ResearchPDF Available

The differentiated effects of health on political participation

Authors:

Abstract and Figures

Based on Canadian General Social Survey - Social Identity (2013; N= 27 695) we examined both the direct and indirect effect of self-rated health and self-rated mental health on 1) national voter turnout; 2) local voter turnout and 3) other forms of political participation. Our analysis is the first attempt to take a comprehensive look at the magnitude of health and political participation in a same country on different levels. The results show that health has a different effect on turnout depending the level of government. General health significantly effects electoral participation on the national and municipal level. Mental health effects electoral participation on the national and municipal level and additionally, people who consider their mental health to be poorer, are more likely to sign an online petition. These elements highlight the necessity of questioning the cost of voting according to the level of government, and that further research into the potential offered by Internet and remote voting, is worthwhile—despite the opinions of critics who eschew these means of voting.
Content may be subject to copyright.
AUTHORS
Jérôme Couture
Postdoctoral Fellow
Research Assistant
Université Laval and INRS-UCS
385 Sherbrooke Street West
Montreal (Quebec)
H2X 1E3 Canada
Email: jerome.couture@pol.ulaval.ca
Tel: 514 499-4059
Sandra Breux*
Professor-Researcher
Institut National de Recherche Scientifique
Centre Urbanisation Culture et Société
385 Sherbrooke Street West
Montreal (Quebec)
H2X 1E3 Canada
Email: sandra.breux@ucs.inrs.ca
Tel: 514 499-4059
* Corresponding author
Title: The differentiated effects of health on political participation
ABSTRACT (250 WORDS MAX):
Based on Canadian General Social Survey - Social Identity (2013; N= 27 695) we examined both
the direct and indirect effect of self-rated health and self-rated mental health on 1) national voter
turnout; 2) local voter turnout and 3) other forms of political participation. Our analysis is the
first attempt to take a comprehensive look at the magnitude of health and political participation in
a same country on different levels. The results show that health has a different effect on turnout
depending the level of government. General health significantly effects electoral participation on
the national and municipal level. Mental health effects electoral participation on the national and
municipal level and additionally, people who consider their mental health to be poorer, are more
likely to sign an online petition. These elements highlight the necessity of questioning the cost of
voting according to the level of government, and that further research into the potential offered
by Internet and remote voting, is worthwhiledespite the opinions of critics who eschew these
means of voting.
Key-words: Self-rated health, self-rated mental health, national voter turnout, local voter turnout,
other forms of political involvement, Internet voting
2
1. INTRODUCTION
Several recent studies have shown that an individual’s state of health can significantly affect their
decision to go to the polls(1). People with physical disabilities are less likely to go to vote(2) and
as a consequence, this means that inequalities in health have political consequences in terms of
partisan attachments and electoral participation(3). These studies, while novel and relevant, also
present certain limitations. In our opinion, three aspects of these studies merit closer analysis.
As is the case with the majority of these types of studies, analysis is only done on a single level of
government; the national level. However, prior research has shown that the determining factors
for voting can differ according to the level of government being considered. We know that the
context and significance of institutions weighs heavily in municipal elections whereas the socio-
demographic profile of the electorate is less significant(4, 5). Additionally, the difference
between participation rates on the national and municipal level appear to show that the dynamics
at work are distinct: in Canada and the United States, there is commonly a difference of 15 to 30
points between the two participation rates(6, 7). It seems plausible therefore that the “health”
variable, just like other variables, may act differently according to the level of government.
Another aspect worth delving into is that these analyses focus on whether the cost of voting is
feasible for the elector compared to other forms of electoral participation such as
demonstrations(3). This statement should be nuanced however, if voting is indeed less
demanding than attending a demonstration, there is indeed a difference in “cost” between voting
on a municipal rather than a federal level. The municipal level is often characterized by a lack of
informational context (8), which for some observers, explains the lower participation rate: voting
is more costly for the elector (compared to other levels of government) as political information is
not as well disseminated and also, there is quite simply not as much information to share(9, 10).
The cost of the vote is not sufficiently uniform to distinguish between levels of government
compared to the other forms of participation that associated with different costs.
The last aspect worth further consideration is that a voter’s general state of health is frequently
used as an independent variable to quantify the relationship with electoral participation. Mental
health, however, is not considered as often in the relationship it has with electoral participation,
despite the fact that its impact has already been shown(11). Since mental health is not connected
to social services (12) and social capital is connected to political participation(6), measuring the
relationship between political participation and mental health seems important.
Using information from the Canadian General Social Survey - Social Identity of 2013, we
examined the influence health had on electoral behaviour on the national and municipal levels.
As previous studies focused on a single level of government, our analysis takes a novel approach
as it will help us to get a better understanding of whether health influences electoral behaviour in
accordance with the level of government being elected and then compare electoral participation
with other forms of political engagement. It will also look at the influence of general health as
well as the evaluation of an individual and their mental health.
