Durkheim’s types of suicide and

social capital: a cross-national

comparison of 53 countries

Matthew D. Moore

Introduction or decreased levels of suicide based on Durkheim’s

conceptualisation.
In 1897 Emile Durkheim wrote Le Suicide, which Examinations of suicide have neglected to
examined suicide using the sociological method connect social capital to Durkheim’s types of sui-
that he laid out earlier in his career. According cide. Furthermore, social capital’s effect on suicide
to Durkheim, there were four has not been a key area of
different types of suicide: study in the social sciences
Matthew D. Moore is in the Department of
egoistic, altruistic, anomic, Sociology, Criminology, and Anthropology,
(Kelly et al. 2009). The current
and fatalistic. The types of University of Central Arkansas. analysis connects Durkheim’s
suicide created by Durkheim Address for correspondence: 3061 lrby Hall, types of suicide to social
suggested that suicide was 201 Donaghey Ave, Conway, AR 72035. capital. Moreover, the differ-
more or less likely to occur Email: mdmoore@[Link]. ent suicide types allow the
due to levels of regulation by current analysis to examine
society and the level of integration of the individual if a certain form of social capital increases the
in society. Since Durkheim’s conceptualisation likelihood of suicide or decreases the likelihood.
of suicide, many studies have been conducted Examining 53 countries cross-nationally, the anal-
examining the types of suicide that Durkheim put ysis explores which factors of social capital and
forward (Kusher and Sterk 2005; Moore et al. which suicide types have the strongest effect on
2014). suicide.
Today, social capital has become a key variable
of study in a variety of areas examining social and
economic outcomes for society (Helliwell 2007; Literature review
Kelly et al. 2009; Knack 2002; Moore et al. 2014;
Putnam 2000; Rice 2001; Rosenfeld et al. 2001; Durkheim examined suicide to demonstrate how
Tiepoh et al. 2004). Social capital is defined as the the sociological method could be applied to a
“connections among individuals – social networks societal problem. At the time of Durkheim’s writ-
and norms of reciprocity and trustworthiness that ing, suicide was seen as a psychological issue.
arise from them . . . in a sense social capital is However, Durkheim believed that suicide had soci-
closely related to what some have called ‘civic ological factors that could be examined using proper
virtue’” (Putnam 2000, p.19). Social capital can statistical methods. Examining different rates of
be understood in the four types of suicide that suicide across many different countries, Durkheim
Durkheim discussed more than a century ago. demonstrated that there were four different types of
A well-integrated and regulated society has high suicide that were created by society. The essence
levels of social capital. Moreover, a well-integrated of the suicide types was the level of integration and
and regulated society would experience increased regulation of the society the individual lived in.

ISSJ 219-220 
C 2017 John Wiley & Sons Ltd
152 Matthew D. Moore

Egoistic suicide Fatalistic suicide

Egoistic suicide occurs when the individual does Fatalistic suicide is not discussed in a major way
not feel as if they are well integrated into society by Durkheim. Ritzer and Goodman (2004) point
(Ritzer and Goodman 2004). An individual that out that fatalistic suicide is only discussed by
is not well integrated into society feels as if they Durkheim in a footnote. Fatalistic suicide occurs
are not part of the community, but the community when regulation of society is excessive. A slave may
also feels that the individual is not part of their commit suicide because of his hopeless position.
lifestyle. The individual then feels isolated and Individuals who feel that they have no control of
does not receive the moral guidance needed to their lives would be more likely to commit suicide.
operate in their social environment. According to Because fatalistic suicide was not discussed in detail
Durkheim, individuals that are not integrated into by Durkheim it will not be examined in the current
society would be more likely to commit suicide. analysis.
This is because the individual would not have the
moral support to get through the downturns life Social capital
often presents us with. Instead, the individual would
feel that they alone feel the way they do. Thus, Social capital has been demonstrated to have a
suicide would be an option for the individual to variety of positive social and economic outcomes
reduce their negative feelings. for society. Social capital has been linked to bet-
ter educational outcomes (Coleman 1988), higher
Altruistic suicide levels of volunteerism (Wilson and Musick 1997),
better quality of life (Recker 2013), decreased crime
Altruistic suicide occurs when the individual is too (Rosenfeld et al. 2001), and lower suicide rates
integrated into society (Ritzer and Goodman 2004). (Kelly et al. 2009). Economically, social capital
Durkheim suggested that when the individual is too has been found to increase economic growth and
integrated the pressure to succeed for the greater prosperity (Goetz and Rupasingha 2006; Tiepoh
good could be overwhelming. Often individuals et al. 2004), improve economic equality and sta-
may commit suicide because they believe it is good bility (Casey and Christ 2005), and produce more
for their society (i.e., suicide bombers). Another successful businesses (Besser 1999). Overall, social
way to examine Durkheim’s concept of altruistic capital has become one of the best indicators of
suicide is that individuals who are expected to community health.
succeed but do not will turn to suicide. A samurai There are many definitions of social capital
warrior who does not live up to expectations will in the literature (Adler and Kwon 2002; Woolcock
commit suicide because he was too integrated into 2001). As a result social capital has “a variety of
society. Thus, when an individual is too integrated interpretations reflecting different trends in domi-
and expected to succeed, but then does not succeed, nate thinking and particular perspective of the user”
suicide should increase. (Wall et al. 1998, p.300). Putnam’s definition of
social capital has become one of the more popular
Anomic suicide versions of the concept. Putnam defined social
capital as the “connections among individuals –
Anomic suicide is likely to occur when the regula- social networks and norms of reciprocity and trust-
tion of society is disrupted (Ritzer and Goodman worthiness that arise from them . . . in a sense social
2004). The disruption can occur during an eco- capital is closely related to what some have called
nomic boom or depression. The disruption leaves ‘civic virtue’” (Putnam 2000, p.19).
individuals with little guidance over what norms Based on Putnam’s definition of social capital,
and values they should be following. People are not the different types of suicide can be operationalised
sure what norms and values still apply and what as a social capital measure. Durkheim’s conceptu-
new norms and values need to be followed. The alisation of suicide is based on the regulation of
lack of regulation leaves some individuals feeling society and the integration of the individual in soci-
overwhelmed by the changes occurring in society ety. The connections of the individual can be seen
and would increase suicide. as the integration level of the individual to society.


