!"#"$"%!$&'#()"#*)($+%$,+)#-"-($'%#()(.#/ $$$.#"#($0)+1'.'+%$"%!$0)'1"#($'%.*)"%2( DATA AND LITERATURE ON MORTGAGE INTEREST: STATE PROVISION AND PRIVATE INSURANCE An Evaluation Report and Source Book
$$$$$$$$$$$$
In-house report 65
Janet Ford Jude England
962 8557
ACKNOWLEDGEMENTS
The review of literature and data on state provision and private insurance for mortgagors was commissioned by the Department of Social Security. We wish to acknowledge the contribution of Department of Social Security staff to this work and members of the academic community, government departments and industry bodies who provided details of their research. In particular we are grateful to Lucy Spencer, the lead policy customer within the Department of Social Security, and Rachel Trott the Liaison Officer for this project. Sue Easton and Roger Burrows from the Centre for Housing Policy and Julie Williams from the Social Policy Research Unit also contributed to the work, as did Lynne Lonsdale and Jane Allen who prepared the report. Thanks are due to all of them.
THE AUTHORS
Janet Ford is The Joseph Rowntree Professor of Housing Policy at the University of York and Director of the Centre for Housing Policy. Jude England is a research fellow in the Centre for Housing Policy.
Main Contents Page
Part 1 Evaluative Report.......................................................................................................................1
Part 2 Source Book .............................................................................................................................42
PART 1 EVALUATIVE REPORT
Contents
SUMMARY: AN EVALUATION OF THE DATA AND LITERATURE ON MORTGAGE INTEREST: STATE PROVISION AND PRIVATE INSURANCE ............................1 1. INTRODUCTION..........................................................................................................6 1.1 Scope of the report .............................................................................................6 1.2 Recent developments in the owner-occupier market .........................................7 TRENDS IN HOME OWNERSHIP: MORTGAGES, BORROWERS AND LENDERS .......................................................................................................................................9 2.1 Trends in mortgage lending ...............................................................................9 2.1.1 Transactions and lending ....................................................................10 2.1.2 Re-mortgaging......................................................................................10 2.2 Mortgage borrowers .........................................................................................11 2.2.1 Characteristics of borrowers ..............................................................11 2.2.2 Attitudes to home ownership and safety-net provision.........................11 2.2.3 Borrowers’ perceptions of labour market risk ....................................12 2.3 Arrears and possessions ...................................................................................13 2.3.1 Defining and measuring arrears .........................................................13 2.3.2 The incidence and patterns of arrears .................................................14 2.3.3 The characteristics of those in arrears and at risk of not paying .......14 2.3.4 The amount and duration of arrears ....................................................15 2.3.5 Access to resources; borrowers’ strategies to minimise arrears .........16 2.3.6 Lender forbearance ..............................................................................16 2.3.7 Trends in mortgage possession ............................................................17 2.3.8 Voluntary and compulsory mortgage possession.................................17 2.3.9 Ways of avoiding possession: mortgage rescue and in-work benefits .18 2.3.10 The destinations of borrowers giving possession.................................19 2.3.11 The public and private costs of arrears and possessions.....................19 2.4 Advice and information for borrowers who get into difficulty .......................20 THE USE AND EFFECTIVENESS OF PUBLIC AND PRIVATE SAFETY-NETS 21 3.1 The policy context ...........................................................................................21 3.1.1 The state safety-net for mortgagors – Income Support for Mortgage Interest .................................................................................................21 3.1.2 Private safety-net provision for mortgagors-Mortgage Payments Protection Insurance ...........................................................................21 3.2 Income Support Mortgage Interest ..................................................................22 3.2.1 The profile of ISMI recipients and ISMI payments ..............................22 3.2.2 Individuals’ knowledge of ISMI ...........................................................23 3.2.3 The ISMI ‘wait’ period.........................................................................23 3.2.4 The effectiveness of ISMI: ISMI shortfalls ...........................................25 3.3 Mortgage Payments Protection Insurance (MPPI) ...........................................26 3.3.1 Attitudes to private insurance ..............................................................26 3.3.2 Costs, terms and conditions of MPPI...................................................26 3.3.3 Take-up.................................................................................................26
2.
3.
3.3.4 Expectations and claims.......................................................................30 3.3.5 Other forms of insurance safety-net .....................................................31 3.3.6 ISMI and MPPI interface ....................................................................31 4. 5. 6. SAFETY-NETS AND WORK DISINCENTIVES......................................................32 INTERNATIONAL COMPARISONS ........................................................................34 ISSUES FOR RESEARCH..........................................................................................35 6.1.1 The risks to mortgagors .......................................................................35 6.1.2 Issues relating to mortgage arrears .....................................................35 6.1.3 Lender forbearance ..............................................................................36 6.1.4 MPPI ....................................................................................................36 6.1.5 ISMI .....................................................................................................37 6.1.6 Safety-nets and work disincentives.......................................................38 6.1.7 Comparative research ..........................................................................38 6.1.8 Summary and conclusions......................................................................38
TABLES
Table 2.1 Mortgage transactions and lending...........................................................................................10 Table 2.2 Mortgage arrears.......................................................................................................................13 Table 3.1 Recorded percentage take-up of MPPI from a range of sources 1995/6-1997/8......................27 Table 3.2 Take-up of MPPI measured via lenders and insurers...............................................................29
SUMMARY AN EVALUATION OF THE DATA AND LITERATURE ON MORTGAGE INTEREST: STATE PROVISION AND PRIVATE INSURANCE
1. A key objective of current housing policy is sustainable owner occupation. The issue of sustainability has come to the fore as a result of structural changes within the housing market, the labour market, the economy and the rising incidence of relationship breakdown. In a climate of greater risk, real or perceived, the presence and effectiveness of safety-net provision for home owners to provide support with mortgage payments is a significant policy issue. (1.2) The state safety-net was further restricted in October 1995. Assistance is available for post-October 1995 borrowers after nine months of an IS/JSA claim. Private insurance cover (MPPI), available since the 1970s, is currently taken by around 1.85 million mortgagors. (1.2) Mortgage borrowing and borrowers The requirement for safety-net provision is, in part, a function of the characteristics of mortgage lending and mortgagors. An important source of data on mortgage lending is the Survey of Mortgage Lenders (SML) which provides information on five per cent of all loans completed in a year. The survey covers the form of advance, forms of repayment, the extent of re-mortgaging and the incomes of borrowers. It is the source of data on important ratios such as loan: value and income: loan. The number of loans recorded does not equate with the number of housing market transactions as it excludes cash transactions and includes re-mortgaging. Data on the percentage of mortgagors who have ever re-mortgaged are available from the Family Resources Survey (FRS) and the Council of Mortgage Lenders (CML) Annual Housing Finance Survey. (2.1) The Survey of English Housing (SEH), the Family Resources Survey (FRS) and the British Household Panel Survey (BHPS) provide data on the socio-economic characteristics of mortgagors. Income data are more comprehensive in the FRS and the BHPS, whilst the SEH has considerable data on housing mobility. The BHPS and the Family Expenditure Survey (FES) provide substantial information about expenditure and financial commitments. (2.2.1) Information on attitudes to home ownership have been available from the British Social Attitudes Survey (BSAS) since 1986. Data on housing tenure preferences (in two and ten years’ time) are available from the CML Housing Finance Surveys. Both data sets indicate the fall in support for home ownership in the early and mid 1990s, with a recovery from 1996 onwards, although not yet to the levels of support seen in the late 1980s. (2.2.2) Attitudes to public and private safety-nets are not covered routinely by any major national survey. Ad hoc surveys have identified some equivocation about private
2.
3.
4.
5.
6.
1
sector provision, as well as a continuing belief that the state should provide, at least in particular circumstances. (2.2.2) 7. Surveys such as the BSAS have indicated a growing perception that the labour market, in general, is riskier than previously. More detailed ad hoc surveys indicate that such risks are explicitly considered but seen as inevitable. Ad hoc surveys have also indicated a poor match between actuarial assessment of risk and self-assessment. (2.2.3) Mortgage arrears The measurement of arrears is dependent on both the definition of arrears and the source of the data. One major source of data, administrative data from lenders, published twice yearly by the CML, measures those borrowers owing three or more months arrears. The SEH measures arrears as any outstanding missed payments and the data come from borrowers. Unlike the CML data, the SEH enables the data on arrears to be linked to the socio-economic characteristics of borrowers. Longitudinal data on the incidence and patterns of arrears are not routinely available but data from 1994 and 1995 indicated that over a period of time, the incidence of arrears is greater than at any one point in time. A proportion of borrowers move in and out of arrears repeatedly. (2.3.1) The SEH currently provides the most complete data set for the analysis of mortgage arrears, both the characteristics of those in arrears and the causes of arrears. Multivariate analysis of SEH data has identified the significant risk factors associated with arrears. Similarly, the SEH is the main source of data for the analysis of possessions, but the size of the sub-samples are such that the data from a number of years have to be combined for analysis. There is a series of studies focusing on particular ‘at risk’ groups such as the self-employed, younger borrowers and members of ethnic minority groups. (2.2.3) Current data on the amounts owed in arrears are limited and where available should be treated cautiously as they are often based on borrowers’ recall and estimation. More complete information was available in the mid 1990s, but is therefore somewhat out of date. (2.3.4) Surveys such as the SEH confirm ad hoc studies showing that mortgagors both in and out of work experience arrears. These sources also indicate the range of resources (kin, savings and safety-net provision etc.) that borrowers draw upon to try to prevent or recover arrears. Mortgagors in work lack any public or private safety-net provision. (2.3.5) Lender forbearance Studies of lender forbearance are primarily ad hoc studies and in the main date from the early and mid 1990s. More recent publications tend to review this older data, updating where possible to take account of recent policy initiatives. There is little systematic, contemporary data on the ways in which lenders now respond to those with payment difficulties. The limited information available suggests that lenders 2
8.
9.
10.
11.