2. DATA AND METHODS
The Canadian General Social Survey, conducted in 2013 (published in 2015), is an excellent
source of information that allows us to connect stated electoral participation and an individual’s
3
self-assessment of health. The survey was conducted across the entire country with Canadians
aged 15 and older. There were 27,695 respondents in all.
We used three models to analyze the data. In the first model (National Election Turnout, model 1,
2 and 3), the dependant variable considered the respondent's stated national electoral
participation; answering the question: Did you vote in the last federal election? In the second
model (Local Election Turnout, model 4, 5 and 6), the dependant variable considered the
respondent’s stated municipal electoral participation: Did you vote in the last municipal or local
election? In the third model, the dependant variable looked at other forms of political activities
which was divided into three versions represented by the following questions: In the past 12
months, have you done any of the following activities: a) searched for information on a political
issue, b) signed a petition on paper, c) signed an Internet petition. As these are dummy variables,
the participants that responded “yes” were coded 1 and the others were coded 0. Respondents
who did not have the right to vote for one or more levels of government were excluded from
electoral participation analysis.
The two major independent variables were: 1) “Self-reported health” which was measured by the
following question: In general, would you say your health is ...? and 2) “Self-reported mental
health” which was measured by: In general, would you say your mental health is...? The answers
to these questions fell into the following categories: “very good”, “good”, “fair”, “bad” and “very
bad”. These answers were included as dummy variables, with the category “very good” being
used as the reference point. The other independent variables included socio-demographic, socio-
economic and social capital data. We used three measurements for social capital: volunteering
(answer to question: In the past 12 months, did you do unpaid volunteer work for any
organization?); community involvement (answer to the question : In the past 12 months, were
you a member or participant in a union or professional association ; a political party or group ;
a sports or recreational ; organization a cultural, educational or hobby organization ; a
religiousaffiliated group ; a school group, neighbourhood, civic or community association ; a
service club a seniors’ group ; a youth organization ; or an immigrant or ethnic association, and
social relationships "Thinking of [all your friends/your friend], in the past month: how often did
you see [any of your friends/your friend]?" This question had six categories of answers: “not in
the last month”, “Once a month", " 2 or 3 times a month", " Once a week" " A few times a week",
"Every day ". These answers were included as dummy variables with “not in the last month”
acting as the reference point,
Education levels were measured using the following question: What type of educational
institution [are you attending/did you attend]?” The answers to this question were: Elementary,
junior high school or high school; Trade school, college, CEGEP or other nonuniversity
institution, University. This is a dummy variable with the reference value being trade school,
college and CEGEP or other non-university institution. The Canadian General Social Survey did
not directly report on the age of its respondents, only age categories to which they belonged. We
opted to use three categories: under 35, 35 to 64 and 65 and older. These were dummy variables
with the “35 to 65” category being the reference point.
4
3. RESULTS
The results of the multivariate analysis of electoral participation are given in table 1. Models 1 to
3 represent electoral participation in the national elections and models 4-6 reflect participation in
local elections. We began by testing the effect of self-evaluation of general health in conjunction
with the self-evaluation of mental health. We then introduced the socio-economic control
variables We then added variables for social connectedness.
The co-efficient in model 1 showed that an elector in good general health and good mental health
was more likely to vote nationally. In model 4, which addresses municipal participation, this
relationship only held for mental health. When we added the socio-economic variables to model
2, the intensity of the relationship between general health and federal electoral participation
increased significantly. In contrast, the relationship with mental health dropped and only
remained significant for respondents who reported their mental health as being “very bad”. The
addition of control variables to model 5 had a relatively similar effect on municipal participation.
More specifically, the addition of socio-economic factors introduced a positive relationship
between general health and electoral participation which was not seen in model 4. Additionally,
the magnitude of the effect between municipal participation and mental health was less
significant.
In model 3, the introduction of variables to measure social relationships had little effect on the
coefficients of variables related to general health. The significant coefficients of model 2 also
remained significant in model 3. More specifically, we found that there was a slight drop in the
intensity of the relationship between general health and national participation. This supports the
idea that health has an direct effect on electoral participation and that part of the indirect
relationship is related to social connectedness. In contrast, the coefficient of the respondents who
reported poor mental health increased in model 3. We found a similar, albeit more significant,
effect following the introduction of relative variables to the social relationships in model 6 which
looks at municipal participation All the coefficients on general health were reduced. Only two
categories remained significant whereas four categories were significant for model 5.
Additionally, the three coefficients related to mental health that were significant in model 5
increased in intensity in model 6.
This magnitude of effect on health and electoral participation is illustrated in table 2 which shows
the distribution of respondents according to their level of general health
Table 1.
