The persistently high rate of suicide in Lithuania
Suicide is a major global health problem. In addition to the enormous toll that suicide takes on individuals and families, a high suicide rate can be detrimental to the long-run growth of a society, particularly if mostly young people are affected.
Historically, Lithuania has had very high suicide rates, especially among its male population. While the rates have fallen since their peak in the mid-1990s, when the collapse of the USSR exposed Lithuanians (and their neighbours) to a new and unfamiliar social environment (Pray et al. 2013, Värnik et al. 2010), suicide rates in Lithuania are still higher than those reported in neighbouring states such as Latvia and Estonia.
In this column, we identify possible factors related to the high suicide rates in the Baltic states and specifically in Lithuania – drawing from the fields of economics, psychology, and sociology – with an updated dataset covering the mid-1990s to 2016. We do not aim to show any causal link but instead, seek to shed light on potentially important elements to take into account. We draw full comparisons between statistical selection methods and related econometric outcomes to identify the factors that most robustly affect total and male-only suicide rates in the Baltics and specifically in Lithuania.
The contributions of our research (Comunale 2020a,b) are twofold. First, this is one of the very few articles that look specifically at the Lithuanian rate of suicide among the Baltic states from an economic point of view and with economic techniques but that also includes social and psychological aspects and uses an updated dataset. Second, the paper relies on sound statistical and econometric techniques in treating data and providing outcomes. We hope that the results of this study will help policymakers reduce the incidence of suicide in the Baltic states.
Some stylised facts
We show now some stylised facts on suicide rates in Lithuania compared to the rest of the EU. As shown in Figure 1 for the total population and for males in Figure 2, Lithuania is a clear outlier in the EU, also in comparison to its possible peers, Estonia, Finland, and Latvia, which are normally perceived as having high rates of suicides.
Figure 1 Total suicides per 100,000 inhabitants
Source: Eurostat, standardised death rate per 100,000 inhabitants, year 2016. The total refers to both male and female population.
Figure 2 Male suicides per 100,000 inhabitants
Source: Eurostat, standardised death rate per 100,000 inhabitants, year 2016.
This trend is very slowly decreasing over time (Figure 3) for both males and females but still, in 2016, the country experienced the highest suicide rates in the EU (28.3 per 100,000 inhabitants).
Figure 3 Total suicides rates over time, Baltic states
Source: WHO, percentage of suicides over the entire population. The total refers to suicide rates of both men and women.
Not only is gender-specific data informative of the Lithuanian situation, but so are the differences across age groups. Among men and women combined, we do not see a major difference between people younger than 65 years old and people 65 and above in the Baltic states, except for the end of the 2010s. For males, the over-65 cases are an increasing percentage of total male suicides and the figure is always more than the average across total male cases. Disaggregated age-wise, men are more at risk in the working ages of 25–45 years and just before/after retirement at 55–65 years of age, with a suicide rate ranging between 48 and 86 per 100,000 inhabitants.
Possible factors linked to suicide rates
We examine 35 possible factors and their links to suicide rates. Socioeconomic factors include GDP growth, fiscal balance, unemployment, wages, demographics, health expenditure, the percentage of rural population, education levels, and Gini index. For behavioural and psychological factors, we look at marital status, alcohol consumption, an index of extroversion, life satisfaction, number of psychiatrists, and the percentage of people with chronic depression. We also include environmental aspects such as changes in temperature over the years and global factors such as the Index of Global Economic Policy Uncertainty.
We do not claim any causality: highly correlated variables do not guarantee causation, i.e. high correlation may be caused by a similar set of shocks. We do not rely on theoretical models but to avoid spurious selection, we chose variables using statistical methods as the weighted-average least squares criterion proposed by Magnus et al. (2010) and, as a check, the elastic net (Zou and Hastie 2005). The selected variables are then used as independent variables for total and male suicides in frequentist and Bayesian regressions.
We find and analyse some robust results from the statistical selections and regression analysis. The main factors linked to suicides in the region are GDP growth, demographics, alcohol consumption, psychological factors, and climate temperature. Health expenditure appears to relate to reduced suicides but only for the entire population.
