Just wondering if you've deliberately used the New Build figures rather than the Net Additional Dwellings dataset? I only just noticed this difference on your post with every single Authority having their own chart, Brighton and Hove comes out the worst, with only 14.8 completions per 1000 existing households.
Brighton does have a bad building record, but not *that* bad - its Net Additional Dwellings are 4,904 over the decade, against a starting point of around 125k households - so that's more like 39.2 additional dwellings per 1000, so it's a difference of 2.5x or so compared to the data set you're using.
I asked Perplexity to explain the difference between the two sets, it told me:
New Build Completions counts only new house building starts and completions
Net Additional Dwellings measures net change in total housing stock, which includes:
- New build completions
- Conversions (e.g., turning a house into flats)
- Change of use (e.g., office to residential)
- Demolitions (subtracted from the total)
I suspect this probably skews the conclusions a bit - I would guess that the most constrained places are more likely to get dwellings from conversions/change of use than from new builds, so it might change the shape of the chart somewhat. I wonder if the comparison to the US is also changed: I don't know if the US figures you used would include conversions etc.
Fair criticism – the US figures are labelled 'annual units built' so I went straight to new builds. I agree this might distort some places such as Brighton. I also think we should build more, in part because huge amounts of our housing stock is so old and small (so the overarching normative point doesn't change). But yes - the analysis might look different with net dwellings. I'll update the github repo shortly and you can have a look!
Hi! Great work, really interesting topic and read.
I wondered, is there a link if private enterprises are involved in the building that they are choosing to build in places where they are able to sell at a good price relative to those areas?
For example, in Manchester I would be surprised if lots of housing was built in a lesser economically developed neighbourhood as the margins wouldn't be there? Whereas in wealthier areas in the south east, I would expect there are more reasons to come in with a small burst of additional supply and be able to sell at the same rate as other older houses
Thanks! Yes, that's absolutely the case - it's how markets behave. Price of sandwiches goes up, more people are incentivised to make sandwiches. If sandwiches are more expensive in Greenwich than Glasgow, more people sell them in Greenwich. Exactly what we see.
And this is what we want - if sandwiches are more expensive in Greenwich, it's because more people want them in Greenwich. Much better we make sandwiches where people are hungry.
The trouble is that the planning system says "no, you can't make sandwiches here, you have to make them in Glasgow" in various ways and for various permutations of "sandwiches" and "here" and "make". There are some good reasons why we – the people – would want to control some of that. But there are lots of problems too.
Just wondering if you've deliberately used the New Build figures rather than the Net Additional Dwellings dataset? I only just noticed this difference on your post with every single Authority having their own chart, Brighton and Hove comes out the worst, with only 14.8 completions per 1000 existing households.
Brighton does have a bad building record, but not *that* bad - its Net Additional Dwellings are 4,904 over the decade, against a starting point of around 125k households - so that's more like 39.2 additional dwellings per 1000, so it's a difference of 2.5x or so compared to the data set you're using.
I asked Perplexity to explain the difference between the two sets, it told me:
New Build Completions counts only new house building starts and completions
Net Additional Dwellings measures net change in total housing stock, which includes:
- New build completions
- Conversions (e.g., turning a house into flats)
- Change of use (e.g., office to residential)
- Demolitions (subtracted from the total)
I suspect this probably skews the conclusions a bit - I would guess that the most constrained places are more likely to get dwellings from conversions/change of use than from new builds, so it might change the shape of the chart somewhat. I wonder if the comparison to the US is also changed: I don't know if the US figures you used would include conversions etc.
Net Additional Dwellings can be found here: https://www.gov.uk/government/collections/net-supply-of-housing
Thanks for the post, btw! I'm thinking i will try to extend your analysis by taking into account some of the "housing need" figures, although so far I've only managed to find current ones rather than historical (https://www.stantec.com/uk/ideas/analysis-englands-local-housing-needs-2024-by-stantecs-development-economics-team)
Fair criticism – the US figures are labelled 'annual units built' so I went straight to new builds. I agree this might distort some places such as Brighton. I also think we should build more, in part because huge amounts of our housing stock is so old and small (so the overarching normative point doesn't change). But yes - the analysis might look different with net dwellings. I'll update the github repo shortly and you can have a look!
Hi! Great work, really interesting topic and read.
I wondered, is there a link if private enterprises are involved in the building that they are choosing to build in places where they are able to sell at a good price relative to those areas?
For example, in Manchester I would be surprised if lots of housing was built in a lesser economically developed neighbourhood as the margins wouldn't be there? Whereas in wealthier areas in the south east, I would expect there are more reasons to come in with a small burst of additional supply and be able to sell at the same rate as other older houses
Thanks! Yes, that's absolutely the case - it's how markets behave. Price of sandwiches goes up, more people are incentivised to make sandwiches. If sandwiches are more expensive in Greenwich than Glasgow, more people sell them in Greenwich. Exactly what we see.
And this is what we want - if sandwiches are more expensive in Greenwich, it's because more people want them in Greenwich. Much better we make sandwiches where people are hungry.
The trouble is that the planning system says "no, you can't make sandwiches here, you have to make them in Glasgow" in various ways and for various permutations of "sandwiches" and "here" and "make". There are some good reasons why we – the people – would want to control some of that. But there are lots of problems too.