Meet the sector’s junior researcher of the year

For the prize winner, the research idea was not just about boosting her profile or advancing her career; it was also to get something off her chest.

“I grew up in social housing…I take personal offence when people say social housing has a negative impact on you growing up,” says Ms Martin.

The paper sought to challenge the findings of previous studies that typically found children in social housing have worse development outcomes than those in other tenures - and these negative outcomes are a result of the tenure itself.

Ms Martin used data from the Millennium Cohort Study, an ongoing study tracking the lives of 19,000 children born in 2000-01. The dataset includes information about parenting, education, cognitive performance, health and parents’ employment and income.

“I grew up in social housing…I take personal offence when people say social housing has a negative impact on you growing up,” Anya Martin

Ms Martin used a technique called regression analysis to control for 16 variables, including income, age, benefit claims, health and ethnicity. This enabled her to determine whether differences in childhood development outcomes are likely to be caused by the tenure itself, or whether other factors, such as selection bias, are in place.

“People are often allocated social housing because they’ve experienced some kind of personal, economic shock or crisis,” she says.

The research found that once these factors are controlled for - ie held to be equal - children of social housing tenants are no worse off than those of private renters. The conclusion differs from previous studies, including a piece of research by US academics in 2002, that have found social tenants are worse off than those in other tenures even after factors have been controlled for.

The paper says: “The fact these relationships disappear when a set of control variables are introduced means there is not any evidence of a negative tenure effect from social housing.”

The research also found that there was no positive tenure effect either, as may be expected from lower rents and better security.

Ms Martin says that the tenure effect might be “non-existent”, and in fact differences in outcomes are due to the demographcis of people in social housing.

Another explanation for her results, the paper says, is that there is a tenure effect and the fact there is appears to be no different between outcomes among social and private renters, where previously there was, may be to do with improvements following the Decent Homes programme.

Click below to download Anya Martin's full research paper

Mr Aldridge was fulsome in his praise for Ms Martin’s work, even going as far as saying he could envisage it as the type of report that could be used to aid policymaking at MHCLG.

He said: “There has been a lot of debate about this issue (the impact of social housing on outcomes) with people trying to get a handle on the facts. Behind the clear explanation is good, robust analysis.”

Mr Aldridge added that he had seen regression analysis methods used in similar ways to measure unemployment, but never before to track children’s development.

Professor Gibb was impressed at the use of cohort study data and said the research findings “should be a warning to policymakers not to make simplistic assumptions”.

Each paper was scored one to five on six aspects: writing style and clarity; engagement with literature and theory; methods; empirical rigour and theoretical depth; strength of conclusions; extent to which research was outcome and impact-focused.

Professor Ken Gibb, professor in housing economics, University of Glasgow and Thinkhouse editorial panel member

Dr Beth Watts, senior research fellow, Heriot-Watt University in Edinburgh and Thinkhouse editorial panel member

Gemma Duggan, head of compliance and performance, ExtraCare Charitable Trust and Thinkhouse editorial panel member

Richard Hyde, chair of the Thinkhouse editorial board and chief executive of a small business selling tools to the sector

Stephen Aldridge, director of analysis and data, Ministry of Housing, Communities and Local Government

Steve Douglas, co-chief executive, Altair

Emily Pumford, a researcher at Riverside, finished as runner-up with her use of a linguistic technique called ‘politeness theory’ to gauge whether comments made by ministers signal that the government’s attitude towards social housing has shifted.

Ms Pumford, 26, has a “passion for linguistics” and decided to analyse speeches made at the National Housing Federation’s annual conference over the past few years, including addresses given by ministers Greg Clark, Gavin Barwell and Sajid Javid, as well as Theresa May’s appearance at the conference this year. She looked at the extent to which the politicians used ‘negative politeness’ to distance themselves from social housing and ‘positive politeness’ to establish closeness with the audience.

"Ms Pumford says her analysis shows there has been a positive shift in government’s attitude towards social housing, however, it is difficult to see how this fits with a continued focus on home ownership"

Negative politeness strategies can include “impersonalisng the speaker and hearer, by avoiding using pronouns ‘I’ and ‘you’”. She cites passages of Mr Clark’s speech in 2015, in which he avoids using ‘I’ and ‘you’ and instead refers to the “housing association sector.”

Positive politeness strategies involve “establishing a sense of commonality and familiarity between speaker and hearer” and could include using the pronoun ‘we’ and the expression of common interests. An example of this is Theresa May earlier this year, who said: “It is a challenge we must rise to together.”

The conclusion? Ms Pumford says her analysis shows there has been a positive shift in the government’s attitude towards social housing; however, it is difficult to see how this fits with a continued focus on homeownership.

Mr Aldridge describes Ms Pumford’s work as “an interesting application of a theoretical concept” but says that it is more difficult to see how the work could be applied than the winning entry.

Click below to download Emily Pumford's paper

"Given that social housing is allocated to those in need, and typically after an extended period of deprivation, it is perhaps unsurprising that on some baseline measurements children in social housing have poorer outcomes. It is widely acknowledged and documented that poorer children tend to have poorer outcomes. If anything, it is more surprising that the baseline differences are so minor.

The fact that these relationships, where they exist, disappear when a set of control variables are introduced means that there is not any evidence of a negative tenure effect from social housing. In other words, social housing itself does not appear to create worse outcomes. The worse outcomes of children in social housing on some variables can be explained by the fact that they are disproportionately deprived, not because their housing situation negatively impacts on their outcomes. However, there was also no evidence of a positive tenure effect on children, as this paper hypothesised may originate from having lower rents and better security.

