This paper uses the British Household Panel Survey (1991-2008) and the UK Census (waves 1991, 2001 and 2011) to examine the association between gentrification and displacement in English cities. Gentrification is the phenomena of a large and relatively sudden in-migration of wealthy or middle class residents into a previously poor or working class neighbourhood. I measure this using the change in share of neighbourhood population that holds a degree certificate conditional on the initial share. The effect is found to depend crucially on household income and homeownership status as well as how gentrified the initial share. In a ward with an initial degree share of 5%, a renter household with an income of 0.5 times travel-to-work-area (TTWA) median is found to be 1.5 percentage points more likely to exit for each percentage point increase in degree share, after controlling for other factors. However, if the household income is 1.5 times the TTWA average then this effect becomes statistically insignificant. For homeowners, on the other hand, the effect remains stable at around 1pp across all income levels. This result may be explained by institutional context since, in the UK, property tax is only applied at the point of sale. This means that a low-income homeowner facing increasing property value is not forced to sell due to escalating tax assessments, as may be the case in the US. So whilst the rising housing costs associated with gentrification do induce homeowners to sell, there is no income dimension to this effect as there is with rising rental costs. These results provide a significant contribution to the literature on displacement, which has previously found little evidence for the effect.
