The Pew report is pretty interesting.
https://www.pewresearch.org/religion/20 ... n-america/ The report examines several scenarios, all of which vary only in what they assume about rates of switching. The first scenario shows what would happen if no one switched in the future, to show the impact of demographics. Even then, Christianity's percentage continues to fall because the average Christian is older than the average unaffiliated person. The effect of age distribution swamps any effect of religious belief on birth rates. This scenario, which no one expects to occur, is the only one that shows Christians still representing a majority (54%) in the U.S. in 2070.
The second scenario assumes that the rate of in and out switching stays steady over time at the same rate we see today.
Because the rate of out switching from Christianity is higher than that for the unaffiliated, the percentage of Christians continues to fall and of unaffiliated continues to rise. In 2070, Christians are projected to be 46% of the U.S. population.
The third scenario takes into account the trends in the rate of switching over time. However, it caps the rate of out switching at the highest rates observed in european countries (50-55%). That results in a projection of 39% for Christians in 2070.
And if the caps are removed, the projection is that Christians will represent only 35% of the population in 2070.
Under the first two scenarios (no switching and constant switching), Christians will still outnumber the unaffiliated in 2070. But under the second two (trends in switching with limits and without limits), the projection is reversed: the unaffiliated will outnumber Christians in 2070.
Under all four scenarios, "other religions" will approximately double their share by 2070.
The Pew study did not examine causes.