indonesia. It focuses on identifying socio-economic and demographic variables that influence over-education and under-education. By utilizing individual-level data, the research aims to uncover patterns and predictors that explain the misalignment between educational qualifications and job roles, providing a comprehensive understanding of the determinants of educational mismatch.

The analysis is based on individual-level data from the National Labor Force Survey (Sakernas), conducted biannually by Indonesia's Central Statistics Agency (BPS) to collect employment-related information. The February and August Sakernas surveys differ in scope, with the February survey representing provincial and national levels, while the August survey, with a larger sample size of approximately 300,000 households, allows for district-level estimates. This study uses data from the February 2022 Sakernas survey to construct and measure education-job mismatches among educated workers in Indonesia.

For this research, educated workers are defined as individuals aged 15-64 who are currently employed and have completed at least a high school diploma or its equivalent. Military (TNI) and police (Polri) personnel are excluded from the analysis. Applying these criteria, the final sample consists of 50,647 individuals, providing a robust dataset for analyzing educational mismatches.

The data were categorized into three vertical mismatch categories to match workers' educational levels with the required qualifications for their jobs: matched, mismatch over-education, and mismatch under-education. Examples include individuals with elementary education (SD) working in manual labor roles (matched) or those with higher education (e.g., bachelor's degree) working as managers or professionals (matched). Classification of these vertical mismatch categories follows international standards established by the International Labour Organization (ILO) (see ILO, 2018). Details of the variables used in this study are provided in Table 1 below.

Inferential analysis in this study employs Multilevel Multinomial Logit Regression (MMNL). The multilevel multinomial logit model is suitable for analyzing non-binary, discrete dependent variables and control individual and regional characteristics. This study regressors include individual demographic characteristics, labor demand factors (supply-side), labor market conditions, and spatial characteristics.

The dependent variable is an unordered categorical variable indicating individual vertical mismatch status: matched = 1, mismatch over-education = 2, mismatch under-education = 3. The multinomial logit model is formulated mathematically as follows:

(1)

Where Yi represents the probability of individual i choice among j = 1, 2, …, J available options. Furthermore, Greene (2018) developed a model that accommodates individual characteristics and attributes of choices, where in the context of this study, attributes relate to individual choices concerning labor market conditions, household socio-economic characteristics and provincial variance characteristics. Mathematically, this is expressed as follows:

(2)

Here xij′ represents the attributes related to the choice (labor market conditions and regional characteristics), wi′ denotes variables associated with individual demographic characteristics and zi′γ denotes variables at the provincial level to control variance macro characteristics.

Taking these considerations and adapting from the model by (Grilli & Rampichini, 2007), the specification of the multilevel multinomial logit regression model in this study is as follows:

Yi= β0+ β1Xi + β1Yi + β1Zi + β1Wi + μ (3)

Where Yi is the dependent variable representing vertical mismatch status with three categories: matched, mismatched under-education, and over-education. The independent variables: Xi individual demographic characteristics, Yi labor demand, Zi labor market conditions, and Wi regional characteristics. Further operational definitions of these variables are detailed in Table 1 below.

In conclusion, the determinants of education-job mismatch among educated workers in Indonesia are multifaceted and deeply interconnected. Age and gender are significant factors, with younger workers and women more likely to experience over-education due to structural barriers and informational asymmetries in the labor market. While higher levels of education are often perceived as advantageous, they do not always translate into appropriate job matches, frequently resulting in over-education. Regional economic disparities and sectoral characteristics further complicate the dynamics of mismatch, underlining the need for targeted and localized policy interventions.

To address these challenges effectively, several policy recommendations tailored to the Indonesian context are proposed. First, strengthening career counseling and guidance services within educational institutions is essential. Providing students with comprehensive labor market information, including in-demand skills and employment trends, will enable them to make more informed decisions regarding their education and career paths. Additionally, promoting vocational training and lifelong learning programs, particularly in sectors undergoing rapid technological advancements, will help bridge the gap between educational qualifications and job market requirements. Collaborations between educational institutions and industry stakeholders to align curricula with industry needs and the implementation of work-integrated learning programs will further equip graduates with practical skills and reduce both over-education and under-education.

Addressing regional disparities is particularly critical in Indonesia. Targeted regional development programs aimed at creating job opportunities in rural and underdeveloped areas are needed to reduce under-education and enhance job matching outcomes in these regions. Sector-specific strategies to foster innovation and skills development in agriculture, industry, and services should also be prioritized. These initiatives can align labor supply with demand, ensuring a more effective utilization of human capital.

Gender disparities in job matching require targeted interventions. Policies that promote gender equity, flexible work arrangements, and leadership opportunities for women can help address the higher incidence of over-education among female workers. In addition, enhancing labor market information systems to provide accurate, timely, and accessible data on job vacancies, skill requirements, and employment trends will reduce informational asymmetries and improve the overall efficiency of the job-matching process. Developing digital job-matching platforms tailored to Indonesia’s labor market conditions can further support this effort.