Globalization Fails Young Women in South Africa as Jobless Growth Deepens

South Africa has spent more than three decades opening itself to the world, signing trade agreements, welcoming foreign capital, and plugging into global networks of information and ideas. Yet for young women, the promise of globalization has largely failed to materialize. A new econometric study of data spanning 1990 to 2022 finds that, in the long run, international integration has actually raised the female youth unemployment rate rather than lowering it, with the damage concentrated in the very dimensions of globalization most often celebrated as engines of opportunity. The research, published in Discover Global Society by economist John Bosco Nnyanzi of Makerere University, uses the autoregressive distributed lag (ARDL) bounds testing approach to disentangle short-term from long-term effects across nine distinct globalization indices, and its findings challenge the comfortable assumption that openness alone can solve one of the world’s most stubborn unemployment crises.

The scale of the problem is difficult to overstate. In the first quarter of 2025, the NEET rate, the share of young people not in employment, education or training, stood at 37.1 percent for South Africans aged 15 to 24, with young women at 37.5 percent slightly worse affected than young men at 36.7 percent. In the broader 15 to 34 age group, the rate climbed to 45.1 percent, with 48.1 percent of women affected compared with 42.2 percent of men. These figures are barely improved from 2013, when 37 percent of female youth were outside work and education against 30 percent of males, despite government interventions including the Expanded Public Works Program, the National Youth Development Program, and the Employment Equity Act. Against this backdrop, Nnyanzi set out to test whether globalization, in its many forms, has been part of the solution or part of the problem.

The study’s methodological core is the ARDL framework developed by Pesaran, Shin and Smith, a technique prized for its flexibility with small samples and mixed orders of integration, meaning it can handle variables that are stationary, non-stationary, or a combination of both. By estimating both short-run dynamics and long-run equilibrium relationships within a single model, and by validating the results with unit root tests, bounds cointegration tests, and a battery of diagnostics including Breusch-Godfrey, White, and Ramsey RESET tests alongside CUSUM stability checks, the analysis provides unusually robust evidence for a single-country time series. The dependent variable is the female youth unemployment rate, defined as the share of the female labour force aged 15 to 24 without work but available for and seeking employment, which averaged a striking 51.52 percent over the study period, ranging from 44.77 to 69.79 percent.

Globalization itself was measured using the KOF index, disaggregated into economic, social, political, informational, interpersonal, cultural, financial, and trade dimensions, each calculated as the average of de facto and de jure components. This disaggregation is the study’s central innovation, because aggregate indices can mask sharply divergent effects. The model also controlled for foreign direct investment inflows, net migration, output volatility measured as the standard deviation of growth rates, fertility rates, and female education captured by the gender parity index, guarding against omitted variable bias and spurious correlation. Variance inflation factors confirmed that multicollinearity was not a problem, with no individual VIF exceeding 5.

The long-run results are sobering. A 1 percent increase in the overall globalization index was associated with a 0.26 percent increase in female youth unemployment, holding other factors constant. Disaggregated further, trade globalization exerted the largest effect, with a 1 percent increase linked to a 0.54 percent rise in female youth joblessness, followed by informational globalization at 0.50 percent, social globalization at 0.45 percent, interpersonal globalization at 0.34 percent, economic globalization at 0.29 percent, and political globalization at 0.22 percent. Only financial and cultural globalization broke the pattern, showing no significant long-run harm. These findings align with the skeptical school of globalization theory, which holds that the benefits of integration are overstated and unevenly distributed, and they echo prior evidence that South Africa’s rapid tariff reductions exposed labour-intensive manufacturing to cheaper imports without adequate support or reskilling for displaced workers.

The short-run picture, however, is strikingly different, and this temporal asymmetry is perhaps the study’s most policy-relevant insight. Financial globalization at its first lag reduced female youth unemployment by 0.08 percent for each 1 percent increase, plausibly by easing financial vulnerabilities and promoting inclusion. Interpersonal globalization, reflecting the growing interconnectedness of people through communication, travel, and migration, cut female youth unemployment by 0.22 percent per 1 percent increase. Political globalization showed the largest short-run benefit, with a 100 percent increase associated with a 19 percent reduction in female youth joblessness. These short-run gains support the hyperglobalist view that integration creates opportunities that would not otherwise exist, particularly for women seeking entry into formal labour markets.

The error correction models revealed that between 14 and 25 percent of any deviation from long-run equilibrium unemployment is corrected within a single year, indicating a reasonably fast adjustment process. But the study went further, testing for non-linearity by introducing squared globalization terms. The results uncovered a U-shaped relationship in the long run for the overall index as well as the social, informational, cultural, and political dimensions, meaning that globalization reduced female youth unemployment up to an optimal threshold, beyond which the effect turned harmful. The calculated thresholds were 40.37 for overall globalization, 38.55 for social, 39.08 for informational, and 45.88 for political globalization. Comparing these with 2022 values shows that every dimension except cultural globalization has already surpassed its threshold, suggesting South Africa has exhausted the employment benefits of these forms of integration and may now be experiencing their detrimental side, possibly through automation and competitive displacement of female workers.

The control variables added further nuance. Net migration and female education both increased female youth unemployment, the latter perhaps reflecting higher reservation wages among educated women, skills mismatches with available jobs, and rigid social norms that leave many qualified young women searching indefinitely. Fertility rates showed a negative association, likely because young motherhood curtails formal education and pushes women into the informal economy, where underemployment goes uncounted in official statistics. Foreign direct investment, contrary to expectations, raised long-run female youth unemployment, possibly due to capital-intensive investment patterns and the crowding out of female-dominated informal enterprises. GDP volatility increased unemployment in the short run, consistent with cyclical layoffs, but reduced it in the long run, a pattern the study attributes to creative destruction, as recessions eliminate inefficient firms and reallocate labour toward more productive ones, while women’s concentration in counter-cyclical sectors such as healthcare, education, and public administration shields them from the worst hiring-and-firing cycles.

The policy implications are pointed. Nnyanzi recommends embedding gender impact assessments into trade agreements so that liberalization does not devalue women’s labour, supporting women-owned enterprises through tax incentives and access to finance, and closing the gender digital divide by making internet connectivity and smart devices affordable for women, particularly in rural areas. He also calls for incentives encouraging women to use e-commerce and fintech platforms to reach global markets, alongside gender-sensitive strategies for cultural exchange and mobility. Because interpersonal, political, and financial globalization deliver short-run gains, the study argues that governments should deepen capital flow liberalization, strengthen international institutions, and foster information exchange, while remaining alert to the long-run thresholds at which these benefits reverse. The overarching message is that a one-size-fits-all approach to globalization policy will not stabilize South Africa’s labour market; only gender-responsive, dimension-specific, and time-aware interventions can convert global integration from a driver of female youth joblessness into a genuine engine of inclusive employment.

Subject of Research: The effects of globalization and its sub-dimensions on female youth unemployment in South Africa

Article Title: Exploring the female youth unemployment effects of globalization in South Africa using the standard ARDL approach

Article References: Nnyanzi, J. B. (2026). Exploring the female youth unemployment effects of globalization in South Africa using the standard ARDL approach. Discover Global Society, 4(1), Article 223. https://doi.org/10.1007/s44282-026-00571-8

Image Credits: AI Generated

DOI: 10.1007/s44282-026-00571-8

Keywords: globalization, female youth unemployment, South Africa, ARDL, KOF index, trade liberalization, NEET, threshold effects, labour market, gender inequality, foreign direct investment, economic policy