| Abstract: | The author considers three methods of counting maternal deaths, and hence of calculating maternal mortality levels, in the nineteenth century. The first method counts maternal deaths classified in contemporary statistical publications. The second method examines nineteenth-century death registers for records of maternal deaths. The third method identifies maternal deaths through family reconstitution. The author finds that each method, in isolation, leads to underestimating maternal mortality levels because maternal deaths were often misregistered or misclassified under nonmaternal causes and were often not associated with live, registered births, as required for family reconstitution. |
