We are surrounded by examples, from the sublime to the murderous, that illustrate the importance in human life of the systems of belief we construct and take to heart. Michelangelo painted astonishing images on the ceiling of the Sistine Chapel, inspired by beliefs about man's relationship with God. Einstein constructed the Special Theory of Relativity, driven by the conviction that the laws of nature must be the same everywhere.
Terrorists crashed airliners into the World Trade Center and the Pentagon, believing that they did the work of Allah.Whether implicitly or explicitly, human beings live, and die, in accordance with fundamental philosophical premises. In fact, the whole point of being explicit about our philosophical systems is to codify basic principles that guide our understanding of the world and our actions in it. Ultimately, the goal is to live the good life, although exactly what that means itself varies across philosophical systems.
There are many ways to classify the myriad systems of belief about the nature of the world and humanity's relationship to it, but for present purposes a very simple distinction between two broad categories will do. In the accompanying table, I include a third category for the sake of completeness. I will return to it briefly later, but will not characterize it further here.
I've characterized the first category as the Modern/Enlightenment perspective. Originating in the West during the period of the Enlightenment, it establishes personal happiness in this life as the fundamental goal of living. The Declaration of Independence, an Enlightenment document par excellence, codifies this idea in its listing of unalienable rights: "life, liberty, and the pursuit of happiness." The happiness to be achieved is not mere hedonic pleasure but something more akin to Aristotle's eudaimonia, roughly translated as "flourishing." The means to that end is to live in accordance with the best in human nature, which requires the exercise of reason in pursuit of one's chosen goals. Reason proceeds ultimately from the evidence of our senses, which deliver to us the information we need to establish knowledge of the universe in which we find ourselves. By careful observation and the exercise of our rational faculties, human beings can discover a set of basic moral principles that help us to make choices consistent with the goal of achieving long-term happiness and life-satisfaction. A passionate belief in the correctness of these principles motivates action in pursuit of that goal.
Contrast this set of beliefs with the Pre-Modern/Religious one. That differs strongly on two key dimensions. First, one's ultimate goals are not to be found on the corporeal plane at all; life as we all know it is but the means to a transcendent end, to be realized only in a mystical realm where the non-material essence of the person—one's soul—is immortal. Living properly on the corporeal plane prepares one for acceptance into the loftier, more pleasant neighborhoods of the afterlife, whether these be conceived as abstract states of utter un-self-consciousness, as in the case of some Eastern systems, or as actual places where evil, pain, and suffering are unknown. Second, the means of knowing how to live life properly is to surrender one's judgment to the inspirations of faith. The idea is that—either directly, through a personal relationship with God or the Way, or indirectly, through the teachings of other, more enlightened people—one comes to understand God's plan for one's life. God is the authoritative source for the rules by which humans should live. Physical reality and knowledge of it are but transient noise against the background of the truly significant spiritual realm to which all good souls aspire.
Obviously these are not monolithic categories. Muslims and Buddhists share very few specific beliefs about the nature of the non-corporeal realm or the means to achieve access to it. Modernists are in greater accord about basic metaphysics and the means to knowledge but may be seriously at odds in matters of politics and morality. I'll return later to a discussion of how some of these ideological differences might make a psychological difference, but for now it is important to grasp the core distinctions between these two philosophical systems. For those who live a modern, secular life, reasoning one's way to happiness in this life is the essence of humanity. For those who live a pre-modern, religious life, the ultimate goal of everlasting spiritual life can be attained only by surrendering concerns about this world to faith in the powers of otherworldly spirits.
ReligiousPostmodernUltimate GoalHappiness on EarthTranscendental ImmortalitySelf-selected (personal power or "voice"?)MeansReason operating on the evidence of our sensesFaithIntuitive (Social Intelligence?)
Table 1. This is a crude classification of broad approaches to the problem of organizing a philosophical system as a guide to living. Any such system identifies an overarching goal that, if achieved, constitutes the good life. It also describes a procedure—a psychological modus operandi—that, if followed, will maximize the likelihood of achieving the ultimate goal. A more complete version of this classification was presented by David Kelley at The Objectivist Center's 1998 "Objectivism Today" conference: The Real Culture Wars . For details of Kelley's schema, see the December 1998 issue of Navigator, or search for "The Real Culture Wars" at the TOC Web site: www.ObjectivistCenter.org.