Our results are in line with Pereira dos Santos et al. (2016), which studies the relationship between suicide rates and economic and social factors mainly taken from Durkheim (1897). For instance, we confirm the negative correlation between suicide rates and real cycles (or GDP growth rates).
Alcohol consumption is especially relevant for male suicide rates as traditional norms of masculinity can be also associated with excessive drinking (Baranov et al. 2020). It may be excessive drinking rather than regular consumption that triggers suicidal behaviours.
The percentage of people with reported chronic depression over total population is also positively linked to male suicides. We believe, however, that the effective number of people with depression is underestimated, especially among the male population. As highlighted recently in Baranov et al. (2020), traditional masculinity norms have profound economic and social effects and can influence the willingness to ask for help, thus attaching a stronger stigma to mental health problems. However, it is good that in the past few years, Lithuania has opened up the discussion about mental health issues and hotlines are being developed more actively, together with reforms in hospitals and mental health centres (as reported by the EU Commission Country Health Profile 2017).
Lastly, there is an extensive literature on how environmental and meteorological variables affect mental health, with more recent studies on the link between suicides and climate change or global warming (see for instance Fountoulakis et al. 2016). We do find here a positive link between suicide rates and global warming.
In the case of Lithuania, macroeconomic and labour-market conditions appear to be strongly linked to total and males suicides. The importance of the labour market is also confirmed in the alternative selection method and is in line with Stuckler et al. (2009), who indicate that an active labour market might mitigate the adverse effect of unemployment on health.
The percentage of rural population does not seem to be a key robust factor for Lithuania, even though it was recognised as such in previous studies (Kalėdienė and Petrauskiene 2004 for Lithuania, and Hirsch 2007 more in general), as rural living can lead to social isolation and the stigma toward mental health issues may be higher.
Authors’ note: The views expressed in this column are those of the authors and do not necessarily represent those of the ECB, the Bank of Lithuania or the ESCB.
Baranov, V, R De Haas and P Grosjean (2020a), “Men”, VoxEU.org, 2 April.
Comunale, M (2020), “The persistently high rate of suicide in Lithuania: An updated view”, Bank of Lithuania Discussion Paper No. 21/2020.
Durkheim, E (1897), Le suicide: étude de sociologie, Paris.
Fountoulakis, K, I Chatzikosta, et al. (2016), “Relationship of suicide rates with climate and economic variables in Europe during 2000–2012”, Annals of General Psychiatry 15: 19.
Hirsch, J K (2006), “A review of the literature on rural suicide”, Crisis: The Journal of Crisis Intervention and Suicide Prevention 27(4): 189–99.
Kalėdienė, R, and J Petrauskiene (2004), “Socio-economic transition, inequality, and mortality in Lithuania”, Economics and Human Biology 2(1): 87–95.
Magnus, J R, O Powell and P Präufer (2010), “A comparison of two model averaging techniques with an application to growth empirics”, Journal of Econometrics 154: 139–53.
Pereira dos Santos, J, M Tavares, P Pita Barros (2016), “More than just numbers: Suicide rates and the economic cycle in Portugal (1910–2013)”, SSM – Population Health 2(2016): 14–23.
Pray, L, C Cohen, I Henrik Mäkinen, A Värnik and F Landis MacKellar (eds.) (2013), “Suicide in Eastern Europe, the CIS, and the Baltic Countries: Social and public health determinants. A foundation for designing interventions”, International Institute for Applied Systems Analysis (IIASA) Health and Global Change Unit, Laxenburg, Austria.
Stuckler, D, S Basu, M Suhrcke, A Coutts, et al. (2009), “The public health effect of economic crises and alternative policy responses in Europe: An empirical analysis”, Lancet 374(9686): 315–23.
Värnik, A, M Sisask and P Värnik (2010), “Baltic suicide paradox”, Tallinn University Press /Tallinna Ülikooli Kirjastus.
Zou, H, and T Hastie (2005), “Regularization and variable selection via the elastic net”, Journal of the Royal Statistical Society Series B: Statistical Methodology 67(2): 301–20.