My wife Elma came home from a dinner unsettled. One of the other guests had spent the evening being very negative about Isleworth — the west London suburb where we live. It’s all got worse. The streets, the people, the place. Elma wasn’t annoyed that he felt it. She was bothered that the feeling and the facts seemed to have drifted apart.

It turns out that’s one of the more interesting questions you can ask about a place. Has it actually got worse, or does it just feel that way? People answer it reflexively, usually without the data. So I went and got the data.

The short answer

Most things are better. A few are worse. And which is true depends on who you are.

What surprised me was how many people who’ve lived here for decades were ready to tell their story the moment anyone asked. That’s where I started.

Compared with 1976, the streets are dramatically safer. In 1976 roughly 6,500 people died on UK roads. Last year it was around 1,600 — with twice as many cars on the road. That’s nearly 5,000 people a year who would be dead if the roads had stayed as deadly as they were. Fifty years of seatbelts, drink-drive limits, 20mph zones and better cars — a safety dividend barely mentioned at dinner.

Infant mortality has collapsed. In 1980, roughly 12 out of every 1,000 babies born in England didn’t see their first birthday. Today it’s closer to 3. If the single most important fact about becoming a parent in Isleworth has changed since 1976, it’s that your baby is about four times more likely to live.

Violent crime is down hard. The Crime Survey puts it at roughly a 75% fall since its mid-1990s peak. London recorded its lowest annual homicide total since 2014 last year. Burglary is at historic lows. This isn’t the story people tell, but it’s what the 31,000-interview-a-wave national survey keeps finding.

Reading all that through, I felt some reassurance. On public safety and public health especially, things have genuinely improved, and we’ve stopped noticing.

Against that, some things genuinely got worse. Young people’s mental health is the hardest one. Between 2017 and 2023, the share of 17 to 19-year-olds with a probable mental disorder nearly doubled, from about one in ten to almost one in four, and it’s hit young women hardest. If you’re twenty in Isleworth in 2026, you’re not living in the same country as the seventy-year-old neighbour who thinks it’s all got worse. You might be having a harder time than she is.

As a parent, that’s the hardest paragraph in the piece to read — I’ve got two daughters in exactly that age range. What I keep asking is: what do they need from me, from Elma, from the community around them? The data doesn’t answer that.

And rough sleeping in London is at a record. The latest quarterly count put nearly 5,000 people on the streets, up year on year — with the long-term street-sleeping count up 18%. When a neighbour says they see more rough sleepers, they’re right. I walk the same streets, and I see it too.

Then there’s the honest bit that most of these arguments skip. The aggregate answer depends on who you are. A retired accountant in Osterley and a young Black woman giving birth in a Hounslow hospital are not living in the same 2026. Life expectancy in the poorest parts of the country has stalled and in some places fallen. The single starkest inequality in the NHS — that Black women are still far more likely than white women to die around childbirth — narrowed for a decade and then widened again in the most recent data. Women walking home after dark are running a different calculation from the one the aggregate crime fall describes, and they’re right to. The aggregate says better. For some of your neighbours, it isn’t. That inequity is stark, and on the doorstep, and it was difficult to hear. We’ve still got a lot to improve on.

Why we don’t believe the good news

There’s a reason the safer-and-healthier story doesn’t feel true. In 2023 two researchers, Adam Mastroianni and Dan Gilbert, published a study in Nature that pulled together decades of surveys across dozens of countries. In every era they looked at, majorities said people today were less kind, less honest, less moral than before. The 1940s respondent said it. The 1970s respondent said it. The 2020s respondent says it. They can’t all be right.

Their explanation has two halves. What we see now is oversampled for bad news — the news business runs on the vivid and the fresh. What we remember from before has been quietly airbrushed. Bad memories fade faster than good ones. Now looks bad because of what we see. Then looks good because of what we’ve forgotten. It’s not stupidity. It’s how normal minds work.

The UK evidence backs this up. Ipsos has asked Britons for years whether crime is rising. Most have said yes — during periods when it’s been falling fast.

The gap runs both ways

The perception gap isn’t just pessimism. It also hides real improvements we don’t give ourselves credit for. London’s air is measurably cleaner than it was a decade ago — roadside nitrogen dioxide is down by roughly half since 2016 — and most Londoners will tell you the air feels the same or worse. Fraud, meanwhile, has become the most common crime against individuals in England. In 2026 you’re more likely to be defrauded than burgled, mugged or have your car stolen. Yet most of us still fear the street attack more. One real thing is under-noticed. Another real thing is under-feared.

People need more help with fraud and cybercrime. That’s one gap where the data and the feeling are pointing the same way — it just hasn’t arrived in public conversation with the weight it deserves.

So — has it got worse?

Back to the dinner table.

“Isleworth has got worse” is a sentence that feels true because of how minds work, and because some of it is true for some people. Both of those things matter. Neither is the whole story. A child born here in 2026 will almost certainly live longer than one born here in 1976, be far less likely to die in infancy, far less likely to be a victim of violence. And some of their neighbours are having a harder time than their grandparents did.

I used to nod along with the guy at Elma’s dinner. I don’t any more, because I went and looked.

So here’s the ask. Next time someone at your table says it’s all got worse — including, occasionally, you — try three questions before you agree.

Which bit? For whom? Compared to when?

The answer is almost always more interesting than the reflex, and usually more useful.

And then a harder one, for the people who represent us — in the council, in Parliament. How are they addressing the things that genuinely are worse? How can we help the people in our own community who need more?

Focus on the things that actually need fixing, not on the loudest version of the story. Hand on the questions, not the verdict.

A longer version of this piece, with the full numbers, methodology and citations, is available as a companion reference.