Does Religious Mix Affect Local House Price Dynamics?

Connected Neighbourhoods
Does Religious Mix Affect Local House Price Dynamics?
Nema Deana and Gwilym Pryceb (AQMeN Urban Segregation & Inequality Project)
a
School of Mathematics and Statistics, University of Glagow; b Sheffield Methods Institute, and Department of Urban Studies and Planning, University of Sheffield
Homophily in the housing market
Religious homophily
 Why do cities tend to self-segregate?
 One theory is that it’s due to “homophily”—the
tendency for "birds of a feather to flock together"
 Homophily is our propensity to be attracted to people
most like ourselves.
 Homophily can also be applied to connections
between neighbourhoods.
 We look at the extent to which neighbourhoods with
similar religious make-up tend to be connected in
terms of house price movements.
 Is it only the physical similarity of dwellings and
neighbourhoods that determines whether buyers consider
them to be close substitutes? Or does religious mix also
affect perceived substitutability?
 We measure perceived substitutability by computing cross
price elasticities which measure how strong changes in
house prices are connected between neighbourhoods.
 Religious homophily exists when neighbourhoods of
similar religious mix are more likely to be perceived as
close substitutes.
New method for measuring perceived homophily between neighbourhoods
 We first estimate for each pair of locations the strength of relationship in house price movements by computing a
substitutability map for each location (see below plotted for one of the 10,000 postcodes in the data)
 Each map is based on the cross-price elasticities between one location and all the others in the data.
 We then apply social network analysis to the 10,000 nodes, where each node is a postcode.
 If the cross price elasticity between any pair of postcodes is greater than 0.9 then they are said to be connected (i.e. a
pair of nodes are connected if they are perceived by the market to be close substitutes).
 We then estimate homophily coefficients to see whether similarity in neighbourhood religious mix increases the
likelihood that a pair of postcodes are perceived to be close substitutes.
Substitutability surface for location D120
Homophily coefficients
 Map shows perceived substitutability of other
locations in Glasgow respect to location D120
 Substitutability is measured by how closely price
movements of those other locations are related to
the prices of dwellings at D120
 Why are some locations perceived to be much
closer substitutes to D120 than others?
Want to know more?
 Dean, N. and Pryce, G. (2015) “Connected
Neighbourhoods: Does Religious Mix Affect Local
House Price Dynamics?”, AQMeN Research Briefing
9, www.aqmen.ac.uk.
Religious
Attributes
Physical
Attributes
Dwelling D120
(Positive values indicate homophily. All coefficients are statistically significant)
Key findings:
 Our results suggest that the Glasgow housing market is
not blind to religion – closely related price movements
between areas (our way of measuring perceived
substitutability) are less likely to occur if those areas are
different in terms of religious mix.
 The homophily effect is as strong for certain types of
religious mix as it is for physical characteristics such as
type of dwelling and proximity to railway stations.
House price data are from GSPC sales. Neighbourhood attribute data are based on Census, Meridian, and Experian data. This work was funded by the Economic & Social
Research Council through the Applied Quantitative Methods Network: Phase II, grant number ES/K006460/1. Contact: Prof Gwilym Pryce, Sheffield Methods Institute, University
of Sheffield. Email: g.pryce@sheffield .ac.uk.