How to persuade opponents of local housebuilding
Stack’s expertise is in helping campaigns and communicators to identify, understand and persuade their audiences.
Identifying and understanding audiences
This requires that we look past demography, for the simple and intuitive reason that not every person of a given age (or ethnicity, or sex, or tenure) thinks identically. Attitudes are the product of a much more complicated mixture of inputs from childhood, place, and our wider social and cultural context.
On the subject of development, our analysis is clear: the best predictors of whether someone will support or oppose local housebuilding are attitudinal not demographic. We ran a logistic regression model incorporating variables like age, home ownership and income, alongside attitudinal variables and found that the top predictors of opposition relate to how someone feels about their area. The graphic below ranks these variables by their predictive strength:
Relative influence of different demographic and attitudinal variables on opposition to new development. The size of each bar reflects the strength of the variable’s coefficient in a logistic regression model.
It is particularly striking that the more development someone thinks has happened in their local area, the likelier they are to oppose more of it. This is important because we have also found that regardless of how much development has actually happened, around half of people think there has been “A great deal, there has been constant development and major new projects” or “A fair amount, there have been several noticeable new developments”:
“How much new development (e.g., large building sites, new housing projects, infrastructure) have you seen happening in your area in recent years?”
This poses a real challenge to developers and local authorities, especially in those areas where there has been relatively little actual development (left-most in the graphic above). Each person’s perception of local development is of course unique to them, and is therefore relative, but it is also objectively true that some places have had much less development than others. We are testing the extent to which pointing that out can start to break down opposition, watch this space.
More widely, we also find that those opposed to development differ attitudinally in important ways from those who support it. To explain this, we should first set out how we segment the population.
For this exercise, we have adopted quite a simple segmentation: we factor in your view about national housebuilding (should there be more, the same or less), about local housebuilding, and about Green Belt development:
As an aside, a good way to highlight how attitudes transcend demography is to choose a demographic group and see how they split across this sort of attitudinal segmentation. Taking respondents aged 65+, we might assume they are almost entirely opposed to development, but in reality we find that 25% of this group are supporters. Or homeowners, amongst whom 31% are supporters:
Intuitively, those who support local housebuilding (YIMBYs and ABBAs) feel positive about change in their local area whereas opponents (BANANAs and NIMBYs) don’t, as shown in the spider graphic below. Interestingly, supporters and opponents are just as likely to feel their local area is beautiful, and to like it:
There is a stark difference between supporters and opponents on the question of whether their local area has got better or worse.
Persuading audiences
Having identified and started to understand our audiences, we can now try to work out what will persuade them in a given place.
But there’s a complication: a BANANA in Yorkshire is not identical to a BANANA in Norfolk. Understanding the relationship between attitude and location is fundamentally important in the development context, for obvious reasons. The winning argument in one place could well alienate people in another.
This is a problem political campaigns have been trying to solve for decades, and long before it became popular they have been using MRP to do it. Multi-level regression and poststratification (MRP) is a technique that allows you to accurately project findings from large-sample national-level surveys down to local level. Our modelling goes down to Output Area, the smallest unit at which the ONS produces census statistics, meaning we can estimate differences of opinion from street to street in a given ward.
To illustrate how this works, let’s take two equally attractive brownfield sites from LandTech’s analysis of opportunities. One is in Bulwell (Nottingham), the other is in Upminster (London), and in planning terms they are essentially indistinguishable. But attitudinally, they could not be further apart. The largest segment in most Output Areas in Bulwell are YIMBYs, followed by MIMBYs and in a few places NIMBYs. By contrast, Upminster is much more NIMBY, with the persuadable BYBY group second-largest across most of the area.
Stack’s segment mapped by Output Area around two LandInsight identified low density brownfield sites
We also know from our large and growing dataset of local attitudes that what people would prioritise for investment varies from place to place, and segment by segment, as do their concerns about development, the amenities they would prefer in a new mixed-use development, and so on.
In Bulwell, a NIMBY will be much more responsive to an argument about affordability for the next generation than in Upminster, where the winning argument will be about protecting green space and minimising the amount of land needed to meet a housing target, for example.
A pervasive concern nationally is that new local housebuilding will exacerbate existing pressures on local public services. This ranks as the top priority for investment in most parts of the country, in particular among opponents.
We have recently tested a specific argument to try and isolate trade-offs in a relatable way, using deliberative polling. We polled 6,731 UK adults in May 2025, to test an argument around the spiralling temporary accommodation bills faced by many local authorities, and the results are striking.
First, we baselined support for local housebuilding with our standard question: “To what extent would you support or oppose more homes being built near to where you live?” At net support of +17% this is in line with our megapoll analysis described in other blog posts.
We then presented respondents with new information, and re-asked support to see how effectively the information changed their mind:
Councils spent £2.29 billion last year on temporary accommodation for people without a home.
That cost is equivalent to weekly bin pickups for 30 million homes a year, or three years’ funding for every library in the country, or fixing over 30 million potholes.
If building more homes in your local area meant the Council spent less money on temporary accommodation, to what extent would you support or oppose more homes being built near to where you live?
The effect is huge: net support increases to +37%, and performs well across the age and political spectrum. Whilst it doesn’t convert the most vociferous opponents into net-supportive territory overall, it moves them a very long way:
Net support for local housebuilding before and after the respondent is shown an argument about temporary accommodation expenditure, by demographic and political cross-breaks
When it comes to persuading opponents, the truth is that there is no single silver bullet argument - what persuades opponents in one place may not work in another. We have just released an updated version of our strata app that gives you instant access to granular insight at a local level about people’s attitudes to development, to help you work out how to engage communities with their concerns in mind.
All of Stack’s data relating to development, planning and the built environment is now accessible via strata: check it out for free!
Explore the series: