The two main theories of religious change are the secularization paradigm and the economic model of religion. The secularization paradigm's main premise is that modernization weakens the power and authority of the church as an institution and reduces the importance of religion in the daily lives of the population. This paradigm applies well to Europe, but the United States acts as a powerful counter-example. Since the 1940s, religious attendance in the United States has remained generally stable, with approximately 40% of the population claiming to attend religious services, mostly in Christian churches, at least once a week. American sociologists explain this relative vitality with reference to an open and competitive religious marketplace, claiming that the innate desire for spirituality is met by the sheer diversity of religious groups in the United States. This economic model of religion applies poorly to the European situation. This thesis examines these apparent contradictions by considering the similarities and differences between the dynamics of religious change in five western countries since the 1970s or 1980s; the countries are Britain, Australia, New Zealand, Canada and the United States. The key question is which combination of dynamics is responsible for religious change in western countries. The effects of migration and fertility aside, all population change must be related to some combination of age, period or cohort effects; age effects are those that occur as people age, period effects are those that affect the whole population regardless of age, and cohort effects are often attributed to circumstances or events during youth. These different dynamics of religious change would each indicate different sources of religious change at the individual level, which may lend support to one theory of religious change over another.I show that there is overwhelming evidence that most religious change in Britain, Australia, New Zealand and Canada is due to differences between generations in the time periods observed with some slight downward period effects. The main drivers of change in the United States are either downward cohort with upward period effects, upward age effects, or a combination of these three effects. I conclude that the changing conditions of socialization in youth, both formal and informal, related to modernization and cultural shifts can explain the dynamics in Britain, Canada, Australia and New Zealand. Similar effects in the United States may be counteracted by the high social desirability of religion in that country by contrast with the other countries and the ability of particularly conservative Protestants in the United States to isolate themselves from views that conflict with their own; these groups are aided in this by numerical strength and by the ability to socialize, work and view media all of which enhance their religious worldview.