Content uploaded by Ryan T Cragun
Author content
All content in this area was uploaded by Ryan T Cragun on Mar 09, 2022
Content may be subject to copyright.
11
Introduction
There are a number of studies that suggest religiosity is related to happiness
or well-being and that less religious or nonreligious individuals are not as
happy as are religious individuals (Fenelon and Danielsen 2016). However,
the findings on the relationship between religiosity and happiness are mixed,
with some studies finding that higher levels of religiosity are related to
higher levels of happiness (Krause 2003), others find the opposite (Brown
and Tierney 2009), some find a more complex relationship (Diener, Tay,
and Myers 2011), and some find no relationship at all (Dahl and Galliher
2010; Edling, Rydgren, and Bohman 2014; Norton et al. 2006). While this
chapter is supposed to focus on nonreligion and happiness, readers will note
that most of the focus of the chapter is actually on the relationship between
religiosity and happiness. We have done this because our argument is that
there is not a substantial or meaningful relationship between either religi-
osity or nonreligiosity and happiness. By illustrating that higher levels of
religiosity are not meaningfully correlated with higher levels of happiness,
we ultimately illustrate that nonreligious individuals are not less happy.
This chapter begins with a review of prior research on the relationship
between religiosity and happiness. In reviewing that literature, we focus
on the limitations, namely: (a) much of the research focuses on the United
States; (b) there are very few representative studies that find a positive rela-
tionship between religiosity and happiness; (c) most of the studies that find
a relationship between religiosity and happiness fail to detail that, while the
relationship is statistically significant, the effect size is negligible or small,
meaning the relationship does not matter; (d) there are very few studies that
illustrate a causal relationship between religiosity and happiness; and (e)
very few studies propose a theoretical explanation for why there should be
a relationship between religiosity and happiness. We build on our literature
review by examining a large, international dataset, the World Values Survey,
exploring the relationship between three aspects of religiosity – identity,
behaviour, and belief – in more than 50 countries. Our goal is to determine
if there is a statistically significant relationship and whether this relationship
Religiosity and happiness
much ado about nothing
Ryan T. Cragun and David Speed
11
DOI: 10.4324/9781003045540-11
10.4324/9781003045540-11
168 Ryan T. Cragun and David Speed
is meaningful. We find that religiosity is significantly related with happiness
in a minority of countries and there is no discernible pattern to why that
relationship exists where it does (we also note several negative associations).
We note that the effect size of religiosity on happiness is universally small
or negligible, meaning that, even when there is a statistical relationship, the
effect does not warrant further scrutiny. We conclude that while some indi-
viduals in some very specific circumstances consider religion beneficial and
believe religion contributes to their happiness, there is no reason to believe
that religion should be considered a clinical treatment for nonreligious indi-
viduals. In short, the line of research suggesting that religiosity boosts hap-
piness is more wishful thinking than empirical reality.
Literature review
There are many studies that suggest higher levels of religiosity are related
to higher levels of happiness. This has been shown among older black and
white individuals in the United States (Krause 2003), among older Mexican
immigrants to the United States (Krause and Bastida 2011), and for elderly
individuals in Vietnam (Tran et al. 2017). Religiosity has also been shown
to be related to lower levels of depressive symptoms among US individu-
als with chronic health conditions (Lucette et al. 2016), for elderly Korean
immigrants to the United States (Mui and Lee 2014), and for female mem-
bers of the LDS Church in Utah but not for male members of the faith
(Norton et al. 2006).
The reader may note something odd about this cross-section of findings:
the relationships described are extremely particular. It is not that ‘religi-
osity = happiness’, it is more along the lines of ‘provided X, Y, and Z are
true, then religiosity may = happiness’. It is quite uncommon to find studies
that show a clear relationship between religiosity and happiness in general
populations rather than in very specific, niche groups. Studies that have
looked at broad populations often find very different results. Diener et al.
(2011), using large samples from the United States and around the world,
find a much more complex relationship between religiosity and subjective
well-being. While there is a small effect of religiosity on well-being in the
United States, the results are mixed around the world. They find that, in
less developed countries with high levels of religiosity, people who are reli-
gious are happier. But in developed countries where people are less reli-
gious and life circumstances are not as challenging, there are negligible or
no differences in happiness between religious and nonreligious individuals.
They also note that there appears to be a contextual component to this
relationship: highly religious individuals in highly religious societies benefit
from their fit with prevailing social norms while nonreligious individuals in
highly religious societies suffer from their lack of fit with prevailing norms.
Subsequent studies have found similar results. Stavrova et al. (2013), using
multi-level models and data from the fourth and fifth waves of the World
Religiosity and happiness 169
Values Survey (WVS), find support for the person-fit idea; when a society
values religiosity, religiosity contributes to peoples’ happiness. But that is
not the case when societies do not value religiosity. Similarly, Speed and
Fowler (2017), using a large, representative sample from Ontario, Canada,
and, in a more recent study, a representative sample of Canadian atheists
in several provinces (Speed in press), find that atheists do not substantively
differ from religious groups in terms of health outcomes.
These somewhat conflicting findings lead us to consider the limita-
tions of prior research on the relationship between religiosity and happi-
ness. The first limitation we note is that much of the research on religiosity
and happiness uses data from the United States. The United States does
differ somewhat from other highly developed Western countries in that it
remains more religious (Chaves 2011; Voas and Chaves 2016). But there
is a growing realisation in the social sciences that much of the research
using data from the United States and other highly developed countries is
not representative of the rest of the world (Haidt 2013; Henrich, Heine,
and Norenzayan 2010). Thus, when research finds that religious individu-
als in the United States have higher levels of happiness, as is the case for
many studies in this area of research (Childs 2010; Lim and Putnam 2010;
Pollner 1989; Williams et al. 2020; Yang and Park 2019), there are reasons
to be sceptical that such relationships and findings will generalise from the
United States to other countries. Prior research has found that data from the
United States uniquely yield a statistically significant relationship between
religiosity and happiness, which is not the case in other countries (Diener
et al. 2011; Edling, Rydgren, and Bohman 2014; Stavrova et al. 2013). For
instance, while studies have found a positive relationship between religiosity
and happiness among the elderly in the United States (Ballew et al. 2012;
Krause 2003), other research has found the opposite relationship among the
elderly in China (Brown and Tierney 2009). Likewise, Edling et al. (2014)
found that religiosity was unrelated to happiness after controlling for social
networks in Sweden. Assuming what may be true in the United States is true
around the world is a serious limitation of prior research on the relationship
between religiosity and happiness.
Another limitation of prior research is that many of the studies that find
a positive relationship between religiosity and happiness are in subsets of
a population but not in the general population (Ellison and Fan 2008).
This is akin to saying that everyone would benefit from electro-convul-
sive therapy, anti-depressants, or LSD even though these have only been
shown to help individuals suffering from major depression (Dierckx et al.
2012; Dos Santos et al. 2016). Studies have found a positive relationship
between religiosity and happiness/well-being among older people in the
United States (Ballew et al. 2012; Krause 2003); among religious black
female survivors of hurricanes in the United States (Chan, Rhodes, and
Pérez 2012); among Mexican-origin adults living in Fresno, California
(Ellison et al. 2009; though religious attendance had detrimental effects
170 Ryan T. Cragun and David Speed
on well-being); for a sample of 100 older Korean immigrants to the
United States living in New York but not for older Chinese individuals
(Mui and Lee 2014); among a small sample of black Americans (Ellison
and Flannelly 2009; religious attendance and church support were unre-
lated to depression, only perceived guidance from religion was correlated
with depression); and among a sample of older black individuals living
in Hillsborough County, Florida (Jang et al. 2006). If the relationship
between religiosity and happiness was robust and pervasive, there should
be numerous representative studies illustrating the positive benefits of
religiosity on well-being.
To illustrate this limitation, imagine you have a headache and you want
to take an over-the-counter medication to relieve your symptoms. When
reading the dosage instructions you realise that the medication works 20%
of the time for 5% of the people and no one is sure why this is the case: how
confident would you be in this medication? Even if it did work for you spe-
cifically, would you recommend it to friends, families, and strangers? The
answer should be obvious: of course not. One of the few studies we were
able to find that used population-wide data to examine the relationship
between religiosity and health (Ellison and Fan 2008) utilised a measure
of spirituality that is flawed when used with nonreligious participants – the
Daily Spiritual Experiences Scale (DSES) – which exhibits substantial meas-
urement variance when answered by nonreligious individuals (Hammer and
Cragun 2019). In other words, the nonreligious ‘interpret’ the questions dif-
ferently from the religious. Additionally, Ellison and Fan provide no meas-
ures of the amount of variation explained by their problematic measure of
spirituality and do not report effect sizes.
