NBER Working Papers by Louis Cain

Making use of those Union Army veterans for whom death certificates are available, we compare the conditions with which they were diagnosed by Civil War pension surgeons to the causes of death on the certificates. We divide the data between those veterans who entered the pension system early because of war injuries and those who entered the pension system after the 1890 reform that made it available to many more veterans. We examine the correlation between specific conditions and death causes to gauge support for the hypothesis that death is attributable to something specific. We also examine the correlation between the accumulation of rated conditions to time until death to gauge support for the “insult hypothesis.” In general, we find support for both hypotheses. Examining the hazard r...