Abstract：
BACKGROUND: Universal Health Coverage (UHC) is a widely accepted objective among entities providing development assistance for health (DAH) and DAH recipient governments. One key metric to assess progress with UHC is financial risk protection, but empirical evidence on the extent to which DAH is associated to financial risk protection (and hence UHC) is scarce.
METHODS: Our sample is comprised of 65 countries whose DAH per capita is above the population -weighted average DAH per capita across all countries. The sample comprises of 1.7 million household observations, for the period 2000-2016. We run country and year fixed effects regressions, and pseudo-panel models, to assess the association between DAH and three measures of financial risk protection: catastrophic health expenditure (i.e., out-of-pocket health expenditures larger than 10% of total household expenditures [\'CHE10%\']), out-of-pocket health expenditure as a share of total expenditure (\'OOP%\'), and impoverishment due to health expenditures, at the 1.90US$ per day poverty line (\'IMP190\').
RESULTS: on average, DAH investment does not appear to be significantly associated with financial risk protection outcomes. However, we find suggestive evidence that a 1 US$ increase in DAH per capita is negatively associated (i.e., an improvement) with at least one financial risk protection outcome for the poorest household quintile within countries (in fixed effects models, IMP190: 0.05 percentage points, p < 0.1; in pseudo-panel models, CHE10%: 0.12 percentage points, p < 0.01). DAH is also negatively associated (i.e., an improvement) with most financial risk protection outcomes when it is largely channelled via government systems (i.e., when it is \"on-budget\") (CHE10%: 0.68 percentage points, p < 0.05). Several robustness checks confirm these results.
CONCLUSIONS: DAH investments require careful planning to improve financial risk protection. For example, positive DAH effects for the poorest quintiles of the population might be driven by DAH targeting poorer populations and doing so effectively. Our results also suggest that channelling more resources via governments might be a promising avenue to enhance the impact of DAH on financial risk protection.
METHODS: Our sample is comprised of 65 countries whose DAH per capita is above the population -weighted average DAH per capita across all countries. The sample comprises of 1.7 million household observations, for the period 2000-2016. We run country and year fixed effects regressions, and pseudo-panel models, to assess the association between DAH and three measures of financial risk protection: catastrophic health expenditure (i.e., out-of-pocket health expenditures larger than 10% of total household expenditures [\'CHE10%\']), out-of-pocket health expenditure as a share of total expenditure (\'OOP%\'), and impoverishment due to health expenditures, at the 1.90US$ per day poverty line (\'IMP190\').
RESULTS: on average, DAH investment does not appear to be significantly associated with financial risk protection outcomes. However, we find suggestive evidence that a 1 US$ increase in DAH per capita is negatively associated (i.e., an improvement) with at least one financial risk protection outcome for the poorest household quintile within countries (in fixed effects models, IMP190: 0.05 percentage points, p < 0.1; in pseudo-panel models, CHE10%: 0.12 percentage points, p < 0.01). DAH is also negatively associated (i.e., an improvement) with most financial risk protection outcomes when it is largely channelled via government systems (i.e., when it is \"on-budget\") (CHE10%: 0.68 percentage points, p < 0.05). Several robustness checks confirm these results.
CONCLUSIONS: DAH investments require careful planning to improve financial risk protection. For example, positive DAH effects for the poorest quintiles of the population might be driven by DAH targeting poorer populations and doing so effectively. Our results also suggest that channelling more resources via governments might be a promising avenue to enhance the impact of DAH on financial risk protection.
摘要：
背景：全民健康覆盖（UHC）是为卫生（DAH）和DAH受援国政府提供发展援助的实体之间广泛接受的目标。评估UHC进展的一个关键指标是金融风险保护，但关于DAH与金融风险保护(以及UHC)相关程度的经验证据很少。
方法：我们的样本包括65个国家，这些国家的人均DAH高于所有国家的人口加权平均人均DAH。样本包括170万个家庭观察，2000-2016年期间。我们运行国家和年份固定效应回归，和伪面板模型，评估DAH与三项财务风险保护措施之间的关联：灾难性卫生支出(即，自付医疗支出超过家庭总支出的10%[\'CHE10%\'])，自付医疗支出占总支出的比例(“OOP%”)，以及医疗支出导致的贫困，在每天1.90美元的贫困线(“IMP190”)。
结果：平均，DAH投资似乎与金融风险保护结果没有显着关联。然而,我们发现暗示性证据表明，人均DAH增加1美元是负相关的(即，改进)对国家内最贫穷的五分之一家庭至少有一个金融风险保护结果(在固定效应模型中，IMP190：0.05个百分点，p<0.1；在伪面板模型中，CHE10%:0.12个百分点,p<0.01)。DAH也是负相关的(即，一种改进)，当它主要通过政府系统引导时，大多数金融风险保护结果(即，当它是“预算内”时)(10%：0.68个百分点，p<0.05)。一些健壮性检查证实了这些结果。
结论：DAH投资需要仔细规划，以提高金融风险保护。例如,DAH对最贫穷的五分之一人口的积极影响可能是由于DAH针对较贫穷的人口并有效地做到了这一点。我们的结果还表明，通过政府提供更多资源可能是增强DAH对金融风险保护影响的有希望的途径。
方法：我们的样本包括65个国家，这些国家的人均DAH高于所有国家的人口加权平均人均DAH。样本包括170万个家庭观察，2000-2016年期间。我们运行国家和年份固定效应回归，和伪面板模型，评估DAH与三项财务风险保护措施之间的关联：灾难性卫生支出(即，自付医疗支出超过家庭总支出的10%[\'CHE10%\'])，自付医疗支出占总支出的比例(“OOP%”)，以及医疗支出导致的贫困，在每天1.90美元的贫困线(“IMP190”)。
结果：平均，DAH投资似乎与金融风险保护结果没有显着关联。然而,我们发现暗示性证据表明，人均DAH增加1美元是负相关的(即，改进)对国家内最贫穷的五分之一家庭至少有一个金融风险保护结果(在固定效应模型中，IMP190：0.05个百分点，p<0.1；在伪面板模型中，CHE10%:0.12个百分点,p<0.01)。DAH也是负相关的(即，一种改进)，当它主要通过政府系统引导时，大多数金融风险保护结果(即，当它是“预算内”时)(10%：0.68个百分点，p<0.05)。一些健壮性检查证实了这些结果。
结论：DAH投资需要仔细规划，以提高金融风险保护。例如,DAH对最贫穷的五分之一人口的积极影响可能是由于DAH针对较贫穷的人口并有效地做到了这一点。我们的结果还表明，通过政府提供更多资源可能是增强DAH对金融风险保护影响的有希望的途径。