National election turnout
National election turnout
Model 1
Model 2
Model 3
Model 4
Model 6
Constant
1.34***
(0.04)
1.07***
(0.06)
0.53***
(0.09)
0.55***
(0.03)
-0.06
(0.08)
Good health
0.01
(0.05)
-0.09
(0.06)
-0.07
(0.06)
-0.02
(0.04)
-0.10*
(0.046)
Fair health
-0.09
(0.06)
-0.26***
(0.06)
-0.23***
(0.06)
0.03
(0.04)
-0.09
(0.05)
Bad health
-0.15*
(0.07)
-0.43***
(0.08)
-0.32***
(0.08)
-0.006
(0.06)
-0.20**
(0.067)
Very bad health
-0.23*
(0.10)
-0.53***
(0.11)
-0.44***
(0.11)
-0.02
(0.09)
-0.13
(0.10)
Good mental
0.05
(0.05)
0.13**
(0.048)
0.11*
(0.05)
-0.05
(0.04)
-0.02
(0.04)
Fair mental
-0.15**
0.02
0.01
-0.23***
-0.14**
5
(0.052)
(0.06)
(0.06)
(0.04)
(0.047)
Bad mental
-0.45***
(0.08)
-0.10
(0.08)
-0.12
(0.09)
-0.51***
(0.07)
-0.25***
(0.077)
Very bad mental
-0.96***
(0.15)
-0.48**
(0.153)
-0.52***
(0.159)
-0.82***
(0.14)
-0.43**
(0.15)
Age (65 +)
1.16***
(0.06)
1.17***
(0.06)
0.79***
(0.05)
Age (Under 35)
-0.89***
(0.04)
-0.92***
(0.04)
-0.90***
(0.05)
Gender (male)
0.05
(0.04)
(0.08)
(0.04)
-0.08***
(0.03)
Education
(No diploma)
-0.52***
(0.06)
-0.45***
(0.06)
-0.18***
(0.05)
Education
(University
degree)
0.54***
0.47***
(0.05)
0.03
(0.04)
Co-habitation
0.41***
(0.05)
0.39***
(0.04)
0.36***
(0.03)
Employed
-0.01
(0.04)
-0.03
(0.04)
-0.08*
(0.037)
Volunteering
0.31***
(0.04)
0.49***
(0.04)
Community
involvement
0.29***
(0.04)
0.22***
(0.03)
Meet friends:
once a month
0.32***
(0.08)
0.13
(0.07)
Meet friends:
several times a
month
0.39***
(0.08)
0.32***
(0.07)
Meet friends:
once a week
0.23**
(0.08)
0.18**
(0.07)
Meet friends:
several times a
week
0.32***
(0.07)
0.26***
(0.06)
Meet friends:
every day
0.31***
(0.08)
0.26***
(0.07)
Log likelihood
-10.717
-9.778
-9.219
-14.168
-12.467
Observations
22.497
22.255
21.336
22.100
20.955
Note: standard errors in parentheses, *** p < 0.001; ** p < 0.01; * p < 0.05. General Health, Mental Health, Meeting Friend,
Volunteering, Community involvement, Age, Gender, Education, Co-habitation and Employed are dummy variables with the
following reference categories: Very good health, very good mental, not in the last month, not volunteer in the past year, is not a
member of an organization in past years, 35-64, Male, Diploma but lower than university degree, does not live with spouse or law
spouse, does not work.
Mental health and the likelihood of participating in elections on various levels of health as
estimated from models 3 and 6. The estimates show a difference between the two levels of
government. The effect of general health is higher for national electoral participation than it is for
municipal participation. More specifically, there was a reduction in participation of 6.2
percentage points between the respondents who reported very good health and those who
reported being in very bad health. We also found that the estimated drop in national electoral
participation is 4.3 percentage points for respondents who reported their health as being “bad”
and 3.0 percentage points for those whose general state of health was rated as being “fair”. The
differences are less significant however for local participation. The only significant difference
was 4.2 percentage points for respondents who reported their general state of health as being
“bad”.
6
In terms of the magnitude of effect on mental health, we found the opposite to be true; with it
being more significant on a local level. More specifically, there was a reduction in participation
of 9.1 percentage points for local elections between the respondents who reported “very good”
health and those who reported being in “very bad” health. Additionally, the estimated drop in
participation was 5.3 percentage points for those who reported having “bad” mental health and
2.9 points for those who responded “fair”. The differences are less significant however for
national participation. The only significant difference was 8.2 percentage points between the
respondents who reported having “very good” mental health and those who reported being in
“very bad” health for national participation.
Table 2.