Given the otherworldly focus of the latter group, one might expect that the consequences in this world of adhering to a system of religious beliefs would be of little concern. After all, this life is but the means to a greater end. It comes as something of a surprise, then, to find so much attention being given in the academic and popular press of late to the benefits for health and personal happiness of what is awkwardly called "religiosity." The goal of happiness and self-satisfaction is supposed to belong to those of us with a humanist perspective, yet increasingly one sees reports of research findings claiming to show that the more one embraces religion, the greater one's well-being in this life. Table 2 lists the benefits most frequently claimed on behalf of those who live a religious life.
This is an interesting list, given the core philosophical beliefs of the religious. Why is greater longevity on the physical plane desirable for those who believe that real life only begins when one has left it? How do we square the finding that one achieves greater self-esteem by being religious with, for example, the Christian belief that pride is a mortal sin? What are we to make of the finding that one sees a greater sense of self-control and efficacy among people who are told that virtue lies in complete surrender of the course of one's life to God's will? Many of these benefits follow naturally from a modern, humanist philosophy but seem more at odds with a pre-modern, religious one. Nevertheless, many contemporary religious leaders routinely tout these benefits as reasons to reject the nonreligious life in favor of a religious one.
Claimed Benefits of Religiosity
1. Greater physical health and longevity
2. Increased subjective well-being and happiness
3. Greater expressed satisfaction with one's life
4. Higher self-esteem
5. Lower levels of anxiety
6. Lower rates of suicide
7. Less premarital sex
8. Less alcohol and drug consumption
9. Higher school achievement
10. A greater sense of efficacy and control
Table 2. Researchers have claimed a number of benefits of living a religious life. These are the most frequently mentioned in recent reports in the popular press. All are based on research studies reported in academic journals of psychology, religion/theology, and sociology.
What is the quality of the evidence on which they are based? As with much research involving complex human behavior and thought, the quality of the evidence is highly variable. Most of the research is correlational in character. That is, one attempts to quantify religiosity in some way, then procures measures of health, self-esteem, well-being, life-satisfaction, and so forth. One then uses various correlational techniques in an effort to discover how much of the variability in the things we wish to predict—such as health and well-being—is explained by our predictors (religiosity, in this case).
One of the first things one notices in reading this literature is how often the variables of interest are quantified in simple, unidimensional ways. For example, religiosity is commonly quantified as the response to a single question, typically about something like frequency of church attendance. 1 Similarly, measures of well-being may consist of asking respondents to indicate on a six- or seven-point scale how satisfied they are with their lives or how happy they are overall. It is of course an empirical question whether these simple measures actually capture the variables of interest, and the literature is filled with controversy over just this question.
Even if such simple measures capture something essential, however, the correlational character of the research means that the results based on those measures are open to multiple interpretations. Proponents of the religious perspective want to claim that the evidence reveals God's presence in the lives of the faithful. But there are more straightforward explanations. Most obviously, it could be that effects of religiosity on things like anxiety, happiness, and well-being are indirect, resulting from something other than the system of religious beliefs itself.
One perfectly plausible alternative account of the data is that church attendance, a very common measure of religiosity, is by its nature a social activity. Going to a place of worship puts one in contact with other, like-minded people in a setting built around concrete and predictable rituals in which one shares. This increases one's sense of connection to a community, strengthens one's social network, and thus builds into one's life a reservoir of support and stability that is available in times of stress. It also tends to increase the level of one's activity on behalf of the group (volunteerism). All these things are well known—from the laboratory literature in psychology and from the sociological literature—to increase the degree of one's resistance to stress and anxiety, thereby increasing physical health, sense of security, and one's sense of satisfaction with life.
Thus, there is an alternative explanation, entirely consistent with the evidence based on measures of church going, that it isn't the set of religious beliefs per se that matters. Rather, it might just as well be the set of actions that one takes as a result of engagement in a community of fellow believers. No otherworldly mysticism is implicated. Being part of a religious community confers concrete benefits entirely attributable to ordinary social-psychological processes, not to the intervention of God in people's lives. The mystical beliefs serve to motivate social behaviors that are the actual causes of positive outcomes. That's the sense in which the effects of religiosity are indirect. Other patterns of activity can confer the same benefits so long as they tap into these basic social-psychological mechanisms. To bring the point home, consider that the benefits listed in Table 3 are those that supposedly accrue to owners of cats and dogs.