This leads to the next limitation with much of the prior literature on
religiosity and happiness – the failure to examine the magnitude of any rela-
tionships found. While establishing the existence of a relationship statis-
tically is important, determining the magnitude of the relationship is far
more important. An analogy may help illustrate the importance. Finding
that a relationship is statistically significant is like winning the lottery –
if you find that a relationship between two variables is below one of the
standard alphas used to determine if a relationship is statistically significant
(e.g. p < .05), you have won. However, far more important than winning is
how much someone won. The effect size in this analogy is how much you
won in the lottery. For instance, in 2020 US dollars, a negligible effect size
might be winning US$ 2; a small effect size might be winning US$ 100; a
medium effect size could be winning US$ 1,000; and a large effect would be
winning US$ 1 million or more. Winning US$ 2 is still winning, but no one
cares about winning US$ 2; most people would care if they won $1 million.
Studies that only report that they won the lottery without telling readers
how much they won are highly suspect. Yet, this is precisely the situation
with most of the prior research examining the relationship between religios-
ity and happiness/well-being.
Religiosity and happiness 171
For instance, in their population-wide survey, Ellison and Fan (2008)
do not report measures of variation explained (R2) or effect sizes. Thus,
while they may have found some statistically significant relationships, there
is no indication what the magnitude of those relationships are. Ellison et al.
(2009), using three measures of religiosity – importance of religion, religious
attendance, and seeking spiritual comfort – in a regression with standard
demographic measures looking at depression, find that all of these vari-
ables (religiosity measures and demographics) account for 8% of the vari-
ation in depression (R2 = .084), which is an F2 of .091 or a small effect size.
However, they do not report the contribution of the religiosity variables
separately, leaving the reader to wonder whether the variation explained is
due to the demographic variables or the religiosity measures. Williams et al.
(2020), looking at the contributions of social support and religious support
on depression, find that religious social support accounts for between 2%
and 3% of the variation explained beyond the demographic variables in
their models. Chan et al. (2012) find that religiosity had no direct effect on
distress but was mediated by optimism and purpose. They do not ultimately
calculate the effect of religiosity on distress nor do they indicate how much
of the variation in their dependent variable is accounted for by religiosity.
However, their zero-order correlation suggests that religiousness accounted
for a paltry 0.8% of distress. Krause (2003) finds that church attendance and
private prayer along with a number of demographic variables account for
3.7% of the variation in life satisfaction. His more complex variable of reli-
gious meaning (which is really a measure of meaning, generally)1, accounts
for 5.5% of the variation in life satisfaction, which would be a small effect
size (F2 = .058). In Krause’s most comprehensive models, he accounts for
just 10% of the variation in life satisfaction (combining demographic and
religiosity measures).
In Pollner’s (1989) study of the relationship between religiosity on well-
being, he finds that church attendance accounts for just .5% of the variation
in global happiness, none of the variation in life satisfaction or life excite-
ment, and 1.4% of the variation in marital happiness. Likewise, Pollner
finds that divine relations account for 2% of the variation in global hap-
piness, 2.3% of the variation in life satisfaction, 2.6% of the variation in
life excitement, and .4% of the variation in marital happiness. Collectively,
Pollner is able to account for between 2.3% and 9.6% of the variation in
the outcomes examined. Converted to effect sizes, these are all negligible.
We could continue describing the problems with studies examining the
relationship between religiosity and happiness/well-being but we hope the
point is clear. Many studies do not provide a clear indication of how much
of the variation in the dependent variable – happiness, well-being, depres-
sion – is accounted for with measures of religiosity. When measures of
reduction in error (i.e. R2) are provided, they are typically for an entire
model that includes measures other than religiosity. In the rare instance
when measures of reduction in error are provided for religiosity variables,
172 Ryan T. Cragun and David Speed
the effect sizes are typically negligible or occasionally small. Yet, in all of the
above papers just examined in detail, the authors conclude that religiosity is
an important contributor to happiness, well-being, or depression when their
data indicate just the opposite.
Another problem with most of the research in this area is that the data
are cross-sectional and not longitudinal. In those cases the authors could
only argue correlation and not causation. While several studies have used
longitudinal designs, their results are not impressive. Chan et al. (2012),
looking at pre-disaster religiosity’s effects on post-hurricane distress, found
no direct effect of religiosity, only an indirect effect on optimism. Ballew et
al. (2012), using the problematic DSES, found that DSES was unrelated to
depression for men one or two years later but was related to life satisfaction
one and two years later; DSES was not related to depression or life satis-
faction for women one or two years later. The authors do not report pro-
portionate reduction in error measures for their equations. Childs (2010),
using lagged data analyses to try to determine a causal relationship between
happiness and religious service attendance, finds that the effect of religious
attendance on happiness is .024, which is a small effect. While Childs (2010)
did attempt to control for social support/capital in their model, unvalidated
proxy measures were employed, which likely indicates the 2.4% figure is an
overestimate. Overall there is mixed and scant evidence for a causal rela-
tionship between religiosity and happiness.
This leads to our final concern with existing research on the relationship
between religiosity and happiness – the lack of a clear theoretical argument
for why there might be a relationship. Much of the research on this topic
describes vaguely the possibility that religious attendance might bolster
social support (Williams et al. 2020). Alternatively, some of the research
suggests religiosity might provide a sense of meaning that will then lead to
greater happiness (Krause 2003). In both of these theoretical scenarios, it
is not actually religiosity that is contributing to higher levels of happiness,
rather it is something else that religiosity might contribute to. Religion is not
necessary for social support or for meaning: both of these can and often are
derived outside of a religious context. Thus, for much of the prior research
examining the relationship between religiosity and happiness, the theorised
causal pathway is indirect and only suggests that religion might facilitate
higher levels of meaning or social support. Thus, there is no clear theoretical
reason to expect that higher levels of religiosity will contribute directly or
meaningfully to higher levels of happiness.
Based on our familiarity with this literature, we are sceptical that religi-
osity matters for happiness or well-being. When a body of literature con-
sistently finds small or negligible effects and much of the data come from
obscure populations, for objective scholars there is a pretty obvious con-
clusion – the relationship that is being pursued doesn’t really matter or
doesn’t really exist (Schuurmans-Stekhoven 2011). In many ways this body
of research is akin to searching for Bigfoot. People keep looking and looking
Religiosity and happiness 173
and occasionally will find large footprints or hair or maybe even snap a
fuzzy picture that purports to show Bigfoot. Yet, no one seems to be able
to provide compelling evidence of Bigfoot. Likewise, what is missing from
the research on religiosity and happiness are large, robust studies with rep-
resentative samples of country-level populations that show a clear, compel-
ling relationship with large effect sizes. Such studies don’t exist precisely
because religiosity does not meaningfully influence happiness in a way that
is universal or substantial.
To be clear, we are not suggesting in this chapter that religion cannot
contribute to someone’s happiness or well-being or help someone who is
depressed. That is certainly possible. Our primary argument in this paper is
that there is no reason to believe that religiosity will have large, broad effects
on the happiness of a population. Inversely, there is no reason to believe that
nonreligious individuals will be less happy than are religious people.
Data
Given our discussion above and the focus on using data primarily from a
single country, the United States, we turned to a well-known data set, the
World Values Survey (WVS), to examine the relationship between religios-
ity and happiness. The WVS has been repeated six times, starting in 1981.
We use the sixth wave of the WVS, which was fielded from 2010 to 2014.
The sixth wave includes data from 60 countries. However, the three inde-
pendent variables we examined in this chapter were not included in sur-
veys in some countries. None of the independent variables were included
in the survey fielded in Egypt, two dependent variables were not included
in the survey for Kuwait, and one dependent variable was not included in
the surveys fielded in Morocco and Qatar. Thus, our analyses exclude these
countries where data are missing. A complete list of the countries analysed
is shown in Table 11.1.
Methods
The WVS includes a question designed to capture happiness. Participants
are asked to complete the following statement, ‘Taking all things together,
would you say you are…’ They are then given four response options: (1)
not at all happy, (2) not very happy, (3) rather happy, or (4) very happy
(we reversed the coding so higher values indicate greater happiness). This
measure is our dependent variable.
