Its relatively rare for the slightly esoteric field of strategic planning and the rather tortuous process of preparing a new London Plan to make the front page of the Evening Standard. But, sure enough, it did yesterday, as a result of the Secretary of State for Housing’s very public intervention and, after a hiatus of several months, has prompted me to finish writing a post I started I year ago.
I actually have some sympathy for the Mayor insofar as the ES reporting was quite the opposite of the effect of the letter, which was to agree the Mayor’s – lower – housing target for London, in exchange (one assumes, reading not too far between the lines) for some fairly significant changes elsewhere in the plan, along the lines of the Secretary of State’s ‘suggestions’ and a commitment to immediately start revising the plan again following its publication. (I suppose it is a shame we can’t plan to get it right first time.)
However, Mr Brokenshire’s comments about housing statistics slightly piqued my curiosity . As a reminder, he said:
“It is in the public interest that there is much more, and more regular, information in the public domain on housing delivery across London and I have asked my officials for advice on what can be done to increase transparency of the net additions to the housing stock in London.”
Which to me suggests that he has some queries about the data that we currently use. This is something (as someone who has a weak spot for pedantic arguments involving arid tables of data) that has bothered me before.
For all that we talk about building more homes (I suppose I should call it “delivery” although it seems to me that whilst one delivers a pizza, one builds a house) actually finding the stats on housing construction is a lot trickier than it should be.
The graph below shows comparable data from the various sources I’m aware of, that is, the London Plan AMR statistics and two sets of DCLG live tables on net additions and quarterly starts/completions. I’ve discussed these a little further below.
This illustrates the growing difference, in recent years especially, between the Government’s statistics and the Mayor’s, which is not far off 8,000 a year, or 20% of the current annual target.
If one delves into the stats – which I’m not planning on doing now – a similar pattern emerges on affordable housing delivery.
Given the vital importance of the housing targets – to the extent that changes to them are, literally, front page news, the fact that there seems to be such uncertainty over the extent to which they are being achieved is surprising.
There may well be an explanation for this – but I don’t know what it is. If one side effect of the current spat could be making these figures clearer and easier to access, that could only improve the quality of the conversation.
It will be interesting to see what the latest London Plan AMR report (now delayed from July to August) adds to the discussion.
A note on the sources
I am aware of various sources of information:
- The London Plan Annual Monitoring Reports – these tend to only go back 4 years or so, so one has to line up two or three to see long term trends to assemble the data. The headline figures now include non-conventional housing such as hostels and HMOs. ;
- The London Development Database – which the GLA has, to its significant credit, now made almost entirely accessible and on which the AMRs are ostensibly based. Although the data in the LDD does not seem to tie all that closely to the headline figures in the AMR.
- DCLG Live Table 122, giving net change in housing stock by borough, which the accompanying quarterly reports describe as the most complete dataset for overall net additions to housing stock, which is what we’re interested in;
- DCLG Live Table 253 which records quarterly starts and completions, based on building control information (and which, therefore, misses other sources of housing but is an indicator of direction); and
- London First / Grant Thornton’s annual review of housing completions, although this does not include permissions below 10 units and so is more useful for its examination of trends (especially the growing attrition rate), as opposed to the overall headline figures.