12.
tightened their procedures following the changes to the ISMI provision in October 1995, for example removing the more generous treatment of those on ISMI, or making greater use of judicial proceedings (2.3.6) 13. Mortgage possessions The only significant source of data on current possession trends is the CML bi-annual administrative data. It does not distinguish between voluntary and compulsory possessions. The CML also provide time series data to 1980. (2.3.7); (2.3.8) Studies dating mainly from the early and mid 1990s identified potential alternatives to mortgage possession in the form of mortgage rescue schemes. These studies identified a series of legal, administrative and financial constraints facing lenders and housing providers. Studies also estimated the impact of the payment of mortgage interest ‘direct’. However, much of this work was based on assumptions about processes and preferences that proved to be unwarranted or overly simplistic. There are also a small number of studies examining policy options to support low income, in-work home owners. (2.3.9) Research on the public and private costs of mortgage arrears and possessions is limited. A wide range of potential costs for borrowers, lenders, central and local government have been identified but, with few exceptions, not addressed empirically. Ad hoc studies have highlighted the financial costs to borrowers, and their persistent nature and a number of social, emotional and health costs. Data from the Department of Social Security (DSS) monitors the payments associated with ISMI and the official homelessness statistics indicate something of the demands on the local authorities due to mortgage arrears. The costs to other public and voluntary sector organisations involved in responding to arrears and possessions are not known. (2.3.11) Information on the availability and use of advice and information on arrears and possessions is available from providers. This information reflects the nature of the clients they serve and the administrative categories adopted by the organisations. Ad hoc studies of borrowers in arrears also provide some data on those who sought information and advice and how they evaluated it. Very few studies in the 1990s have focused directly on these issues with the exception of a study of court based advice services and an evaluation of the development and functioning of mortgage arrears fora between lenders and local authorities. (2.4) The state safety net for mortgagors The Quarterly Statistical Enquiry (QSE) from the DSS provide the most comprehensive data on the number of IS/JSA claimants, some data on the characteristics of ISMI claimants and the numbers in the wait periods and so potential ISMI claimants. The data source contains more information than is routinely published and is a five per cent sample of all claimants. A small number of smaller ad hoc studies provide more detailed analysis of the socio-economic characteristics of IS/JSA claimants and ISMI claimants. (3.2.1)
14.
15.
16.
17.
3
18.
Both qualitative and quantitative studies indicate the overall poor level of understanding of ISMI. Ad hoc studies have shown how amongst borrowers who thought they knew the regulations, many were mistaken with respect to the costs that were covered, length of the wait periods etc. (3.2.2) Both quantitative and qualitative ad hoc studies from the early 1990s onwards have reported on the extent to which the wait period for ISMI contributes to arrears. Ad hoc studies indicate that MPPI is relatively ineffective in managing the ISMI wait period. (3.2.3) Quantitative data from ad hoc studies show the extent to which ISMI shortfalls contribute to arrears and that the proportion of IS/JSA claimants with shortfalls is increasing. Ad hoc studies have also indicated the central contribution of the application of the standard interest rate. There are only published qualitative data on the financial value of the shortfalls that borrowers have to accommodate, but data from the QSE could provide a more complete picture. (3.2.4) Private insurance for mortgagors Information on the costs, terms and conditions of MPPI either comes from ‘consumer’ research which examines and collates data from lenders or insurers, or from surveys of mortgagors. The FRS and FES ask about the costs of MPPI, but not the terms and conditions. Information on the terms and conditions of the policies taken are not available in any systematic, robust way from surveys or ad hoc studies. Some information is available from studies that look at MPPI claimants. (3.3.2) Estimates of the take-up of MPPI have typically come from surveys of lenders/insurers or borrowers. There are difficulties with both sources. Lenders tend to under-estimate take-up due to their lack of information on borrowers who access MPPI through an intermediary. Borrowers tend to over-estimate take-up sometimes mistaking other kinds of insurance for MPPI. Several surveys ask borrowers about the take-up of MPPI but in a number of cases the questions are problematic. Questions may not adequately screen out or distinguish insurance for death associated with non-endowment policies. They may lack clarity with respect to free insurance. Surveys such as the FRS and CML Annual Housing Finance Survey have sought to clarify some of these issues and so on the face of it provide more robust data. However, there remain discrepancies in the take-up rates reported. Ad hoc surveys have also asked about MPPI take-up. Despite the differences in the figures, all sources indicate a similar trend of steady but limited growth. An initiative in 1999 by the Association of British Insurers (ABI) and the Council of Mortgage Lenders (CML) has resulted in more robust, consolidated data on take-up that responds to some of the difficulties with the existing statistics. (3.3.3) Data on the number of MPPI claims and the percentage acceptance, from 1998 onwards, are available from the ABI/CML statistics, as well as from one or two ad hoc surveys which provide both qualitative and quantitative data. The response rate amongst rejected claimants in the ad.hoc quantitative surveys was much lower than estimates from the insurance industry indicated could be expected. (3.3.4) 4
19.
20.
21.
22.
23.
24.
There are very little data on the ISMI/MPPI interface, and what does exist is only qualitative. None of the routine surveys ask about this issue. (3.3.6) Safety-nets and work disincentives There is a very substantial literature on work disincentives, much of it econometric. Studies focusing on the behavioural responses of mortgagors to the actual or theoretical disincentive posed by ISMI and MPPI are much less frequent. With respect to MPPI, there is no current study on its work disincentive effect. Qualitative research with MPPI claimants has not explored the issue even in passing. Limited research, primarily of a qualitative nature on the impact of ISMI, indicates that mortgagors recognise the work disincentive effect but that their response can sometimes override this recognition. Therefore, ISMI can constitute a work disincentive, but does not always do so. A number of studies have pointed to the role of attitudes and commitment to work in shaping the response to ISMI. (Section 4) Comparative analyses There are no published, systematic, comparative analyses of most of the issues that pertain to safety-net provision. The exception is a consideration of state safety nets where a number of studies have looked at the support to home owners but often briefly. As yet unpublished comparative work is examining social assistance in 12 European countries, including a consideration of ISMI and comparable support for home owners which may be more detailed. A more typical response is for research with a strong policy orientation to examine one or two forms of comparable provision in other countries but not necessarily in detail. (Section 5) Issues for research A number of areas for further research are identified. They include: an assessment of the combined effect of a number of risks on mortgagors; identification of the contemporary incidence and patterning of arrears; analysis of lender forbearance with respect to safety-net provision; analysis of MPPI as a work disincentive, and; the impact of the Working Families Tax Credit on arrears and payment difficulties. (Section 6)
25.
26.
27.
5
1.
1.1
INTRODUCTION
Scope of the report
A key objective of current housing policy is sustainable owner occupation. The issue of sustainability has come to the fore as a result of structural changes within the housing market and other social and economic structures such as the labour market, and household formation and dissolution, that have brought about an increasing risk of mortgage arrears and possessions. In the current environment of change and real and/or perceived uncertainty, the presence and effectiveness of safety-nets for home owners to provide support with mortgage payments at points of significant disruption to income is a key policy issue. Sustainable home ownership is important for several reasons. It underpins the maintenance of a buoyant owner-occupier market and contributes to a successful economy (Maclennan, 1994; Maclennan et al. 1997). Owner occupation also remains the preferred tenure of the majority of households (BSAS, 1998) and housing policy promotes the sector as the main form of housing provision. Unsustainable home ownership results in additional demands on the public sector; at local authority level responsibility for re-housing homeless households, and for central government in terms of the costs associated with Housing Benefit. Further, there is evidence of the detrimental impact of problematic home ownership in terms of the physical and emotional health and financial well-being of households (Nettleton et al. 1999; Ford, 1994; Davis and Dhooge, 1993), all of which points towards support for a policy agenda concerned to prevent or minimise the impact of a range events that are often unforeseeable and unavoidable at the individual and household level. This report reviews the available literature and research relevant to such safety-net provision. It is concerned with both state provision and private insurance; Income Support for Mortgage Interest (ISMI) on the one hand and Mortgage Payments Protection Insurance (MPPI) on the other. This evaluation will assess both what is known about safety-net issues - broadly defined - and how robust such findings are. The report also briefly identifies key issues about which little is known, and potential future topics for research. It also seeks to place discussion of specific safety-net provisions in the context of the owner- occupier market. There are several substantive issues that require evaluation and the report is organised around those topics and key questions. In each case the concern is to evaluate the available literature and data on each of these areas. They are divided into research on the trends in home ownership; mortgages, borrowers and lenders, research directly focussed on the use and effectiveness of the public and private safety-nets, and a small number of consequential issues such as the impact of safety-net provision on attitudes and willingness to work. Each of these three broad themes forms a major section of the report. Within each section, the main relevant findings will be reported and assessed and gaps in knowledge and research identified.
6
This division into areas is a heuristic device. In practice, apparently separated issues are often closely inter-related. At the heart of discussions there is a central policy issue to which they all contribute. This concerns the extent of the support for the proposition that the private sector can insure risks such as unemployment and ill health, and with respect to unemployment, can insure the risk throughout the economic cycle, at costs that individuals and householders can afford and without jeopardy to other Government policies. The structure of this evaluative report is as follows: Section 2 examines recent trends in home ownership, Section 3 considers the available research on public and private safety-nets, Section 4 looks at safety-nets and work disincentives and Section 5 at the international information on safety-nets. Section 6 discusses a number of potential research questions that now need to be considered. The report starts with a brief contextual discussion of recent developments in the owner-occupier market. 1.2 Recent developments in the owner-occupier market The substantial growth in home ownership since the end of the 1970s has resulted in the tenure having a more diverse socio-economic structure than at any previous time. In particular, the 1980s saw the extension of owner occupation throughout the socio-economic scale, as a result of the policy of right-to- buy, financial de-regulation and a competitive mortgage market (Forrest and Murie, 1994). This expansion of home ownership took place over a period in which the owner-occupier market experienced some extreme fluctuations in house prices, in the costs of mortgage finance and in the number of transactions and exhibited more volatility than had been the case for many decades. In the early 1990s negative equity emerged. While this housing market has always been subject to some cyclical processes, the impact of cyclical unemployment became more marked in the 1980s in part because a higher percentage of homeowners were drawn from those groups traditionally most at risk of unemployment and, in part, because unemployment ceased to impact only on manual workers. More structural changes such as a rising incidence of relationship breakdown also posed some difficulties for home owners as has the growth of low paid work, self-employment and more casualised forms of employment. One of the consequences of the combined impact of all these changes can be seen in the overall pattern of mortgage arrears, which shows a rising but cyclical trend, and in the numbers reporting difficulties in meeting payments (Ford and Wilcox, 1998). Low-inflation has also impacted on the housing market, arguably shifting attitudes away from the perception of housing as a means of accumulation towards an emphasis on its current and on-going use. These changes also mean that the amount of equity outstanding on a property is not eroded by inflationary processes for a considerable period of time. Such changes also have implications for the extent to which borrowers who experience labour market disruption or relationship breakdown or ill health and housing problems can sell their way out of difficulties and trade down or out of the market. Other changes in the level of mobility are argued to reflect individuals’ perceptions (or experiences) of growing uncertainties within the labour market (Spencer, 1996; Bryson and MacKay, 1997) which in turn potentially increase their caution with respect to the amount of debt they are willing to incur and which have,
7
amongst other things, delayed the entry of younger households into owner-occupation. (Holmans, 1997). A state safety-net for mortgage interest (ISMI), payable to those in receipt of (currently) Income Support and Jobseekers’ Allowance, was introduced in 1948. A first series of changes were implemented in 1987 (see below for details) but notwithstanding these restrictions, the benefit continued to provide significant support to those eligible for Income Support and contributed to the avoidance of mortgage arrears. In October 1995, further restrictions on ISMI eligibility were introduced and there was a policy expectation that the private insurance market would become the primary source of safety-net provision through the further development of MPPI. On-going initiatives will potentially impact on sustainable home ownership. Supported by the Government, mortgage lenders and insurers are working to improve the cover offered by MPPI and the way it is sold. In October 1999, the Working Families Tax Credit was introduced, increasing support to low-income working households, including mortgagors, with clear scope to limit mortgage arrears. There is also a current debate about the regulation of mortgages and associated products.
