Housing


Improving the allocation of social housing.



Ongoing


PARTNERS

Innovate UK, Enfield Council, Hammersmith and Fulham Council and Bradford Council


Share

Public housing in the UK is allocated through a range of lettings systems. These vary from choice-based lettings systems which allow the prospective tenant to ‘bid’ on selected properties, to systems where council workers directly assign properties. These processes are often mathematically suboptimal, producing undesirable outcomes that damage resident wellbeing.

HAUS is a research project funded by Innovate UK, conducted in collaboration with academics from NYU, Brown University, the University of Chicago, and the University of Oxford, which aims to improve the process of allocation by incorporating insights from matching theory.


Image

From theory to reality

The central problem in housing allocation is very similar to that faced in organ donation; how to allocate goods and services when units are not all compatible, without using prices? Theoretical results in the literature focus on allocation mechanisms which find their way to the best possible outcome given a set of houses, and a set of prospective tenants.

Our proposed solution is based on the Multiple-Waitlist Procedure, which allows applicants to trade off between their preferences over housing characteristics, and their waiting time. This in turn gives us a better understanding of applicant priorities, and ensures people are allocated a house which suits them. The estimated social benefits of moving to this better procedure have been conservatively estimated at approximately £5,000 per household by the academic advisors for this project.

With an intended design in mind, HAUS is focused on the implementation of this procedure in practice, working with councils to understand how they allocate housing, and how the processes could be redesigned to implement the proposed solution. In particular, we are currently working on understanding the interaction of temporary and permanent housing systems.


 
Explore more solutions