We examined three independent variables that capture religious belief,
behaviour, and identity. The question we drew upon to explore whether reli-
gious identity is related to happiness asked participants was ‘Independently
of whether you attend religious services or not, would you say you are …’.
Participants were then given the following three options: (1) a religious per-
son, (2) not a religious person, or (3) an atheist. Given the response options
174 Ryan T. Cragun and David Speed
Table 11.1 Analysis of happiness by religious identity (WVS Wave 6)
Religious
(in %)
Not religious +
atheists (in %)
Chi-Square p-value H-B adjusted
alpha
Cramer’s V
Algeria 79.7 79.0 0.045 0.832 0.006 0.007
Argentina 87.2 88.6 0.382 0.537 0.002 0.020
Armenia 82.3 79.4 0.520 0.471 0.002 0.022
Australia 95.7 92.7 5.811 0.016 0.001 0.063
Azerbaijan 80.9 82.2 0.210 0.647 0.003 0.014
Bahrain 72.6 72.2 0.011 0.916 0.010 0.003
Belarus 69.8 66.0 2.286 0.131 0.001 0.040
Brazil 92.3 89.8 1.648 0.199 0.002 0.034
Chile 86.0 83.5 1.210 0.271 0.002 0.035
China 85.9 85.7 0.010 0.919 0.013 0.002
Colombia 92.5 87.3 7.532 0.006 0.001 0.071
Cyprus 80.2 87.0 6.412 0.011 0.001 0.080
Ecuador 91.5 96.8 10.738 0.001 0.001 0.095
Egypt na na na na na na
Estonia 74.1 77.6 2.134 0.144 0.001 0.039
Georgia 69.6 69.7 0.000 0.994 0.050 0.000
Germany 86.8 86.7 0.011 0.915 0.008 0.002
Ghana 82.6 81.0 0.075 0.784 0.005 0.007
Hong Kong 93.4 88.2 4.559 0.033 0.001 0.068
India 86.9 78.2 40.952 0.000 0.001 *** 0.086
Iraq 69.3 64.1 1.924 0.165 0.002 0.042
Japan 90.7 88.3 2.078 0.149 0.001 0.033
Jordan 86.9 80.5 6.276 0.012 0.001 0.072
Kazakhstan 88.7 88.8 0.003 0.957 0.025 0.001
Kuwait 92.9 93.7 0.260 0.610 0.003 0.015
Kyrgyzstan 96.4 91.2 7.266 0.007 0.001 0.070
Lebanon 78.1 79.0 0.109 0.742 0.004 0.010
Libya 88.5 85.5 2.828 0.093 0.001 0.038
Religiosity and happiness 175
Malaysia 96.8 95.2 2.379 0.123 0.001 0.043
Mexico 94.4 93.8 0.220 0.639 0.003 0.011
Morocco 77.0 82.0 1.525 0.217 0.002 0.037
Netherlands 94.9 91.6 7.759 0.005 0.001 0.065
New Zealand 95.2 94.5 0.185 0.667 0.003 0.016
Nigeria 83.9 81.9 0.190 0.663 0.003 0.010
Pakistan 83.8 100.0 0.578 0.447 0.002 0.022
Palestine 77.1 68.9 6.798 0.009 0.001 0.083
Peru 76.6 76.4 0.005 0.945 0.017 0.002
Philippines 90.1 87.7 1.188 0.276 0.002 0.031
Poland 93.5 94.6 0.192 0.661 0.003 0.014
Qatar 98.2 96.8 0.594 0.441 0.002 0.024
Romania 67.5 68.4 0.077 0.782 0.005 0.007
Russia 76.9 74.3 1.744 0.187 0.002 0.029
Rwanda 90.2 95.2 1.750 0.186 0.002 0.034
Singapore 94.6 92.1 4.529 0.033 0.001 0.051
Slovenia 81.8 86.2 2.894 0.089 0.001 0.054
South Africa 78.3 75.3 3.116 0.078 0.001 0.030
South Korea 93.1 87.8 8.017 0.005 0.001 0.082
Spain 84.3 87.5 2.317 0.128 0.001 0.045
Sweden 96.8 94.8 2.205 0.138 0.001 0.044
Taiwan 88.8 92.3 4.414 0.036 0.001 0.061
Thailand 93.8 93.2 0.144 0.705 0.004 0.011
Trinidad 88.4 86.8 0.372 0.542 0.002 0.020
Tunisia 78. 79.8 0.137 0.711 0.004 0.011
Turkey 86.2 76.5 15.390 0.000 0.001 *** 0.099
Ukraine 72.6 68.2 2.893 0.089 0.001 0.045
Uruguay 84.8 87.8 1.771 0.183 0.002 0.043
US 92.7 87.5 15.851 0.000 0.001 *** 0.085
Uzbekistan 96.4 96.6 0.026 0.872 0.006 0.004
Yemen 71.5 70.8 0.020 0.888 0.007 0.005
Zimbabwe 82.7 74.3 4.738 0.029 0.001 0.056
*** Relationship is statistically significant after adjusting alpha using Holm-Bonferroni correction.
176 Ryan T. Cragun and David Speed
for this question, we chose to analyse it as a nominal or categorical variable.
We also chose to examine the relationship between this variable and the
happiness variable at the bivariate level only. Prior research (Cragun 2019)
has illustrated that this is an extremely problematic question for a variety
of reasons. An atheist who is not a religious person will be forced to choose
between response options 2 and 3. We combined ‘not a religious person’
together with ‘an atheist’ so we had just two response options: (a) a religious
person, and (b) not a religious person and/or an atheist. For this analysis we
also collapsed the happiness question into two categories: very and rather
happy were combined as were not very happy and not at all happy. We
used Chi-Square for the analysis and Cramer’s V to examine effect sizes,
interpreting the effect sizes in line with recommended values: .10 = small,
.30 = medium, .50 = large (Rea and Parker 2014).
The second independent variable asked participants, ‘How important
is God in your life?’ They were then allowed to choose a value from a
scale ranging from (1) Not at all important to (10) Very important. Prior
research has used a similar variable for analysing relationships between reli-
gious beliefs and health outcomes (Homan and Boyatzis 2010), including
happiness (Childs 2010). Given the response options, we chose to analyse
this variable as continuous. We initially ran just a simple bivariate analysis
between this variable and the dependent variable (shown in Table 11.2),
then re-ran the analysis controlling for various demographic variables: age,
sex, social class, educational attainment, and political views. In our models,
we report both R2, an indicator of the amount of variation explained in the
dependent variable by the independent and control variables, and Cohen’s
F2, which is a measure of effect size (Cohen 1988). Cohen’s F2 can be inter-
preted as follows: .02 = small, .15 = medium, and .35 = large.
The third independent variable asked participants, ‘Apart from weddings
and funerals, about how often do you attend religious services these days?’
We reverse coded the response options to facilitate interpretation so they
were as follows: (1) Never, practically never; (2) Less often; (3) Once a year;
(4) Only on special holy days; (5) Once a month; (6) Once a week; (7) More
than once a week. We followed the same analytical approach for this vari-
able as we did for importance of belief in God.