Self-assessed
Health
reports
National election turnout
Local election turnout
Share of
respondents
(%)
Probability of
voting
(%)
95 %
confidence
intervals
Share of
respondents
(%)
Probability
of voting
(%)
95 %
confidence
intervals
Very Good
health
21.4
83.4
82.3-84.6
21.5
67.6
66.2-69.1
Good
health
37.9
82.5
81.7-83.4
37.8
65.5
64.4-66.5
Fair
health
28.0
80.4
79.4-81.4
28.0
65.7
64.5-66.9
Bad
health
9.6
79.1
77.3-80.9
9.6
63.4
61.3-65.6
Very bad
health
3.1
77.2
74.0-80.4
3.1
65.0
61.3-68.8
Very good
mental
33.2
81.2
80.2-82.2
33.3
67.0
65.8-68.1
Good
mental
38.6
82.8
82.0-83.6
38.4
66.6
65.5-67.6
Fair
mental
22.0
81.3
80.2-82.4
22.1
64.1
62.7-65.5
Bad
mental
5.1
79.1
77.3-80.9
5.1
61.7
58.9-64.4
Very bad
mental
1.1
73.0
67.7-78.3
1.1
57.9
51.4-64.4
Table 3 shows the relationship between general health and mental health with other forms of
political participation. The first column of table 3 looks at political information research, the
second looks at the signing a petition in a paper form and the third looks at signing an online
petition. The three models include all the variables presented in table 1. The results show that
health has a differentiated effect according to the type of political participation. General health
had no effect on searches for political information and signing a paper petition. However, it did
have an effect on signing an online petition. More specifically, the better the respondent’s
reported state of health, the more likely they were to sign an online petition. This relationship was
significant for respondents in three of the four categories in general health (good, fair, and bad).
7
Table 3.
Seek political
information
Sign a petition
on paper
Sign a petition
online
Constant
-1.64***
(0.08)
-2.71***
(0.10)
-2.30***
(0.09)
Good health
0.04
(0.04)
0.02
(0.05)
-0.09*
(0.045)
Fair health
-0.06
(0.04)
-0.02
(0.06)
-0.17***
(0.05)
Bad health
-0.11
(0.06)
-0.06
(0.08)
-0.22**
(0.08)
Very bad health
-0.07
(0.10)
0.10
(0.13)
-0.10
(0.11)
Good mental
-0.04
(0.04)
0.07
(0.04)
0.15***
(0.05)
Fair mental
-0.09*
(0.04)
0.15**
(0.05)
0.26***
(0.06)
Bad mental
0.23***
(0.07)
0.18*
(0.09)
0.52***
(0.09)
Very bad mental
-0.01
(0.15)
0.20
(0.19)
0.75***
(0.159)
Age (65 +)
-0.36***
(0.04)
-0.09
(0.05)
-0.32***
(0.05)
Age (Under 35)
0.71***
(0.03)
-0.22***
(0.05)
0.01
(0.04)
Gender (male)
0.46***
(0.03)
-0.01
(0.04)
-0.08*
(0.03)
Education
(No diploma)
-0.68***
(0.05)
-0.64***
(0.06)
-0.98***
(0.07)
Education
(University
degree)
0.86***
(0.03)
0.07
(0.04)
0.63***
(0.03)
Co-habitation
-0.17***
(0.03)
0.01
(0.04)
-0.08*
(0.04)
Employed
0.002
(0.03)
0.06
(0.04)
0.10*
(0.04)
Volunteering
0.51***
(0.03
0.52***
(0.04)
0.41***
(0.03)
Community
involvement
0.53***
(0.03)
0.79***
(0.05)
0.60***
(0.04)
Meet friends:
once a month
0.10
(0.07)
0.12
(0.10)
0.23**
(0.08)
Meet friends:
several times a
month
0.27***
(0.07)
0.31***
(0.09)
0.36***
(0.08)
Meet friends:
Once a week
0.16*
(0.06)
0.20*
(0.09)
0.15
(0.08)
Meet friends:
several times a
week
0.23***
(0.07)
0.45***
(0.08)
0.39***
(0.07)
Meet friends:
every day
0.29***
(0.07)
0.41***
(0.09)
0.27***
(0.08)
Log likelihood
-15.244
-10.851
-12.132
Observations
25.695
25.574
25.561
Note: standard errors in parentheses, *** p < 0.001; ** p < 0.01; * p < 0.05. General Health, Mental Health, Meeting Friend,
Volunteering, Community involvement, Age, Gender, Education, Co-habitation and Employed are dummy variables with the
following reference categories: Very good health, very good mental, not in the last month, not volunteer in the past year, is not a
member of an organization in past years, 35-64, Male, Diploma but lower than university degree, does not live with spouse or law
spouse, does not work.
8
In terms of mental health, it had an inverse effect to the results found for electoral participation.
More precisely, the better the respondents reported state of mental health, the less likely they
were to research political information and sign a paper or online petition. The most significant
effect was found for signing an online petition with all of the mental health categories being
statistically significant. Regarding marginal effects, which were not given in the table, the
probability of signing an online petition was above 12.7 percentage points for those who reported
“very bad” mental health compared to those who reported being in “very good” mental health.
The estimated higher probability is 8.4 percentage points for those who reported “bad” mental
health, 3.9 points for those who responded “fair” and 2.2 points for the respondents who
answered that they were in “good” mental health.
