10 February 2023

Nimby campaigners added a zero to a key stat and nobody noticed (until now)


It really couldn’t have happened to a nicer lobby group. In their zeal to exaggerate the threat that housebuilding poses to our green and pleasant land, CPRE, the countryside charity, made a massive, avoidable and frankly hugely embarrassing error.

In a Westminster Hall debate this week on brownfield and green belt, Margaret Greenwood MP cited CPRE research claiming that ‘development of the highest quality farmland has increased 1,000-fold’. Data points that seem too good to be true almost inevitably are, and a 1,000-fold increase is suspiciously high. So I went over to CPRE’s website to check and found a press release from December 2022 entitled ‘Record breaking number of brownfield sites identified for redevelopment’, which included the 1,000-fold figure.

But the ‘previous research’ it cited from July showed that the amount of ‘Best and Most Versatile’ agricultural land set aside for development increased from 60 hectares to 6,000 hectares a year between 2010 and 2022. Which is, err, not 1,000 times. They simply added another zero to one of their key statistics by mistake, and left it there for months.

The CPRE have now, following my impassioned intervention, corrected their ‘typo’. They maintain that the 100-fold figure they now use is still ‘hugely worrying’. But it needs to be taken with an incredibly heavy dose of context.

The report from July from which the figure is taken cites two different estimates for high quality agricultural land: 2.2m hectares or 3.75m hectares. So that 6,000 hectare figure means we’re losing between 0.16% and 0.27% of our best agricultural land every year. Not really much cause to sound the alarm. Especially when you look at the Government’s figures for the overall amount of agricultural land in England over the last couple of decades, in which the loss of – to quote the CPRE – ‘huge quantities of productive land’ is notable by its near-complete absence.



Anyway, the fun doesn’t stop there. Quite a lot of the land they claim is ‘lost’ is covered by solar panels – which they include in ‘development’ or ‘land lost to housing and industry’. But obviously, if you put panels on, you can take them off. And in many cases keep farming underneath them, thanks to a remarkable invention known as ‘stilts’. (See the section here on ‘agrivoltaics’.)

But even if you include the solar panels, that’s still just 0.6% of good cropland lost over a decade. As the CPRE themselves say, that ‘sounds insignificant’. But then they claim that this equates to losing enough land to produce 250,000 tonnes of vegetables. ‘Oh no!’ you cry, ‘not our precious vegetables!’

But again the underlying data tells a different story. The CPRE report says that just 1% of the UK’s agricultural land was used to produce 52.7% of our vegetables. So those lost 6,000 hectares could be used to provide piles and piles of veggies.

But again, as with the 1,000 times claim, it pays to check out any statistic that seems suspiciously high. This level of agricultural productivity seemed remarkable. Could it, I asked myself, involve some kind of… polytunnels? Or greenhouses? And sure enough, when you follow the claim back to its source, you find that the 52.7% stat includes literally every piece of veg produced in this country. So the answer is yes, yes, it could.

In short: the CPRE erroneously promoted a figure for the amount of good agricultural land lost to development that was 10 times too high. The underlying data was based on defining solar panels on fields as development but not greenhouses or polytunnels. It ignored their report’s own admission that these land loss levels will probably appear ‘insignificant’ to the average observer. And it involved extrapolating from one year, 2020, when there was, as they admit, a ‘particular spike’ in projects – as well as starting the clock running in 2010, when development was at the bottom of its post-crash trough, and was axiomatically bound to increase. (It would be great to see how the figures for the 2010s compare with the 2000s, 1990s or 1980s – but the CPRE don’t make their data available.)

Even that isn’t the end of it. Because the CPRE claims, again and again, that any and all building on greenfield sites is unnecessary because we can fit it all on brownfield. But that claim has been utterly demolished, not least in the latest report from the Centre for Policy Studies, the think tank I run (which also operates CapX).

And even if it were true, the CPRE would still be wrong. Because the same press release from December that included the rogue stat also claimed that ‘the demand for social housing is growing six times faster than the rate of supply in rural areas’. At current rates, it warned, ‘the backlog of low-income families needing accommodation would take over 120 years to clear’.

But one of the most obvious things about brownfield land, as that CPS report pointed out, is that there’s not much of it in the countryside. It’s true that the CPRE do accept that we may need some greenfield housing in the countryside as long as it is ‘primarily affordable homes for local needs’. But they don’t say in that report how much of this they’d accept, or qualify their complaints about greenfield development by subtracting the affordable housing that is already being built. And meanwhile, they consistently denounce essentially insignificant levels of green belt and greenfield release as shocking crimes against the countryside. The result is a policy cocktail that would ship the young rural poor out into the towns and cities. That’s a version of protecting the countryside. But not one that many of us would sign up to.

To make one statistical error looks like carelessness. But CPRE is consistently slippery with the facts. Let’s stop treating them as pure-hearted guardians of the countryside, and give their arguments the scrutiny they deserve.

Robert Colvile is Editor-in-Chief of CapX.