In this case, of course, there's nothing at all mystical about the processes involved.
Claimed benefits of…?
1. Fewer illnesses (colds, flu, etc.)
2. Increased subjective well-being and happiness
3. Lower levels of stress and anxiety
4. Increased survival rates following heart attacks
5. Reduced sense of loneliness, isolation
6. Greater sense of purpose
7. More frequent positive encounters with others
8. Increased attention span in learning-disabled kids
9. Increased sense of responsibility in children
Table 3. Activities other than church attendance can have effects that are remarkably similar on overall health, well-being, and life-satisfaction. Try to guess the alleged source of the benefits above before checking the answer in the above paragraphs.
Wanted: A More Direct Measure
Variables in Ellison study
Life-SatisfactionHappinessAge*n Sex Education Income***p Non-White Married**p***pRural-Urban South
Traumatic Event*n**nVoluntary Associations**p Conservative Protestant Liberal Protestant*p Moderate Protestant Catholic *nMormon/Jehovah's Witnes Nondenominational Protestant***p Non-Christian Religion Religious Attendance Divine Interaction*p Existential Certainty***p*pCertainty x Education***nN/ACertainty x TraumaN/A*p
Table 4. This is a summary of the variables used in C.G. Ellison, "Religious Involvement and Subjective Well-Being." 3 See text for a full explanation of the variables. The number of asterisks indicates the level of statistical significance (p < .05, .01, and .001 for one, two, and three asterisks, respectively). A "p" indicates a positive relationship, in which increases in the value of the variable on the left are associated with increases in the dependent measure listed at the top of the column. The letter "n" indicates that the level of satisfaction or happiness goes down as the value of the variable goes up. Thus, more traumatic events in one's life are associated with lower life-satisfaction and happiness.
At this point, the devoutly religious will surely object that this is unfair to the core hypothesis of direct effects of belief, because this unidimensional measure of church attendance doesn't get to the heart of what makes one a religious person. Can we measure the effects of religious belief more directly? That is a difficult task, and methodologically sophisticated, scientifically respectable efforts of this kind are relatively rare. Many such studies are badly done, but let me give the advantage to the religious side of the argument by focusing on a particular research study performed by Christopher Ellison and reported in the Journal of Health and Social Behavior, a well-respected publication of the American Sociological Association.2 This research is among the best of its kind, and the paper is one of those most frequently cited in support of the claimed benefits of strongly held religious beliefs.
Table 4 summarizes the list of variables Ellison included in his study, which was based on interviews of 948 people in 1988. The two columns labeled "Life-Satisfaction" and "Happiness" name the factors that Ellison is attempting to predict.
The variables listed in the left-most column are the predictor variables, i.e., the things that are hypothesized as causal factors in determining life-satisfaction and happiness. The variables above the heavy horizontal rule reflect basic demographic facts. Those below the line indicate Ellison's effort to tap into the efficacy of actual beliefs. Actual content of the belief set is indicated by reference to religious affiliation. Religious attendance is the same variable so often measured alone in the research summarized above. "Divine interaction" refers to activities like prayer or other indications of direct contact with God (experiences of being in God's presence, transcending the material world during meditative states, etc.). Perhaps the most interesting of these variables is existential certainty, which is a measure of the degree to which people report being free from doubts about their beliefs.
Ellison's approach recognizes that when we are trying to understand something like happiness or life-satisfaction, it is probable that there are multiple causes at work in complex ways. We have to figure out how much each of several different factors contributes to the variation in our summary measure. One way to do this is to use a technique called multiple regression, a statistical technique for finding the unique contribution of each variable to the overall variation in the thing we're trying to predict, in this case life-satisfaction and happiness. Table 4 summarizes the results of Ellison's multiple regression analysis. If a variable has no asterisk next to it, it did not contribute significantly to the variation in life-satisfaction or happiness. The number of asterisks simply indicates how unlikely it is that the observed relationship might have occurred by chance. The letter "p" indicates a positive relationship, "n" a negative one.