For readers unfamiliar with effect sizes, which will figure prominently in
our results and discussion below, we offer a short explanation. In a simple
sense, something is statistically significant when two estimates and their
confidence intervals (typically 95%) do not overlap with each other. When
something is statistically significant, it literally means there is a non-zero
difference between two estimates. Statistical significance could indicate that
Parameter A is 0.0005% different from Parameter B; it could also mean
that Parameter A is 5000.00% different from Parameter B. The kicker is
that all differences will become statistically significant if you include enough
people in a model. So, the way we quantify the difference between groups is
with effect size, which indicates the strength of an association or the size of
Religiosity and happiness 177
Table 11.2 Happiness regressed on importance of belief in God (WVS, Wave 6)
Importance of belief in God only Importance of belief in God with control variables
b†p-value H-B
adjusted
alpha
R2Cohen’s
F2
b†p-value H-B
adjusted
alpha
R2Cohen’s
F2
Algeria –0.020 0.248 0.0022 0.001 0.001 –0.008 0.699 0.0050 0.064 0.068
Argentina 0.004 0.573 0.0083 0.000 0.000 0.002 0.792 0.0071 0.052 0.055
Armenia 0.018 0.128 0.0019 0.002 0.002 0.007 0.647 0.0042 0.087 0.096
Australia 0.011 0.012 0.0014 0.004 0.004 0.007 0.155 0.0019 0.039 0.040
Azerbaijan 0.014 0.524 0.0063 0.000 0.000 0.001 0.680 0.0045 0.053 0.056
Bahrain –0.010 0.341 0.0026 0.001 0.001 –0.002 0.845 0.0125 0.010 0.010
Belarus 0.015 0.028 0.0015 0.003 0.003 0.021 0.002 0.0012 0.091 0.100
Brazil 0.015 0.169 0.0019 0.001 0.001 0.022 0.071 0.0015 0.026 0.027
Chile 0.010 0.292 0.0024 0.001 0.001 0.015 0.141 0.0018 0.079 0.085
China –0.015 0.002 0.0012 0.005 0.005 0.143 0.000 0.0009 *** 0.058 0.061
Colombia 0.011 0.429 0.0033 0.000 0.000 0.009 0.565 0.0036 0.010 0.010
Cyprus 0.018 0.125 0.0018 0.002 0.002 0.032 0.007 0.0013 0.076 0.083
Ecuador –0.008 0.482 0.0038 0.000 0.000 –0.004 0.763 0.0063 0.033 0.034
Egypt na na na na na na na na na na
Estonia 0.002 0.704 0.0125 0.000 0.000 0.009 0.134 0.0017 0.181 0.221
Georgia 0.068 0.000 0.0010 *** 0.019 0.019 0.019 0.256 0.0022 0.156 0.184
Germany 0.003 0.557 0.0071 0.000 0.000 –0.001 0.885 0.0167 0.072 0.078
Ghana 0.025 0.298 0.0025 0.001 0.001 0.018 0.457 0.0029 0.038 0.040
Hong Kong 0.012 0.070 0.0017 0.003 0.003 0.008 0.249 0.0021 0.027 0.028
India 0.032 0.000 0.0009 *** 0.012 0.012 0.033 0.000 0.0009 *** 0.043 0.044
Iraq 0.010 0.711 0.0167 0.000 0.000 0.049 0.048 0.0015 0.176 0.214
Japan 0.021 0.000 0.0011 *** 0.008 0.008 0.011 0.071 0.0016 0.081 0.089
Jordan –0.051 0.009 0.0013 0.006 0.006 0.083 0.000 0.0010 *** 0.043 0.045
Kazakhstan –0.005 0.489 0.0045 0.000 0.000 –0.006 0.342 0.0026 0.029 0.030
(Continued )
178 Ryan T. Cragun and David Speed
Kuwait na na na na na na na na na na
Kyrgyzstan 0.004 0.504 0.0050 0.000 0.000 0.000 0.953 0.0500 0.022 0.022
Lebanon 0.046 0.000 0.0010 *** 0.024 0.025 0.045 0.000 0.0010 *** 0.046 0.049
Libya 0.000 0.973 0.0500 0.001 0.001 –0.015 0.264 0.0023 0.044 0.046
Malaysia 0.057 0.000 0.0009 *** 0.030 0.031 0.051 0.000 0.0009 *** 0.067 0.072
Mexico 0.019 0.019 0.0015 0.003 0.003 0.024 0.004 0.0012 0.024 0.025
Morocco 0.025 0.371 0.0029 0.001 0.001 –0.009 0.803 0.0083 0.101 0.112
Netherlands 0.014 0.001 0.0011 *** 0.006 0.006 0.013 0.006 0.0013 0.046 0.048
New Zealand 0.004 0.456 0.0036 0.001 0.001 –0.002 0.811 0.0100 0.077 0.084
Nigeria –0.021 0.096 0.0017 0.002 0.002 –0.008 0.518 0.0031 0.073 0.079
Pakistan 0.115 0.000 0.0009 *** 0.024 0.025 0.105 0.000 0.0010 *** 0.120 0.136
Palestine 0.056 0.003 0.0012 0.008 0.009 0.057 0.002 0.0012 0.114 0.128
Peru 0.015 0.230 0.0021 0.001 0.001 0.029 0.025 0.0014 0.049 0.052
Philippines –0.009 0.487 0.0042 0.000 0.000 –0.011 0.391 0.0028 0.014 0.014
Poland –0.002 0.792 0.0250 0.000 0.000 0.005 0.580 0.0038 0.040 0.042
Qatar 0.008 0.509 0.0056 0.000 0.000 0.137 0.000 0.0009 *** 0.037 0.038
Romania 0.004 0.681 0.0100 0.000 0.000 0.035 0.001 0.0011 0.123 0.140
Russia 0.010 0.056 0.0016 0.002 0.002 0.014 0.047 0.0014 0.079 0.086
Rwanda –0.047 0.000 0.0009 *** 0.031 0.033 –0.050 0.000 0.0009 *** 0.050 0.053
Singapore 0.025 0.000 0.0010 *** 0.011 0.011 0.083 0.000 0.0009 *** 0.061 0.065
Slovenia –0.006 0.343 0.0028 0.001 0.001 0.009 0.306 0.0025 0.113 0.127
South Africa 0.006 0.371 0.0029 0.000 0.000 –0.001 0.890 0.0250 0.094 0.104
South Korea 0.029 0.000 0.0009 *** 0.025 0.026 0.026 0.000 0.0010 *** 0.070 0.075
Spain –0.007 0.229 0.0020 0.001 0.001 –0.004 0.557 0.0033 0.089 0.098
Importance of belief in God only Importance of belief in God with control variables
b†p-value H-B
adjusted
alpha
R2Cohen’s
F2
b†p-value H-B
adjusted
alpha
R2Cohen’s
F2
Table 11.2 Continued
Religiosity and happiness 179
Sweden 0.011 0.059 0.0016 0.003 0.003 0.007 0.222 0.0020 0.035 0.036
Taiwan 0.019 0.006 0.0013 0.006 0.006 0.016 0.042 0.0014 0.032 0.033
Thailand –0.016 0.013 0.0014 0.006 0.006 –0.010 0.135 0.0017 0.029 0.030
Trinidad 0.083 0.000 0.0011 *** 0.016 0.016 0.078 0.001 0.0011 *** 0.053 0.056
Tunisia –0.059 0.006 0.0012 0.007 0.007 –0.053 0.023 0.0013 0.070 0.075
Turkey 0.055 0.000 0.0010 *** 0.011 0.012 0.046 0.001 0.0011 *** 0.033 0.034
Ukraine 0.010 0.008 0.0013 0.205 0.258 0.013 0.084 0.0016 0.078 0.085
Uruguay 0.018 0.014 0.0014 0.007 0.007 0.029 0.000 0.0011 *** 0.042 0.043
US 0.030 0.000 0.0009 *** 0.019 0.019 0.024 0.000 0.0010 *** 0.073 0.079
Uzbekistan 0.036 0.000 0.0010 *** 0.013 0.013 0.005 0.738 0.0056 0.020 0.020
Yemen –0.126 0.000 0.0011 *** 0.015 0.015 –0.057 0.176 0.0019 0.078 0.084
Zimbabwe –0.022 0.267 0.0023 0.001 0.001 –0.021 0.276 0.0024 0.078 0.084
† Importance of belief in God: 1 = not at all important; 10 = very important.
*** Relationship is statistically significant after adjusting alpha using Holm-Bonferroni correction.
180 Ryan T. Cragun and David Speed
a difference. In a health context specifically, large effect sizes would mean
that a relationship is likely real and likely matters; but a small effect size has
less clear relevance. Finding a small difference in mortality rates may matter
but finding a small difference in a proxy of wellness (e.g. happiness) likely
does not. Finally, given that we are conducting numerous statistical tests,
we employ a Holm-Bonferroni correction with each independent variable to
reduce the risk of committing a Type I error.
Results
Starting with the relationship between religious identity and happiness,
of the 59 countries examined, after our Holm-Bonferroni correction, we
found a statistically significant relationship in just three countries: India, the
United States, and Turkey. The results are shown in Table 11.1. In India,
86.9% of those who reported they are a religious person indicated they
were very or rather happy while 78.2% of the nonreligious and atheists
indicated the same (χ2 = 40.952, p < .001, Cramer’s V = .086). In the United
States, 92.7% of religious individuals indicated they were very or rather
happy while 87.5% of the nonreligious and atheists indicated the same
(χ2 = 15.851, p < .001, Cramer’s V = .085). And in Turkey, 86.2% of reli-
gious individuals indicated they were very or rather happy while 76.5%
of the nonreligious and atheists indicated the same (χ2 = 15.390, p < .001,
Cramer’s V = .099). Thus, out of the 59 countries examined, there was a
multi-comparison adjusted statistically significant relationship between reli-
gious affiliation and happiness in just three countries. However, the effect
sizes for these relationships in all three countries fall below the standard cut-
off for considering the effect small. In other words, while the relationship
between religious identity and happiness is statistically significantly in these
three countries, the difference is so small as to be unimportant.