C 2017 John Wiley & Sons Ltd
Durkheim’s types of suicide and social capital 153

Also, the number of networks and organisations in in areas where the community is highly involved,
the community aid in the regulation of daily life of individuals may turn to altruistic suicide when they
an individual. Thus, changes in social capital could feel they have not succeeded. Social capital is based
produce increases or decreases in suicide based on around improving the community through norms of
the type of change that occurs. reciprocity. If an individual does not feel they can
If individuals become less integrated in the return a favour they may feel too much pressure
community, there would be increased egoistic from the society, thus altruistic suicide can occur.
suicide. Studies of social capital have demon- Anomic suicide occurs when the regulation
strated that areas with more diversity do have less of society breaks down. This breakdown reduces
social capital (Alesina et al. 1999; Alesina and social control as individuals are free to engage
La Ferrara 2000; Putnam 2007). Alesina and La in whatever desire they want. Social capital has
Ferrara (2000) examined the level of participation demonstrated that when the neighbourhood is not
in social activities across the United States. They regulated crime increases (Deller and Deller 2012;
found that the more diverse the locality the less the Rosenfeld et al. 2001). Rosenfeld et al. (2001)
participation in social activities. Moreover, Alesina examined the role social capital plays in reducing
et al. (1999) examined the level of diversity and the violent crime. Using a structural equation model
level of social spending. The scholars found that the authors found that social capital does decrease
as diversity increased, the level of social spending violent crime. Deller and Deller (2012) examined
decreased. Finally, Putnam (2007) suggested that social capital in rural counties and found that
increased diversity would reduce social capital. The social capital reduced burglary and larceny. Other
findings of these studies suggest that as individuals researchers have demonstrated that social capital
feel they are not similar to their neighbours they decreases the fear of crime (Kruger et al. 2007)
withdraw from the neighbourhood, thus they are and decreases firearm violence in a community
less integrated into the society. As individuals feel (Kennedy et al. 1998). These studies illustrate
different from others around them they withdraw that social capital regulates the behaviour of a
from community life. The individual would feel less neighbourhood. Increased social capital reduces the
respected, less satisfied with the life, and have less crime and fear of crime in a neighbourhood. As
purpose because the other neighbourhood residents social capital decreases individuals are not regu-
did not share the same values and norms. Therefore, lated and may begin to commit crimes and other
the individual would not receive positive feedback deviant acts, such as suicide.
about their station in life and egoistic suicide would
become more likely. Types of social capital
Altruistic suicide can occur when the pres-
sure to succeed becomes overwhelming to the Social capital has been used as an umbrella variable
individual. Examples of the pressure to succeed in many studies (Kennedy et al. 1998; Kruger
connection to suicide have been demonstrated et al. 2007; Rosenfeld et al. 2001). However,
in the literature (Cutright and Fernquist 2000; other scholars have explained that there may be
Lester and Lynn 1993). Lester and Lynn (1993) different types of social capital that affect deviant
examined measures of achievement, competition, acts in different ways. One of the more popular
and occupational preference of college students in types of social capital is bridging and bonding.
30 countries. They found that when students Bonding social capital is comprised of networks
expressed higher values to achieve and compete, and norms of reciprocity for individuals among the
the suicide rates increased. In countries where same group, and bridging social capital consists of
individuals feel the pressure to succeed the suicide networks and norms of reciprocity for individuals
rates were higher. Cutright and Fernquist (2000) among individuals in different groups. Bridging
examined multiple measures of integration for 20 social capital is what Granovetter (1973) called
countries from 1955 to 1989 and found that coun- weak ties and bonding is what he called strong
tries that had a more collective orientation had ties. Examining the effect that bonding and bridging
higher suicide rates than less collective countries. social capital have on crime, Beyerlein and Hipp
The findings of these studies demonstrate that if (2005) used evangelical religions as a measure of
individuals feel the pressure to succeed and live bonding social capital and mainline religions as