Diagram 1 (left side) shows the marginal effects produced using model 6 which looked into the
local electoral participation of respondents according to different categories of mental health
including general health and age. The diagram shows the probability on the local level increases
with age, however the difference between respondents who reported being in “very good” mental
health and those who reported being in “very bad” mental health remained relatively constant
between the age categories.
The right side of the diagram shows the marginal effects in model 3 for signing a petition
according to the different categories of mental health according to different categories of mental
health and including general health and age. The diagram shows the probability of signing a
petition decreases with age, especially for respondents aged 65 and up, however, the difference
between those who reported being in “very good” mental health and those who reported being in
“very bad” mental health remained relatively constant between the age categories. These results
show that age does not truly effect the relationship between health and political participation.
.3 .4 .5 .6 .7 .8
Probability Local Voting
Under 35 35-64 65 and over
Age
very poor poor
fair good
very good
Predictive Margins of Age by Mental Health with 95% C.Is
.15 .2 .25 .3 .35 .4
Probability Sign Petition Online
Under 35 35-64 65 and over
Age
very poor poor
fair good
very good
Predictive Margins of Age by Mental Health with 95% C.Is
9
4. CONCLUSIONS
Our research confirms previous studies that established a connection between a person’s state of
health and electoral participation. However, our study stands apart from the other studies because
it highlights how the variable of health influences electoral behaviour of the voter according to
the type of political participation. More precisely, the general state of health influences electoral
participation nationally, however this relationship is less striking when it comes to voting in
municipal elections. For Nakhaie (6), the connection between belonging to a community and
voting is expressed more strongly on the municipal rather than the federal level. However, how
does this finding explain the reduced impact on the general state of health for local participation?
When considering the municipal level, the community seems better mobilized to “turn out” to
vote; or perhaps social pressure is simply better expressed on this level.
Mental health also clearly influences municipal electoral participation. Perhaps the specificities
of the municipal context might contribute to explaining this relationship. Three particularities
appear to mark the Canadian municipal scene: an absence of political parties, which makes
visibility on issues unclear; an unclear understanding by citizens regarding the jurisdiction of the
municipal level and the fact that there is a lack of media coverage for municipal elections(10).
Municipal voting is more susceptible to being a costly act for the voter as it requires more effort
to keep informed and have a proper understanding of the municipal tier of government. Given
this context, the following question is worth considering: is the cost of voting too high for people
who consider their mental health to be poor?(11)
This connection could explain our third finding: perhaps people in a poorer state of mental health
are more likely to sign and online petition because it is a less “costly” political act. An alternative
answer may be that people in a poorer state of mental health are generally more socially isolated.
An online petition does not require as much social capital whereas the act of voting remains a
distinctly social act. This leads us to believe that the initiatives launched over the last few years to
incorporate electronic voting in several Ontario municipalities may constitute a worthy pursuit to
stimulate electoral participation on the municipal level. In fact, this might also be equally
applicable nationally to attenuate the negative effects of general health on electoral participation.
Much more research is still required on this subject and more experiments should be done on
electronic voting(13) (as many have been unsatisfactory) to gain a better understanding of the
connections between health in all of its incarnations and electoral participation.
0
10
Key-points
Health has a differentiated effect on electoral participation according to the level of
government.
Health in general has a greater effect on national electoral participation than on municipal
electoral participation.
Mental health has a greater effect on local electoral participation than on national electoral
participation.
People who reported having worse mental health are more likely to sign online petitions
than others.
In the sphere of politics as well as health, the cost of voting should take into consideration
the potential offered by electronic political engagement as it could contribute to
increasing political participation and may provide positive effects for the participating
individual.
References
1. Pacheco J, Fletcher J. Incorporating Health into Studies of Political Behavior: Evidence
for Turnout and Partisanship. Political Research Quarterly. 2015;68(1):104-16.
2. Schur L, Shields T, Kruse D, Schriner K. Enabling democracy: disability and voter
turnout. Political Research Quaterly. 2002;55(1):167-90.
3. Mattila M, Söderlund P, Wass H, Rapeli L. Healthy voting : The effect of self-reported
health on turnout in 30 countries. Electoral Studies. 2013;32:886-91.
4. Trounstine J. All politics is local: The reemergence of the study of city politics.
Perspectives on Politics. 2009;7(3):611-8.
5. Caren N. Big City, Big Turnout? Electoral Participation in American Cities. Journal of
Urban Affairs. 2007;29(1):31-46.
6. Nakhaie MR. Electoral participation in municipal, provincial and federal elections in
Canada. Canadian Journal of Political Science. 2006;39(2):363.
7. Hajnal Z, Lewis P. Municipal Institutions and Voter Turnout in Local Elections. Urban
Affairs Review. 2003;38(5):645-68.
8. Elmendorf C, Schleicher D. Districting for a Low-Information Electorate. Yale Law
Journal. 2012;121(7):1846-86.
