A New Measure of 19th Century US Suicides

Abstract

Suicides hurt families and the US economy with an annual cost of $69 billion. However, little is known about what determined suicide rates in the past. This is likely due to the lack of consistent data prior to the 20th century. In this article, I propose using newspaper suicide mentions for the period 1840–1910 as a proxy measure for suicide and perform several validation exercises. I show that the stylized facts like suicides drop during wars holds for suicide mentions. I also validate the newspaper suicide mentions against sparse suicide mortality data and a novel valence measure. This new measure can be used to assess the relationship between suicides and numerous policy changes happening in the 19th century that previously could not be explored. It thus offers a new research avenue for quantitative historical analyses, which can inform current policy via novel historical insights.

Economic uncertainty and suicide in the United States

“…Our findings add to those of previous studies that suggest that economic uncertainty is associated with suicides in the UK [30] and the US. [31] They also add to a growing body of literature on the link between the short-term effects of uncertainty, bad financial news and events at the national level on health [17,28,29,[51][52][53].…”

New(spaper) evidence of a reduction in suicide mentions during the 19th century US gold rush

“…Of course, this measure is only helpful if suicide mentions are related to suicides. Kronenberg (2021) reports that for this data source and Maine, there are 1.45 mentions per suicide and that there is substantial covariation between other measures of mental well-being and the suicide mention measure.…”

“…My reading of the literature is that there are two streams. The first considers the relationship between individual changes in income or wealth and how they relate to (mental) health (Eliason & Storrie, 2009a, 2009bKronenberg et al, 2017;Sullivan & von Wachter, 2009a, 2009b. The second stream analyzes how changes in a measure of the aggregate economy relate to changes in an aggregate measure of health (Fishback et al, 2007;Gerdtham & Ruhm, 2006;Gonzalez & Quast, 2011;Miller et al, 2009;Ruhm, 2000Ruhm, , 2015Ruhm, , 2016Stevens et al, 2015).…”

“…Some awardees were funded multiple times, so more newspaper pages from those states are available (see Tables A1 and A2 in the online appendix). For further information regarding the newspaper data, I refer to Kronenberg (2021) and the Chronicling America website. 7 In essence, this dataset is an example of textual corpora available to answer exciting questions relating to economic history (Gutmann et al, 2018).…”

“…2 I use two-way fixed-effect estimation to show that the gold rush reduced suicides. The analysis is performed on a state-year panel that consists of values for mined gold from 1840 to 1860 and suicide mentions per 100,000 pages in regional newspapers as a proxy for suicides (Kronenberg, 2021).…”