The recently exposed Cambridge Analytica scandal, where intrepid Observer journalists revealed that more than 50 million Facebook profiles were harvested without consent for political ends, has shaken the worlds of media, politics, even international relations. Facebook itself has also taken quite a hit, with its share value dropping considerably, and many people, even the co-founder of WhatsApp, joining the #DeleteFacebook movement.

But, then, we all know at least one person who publicly vows to quit Facebook/Twitter/another social network, only to never actually do so, or if they do, quietly return a short time later. Why are social networks so hard to part with? There are many factors at work, but one is that, whether by design or happy accident, they tend to stimulate our brains in fundamental ways.

Social interactions played a key role in the evolution of our unreasonably large and powerful brains, according to many. The ecological dominance, social competition model, for instance, argues that when you’re part of a human group, your group protects you, feeds you, cares for you, helps you, even provides you with mating opportunities. Over time, this meant humans didn’t evolve traits for survival as individuals in the wild, but success in a group. The ability to empathise, to communicate and coordinate, to anticipate and calculate the reactions and intentions of others, and, yes, to deceive and manipulate them, all requires serious neurological processing power. Ergo, bigger, more capable brains. It would explain why ours nearly tripled in size in only 2 million years. But if surviving in a social environment drove the evolution of the human brain, then you’d expect to see this reflected in its workings and structure. And you do. Repeatedly.

Evidence points towards a neural network that governs social interactions, and it’s heavily linked to the mesolimbic reward pathway, that part of the brain that causes us to experience pleasure. It’s far more complex and nuanced than this, but a reasonable conclusion would be that we’ve evolved to really like social interactions. This is supported by studies that show if you interfere with the action of endorphins (natural opiates) in social animals, it affects their socialising behaviour. Social interactions and pleasure are linked at the most basic levels.

So, what if you could experience multiple social interactions, 24/7, at the touch of a button? It would be reasonable to assume our brains would gain a lot of enjoyment from such a thing.

The inverse is also true; a negative social interaction, a rejection, seems to cause a lot of activity in the pain-processing areas, which suggests that names do hurt you, despite the cliche that argues otherwise. It makes evolutionary sense; if your survival depends on being accepted by the group, being rejected by others is literally a matter of life or death, so social interactions always come with an element of risk.

Unless, of course, you have the ability to shut out or block those who don’t agree with you, or shield yourself from contrasting views. Such a system would be very appealing to the typical human, providing a sense of safety and comfort.

We may want to be liked and accepted by others, but we also want to be admired by them. Social status is important for many species, and humans are no different. Our brains are very sensitive to it. Social phobias are by far the most common kinds, so any time you engage with someone new it’s potentially a big deal, requiring a lot of cognitive effort. And maybe because of this, our brains seem predisposed to making sure other people like us as much as possible. For example, the phenomenon of impression management suggests that our brains default to a positive self-image (even if it’s inaccurate), which guides how we present ourselves to others. You may scoff at someone for taking 50 selfies in a row before finding one they’re happy with, but the brain does that all the time.

But, imagine you had total control over how others perceive you. Imagine you could choose exactly how to look and what to say whenever you wish to say it, rather than being at the mercy of the world around you. This would remove much of the risk, reduce uncertainty and provide a very comforting sense of control

So, the brain has all these properties, and many more, seemingly evolved to increase the odds of successful social interactions and gaining the approval of others. But in the real world, all this requires effort, and places demands on the brain’s limited resources. Dunbar’s number (named for the anthropologist Robin Dunbar) argues that the maximum number of stable relationships a human can maintain is 150, and social psychologists have often noted that prolonged social interaction requires periods of withdrawal, privacy, as the brain needs to “recharge”.

What Facebook and other social networks do is provide the social interactions, connections and approval we seemingly crave, but with less risk and less effort (something else our brains really respond to).

Facebook and the like tick a lot of our brains’ boxes, often without us realising, in simple, controllable, measurable ways (a nice chat is generally positive, but 78 likes on a status? That’s quantifiable, increasing a sense of status and control). It’s perhaps no wonder that so many seem unable to tear themselves away from it, despite all its considerable flaws.

Many suggest that we’re “addicted” to Facebook, but that’s a bit extreme. It better resembles the way our brains have evolved to respond positively to things that taste sweet, like fruits, because these provide valuable sugars and energy. But then we figured out how to refine sugar, and now need self-discipline and control to avoid it.

Facebook is like refined sugar for the brain’s social reward systems. And if eating bag after bag of sugar didn’t make you noticeably sick and had no detectable health consequences, it would probably be hard to stop.

• Dean Burnett is the author of The Happy Brain, to be published in May in the UK, US and Canada