*New* Formal Demography Working Group Monthly Meeting (5/16/25)

Posted: 5/8/2025 ()

The Formal Demography Working Group aims to bring together formal demography scholars and those interested in formal demography to discuss recent and classic work, brainstorm new ideas, and to foster new collaborations. The next meeting will be next Friday, 16 May at 11:30am ET (this is 4:30pm in UK). Hampton Gaddy from the London School of Economics will speak about ‘Challenges in Estimating Crisis Mortality: Spatial Heterogeneity, Endogenous Incompleteness, Sample Size, and Ad Hoc Methods.’

Abstract: Accurately estimating the mortality of crises can be methodologically difficult. Researchers often use ad hoc methods with biases that have not been systematically explored, or they apply ‘standard’ demographic methods that rely on assumptions that may be difficult to test in contexts of interest. In this talk, I first review the influence and bias of a class of ad hoc methods that estimate death tolls by extrapolating population counts before and after a crisis and subtracting the projected values (which I term the growth rate discontinuity method, or GRDM). Then, I will describe two interrelated problems that come from estimating excess mortality at different spatial scales. On one hand, when there are multiple mortality crises in a period of interest, spatially disaggregated data may be required for identifying the true baseline mortality for any one crisis; without sufficient disaggregation, death tolls will be underestimated. On the other hand, spatial aggregation is needed to meet the minimum population size required to estimate excess mortality precisely; without sufficient aggregation, death tolls will be underestimated or overestimated, depending on context. Finally, I will briefly discuss biases that arise when the completeness of mortality data in excess mortality models is not assessed.