Table 1.
IV estimates of primary schooling, reading, and basic education.
| Fertility | Child survival | Women's empowerment | Financial behaviors index | |
|---|---|---|---|---|
| Basic education, linear combination of primary schooling and reading (mean, RE meta- analysis weights) | −1.238 | 0.077 | 0.684 | 0.893 |
| Std. Error of mean coefficient | 0.088 | 0.007 | 0.084 | 0.128 |
| Absolute value, ratio mean to std. err. | 14.07 | 11.00 | 8.14 | 6.98 |
| 20th percentile of distribution of results | −2.542 | −0.002 | −0.01 | 0.354 |
| 80th percentile of distribution of results | −0.158 | 0.179 | 1.922 | 1.81 |
| Ratio of IV estimate of basic education to OLS estimate of primary schooling (without controlling for reading) | 3.76 | 3.35 | 4.68 | 3.20 |
| Ratio of IV estimate of basic education to OLS estimate of basic education | 3.32 | 2.85 | 3.54 | 1.78 |
| Schooling (mean, RE meta-analysis weights) | −0.541 | 0.029 | 0.117 | 0.437 |
| Std. Error of mean coefficient | 0.121 | 0.01 | 0.108 | 0.093 |
| Absolute value, ratio mean to std. err. | 4.47 | 2.90 | 1.08 | 4.70 |
| 20th percentile of distribution of results | −2.359 | −0.078 | −0.804 | −0.686 |
| 80th percentile of distribution of results | 0.967 | 0.154 | 1.451 | 0.758 |
| Reading (mean, RE meta-analysis weights) | −0.566 | 0.032 | 0.538 | 0.368 |
| Std. Error of mean coefficient | 0.119 | 0.011 | 0.128 | 0.158 |
| Absolute value, ratio mean to std. err. | 4.76 | 2.91 | 4.20 | 2.33 |
| 20th percentile of distribution of results | −2.27 | −0.078 | −1.004 | −0.396 |
| 80th percentile of distribution of results | 0.905 | 0.22 | 1.525 | 2.416 |
| Ratio of reading coefficient to primary schooling | 1.05 | 1.10 | 4.60 | 0.84 |
| Number of survey rounds | 128 | 128 | 67 | 10 |
Note: Regressions contain controls for age, age squared, age cubed, asset index, a rural/urban dummy, and dummies for regions. Schooling coefficients have been scaled to reflect primary school (six years) completion; reading coefficients are scaled to reflect going from no reading to reading without help.