C 2017 John Wiley & Sons Ltd
154 Matthew D. Moore

a proxy measure for bridging social capital. They literature (Flavin and Radcliff 2009; Kelly et al.
found that bridging social capital decreased crime 2009).
and bonding social capital increased crime. This Also, many countries do not report suicides
study demonstrates how social capital can have a to the WHO or collect information on the number
different effect depending on the type. The type of of suicides for their nation. Taking the average of
social capital present in a community can have a the country allows the current analysis to have a
different effect on suicide in the same way. This large sample size to examine predictors of suicide
could result in the different types of suicide that cross-nationally. Countries that report for two of
Durkheim laid out. the time periods could be included in the analysis,
Other scholars have also demonstrated that dif- whereas if a single year had been selected for the
ferent forms of social capital can affect crime differ- analysis some countries might have been excluded
ently. Deller and Deller (2010) argued that the type from the analysis. For example, the Dominican
of social capital matters in reducing crime in rural Republic reported suicides to the WHO for 2000
areas. They found that social capital that benefitted and 2005. Thus, the country was included in the
the community reduced property crime. Moore and current analysis and the two reported time periods
Recker (2014) created Olsen-Type, Putnam-Type, were averaged together. All countries in the current
and Recreation-Type groups based on the goals analysis have reported at least two time periods
of the organisation. Individuals join Olsen-Type to the WHO. The final sample size of the current
groups because they want something in return. analysis was 53.
Individuals join Putnam-Type groups to help the While the sample size may be small, the
community, and individuals join Recreation-Type sample size of the current analysis is in line with
groups to engage in recreation activities. Moore other studies. Flavin and Radcliff (2009) examined
and Recker (2014) found that crime increased when suicide rates across states in the United States with
there were more Olsen-Type groups in the county a sample size of 48. Again, suicide may not be
and crime decreased when there were more Putnam- reported to the WHO or collected by the country.
Type groups in the county. Other researchers have Thus, the small sample size is due to suicide data
demonstrated that different forms of social capital being unavailable for multiple years for a variety
do affect crime differently (De Coster et al. 2006; of counties. Many studies on cross-national suicide
Hawdon and Ryan 2009). Once again this demon- have a smaller sample of countries than the current
strates that research should examine social capital analysis (Chew and McCleary 1995; Girard 1993;
using different lenses. Suicide and other deviant Huang 1996). Similar problems with cross-national
behaviours may be affected differently, depending data exist with other studies examining social phe-
on the type of social capital present or absent from nomena. For example, cross-national research on
the neighbourhood. homicide often has small sample sizes. Examining
homicide cross-nationally, Pridemore (2008) had
a sample size of 46, Pratt and Godsey (2003) of
Methods 46, Savolainen (2000) of 39, and Fajnzylber et al.
(2002) of 39. While not optimal for statistical
Dependent variable analysis, the current analysis could be used as a
guide for future analyses and is in line with other
The suicide rates were collected from the World cross-national research studies.
Health Organization (WHO). The WHO amasses
the suicide rates for a number of countries across Independent variables
the globe. The WHO reported suicide rates for 2000,
2005, and 2008, which were then averaged for the Measuring social capital can be a tricky process.
current analysis. Suicide displays a wide variation Adler and Kwon (2002) noted that there are as many
from year to year (Flavin and Radcliff 2009). To authors on social capital as there are operational def-
control for the volatility in the suicide data, the initions of social capital. The social capital literature
average of the eight-year time span was used in the has become so clouded with different definitions
current analysis. Averaging suicide rates to control that Woolcock (2001, p.69) stated that this “has
for the wide variation is commonly done in the led many critics to accuse social capital of having


