The paradox of assistance: Chinese aid in Africa and economic expectations

This paper examines how exposure to Chinese development projects shapes economic expectations in Africa. We link AidData's geocoded project records to 70,060 Afrobarometer interviews from 2005 to 2023 and identify effects using a shift–share instrumental variables design that interacts movements in Chinese construction-input costs with pre-sample allocation patterns. We find that living within 25 km of an active Chinese project lowers the probability that respondents expect national economic conditions to improve and, more modestly, reduces optimism about their own living conditions. Mechanism tests show that exposure raises perceived corruption—especially of national officials—and erodes trust in the presidency, parliament, and local councils; higher perceived corruption is monotonically associated with lower economic optimism. Results are robust across alternative instrument lags, distance thresholds, and spatial-spillover controls, and they do not generalize to placebo outcomes or to analogous estimates for World Bank projects. The evidence reconciles mixed views of Chinese aid by highlighting governance signals as the channel through which localized experiences translate into both macro- and micro-oriented expectations.