Table 1 Characteristics of studies on the aggregate effect of structural adjustment programmes on child health
| Â | Outcome measure | IFI | Study period | Sampling | Study setting | Addresses selection bias | Analysis | Effect of structural adjustment |
|---|---|---|---|---|---|---|---|---|
Coburn, Restivo and Shandra, 2015 | Infant mortality | AfDB | 1990–2006 | 5-yearly | 31 African countries | Yes | Fixed-effects OLS regression | Detrimental |
Shandra, Shandra and London, 2012 | Infant mortality | IMF | 1990–2005 | 5-yearly | 32 African countries | Yes | Fixed-effects OLS regression | Detrimental |
Hajro and Joyce, 2009 | Infant mortality | IMF | 1985–2000 | 5-yearly | 82 developing countries | Yes | Fixed-effects OLS regression | Beneficial for non-concessionary loans in interaction with economic growth and no association for concessionary loans |
Shandra, Nobles, London and Williamson, 2004 | Infant mortality | IMF | 1980–1997 | Two-point | 59 developing countries | No | Lagged dependent panel regression | Detrimental in interaction with democracy |
Noorbakhsh and Noorbakhsh, 2006 | Infant mortality | World Bank | 1975–1998 | 5-yearly | 33 African countries | No | OLS regression | No association |
Oliver, 2006 | Infant mortality and under-5 mortality | IMF and World Bank | 1980–2000 | N/A | Argentina and Uruguay | No | Comparative case study | Detrimental |
Pandolfelli and Shandra, 2013 | Child mortality | AfDB | 1990–2005 | 5-yearly | 35 African countries | Yes | Random-effects generalised least-squares regression | Detrimental |
Shandra, Shandra and London, 2011 | Child mortality | World Bank | 1990–2005 | 5-yearly | 31 African countries | No | Fixed-effects OLS regression | Detrimental |
Pongou, Salomon and Ezzati, 2006 | Child malnutrition and maternal health-seeking behaviours | IMF and World Bank | 1991–1998 | Pooled two-point | 1587 and 1923 children in Cameroon | No | OLS regression | Detrimental |
Daoud, Nosrati, Reinsberg, Kentikelenis, Stubbs and King, 2017 | Child malnourishment | IMF | c2000 | Cross-sectional stratified | 67 developing countries, with 1, 941, 734 children | Yes | Multi-level logistic regression | Detrimental in interaction with parental education |