9. Cutler F, Matthews J. The Challenge of Municipal Voting : Vancouver 2002. Canadian
Journal of Political Science. 2005;38(2):359-82.
10. Breux S, Bherer L. Les élections municipales au Québec : enjeux et perspectives. Québec:
Presses de l'Université Laval; 2011.
11. Denny KJ, Doyle OM. "...Take up thy bed, and vote" Measuring the relationship between
voting behaviour and indicators of health. European Journal of Public Health. 2007;17(4):400-1.
12. Kawachi I, Berkman L. Social Ties and Mental Health. Journal of Urban Health : Bulletin
inf the New Tok Academy of Medicine. 2001;78(3):458-67.
13. Goodman N, Pammett JH, DeBardeleben J. Internet Voting : The Canadian Municipal
Experience. Canadian Parliamentary Review. 2010;Automn 2010:13-21.
Article
This study examined associations between diagnoses with five chronic health conditions (diabetes, cancer, heart disease, asthma, and arthritis) and turnout in the 2012 US presidential election. We used cross-sectional survey data from 16 states from the 2013 and 2014 Behavioral Risk Factor Surveillance System. We estimated a logistic regression model with the main dependent variable as a survey item asking respondents if they voted. We also estimated logistic regression models stratified by race/ethnicity to assess whether the chronic health condition–turnout relationships varied within each racial/ethnic group. Results show that individuals diagnosed with diabetes were 7 percentage-points more likely to vote that those who were not. Stratified models revealed these diabetes–turnout relationships are particularly strong among those who identified as Hispanic and multiracial. Other health characteristics demonstrated consistency with previous literature, including lower self-rated health being associated with lower odds of turnout. Our research suggests an intriguing new relationship between the experience of diabetes and a higher propensity to vote and that different chronic health conditions have varying associations with the likelihood to vote, implying that some groups are more vulnerable to being underrepresented in politics.
Article
Depression is the most common mental health disorder. It has consequences not only on individuals but also on social and political levels. We argue that depressive symptoms impair political participation by reducing the motivation and physical energy of sufferers. We test our hypotheses by conducting regression analyses of four nationally representative cross-sectional and longitudinal surveys that collectively span many democracies. Our results are threefold. First, we find that the severest depressive symptoms lower the probability of voting by 0.05–0.25 points, an effect that is exceeded only by education and age. Second, we show that depressive symptoms negatively affect political interest and internal efficacy, thereby confirming that they diminish political motivation. Third, we find that depressive symptoms most strongly affect physically demanding acts, thereby confirming that they reduce the physical energy required for participation. We conclude by urging scholars to take depressive symptoms seriously in the study of political behavior.
Article
Full-text available
Although the genetic influence on voter turnout is substantial (typically 40–50%), the underlying mechanisms remain unclear. Across the social sciences, research suggests that ‘resources for politics’ (as indexed notably by educational attainment and intelligence test performance) constitute a central cluster of factors that predict electoral participation. Educational attainment and intelligence test performance are heritable. This suggests that the genotypes that enhance these phenotypes could positively predict turnout. To test this, we conduct a genome-wide complex trait analysis of individual-level turnout. We use two samples from the Danish iPSYCH case–cohort study, including a nationally representative sample as well as a sample of individuals who are particularly vulnerable to political alienation due to psychiatric conditions (n = 13,884 and n = 33,062, respectively). Using validated individual-level turnout data from the administrative records at the polling station, genetic correlations and Mendelian randomization, we show that there is a substantial genetic overlap between voter turnout and both educational attainment and intelligence test performance.
Article
Full-text available
This study tested factors related to collective action in East Asian and Western countries. Although the social identity model of collective action (SIMCA) could explain which factors relate to collective action in Western countries, previous studies have found that some of the SIMCA variables did not predict collective action in non‐Western countries. We hypothesised that the internal locus of control would be positively related to collective action even in East Asian countries because previous studies have demonstrated that the former can positively influence proactive behaviour when factors inhibiting such behaviours exist. We analysed the 2010 World Values Survey data from seven countries (three East Asian and four Western countries). The results demonstrated that the internal locus of control was positively related to collective action more so among East Asians (Koreans, specifically those with low political interests, the Taiwanese, and Japanese people with lower income) than among Westerners. These results suggest that the internal locus of control is a key individual difference that predicts collective action in the presence of factors that inhibit such participation (e.g., cultural values, low political interest, and low income).
