It disturbs me that my nephew is being raised utterly godless. Respect for tradition ought to at least make the parents raise their kid in the religion they were raised in, or something similar enough to it. If the child wants to lapse or drop out as an adult, that's ultimately their choice. But parents give their children no choice at all by raising them in a religious vacuum.
Moreover, by severing the ties to the past, even if the child does take up religion as an adult, they are less likely to wind up in the same or similar religion or denomination as their ancestors, than if they had had that status passed onto them directly from their parents. Connections to the past can become fractured incredibly quickly.
But will godlessness actually continue to grow and grow as a result of children being raised that way to begin with? Or is there some kind of affinity for religion wired into human nature? That could take the form of the child seeking out religion even if they had been deprived of it during development, or a parent who was raised godless deciding to give their child a religious upbringing.
The General Social Survey asks questions about the religion that the respondent belongs to, that their children belong to, and that their parents belonged to. We can thus investigate how stable the transmission of "no religion" is through your descendants. Only whites will be looked at, to sift out race as a confounding variable.
Let's start at the nuclear level. If your current religious affiliation is "none," you are only about 70% likely to raise your child with no religion. That may still be more likely than not, but it's quite a bit lower than the success of passing on your Protestant or Catholic status (around 80% likely).
Now zoom out to follow the transmission from grandparents to grandchildren. If your father or your mother was raised with no religion, there's only about a 50% chance that your child — their grandchild — will be raised with no religion. That is just what we'd expect from two independent transmissions that each have a success rate of 70% (square it, and you get 50%). As before, the similarity is higher for the presence of religion: Protestant status shows about a 70% chance of the grandparents and grandchildren being the same, while Catholic status shows about a 60% chance of being the same.*
Aside from similarity between parents and offspring, what about the change within the life of a single individual who was raised with no religion? There's only about a 50% chance that they will remain without religion into adulthood. By contrast, someone raised Catholic is over 75% likely to stay Catholic as an adult, and Protestants even more so, at over 85%.
What fills in for the absence of religion, then? In all these cases, a majority turn to Protestantism and a minority toward Catholicism. That is likely due to Protestantism being the dominant form of religion in America during the late 20th and early 21st centuries, rather than any special appeal of Protestantism.
To resolve that question, we could look into the same phenomenon in a majority Catholic country like Italy, or even better Spain, where godlessness is more common and would provide a larger sample to study. I expect Spaniards who are raised with no religion to turn more to Catholicism than to Protestantism, if they seek out religion in adulthood.
But what pulls the seekers toward one form of religion or another is a separate topic. For now we just need to emphasize how less stable it is for parents to transmit the absence of religion through their descendants.
Of course we should also mention that there is a heavy secular trend toward godlessness, which the GSS data confirm. If you look at whites who were raised with no religion, and see whether they remain so in adulthood, it becomes more and more likely with successive generations, especially with the Boomers and later generations. It doesn't matter whether you look at the generations during the same stages of life or not.
To give an idea of the trend, though, if you were raised with no religion, how likely are you to have no religion during your 30s? Only about 20% if you were born from 1935 to 1944, but 35% if you were born from '45 to '54, and nearly 60% if you were born from '55 to '64. At least it doesn't get worse after that — it stays around 60% for the '65-'74 and the '75-'84 cohorts. The snowball appears to have stopped rolling in that direction.
The secular trend shows up among those raised Protestant and Catholic as well. Late X-ers and Millennials who were raised Protestant are between 20-30% likely to abandon religion in adulthood. Among Catholics, just under 20% are likely to abandon the religion they were brought up in. Both of those figures are nearly an order of magnitude higher than they were under the Greatest Generation.
The point of contrasting these two forces is to underline how impotent parents are to affect the course of religious evolution through parenting per se. Religious-minded folks might view that as a downer, but it's worse news for the godless. Religion of one kind or another is easier to pass on to your children, while godlessness will be more stubbornly resisted.
Still, religious folks should not delude themselves about how effective their efforts will be to bring up their children in a religious manner, when the whole rest of social and cultural forces are clearly pushing in the opposite direction. It is those forces that are driving up the retention rates of godlessness from childhood into adulthood among the Boomer-and-after generations — not a greater failure on the part of religious parents, or a greater success on the part of godless parents.
If parents want future generations to grow up in a world that's more welcoming of religion, they need to change those broader societal trends, not to merely adopt a conservative brand of helicopter parenting. Otherwise their best efforts to personally transmit religion to their children, and their children's children, could get swamped by hostile outside forces.
* Those figures are slightly higher and slightly lower, respectively, of what you'd get from independent transmissions each with an 80% chance of success, but I don't think there's too much to read into that.
GSS variables: kd1relig, relig, parelkid, marelkid, relig16, race, cohort, age