The US has long been considered an outlier when it comes to religious beliefs, with much higher numbers of believers and churchgoers than in other parts of the developed world. But a new study says that that gap seems to be closing, and once again it is the young who are leading the way.

Religion in the United States is declining and mirroring patterns found across the western world, according to new research from UCL and Duke University in the United States.
…

According to the study, 68% Americans aged 65 and over said they had no doubt God existed compared to 45% of young adults, aged 18-30. 41% of people 70 and older said they attend church services at least once a month, compared to 18% of people 60 and below.

The authors compare populations in Australia, New Zealand, Canada, and the United States and the abstract summarizes their conclusions.

Virtually every discussion of secularization asserts that high levels of religiosity in the United States make it a decisive counterexample to the claim that modern societies are prone to secularization. Focusing on trends rather than levels, the authors maintain that, for two straightforward empirical reasons, the United States should no longer be considered a counterexample. First, it has recently become clear that American religiosity has been declining for decades. Second, this decline has been produced by the generational patterns underlying religious decline elsewhere in the West: each successive cohort is less religious than the preceding one. America is not an exception. These findings change the theoretical import of the United States for debates about secularization.

They state their three main conclusions.

We have established three central empirical claims. First, religiosity has been declining in the United States for decades, albeit slowly and from high levels. Second, religious commitment is weakening from one generation to the next in the countries with which the United States has most in common, and generational differences are the main driver of the aggregate decline. Third, the same pattern of cohort replacement is behind American religious decline. This decline seems to have begun with cohorts born early in the 20th century. At least since then, strong religious affiliation, church attendance, and firm belief in God have all fallen from one birth cohort to the next. None of these declines is happening fast, and levels of religious involvement in the United States remain high by world standards. But the signs of both aggregate decline and generational differences are now unmistakable.

The authors say that their results undermine the argument that modernity does not necessarily lead to increased secularization.

Whether the United States is a counterexample thus opens into the larger question whether modernity, sooner or later, will bring secularization. This is of course a classic question in sociology. For a long time the majority of social scientists answered affirmatively, but the critics of secularization turned the tide so effectively that today the weight of scholarly opinion is on the other side. America’s accepted status as a counterexample did much to shift opinion in that direction. We are not prepared to say that removing that status should by itself reinstate the old idea that modernity everywhere will bring secularization eventually. We are prepared to say, however, that since it is no longer clear that the United States is on a qualitatively different religious trajectory than Europe, it is too soon to assert that the secularization thesis does not apply outside of Europe, Canada, and Australia. It now seems that the classic question—does modernization undermine religion?—has been prematurely answered, “not in general.” That answer should be reconsidered in light of the evidence we have presented here.

The paper titled Is the United States a Counterexample to the Secularization Thesis? by David Voas (University College, London) and Mark Chaves (Duke University) published in the American Journal of Sociology, Volume 121, Number 5 (March 2016): 1517–56 can be read here (subscription required).