Article
Full-text available
Depression is the most common mental illness and its profound impact on cognition and decision-making has implications for political judgement. However, those implications are unclear in the case of referendums offering a choice between status quo and change. On one hand, one component of depression is the kind of life dissatisfaction associated with voting for change. Yet cognitive models also portray depression sufferers as biased towards the status quo: they are less inclined to research change, more pessimistic about its benefits and more likely to exaggerate its potential costs. In this paper, we use data from Understanding Society to examine the impact of those cross-pressures on support for Brexit. Prior to the referendum, while life dissatisfaction and generally poor health predicted support for Leaving the EU, those diagnosed with depression were disproportionately likely to support Remain. Supporting our claim that the latter was a sign of status quo bias, this difference disappeared once the result was in and leaving the EU had become the widespread expectation. The study highlights the unexplored importance of mental health for political judgements, emphasises the multidimensionality of conditions like depression, and illustrates the psychological role of status quo bias in referendum voting.
Article
Poor health is generally believed to cause political passivity. Prior studies that satisfactorily acknowledge the causality problems involved are mainly limited to considering turnout and the U.S.A., so we lack knowledge of how non-electoral participation is affected in other countries. This article considers Sweden, characterized by a generous welfare state and an extensive public health system. Using unique panel data, which allow more thorough analyses of causality, poor health was found to have a negative effect on voting but not on non-electoral participation. By primarily focusing on other countries than Sweden and the U.S.A., it is a task for future longitudinal research to show whether the belief that poor health lead to political passivity is incorrect—or whether Sweden is an exceptional case, due to the barriers to participation being particularly low there.
Article
Recently, awareness of the importance of health in explaining political participation has grown considerably. Studies have focused on individual participation forms but not on broader participation modes. Furthermore, analyses of the mechanisms explaining the health effects have been lacking. Here, structural equation models are employed to study the relationship between health, political trust, and institutional and non-institutional participation using data from Finland. Poor health is related to increased non-institutional participation, while good health boosts traditional institutional participation, although the latter relationship is very weak. These observations are explained by differences in political trust. Those in good health have stronger trust in the political system, while poor health is connected with reduced trust. These differences manifest themselves in varying behaviour. Poor health decreases trust which leads to increased non-institutional participation, while good health leads to a high trust and institutional activities.
Article
Taking advantage of China's send‐down movement as a natural experiment, we investigate how experiencing a political movement during adolescence affects political participation decades later. Using data from the China Family Panel Survey and the regression discontinuity design, we find that the send‐down experience significantly reduces individuals’ political participation, measured by their participation in community committee elections, time spent on community service activities and how much they care about public news. Further analysis suggests that the send‐down experience negatively affects political participation through poorer mental health and less trust in local government.
Article
Full-text available
This study examines the association between self-reported health and the propensity for supporting citizens’ initiatives in Finland. Democratic innovations such as the citizens’ initiative provide novel ways for citizens to express their preferences, but whether people in poor health make use of such possibilities remains unclear. The data come from the Finnish National Election Study (FNES2015), a cross-sectional representative sample of the Finnish population. The results suggest that self-reported health affects the propensity to sign citizens’ initiatives, but the effect depends on age since it mobilizes young citizens in poor health, whereas the impact on older generations is negligible.
Article
Previous studies show that people with poor health have a lower propensity to vote. With individual-level register data on sickness allowance episodes and voting in three Finnish elections, we address the following questions: (1) What degree of sickness allowance days negatively influences turnout? (2) Are sickness absences on election day more harmful than absences that occur before the elections? (3) What is the effect of cumulative sickness allowance spells before the elections over a period of several years? We use a threefold categorisation approach, which differentiates between immediate, short-term and long-term health effects on voting. The results show that multiple sickness allowance spells over several years are more strongly connected to turnout than health problems experienced only in the year prior to the elections. Falling ill at the time of the elections had no consistent additional negative relationship with voting. We suggest that the demobilising effects of immediate health problems are associated with tangible factors, while long-term effects are related to lowered levels of political efficacy, interest and social connectedness.
Article
Full-text available
We argue that research on political behavior, including political participation, public opinion, policy responsiveness, and political inequality will be strengthened by studying the role of health. We then provide evidence that self-rated health status (SRHS) is associated with voter turnout and partisanship. Using the National Longitudinal Study of Adolescent Health (Add Health) and General Social Survey (GSS), we find that people who report excellent health are more likely to vote and more likely to identify with the Republican Party. Moreover, the effects of health on voter turnout and partisanship appear to have both developmental and contemporaneous components. Taken together, our findings suggest that health inequalities may have significant political consequences.
Article
Full-text available
Most commentary on redistricting is concerned with fairness to groups, be they racial, political, or geographic. This Essay highlights another facet of the redistricting problem: how the configuration of districts affects the ability of low-information voters to secure responsive, accountable governance. We show that attention to the problem of voter ignorance can illuminate longstanding legal-academic debates about redistricting, and that it brings into view a set of questions that deserve our attention but have received little so far. District designers should be asking how alternative maps are likely to affect local media coverage of representatives, as well as the “branding” strategies of political party elites. Bearing these questions in mind, we offer some tentative suggestions for reform.
