Every food business in England has to register with its local council. Not optional, not a directory listing anyone can edit: a legal requirement, going back to the Food Safety Act.
That makes the Food Standards Agency's register something unusual. It's a complete count of a thing nobody normally counts.
So we counted it. Across 264 English districts there are 43,912 takeaways and 98,679 restaurants and cafés.
The ratio that tells you something
Raw counts mostly just measure how big a place is. The interesting number is the ratio: how many takeaways a district has for every restaurant or café.
The English average is 0.49. The spread around it is enormous.
| District | Takeaways | Restaurants | Ratio |
|---|---|---|---|
| Newcastle-under-Lyme | 120 | 86 | 1.40 |
| Rochdale | 327 | 241 | 1.36 |
| Stevenage | 167 | 131 | 1.27 |
| Barking and Dagenham | 230 | 232 | 0.99 |
| Sunderland | 372 | 376 | 0.99 |
| Richmond upon Thames | 66 | 730 | 0.09 |
| Kensington and Chelsea | 84 | 971 | 0.09 |
| Westminster | 284 | 3,504 | 0.08 |
Newcastle-under-Lyme has 1.40 takeaways for every sit-down place. Westminster has 0.08. That's not a small difference in degree, it's two completely different high streets.
Yes, it tracks deprivation. No, it doesn't explain it.
Run the ratio against the official index of multiple deprivation across all 264 districts and you get a correlation of about 0.53. The most deprived fifth of districts averages roughly double the ratio of the least deprived fifth, and the pattern holds cleanly through every quintile in between.
None of which is new. Researchers at Cambridge's CEDAR unit tracked takeaway density in Norfolk from 1990 to 2008 and found outlets were significantly more common in deprived areas at every point measured, and that the gap widened over those eighteen years (Maguire, Burgoine and Monsivais, 2015). An earlier Glasgow study found much the same (Macintyre et al.). We are confirming a known pattern at national scale, not discovering one.
Which is usually where the reporting stops. But the exceptions are more interesting than the trend.
Newcastle-under-Lyme has the highest ratio in England on below-average deprivation. Stevenage is third highest and is not a deprived town. Meanwhile Westminster's ratio is the lowest in the country despite middling deprivation scores, because central London's 3,504 restaurants feed tourists and office workers rather than residents.
Which is the point. The ratio describes what's on a high street. It doesn't diagnose why, and anyone using it to explain a town's problems is reading more into it than the data holds.
What we can't count
Charity shops. There's no register of them, because nobody has to tell anyone they've opened one. We checked three sources: the FSA register doesn't cover them, OpenStreetMap found 52 in Bolton but its coverage of other categories ranged from 44% to 99% complete, and official business counts file them under the parent charity's industry code, so they show as zero.
Britain genuinely does not know how many charity shops it has. Same for hairdressers, until you go to the business register, which counts premises rather than shopfronts and gives a different answer again.
The full district-by-district table, with per-resident rates, is on the takeaways and restaurants page.