Does living near a mega farm affect your house price - a UK analysis
3.4. Reductions in property values
Contrary to previous studies, we did not find that houses closer to Industrial Meat Producers sold for significantly less than houses further away from IMPs.
However, we found that participants reporting living near an IMP were significantly less likely to agree that buying a house in their area was a good investment.
Respondents that lived near an IMP were no less likely to agree that their area is safe to raise a family, and did not note any loss of community or belonging as a result of living near an IMP. Lastly, most of them disagreed that their house would be worth more if there was not an IMP nearby.
Whilst survey respondents living closer to IMPs were less likely to say that buying a house in their area was a good investment, They did not think the IMP was the primary cause. This contrasts with research from the USA, which finds losses of 3-26%. It is possible that IMPs do damage house prices, but only in certain conditions that we could not measure; it may be that houses only lose value when the IMP is downwind (as this carries odours and pollution), or only for the largest IMPs.
Another explanation for this is that most IMPs in the UK have been established in the last decade or so, whereas US confined animal feeding operations have been “hollowing out” rural economies since the 1960s. Respondents in our survey who lived near IMPs did not report loss of community or belonging and did not feel their area was less safe for their families. Whilst this is a positive sign, we caution that this should not be taken for granted: ongoing monitoring trends in house prices near IMPs is advised.
Estimating reductions in property values due to proximity to industrial meat production facilities
To estimate the effect of proximity to industrial meat production facilities on house prices, we used multilevel regression modelling to test whether houses closer to IMPs sold for lower values than those further away from IMPs. We used our sample of British households that sold in 2023 (see Appendix B), but restricted our analysis to houses with at least 1 IMP within 10km (67,989 houses). We removed a further 1,045 houses with a sale value of over £1M as these extreme outliers skewed the distribution of house prices.
We fit a linear mixed effects model predicting house selling price from the distance to nearest IMP. We also included a binary variable of whether the house was a new build, as well as dummy variables for property type (flat, terrace, semi-detached, detached and other). We included random intercepts for district of the UK, postcode of the house and the identity of the nearest IMP. These random intercepts account for the fact that many houses have the same postcode, many districts of the UK naturally contain higher or lower house prices for reasons unrelated to IMPs, and some IMPs may have particularly strong effects on house prices. We also included a random slope of distance from IMP on the identity of the IMP. This models the possibility that different IMPs have different effects on house prices. This is because some IMPs are significantly larger than others so are likely to have stronger effects.
However, we did not find a significant effect of proximity to IMPs on selling price (p > 0.05). Including an effect of distance to nearest IMP squared (which models the hypothesis that the negative effect fades quickly with distance from a given IMP) also did not yield significant results. Additionally, natural log transforming house price (which represents the hypothesis that each km closer to an IMP decreases house price by a fixed percentage) also did not show results.