Article
Although low voter turnout in national elections has garnered considerable attention and concern, much lower turnout in municipal elections has often been largely ignored. Using a survey of cities in California, this article examines a series of institutional remedies to low turnout in mayoral and city council elections. Moving local elections to coincide with the dates of national elections would have by far the largest impact on voter turnout, but other institutional changes that tend to raise the stakes of local elections also increase turnout. Specifically, less outsourcing of city services, the use of direct democracy, and more control in the hands of elected rather than appointed officials all tend to increase turnout.
Article
How likely are the millions of Americans with disabilities to participate in politics? What insights do their experiences provide into overall participation levels and determinants? This article reports the results of a nationally representative household telephone survey of 1,240 peoplestratified to include 700 people with disabilities-following the November 1998 elections. Voter turnout is found to be 20 percentage points lower among people with disabilities than among people without disabilities who have otherwise-similar demographic characteristics. Other standard predictors of turnout such as political efficacy and mobilization explain only a small portion of this gap. There is great variation within the disability sample: the lower turnout is concentrated among people with disabilities who are not employed or who are age 65 or older, who have had recent onset of a disabling condition, and who have difficulty going outside alone (despite the availability of absentee ballots). The findings suggest that disability, apart from imposing resource constraints, often has social and psychological effects that decrease voter turnout through decreased social capital and identification with mainstream society, particularly among senior citizens. The findings also support the idea that general mobility and major life transitions can be important influences on voter turnout in general, and raise questions on the causal relations among age, employment, efficacy, and voter turnout that should be a focus of future research.
Article
The purpose of this paper is to highlight the importance of social relations or social capital for voting turnout at three levels of Canadian government, paying particular attention to social contexts, socio-demographics and socio-economic forces. The data source is the Public Use Microdata File from the National Survey of Giving, Volunteering and Participation, administered by Statistics Canada (2001). Results provide support for social capital theory. Those who donate to charities and/or volunteer have a stronger propensity to vote than their counterparts. Two other measures of social capital, social networks and participation in religious activities, are also related to turnout. However, their effects are comparatively modest. Among the social bases of social capital, community rootedness is an important predictor of turnout. Civic engagement or attentiveness to current affairs also significantly increases voter turnout at all levels of Canadian government. Finally, standard socio-economic and demographic predictors of political participation do show independent effects on turnout. However, with the exception of age, these predictors are not as consistent or as strong as social capital measures in explaining turnout. Theoretical and policy implications of the findings are discussed.
Article
The study of local politics has been relegated to the periphery of political science and many explanations have been offered for the marginalization of the subfield. I offer three related arguments for why scholars should revisit the study of sub-state politics. First, the local level is the source of numerous political outcomes that matter because they represent a large proportion of political events in the United States. Secondly, there are methodological advantages to studying local politics. Finally, analyzing politics at the sub-state level can generate thoroughly different kinds of questions than a purely national-level focus and can offer different answers to questions that apply more generally. Research on local politics can and should contribute to broader debates in political science and ensure that we understand both how and why cities are unique.
Article
As compared with federal and provincial elections, municipal elections in Canada present voters with challenges of informational quantity and quality. These unique challenges have implications for the psychological structure of citizens' voting calculus. Using a survey of voters conducted after the city of Vancouver civic election of 2002, we estimate a model of vote choice for mayor. We show that voters respond to the different context in predictable ways: their choices are determined largely by ideological orientations and provincial partisanship, with local economic evaluations and local issues playing only a very small role.
Article
This article seeks to describe and explain variation in voter turnout in American big city municipal elections using data from 332 mayoral elections in 38 large U.S. cities over 25 years. In my cross-sectional time-series analysis of turnout in mayoral elections, I find that city-level demographic factors are only weakly correlated with turnout. By contrast, institutional and campaign factors explain much of the variation. The effect of Progressive era reforms on depressing turnout is greatest in the most competitive elections. I conclude by discussing the implication of the overall downward trend in turnout and changes cities can make to increase participation.
Article
It is generally agreed that social ties play a beneficial role in the maintenance of psychological well-being. In this targeted review, we highlight four sets of insights that emerge from the literature on social ties and mental health outcomes (defined as stress reactions, psychological well-being, and psychological distress, including depressive symptoms and anxiety). First, the pathways by which social networks and social supports influence mental health can be described by two alternative (although not mutually exclusive) causal models-the main effect model and the stress-buffering model. Second, the protective effects of social ties on mental health are not uniform across groups in society. Gender differences in support derived from social network participation may partly account for the higher prevalence of psychological distress among women compared to men. Social connections may paradoxically increase levels of mental illness symptoms among women with low resources, especially if such connections entail role strain associated with obligations to provide social support to others. Third, egocentric networks are nested within a broader structure of social relationships. The notion of social capital embraces the embeddedness of individual social ties within the broader social structure. Fourth, despite some successes reported in social support interventions to enhance mental health, further work is needed to deepen our understanding of the design, timing, and dose of interventions that work, as well as the characteristics of individuals who benefit the most.