8
2.
TRENDS IN HOME OWNERSHIP: MORTGAGES, BORROWERS AND LENDERS
A wide range of factors impact on the development of unsustainable home ownership and so are central to any discussion of the need for, and form of, safety-net provision. In particular, the numbers entering home ownership, their characteristics, their commitments and access to resources and the risks they face in the labour market all form a dynamic environment with the potential to increase or constrain the likely need for safety-net provision. Further, changes to the demands on the state safety-net, both the number of claimants and the costs, are affected by the extent of new borrowing and re-mortgaging, as well as by current interest rates and the prevailing rate of unemployment. The ways in which mortgage lenders view and respond to payment difficulties are also part of the context in which safety-nets operate. These economic and administrative issues interplay with the personal characteristics, attitudes and aptitudes of mortgagors to increase or decrease the likelihood of payment difficulties and recovery from them. The first section of this report therefore reviews and evaluates the available information on: • • • • the trends in mortgage lending; the characteristics of borrowers, their perceptions of labour market risks and their attitudes to safety-net provision; the extent and characteristics of those in mortgage arrears and giving possession, borrowers’ access to resources and lenders’ forbearance; and the advice and information available to borrowers who get into difficulty.
2.1 Trends in mortgage lending There are a number of major sources of information on trends in home ownership1. The key data source is the Survey of Mortgage Lenders (SML). The SML is a continuous survey based on a 5 per cent sample of all loans in the UK made by a representative sample of more than 50 mortgage lenders, including all but two of the largest mortgage lenders. This survey has replaced the previous 5 per cent sample of Building Society Mortgage Completions. It is administered by the Council of Mortgage Lenders. In recent years between 60,000 and 70,000 loans have been sampled annually. The size of the SML sample allows disaggregated analysis by countries within the UK, by English regions, by lending organisation and by key socio-economic variables, although the focus of differentiation in the reports of the SML tends to be between first and repeat buyers. Time series data are available although prior to 1994, the data were drawn from a more restricted group of lenders. Whilst other surveys do provide limited information on aspects of mortgage lending (FRS, SEH,) they are neither as comprehensive nor based on as large a sample as the SML. The SEH’s coverage is also restricted to England.
3
The major surveys considered are the Survey of English Housing (SEH), Family Resources Survey (FRS), Scottish House Condition Survey (SHCS), Scottish Household Survey (SHS), British Household Panel Survey (BHPS), Survey of Mortgage Lenders (SML), British Social Attitudes Survey (BSAS), Family Expenditure Survey (FES), General Household Survey (GHS), Council of Mortgage Lenders, Annual Housing Finance Survey (HFS) and the Council of Mortgage Lenders (CML) statistics on arrears and possessions.
9
2.1.1 Transactions and lending The SML provides data on the loan made: the amount; the type of advance (new mortgage , re-mortgage, further advance); its purpose; repayment period; gross rate of interest; form of interest (variable/fixed, deferred or discounted); repayment method; source of introduction. Information is also available on aspects of the property including its price, and post-code location, and on aspects of borrowers including their income at the point of application for a loan and the amount borrowed. Some key data is shown in Table 2.1.
Table 2.1 Mortgage transactions and lending
1991 No. of mortgagors No. of mortgages advanced No. of mortgages to first time buyers Average advance as% of selling price for first time buyers
Source: Survey of Mortgage Lenders
1995 10,521,000 909,000 419,000 89.0
1998 10,821,000 1,033,000 527,000 83.0
9,815,000 1,060,000 330,000 82.7
The data are used to calculate key indicators such as average advances to first-time and repeat buyers, average loan to value ratios, average loan to income ratios, average monthly repayments as a percentage of income and annual changes in any of these measures. Data on mortgage loans (even new loans) do not necessarily equate with the number of residential transactions as the data will exclude cash transactions. Thus the SML is not a source of data on all transactions although it can be used to estimate the number of transactions with a mortgage, a figure relevant to discussions of safety-net provision. More complete data are available from the Inland Revenue Statistics on residential property transactions although these are restricted to England and Wales. 2.1.2 Re-mortgaging Re-mortgaging occurs when borrowers remain within their current property but repay an existing mortgage and take another one, either with the same lender or with a different lender. The process is often motivated by access to a lower interest rate and/or to withdraw equity. In terms of safety-net provision, the process of re-mortgaging became significant in October 1995 when changes to the terms and conditions of ISMI defined re-mortgages as new mortgages. Borrowers re-mortgaging from October 1995 onwards were therefore subject to a nine month wait period before the receipt of ISMI. Information on the incidence of re-mortgaging (as opposed to additional or further loans) can be obtained from the SML, which records re-mortgages in any one year. The FRS and the smaller CML Annual Housing Finance Survey record all those who have ever re-mortgaged in their current property. There are slight discrepancies between these two data sources but the order of magnitude is the same. Both provide some limited additional information about the source of the re-mortgage (CML) or the motivation for the re-mortgage (FRS).
10
The data sources are consistent in identifying that the extent of re-mortgaging increased substantially after 1993 when around six per cent of households re-mortgaged (HFS). It increased steadily, reaching around 17 per cent in 1996 (reflecting the highly competitive mortgage market) before falling to 14 per cent in 1997/8. These data suggest that either the changes to ISMI are not widely recognised or understood or that those re-mortgaging are borrowers who believe that they are unlikely to be eligible for ISMI (see below). 2.2 Mortgage borrowers 2.2.1 Characteristics of borrowers All the major surveys provide details on the socio-economic characteristics of mortgagors. Broadly, they provide a comparable range of variables (age, household structure, economic status, ethnicity), although the large SML is more restricted, lacking data in the key areas of household structure, ethnicity, and employment. More detail on employment is collected in the FRS and BHPS than the SEH. There are also key differences with respect to income and expenditure data. Income data are more comprehensive in the FRS and the BHPS, than in the SEH and SML. The BHPS and the FES provide much greater detail on expenditure commitments (including credit and debt) than the other surveys. However, despite some limitations, the SEH offers some clear advantages in considering the characteristics of mortgagors. The sample of mortgagors is large (over 8,000) including data on both demographic and attitudinal issues. It has consistently examined housing mobility and so reported on the characteristics of recently moving home buyers as well as all mortgagors. 2.2.2 Attitudes to home ownership and safety-net provision Long-standing, consistent data on attitudes to home ownership come from the British Social Attitudes Survey (1983) and the CML Annual Housing Finance Survey. In both surveys individuals are the respondents. The CML uses questions on tenure preference in 2 and 10 years time to identify support for home ownership. The BSAS uses a standard sequence of questions to identify attitudes. The questions were asked in 1986, 1989, 1990/1/6/7/8. It asks respondents to imagine a newly married couple with stable jobs and whether they would advise them to buy without delay, to wait and then buy, or not to buy at all. This is followed by a further set of questions on different attributes of home ownership. Recently, these dimensions have been used to construct a composite scale of attitudes to home ownership (Ford and Burrows, 1999). Both the BSAS and the CML data on attitudes to home ownership show that the very positive attitudes of the late 1980s were less positive by 1996. Since then, attitudes have become more positive but the BSAS show they have not returned to the support seen in 1986, suggesting greater caution on the part of both existing and potential home owners. On the CML measure, attitudes have almost regained their mid 1980s strength. Information on attitudes to mortgage safety-net provision per se are not routinely collected in the large surveys. However, this issue was addressed as part of a review by the DSS of a number of existing and specially commissioned studies (Williams et al. 1999 Section 8.4). More qualitative research has also considered this question (Ford and Kempson, 1997; Munro, et al. 1997) and it is a central issue of on-going work by Walker et al. This latter study is examining the ways in which families confront the risk of unemployment, including their current and prospective willingness to take income protection insurance or MPPI. These 11
studies all show continuing support for a state safety-net for mortgagors along with some recognition that individuals must try to take increasing responsibility for their own social protection where they can. Ford and Kempson (1997) and Walker et al. (on-going) have also provided tentative evidence of ‘pre-dispositions’ towards a willingness to insure which influences attitudes to private insurance in general and MPPI in particular. There is evidence of some equivocal attitudes to the private MPPI market rather than unquestioning support from all these studies. Attitudes to ISMI and MPPI are, however, one aspect of the much larger issue of attitudes to state and private provision and the re-structuring of welfare in general. Here the British Social Attitudes Survey is a source of information and based on that data and other sources there is an emerging, broadly based literature that encompasses health and long-term care provision and pension privatisation. These studies are not reviewed here. 2.2.3 Borrowers’ perceptions of labour market risk After a period of equivocation, there is a now growing consensus that the labour market is a riskier environment than it was, and that risk can be measured in a number of different ways (see Dex and McCulloch, 1997 or Burchell et al. 1999 for overviews). In terms of safety-net provision, a more important issue is the extent to which mortgagors perceive that they are at risk in the labour market, how accurate their perceptions are, and whether this leads to a reconsideration of their safety-net provision, especially MPPI. There are studies that consider perceptions of labour market risk, and the British Social Attitudes Survey has been a source of much of the data used (Spencer, 1996; Bryson and Mackay, 1997). Mortgagors have occasionally been a specific focus of interest (Doling and Ford, 1991). The questions used have, however, typically asked about perceptions of labour market risk in general rather than about individuals’ perception of the risk to their own jobs. The latter questions were asked in the 1997 BSAS for the first time, but have not been repeated. Since 1997, the CML Annual Housing Finance Survey has asked a range of questions about the incidence of a range of risks including unemployment. There have been a number of ad hoc surveys. For example, Ford and Kempson (1997) asked 1000 mortgagors about a range of experienced and perceived labour market risks including how secure their jobs were over the coming 12 months. However, this was not a random sample survey, although the characteristics of respondents were representative of all mortgagors in England. Munro (1997) asked 800 owner-occupiers in two locations about labour market risks and concluded that the risks were recognised and explicitly considered but viewed as inevitable. On-going research by Walker et al. is also examining perceptions of risk and their ‘accuracy’ and includes the use of a random sample Omnibus survey of 1000 people in employment and questions from the BHPS. Notwithstanding different methodologies, these studies tend to concur that perception of risk in general is widespread, but that perception of, and concern about, self-risk is contingent on a range of factors and only poorly related to any actuarial calculation.