A careful analysis of Table 11.1 also reveals that a higher proportion of
nonreligious and atheist individuals reported being very or rather happy in
20 of the 59 countries (34%). However, in none of these countries was the
relationship statistically significant after the multi-comparison alpha adjust-
ment. We note this for two reasons. First, if we were not being extremely
careful with our statistical analyses, we could argue that the nonreligious and
atheists are happier than are the religious in some countries (e.g. Cyprus and
Taiwan). Second, that a slightly higher percentage of nonreligious and athe-
ist individuals in just over ⅓ of the countries examined report being happier
than religious individuals strongly suggests that the relationship between
religious identity and happiness is neither universal nor robust. When a rela-
tionship regularly changes directions, with typically negligible effects, the
most obvious conclusion is that there is no meaningful relationship.
Table 11.2 presents the results of two regression analyses exploring to
what extent the importance of someone’s belief in God is related to hap-
piness. The first model in the table is a bivariate analysis that examines
Religiosity and happiness 181
just the relationship between importance of belief in God and happiness.
The second model includes demographic variables. Focusing first on just the
bivariate analyses, in 15 out of the 58 countries (26%), importance of belief
in God was significantly related to happiness after the Holm-Bonferroni
correction. However, in two of those countries, Rwanda and Yemen, the
relationship was the opposite of what might be expected, with individuals
who reported greater importance of their belief in God indicating lower
levels of happiness. Thus, in 13 out of the 58 countries (22%), belief in God
did suggest higher levels of happiness. Of those countries, four had effect
sizes that could be considered small (Lebanon, F2 = .025, Malaysia; F2 = .030;
Pakistan, F2 = .024; South Korea, F2 = .025) and one, Ukraine, had a medium
effect (F2 = .205). Of note, the effect size in Rwanda where the relationship
was reversed was also small (F2 = .031).
Also included in Table 11.2 is the R2, which can be interpreted as the
proportion of the variation in the dependent variable, happiness, that is
explained by the independent variable, importance of God. The variation
explained in most countries is less than 1% (i.e. < .01). The second model
in Table 11.2, which includes demographic variables, also includes both R2
and Cohen’s F2, which provide a useful comparison to illustrate that impor-
tance of one’s belief in God is generally unimportant for happiness. The
demographic variables almost universally account for more variation than
importance of belief in God. For instance, in Chile, importance of belief
in God alone accounts for just 0.1% of the variation in happiness; when
demographic variables are included in the model, they account for 5.8% of
the variation. The one exception is in Ukraine where, with the demographic
variables included in the model, the R2 and F2 are both lower, suggesting
that there may be a statistical anomaly in the dataset. In short, the contribu-
tion of someone’s belief in God to their happiness generally is meaningless
and is undoubtedly less important than age, sex, income, education, and
political views.
Table 11.3 uses an approach similar to Table 11.2, with two sets of
models but with religious service attendance as the dependent variable.
In the first model, happiness is regressed on just religious service attend-
ance. In 14 countries (25%), after adjusting for multiple comparisons,
religious service attendance is significantly correlated with happiness.
However, for almost every country, the R2 is less than .01, as is Cohen’s
F2. In just one country of the 56 where data were available – the United
States – was the effect size small (b = .048, p < .001, R2 = .03, F2 = .031).
In the other 55 countries, the effect size was negligible. The second set
of models include the demographic variables in the regression and, not
surprisingly, the demographic variables account for more variation than
religious service attendance does in every country. For instance, religious
attendance in Germany accounts for 0.6% of the variation in happiness;
the demographic variables plus religious attendance account for 7.5% of
the variation in happiness.
182 Ryan T. Cragun and David Speed
Table 11.3 Happiness regressed on religious service attendance (WVS, Wave 6)
Religious attendance only Religious attendance with control variables
b†p-value H-B
adjusted
alpha
R2Cohen’s
F2
b†p-value H-B
adjusted
alpha
R2Cohen’s
F2
Algeria –0.018 0.032 0.0016 0.004 0.004 –0.018 0.169 0.0017 0.069 0.074
Argentina 0.005 0.609 0.0029 0.000 0.000 0.002 0.859 0.0071 0.051 0.053
Armenia 0.016 0.250 0.0022 0.001 0.001 0.001 0.955 0.0250 0.084 0.092
Australia 0.031 0.000 0.0010 *** 0.011 0.011 0.035 0.000 0.0010 *** 0.050 0.052
Azerbaijan 0.005 0.707 0.0038 0.000 0.000 0.001 0.599 0.0031 0.053 0.056
Bahrain –0.008 0.639 0.0036 0.000 0.000 –0.008 0.968 0.0500 0.009 0.009
Belarus 0.006 0.579 0.0028 0.000 0.000 0.015 0.202 0.0017 0.086 0.094
Brazil 0.035 0.000 0.0010 *** 0.013 0.013 0.047 0.000 0.0009 *** 0.043 0.045
Chile –0.001 0.928 0.0083 0.000 0.000 0.003 0.768 0.0063 0.071 0.076
China‡–0.021 0.071 0.0018 0.001 0.001 –0.021 0.076 0.0016 0.044 0.046
Colombia 0.022 0.009 0.0014 0.005 0.005 0.030 0.001 0.0011 0.018 0.018
Cyprus –0.034 0.006 0.0013 0.008 0.008 –0.007 0.608 0.0036 0.068 0.074
Ecuador 0.001 0.949 0.0100 0.000 0.000 0.008 0.491 0.0028 0.033 0.034
Egypt na na na na na na na na na na
Estonia 0.026 0.015 0.0015 0.004 0.004 0.045 0.000 0.0010 *** 0.190 0.234
Georgia 0.045 0.001 0.0011 *** 0.009 0.009 –0.016 0.274 0.0019 0.156 0.185
Germany 0.028 0.001 0.0011 *** 0.006 0.006 0.019 0.026 0.0014 0.075 0.081
Ghana 0.003 0.840 0.0050 0.000 0.000 –0.001 0.938 0.0167 0.038 0.039
Hong Kong 0.008 0.475 0.0026 0.001 0.001 –0.004 0.720 0.0050 0.026 0.026
India 0.023 0.000 0.0010 *** 0.012 0.012 0.028 0.000 0.0009 *** 0.034 0.035
Iraq –0.002 0.812 0.0045 0.000 0.000 0.021 0.072 0.0015 0.175 0.213
Japan 0.044 0.000 0.0010 *** 0.007 0.007 0.039 0.003 0.0012 0.087 0.096
Jordan‡0.000 0.981 0.0167 0.000 0.000 0.030 0.016 0.0013 0.033 0.035
Religiosity and happiness 183
Kazakhstan 0.010 0.263 0.0024 0.001 0.001 0.007 0.424 0.0024 0.029 0.030
Kuwait na na na na na na na na na na
Kyrgyzstan –0.003 0.626 0.0031 0.000 0.000 –0.005 0.472 0.0026 0.022 0.022
Lebanon 0.005 0.635 0.0033 0.000 0.000 0.006 0.665 0.0045 0.022 0.022
Libya –0.001 0.882 0.0071 0.000 0.000 0.001 0.900 0.0100 0.046 0.048
Malaysia 0.036 0.000 0.0009 *** 0.015 0.016 0.034 0.000 0.0010 *** 0.056 0.059
Mexico 0.026 0.000 0.0011 *** 0.007 0.007 0.027 0.000 0.0011 *** 0.027 0.028
Morocco na na na na na na na na na na
Netherlands 0.024 0.001 0.0012 *** 0.006 0.006 0.022 0.005 0.0012 0.047 0.049
New Zealand 0.020 0.051 0.0016 0.005 0.005 0.008 0.585 0.0029 0.080 0.086
Nigeria 0.012 0.438 0.0025 0.000 0.000 0.005 0.728 0.0056 0.073 0.079
Pakistan –0.039 0.004 0.0013 0.007 0.007 –0.057 0.000 0.0010 *** 0.114 0.129
Palestine –0.002 0.802 0.0042 0.000 0.000 –0.002 0.878 0.0083 0.103 0.115
Peru 0.020 0.069 0.0017 0.003 0.003 0.023 0.052 0.0015 0.044 0.046
Philippines 0.000 0.990 0.0500 0.000 0.000 –0.007 0.639 0.0038 0.014 0.014
Poland 0.030 0.007 0.0014 0.008 0.008 0.030 0.023 0.0013 0.043 0.045
Qatar na na na na na na na na na na
Romania 0.000 0.983 0.0250 0.000 0.000 0.029 0.036 0.0014 0.118 0.134
Russia 0.014 0.094 0.0019 0.001 0.001 0.011 0.328 0.0020 0.083 0.091
Rwanda 0.044 0.000 0.0009 *** 0.015 0.015 0.045 0.000 0.0009 *** 0.032 0.033
Singapore‡0.028 0.000 0.0010 *** 0.009 0.009 0.025 0.000 0.0011 *** 0.045 0.047
Slovenia –0.002 0.866 0.0056 0.000 0.000 0.010 0.462 0.0025 0.111 0.125
South Africa 0.001 0.872 0.0063 0.000 0.000 –0.009 0.285 0.0019 0.096 0.106
South Korea 0.031 0.000 0.0009 *** 0.018 0.018 0.028 0.000 0.0010 *** 0.064 0.068