Many of the findings from the Ellison study are entirely consistent with a secular perspective on life. Among the most powerful predictors of life-satisfaction, for example, are income, marital status, and involvement in voluntary associations. This is entirely consistent with the social-psychological interpretation of the effects of church going offered above. Furthermore, this study reveals no independent or unique contribution to either happiness or life-satisfaction from church attendance, suggesting that this really is just a proxy for other, truly causal factors. The data from this study are quite rich, and the reader is invited to mine the table for other insights. But for the remainder of this discussion, let me focus on a few of the more theoretically provocative findings.
The first of these is the lack of systematic denominational effects. This is a common pattern in this literature, suggesting that it doesn't matter what the detailed content of your beliefs is, so long as they have the right quality or character. For strong adherents to specific belief systems, this result is often quite disturbing, but it has been replicated many times. So what is the quality or character that a belief system must have to be efficacious? Ellison points to the findings involving existential certainty as key, and that is the second thing worth noting here. What Ellison calls existential certainty is the degree to which one accepts one's religious belief system without doubt or hesitation. As the table shows, the higher one's level of existential certainty, the greater the level of life-satisfaction and happiness. Less conviction—greater doubt and skepticism—will be associated with less happiness and satisfaction.
We can gain a deeper understanding of this effect by examining two other findings that involve what are known as interaction effects, shown in the last two lines of the table. Certainty interacts with level of education to produce a negative relationship to life-satisfaction. In other words, existential certainty is more important for life-satisfaction the lower the level of one's education. This is consistent with research that seems to show that having available a strong religious interpretation of one's life experiences helps to compensate for lower levels of intelligence or poorer cognitive skills. Religion provides a framework for giving coherence and meaning to life when one lacks the intellectual skills to sort it all out on one's own.
With respect to personal happiness, existential certainty interacts with the number of traumatic events in one's life. High levels of certainty are more strongly associated with happiness the more traumatic events one has experienced, perhaps because it again provides a meaningful framework for coping with these horrible events (as in, "It's all part of God's mysterious plan").
Ellison's study is not without its own technical problems and limitations, but for the sake of argument I'd like to give the religionists the benefit of the doubt and assume that there is something to be gained from religious involvement and that some of that benefit is associated with the direct effects of being a believer, and not just with social-psychological factors. Does it follow that the system of beliefs has to be religious or mystical to produce the same effects?
That question brings us to a particularly provocative finding often tucked away in the literature on religiosity. The findings we've been discussing so far tend to rely on statistical procedures that assume that any trends in the data will be linear. In other words, the procedures assume that having more religiosity will be associated in a simple, direct way with having more (or less) of something else, like happiness or life-satisfaction. But if the pattern in the data is curvilinear, not linear, the results of a linear analysis may be very misleading.
An interesting example of such non-linearity can be found in a study by Shaver, Lenauer, and Sadd published in the American Journal of Psychiatry.4 Entitled "Religiousness, Conversion, and Subjective Well-Being: The 'Healthy-Minded' Religion of Modern American Women," this study reveals that the anti-religious have the same good health and life-satisfaction pattern as the very religious, while the nonreligious and only moderately religious have the highest rates of low happiness and poor health (see Figures 1a and 1b).6
Figure 1a. The relationship between religiosity and unhappiness scores—the higher the score, the less happiness (more unhappiness). Adapted from Shaver, Lenauer, and Sadd (1980), "Religiousness, Conversion, and Subjective Well-Being: The 'Healthy-Minded' Religion of Modern American Women."5
Figure 1b. The relationship between religiosity and symptoms of ill health—the higher the score, the more unhealthy one is. Adapted from Shaver, Lenauer, and Sadd (1980), "Religiousness, Conversion, and Subjective Well-Being: The 'Healthy-Minded' Religion of Modern American Women."7
Similar patterns have been found among those being treated for depression, with the most severe levels of depression found among those in the middle range of religiosity.
Note that if one removes the anti-religious from these graphs, one sees a pattern very much like that already discussed: more religiosity means more happiness (less unhappiness) and better health. But very few studies include the anti-religious category; Ellison did not include the anti-religious in his work, for example. If this group is excluded, as it almost always is, the true, curvilinear pattern is not revealed.
Here we have to look at a broader literature and make some inferences, but a pattern does seem to emerge. First, agnostics and atheists tend to score higher than believers on tests of intellectual ability such as standard IQ tests. Religious skeptics score more highly than believers on both the verbal and the quantitative sections of the SAT. Add to this pattern Ellison's finding that existential certainty among the religious compensates to some degree for lower levels of education and intelligence in support of higher life-satisfaction scores. Suddenly, we have available a hypothesis about the role of religious belief in the quality of life that is very different from the one that religionists want to promote.
