APPENDIX TABLE 5.

REGRESSION RESULTS BY THE MEASURE OF MALARIA RISK

	Malaria Risk Estimation Method					Model (5) in Table 2 Effect on Height at Enlistment		Model (5) in Table 3 Effect on Susceptibility to Infections during the Civil War
Model	Which Period and its Forts are Used to Estimate County Malaria Risk?	Which Type of Fever in Fort Data is Used as Malarial Fever?	Statistics of Estimated Risk
			Mean	Maximum	Minimum	Estimate	t-value	dx/dy	z-value
(1)	1829–1838 / 1871–1874	intermittent / remittent	0.1601	0.4788	0.0138	−2.5749	−2.04	0.2902	4.66
(2)	1829–1838 / 1871–1874	intermittent	0.1256	0.3956	0.0079	−2.9729	−1.97	0.3580	4.79
(3)	1829–1838 / 1871–1874	remittent	0.0583	0.3579	0.0026	−6.0801	−2.55	0.5680	4.24
(4)	1829–1838	intermittent / remittent	0.1337	1.1156	0.0005	−1.1414	−1.40	0.1413	2.89
(5)	1871–1874	intermittent / remittent	0.1656	0.5603	0.0149	−2.3238	−1.83	0.3279	5.05
(6)	County population weighted measure		0.1741	0.4491	0.0080	−1.9169	−1.93	0.3176	6.38

Notes: In model (3), I rescaled the risk variable with multiplying by 10. As weights in model (6), I used county population in the 1850 census. Model (6) uses all available forts in both periods, and both types of fevers.