Perform forward stepwise-selection on all variables included in Robert’s canonical specification (“pre_coverage”, “rGDPg2020”, “digIDcov”, “dig_reg”, “fsi”, “eap”, “log_deaths”). At each iteration, the variable which would have the the smallest p-value if added to the model is added. I add in the few variables which are not statistically significant all at once.
Note that the final regression on the right weights each observation by the country population.
##
## Stepwise Selection Summary
## -----------------------------------------------------------------------------------------
## Added/ Adj.
## Step Variable Removed R-Square R-Square C(p) AIC RMSE
## -----------------------------------------------------------------------------------------
## 1 pre_coverage addition 0.505 0.499 52.0620 -8.9381 0.2239
## 2 dig_reg addition 0.644 0.635 17.2740 -34.2834 0.1910
## 3 eap addition 0.675 0.663 11.0240 -39.8843 0.1836
## 4 log_deaths addition 0.711 0.696 3.5080 -47.6148 0.1743
## -----------------------------------------------------------------------------------------
| Dependent variable: | |||||||
| coverage | |||||||
| (1) | (2) | (3) | (4) | (5) | (6) | (7) | |
| pre_coverage | 1.147*** | 0.849*** | 0.731*** | 0.673*** | 0.637*** | 0.620*** | 0.635*** |
| (0.117) | (0.138) | (0.134) | (0.121) | (0.119) | (0.120) | (0.068) | |
| rGDPg2020 | -0.005 | -0.005 | -0.005 | ||||
| (0.004) | (0.004) | (0.003) | |||||
| dig_id | 0.069 | 0.078 | -0.193** | ||||
| (0.078) | (0.079) | (0.075) | |||||
| dig_reg | 0.262*** | 0.267*** | 0.204*** | 0.195*** | 0.193*** | 0.301*** | |
| (0.057) | (0.054) | (0.059) | (0.063) | (0.062) | (0.041) | ||
| fsi | 0.006 | 0.006 | 0.007 | ||||
| (0.017) | (0.017) | (0.020) | |||||
| eap | 0.164** | 0.279*** | 0.277*** | 0.295*** | 0.279*** | ||
| (0.082) | (0.077) | (0.091) | (0.092) | (0.048) | |||
| log_deaths | 0.040*** | 0.035** | 0.036** | 0.058*** | |||
| (0.013) | (0.017) | (0.017) | (0.017) | ||||
| pop_2018 | -0.000 | ||||||
| (0.000) | |||||||
| Constant | 0.131*** | 0.055** | 0.046** | -0.070* | -0.144 | -0.147 | -0.134 |
| (0.028) | (0.022) | (0.019) | (0.037) | (0.174) | (0.176) | (0.224) | |
| Observations | 83 | 83 | 83 | 83 | 82 | 82 | 82 |
| R2 | 0.505 | 0.644 | 0.675 | 0.711 | 0.717 | 0.720 | 0.893 |
| Adjusted R2 | 0.499 | 0.635 | 0.663 | 0.696 | 0.690 | 0.689 | 0.883 |
| Residual Std. Error | 0.224 (df = 81) | 0.191 (df = 80) | 0.184 (df = 79) | 0.174 (df = 78) | 0.176 (df = 74) | 0.176 (df = 73) | 771.234 (df = 74) |
| F Statistic | 82.746*** (df = 1; 81) | 72.416*** (df = 2; 80) | 54.771*** (df = 3; 79) | 48.029*** (df = 4; 78) | 26.751*** (df = 7; 74) | 23.477*** (df = 8; 73) | 88.067*** (df = 7; 74) |
| Note: | p<0.1; p<0.05; p<0.01 | ||||||
The following regressions include only one of the variables digital ID coverage, digital ID registration, and FSI each.
