Summary Statistics for Ecuador-PES.

*p<0.1;
**p<0.05;
***p<0.01
Standard deviations in parentheses. T-tests test for differences in means assuming unequal variances. “Before matching” uses the full sample; “After matching” is based on 1-to-1 propensity score matching without replacement and limiting the maximum distance between matches with a caliper. Normalized differences in means are calculated as recommended by [5]. There is no test for statistical significance associated with normalized differences in means but a rule of thumb is that sizes larger than 0.25 can bias simple ordinary least squares regression.
Summary Statistics for Ecuador-PES.

* 代表p < 0.1；** 代表p < 0.05；*** 代表p < 0.01。括号内数值为标准差。本研究采用方差不等的t检验进行均值差异显著性检验。「匹配前」分析使用全样本；「匹配后」分析则基于无放回的1:1倾向得分匹配（propensity score matching），并通过卡尺（caliper）限定匹配项间的最大距离。均值标准化差异按照文献[5]的建议计算。目前尚无针对均值标准化差异的统计显著性检验方法，但经验法则表明，当标准化差异大于0.25时，会对普通最小二乘回归的结果产生偏误。厄瓜多尔-PES（Ecuador-PES）的汇总统计量。