Table 7.

The short run effect of healthcare expenditure on malaria incidence

Dependent variable:$\mathit{lnmi}$	(1)	(2)	(3)	(4)
Malaria incidence (lag 1)	0.581***	0.535***	0.572***	0.418***
	(6.37)	(5.86)	(6.18)	(3.84)
Public health expenditure	− 0.158***	− 0.134***	− 0.162***	− 0.165***
	(− 4.84)	(− 4.11)	(− 4.86)	(− 5.10)
Public health expenditure (lag 1)	0.159***	0.147***	0.155***	0.130***
	(4.21)	(4.03)	(4.02)	(3.39)
Urban population growth	0.820***	0.481**	0.803***	0.597***
	(3.92)	(2.31)	(3.79)	(2.98)
Urban population growth (lag 1)	− 0.809***	− 0.679***	− 0.805***	− 0.646***
	(− 3.48)	(− 3.11)	(− 3.45)	(− 2.95)
Female-head Households	0.924***	0.904***	0.925***	0.916***
	(7.41)	(6.95)	(7.39)	(7.46)
Female-head household (lag 1)	− 0.803***	− 0.771***	− 0.799***	− 0.643***
	(− 5.42)	(− 5.32)	(− 5.37)	(− 4.18)
Regulatory quality	− 0.129**		− 0.129**
	(− 2.47)		(− 2.46)
GINI coefficient		− 1.503***		− 0.977***
		(− 3.32)		(− 2.72)
Female literacy		0.787**
		(2.54)
$ECT(-1)$	− 0.166***	− 0.168***	− 0.169***	− 0.147***
	(− 4.26)	(− 4.27)	(− 4.30)	(− 3.57)

Lags are chosen automatically based on AIC; t-values in parentheses; ***p < 0.01, **p < 0.05, *p < 0.1. $\mathit{lnpvhe}$, $lnue\_fem$, $\mathit{lngdppc}$, and $\mathit{lngini}$ do not appear because they have zero lags and have no impact in the short run. ECT(-1) is error correction term