Spain –0.010 0.259 0.0023 0.001 0.001 0.005 0.643 0.0042 0.088 0.097
Sweden 0.031 0.007 0.0014 0.006 0.006 0.018 0.143 0.0016 0.036 0.037
Taiwan 0.013 0.196 0.0020 0.001 0.001 0.001 0.600 0.0033 0.029 0.030
Thailand 0.041 0.000 0.0011 *** 0.012 0.012 0.033 0.005 0.0012 0.034 0.035
Trinidad 0.033 0.005 0.0013 0.008 0.008 0.014 0.364 0.0023 0.034 0.035
Tunisia 0.000 0.968 0.0125 0.000 0.000 0.001 0.901 0.0125 0.061 0.064
Turkey 0.028 0.003 0.0012 0.006 0.006 0.023 0.035 0.0014 0.030 0.030
(Continued )
184 Ryan T. Cragun and David Speed
Ukraine 0.039 0.002 0.0012 0.007 0.007 0.038 0.002 0.0011 0.082 0.089
Uruguay 0.022 0.062 0.0017 0.004 0.004 0.035 0.006 0.0013 0.035 0.036
US 0.048 0.000 0.0009 *** 0.030 0.031 0.004 0.000 0.0009 *** 0.079 0.086
Uzbekistan –0.013 0.139 0.0019 0.001 0.001 –0.014 0.330 0.0021 0.021 0.022
Yemen –0.010 0.236 0.0021 0.001 0.001 0.034 0.343 0.0022 0.075 0.081
Zimbabwe 0.028 0.023 0.0015 0.003 0.003 0.014 0.250 0.0018 0.078 0.084
† Religious service attendance: 1 = never; 7 = more than once a week.
*** Relationship is statistically significant after adjusting alpha using Holm-Bonferroni correction.
‡ The variable asking about political identity was not asked in this country and is not included in the multivariate model.
Religious attendance only Religious attendance with control variables
b†p-value H-B
adjusted
alpha
R2Cohen’s
F2
b†p-value H-B
adjusted
alpha
R2Cohen’s
F2
Table 11.3 Continued
Religiosity and happiness 185
Discussion
In our analyses, we examined to what extent religious identity, religious
belief, and religious attendance are related to happiness in more than 50
countries around the world using large, representative surveys. In light
of problems we detailed above, we used a Holm-Bonferroni correction to
adjust for multiple comparisons and focused closely on the amount of varia-
tion our independent variables explained as well as effect sizes to determine
how important religiosity is to individual happiness.
Of the 59 countries where we explored the relationship between religious
identity and happiness, the relationship was statistically significant in just
three of them and in not a single country did the effect size reach the clas-
sification of ‘small’. In other words, religious identity is not meaningfully
related to happiness in these 59 countries. Of the 58 countries where we
explored the relationship between religious belief and happiness, the rela-
tionship was statistically significant, but the effect size was negligible in all
but five of those countries and it was small in four of those countries. In
other words, religious belief is, at best, marginally related to happiness in
just a handful of countries around the world. Finally, of the 55 countries
where we examined the relationship between religious attendance and hap-
piness, the relationship was statistically significant in 14 of those countries.
However, in 13 of those 14 countries, the effect size was negligible. In just
one country, the United States, was the effect size small.
This leads us back to the concerns about research on religion and happi-
ness we noted at the beginning of this chapter. The first concern we raised
is that much of the research on the relationship between religiosity and
happiness draws on data from the United States. As our analyses show,
this is extremely problematic. The United States is very much the excep-
tion and not the rule when it comes to religiosity’s correlation with hap-
piness. Religious identity, religious belief, and religious behaviour were all
significantly related to happiness in the United States, but not in most of the
other countries in our analysis. A strong case could be made that the United
States is exceptional when it comes to the relationship between religiosity
and happiness. Yet, much of the prior research on this relationship focuses
exclusively on the United States when arguing that religiosity is related with
happiness (Ellison and Fan 2008; Krause 2003). Our analyses suggest this is
an extremely problematic assertion. Future research should, at a minimum,
recognise that the relationship may be unique to the United States. Future
research on this topic should also include international comparisons with a
variety of countries, particularly countries that vary in both degree of religi-
osity (from more to less religious) and in kind (e.g. predominantly Muslim,
predominantly Buddhist, predominantly atheist, etc.).
The second concern that we noted with prior research on the relationship
between religiosity and happiness is that many of the studies do not use rep-
resentative samples but rather use targeted samples or convenience samples
186 Ryan T. Cragun and David Speed
(Chan et al. 2012; Jang et al. 2006). Our analyses address this concern
by using the World Values Survey. As we illustrate, representative samples
from countries around the world show that, in most countries, there is not
a statistically significant relationship between religiosity and happiness, and
in the few countries where the relationship is statistically significant, for
very few of them is the relationship large enough to be considered even a
small effect. In other words, in large surveys using representative samples
of entire countries, religiosity is largely unrelated to happiness. It is possibly
the case, as much of the prior research on this topic illustrates, that religios-
ity is related to happiness for certain subgroups of populations – older indi-
viduals, racial/ethnic minorities, marginalised individuals, etc. Our study
cannot rule out the possibility that religiosity contributes to the happiness
of some individuals. But, in general, happiness is not reliant upon religiosity.
Our third concern with prior research on the relationship between religi-
osity and happiness is that most of the prior studies focus exclusively on sta-
tistical significance and ignore effect sizes. Our analyses illustrate just how
important effect sizes are in understanding this relationship. For religious
identity, we found three countries where the relationship was statistically
significant, but the effect sizes were negligible in all three of those coun-
tries. For both religious belief and religious service attendance, we found
a statistically significant relationship in about 20–25% of the countries we
examined, but in almost every one of those countries where the relationship
was significant, the effect size was negligible. Our analyses indicate that,
with very few exceptions, religiosity’s relationship with happiness has such
a small effect size that it does not matter.
Unfortunately, our research does not address one of the concerns we
raised about prior research on this topic – the lack of longitudinal data
to illustrate causality. That would be a fruitful area of inquiry for future
researchers if there was a meaningful relationship between religiosity and
happiness. Our analyses can help us address our final concern – the lack
of theorising surrounding the relationship between religiosity and hap-
piness. Prior research examining the relationship between religiosity and
health (Cragun et al. 2016; Speed, Barry, and Cragun 2020) has argued
that religiosity may contribute to social health but there is no reason to
believe that it would or should contribute to other forms of health (e.g.
physical, mental, or sexual health). However, there are some theoretical
reasons why religiosity might be associated with higher levels of happiness.
Ceteris paribus, there is no reason to believe that simply affiliating with a
religion would contribute to happiness except in countries where identifying
with a minority religion can result in persecution or discrimination (Diener
et al. 2011; Eichhorn 2012; Stavrova et al. 2013). In such contexts, it is not
that religious identity increases happiness; it’s that religious individuals dis-
criminate against nonreligious individuals, which reduces the happiness of
nonreligious individuals (Cragun et al. 2012; Edgell, Gerteis, and Hartmann
2006; Heiner 1992).
