Reproducibility package for Six Years Of High-Frequency Phone Surveys: Lessons From Sub-Saharan Africa

Phone surveys emerged as a critical data collection tool in low- and middle-income countries (LMICs) during the COVID-19 pandemic and have since become a permanent feature of the data landscape in many settings. This paper draws on the experience ofthe Living Standards Measurement Study High-Frequency Phone Surveys (LSMS-HFPS) — one of the longest-running and largest longitudinal phone survey initiatives in LMICs — to examine three methodological challenges central to their sustainable implementation: individual-level data collection and respondent selection, attrition, and the integration of phone and in-person surveys in mixed-mode data systems. The LSMS-HFPS initiative has provided continuous monitoring since April 2020 across six African countries, supporting over 120 phone survey rounds and more than 200,000 interviews alongside a parallel system of in-person surveys. We show that survey design choices can mitigate biases in individual-level data collection, though at the cost of interview length and logistical complexity. Attrition has been non-negligible but has not translated into significant sample bias. The paper discusses analytical and survey design strategies for addressing coverage bias and potential mode effects, and identify priorities for future research, including ex-post adjustment methods for mode effects, high-frequency welfare measurement through proxy-based imputation, and the cost-quality-feasibility trade-offs of mixed-mode survey design. The LSMS-HFPS experience demonstrates that the infrastructure, methods, and evidence base now exist to build genuinely integrated survey systems delivering reliable, timely, high-frequency data at substantially lower unit cost than in-person surveys alone.