Housing Need and the Need for Housing

Richard Fordham, Stephen Finlay, Justin Gardener, Angus Macmillan, Cecilia Muldoon, Geoff Taylor, Gary Welch

• Publisher: Taylor & Francis
• Year: 2019
• Print ISBN: 9781138340640
• eBook ISBN: 9780429803154
• Edition: 1
• Topic: Social Sciences
• Subtopic: Sociology
• Index: Social Sciences

First published in 1998, this book addresses a central issue in housing: that of 'need' and how to meet it. Need refers either to individual household circumstances or to the aggregate requirement for new (social and market) housing. The book develops a new policy mechanism to address the present fragmented policy situation. The mechanism now used to allocate finance and new social housing numbers is complex and ineffective in terms of meeting needs. The book's proposals are rooted in clear definitions of the key terms, especially that of housing need, which suffers from lack of any clear definition. The analysis is developed using a large database of survey information covering England and Wales. The proposed policy mechanism should be of great interest at a time when radical change to local government are being considered. The proposed mechanism would greatly increase value for money of public time and expenditure on (planning for) social housing.

Appendix 1 Data Sources Used for Studies in This Book

A1.l Introduction

A number of analyses have been carried out using the large database derived from our housing needs surveys. These analyses have been done at various times and using various sets of data. This appendix describes the data sets which have been used.

Throughout this book, reference has been made to various examples drawn from data sets collected in the housing needs surveys carried out by Fordham Research. By the time of writing, Fordham Research has been commissioned to undertake about 70 housing needs surveys, therefore the amount of data available is vast and allows accurate analysis of factors relating to housing need. Not all data sets used for analysis in the book are the same, so it is worth quickly outlining the various sources used, and the reasons for these.

The data sets and where they have been used in the book are summarised below. We have identified them as ‘Database 1, 2, etc.’ for ease of reference. Each is intended for a slightly different purpose.

A1.2 Benchmark Study: Database 1

Database 1 was a study of comparable data already available from the first 49 housing needs surveys carried out by Fordham Research. Although the questions asked (both in numbers and wording) differed slightly between councils studied, the general results collected are of a comparable nature.

The benchmark study considered housing need and factors relating to it. In addition, councils have been grouped in relation to size, type (urban/rural) and geographical location, to find patterns between these. This set of data has been widely used as a comparison with already available sources of information, such as the GNI at the local authority level.

A1.3 Combined Data Set: Database 2

When we set out to write this book, we were aware that considerable additional analysis would need to be carried out to support our arguments. Although on an individual survey level results are accurate, it was felt necessary to use data from a number of councils, so as to boost sample sizes (and hence statistical accuracy) and also to cover different areas with differing characteristics (e.g. high/low need, high/low proportions of owner-occupiers) as well as different geographical locations.

As a result, 13 different councils were chosen, where an element of personal interviews were included within the data collection process. The data from these councils was then amalgamated into one large data set, comprising 8,252 interviews. The data includes councils with high and low levels of need, councils from the North and South of England, London and Wales. In general, it was thought to give a representative picture of the surveys carried out by Fordham Research, and is representative of over 770,000 households across the country.

All the data was weighted to represent the estimates of the number of households in each of the areas studied. Due to the fact that the two London boroughs included were far larger than any of the other individual councils, the results are slightly biased towards these. This gives the impression of a slightly higher level of housing need than we had been expecting from this data set. This does not seriously effect the overall results from this source.

A1.4 Comparing Methodologies: Database 3

Section 4.2.3 is based upon the original version of an article which appeared in Housing (Fordham and Gardner, 1997). As the original article was written some months before the conception of this book, the Councils included in the dataset are different from those included in database 2 above (although if the same analysis were to be carried out using database 2, the results would almost certainly be similar). The data comes from six surveys with a significant number of personal interviews (4,800 in total). These were chosen to be geographically distinct, and also distinct in terms of rural/urban nature, and are thought to be fairly representative of the housing needs surveys carried out by Fordham Research.

