#http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2168000
My finding is simple: people who trust God, don't trust people. The more you trust God, the less you trust human
species. It is a specific dimension of religiosity - believing in God, not religiosity as a whole - that predicts lower trust.
Belief in God depresses trust in men in the US and in the World as shown using General Social Survey (GSS) and
World Values Survey (WVS). I suggest directions for the future research.

I am lucky (as a social scientist) to know many people who fiercely believe in God and many who don't believe
at all. And it strikes me that the believers are quite suspicious of other people - they don't seem to be trusting, they
think that other people are up to something. In fact, whenever I meet any religious people, the same pattern strikes
me. People who are highly devoted to God, don't trust fellow human beings. Specifically, I observed that it is belief
in (closeness/devotion to) God that makes people less trusting in human kind. The strength of believing matters, too.
People who unconditionally trust God will not trust humans. That is, the very strong belief in God seems to prevent
people from believing in others. Moderate belief in God, on the other hand, will not have a big effect.
Following Okulicz-Kozaryn (2010), I define two dimensions of religiosity: social and individual. Social religiosity is
about social interaction and takes forms such as church attendance and religious meetings. Individual religiosity, on the
other hand, is about personal interaction with God and takes forms such as prayer or feeling of closeness to God. The
effect of the two types of religiosity on trust is similar to their effect on happiness: Individual religiosity depresses both
happiness and trust while social religiosity boosts both happiness and trust. So it is not religiosity in general, but a
specific dimension of it, belief in God, that depresses trust in man. Further, I think that Abraham Maslow was right in
saying that peak experience (spirituality)1 is actually great for people, while organized religion is not so great. Organized
religion, say Christianity or Islam, deviated too far from its (spiritual) origins, and became institutionalized, hierarchical,
and even bureaucratic.
By trust I mean generalized or social trust, i.e. trust in other people. Fundamentally, such trust helps with com-
munication, lowers transaction cost and risk. Trust has great consequences: economic growth, improved
commerce and trade, happiness, political and civic involvement, crime prevention and better health. It is also becoming increasingly difficult to live if you do not trust people, because economy and
everyday life is more and more about collaborating, sharing and trusting other Organized religion provides social capital in a form of church attendance, religious meetings, and so forth. Furthermore,
religiosity boosts some other forms of social capital: For instance, religiosity increases charity and volunteering (Putnam
2001). Hence, some forms of religiosity, i.e. social religiosity, can actually increase trust. Yet, both types of religiosity,
social and individual, draw a circle of trust. There is a fundamental problem with circles: There will be people inside
but also outside of it, otherwise, what's the point of a circle? Circles of trust are based on in-group trust (binding, not
bridging social capital). Individual religiosity is bad for trust because there is only an adherent and her God in the circle.
Social religiosity adds other adherents in the circle. In other words, individual religiosity draws a circle that separates a
person from the mankind. And social religiosity draws a circle, that separates believers in some God from nonbelievers
in that God. Adherence to a religion separates a person from other religions.
There is little doubt that religion promotes intra-group (bonding as opposed to bridging) trust,4 but its link to
generalized/social trust is complex. For instance, Sosis shows fascinating examples of strong bonding/intra-
group trust for Jewish diamond merchants, Maghribi and Muslim traders that got them a comparative advantage over
competition in doing business. There is a comparative disadvantage, too. It is likely that everybody else would trust less
these groups, and these groups would also trust less everybody else. The strong trusting relationships within the group
are formed at the expense of out-of-the-group. To paraphrase Banfield, the individual religiosity could
be called "amoral religiosity": it kills generalized trust.
Let me make a general note about the very strong levels of trust. Very strong or unconditional trust happens within
a circle{nobody trusts unconditionally everybody else, but some trust unconditionally God. Such unconditional trust
happens at the expense of the generalized trust{you trust unconditionally within your circle either because you cannot
trust anybody else, or you don't have to trust anybody else, because you trust unconditionally within your circle. Either
way: there is a tradeoff, as I document in this study.
There are several studies looking at the religiosity-trust link, but they all do it superficially. Alesina and Ferrara
use GSS for 1974-94 and operationalize religion as denomination and find no effect on trust. Berggren and Bjornskov
quote findings from two studies: being a protestant in Germany, or being a Catholic in Latin America increases
trust. I will touch on that in the appendix, but the idea is that you'll trust most people if you are a member of a
denomination that most people in your country belong to. For instance, there are not many atheists in the US, and hence
they are mistrusted. A very similar study to this one is Welch et al., however, they do not
even notice6 that there is a difference in their regression between social and individual religiosity and this is really the key
finding of my study. Helliwell et al. do notice the thesis of this study, but do not notice its significance
and do not elaborate beyond this:

Again, studies about religiosity-trust relationship focus on the religiosity in general, while religiosity is clearly a complex
concept. In the next two sections I will elaborate religiosity using multiple measures from General Social Survey (GSS)
and World Values Survey (WVS).