12
2.3 Arrears and possessions 2.3.1 Defining and measuring arrears During the 1990s, the availability and quality of data on mortgage arrears increased significantly. There are three routine sources of annual data on mortgage arrears at a national level. These are the CML administrative statistics, the Survey of English Housing and between 1985-1998, an annual survey of arrears undertaken by Ford. In addition, a number of ad hoc surveys have provided data on the percentage of mortgagors in arrears in particular places at particular times (for example, the studies of Luton, Bristol and Glasgow reported in Maclennan, 1994). These surveys and studies differ in terms of the source of the data on arrears, the definition of an account in arrears, and whether the data are cross sectional or provide some evidence on incidence over a period of time. Information can be drawn either from borrowers or lenders. The information from the CML and Ford is based on administrative returns from a sample of lenders. The data are cross sectional and indicate the number of mortgagors with arrears at six monthly intervals or annually. The figures are grossed up to provide a national estimate at a point in time. The CML restricts its definition of arrears to those of three and more months while Ford included data on arrears of two to three months. Where these two sources use the same categories of measurement the results are broadly comparable, although they are based on slightly different samples of lenders. Both data sources are UK based and provide time series data.
Table 2.2 Mortgage arrears
Arrears of No. of mortgages 000's 1991 1995 1998 9,815,000 10,521,000 10,821,000 3-6 months 6-12 months 12+ months 91,740 85,200 34,880 (0.9%) (0.8%) (0.3%)
not published 177,910 129,090 (1.7%) (1.2%)
183,610 126,670 74,040
(1.9%) (1.2%) (0.7%)
Source: Council of Mortgage Lenders Percentage figures have been rounded to one decimal point
While the number of lenders contributing to the CML survey has grown since the early 1990s, lenders are not systematically sampled and the survey excludes smaller, often local, lenders. There is no discussion of the possible impact of this, although with time the issue is of diminishing importance as many smaller lenders are merged with or taken over by larger lenders. Other limitations of the data are the inability to disaggregate by, for example, lender type or region. In contrast, the Survey of English Housing and the majority of the ad hoc surveys obtain data on the number in arrears from mortgagors. There is a series of recognised limitations associated with this approach. For example, borrowers may be unwilling to admit to arrears, may not know about any arrears, underestimate them, or believe that payments from DSS 13
mean they are no longer in arrears or that an arrangement to repay their lender means they are not in arrears. As a result, the proportion of borrowers in arrears (however defined) is lower than that reported by lenders. A number of studies have recognised these potential problems and attempted to address them. The SEH in particular asks people if, for example, they are receiving payments from the DSS, whether they can make up a shortfall on interest or meet capital payments due and whether they are making reduced payments to their lender (SEH 1993/4-1997/8, but see 1997/8 for a full discussion). As a result of adding questions, and adjusting for potential under-reporting, the SEH indicates considerable agreement between the data obtained from borrowers and lenders. However, whereas the development of additional questions provides robust data, the adjustment for under-reporting is not empirically grounded and needs to be treated with caution. Notwithstanding the last comment, the SEH is an increasingly robust source of data on the number of borrowers with mortgage arrears. Additional advantages are that the data can be analysed in relation to the socio-economic characteristics of borrowers, something not possible with the data from Ford (1985-1998) or the CML. The SEH also captures the range of ways in which arrears can arise, again something currently not available to the same extent through other data sources. 2.3.2 The incidence and patterns of arrears Longitudinal data on the ‘careers’ of borrowers in arrears are less well developed. A study by Ford et al. (1995 p.18) included data from a random sample of over 600 mortgagors and showed that over the period 1990-1994, one in five of all home buyers had arrears. They identified whether mortgagors had defaulted on one or more occasion, and their current payment position. Such information is pertinent to understanding whether there are a group of people who move in and out of arrears relatively frequently, who are in some sense at ‘permanent’ risk, or whether arrears are typically a ‘one-off’ experience. The data from Ford et al. are, however, now old. Since the early 1990s, both the CML and SEH have improved the data on the proportion of borrowers in arrears, now providing a more complete picture and a convergence of results. The SEH has the advantage of allowing a range of additional analyses to be undertaken. However, the SEH is a relatively new data source. CML provides the only time series data but it remains limited to arrears of three and more months (and to six and more months prior to 1993) and draws only on the experience of the largest lenders. Other surveys do provide data on the numbers who have mortgage arrears (CML Housing Finance Survey, BHPS) but the numbers are small and/or the range of questions more limited than found in the SEH. Longitudinal data are valuable but less frequently available and the BHPS is likely to offer the best basis for such analyses but with limited numbers of respondents and restricted questions. 2.3.3 The characteristics of those in arrears and at risk of not paying Prior to the mid 1990s, research on the characteristics of those with mortgage arrears relied primarily on bi-variate analysis. One exception was a study in the mid 1980s (Doling and Stafford, 1986), but this was constrained by the limited nature of the data on the characteristics of those in mortgage arrears available at that time. A number of sample surveys in the 1990s (Ford and Wilcox, 1992; Ford et al. 1995) showed both how arrears were 14
concentrated amongst particular groups and the way the causes of arrears (and so the characteristics of those in arrears) changed over time. One example was the presence of a higher percentage of white collar, professional households in arrears in 1995 compared to 1992, reflecting the extent to which the recession of the early and mid 1990s impacted on white collar workers. Only in the case of Ford et al. (1995) were national data used. Analysis of the characteristics of those with mortgage arrears has become more robust. This is largely due to the availability of random sample survey data via the Survey of English Housing and the work of the authors of the SEH reports (Green et al. 1994-97) and Burrows (1997, 1998), and the use of logistic regression analysis. This enables the significance of each variable identified at the bi-variate level to be assessed after allowing for the effect of all other variables. As a result, some of the characteristics of mortgagors identified in bi-variate analysis were seen to be spurious. Logistic regression is also used to indicate the odds of those with particular characteristics developing arrears, vis-à-vis a reference group (for example, Burrows, 1997, 1998). Comparable analyses have been undertaken for borrowers finding it difficult to pay but not yet in arrears, and for mortgagors giving possession of their property (Nettleton et al. 1999). A series of analyses have focussed on particular ‘at risk’ groups; the self-employed (Burrows and Ford, 1998), borrowers in ‘non-standard’ forms of employment (Ford, 1998), young people (Ford, 1999), and ethnic minority groups (Green et al. 1999), all based, at least in part, on SEH data. Earlier work of a more qualitative nature focussed on the link between casual work and arrears (Ford, 1989) and arrears and relationship breakdown (Bull, 1995). On-going work by Smith and Munro (supported by the Economic and Social Research Council) is considering a number of dimensions of the relationship between health and housing, including arrears. Substantial insights into the processes whereby arrears develop and either persist or are cleared amongst particular groups result where qualitative and quantitative data are used together (for example, Burrows and Ford’s [1998] consideration of arrears amongst selfemployed heads of household or the Kempson et al. [1999] examination of the combination of arrears and safety-net provision). 2.3.4 The amount and duration of arrears Information on the amount of arrears owed is limited. None of the large national surveys collect this information directly. Some approximations might be cautiously derived where surveys report the current monthly payment and the number of months’ payments owed. Data on both the amount of arrears and the duration of arrears were collected by the random sample survey of borrowers in arrears and possession undertaken by Ford et al. (1995), but should be treated cautiously because information from borrowers is often subject to recall and error. There are difficulties in knowing how systematically borrowers have included, or can include, a whole range of payments (lenders’ administration costs, court costs etc) that increase arrears. Figures are not in any case an on-going guide to the amounts owed in arrears (even for the same number of months in arrears) as interest rates, and lenders’ decisions about income: loan multiples can change quickly. Some qualitative studies also indicate the amounts owed by individual households but they are subject to the same limitations unless detailed documentation is available. Figures on the length of time people have been in arrears (as opposed to how many months’ payments they owe) are similarly highly specific to
15
particular times. There is however consistent evidence that for any number of months owed, the period over which arrears persist is longer. Since 1993, the Council of Mortgage Lenders has provided a ‘percentage of balance’ analysis of mortgage arrears. This shows the number of mortgages with 2.5% to <5% (and further increments up to more than 10%) of the mortgage balance in arrears, and provides one overall measure of the ‘seriousness’ of mortgage default. These figures, though, are not related to household financial data and they are rarely quoted or used. 2.3.5 Access to resources; borrowers’ strategies to minimise arrears The research evidence indicates that households both in and out of work can face difficulties paying their mortgages (SEH, 1993-1998). The evidence also indicates that arrears are not an inevitable consequence of a loss of income or low income (SEH, 1997/98). Findings reported by a number of more ad hoc surveys (Ford and Wilcox, 1992; Maclennan, 1994) and by qualitative studies (Ford, 1989, Davis and Dhooge, 1993) are consistent with those from the SEH. For unemployed households, surveys and other studies show that payment may be made from one or more of a wide range of resources, including the public-private safety-nets. The evidence shows consistently that such ‘benefits’ may prevent arrears but do not necessarily do so. For example, Ford et al. (1995) and Kempson et al. (1999) indicated that amongst those in receipt of ISMI or MPPI, almost a half and one in three respectively developed arrears. Unemployed households without access to a safety-net, with only partial access or in the wait periods before safety-nets kick in, also use a range of means to pay their mortgages including: redundancy payments, savings, financial help from relatives and friends and further borrowing. Use of these resources sometimes enables them to avoid mortgage arrears but not always. Other strategies include cutbacks in domestic and personal expenditure or the payment of the mortgage at the expense of paying other commitments. Those in work can also have problems paying their mortgages and may develop arrears (Ford,1989; SEH, 1993/5-1997/8). Studies show that this group draw on some of the same range of resources as unemployed mortgagors to try to prevent arrears or to minimise their size. A key difference, however, is that mortgagors in work lack access to any support with housing costs. There is some evidence that mortgagors treat their mortgage as the central commitment and seek to pay this before other commitments (Berthoud and Kempson, 1992). However, other studies suggest that payment patterns are more contingent (and hence variable), responsive to a range of influences including external factors such as the actions of creditors, the attitudes and commitment of individuals to home ownership and the extent to which they understand the consequences of not paying (Ford, 1989; Davis and Dhooge, 1993). 2.3.6 Lender forbearance Studies of lender forbearance have provided consistent findings. However, in most instances they were undertaken during the early and mid 1990s (Ford, 1989; Jenkinson, 1992; Ford and Wilcox, 1992; Ford 1994; Ford et al. 1995). More recent publications (Wilcox and Williams, 1996; Wilcox and Sutherland, 1997) have revisited previous material and up-dated where 16
