<![CDATA[Entrepreneurial activity is widely regarded as a central driver of economic development through its contribution to innovation, employment creation, and productivity growth. This study examines the long-run and short-run relationships between internet usage, human capital, gross domestic product (GDP) per capita, unemployment, and self-employment as a proxy for entrepreneurial activity in Indonesia and the United Kingdom, within the push–pull entrepreneurship framework. Annual time-series data covering 1994–2023 for Indonesia and 2001–2023 for the United Kingdom are analyzed using country-specific Autoregressive Distributed Lag (ARDL) bounds testing models. The results reveal pronounced heterogeneity in the long-run determinants of self-employment across the two economies. In Indonesia, internet usage is positively and significantly associated with self-employment in the long run (β = 0.223, p < 0.01), whereas human capital (β = −0.493, p < 0.10) and GDP per capita (β = −0.006, p < 0.01) are negatively associated with self-employment, with a rapid error-correction adjustment of 96.7 percent per period. In the United Kingdom, unemployment is the only variable exhibiting a statistically significant long-run association with self-employment (β = 0.892, p < 0.10), while internet usage, human capital growth, and GDP per capita show no significant long-run relationship; instead, internet usage, human capital growth, and unemployment exert significant short-run effects, with a comparatively gradual adjustment of 31.7 percent per period. The novelty of this study lies in providing the first direct, country-specific ARDL comparison of a developing and a developed economy within a unified push–pull framework, demonstrating that identical macroeconomic factors operate through opposing directions of association and fundamentally different temporal horizons depending on a country's stage of economic development. Given the observational nature of the study and the use of self-employment as a proxy for entrepreneurial activity, these findings should be interpreted as macro-level associations rather than definitive causal effects.]]>
Copyrights © 2026
