Headlines about the health effects of loneliness are getting more frequent. But these articles always strike me as odd. Loneliness is associated with poor health; but poor health can be the result of loneliness, rather than the cause. It’s like talking about the effects of a fire without ever mentioning the person with the match.

We have trouble talking about the causes of loneliness. The condition is usually presented as a mix of bad luck, genetics, and other factors too ephemeral to pin down. But the notion of loneliness as inexplicable isn’t true.

One factor that’s been consistently linked to loneliness is income. Here’s how it works: as your income falls, you get lonelier. An Australian study showed that people earning less than $600 a week were significantly lonelier than those earning over $1,000 per week. A large Dutch study came to the same conclusion: people with low incomes were twice as likely to be lonely and six times more likely to be socially isolated.

Researchers have become so interested in the relationship between loneliness and income that they’ve started studying it in the abstract. What happens if you give someone 80 $100 bills to count, and then exclude them from a team game? They don’t really care. What if you give them 80 blank pieces of paper, and then exclude them? They feel rotten.

Academics explain this in terms of “financial resources promoting social participation”. At a very basic level, this means that if you have a lot of disposable income, you can spend it on group activities, like yoga classes and wine-tasting clubs. But it also means more than that.

We live in a culture where it’s increasingly difficult to find ways to connect for free. We devalue and underfund libraries, parks and community centres. We commodify neighbourhoods, and then ask those who can no longer afford them to leave. We pull away from traditional sources of inclusion such as unions, places of worship, and groups such as the parent-teacher association (PTA). You might celebrate the death of the PTA – but that still leaves you with the prospect of people riding buses all day because they have no other way to connect.

It’s not just low incomes that cut people off from increasingly market-oriented sociability. My local paper, the Toronto Star, just introduced readers to the word “precarity”, or unsteady employment in all its forms: freelance work, flipped contracts and jobs with constantly changing shifts.

I’ve come to think that, when it comes to loneliness, precarious employment may be worse than having no job at all. If you’re not working, you might be able to rally yourself to volunteer at a community garden or animal shelter. If your schedule is constantly changing, or if you’re in a panic about landing your next freelance gig, you might be too anxious to do anything.

The London School of Economics sociologist Richard Sennett calls this “economically induced withdrawal” – you’ve got so much uncertainty in your life that your appetite for going out and meeting new people is basically nil.

Experts have been linking economic inequality and social inequality for a long time, arguing that healthy social networks and a lot of face-to-face contact are new forms of privilege – with the rich having more of them and the rest of us having less. At a policy level, though, we’re still not connecting the dots. An old church near me was just converted into condos without anyone asking if there was a way of relocating the communal space. Welfare is discussed without any reference to people’s ability to cook for others or meet a friend for coffee.

Socialising is increasingly seen as a “perk” – if you have the resources for that pilates class, then by all means enjoy it; if you don’t, then deal with feelings of isolation on your own. But loneliness and sociability are not private issues. Even if you have a steady job, you’re not going to see much of your friends if they’re maxed out searching for work.

Once your neighbours leave for cheaper locations, they’re gone for good. And we all suffer when people retreat from public life: it’s not much of a party if you’re the only one showing up.

It’s time to start linking the health effects of loneliness to the causes of loneliness. We know that loneliness piggybacks on low incomes, precarious work and unemployment. These issues are more clear cut than the health effects of loneliness: requiring predictable work hours is a lot simpler than halting the effects of dementia.

It’s also time to recognise that rising inequality has a social edge. Increasingly, people with low or even middle-incomes have fewer social opportunities. But belonging and inclusion support our emotional, physical and mental health. We shouldn’t have to buy them, we shouldn’t have to scrimp, and we shouldn’t have to watch some people participate fully while so many are being left out.