A1.5 Mode of Survey: Database 4

The data set used for this is different again to those above. This work was also carried out before the beginnings of this book, and so differs from the combined data set. At the outset of comparing personal and postal results, it was just a study for internal reasons, following a request from one council to carry out a postal-only survey (something we had previously been wary of doing). Hence, the results come from a random selection of six councils which had an element of both postal and interview survey types administered to the same population at the same time. Whilst this is not the most scientific way of choosing data sets, it should prove adequate for the purpose for which it is being used (there is no evidence that any geographical area is better or worse at filling in forms than any other).

A1.6 Equity Analysis: Database 5

The analysis used comes from an individual council where levels of owner-occupation were high, as were the numbers of households responding to the question of existing equity. It is not intended to show definitive results, but merely to highlight the potential for existing equity to solve some households’ housing need. The larger combined data set was not used for this purpose because, in a number of cases, a question about existing equity was not asked (indeed, it is only over the past couple of years that we have regularly attempted to collect such information).

Appendix 2 Mode of Survey

A2.1 Introduction

The following is a discussion of the various modes of surveys commonly employed in the assessment of housing need. The topic is peripheral to the main concerns of the book, but is worth including as an appendix because, in the end, need will be measured by one or other of these modes of survey, and it is important to use these modes to the best effect in such measurement.

There have been many arguments for and against different types of survey mode in housing needs surveys (see, for example, Couttie, 1996). Indeed, due to its prominence, ‘methodology’ is often mistakenly taken to refer to the choice of mode, rather than to the method of approach in defining need. Choice of mode of survey is the other main issue in deciding the form of a housing needs survey, once the broad methodology addressed in Appendix 1 is decided.

Fordham Research have carried out over seventy housing needs surveys, over half of which have included some element of both interviews and postal survey. This provides an unrivalled database to study the effect of mode of survey. This appendix sets out the results of our analysis.

The sample for the following analysis (Database 4 - see Appendix 1, Section A1.5) results from six surveys where both modes of survey were used. The database for the analysis includes both urban and rural areas.

A2.2 Context for Choice of Mode

Essentially, there are two primary survey methods: personal interviews and postal surveys. The only other choice, telephone surveys, has not to our knowledge been used in such surveys, and they are certainly not the best method of data collection.

The key issues in the choice of survey mode are cost and response rates. Personal interview surveys are more expensive than postal ones (at least six times more), but achieve higher response rates. For surveys of our level of complexity, response rates for interview surveys are around 75%, and those for postal surveys 40%. As postal survey response rates in particular vary depending on the length of questionnaire, this finding refers specifically to our type of questionnaire.

It has been suggested that of those not responding in the postal survey, a disproportionate number of households are not in need: since the survey is called a ‘housing needs survey’, households which consider they have no housing need might be less likely to reply. This issue is discussed below.

A2.3 Accuracy of Income Data

A2.3.1 Background

Income data is something that any housing needs survey must collect. It is used to assess affordability, so accuracy is crucial. This section looks at the merit of each mode of survey for collecting such income data.

Accuracy in the collection of income data has often been cited as one of the main reasons for using either a postal or interview method, indeed different writers have argued in favour of both methods. We have argued that an interview survey is a better method for collecting this information because an interviewer is able to explain exactly what is to be included as ‘income’ (e.g. earned income and that from private pensions, but not benefit received or state pensions). The other argument is that postal surveys are a better way to get income information because respondents are more likely to be honest in the more anonymous setting of a self-completion questionnaire.

Our initial comparisons of the two survey methods show that average income in postal surveys is over £2,000 more than in interviews (£14,445 compared with £12,360). This difference is fairly consistent across all our surveys, and the reasons for the difference call for examination.

A2.3.2 Zero-income Households

There is evidence that interviews are a better medium for studying income, especially w...

Table of contents

• Cover
• Half Title
• Dedication
• Title Page
• Copyright Page
• Table of Contents
• List of Figures
• List of Tables
• List of Boxes
• List of Abbreviations
• Preface
• Part A Introduction
• Part B Defining Housing Need
• Part C Investigations Using Housing Needs Data
• Part D Policy Context of Housing Needs
• Part E Requirement for New Housing
• Part F Recommendations
• Appendix 1 Data Sources Used for Studies in This Book
• Appendix 2 Mode of Survey
• Appendix 3 Summary of Housing Needs Survey Types
• Appendix 4 Planning for Housing in South East England
• Appendix 5 Effect of Equity in Assessing Housing Need
• References
• Glossary