possible using industry documents or codes of good practice, but the data are not based on new empirical research. Thus, there is relatively little systematic, contemporary data of the ways in which lenders now respond to borrowers who have payment difficulties or arrears, and whether they have widened their use of forbearance strategies from their previously rather narrow approach. In particular, little is known of the way lenders respond to those who are in the wait periods for either ISMI or MPPI. Oldman and Kemp (1996), Ford and Kempson (1997) and Kempson et al. (1999) provide some limited data from small scale surveys of lenders and qualitative interviews with claimants. These data indicate that lenders have tightened their procedures because of the cuts to ISMI. However, as the data are limited, they are not an adequate basis from which to draw other than tentative conclusions about the extent and form of a stricter forbearance regime. Given the changes in safety-net provision, these questions are key ones. Lender forbearance through the judicial system was also a focus of interest in the early 1990s (following much earlier interest in the early 1980s in Karn’s work for example). Studies by Ford (1994 ch.2) and Ford et al. (1995 ch. 4) provided qualitative data from borrowers on their experiences of the court process and data from interviews with county court judges. In addition, as part of the study, a postal survey and interviews with mortgage lenders collected the lenders’ perspective on the use of the judicial system. The key findings from these studies indicated the use by lenders of the judicial system to ‘underpin’ informal forbearance agreements; the differences in the strategies adopted by lenders in their use of judicial proceedings (applications for court orders made at different times, at different levels of arrears etc); and the shift amongst judges to greater reliance on suspended orders in the early 1990s. These findings were confirmed in a study undertaken by Nixon et al. (1996) concerned primarily with the provisions and role of advice in the court process (and discussed under the section on advice and information). 2.3.7 Trends in mortgage possession The only significant historical and contemporary source of data on the number of households taken into mortgage possession through the court or voluntary process is the CML bi-annual administrative data. As with the statistics on arrears, these figures are returned by lenders and grossed up to provide a UK incidence of possession during the preceding six month period. Data have been available since 1980. The data are not always disaggregated into voluntary and compulsory possessions, but intermittently some estimate, based on lenders information, is provided. Voluntary possession is defined administratively as possession without a court order. Compulsory possession is via a court order. Data collected by the SEH identify any member of a household who has lost their property through the possession process at any point in the past. These data are therefore not comparable with the CML data, but the numbers identified are of an adequate size to allow analysis of the reasons for possession as well as analysis of the re-housing outcomes of households in possession (Nettleton et al., 1999). 2.3.8 Voluntary and compulsory mortgage possession The nature of the statistics on possession has been reviewed already. Possession can be either voluntary or compulsory, and each can in turn take a number of forms. The SEH, Dodds and 17
Hunter (1990), Ford (1993), Ford et al.(1995) and Nettleton et al. (1999) note that forms of possession include: borrower initiated sale to avoid court action; abandonment; an outright possession order with the borrower leaving before eviction; or physical eviction. Ford (1993) argued that the statistics on possession did not necessarily capture all these forms of leaving the property and so potentially under-estimated the extent of exit from owner occupation because of payment difficulties. The question of mobility out of the tenure based on a ‘feared’ inability to pay has not received detailed, systematic attention since 1990 (Dodds and Hunter). 2.3.9 Ways of avoiding possession: mortgage rescue and in-work benefits Research addressing the ways in which possession might be avoided has emphasised the importance of lenders’ choices about forbearance and the timing of its implementation, both of which are shown to vary. As a result borrowers with the same level of arrears have different risks of possession. Research has also considered a range of current or potential policy initiatives that might mitigate against possession: the provision of state and private safety-nets (Ford and Kempson 1997; Kempson, et al., 1999); mortgage rescue (Ford and Wilcox, 1992; Foster, 1992; HACAS, 1995); and, in-work housing benefits for mortgagors (Webb and Wilcox, 1991; Wilcox and Williams, 1996; Wilcox and Sutherland, 1997; Holmans, 1997). The research on public and private safety nets is outlined and evaluated in a later section. Anticipating that discussion, while it is clear that safety-nets make a contribution to avoiding possession, they do not necessarily always prevent it because they do not necessarily prevent arrears. Mortgage rescue (defined as schemes that allow borrowers to remain in their homes but as tenants or shared owners with access to Housing Benefit) gained some prominence in 1991/2 following fears that possessions might exceed 100,000. Research focussed on one or more of three main issues: the constraints and costs involved in a number of proposed schemes, from both borrower and provider perspectives (Foster, 1992; Ford and Wilcox, 1992); modelling and estimating the likely demand for, and potential of, different mortgage rescue schemes (Foster, 1992; Ford and Wilcox, 1992; Ford et al., 1995 ch. 4); and the development of more effective financial and administrative models for mortgage rescue schemes (HACAS, 1995). The modelling and estimation work on the demand for, and potential of, mortgage rescue was often based on assumptions about processes and preferences that proved unwarranted and simplistic. For example, calculations of the likely impact on arrears of the implementation of direct payment of mortgage interest were based on the unwarranted assumption that prior to direct payment borrowers were failing to pass on the benefit to their lenders. There was also an absence of a recognition that ISMI shortfalls could be substantial. Many of these assumptions were challenged by the Ford and Wilcox study in 1992, but here too the modelling was constrained by insufficient data. Overall, the initial research interest in mortgage rescue in 1991/2 has not been sustained at any significant level. Searches have identified only one recent study by an MA student which evaluated the mortgage rescue scheme operated by a large housing association. However, the issues addressed by mortgage rescue schemes – the necessity and costs of re-housing 18
households in a different tenure following possession – remain pertinent, as in 1998 there were still 32,000 households who gave possession. Little is known about the views of those who experienced such schemes or the ways in which, in a different housing market and financial climate, mortgage rescue might be more feasible and more acceptable. There is little work that considers the place of such schemes (if any) in the context of other policy initiatives to support home owners. Finally, there are a number of studies that have developed policy proposals (often at a level of costed detail) to prevent arrears and so possessions amongst those in work by arguing for an in-work housing benefit for low income mortgagors (Webb and Wilcox, 1991; Sutherland and Wilcox, 1997; Holmans, 1997; Wilcox, 1999). Most proposals are some variant on the first proposal for a mortgage benefit outlined in Webb and Wilcox, but taking on board the changing tax structures and fiscal policies of successive Governments. 2.3.10 The destinations of borrowers giving possession Data on the re-housing of households taken into possession are available from borrowers and from administrative sources. For many years local authority statistics on the rehousing of homeless people have been available. However, the figures relating to rehousing as a result of mortgage arrears are likely to understate the extent of re-housing as households in this position may first go to family and friends and present themselves to be re-housed when they can no longer stay there. There is no way of identifying which of those re-housed under this category have given possession of their property. Acceptance as homeless following possession can also vary by local authority area depending on the interpretation of intentional and unintentional homelessness (see Ford, 1997 for a discussion of the statistics). Ford et al. (1995 ch. 6) identified the range of housing destinations for 111 households in possession between 1990 and 1994, indicating both initial and subsequent destinations. The degree of housing mobility noted was substantial. The SEH offers the basis for examining rehousing in a similar way for a larger pool of cases. Qualitative research also provides analysis of the process of re-housing (for example, Ford, 1994; Nettleton et al.1999) indicating a range of problems that include unsatisfactory housing, limited information, the absence of choice and delays in decisions about likely acceptance as homeless but cannot indicate the national distribution of these problems. 2.3.11 The public and private costs of arrears and possessions Research on the public and private costs resulting from mortgage arrears and possession is limited and piecemeal. With respect to the private costs to individuals and households, many studies indicate the emotional and/or financial costs that are associated with the experience of the arrears and possession process (Ford, 1998; 1993, Davis and Dhooge, 1993; Ford et al. 1995). However, studies whose primary focus is on the costs are infrequent. Nettleton et al. (1999) is one exception where the research was designed to explore the social and emotional costs to adults and children rehoused by social housing organisations following possession. This qualitative research with 30 families, including 17 children, indicated high levels of stress and shame, a heightened incidence of physical and mental ill health and the continuing impact of possession on their lives. The health related impact of arrears and possessions was also explored by Nettleton and Burrows (1998), using data from the BHPS. The findings are consistent with those from the qualitative studies. Ford and Burrows (1999) have also 19
developed a schematic framework identifying the range of costs associated with arrears and possessions for borrowers, lenders and a range of other parties and exploring in detail the financial costs to borrowers. Overall, however, this research has only begun consideration of a set of issues that require much greater attention, and where the costs to other parties have been particularly neglected (see below). Studies examining the public costs of possessions have tended to be restricted to a consideration of the re-housing costs either to local authorities in terms of additional demand for temporary and permanent accommodation from those made homeless or to central government in the form of additional claims for Housing Benefit. There are also discussions of the costs to central government associated with arrears and the payment of ISMI. However, research has focussed on these issues at particular times and in relation to particular issues (for example, research on mortgage rescue, or the cost-effectiveness of a mortgage benefit), rather than as an on-going concern with a comprehensive identification and measurement of costs. Further, the public costs to non-housing organisations such as, for example, primary care providers, have not been explored, although there is evidence as noted above that arrears and possessions have a health dimension. Similarly, there has been little or no emphasis on the costs to advice agencies of arrears and possessions per se. 2.4 Advice and information for borrowers who get into difficulty Data on the advice and information for borrowers in difficulty is available from three sources. First, from advice agencies such as NACAB or Shelter who report on their case loads. These reports often highlight emerging issues (for example NACAB, 1995) and the demands on particular agencies for advice but they are, inevitably, a partial account, given that seeking advice is both voluntary and affected by access to advice agencies. The data are also processed and reported under administrative categories developed to serve the organisation’s purpose. An on-going evaluation of advice provision and delivery by Goodlad et al., funded by The Joseph Rowntree Foundation, may provide a more systematic account of the availability and quality of advice to mortgagors but the study is geographically limited although it does encompass a range of providers and forms of delivery. Second, both quantitative and qualitative studies of borrowers in arrears and possessions provide information on whether or not people sought advice, who from and how they assessed it (Ford et al. 1995; Nettleton et al. 1999). These sometimes include detailed case studies, but neither of the latter studies were designed to focus on advice issues as a central concern. Only Ford et al. (1995) had a large, national (England) sample. Third, there are a number of studies focussing directly on the provision and nature of advice and information to borrowers. Nixon et al. (1996) examined the provision of advice in the county courts in England using a postal survey of all courts and an examination of records, procedures and outcomes in a smaller number of courts. This research concluded that the availability of advice had a positive impact on the outcome of the court case for borrowers, but these findings are supported qualitatively rather than by statistical procedures. Other studies (for example, Evans, 1998), have looked at the development of structures and procedures whereby lenders and local authorities can work together more effectively.
