The 1918 influenza pandemic and suicide rates in Norway

Several reports have focused on economic and health outcomes of the 1918 flu pandemic (Garrett, 2007, 2008; Obrizan et al., 2020) as well as race and disabilities (Dimka & Mamelund, 2020; Okland & Mamelund, 2019), while a brief report by Rozen (2020) also demonstrated long-term effects of pandemics that occur long after infection has occurred. Similar long-term effects are also discussed by Aassve et al. (2020) with regards to consequences to levels of social trust. Across these studies, there is a consistent indication that pandemics are associated with negative long-term social outcomes. However, there is a lack of robust research on precise levels of short-term social effects such as suicides as a result of the pandemic in Norway between 1910-1920. This is of particular interest since Norway is a country that remained neutral during WW1.

This proposed article will be based on previous empirical research regarding the 1918 influenza pandemic (commonly known as the Spanish flu) in Norway and will include the key debates that surround the state of the art in this field. Recent developments due to the COVID-19 pandemic, have highlighted that research of this nature is in high demand. There is a great need for pandemic research in order to aid the development of social policies that can mitigate the negative effects of these global events. Hence, in order to understand the current surroundings within the discipline of social science, it is crucial to understand previous occurrences. Consequently, this proposed article-based thesis aims to investigate possible effects on the prevalence of suicide rates as a result of the 1918 flu pandemic. Due to Norway’s neutral standpoint during WW1, there will be no requirement to control for WW1 deaths. This research will push the front as it is the first attempt to explore this association in a neutral context. In more recent years, the debate that surrounds this area of research relates to what extent social aspects were affected by the 1918 flu pandemic. Due to the current day COVID-19 pandemic, research of this nature has become of considerable interest.

Numerous exogenous social and political events have been associated with the 1918 flu pandemic, with some relation to rates of suicide. The relationship between suicide rates and the 1918 flu pandemic has been well-established in a journal article by Wasserman (1992) concluding that suicide rates were more greatly affected by the flu pandemic than the by the coinciding world war. Wasserman suggested four mechanisms that may explain increased suicide rates: fear, experience of the disease itself, bereavement of lost loved ones and the reduction of social integration (Wasserman, 1992). This research will aim to support these explanations by examining rates of mortality within particular municipalities in Norway during the 1918 flu pandemic and the individual lockdown interventions of these locations.

Recently, the changing dynamics of pandemic research have produced an adequate number of studies regarding suicide rates during pandemics in countries such as the United States and United Kingdom (Honigsbaum, 2010; Honigsbaum & Krishnan, 2020; Miller, 2008). Though these findings can be generalised for global pandemic research, this paper intends to fine tune previous literature in relation to suicide rates, specifically in Norway between 1910-1920. Previous suicide literature has increased researchers’ concerns with other specific effects that occur as a result of pandemics (Kotsadam et al., 2020). For example, Dimka and Mamelund (2020) support remarks made by Wasserman (1992), by further stating that there are numerous social conditions and biological variables that can impact on an individuals’ overall health during pandemics. This is in addition to the reduced social integration that is often involved during nonpharmaceutical interventions (NPIs).

Wasserman (1992) used a fixed effects model and much of the methodology in this article will take influence from this paper. The key concepts that will be explored in this paper include rates of suicide and the mental health aspects that coincide with global pandemics. However, unlike Wasserman (1992), this paper will focus solely on Norway. Previous research has proposed that the reduced social integration that occurred alongside the 1918 flu pandemic increased levels of suicide in countries such as the United States of America. Therefore, the objective of this article is to investigate whether this occurred in Norway during the same period. It is crucial to understand that there were other cultural aspects that occurred in the U.S. at this time that were not as prevalent in Norway and therefore results may differ. In order to answer the research question, it is essential to explore the central concepts and discuss theories that will examine the relationships that may be present. These theories, hypotheses and models will be discussed further within the literature review in this thesis.

Participants

The number of participants was measured by the number of inhabitants in each municipality each year in Norway between 1910 – 1920. It is unclear if all individuals accounted for in this data are Norwegian citizens as ethnic backgrounds were not recorded in the data for in this research. Furthermore, this data includes between 583 and 704 municipalities (depending on year) across Norway’s 20 counties at the time of reporting.

Sampling

All historical mortality statistics and population data were extracted from comprehensive annual reports on the health conditions in Norway as well as data available in the Norwegian Centre for Research Data (NSD) as part of the municipal database. This data was also previously used by Kotsadam et al. (2017). Therefore, the sampling population included in this study is the general population of Norway between the years 1910-1920.

The independent variables for this research will be (a) year, (b) municipality, (c) Number of doctors, (d) Number of vaccinators and (e) Total deaths. The dependent variable in this study will be Suicide rates (selvmord). This study will explore existing panel data regarding suicide rates and mortality statistics specific to the time surrounding the 1918 flu pandemic (1910-1920) (See Appendix A for analysis model).

The study will be using quantitative methods to examine the associations between suicide rates and the intensity of the 1918 flu pandemic across time and space with Norwegian municipalities as the unit of analysis. The intended methods will include panel-data regression models with fixed effect (changes within the study unit of municipalities) and random effects (between and within differences).

Data Analysis

Data analysis will be performed using STATA 16.1. Panel data analysis will a regression model including fixed effect and random effects.

Measures

Mortality was measured by the number of deaths and suicides in Norway during this time. Additionally, medical staff were measured by the number of doctors and vaccinators in each municipality during 1910-1920.

Design

This study will use longitudinal panel data collected between 1910-1920. The panel data observes cross sections across the time at which data was collected. Therefore, those included within this sample include the general public of Norway during these dates and some demographic information will be controlled.