In the last 30 years the world has conducted a natural experiment involving the widespread adoption of internet and mobile technology, with all of the function that comes along with it (such as social media). At the same time the world seems to be a more complex, stress-inducing, and dysfunctional place. This has lead to the common assumption that the two are somehow causally related. The conventional wisdom is that we have made our lives more stressful by constantly being online. Social media specifically gets a lot of blame for creating a generation of anxious neurotic people.
But of course, conventional wisdom is often wrong. We need data to see if this assumption is actually true. This is an extremely complex social question, and no single study is going to give a definitive answer. As we often discuss here, there are many ways to conduct such research and to slice and dice the data. There are also many questions we might ask – are we talking about typical internet use, internet “addiction”, in which demographic group, and with which specific outcome?
Existing data is complex and mixed, without any clear conclusion to date. There is some suggestion of an association between “excessive” internet use and depression and loneliness, but this also depends on age and gender. It’s also very possible that being lonely and depressed might lead to internet addiction, rather than the other way around.
A recent study is a significant addition to the literature, partly because it is the largest such study to date – Global Well-Being and Mental Health in the Internet Age. This study is not a collection of new data but an analysis of existing data, from the Internet users and mobile-broadband subscriptions databases in 168 countries, and the Global Burden of Disease (GBD) study 2019 data, and a synthetic version of the Gallup World Poll (GWP) data.
This is a population-level study, so it does not include information about individual people. They are asking the question – over the last 20 years, as internet and mobile technology use skyrocketed around the world, was there also a population-level increase in mental illness or change in markers of well-being? They mainly correlate this data between countries, but also do some within-country analysis.
The paper is divided into two studies, the first dealing with well-being (overall life satisfaction, positive experiences, and negative experiences), and the second dealing with mental health (anxiety, depression, self-harm). They looked at overall trends for each country, and also divided the data by gender and age. They looked at changes over time, and differences between countries.
I will break it down a bit, but here is the overall bottom line – there were no significant correlations between internet use, mobile technology use, and any measure of well-being or mental illness. There simply is no signal, and not what we would expect if internet use in any significant way decreased well-being or increased mental illness. There were trends, which I will get to, but the authors used a region of practical equivalence (ROPE) analysis. This is a determined region of effect size that is statistically and practically indistinguishable from zero. Within this range, the results might as well be zero. All effects were within this range.
Any discussion of trends within the data must explicitly recognize that these effects were essentially nothing, consistent with the null-hypothesis. Over time there was a small increase in both positive and negative experiences (with a larger increase in negative experiences), while overall well-being was essentially flat. So with internet access people experience more stuff, with a bias toward negative experiences, but all non-significant trends, with flat-line well-being over time.
If we look at between country data, internet access correlates with increased life-satisfaction, and both positive and negative experiences. Meanwhile, with mobile access, there was a trend toward increased satisfaction, decreased negative, and slightly decreased positive experiences.
Again, these trends were all indistinguishable from zero. If there is an underlying reality to them, there is also no way to draw an inference to cause and effect from this data. It is perhaps more likely that things like socioeconomic status are driving both these well-being trends and internet and mobile access.
The data on mental illness (anxiety, depression, self-harm) are easier to summarize – all the trend lines are pretty flat, either over time or associated with internet or mobile access. Tiny trends are not even worth pointing out, and they are all consistent with the null hypothesis.
To emphasize, no one way of looking at data will give a complete or definitive picture of such a complex question as the effect of a massive social change on society and various demographics within that society. But the window provided by this data, which included data on “2,434,203 individuals age 15 to 89 across 168 countries from 2005 to 2022”, is fairly reassuring. Zooming in on individual-level data has its own strengths, but also comes with a host of potential confounding factors.
This study zooms all the way out, and asks a fairly basic question. As the world has adopted internet and mobile technology over the last 20 years, has there been a corresponding decrease in well-being or increase in mental illness overall? The data says no.
Of course, in this study we are dealing with averages. It is possible that internet use, for example, increases well-being in some people while decreasing it in others, leading to no change in the population-level average. Breaking down the data by gender and age in this study did not reveal any such pattern, but a more complex pattern may still be hiding in the data. This study also did not look at individual level of use (internet addiction), or the way individuals use the internet (for work, play, or social interaction). So this study does not mean there is no effect on one’s life from using the internet.
The authors point out that while population level data like the data they analysis is freely available, individual-level data is not available. This data exists – the giant tech companies gather a lot of data from their users for marketing purposes – but they do not make this data available to independent researchers. Doing so would allow for more detailed research into this question.
But for now we can be a little reassured that the data does not support the notion that the world is becoming more anxious or depressed and less satisfied because of the internet.