20
3.
THE USE AND EFFECTIVENESS OF PUBLIC AND PRIVATE SAFETY-NETS
3.1 The policy context 3.1.1 The state safety-net for mortgagors – Income Support for Mortgage Interest Prior to 1987, mortgagors in receipt of subsistence benefits were able to have their full mortgage interest payments met, subject to a number of eligibility criteria. In 1987, legislation was passed that restricted payment during the first two months of any claim to 50 per cent of eligible interest and thereafter full eligible interest. Eligible interest was restricted in several ways (see Ford and Griffith, 1994 for a fuller discussion). The November 1994 budget announced that from October 1995 there would be further restrictions to ISMI. State support would be provided as follows for borrowers with loans taken out prior to October 1995 and accepted as an Income Support or Job Seekers Allowance claimant: • • • no support for eight weeks; up to 50 per cent of eligible interest for the following 18 weeks; full eligible interest thereafter.
For borrowers taking mortgages after October 1995 (including re-mortgages) there was: • • no support for the first 39 weeks of a claim; full eligible interest thereafter.
Mortgagors aged 60 and over are, however, exempt from these waiting periods, whilst lone parents claiming IS following the death of a partner or relationship breakdown, and carers claiming IS are exempt from the longer waiting period. Other changes included: a £100,000 ceiling on eligible interest; interest payments subjected to a ‘standard’ rate of interest; and claims treated as ‘linked’ and so not requiring a nine month re-qualification period, provided the second ISMI claim occurred within 12 weeks of the termination of the first claim. These changes to ISMI, introduced in October 1995, were accompanied by a clear shift in policy towards safety-net provision whereby mortgagors were expected to provide cover against unemployment, sickness and accident in the private insurance market, at least in the first instance. This policy was confirmed again in 1997 with the publication of A New Contract for Welfare (DSS, 1997). Further influences were the growing costs of ISMI and the belief that the continuation of ISMI was impeding the development of MPPI. 3.1.2 Private safety-net provision for mortgagors-Mortgage Payments Protection Insurance The market for private mortgage payments protection insurance dates back to the 1970s, although take-up in 1995 had not reached 20 per cent. Policies typically offer cover for mortgage and related costs (endowments premiums, MPPI premiums etc.), and can cover both the main breadwinner and a second earner for one or more of accident, sickness and 21
unemployment. Borrowers insure for a particular sum of money (usually the monthly mortgage costs at the point of taking the policy). Policies typically pay out for up to 12 months following a two month deferral period. MPPI premiums have to be paid during the claiming process. Following a claim there is usually a re-qualifying period of 12 months. The majority of MPPI policies are ‘block’ polices whereby all mortgagors with a particular lender are offered the same terms and conditions at the same price per £100 insured. Indeed, the lender or mortgage broker typically sells MPPI on behalf of the insurer. The direct market for MPPI is currently very small, but is predicted to grow. One implication of block policies is that the underwriting takes place at the point of a claim. During 1998, following some reservations about MPPI policies, the Council of Mortgage Lenders (CML) and Association of British Insures (ABI) supported by both DETR and DSS, launched a benchmark product. This was designed to set a minimum standard for MPPI, identifying for example conditions on which temporary contract workers could be covered, limiting the number of times per year premiums could be adjusted and making transparent the rules linking claims etc. There are no data available as yet on the impact and effectiveness of this initiative. Since 1995, a number of studies have evaluated aspects of the changing pattern of safety-net provision for mortgagors. Key questions that policy makers and researchers have addressed include: to what extent has the take-up of MPPI been developed and have those at risk been most willing to insure? Is there evidence that the continuation of ISMI ‘crowds out’ the development of MPPI? What has been the impact of the standard rate of interest for those in receipt of ISMI? Is MPPI effective in enabling mortgagors to avoid arrears? Equally, what has been the impact of the restrictions to ISMI? 3.2 Income Support Mortgage Interest 3.2.1 The profile of ISMI recipients and ISMI payments The Department of Social Security (DSS) through the Quarterly Statistical Enquiry (QSE) provides the most comprehensive source of information on the numbers and characteristics of those receiving ISMI. Routinely, the number of claimants, the circumstances of the claimant (unemployed, lone parent etc.), the average amounts received, the range of payments, and the duration of ISMI receipt are available. Based on the QSE, the number of borrowers in receipt of ISMI has fallen since a peak of 555,000 in 1993 to 317,000 in February 1999. The overall costs of ISMI also peaked in 1993. The amount of the weekly benefit to claimants is affected by the level of interest rates and the eligibility criteria applied. The data show the decline in the proportion of claimants due to unemployment since 1995, consequent on falling unemployment per se and the likely increase in the proportion of mortgagors on JSA who return to work before they are eligible for ISMI. In February 1999, only 18 per cent of ISMI claimants were on JSA. These issues are discussed further in the section on work disincentives. The DSS Quarterly Statistical Enquiry collects a wider range of variables on mortgagors and ISMI than are routinely reported. Potentially, the QSE could be used to examine a wider range of issues, in particular, the factors that contribute to ISMI shortfalls. Information on the percentage of ISMI claimants is also available from the FRS and SEH but sample sizes are smaller and comparisons of the FRS and QSE suggests that the FRS under reports IS/JSA and so potentially ISMI respondents. 22
Other studies also provide information on the characteristics of ISMI claimants and ISMI payments. Most recently, quantitative data are available from Kempson et al. (1999) following a survey of 1,684 IS/JSA recipients and ISMI claimants drawn from the February 1998 DSS quarterly statistics. The survey was specifically concerned with the effectiveness of both public and private safety-nets. This survey provided data on the different claimants by a wide range of socio-economic variables including age, household type, whether the head of household had a partner, date of mortgage, monthly mortgage payments, savings and whether the claimant had an MPPI policy. 3.2.2 Individuals’ knowledge of ISMI Both qualitative and quantitative research consistently indicates the overall poor level of understanding about ISMI. Qualitative research shows that individuals and households vary with respect to their knowledge of ISMI (Kemp et al., 1994; Oldman and Kemp 1996; Ford et al., 1996; Ford and Kempson, 1997; Walker et al. (on-going research). All these studies report a sizeable proportion of borrowers with very limited and sometimes no knowledge of ISMI. Some researchers have drawn a distinction between mortgagors’ ‘awareness’ of a benefit that assists with housing costs and the ‘accuracy’ of people’s knowledge about what the benefit covers (Ford et al. 1996). Quantitative data on this issue comes from a postal survey of 2,000 mortgagors (Ford and Kempson, 1997). While the survey had a relatively low response rate of 48 per cent, and the sample was drawn from only three large mortgage lenders, the characteristics of respondents were checked against the characteristics of mortgagors in the SEH and shown to be representative of all mortgagors. Borrowers had a very poor knowledge of ISMI. Only one in five knew accurately what help they could expect to receive. A third of respondents thought they knew but were wrong. Recent experience of a claim did not appear to improve the level of knowledge. The results confirm the more qualitative research noted above. Qualitative interviews also indicated the areas where knowledge was lacking or inaccurate. These included what was covered by ISMI, the length of the wait period and the length of time that payment would continue. Interviews with lenders conducted as part of the Ford and Kempson study indicated that a majority of lenders perceived that borrowers were not very aware of the ISMI provisions. 3.2.3 The ISMI ‘wait’ period Those on IS or JSA currently become eligible for ISMI after a ‘wait’ period of either 8 weeks or 39 weeks, depending on when they took their mortgage. The DSS Quarterly Statistical Enquiry can identify the balance of pre and post-October 1995 borrowers and so the balance of wait periods likely to be experienced. In February 1998, more than three-quarters of then IS/JSA and ISMI claims were, or had been, subject to the shorter wait period. From both the borrower and lender perspective, the ability to manage the wait period is important and could potentially undermine the impact of ISMI if arrears reached an ‘unmanageable’ level. For this reason, research pertaining to the wait period is discussed first below and then research focussing on the implementation of ISMI. 23
Prior to the changes to ISMI in October 1995, research had noted that the 50 per cent payments for the first two months could prove problematic, but had not considered the wait period in detail. The first published study on the impact of both the shorter and longer wait periods is Kempson et al. (1999). The evidence indicates that as borrowers move through the wait period the risk of not paying increases. However, arrears only characterise a minority of claimants. The majority managed to pay during the wait period although often with difficulty. The response rate in this survey was 41 per cent for IS/JSA claimants and 50 per cent for those receiving ISMI and, ideally, higher response rates are needed for this picture to be confirmed. These figures may also be an under-estimate of arrears as some of these IS/JSA claimants were only part-way through their wait period. Amongst those now receiving ISMI, a higher proportion had experienced arrears in the wait period, but more were pre-October 1995 borrowers than post-October borrowers. The finding that the majority of IS/JSA claimants do not develop arrears but pay with difficulty directs attention to how such households pay. Kempson et al. (1999) also explored this issue. The quantitative findings indicate that with one or two exceptions, claimants and non-claimants seek to meet their commitments in similar ways. Typically they draw on any savings, on family and friends and redundancy payments. Claimants also draw on social security benefits. More detailed information on the experience of the wait period was available from 29 interviews with IS/JSA and ISMI claimants. Research by Walker et al. will shortly report on interviews with mortgagors claiming IS/JSA as a result of unemployment. Preliminary findings indicate that the findings of Walker’s study are consistent with those of Kempson et al. (1999), and that both reinforce the findings of the quantitative research. A key question for research and policy is the extent to which the ISMI wait period is managed by a successful MPPI claim. This requires evidence on the take-up of MPPI amongst claimants, the reason for receipt of IS/JSA and outcome of claimants’ claims. The most direct evidence here is the research by Kempson et al. (1999). The evidence indicates that only a minority of IS/JSA claimants had MPPI (25%), that only a small minority benefited from the policy (5%) but that post-October borrowers were more likely to be helped by MPPI in the wait period. MPPI is of little help to those in the eight week wait period as it typically has a similar deferral before pay-out. Kempson et al. (1999) also provide information on the outcome of claims amongst those on IS/JSA. This group benefit less from MPPI than do borrowers in general. A substantial minority could not claim as the grounds for their receiving IS/JSA were not ones covered by their MPPI polices. Amongst submitted claims, a sizable minority were rejected. However, those making successful claims had to manage the wait period, and MPPI shortfalls were not unknown. (p. 52-55) Studies such as those by Kempson et al. (1999) and Walker et al. only consider those either still in the wait period or in receipt of ISMI. They do not provide evidence on those whose wait period is terminated due to leaving the register. This issue is nevertheless relevant to a consideration of safety-net provision as such claimants may leave with arrears (particularly those in the 39 week wait period), and so face a risk of possession if they are unable to clear