This is significant as most existing literature (based on older datasets) typically finds there is still a difference even after other socioeconomic characteristics are controlled for. Previous authors (Lupton et al., 2009) have observed that the correlation between tenure and life outcomes has changed over time, going from no substantial difference in later life outcomes (after controls) for those born in 1946, when social housing in childhood was far more widespread, to more substantial differences in the 1970-born cohort as social housing has been residualised as a safety net for the most in need.

"That this paper no longer find evidence for correlations that have been observed in the past could mean one of two things. One possibility, fitting the tenure effect hypothesis, is that the relationship between tenure and outcomes has changed. When social housing was more widespread in the mid-20^th century, the housing was of good quality relative to the private sector, meaning no observable negative tenure effects. As the original buildings aged (and were not adequately maintained) near the end of the 20^th century, this impacted children growing up there negatively. And as the Decent Homes Programme improved social housing (and indeed pushed it above the quality of private rented housing) in the early 21^st century, social housing again sees itself compare favourably to other tenures.

"Another possible explanation, fitting the selection effect hypothesis, is that the tenure effect is non-existent, and the changing fortunes of children in social housing relative to other tenures is a result of selection bias. As identified earlier, social housing allocation used to be more widespread and less related to need, meaning the households in social housing were more or less average in their socioeconomic characteristics. They would therefore be expected to be more or less average in child outcomes. As supply dropped off in the 1980s and 1990s, social housing became a more needs-based tenure, and as a result the people entering social housing were on average of lower socioeconomic status than previously. This may explain the worsened outcomes of children in social housing as compared to other tenures.

"Additionally, given the lack of observable relationship between tenure and the set of child developmental outcomes, it is unlikely that tenure is a significant causative or deciding factor influencing these outcomes. This is important to emphasise, as even where statistical studies of this magnitude can highlight interesting statistical differences, it is easy to overestimate the importance of tenure to any one individual. In reality, children’s lives and outcomes are impacted by a range of much more substantial and personal issues than whether their parent’s landlord is a private individual or a registered social provider.

"It is possible that both the tenure effect and the selection effect are at play to some extent (and indeed the data appears to match the historical rationale behind both). It is not possible to come down in favour of one or the other. However, this research has been able to re-assess the observation that children in social housing are worse off relative to their privately renting peers, in the light of more up-to-date data from a major national dataset. That the relationship between tenure and child outcome appears to be historically contingent challenges the notion that social housing is an inherent cause of worse outcomes, and provides impetus to social housing providers to continue their work to improve the quality of the estates and properties they manage."

Extract from The impact of social housing on child development outcomes, by Anya Martin

Thinkhouse was formally launched in spring 2018, and aims to “provide a single location and summary of the best and most innovative research pieces, policy publications and case studies”.

It specifically looks at reports that propose ways to boost the amount and quality of housing and the economic, social and community issues of not doing this.

The Thinkhouse editorial panel highlights the ‘must-read’ reports, blogs about them and runs the annual Early Career Researcher’s Prize.

The panel includes current and former housing association chief executives, academics, lawyers, economists and consultants. It is chaired by Richard Hyde, chief executive of a business that sells construction hand tools.

Who is on the panel?

What should have been covered in 2017 but was not? Here is what the Thinkhouse editorial panel members think:

• Gemma Duggan: Have the reports in 2017 recognised that the (degree of) shift by the government towards again accepting social housing as an important part of policy, which started after the referendum and change of PM in 2016, was energised further following the Grenfell Tower catastrophe? What scope is there, at this late stage, to promote relevant research for the Social Housing Green Paper due in early 2018?
• Suzanne Benson: Research into the direct experiences of social housing tenants and those renting from private landlords seemed to be well covered in academic circles, but did not seem to hit the mainstream media. How can the homeownership and rental tenures (both social, affordable and private rented sector) be more effectively linked together to recognise that individuals do not have static housing needs? More thinking on the linkages between the tenures that work in the market and how they can be used more dynamically to benefit people as they move through different life stages and income brackets would help with both monitoring the demand side of the housing question and planning a more targeted supply.
• Kerri Farnsworth: It would have been helpful to have had more on the cost of housing and the impact on living standards. This area of study seems to have become fallow since John Hill’s comprehensive report 10 years ago. In particular providing robust cost-benefit analyses evidence and monetisation of the linkages between housing and health would have been useful. From a macro perspective, what about the linkages between housing, agglomeration and economic growth?
• Boris Worrall: Given the amount of public land and the failure to release significant chunks of it without the public sector scrapping to get ‘best value’, it would have been interesting to have some further research, which builds on recent pieces from Civitas, on how this has held back housing.
• Martin Wheatley: As the government starts, post-Grenfell, to look again at estate regeneration, perhaps it is time to widen the debate beyond just bricks and mortar to looking at the benefits of regeneration in a wider social, physical and economic context that is not restricted/can cross local authority boundaries. This may also be pertinent in helping shed light on some of the current battles between residents and councils battling over radical estate demolition/rebuild.
• Chris Walker: Are there some devolution lessons that can be applied nationally? For instance the Scottish rent pressure zones?
• Mick Laverty: Does the fragmented housing association sector create a structural deadweight that limits the actual number of houses it can build?
• Beth Watts: Finally, although such lists are never really final, are we promoting new and young researchers to take a new/diverse viewpoint? Should Thinkhouse offer a cash prize for the best piece of housing research/analysis completed by someone under 30?