38のOECD諸国におけるCOVID-19データの実証分析：接触追跡を成功させる要因は何か

Despite widespread adoption of contact tracing during the COVID-19 pandemic, countries exhibited marked differences in tracing performance. Existing studies have documented policy variation but lack standardized, comparable outcome indicators of tracing effectiveness. Recent reviews emphasize the need for quantitative, cross-national, outcome-based measures and broader frameworks that account for contextual factors such as governance, education, and culture. To address this gap, this study aims to validate the time lag (τ) between peaks in confirmed cases and testing as a proxy for tracing responsiveness and examines how government policies, socioeconomic conditions, and cultural dimensions jointly shape τ across OECD countries.

Panel data for 38 OECD countries (Jan 2020-Jun 2022) were analyzed. τ was computed through rolling cross-correlation between daily cases and tests. We estimated fixed-effects and pooled OLS models to identify policy and contextual effects and ran mixed-effects models as a robustness extension. Additional robustness tests included two-way fixed effects, exclusion of high-incidence periods, and winsorization of τ, alongside variance-inflation and residual diagnostics, to ensure results were not driven by outliers or specification bias.

Stricter containment and stronger tracing policies were associated with longer τ, while broader testing and higher epidemic severity shortened it. Among contextual variables, higher education lengthened τ, whereas greater urbanization reduced it. Cultural dimensions showed smaller, less stable effects. Despite modest explanatory power (R² ≈ 0.08-0.11), results were consistent across alternative specifications and model types.

The case-test peak lag (τ) provides a practical, comparable metric of tracing performance across countries. However, our findings suggest that its utility as an indicator of tracing responsiveness is context-dependent and should be interpreted with caution. Effective tracing depends on coherent policy coordination supported by socioeconomic and institutional environments that foster technological participation and compliance. Limitations include τ's indirect nature and remaining unobserved heterogeneity.