C 2017 John Wiley & Sons Ltd
Durkheim’s types of suicide and social capital 155

become all things to all people, and hence nothing feelings from the community members. Thus, the
to anyone”. Thus, researchers have to be clear on individual may believe that they do not have a
why they are selecting variables that operationalise satisfying or purposeful life. Moreover, Stack and
social capital. To operationalise social capital, we Kposowa (2008) used a life satisfaction measure
selected variables that are in line with the definition in their analysis of acceptance of suicide cross-
of social capital and can be operationalised as nationally. Therefore, purposeful life and life satis-
a form of suicide that Durkheim laid out. All faction represent the feeling of trust and networks
social capital variables were collected from United within the community. Respect, purposeful life, and
Nations Development Programme (UNDP) Human life satisfaction all represent Durkheim’s concept
development report for 2010. of egoistic suicide. Egoistic suicide occurs when
The first variable used to measure social capi- the individual is not integrated into the society
tal was the perception of safety. If individuals do (Ritzer and Goodman 2004). As discussed above,
not feel safe from violence and harm they will these three variables fluctuate based on how the
withdraw from community life. This withdrawal individual is integrated and viewed within their
from community life decreases the level of social community. Low integration should produce low
capital in the neighbourhood. Studies examining levels of respect, purposeful life, and life satisfac-
crime and social capital have found that increased tion and increased suicide.
social capital decreased crime (Deller and Deller Finally, the level of social support the indi-
2010, 2012; Rosenfeld et al. 2001). Kruger et al. vidual felt was included as a social capital variable
(2007) found that as social capital increased, fear in the current analysis. Putnam (2000) argued that
of crime decreased. Thus, the perception of safety social capital included the social networks an indi-
does indicate the level of social capital present vidual belongs to. The level of social support is
in the neighbourhood. Furthermore, the perception the number of social networks the individual feels
of safety is a measure of Durkheim’s concept of they are part of. From the UNDP, the social support
anomic suicide. According to Durkheim, anomic variable was percentage of individuals saying they
suicide occurred when the regulatory powers of felt they had social support. Social support also
society became disrupted. Because of this dis- represents Durkheim’s concept of altruistic suicide.
ruption in regulation individuals would be free Altruistic suicide occurs when the integration of
to do what they wanted. Therefore, individuals an individual in society is too strong (Ritzer and
in an anomic society would engage in all forms Goodman 2004). If people have too much integra-
of deviance, including crime and suicide, because tion in society they will feel pressured to achieve.
the society could not stop them. If individuals Moore et al. (2014) found that counties in the
perceived society as safe, that would indicate a United States with individuals who are expected to
regulated society and vice versa. achieve a high level of success had higher suicide
Respect, purposeful life, and life satisfaction rates. Therefore, individuals who have a large social
were all included in the analysis for their connection support network may be more likely to commit
to social capital and as a representation of egoistic suicide based on Durkheim’s conceptualisation of
suicide. Individuals were asked if they felt that they altruistic suicide.
were respected by their peers, had a purposeful
life, and were satisfied with their life. Respect Control variables
was included as a social capital because of the
trust networks and then norms of reciprocity that The indicators for the present analysis were used
is included in the definition of social capital by based on previous literature and for the variables
Putnam (2000). If an individual has respect from connection to Durkheim’s conceptualisation of the
their neighbours, the neighbours will in turn trust predictors of suicide. The first variable selected for
and reciprocate good feelings toward that individ- the current analysis was educational spending as a
ual. Thus, the individual will feel that they are percentage of the gross domestic product from 2000
part of the community. The community also has a to 2007 (see Table 1). The average of educational
large impact on how an individual feels about their spending was used in the analysis. The average
station in life. If the community does not trust an allows the analysis to include more countries and
individual, the individual will experience negative to control for spikes in spending from one year to


C 2017 John Wiley & Sons Ltd
156 Matthew D. Moore

TABLE 1. Descriptive statistics for the variables in the 1996; Johnson et al. 2000; Stack and Kposowa
analysis (N = 53) 2008) and in studies of suicide in the United States
(Flavin and Radcliff 2009; Moore et al. 2014). The
Standard unemployment rate is an indicator of poverty. As
Mean deviation
poverty and unemployment levels increased, indi-
Suicide rate 11.828 8.332 viduals would experience more strain. This strain
Education 4.756 1.769 could lead to deviant behaviours such as suicide.
Population density 137.456 153.456
Infant mortality rate 12.117 10.856
Unemployment rate 8.265 4.537 Findings
Urban % 71.097 15.199
Perception of safety 58.018 15.645
Respect 86.964 9.843 A regression analysis was performed with the inde-
Purposeful life 86.981 8.486 pendent variables and the control variables. With
Life satisfaction 6.894 .966 a small sample size, the model may be susceptible
Social support 87.018 7.080 to multicollinearity and to test for the presence of
multicollinearity, Variance Inflation (VIF) scores
are used. VIF scores above 10 indicate the presence
another. The education variable was included as a of multicollinearity (Neter et al. 1989). However,
control for level of development for each country. other scholars in macrosociology have suggested
Countries that have a robust educational system that VIF scores above 5 (Ouimet 2012) indicate
would be expected to experience more innovation the presence of multicollinearity. No model in the
and have returns on that investment. current analysis had a VIF score above 5. The R2 =
Population density and the percentage of the .632 indicated a good fitting model. Furthermore,
population that resides in urban areas were included the model explained 54.5 per cent of the variance
in the current analysis as modernisation measures. in suicide rates for the countries in the model. This
Durkheim suggested that as society modernised level of variance suggests an acceptable goodness-
people would experience anomie. This feeling of of-fit to our model, considering the wide variation in
loneliness would lead to various forms of deviance, heterogeneity found in each country and the cross-
including suicide. As societies modernised, they sectional design of the current study.
would become more densely packed as more None of the control variables were significant
individuals moved to the society and individuals in the regression analysis (see Table 2). These
would move to urban centres to obtain employ- control variables were selected for the current
ment. The infant mortality rate was included as a analysis because of their prevalence in other cross-
proxy measure for poverty. Many studies use the national studies (Huang 1996; Johnson et al. 2000;
gross domestic product (GDP) as a measure of Stack and Kposowa 2008). The current analysis
poverty (Chew and McCleary 1995; Girard 1993; is in line with previous studies on suicide cross-
Huang 1996). However, examining homicide cross- nationally. Johnson et al. (2000) explained that
nationally, Pridemore (2008) argued that the GDP the findings on unemployment effect on suicide
is a poor measure of poverty. Countries can have a have been mixed. The current analysis did not find
wide variation of poverty but have similar levels a significant effect for unemployment. Thus, the
of GDP. Instead, the infant mortality rate pro- effect that unemployment has on suicide continues
vides a better indicator of poverty, as countries to be cloudy.
that have larger social support will have lower Moreover, the other control variables were not
infant mortality rates. Pridemore (2008) pointed significant, even though they have been demon-
out that many cross-national studies use the infant strated to be significant predictors in other cross-
mortality rate as a proxy measure for poverty. In national studies of deviant behaviours. Pridemore
fact, the association between poverty and the infant (2008) found that the infant mortality rate was a
mortality rate is strikingly consistent (Frey and stronger predictor of homicide than GDP. However,
Field 2000; Pridemore 2008). The last variable the current analysis did not find the infant mortality
used as a control was the unemployment rate. The rate to be significant. Other cross-national studies
unemployment rate has been used in a number of have found that the percentage of the population in
studies examining suicide cross-nationally (Huang urban areas as a significant predictor of homicide


