Heterogeneity effect of P4P
| Variableˆ | N | Difference-in-difference, short-term effect | Difference-in-difference, long-term effect |
| Beta* (p value) | Beta* (p value) | ||
| Outcome 1: facility-based delivery coverage | |||
| P4P effect × tercile 1 (poorest population)* | 8728 | 8.91 (0.009) | 7.84 (0.009) |
| P4P effect × tercile 2 (middle wealth)* | 8728 | 5.56 (0.065) | 0.82 (0.734) |
| P4P effect in public facility × tercile 1 (poorest)* | 8728 | 10.8 (0.007) | 11.2 (0.005) |
| P4P effect in public facility × tercile 2 (middle wealth)* | 8728 | 3.2 (0.365) | 1.19 (0.722) |
| P4P effect × public facility | 445 | 4.12 (0.405) | 1.50 (0.756) |
| P4P effect × dispensary facility | 445 | 1.63 (0.723) | 1.40 (0.717) |
| P4P effect × with available utilities | 450 | −2.99 (0.534) | 1.19 (0.774) |
| P4P effect × low availability of drugs& | 435 | 4.59 (0.317) | 2.49 (0.564) |
| P4P effect × lower baseline performer | 450 | 12.3 (0.005) | 4.42 (0.208) |
| P4P effect × rural facilities | 445 | 9.07 (0.046) | 7.48 (0.097) |
| Outcome 2: IPT2 coverage | |||
| P4P effect × tercile 1 (poorest population)* | 7362 | 0.61 (0.903) | 6.07 (0.228) |
| P4P effect × tercile 2 (middle wealth)* | 7362 | 5.05 (0.263) | 7.05 (0.112) |
| P4P effect × public facility | 445 | 3.62 (0.704) | 3.62 (0.679) |
| P4P effect × dispensary facility | 445 | −9.14 (0.137) | −6.20 (0.368) |
| P4P effect × with available utilities | 450 | −0.99 (0.875) | 9.10 (0.201) |
| P4P effect × low availability of drugs& | 435 | −1.75 (0.777) | −7.32 (0.283) |
| P4P effect × lower baseline performer | 450 | 7.62 (0.127) | 5.24 (0.389) |
| P4P effect × rural facilities | 445 | 4.64 (0.574) | −0.89 (0.925) |
*All specifications lead to an estimated beta showing percentage point change after controlling for a year dummy, facility-fixed effects and facility-level covariates (availability of utilities and wealth status of the catchment population).
&Availability of drugs include 37 drugs and vaccines, and analysis used a dummy variable classified in each arm separately based on baseline availability distribution (=1 for availability below the median/bottom half and 0, otherwise)
ˆreference category in brackets: public (vs non-public), dispensary (vs health centre and hospital), with electricity and water supply at baseline (vs none), baseline availability of drugs below the median/in bottom half (vs top half), baseline lower performer/below the median (vs higher performer), rural (vs urban district) and poorest/middle wealth (vs least poor).
*Data from household survey.
IPT, intermittent presumptive treatment; P4P, payment for performance.