That alternative hypothesis is that the direct effects of religious belief on happiness and life-satisfaction are explainable as the consequence of a strong commitment to a system for organizing one's life experiences into a meaningful, coherent whole. Major religious systems offer answers to Big Questions about the origin of life and the universe, and do so in a way that offers a framework for organizing life's goals and explaining personal tragedies as well as good fortune. Most such systems are hierarchical, with clear lines of authority that reduce or eliminate ambiguity about how one should act in all the spheres of one's life. They offer, in other words, a pre-packaged blueprint for living, anchored in social structures that have evolved over centuries. Such systems operate successfully, however, only to the degree that one makes a strong commitment to them—hence the relationship between degree of religiosity and the good life.
To round out the story, we only need to recognize the greater intellectual challenge associated with taking responsibility for formulating one's own sense of life's purpose and finding a reasonable path to one's chosen goals—hence the relationship between an anti-religious worldview and a high level of cognitive functioning. The anti-religious also have recourse to a historical body of philosophical work, as well as modern science, to aid in their journey. But in a world where being anti-religious requires swimming against the current of mainstream culture, it is almost certain that there must be a high degree of commitment to the nonreligious worldview, borne of personal intellectual struggle and the need to defend one's beliefs in the face of both implicit and explicit criticism. This strength of conviction easily comes to equal that of the most devout among the religious and provides the same psychological benefits. It carries the added benefit that it doesn't require becoming a permanently dissociative personality within whom life in the real world must compete always for attention with the pull of the mystical world, where reason and evidence are banned.
So we come back to our original contrast between Enlightenment and religionist worldviews and discover that popular and academic claims for the benefits of religious adherence need to be heavily qualified. Many such effects are social-psychological in origin, not spiritual, and the more direct psychological effects appear to be strongest either for people with relatively limited cognitive coping skills or for those who have experienced greater than normal levels of trauma in life. Furthermore, being anti-religious is associated with the same physical and psychological benefits as being very religious, and is healthier than being only moderately so.
The paucity of research on the anti-religious means that many of my conclusions are hypothetical at this point. A great deal of additional work needs to be done to establish the validity of this alternative interpretation. Simply repeating many of the better studies already done on religiosity with anti-religious people included in the samples would be a major step forward. With large enough samples, one might also begin to explore some fascinating implications of these hypotheses for yet other perspectives.
For example, Table 1 recognizes a third worldview, that of the postmodernists, about which we have essentially no psychological data. Postmodernists include both the religiously inclined and the nonreligious, but their philosophy denies that a concept like truth has the kind of stability of meaning necessary to develop real conviction about anything, regardless of whether the claim to truth is based in reason or faith. It would seem to follow that adopting a postmodern worldview should produce a psychological state most like that of the only slightly or moderately religious; from this it would follow that one ought to see lower levels of life-satisfaction and happiness than in either the deeply religious pre-modernist or the anti-religious modernist.
Whatever the outcome of future research in this area, it is clear that the current literature does not establish that religion is necessary for health and happiness in this world. This fact will have little effect on the attitudes and beliefs of the truly faithful, precisely because faith, not evidence or reason, is seen as the ultimate arbiter of truth. I hope that I have shown, however, that the nonreligious enjoy an abundance of perfectly plausible alternative explanations for the much ballyhooed findings that religion is good for you—all of which are entirely consistent with living the good life without any religion at all.
1. "Church attendance" is the shorthand in almost all of this literature for attending services at any place of religious worship, whether mosque, synagogue, cathedral, temple, or what have you.
2. Christopher G. Ellison, "Religious Involvement and Subjective Well-Being," Journal of Health and Social Behavior, no. 32, March 1991, pp. 80-99.
4. Shaver, Lenauer, and Sadd, "Religiousness, Conversion, and Subjective Well-Being: The 'Healthy-Minded' Religion of Modern American Women," American Journal of Psychiatry, no.137, 1980, pp. 1563-1658.
6. This research was limited to women, so further work is needed to confirm the results with men. There is, however, nothing in the psychological literature that would suggest that it should not generalize to men.
7. Shaver, Lenauer, and Sadd, "Religiousness, Conversion, and Subjective Well-Being."