24
their arrears. This issue is discussed further in the section on the interaction of safety-nets and employment. 3.2.4 The effectiveness of ISMI: ISMI shortfalls The eligibility criteria associated with ISMI can result in ISMI claimants receiving less than the full mortgage interest they have to pay to their lender. This potential difference, between the interest due to the lender and the interest paid by ISMI, is referred to as the ISMI shortfall. In 1995, additional changes to ISMI resulted in its payment at a standard rate. This procedure was introduced in order to reduce the large number of under and over payments that were occurring. Payment at a standard interest rate can contribute to the ISMI shortfall where an individual’s actual interest rate exceeds the standard rate (for example, where a fixed rate mortgage becomes uncompetitive as interest rates fall). Research from the early 1990s onwards has identified the ways in which the structure of ISMI contributes to shortfalls, the number of claimants facing shortfalls and the consequences of ISMI shortfalls in terms of arrears and lenders’ forbearance. Qualitative research in 1984 by the Social Security Inspectorate first drew attention to the shortfall issue in an examination of 300 ISMI cases. Subsequent studies (Ford and Wilcox, 1992; Ford et al. 1995 and Kempson et al. 1999, p. 57) have shown the steady growth in the percentage experiencing an ISMI shortfall to around 80 per cent. Studies also indicate the factors that contribute to shortfalls and note the significant impact of the introduction of the standard rate of interest. Kempson et al. (1999) report that this was a problem for approaching half of all those with shortfalls. Excessive housing costs and ineligible second loans were less frequently a problem. Neither of the two surveys discussed above provide data on the monetary value of the shortfall. Although this information is not routinely available from the QSE, it is a possible source of more detail on the value and reasons for shortfalls. Qualitative interviews conducted as part of the Kempson et al. study noted that shortfalls ranged from £17 to £279 per month. Further data from a small number of qualitative interviews may be available shortly from Walker et al. The Kempson et al. study also discussed the ways in which people attempted to meet the shortfall and their success in avoiding arrears. Shortfalls do not necessarily lead to arrears but did so for just over a third of ISMI recipients. Research better able to specify the value of shortfalls is needed if the effectiveness of ISMI as a longer term safety-net is to be identified. The data would also contribute to policy discussion on both forbearance and the extent to which households can realistically cover shortfalls. As already noted, (1.3.6) there is evidence from a number of surveys and interview studies of lenders that many have adopted stricter forbearance following the ISMI changes in 1995. There is, however, none which provides a systematic documentation and analysis of lenders’ current policies and practices with respect to forbearance or the implications of these changes for those in receipt of ISMI. Such evidence as there is comes from surveys of claimants and borrowers in arrears. These include the finding that over a fifth of ISMI claimants had a suspended order against them (Kempson et al, 1999 p62 ). However, studies based on samples of ISMI claimants are, by definition, not able to indicate whether those in receipt of ISMI do, in time, have their property repossessed. 25
3.3 Mortgage Payments Protection Insurance (MPPI) 3.3.1 Attitudes to private insurance The available information was discussed in the section on trends in home ownership. The conclusion was that there are relatively few studies of attitudes to MPPI and that mortgagors are typically ambivalent. 3.3.2 Costs, terms and conditions of MPPI Information on the costs of MPPI have been assembled in two main ways. First by ‘consumer research’ examining information from lenders or collated by publications such as Which Mortgage? Since 1999, average costs for MPPI have also been reported by the Association of British Insurers (ABI) and the CML. Secondly through surveys of mortgagors such as that by Ford and Kempson (1997), or the more robust FRS which collects data on the current monthly costs of MPPI. The 1997/8 FRS survey identified approximately 2,500 mortgagors with MPPI (defined as excluding non-endowment death benefit). The range of costs can also be identified. As already noted the FRS has the advantage of being a large national survey. Data on costs are largely consistent over comparable time periods. Information on the terms and conditions of MPPI at the point of take-up are not available in any systematic robust way from surveys or research studies. For example, data on the exclusion of certain conditions, on deferral periods prior to payment, initial and requalification periods are not collected by the large surveys. Some studies provide information on terms and conditions from an exploration of the claims process (Kempson et al., 1999; Walker et al., on-going; Ombudsman’s reports) but these data only refer to those who have claimed, and overwhelmingly, those whose claims are rejected. They offer a guide to the way in which the interaction between events and the terms and conditions of an MPPI can be problematic, and provide evidence of respondents’ misunderstanding of the terms of their MPPI policy, and of potential mis-selling, but they are not evidence of the range of terms and conditions that pertain across all policies. Again, such data are more readily accessed through the market intelligence publications. A more qualitative assessment was available from Kempson and Ford (1997) who asked lenders to indicate the extent to which their policies included certain groups of people and on what terms and conditions (the inclusion of self-employed workers, of temporary workers etc). Nineteen policies were examined in this way, but the data refer to 1997, and represent only a small proportion of all the MPPI policies available. 3.3.3 Take-up The available information on the level of take-up of MPPI can be drawn from two sources; providers (lenders/insurers) and mortgagors. Until very recently, information from providers was not available on any routine basis. Organisations such as the Association of British Insurers (ABI) occasionally provided statistics of take-up and there have been a small number of surveys of mortgage lenders seeking similar data (KPMG, 1996; Ford and Kempson, 1997). However, the availability of information from providers changed in 1999 when, as a result of an initiative between the ABI and the Council of Mortgage Lenders (CML), for the first time statistics from lenders and insurers were assembled and consolidated. Statistics on take-up of MPPI in this form are available from the start of 1998.
26
Data on take-up are also collected from mortgagors by some of the routine surveys; for example, the FRS and the CML Annual Housing Finance Survey. One-off research studies have also provided data from mortgagors (for example, Maclennan et al.1997, Ford and Kempson, 1997) but while the data are broadly consistent in terms of the trend, take-up percentages do vary. Neither the SEH nor the BHPS offer a reliable guide to MPPI. Both surveys ask respondents if MPPI is a payment included in their mortgage payment and the question is designed as a check on the composition of mortgage payments and not on MPPI per se. As other evidence indicates that approaching half of mortgagors pay their MPPI separately from their mortgage, these figures cannot be taken as a reliable indicator of MPPI take-up. Table 3.1 indicates the figures for take-up from a range of sources
Table 3.1 Recorded percentage take-up of MPPI from a range of sources 1995/6-1997/8 Survey Borrowers as respondents Family Resources Survey CML Annual Housing Finance Survey Cebulla (Omnibus) Ford and Kempson Lenders/insurers as respondents Association of British Insurers (ABI) Ford and Kempson ABI/CML initiative
* **
1996/97 % 24 (44) and 23 * N/A 21
1997/98 % 27 (38) and 28 * 31** N/A
1999
N/A (41) and 31*
20 (Est) 16 N/A 17 17
The figure in brackets are the percentages who indicated they have MPPI in response to a general question about insurance products. The second set of figures are the percentage noting that they could draw on MPPI if unemployed in the furture This survey only considered employed people
N/A
Not available
As shown in Table 3.1, these studies and surveys indicate that: • • • Percentage take-up of MPPI is rising, but year on year increases are modest; Data from surveys of borrowers at any one point in time show some variation in takeup; Lower levels of take-up are reported by lenders and insurers than by borrowers.
A number of considerations are relevant. First, the geographical coverage of the surveys may vary. Second, the nature of the questions asked varies. In particular, the FRS asks a question that combines MPPI and the possession of insurance for death for those with non-endowment 27
mortgages. Polices covering death have then to be excluded to give a figure for take-up of 24 per cent in 1996/7. This figure is close to that of Ford and Kempson (1997). In contrast, the CML Housing Finance Survey typically asks respondents to indicate from a list of insurances which products they have. MPPI is one entry on the list. As Table 3.1 shows, the use of this question results in a level of MPPI take-up of 38 per cent in 1997/8, a considerably higher figure than reported via the FRS. These discrepancies raise questions about how reliable the figures are. In particular, there are suggestions that the CML figures (and others) may be inflated as a result of respondents’ lack of understanding of what products they have in fact purchased. This possibility is supported by the response to a subsequent question in the CML Housing Finance Surveys which seeks to identify those who could draw on MPPI if they were unemployed for any length of time in the future. The response - 28 percent in 1998- is judged to be a much better guide to the level of MPPI take-up. Third, the sampling frames differ. For example, the FRS uses a random sample, the CML Annual Housing Finance Surveys a quota sample while Ford and Kempson sampled borrowers drawn from only a small number of albeit national mortgage lenders (although as noted, a comparison of mortgagor characteristics with the SEH data indicated no substantial bias in the characteristics of those surveyed). Fourth, the surveys are administered in different ways. Not unexpectedly, the interview based surveys (FRS, CML, Cebulla) offer clear advantages over the postal survey (Ford and Kempson, 1997) in terms of an ability to clarify questions, check documentation and so reliability and response rates. Fifth, the treatment of ‘free’ policies (offered by lenders to some first-time borrowers) is uncertain. As suggested above, there is some evidence that, to date, the responses from borrowers are likely to over-estimate MPPI take-up. Equally, responses from lenders have potentially been an under-estimate. This is important in accounting for the discrepancies between the results from lenders and those from borrowers. Under-estimates are, in part, a consequence of the variation in rates of MPPI take-up amongst lenders and hence the particular pool of lenders surveyed will affect the results. There is also considerable variation in the extent and accuracy of lenders’ records and, in addition, lenders may not be aware of the extent to which borrowers have taken MPPI from another source. Lenders may differ in reporting ‘free’ MPPI. Over-estimates by borrowers stem from poor information and knowledge about the products they have taken. McLennan et al (1997) have suggested that borrowers’ estimates of take-up are likely to be 0.67 of the reported level. Studies have shown that some believe they have MPPI when they do not, others confuse MPPI with the Mortgage Insurance Guarantee (MIG) or with insurance to repay the mortgage in the event of death. The significance of some of these issues may be lessening. For example, there has been a reduction in the number of lenders who require MIG and so MIG should be less of a source of confusion. Surveys such as the FRS and CML Annual Housing Finance Survey have also modified the questions they ask in order to try to enhance the clarity about which type of insurance the respondent is referring to.
