The effects of government-led cash transfer programmes on behavioural and health determinants of mortality: a difference-in-differences study.

BACKGROUND: Poverty is strongly associated with numerous adverse health outcomes. Government-led cash transfer programmes are crucial to poverty reduction strategies in many low-income and middle-income countries (LMICs). Although extensive research from individual programmes exists on the effects of cash transfers on beneficiaries, evidence of these programmes' population-wide health effects remains scarce. Previously, we showed that cash transfer programmes are associated with substantially reduced mortality rates among women and young children at the population level in LMICs. In this study, we aimed to explore the mechanisms underlying these reductions. METHODS: In this two-stage difference-in-differences study, we combined individual-level data from Demographic and Health Surveys in 37 LMICs with a comprehensive database of government-led cash transfer programmes between 2000 to 2019 to compare countries with cash transfer programmes and countries without these programmes before and after the introduction of them. We evaluated the effects of cash transfer programmes on 17 outcomes related to maternal health service use, fertility and reproductive decision making, caregiver health behaviours, and child health and nutrition. Outcomes were assessed across all respondents of the surveys but could not be differentiated between recipients and non-recipients for countries with cash-transfer programmes. FINDINGS: Among 37 countries included in the study, 20 introduced large-scale cash transfer programmes during the study period. We included data from 2 156 464 livebirths, and 954 202 children younger than 5 years. We identified statistically significant effects of cash transfer programmes on early antenatal care (5·0 percentage points, 95% CI 2·1 to 7·9; padjusted=0·0019), facility deliveries (7·3 percentage points, 3·2 to 11·3; padjusted=0·0014), delivery by a skilled birth attendant (7·9 percentage points, 3·2 to 12·6; padjusted=0·0027), desired pregnancies (1·9 percentage points, 0·5 to 3·2; padjusted=0·014), interdelivery intervals (2·5 months, 1·8 to 3·1; padjusted=0·0017), unmet needs for contraception (-10·3 percentage points, -15·2 to -5·3; padjusted=0·0006), exclusive breastfeeding (14·4 percentage points, 13·3 to 15·5; padjusted=0·0004), minimum acceptable diets (7·5 percentage points, 5·5-9·5; padjusted=0·0009), measles vaccinations (5·3 percentage points, 1·6 to 8·9; padjusted=0·026), male twin livebirths (0·8 per 1000 male livebirths, 0·3 to 1·4; padjusted=0·023), diarrhoeal illness (-6·4 percentage points, -11·7 to -1·1; padjusted=0·038), and underweight nutritional status (-2·0 percentage points, -3·6 to -0·4; padjusted=0·029). There were no statistically significant effects on age at first birth (1·6 months, -1·3 to 4·4; padjusted=0·48), intended pregnancies (-0·2 percentage points, -2·8 to 2·3; padjusted=0·86), small birth sizes (0·4 percentage points, -0·7 to 1·4; padjusted=0·53), wasting (-2·1 percentage points, -5·0 to 0·9; padjusted=0·17), and stunting (4·3 percentage points, -0·2 to 8·7; padjusted=0·10). INTERPRETATION: As many countries consider the future of their cash transfer programmes, including whether to embrace approaches such as basic or guaranteed incomes, these findings provide new evidence on the numerous ways in which such programmes can improve population health. FUNDING: US National Institutes of Health.