Derivation of preferred specifications for public health expenditure, 2013/2014
| Variables | (1) All causes 2013/2014 PH spend SYLLR 2013/2014/2015 Outcome model Instrument PH spend Weighted IV second stage Full specification | (2) All causes 2013/2014 PH spend SYLLR 2013/14/15 Outcome model Instrument PH spend Weighted IV second stage Derived specification | (3) All causes 2013/2014 PH & PB spend SYLLR 2013/14/15 Outcome model Instrument PH&PB spend Weighted IV second stage Full specification Backward selection | (4) All causes 2013/2014 PH & PB spend SYLLR 2013/14/15 Outcome model Instrument PH&PB spend Weighted IV second stage Derived specification Backward selection | (5) All causes 2013/2014 PH & PB spend SYLLR 2013/14/15 Outcome model Instrument PH&PB spend Weighted IV second stage Initial specification Forward selection | (6) All causes 2013/2014 PH & PB spend SYLLR 2013/14/15 Outcome model Instrument PH&PB spend Weighted IV second stage Derived specification Forward selection |
| Public health spend per person | −0.084** (0.041) | −0.115** (0.048) | −0.024 (0.037) | −0.081** (0.034) | −0.006 (0.025) | −0.144*** (0.040) |
| Healthcare spend per person | −0.551 (0.413) | −0.672*** (0.233) | −1.012*** (0.244) | −0.837*** (0.269) | ||
| IMD 2010 | 0.203*** (0.075) | −0.505*** (0.157) | 0.253*** (0.062) | 0.221*** (0.063) | ||
| Proportion of all residents born outside the EU | −0.016 (0.018) | −0.043* (0.024) | −0.084*** (0.019) | −0.070*** (0.019) | ||
| Proportion of population in white ethnic group | 0.246*** (0.060) | 0.226*** (0.051) | ||||
| Proportion of population providing unpaid care | −0.439*** (0.167) | −0.231** (0.091) | −0.399*** (0.144) | −0.479*** (0.096) | −0.547*** (0.122) | |
| Proportion of population aged 16–74 with no qualifications | −0.034 (0.112) | −0.111 (0.105) | ||||
| Proportion of households without a car | −0.062 (0.072) | −0.033 (0.087) | ||||
| Proportion of households that are owner occupied | 0.129* (0.071) | 0.090 (0.075) | ||||
| Proportion of households that are one pensioner households | −0.082 (0.084) | −0.023 (0.079) | ||||
| Lone parent households with dependent children | 0.056 (0.060) | −0.048 (0.082) | ||||
| Proportion of population aged 16–74 that are permanently sick | 0.315*** (0.070) | 0.475*** (0.068) | 0.237*** (0.068) | 1.187*** (0.331) | 0.554*** (0.031) | 0.601*** (0.051) |
| Proportion of those aged 16–74 that are long-term unemployed | 0.039 (0.057) | 0.085 (0.060) | 0.156*** (0.040) | |||
| Proportion of those aged 16–74 working agriculture | −0.015 (0.010) | −0.007 (0.013) | ||||
| Proportion of those aged 16–74 in professional occupations | −0.201*** (0.077) | −0.205*** (0.049) | −0.259*** (0.072) | −0.194*** (0.045) | ||
| IMD 2010, squared | 0.092*** (0.028) | |||||
| Proportion of population aged 16–74 permanently sick, squared | 0.138*** (0.052) | |||||
| Constant | 5.532*** (0.649) | 7.936*** (0.402) | 8.714*** (2.852) | 11.286*** (1.409) | 15.008*** (1.756) | 13.666*** (1.762) |
| Observations | 151 | 151 | 150 | 150 | 150 | 150 |
| Endogeneity test statistic | 11.369 | 10.579 | 5.928 | 17.683 | 6.137 | 22.853 |
| Endogeneity p value | 0.001 | 0.001 | 0.052 | 0.000 | 0.046 | 0.000 |
| Hansen-Sargan test statistic | 14.750 | 20.849 | 1.667 | 23.78 | 1.465 | |
| Hansen-Sargan p value | 0.000 | 0.000 | 0.197 | 0.000 | 0.226 | |
| Kleibergen-Paap LM test statistic | 26.821 | 32.762 | 9.027 | 16.034 | 24.002 | 18.331 |
| Kleibergen-Paap p value | 0.000 | 0.000 | 0.060 | 0.000 | 0.000 | 0.000 |
| Kleibergen-Paap F statistic | 69.320 | 120.521 | 2.323 | 8.979 | 7.220 | 11.627 |
| Pesaran-Taylor reset statistic | 10.116 | 2.456 | 1.405 | 0.175 | 0.073 | 0.466 |
| Pesaran-Taylor p value | 0.001 | 0.117 | 0.236 | 0.676 | 0.788 | 0.495 |
| SW_PH F-statistic | n/a | n/a | 70.796 | 70.796 | 100.608 | 57.002 |
| SW_PH p value | n/a | n/a | 0.000 | 0.000 | 0.000 | 0.000 |
| SW_PB F-statistic | n/a | n/a | 13.469 | 13.469 | 9.052 | 17.375 |
| SW_PB p value | n/a | n/a | 0.000 | 0.000 | 0.000 | 0.000 |
Robust SE in brackets.
*P<0.05, **P<0.01, ***P<0.001.
EU, European Union; IMD, Index of Multiple Deprivation; n/a, not applicable; SW_PB, Sanderson-Windmeijer test statistic for programme budgeting healthcare expenditure; SW_PH, Sanderson-Windmeijer test statistic for public health expenditure; SYLLR, standardised under 75 years of life lost rate.