C 2017 John Wiley & Sons Ltd
Durkheim’s types of suicide and social capital 157

TABLE 2. Regression model with suicide rates for 53 suicide rate (see Table 2). Moreover, Durkheim’s
[Link] concept of anomic suicide does not occur within
the context of social capital. Society’s regulation
B VIF of individuals does not affect their likelihood of
Education .065 2.259 suicide. Purposeful life was also not significant
(.949) in the model. Purposeful life was theorised to
Population density −.060 1.587 be a measure of egoistic suicide. This finding
(.008)
Infant mortality rate −.135 2.945 may suggest that not being integrated into society
(.123) does not affect suicide. However, both respect
Unemployment rate −.045 1.456 (β = −.427; p < .01) and life satisfaction (β =
(.207)
Urban % −.077 1.778
−.374; p < .05) were significant and negative.
(.076) As respect increased, the suicide rate decreased
Social capital measures and as life satisfaction increased, the suicide rate
Anomic suicide decreased. These findings suggest that social capital
Perception of safety −.130 1.333
(.059) does reduce suicide and that social capital may
Egoistic suicide help reduce crime through egoistic (integration)
Respect −.427** 1.852 suicide. Individuals who were integrated into the
(.110) society felt respected and satisfied with their lives
Purposeful life −.257 2.310
(.144) did not commit suicide. Stack and Kposowa (2008,
Life satisfaction −.374* 3.262 2011) found that life satisfaction reduced suicide
(1.496) acceptability, thus the current analysis is in line
Altruistic suicide with previous research.
Social support .448* 4.050
(.230) Finally, social support was significant and
R2 .632 positive (β = .448; p < .05). As social support
increased, the suicide rate increased. This finding
*p < .05; **p < .01; standardised coefficients (standard error).
suggests that high levels of social support increase
suicide. The level of social support was concep-
tualized to be an example of altruistic suicide. As
(Jacobs and Richardson 2008; Pratt and Godsey a person becomes over-integrated into society the
2003), and other researchers have argued that individual begins to feel pressure to succeed and
population density could be used as a modernisa- to fit into the norms of that society. Too much
tion variable (Avison and Loring 1986; Hansmann social support can put pressure on the individual,
and Quigley 1982; Lafree and Kick 1986; Lee who then turns to suicide when they cannot live
2001; Messner 1982; Messner and Rosenfeld 1997; up to the expectations of the society. Moore et al.
Neapolitan 1996). The current analysis did not find (2014) found that too much integration leads to
a significant effect for percentage of the population increased suicide in the United States. Furthermore,
living in urban areas or for population density. from a social capital perspective, being watched
Finally, Haddad et al. (1990, p.1) explained that and judged by the community that an individual
“education has been recognized as a cornerstone lives in can create pressure when the individual
of economic and social development”. Moore and does not live up to those judgments. Thus, being
Daday (2010) demonstrated in a cross-national too integrated has the same effect on suicide as not
analysis that human capital could be increased being integrated into society.
through education. The current analysis did not
find that educational spending was a significant Conclusions
predictor of suicide. The findings for the control
variables suggest that suicide is affected by other Since Durkheim’s research on suicide there have
variables than have been used in previous cross- been a number of studies that have examined his
national analyses. conceptualisation of the types of suicide that he
When examining the social capital measures created (Kusher and Sterk 2005; Moore et al. 2014).
in the current analysis, the perception of safety was At the same time, a number of studies in social
not significant. Feelings of safety do not affect the capital have demonstrated how social capital is


