Extended Summary

This study empirically examines the relationship between human capital accumulation and unemployment dynamics in Türkiye during the 2001-2022 period and re-evaluates the validity of human capital theory under current conditions. According to the classical human capital approach, investments in education, health, and skills should increase individuals’ productivity and reduce unemployment; however, the persistent high unemployment rate despite the rapid expansion of higher education supply in the last two decades has necessitated questioning this relationship in the Turkish context. Therefore, the effects of human capital and economic development factors on unemployment were analyzed using Principal Component Analysis (PCA), which allows for the reduction of multi-dimensional indicators into composite indices, and the ARDL bounds testing approach, which enables testing long-run relationships among series with different integration orders. Firstly, the Human Capital Index (HUMAN_PC), representing the human capital dimension, was constructed from the log difference of the net enrollment rate in higher education (D_LN_TER_ENR), life expectancy at birth (LN_LIFE_EXP), and the working-age population ratio (LN_POP1564) variables; the Economic Development Index (ECON_PC), representing the economic development dimension, was constructed from per capita income (LN_GDP_PC), industry share (LN_IND_SHARE), R&D expenditures as a share of GDP (LN_RD_GDP), and total labor force (LN_LAB_FORCE) variables using the PCA method. This approach eliminated the multicollinearity problem caused by high correlation and VIF values among variables, while also strengthening conceptual consistency. The results of the ADF unit root tests determined that no series was at the I(2) level, thus confirming the applicability of the ARDL bounds test. The analysis was conducted under two separate models: Model A, examining the relationship between the human capital index and unemployment, and Model B, testing the relationship between the economic development index and unemployment. The Bounds test results indicated no long-run cointegration relationship in Model A, but a long-run and statistically significant relationship in Model B. In Model A, the F-statistic (4.06) remained below the critical upper bound; thus, it was concluded that human capital investments do not permanently reduce unemployment in Türkiye. Conversely, in Model B, the F-statistic (10.53) was found to be above the critical values of both Pesaran et al. (2001) and Kripfganz and Schneider (2020), confirming a long-run relationship. The long-run coefficient was calculated as 0.3944, and the error correction term as -0.5808; this value indicates that increases in economic development generate a positive effect on the unemployment rate and that short-run disequilibrium is corrected over approximately a two-year period. The findings exhibit full consistency with the “jobless growth” phenomenon observed in the Turkish economy: while per capita income, industrial production, and R&D expenditures increase, the capacity to create employment remains limited; the growth process is characterized by capital and technology intensity rather than labor intensity. The absence of a long-run cointegration relationship for the human capital index can be explained by the insufficient expansion of labor demand despite the rapid increase in the number of higher education graduates, skill mismatch, regional inequalities, and the “educated unemployment” mechanism. This situation suggests that human capital accumulation can only transform into an employment-generating effect when integrated with an inclusive and innovation-based transformation of the production structure. Diagnostic tests (Breusch-Godfrey, Breusch-Pagan-Godfrey, White, Jarque-Bera, CUSUM, and CUSUMSQ) confirmed that both models satisfy the econometric assumptions and that the parameters are stable. The findings show that human capital investments alone are not sufficient to reduce unemployment; improvements in education, health, and demographics yield meaningful results only when aligned with the structure of economic development. At the policy level, a holistic approach is recommended, integrating human capital with labor demand and prioritizing demand-side employment policies over supply-side strategies. Within this scope, strengthening education-employment alignment, integrating vocational and technical training with the production structure, scaling up reskilling programs, regional employment incentives, and creating new job areas in green-digital sectors should be among the key priorities. Consequently, it is understood that the human capital hypothesis does not generally hold true in the Turkish context; unemployment dynamics are primarily determined by the composition of economic development, technological intensity, and institutional structure conditions. Therefore, when designing sustainable employment policies, Türkiye needs to treat human capital increase not merely as a supply-side policy tool, but as a strategic area integrated with the economic structure, innovation capacity, and the quality of production technology.