Religiosity and happiness 187
Why religious belief would contribute to happiness is theoretically
unclear. If the suggestion is that happiness is entirely determined by an indi-
vidual’s beliefs, which research suggests is not the case (Diener and Suh
1997), then holding religious beliefs might contribute to happiness. But that
seems theoretically tenuous at best. We argue that there is no theoretical
reason to think that religious beliefs would contribute to happiness. Finally,
as we have shown elsewhere (Speed et al. 2020), religiousness can increase
social support, and social support has been shown to be related to happiness
(Sehmi et al. 2019). However, once social support is controlled for statisti-
cally, much of the effect of religiosity on health is mediated. We suggest the
same is likely for happiness when and where religious service attendance
is related to happiness. In other words, from a theoretical perspective, if
religious behaviour has any effect on happiness, it is likely an indirect effect
through social support. We did not control for social support in this study.
This, too, would be a good question for future research to examine.
There is a prevalent assertion that religion is associated with salutary
effects, and that higher levels of religiosity are at one end of a continuum
and nonreligion is at the other end of that continuum. That assumption
is not a safe one for a variety of reasons (Hwang, Hammer, and Cragun
2011), and a growing body of research suggests these are two very differ-
ent groups (Galen and Kloet 2010; Thiessen and Wilkins-Laflamme 2020).
While we have made a similar error, we have done so to illustrate there is
no real reason to think that nonreligious individuals are at risk of being less
happy. We can make this assertion based on our findings as we have shown
that, in most of the countries we analysed, the relationship between religios-
ity and happiness either does not exist statistically or is so small as to lack
any degree of importance. Based on our analyses, we suggest the following
questions: should a therapist or counsellor suggest to someone who is non-
religious that they join a religion to boost their happiness? Should they pro-
pose to someone who does not believe in a god/gods to become a religious
believer to improve their happiness? Should they propose to someone who
never attends religious services that they start attending to improve their
happiness? Our results tell us that religiosity’s relationship to happiness is
so marginal that it is subclinical. Nonreligious individuals do not need reli-
gion in order to be happier because there is scant evidence that they are less
happy as it is.
Conclusion
We have shown in this chapter that prior research examining the relation-
ship between religiosity and happiness is quite problematic. Much of the
research draws on data from the United States or uses obscure sub-samples
of populations rather than representative samples. Most of the studies in
this area do not report effect sizes and, when measures of the importance of
religiosity on happiness are included, they indicate the relationship is either
188 Ryan T. Cragun and David Speed
negligible or small. There is very little longitudinal research and virtually
no research building a clear case for a causal relationship for this under-
theorised claimed pathway.
If there is future research examining the relationship between religiosity
and happiness, before it should be taken seriously, it must clear the hurdles
we set out in this chapter. The data should not come exclusively from the
United States and it should not be from an obscure subset of a population.
Additionally, clear measures of effect sizes must be included. If scholars are
able to tease out a causal relationship, that would be noteworthy.
Given our findings and the findings of prior research in this area, there is
no reason to think religion matters for happiness, well-being, or depression.
Nonreligious individuals are not at risk of being less happy than are reli-
gious individuals because religion does not generally matter for happiness.
Note
1 Using measures that purport to capture spirituality or religiosity but actually
capture some aspect of mental health like purpose or meaning in life or healthy
behaviours is fairly common in this line of research and often conflates the inde-
pendent and dependent variables (de Jager Meezenbroek et al. 2012).
References
Ballew, Shoshana H., Susan M. Hannum, Jean M. Gaines, Katherine A. Marx, and
John M. Parrish. 2012. ‘The Role of Spiritual Experiences and Activities in the
Relationship between Chronic Illness and Psychological Well-Being.’ Journal of
Religion and Health 51(4):1386–96. doi: 10.1007/s10943-011-9498-0.
Brown, Philip H., and Brian Tierney. 2009. ‘Religion and Subjective Well-Being
among the Elderly in China.’ The Journal of Socio-Economics 38(2):310–19. doi:
10.1016/j.socec.2008.07.014.
Chan, Christian S., Jean E. Rhodes, and John E. Pérez. 2012. ‘A Prospective Study of
Religiousness and Psychological Distress among Female Survivors of Hurricanes
Katrina and Rita.’ American Journal of Community Psychology 49(1–2):168–81.
doi: 10.1007/s10464-011-9445-y.
Chaves, Mark. 2011. American Religion: Contemporary Trends. Princeton, NJ:
Princeton University Press.
Childs, E. 2010. ‘Religious Attendance and Happiness: Examining Gaps in the
Current Literature—A Research Note.’ Journal for the Scientific Study of Religion
49(3):550–60.
Cohen, Jacob. 1988. Statistical Power Analysis for the Behavioral Sciences. Mahwah,
NJ: Lawrence Erlbaum.
Cragun, Deborah, Ryan T. Cragun, Brian Nathan, J. E. Sumerau, and Alexandra C.
H. Nowakowski. 2016. ‘Do Religiosity and Spirituality Really Matter for Social,
Mental, and Physical Health?: A Tale of Two Samples.’ Sociological Spectrum
36(6):359–77. doi: 10.1080/02732173.2016.1198949.
Cragun, Ryan T. 2019. ‘Questions You Should Never Ask an Atheist: Towards
Better Measures of Nonreligion and Secularity.’ Secularism and Nonreligion 8:1–
6. doi: 10.5334/snr.122.
Religiosity and happiness 189
Cragun, Ryan T., Barry A. Kosmin, Ariela Keysar, Joseph H. Hammer, and
Michael E. Nielsen. 2012. ‘On the Receiving End: Discrimination toward
the Non-Religious.’ Journal of Contemporary Religion 27(1):105–27. doi:
10.1080/13537903.2012.642741.
Dahl, Angie, and Renee Galliher. 2010. ‘Sexual Minority Young Adult
Religiosity, Sexual Orientation Conflict, Self-Esteem and Depressive
Symptoms.’ Journal of Gay & Lesbian Mental Health 14(4):271–90. doi:
10.1080/19359705.2010.507413.
Diener, Ed, and Eunkook Suh. 1997. ‘Measuring Quality of Life: Economic, Social,
and Subjective Indicators.’ Social Indicators Research 40:189–216.
Diener, Ed, Louis Tay, and David G. Myers. 2011. ‘The Religion Paradox: If Religion
Makes People Happy, Why Are so Many Dropping Out?‘ Journal of Personality
and Social Psychology 101(6):1278–90. doi: 10.1037/a0024402.
Dierckx, B., W. T. Heijnen, W. W. van den Broek, and T. K. Birkenhäger.
2012. ‘Efficacy of Electroconvulsive Therapy in Bipolar versus Unipolar
Major Depression: A Meta-Analysis.‘ Bipolar Disorders 14(2):146–50. doi:
10.1111/j.1399-5618.2012.00997.x.
Dos Santos, R. G., F. L. Osório, J. A. Crippa, J. Riba, A. W. Zuardi, and J. E.
Hallak. 2016. ‘Antidepressive, Anxiolytic, and Antiaddictive Effects of
Ayahuasca, Psilocybin and Lysergic Acid Diethylamide (LSD): A Systematic
Review of Clinical Trials Published in the Last 25 Years.‘ Therapeutic Advances
in Psychopharmacology 6(3):193–213. doi: 10.1177/2045125316638008.
Edgell, Penny, Joseph Gerteis, and Douglas Hartmann. 2006. ‘Atheists as ‘Other’:
Moral Boundaries and Cultural Membership in American Society.’ American
Sociological Review 71:211–34.
Edling, Christofer, Jens Rydgren, and Love Bohman. 2014. ‘Faith or Social Foci?
Happiness, Religion, and Social Networks in Sweden.’ European Sociological
Review 30(5):615–26. doi: 10.1093/esr/jcu062.
Eichhorn, Jan. 2012. ‘Happiness for Believers? Contextualizing the Effects of
Religiosity on Life-Satisfaction.’ European Sociological Review 28(5):583–93.
Ellison, Christopher G., and Daisy Fan. 2008. ‘Daily Spiritual Experiences and
Psychological Well-Being among US Adults.’ Social Indicators Research
88(2):247–71. doi: 10.1007/s11205-007-9187-2.