C 2017 John Wiley & Sons Ltd
158 Matthew D. Moore

a benefit to a community (Helliwell 2007; Kelly analyses have found that areas that have too little
et al. 2009; Knack 2002; Messner et al. 2001; integration have higher suicide rates (Kelly et al.
Moore et al. 2014; Putnam 2000; Rice 2001; Tiepoh 2009; Recker and Moore 2016). Previous analyses
et al. 2004). However, social capital’s connection have also illustrated that too much integration
to suicide has not been a main area of interest increases suicide (Cutright and Fernquist 2000;
in studies examining suicide (Kelly et al. 2009). Lester and Lynn 1993; Moore et al. 2014). The
Moreover, no studies have attempted to connect findings of the current analysis demonstrate that
Durkheim’s conceptualisation of egoistic, anomic, too little or too much integration can lead to
and altruistic suicide with social capital. The cur- suicide.
rent analysis operationalised social capital using Another interesting finding from the current
Durkheim’s types of suicide at the cross-national analysis was how social capital should be stud-
level of analysis. ied using different types. Social capital has been
The findings of the current analysis do shed examined as a singular concept (Kennedy et al.
light on how social capital can be used to reduce 1998; Kruger et al. 2007; Rosenfeld et al. 2001).
suicide. First, the level of integration in society These studies have provided a great deal of infor-
is a significant predictor of suicide. Egoistic sui- mation in the role social capital plays in social and
cide occurs as individuals do not feel that they economic development. However, recent studies
are integrated into their community. The level of have begun to illustrate that social capital can be
respect and life satisfaction significantly reduced broken down into different types (Beyerlein and
suicide in the current analysis. Previous studies Hipp 2005; Deller and Deller 2010; Moore and
on social capital demonstrated that as individuals Recker 2014). The current analysis demonstrated
feel that they are different from the other members that social capital can both decrease and increase
of their community they withdraw (Alesina et al. suicide. Societies with too low a level of social
1999; Alesina and La Ferrara 2000; Putnam 2007). capital have increased rates of suicide and societies
Therefore, individuals have to feel adequately inte- with too high a level of social capital have increased
grated into their community or the community will suicide rates. This demonstrates that social capital
experience increased egoistic suicide because of the continues to be a hazy concept that is in need of
lack of social capital present. continued research. However, studies now indicate
Next, the level of social support the individual that social capital should not be seen as a singular
felt increased suicide. The level of social support concept.
may become overwhelming to an individual as they The current analysis also demonstrated that
feel they cannot live up to expectations. Durkheim suicide is different than other deviant behaviours.
named the over-integration of society altruistic The current analysis used control variables that
suicide. Countries where individuals felt a large have been found to be significant in previous
amount of social support had increased suicide cross-national studies. Pridemore (2008) argued
rates. This finding is similar to previous research that the infant mortality rate is a more significant
that demonstrated that competition, achievement, predictor of homicide than the GDP in cross-
and collectivism were the dominate values (Cutright national studies. The infant mortality rate was not
and Fernquist 2000; Lester and Lynn 1993). The significant in the current analysis. Moreover, other
concept of social capital is based on increased levels variables that measured modernisation and poverty
of social support. Therefore, societies where the were not significant. The reason individuals commit
individual feels pressure to succeed and is over- suicide is different than other studies examining
integrated would have increased altruistic suicide cross-national deviant behaviours.
because the level of social capital is too high. The current analysis does suffer from a few
From the findings in the current analysis, limitations. First, the sample size is small. However,
countries that provided too little integration and the current analysis is in line with previous stud-
too much integration have higher suicide rates. ies on suicide (Chew and McCleary 1995; Flavin
Therefore, communities need to find a way to and Radcliff 2009; Girard 1993; Huang 1996)
provide enough integration for individuals, but and other cross-national deviant behaviour studies
not too much integration where the community (Fajnzylber et al. 2002; Pratt and Godsey 2003;
becomes overwhelming to the individual. Previous Pridemore 2008; Savolainen 2000). Second, the


C 2017 John Wiley & Sons Ltd
Durkheim’s types of suicide and social capital 159

current analysis is the first attempt at connecting counties included in the analysis may not represent
social capital with Durkheim’s types of suicide. the globe. Analyses with a different set of countries
Thus, other measures should be sought out to could come to different conclusions. Thus, better
examine if the findings of the current analysis are measures of suicide and social capital could clear
in line with other research designs. Finally, the up any contradictions in the suicide literature.

