Results of the two stage regression analysis
| Variable | Coefficient | SE | t Value | Prob >t | VIF | Partial r |
|---|---|---|---|---|---|---|
| *Transformed with natural log. | ||||||
| Stage 1 Predicting mortality as a proxy for need controlling for past expenditures | ||||||
| Constant | 205.51 | 20.42 | 16.74 | 0.000 | ||
| Health expend t−1 | 0.011544 | 0.004556 | 2.53 | 0.015 | 1.1 | 0.36 |
| Education | −101.24 | 21.24 | −4.77 | 0.000 | 1.7 | −0.59 |
| Primary industry* | 4.72 | 2.13 | 2.22 | 0.032 | 1.4 | 0.32 |
| Manufacturing | −17.07 | 6.27 | −2.72 | 0.009 | 2.0 | −0.39 |
| Adj r2 65.4% | ||||||
| Stage 2 Predicting current expenditures | ||||||
| Constant | 2604.25 | 1295.50 | 2.01 | 0.051 | ||
| Predicted mortality | 10.88 | 2.52 | 4.32 | 0.000 | 1.1 | 0.56 |
| Teaching hospital | 1092.14 | 121.82 | 8.97 | 0.000 | 1.6 | 0.81 |
| Toxic pollution* | 45.43 | 12.87 | 3.53 | 0.001 | 1.2 | 0.48 |
| Environ expend* | −431.54 | 204.86 | −2.11 | 0.041 | 1.2 | −0.31 |
| Adj r2sq 67.1% | ||||||