Table A4.
Summary statistics for 2008, by later P4P status
| Ever PcV P4P | Never PcV P4P | p‐value of | |||
|---|---|---|---|---|---|
| Variable | Mean | SD | Mean | SD | difference |
| share children (%) | 11.6 | (1.5) | 10.6 | (1.1) | .072 |
| share elderly (%) | 17.0 | (3.3) | 18.5 | (3.2) | .178 |
| log(population) | 11.3 | (1.3) | 10.8 | (1.3) | .043 |
| mean income | 250.9 | (40.0) | 230.0 | (15.8) | .191 |
| share secondary edu (%) | 42.8 | (6.4) | 46.3 | (5.4) | .056 |
| share tertiary edu (%) | 33.3 | (10.6) | 28.5 | (9.1) | .074 |
| other P4P a | 0.05 | (0.226) | 0 | (0) | .001 |
| choicereform a | 0.42 | (0.50) | 0.05 | (0.23) | .330 |
| cost responsibility a | 0.15 | (0.35) | 0.80 | (0.40) | .000 |
| Observations | 108 | 181 | |||
County‐council level variable. Summary statistics for 2008, separate for the municipalities in counties that did ("ever") or did not ("never") introduce P4P tied to the PcV share during the period 2009–2013. PcV share = penicillin V prescriptions' share of total consumption of antibiotics prescribed for respiratory tract infections (J01CE02/J01AA02/J01CA04/J01CR02/J01DB‐DE/J01FA), children 0–6 years. Column p‐value of difference shows p‐values from regressions of each variable on a dummy for PcV P4P county council, that is, the p‐values of the difference between the treated group (Ever PcV P4P) and the control group (Never PcV P4P). The regressions are weighted by the square root of the population size and standard errors are clustered at the county council level. The difference between P4P and control counties is not statistically significant when contrasted in t‐tests at the county level (n = 21). PcV = penicillin V; P4P = pay‐for‐performance.