References

ADLER, P. S. and KWON, S., 2002. Social Science Research, 29 (1), HADDAD, W. D., CARNOY, M., RINALDI,
Social capital: prospects for a new 148–172. R. and REGEL, O., 1990. Education and
concept. Academy of Management development. World Bank Discussion
Review, 27 (1), 17–40. De COSTER, S., HEIMER, K. and Papers, 95.
WITTROCK, S. M., 2006. Neighborhood
ALESINA, A. and La FERRARA, E., 2000. disadvantage, social capital, street HANSMANN, H. B. and QUIGLEY, J. M.,
Participation in heterogeneous context, and youth violence. 1982. Population heterogeneity and the
communities. Quarterly Journal of Sociological Quarterly, 47, sociogenesis of homicide. Social
Economics, 115 (3), 847–904. 723–753. Forces, 61 (1), 206–224.
ALESINA, A., BAQIR, R. and EASTERLY, DELLER, S. C. and DELLER, M. A., HAWDON, J. and RYAN, J., 2009. Social
W., 1999. Public goods and ethnic 2010. Rural crime and social capital. capital, social control, and changes in
divisions. Quarterly Journal of Growth and Change, 41, 221–275. victimization rates. Crime &
Economics, 114 (4), 1243–1284. Delinquency, 55, 526–549.
DELLER, S. C. and DELLER, M. A.,
AVISON, W. R. and LORING, P. L., 1986. 2012. Spatial heterogeneity, social HELLIWELL, J. F. and PUTNAM, R. D.,
Population diversity and cross-national capital, and rural larceny. Rural 2004. The social context of well-being.
homicide: the effects of inequality and Sociology, 77, 225–253. Philosophical Transactions of the
heterogeneity. Criminology, 24 (4), Royal Society of Biological Sciences,
733–749. DURKHEIM, E., [1897]1951. Suicide: a 359 (1449), 1435–1446.
study in sociology, transl. by J. A.
BESSER, T., 1999. Community Spaulding and G. Simpson. Glencoe, HUANG, W., 1996. Religion, culture,
involvement and perceptions of IL: Free Press. economic and sociological correlates
success small businessoperators in of suicide rates: a cross-national
small towns. Journal of Small Business FAJNZYLBER, P., LEDERMAN, D. and analysis. Applied Economics Letters, 3,
Management, 37 (4), 16–30. LOAYZA, N., 2002. Inequality and 779–782.
violent crime. Journal of Law and
BEYERLEIN, K. and HIPP, J. R., 2005. Economics, 45 (1), 1–40. JACOBS, D. and RICHARDSON, A. M.,
Social capital, too much of a good 2008. Economic inequality and
thing? American religious traditions FLAVIN, P. and RADCLIFF, B., 2009. homicide in the developed nations
and community crime. Social Forces, Public policies and suicide in the from 1975 to 1995. Homicide Studies,
84 (2), 995–1013. American states, Social Indicators 12 (1), 28–45.
Research, 90 (3), 195–209.
CASEY, T. and CHRIST, K., 2005. Social JOHNSON, G. R., KRUG, E. G. and
capital and economic performance in FREY, R. S. and FIELD, C., 2000. The POTTER, L. B., 2000. Suicide among
the American states. Social Science determinants of infant mortality in the adolescents and young adults: a
Quarterly, 86 (4), 826–845. less developed world: a cross-sectional cross-national comparison of 34
test of five theories. Social Indicators countries. Suicide and
CHEW, K. S. and MCCLEARY, R., 1995. Research, 52, 215–235. Life-Threatening Behavior, 30 (1),
The spring peak in suicides: a 74–82.
cross-national analysis. Social Science GIRARD, C., 1993. Age, gender, and
Medicine, 40 (2), 223–230. suicide: a cross-national analysis. KELLY, B. D., DAVOREN, M., MHAOLAIN,
American Sociological Review, 58, A. N., BREEN, E. G. and CASEY, P.,
COLEMAN, J., 1998. Social capital and 553–574. 2009. Social capital and suicide in 11
the creation of human capital. European countries: an ecological
American Journal of Sociology, 94, GOETZ, S. J. and RUPASINGHA, A., 2006.
Wal-mart and social capital. American analysis, Social Psychiatry
S95–S120. Epidemiology, 44, 971–977.
Journal of Agricultural Economics, 88
CUTRIGHT, P. and FERNQUIST, R. M., (5), 1304–1310. KENNEDY, B. P., KAWACHI, I.,
2000. Effects of societal integration, PROTHROW-STITH, D., LOCHNER, K. and
period, region, and culture of suicide GRANOVETTER, M., 1973. The strength
of weak ties. American Journal of GUPTA, V., 1998. Social capital, income
on male age-specific suicide rates: 20 inequality, and firearm violent crime.
developed countries, 1955–1989. Sociology, 78 (6), 1360–1380.


