Science on Religion

Exploring the nexus of culture, mind & religion

Internet usage correlates with no religious affiliation

Internet religionThe Internet has undoubtedly revolutionized society. From the ability to communicate across the world, to being able to buy practically anything, and to perhaps shadier things, the Internet plays a vital role in modern daily life. Computer scientist Allen B. Downey (Franklin W. Olin College of Engineering) wonders whether the Internet could have an effect on religion. More specifically, does frequent Internet usage have any effect on one's religious affiliation? In his research, Downey concludes that, indeed, the more one uses the Internet the less likely one is to be religious.

Downey noticed that from 1990 to 2010 the percentage of Americans with no religious affiliation doubled, from 8% to 18%. During the same timeframe, Internet usage increased from practically 0% to almost 80%. While this could be a mere coincidence, Downey suspected a connection between Internet usage and religious affiliation.

To check whether any such connection exists, Downey relied on data from the General Social Survey (GSS) for the years 2000, 2002, 2004, 2006, and 2010. Pertinent to his study, Downey mainly focused on the survey's participants' religious preference, the religion they were raised in, their education, their income, and their Internet usage per week.

In order to interpret the results of his data crunching, Downey based his study's probabilities on the hypothetical person with no religion, born in 1960, having a high school without a college education, and little Internet usage (less than two hours a week). This person would have had a 44% chance of having a religious affiliation as an adult. Extrapolating from the GSS data, Downey finds that if this person (a) had been raised religiously, their chance of having a religious affiliation would increase to 89%, (b) then if they income in the top 75%, their chance of having a religious affiliation would have increased to 91%, (c) was instead born in 1970, their chance of having a religious affiliation would then decrease to 89%, (d) went to college, would decrease the probability of religious affiliation to 87%, and finally (e) use the Internet at least two more hours per week, the chance of them being religious falls to 84%.

In this hypothetical case, the Internet overall decreased the probability of religious affiliation by 7%. Conversely, if the Internet had never existed, there would be a 20% increase in people who affiliate with a religion. This effect is of comparable strength to that of religious upbringing.

Downey freely admits that correlation does not mean causation, but, at the same time, he argues that correlation does provide evidence for causation, especially when alternative explanations are lacking. He cites two ways in which Internet usage could lead one to disassociate from religion. First, the Internet enables communication to find out about people across the religious spectrum and interact with them personally. This could make one's own religious commitments implausible. Second, those with religious doubts could find Internet communities of people who share the same exact doubts, which could exacerbate one's disbelief. In other words, whereas in a world without the Internet people with doubts would be surrounded by community of the faithful and therefore would have their doubts in some way assuaged, with the advent of the Internet, people can find those who believe as they do and so find community in not believing.

Furthermore, Downey fails to find any plausible alternative explanations for this correlation between increased Internet usage and decreased religious affiliation. With the lack of alternatives, he reasons that one can at least tentatively and not naïvely conclude a causal relationship.

Even if at best Downey has only proven correlation rather than causation, his results still indicate a statistically significant relationship. It would appear, for whatever reason, that individuals who frequently use the Internet are more likely to be individuals without any religious affiliation. It might be too simplistic too draw hard-and-fast conclusions from this research, but if you're a religious person, close your browser – and if you're not, keep on browsing!

For more, see “Religious affiliation, education and Internet use” published by arXiv.


Add comment

Security code

You are here: Home Research News Research Updates Internet usage correlates with no religious affiliation