| Dependent variable: | |||
| coverage | |||
| (1) | (2) | (3) | |
| pre_coverage | 0.739*** | 0.641*** | 0.735*** |
| (0.115) | (0.112) | (0.121) | |
| rGDPg2020 | -0.004 | -0.005 | -0.004 |
| (0.004) | (0.004) | (0.005) | |
| eap | 0.318*** | 0.268*** | 0.287*** |
| (0.073) | (0.076) | (0.078) | |
| log_deaths | 0.052*** | 0.035** | 0.052*** |
| (0.013) | (0.015) | (0.015) | |
| dig_id | 0.139** | ||
| (0.057) | |||
| dig_reg | 0.205*** | ||
| (0.059) | |||
| fsi | -0.019 | ||
| (0.016) | |||
| Constant | -0.141*** | -0.063* | 0.086 |
| (0.038) | (0.038) | (0.159) | |
| Observations | 83 | 83 | 82 |
| R2 | 0.667 | 0.717 | 0.653 |
| Adjusted R2 | 0.646 | 0.699 | 0.630 |
| Residual Std. Error | 0.188 (df = 77) | 0.174 (df = 77) | 0.192 (df = 76) |
| F Statistic | 30.898*** (df = 5; 77) | 39.021*** (df = 5; 77) | 28.559*** (df = 5; 76) |
| Note: | p<0.1; p<0.05; p<0.01 | ||
The following regressions only include either GDP growth and deaths but not both.
| Dependent variable: | |||
| coverage | |||
| (1) | (2) | (3) | |
| pre_coverage | 0.668*** | 0.673*** | 0.735*** |
| (0.127) | (0.121) | (0.121) | |
| dig_reg | 0.257*** | 0.204*** | |
| (0.051) | (0.059) | ||
| eap | 0.169** | 0.279*** | 0.287*** |
| (0.084) | (0.077) | (0.078) | |
| rGDPg2020 | -0.008* | -0.004 | |
| (0.004) | (0.005) | ||
| log_deaths | 0.040*** | 0.052*** | |
| (0.013) | (0.015) | ||
| fsi | -0.019 | ||
| (0.016) | |||
| Constant | 0.034* | -0.070* | 0.086 |
| (0.019) | (0.037) | (0.159) | |
| Observations | 83 | 83 | 82 |
| R2 | 0.692 | 0.711 | 0.653 |
| Adjusted R2 | 0.676 | 0.696 | 0.630 |
| Residual Std. Error | 0.180 (df = 78) | 0.174 (df = 78) | 0.192 (df = 76) |
| F Statistic | 43.793*** (df = 4; 78) | 48.029*** (df = 4; 78) | 28.559*** (df = 5; 76) |
| Note: | p<0.1; p<0.05; p<0.01 | ||
The first figure below shows predicted versus actual coverage. I wasn’t able to see any pattern in the outliers but I could be missing something. The next few graphs are partial adjustment graphs which are sometimes useful for checking if there are any important non-linearities that we should be taking into account. (There don’t seem to be any.)
Perform stepwise forward selection to select independent variables in a regression of spending on other vars. Do the same thing with spending / coverage.
##
## Stepwise Selection Summary
## ------------------------------------------------------------------------------------------
## Added/ Adj.
## Step Variable Removed R-Square R-Square C(p) AIC RMSE
## ------------------------------------------------------------------------------------------
## 1 pre_coverage addition 0.198 0.188 2.3480 -550.7298 0.0072
## 2 dig_reg addition 0.235 0.215 0.7880 -552.4523 0.0071
## 3 pre_spending addition 0.331 0.300 -0.6580 -484.0511 0.0069
## ------------------------------------------------------------------------------------------
##
## Stepwise Selection Summary
## -------------------------------------------------------------------------------------
## Added/ Adj.
## Step Variable Removed R-Square R-Square C(p) AIC RMSE
## -------------------------------------------------------------------------------------
## 1 dig_reg addition 0.054 0.040 3.8880 -341.5569 0.0198
## -------------------------------------------------------------------------------------
| Dependent variable: | |||
| spending | spending_pc | ||
| (1) | (2) | (3) | |
| pre_coverage | 0.011** | -0.019 | |
| (0.005) | (0.011) | ||
| dig_reg | 0.005*** | -0.009* | -0.003 |
| (0.002) | (0.005) | (0.006) | |
| pre_spending | 0.002 | 0.001 | |
| (0.002) | (0.002) | ||
| expansion | -0.006 | ||
| (0.011) | |||
| Constant | 0.0003 | 0.028*** | 0.028*** |
| (0.001) | (0.005) | (0.006) | |
| Observations | 69 | 69 | 60 |
| R2 | 0.331 | 0.054 | 0.056 |
| Adjusted R2 | 0.300 | 0.040 | -0.013 |
| Residual Std. Error | 0.007 (df = 65) | 0.020 (df = 67) | 0.019 (df = 55) |
| F Statistic | 10.728*** (df = 3; 65) | 3.800* (df = 1; 67) | 0.812 (df = 4; 55) |
| Note: | p<0.1; p<0.05; p<0.01 | ||