C 2017 John Wiley & Sons Ltd
160 Matthew D. Moore

Social Science & Medicine, 47, MOORE, M. D. and RECKER, N. L., RECKER, N. L. and MOORE, M. D.,
7–17. forthcoming. Social capital, type of 2016. The influence of social capital,
crime, and social control. Crime and social factors, and economic
KNACK, S., 2002. Social capital and Delinquency. [Link] characteristics on suicide rates across
quality of government: evidence from 0011128713510082. U.S. counties. Health Sociology
the states. American Journal of Review, 25 (1), 78–91.
Political Science, 46 (4), 772–785. MOORE, M. D., RECKER, N. L. and
HEIRIGS, M. H., forthcoming. Suicide RICE, T., 2001. Social capital and
KRUGER, D. J., HUTCHISON, P., MONROE, and the creative class. Social government performance in Iowa
M. G., REISCHL, T. and MORREL- Indicators Research. [Link] communities. Journal of Urban
SAMUELS, S., 2007. Assault injury 10.1007/s11205-013-0566-6. Affairs, 23, 375–390.
rates, social capital, and fear of
neighborhood crime. Journal of NEAPOLITAN, J. L., 1996. Cross-national RITZER, G. and GOODMAN, D. J., 2004.
Community Psychology, 35, 483–498. crime data: some unaddressed Sociological theory (6th edition). New
problems. Journal of Crime and York: McGraw Hill.
KUSHNER, H. I. and C. E. STERK, 2005. Justice, 19, 95–112.
The limits of social capital: Durkheim, ROSENFELD, R., MESSNER, S. F. and
suicide, and social cohesion. American NETER, J., WASSERMAN, W. and KUTER, BAUMER, E. P., 2001. Social capital and
Journal of Public Health, 95 (7), M., 1989. Applied liner regression homicide. Social Forces, 80 (1),
1139–1143. models. Homewood, IL: Irwin. 283–310.
LAFREE, G. and KICK, E. L., 1986. OUIMET, M., 2012. A world of SAVOLAINEN, J., 2000. Inequality,
Cross-national effects of development, homicides. The effect of economic welfare state, and homicide: further
distributional, and demographic development, income inequality, and support for institutional anomie theory.
variables on crime: a review and excess infant mortality on the homicide Criminology, 28, 1021–1042.
analysis. International Annals of rate for 165 countries in 2010.
Criminology, 24, 213–235. Homicide Studies, 16 (3), 238–258. STACK, S. and KPOSOWA, A. J., 2008.
The association of suicide rates with
LEE, M. R., 2001. Population growth, PRATT, T. C. and GODSEY, T. W., 2003. individual-level suicide attitudes: a
economic inequality, and homicide. Social support, inequality, and cross-national analysis. Social Science
Deviant Behavior: An Interdisciplinary homicide: a cross-national test of Quarterly, 89 (1), 39–59.
Journal, 22, 491–516. integrated theoretical model.
Criminology, 41 (3), 611–643. TIEPOH, M., NAH, G. and REIMER, B.,
LESTER, D. and LYNN, R., 1993. 2004. Social capital, information
National character, suicide and PRIDEMORE, W. A., 2008. A flows, and income creation in rural
homicide. Personality and Individual methodological addition to the Canada: a cross community analysis.
Differences, 14 (6), 853–855. cross-national empirical literature on Journal of Socio-Economics, 33 (4),
social structure and homicide: a first 427–448.
MESSNER, S., 1982. Societal test of the poverty-homicide thesis.
development, social equality, and Criminology, 46 (1), 133–154. WALL, E., FERRAZZI, G. and SCHRYER,
homicide: a cross-national test of the F., 1998. Getting the goods on social
Durkheimian model. Social Forces, 61 PUTNAM, R. D., 2000. Bowling alone: capital. Rural Sociology, 63 (2),
(1), 225–240. the collapse and revival of American 300–322.
community. New York: Simon and
MESSNER, S. F. and ROSENFELD, R., Schuster. WILSON, J. and MUSICK, M., 1997. Who
1997. Political restraint of the market cares? Toward an integrated theory of
levels of criminal homicide: a PUTNAM, R. D., 2007. E pluribus unum: volunteer work. American Sociological
cross-national application of diversity and community in the Review, 62 (5), 694–713.
institutional-anomie theory. Social twenty-first century: the 2006 Johan
Forces, 75, 1393–1416. Skytte Prize Lecture. Scandinavian WOOLCOCK, M., 2001. The place of
Political Studies, 30 (2), 137–174. social capital in understanding social
MOORE, M. D. and DADAY, J., 2010. and economic outcomes. Canadian
Barriers to human capital development: RECKER, N. L., 2013. Bonds, bridges Journal of Policy Research, 2 (1),
case studies in Swaziland, Cameroon, and quality of life in small towns. 65–88.
and Kenya. African Education Review, Applied Research in Quality of Life, 8
7 (2), 283–304. (1), 63–75.


C 2017 John Wiley & Sons Ltd
Durkheim’s types of suicide and social capital 161

Appendix A: countries used in the analysis

Argentina Estonia New Zealand

Armenia Finland Nicaragua
Austria France Norway
Azerbaijan Germany Panama
Belgium Greece Paraguay
Belize Guatemala Poland
Brazil Hungary Portugal
Bulgaria India Romania
Canada Ireland Russian Federation
Chile Israel Serbia
Colombia Italy Spain
Costa Rica Japan Sweden
Croatia Kazakhstan Switzerland
Czech Republic Korea, Rep. Ukraine
Denmark Kuwait United Kingdom
Dominican Republic Lithuania United States
Ecuador Mexico Venezuela
El Salvador Netherlands


C 2017 John Wiley & Sons Ltd