His introduction eviscerates the conventional wisdom about religions and religious organizations in sociology. Is religion the way the oppressed cope? Nope--the elites tend to be more religious than the masses. Were religions designed as means of social control? Hardly. Do primitive tribes have primitive concepts of God(s)? Nope--There's a High God tradition almost everywhere. Many popular attempts to define religion and its role in society and personal life use name-calling rather than analysis, and lots of non-falsifiable definitions. (In a later chapter he apologizes for and then demolishes one of his own early works.)
He says he will do it differently--and he does. He lays out proposition and definition after proposition and definition, with explanations for why along the way.
Jesus is one of his examples of religious founders, of course, and Stark is so dubious about his details that I wonder if he felt he had to act more cynical than the facts actually warrant. Mohammad's life is attested only quite late, but he cites it without much caveat.
At any rate, he builds his theories on an entirely humanist and economic model. He is aware of The Idea of the Holy but rejects the principle for his purposes. That makes his framework shaky, because insofar as a human actually is in contact with the transcendent, the usual economic concerns no longer work the same way--and feedback can shape the nature of the beliefs and organizations that result.
On the other hand, the framework fits the way religions work around the world, how they evolve and fission--and sometimes merge. One set of his propositions (illustrated by examples) hold that it is the ecclesiastics, and not the average church-goer, who tries to move a church from a "high tension" relationship to the rest of society to a "lower tension" relationship. The clerics are more likely to turn atheist than the congregants. And one of his final conclusions is that some form of religion is a rational position.
One of the things that people have noticed for thousands of years is that women are more religious than men. He looks at that a little closely, and connects it with another way men differ, on the average, from women--much greater risk-taking, especially in youth. "Research on 1,148 newly ordained clergy in the Church of England found that their average score was well below the national average for English men on a scale of risk taking. Finally, the gender difference in religiousness is very large among Orthodox Jews but is nonexistent among Jews who do not believe in life after death and who therefore percieve no risk in being irreligious."
I'd like to see those distributions myself before I sign onto that hypothesis. For example, what is the distribution of religiousness among those Jews who don't believe in life after death? Is it skewed so low that male/female differences are no longer significant? Does male/female religiousness equalize with age, or is set early on and stable thereafter?
One of his observations is that the rich and powerful are more likely to get involved in a new religion, or more deeply in an old one--so presumably they feel a spiritual lack--presumably because they have the leisure to feel it and do try to do something about it. The poor have lots of this-world existential worries.
He insists on a definition of religion in which there is a supernatural, which excludes quasi-religions like worship of the state or political party or the glorious leader. While this is fitting and proper, the erzatz religions need to be accounted for somehow, since they fill the same sort of role. I suspect they follow the same sorts of principles too, and will, over time, evolve in the same sorts of ways. Devout environmentalists don't seem to come from the poorer classes (match)--are there splits in groups as "high tension" devotees become dissatisified with the "medium tension" big tent organization? (If so, another match)
Read it.