Investigating the persistence of infant mortality and the Matthew effect in post-Soviet countries: evidence from nonlinear and Fourier-based time-series methods.

This study investigates long-term trends in infant mortality rates (IMRs) in the case of 15 post-Soviet countries, with the aim of testing the Matthew Effect hypothesis, which suggests that inequality in health outcomes tend to persist or even widen over time. To assess these trends comprehensively, we employed various unit root tests, including traditional augmented Dickey-Fuller (ADF), non-linear (KSS, Kruse, Sollis, and Hepsag), and Fourier-based approaches (Fourier KPSS, Fourier ADF, flexible Fourier ADF, flexible fractional Fourier ADF, Fourier KSS, Fourier Kruse, and Fourier Sollis) to the data period from 1982 to 2022. Additionally, the multi-criteria decision analysis (MCDA) methods, such as equal weighting, entropy weighting, and TOPSIS, were utilized for a robustness check. The findings demonstrated that stationarity is validated in Armenia, Azerbaijan, Georgia, Kazakhstan, Kyrgyzstan, Lithuania, Latvia, Tajikistan, Turkmenistan, Ukraine, and Uzbekistan, thereby corroborating the Matthew Effect. Conversely, Belarus, Estonia, Moldova, and Russia exhibit non-stationary IMR series, indicating continued fluctuations or structural changes in their infant mortality patterns. The findings indicate that infant mortality has plateaued in several countries, pointing to stagnation in policy effectiveness. Conversely, some countries show dynamic change, possibly indicating responsive health systems. These insights provide important direction for country-specific health policy design to reduce IMR and health inequality more effectively.