Ellison, Christopher G., Brian K. Finch, Daniell Nicole Ryan, and Jennifer J. Salinas.
2009. ‘Religious Involvement and Depressive Symptoms among Mexican-Origin
Adults in California.’ Journal of Community Psychology 37(2):171–93. doi:
10.1002/jcop.20287.
Ellison, Christopher G., and Kevin J. Flannelly. 2009. ‘Religious Involvement
and Risk of Major Depression in a Prospective Nationwide Study of African
American Adults.’ The Journal of Nervous and Mental Disease 197(8):568–73.
doi: 10.1097/NMD.0b013e3181b08f45.
Fenelon, Andrew, and Sabrina Danielsen. 2016. ‘Leaving My Religion: Understanding
the Relationship between Religious Disaffiliation, Health, and Well-Being.’ Social
Science Research 57:49–62. doi: 10.1016/j.ssresearch.2016.01.007.
Galen, Luke William, and James Kloet. 2010. ‘Mental Well-Being in the Religious
and the Non-Religious: Evidence for a Curvilinear Relationship.’ Mental Health,
Religion & Culture 14(7):673–89. doi: 10.1080/13674676.2010.510829.
Haidt, Jonathan. 2013. The Righteous Mind: Why Good People Are Divided by
Politics and Religion. New York: Vintage Books.
190 Ryan T. Cragun and David Speed
Hammer, Joseph H., and Ryan T. Cragun. 2019. ‘Daily Spiritual Experiences
and Well-Being among the Nonreligious, Spiritual, and Religious: A Bifactor
Analysis.’ Psychology of Religion and Spirituality 11(4):463–73. doi: 10.1037/
rel0000248.
Heiner, Robert. 1992. ‘Evangelical Heathens: The Deviant Status of Freethinkers in
Southland.’ Deviant Behavior: An Interdisciplinary Journal 13:1–20.
Henrich, Joseph, Steven J. Heine, and Ara Norenzayan. 2010. ‘Most People Are Not
WEIRD.’ Nature 466(7302):29–29. doi: 10.1038/466029a.
Homan, Kristin J., and Chris J. Boyatzis. 2010. ‘Religiosity, Sense of Meaning, and
Health Behavior in Older Adults.’ International Journal for the Psychology of
Religion 20(3):173–86. doi: 10.1080/10508619.2010.481225.
Hwang, Karen, Joseph H. Hammer, and Ryan T. Cragun. 2011. ‘Extending Religion-
Health Research to Nontheistic Minorities: Issues and Concerns.’ Journal of
Religion and Health 50(3):608–22.
de Jager Meezenbroek, Eltica, Bert Garssen, Machteld van den Berg, Dirk van
Dierendonck, Adriaan Visser, and Wilmar B. Schaufeli. 2012. ‘Measuring
Spirituality as a Universal Human Experience: A Review of Spirituality
Questionnaires.’ Journal of Religion and Health 51(2):336–54. doi: 10.1007/
s10943-010-9376-1.
Jang, Yuri, Amy R. Borenstein, David A. Chiriboga, Karon Phillips, and James
A. Mortimer. 2006. ‘Religiosity, Adherence to Traditional Culture, and
Psychological Well-Being among African American Elders.’ Journal of Applied
Gerontology 25(5):343–55. doi: 10.1177/0733464806291934.
Krause, Neal. 2003. ‘Religious Meaning and Subjective Well-Being in Late Life.’ The
Journals of Gerontology: Series B 58(3):S160–70. doi: 10.1093/geronb/58.3.S160.
Krause, Neal, and Elena Bastida. 2011. ‘Financial Strain, Religious Involvement,
and Life Satisfaction among Older Mexican Americans.’ Research on Aging
33(4):397–409. doi: 10.1177/0164027511400433.
Lim, Chaeyoon, and Robert D. Putnam. 2010. ‘Religion, Social Networks,
and Life Satisfaction.’ American Sociological Review 75(6):914–33. doi:
10.1177/0003122410386686.
Lucette, Aurelie, Gail Ironson, Kenneth I. Pargament, and Neal Krause. 2016.
‘Spirituality and Religiousness Are Associated with Fewer Depressive Symptoms
in Individuals with Medical Conditions.’ Psychosomatics 57(5):505–13. doi:
10.1016/j.psym.2016.03.005.
Mui, Ada C., and Elsa S. Lee. 2014. ‘Correlates of Depression among Chinese and
Korean Immigrant Elders in the United States.’ Ageing International 39(3):274–
88. doi: 10.1007/s12126-013-9183-6.
Norton, Maria C., Ingmar Skoog, Lynn M. Franklin, Christopher Corcoran, JoAnn
T. Tschanz, Peter P. Zandi, John C. S. Breitner, Kathleen A. Welsh-Bohmer, and
David C. Steffens. 2006. ‘Gender Differences in the Association between Religious
Involvement and Depression: The Cache County (Utah) Study.’ The Journals of
Gerontology: Series B 61(3):P129–36. doi: 10.1093/geronb/61.3.P129.
Pollner, M. 1989. ‘Divine Relations, Social Relations, and Well-Being.’ Journal of
Health and Social Behavior 30(1):92–104.
Rea, Louis M., and Richard A. Parker. 2014. Designing and Conducting Survey
Research: A Comprehensive Guide. 4 edition. San Francisco, CA: Jossey-Bass.
Schuurmans-Stekhoven, James Benjamin. 2011. ‘Is it God or Just the Data that
Moves in Mysterious Ways? How Well-Being Research may Be Mistaking
Religiosity and happiness 191
Faith for Virtue.’ Social Indices Research 100(2):313–330. doi: 10.1007/
s11205-010-9630-7.
Sehmi, R., B. Maughan, T. Matthews, and L. Arseneault. 2019. ‘No Man is an
Island: Social Resources, Stress and Mental Health at Mid-Life.’ The British
Journal of Psychiatry, 4:1–17. doi: 10.1192/bjp.2019.25.
Speed, David. 2021. ‘Godless in the Great White North: Assessing the Health of
Canadian Atheists Using Data from the 2011/2012 Canadian Community Health
Survey.‘ Journal of Religion and Health. doi: 10.1007/s10943-020-01169-3
Speed, David, Caitlin Barry, and Ryan Cragun. 2020. ‘With a Little Help from
My (Canadian) Friends: Health Differences between Minimal and Maximal
Religiosity/Spirituality Are Partially Mediated by Social Support.’ Social Science
& Medicine, 265:1–9. doi: 10.1016/j.socscimed.2020.113387.
Speed, David, and Ken Fowler. 2017. ‘Good for All? Hardly! Attending Church
Does Not Benefit Religiously Unaffiliated.’ Journal of Religion and Health
56(3):986–1002. doi: 10.1007/s10943-016-0317-5.
Stavrova, Olga, Detlef Fetchenhauer, and Thomas Schlösser. 2013. ‘Why Are Religious
People Happy? The Effect of the Social Norm of Religiosity across Countries.’
Social Science Research 42(1):90–105. doi: 10.1016/j.ssresearch.2012.07.002.
Thiessen, Joel, and Sarah Wilkins-Laflamme. 2020. None of the Above: Nonreligious
Identity in the US and Canada. New York: NYU Press.
Tran, Tuyen Quang, Thanh Quy Nguyen, Huong Van Vu, and Tinh Thanh Doan.
2017. ‘Religiosity and Subjective Well-Being among Old People: Evidence from
a Transitional Country.’ Applied Research in Quality of Life 12(4):947–62. doi:
10.1007/s11482-016-9500-9.
Voas, David, and Mark Chaves. 2016. ‘Is the United States a Counterexample to
the Secularization Thesis?‘ American Journal of Sociology 121(5):1517–56. doi:
10.1086/684202.
Williams, Beverly Rosa, Randi M. Williams, Eddie M. Clark, Crystal L. Park, Emily
Schulz, Debarchana Ghosh, and Cheryl L. Knott. 2020. ‘Marital Status and
Depressive Symptoms in African Americans: The Moderating Role of Social and
Religious Resources.’ Journal of Family Issues 42(4):0192513X20942819. doi:
10.1177/0192513X20942819.
Yang, Tse-Chuan, and Kiwoong Park. 2019. ‘Racial/Ethnic Disparities in Depression:
Investigating How Sources of Support and Types of Integration Matter.’ Social
Science Research 82:59–71. doi: 10.1016/j.ssresearch.2019.04.002.