28
In the light of the above assessment of most of the data sources used to identify MPPI takeup, the recent ABI/CML initiative is timely and offers the prospect of more robust data. It not only draws data from a very substantial proportion of providers but it seeks to include polices sold through intermediaries, to include the provision of ‘free’ insurance and to exclude lapsed and cancelled policies. Table 3.2 outlines the steps taken under this initiative in order to calculate the current figure for the take-up of MPPI.
Table 3.2 Take -up of MPPI measured via lenders and insurers Mortgage and MPPI transactions in stated period Total mortgage loans outstanding at end of period Total mortgage loans advanced in period Estimated number of MPPI polices in force at start of period as % of loans outstanding Estimated number of MPPI policies sold or provided free in the period as % of mortgage loans advanced in period Estimated number of MPPI polices ended in period as % of polices in force at start of period Estimated number of MPPI policies in force at end of period as a % of mortgage loans outstanding Source: CML/ABI (1999) 1998 10,821,000 1,554,000 N/a 23.4% (363,100) N/a 16.6% (1,799,000) 1999 (to June 31st) 10,901,000 807,000 16.6% (1,799,000) 25.3% (204,000) 8.9% 16.9% (1,842,000)
The ABI/CML figures on take-up indicate that roughly 17 per cent of mortgagors had a live policy in June 1999. This methodology and resulting figures clearly have to be tested over time, and problems may emerge, but the early indications are that the ‘true’ level of take-up is well below the levels reported by borrowers (even when adjusted for mis-information), reinforcing the points made above about the impact of their poor understanding of insurance products. The figures for 1999 are also lower than previous estimates from lenders. These findings clearly have implications for the series of questions on MPPI used in individual and household surveys, as they will need to consider how to identify MPPI more effectively (see Section 6.1.4). Moving beyond any simple counts of safety-net take-up to consider additional analysis (for example, the identification of the characteristics of those taking MPPI or claiming on MPPI etc), the ABI/CML figures cannot indicate anything of the socio-economic characteristics of those taking insurance. This information has to come from survey data where the number of survey respondents with MPPI becomes critical. Here the FRS offers the strongest numerical base for analysis with 2526 respondents (27 per cent) reporting MPPI in 1997/8. The CML Housing Finance Survey records 397.
29
3.3.4 Expectations and claims There are few discussions of mortgagors’ expectations of the insurance. Kempson et al. (1999), Walker et al. and Munro et al. (1997) provide some qualitative data on attitudes to MPPI and indicate a range of expectations of MPPI but a considerable degree of scepticism as to whether claims will be met. Walker provides the evidence from 90 interviews and six focus groups which included mortgagors and renters, those with and without insurance. Data are also available from 30 mortgagors who have claimed on their insurance, 10 unsuccessfully and 20 successfully. Kempson et al. undertook 32 telephone interviews with claimants (including 14 rejected claims). Studies typically identify expectations of the insurance retrospectively, at the point of claim, but not pre-purchase. Information on the number of claims on MPPI as a percentage of live policies is provided by the ABI/CML consolidated statistics. The figure for the first half of 1999 was 5 per cent. Kempson et al. (1999) provide the only focussed, systematic study of the claims process to date. The study provides both a quantitative and a qualitative analysis of the claiming process, and the outcomes and consequences of claims. Data on claims were drawn from a sample of 1000 claimants accessed through two national insurance companies to reflect the balance of unemployment and sickness claims. Some information on claiming also came from those in the sample of IS/JSA claimants who had (or previously had) an MPPI policy. The response rate was not out of line with those for many postal surveys (at 47 per cent) but a higher response rate would clearly have made for a more reliable set of results. However, the response rate amongst rejected claimants (who should have formed around 30 per cent of responses if they were to reflect the insurers’ account of the balance of acceptances and rejections) was 15 per cent. Consequently the authors treated the responses qualitatively. Establishing some robust data on the percentage of rejected claimants remains a priority. The ABI/CML statistics start this process and indicate that in 1998, and the first half of 1999, 18 per cent and 22 per cent respectively were rejected. On-going work by Walker et al. provides further qualitative data on MPPI claims. Claimant households were accessed principally through one insurer (a different one to those used by Kempson et al.) while a minority of respondents came from a ‘reserve’ pool from the Kempson et al. study. The results of the two studies are broadly comparable, with rejected claimants prone to arrears and cancellation of their MPPI policies. Equally, successful claimants experience MPPI shortfall payments, and may not avoid arrears. The process of claiming is not always easy, the product can fail to meet their expectations of it and attitudes to private insurance post claiming were, in general, less favourable even amongst those with a successful claim. Currently, it is not possible to analyse claiming processes or the outcome of claims using data from the routine national surveys (for example, SEH, BHPS, FRS, FES, SHCS). There are a number of reasons for this (although not all reasons apply equally to every survey). First, the questions used to identify those with MPPI can be problematic, conflating forms of insurance (see 3.3.3). Other surveys ask questions which identify only a proportion of MPPI mortgagors; for example, the SEH and BHPS only identify those mortgagors who pay for the insurance with their mortgage payment, so excluding the estimated 40 per cent who pay 30
separately. Second, where claiming is considered, only successful claims are noted and the numbers identified can be very small (4 in the 1997/8 FRS). Third, no details of the duration of claims are available. Fourth, questions about the receipt of MPPI are sometimes restricted to those in mortgage arrears as is the case for the SEH. (Further discussion can be found in 6.1.4). 3.3.5 Other forms of insurance safety-net In evaluating the data on take-up, the evidence that MPPI is only one of a number of products which can assist people to meet their mortgage costs following unemployment due to redundancy, sickness or accident needs to be recognised. Critical illness policies or shortterm income replacement cover may be equally appropriate. Thus the MPPI data are an important component of private safety-net provision but other provision and the extent to which it extends the pool of mortgagors with cover, or merely duplicates MPPI cover, needs to be explored. Not all surveys which provide data on MPPI also provide data on other insurance products, but the FRS and CML Housing Finance Surveys do. In principle, therefore, they offer the most comprehensive, on-going account of safety-net provision with the larger cell sizes of the FRS offering a clear analytical advantage. 3.3.6 ISMI and MPPI interface Kempson et al. (1999) provide both qualitative and quantitative information on the ISMI/MPPI interface. The numbers experiencing the interface are, however, relatively low as the take-up of MPPI is not concentrated amongst those with eligibility for IS/JSA. Further, amongst those with MPPI, who in time would be eligible for ISMI, Kempson et al (1999, p. 63) noted that a third were no longer claiming and had returned to work. Potentially, the large surveys could identify people with an experience of the interface, but the small numbers would preclude any sensible analysis. None of the routine surveys currently ask any questions on the interface.
31
4.
SAFETY-NETS AND WORK DISINCENTIVES
There is a very substantial literature on work disincentives, much of it econometric. This literature is not reviewed here. Rather, studies focussing on the behavioural responses of mortgagors in the face of theoretical disincentives associated with the housing related assistance in the form of ISMI and MPPI are considered. There is no systematic information available currently on the impact of MPPI on the decision to accept or reject employment. The Kempson et al. (1999, p. 43) study does, however, report that a proportion of people returned to work before the end of their MPPI claim. Indicative evidence from insurers and lenders also supports this. There are more data available about the impact of ISMI on the decision to work. In principle, this decision is a critical one for mortgagors as the absence of any in-work housing related benefit means they face a severe unemployment trap and need to return to work at relatively high wages in order to match or exceed the combined IS/JSA and ISMI payment. Theoretically, ISMI is a work disincentive. Studies such as Webb and Wilcox (1991) have used secondary data analysis to indicate the extent to which the ISMI disincentive (the unemployment trap) is sensitive to mortgage interest rate changes as well as to wage levels and the ISMI regulations on repeat claims. Empirical data raise some questions over the disincentive argument and studies have begun to suggest that the calculations made around ISMI and employment are not exclusively economic ones for all households. First, there is evidence that a considerable number of mortgagors are in work at incomes below IS levels. A report to the House of Commons Social Security Committee, 1994/95, showed that 600,000 working mortgagor households had incomes below IS/JSA level. Second, several studies have indicated the impact of low wage employment on the development of mortgage arrears which may have been avoided had borrowers not left unemployment or had been willing to withdraw from the labour market (Ford, 1989; Davis and Dhooge, 1993; Ford et al. 1996). Ford et al. (1996) were directly concerned to examine the impact of housing costs on the decision to work and presented qualitative data on 40 households, 16 of whom were mortgagor households. The study challenges the universal application of a ‘better off’ calculation. It provides evidence that some mortgagors returned to work at incomes equal to or worse than benefit levels (as did some tenants). Similar findings were reported by Oldman and Kemp (1996). These studies suggest that mortgagors are highly committed to work, with a proportion responding to a belief that any job will then enable them to obtain a better job, findings also available from other qualitative studies such as Kempson et al. (1994). Equally, there are interviews with mortgagors that do show the disincentive effect of ISMI (Ford, 1995). Here, important aspects of the decision not to work are a fear that employment resulting in an income below the resources available through JSA and ISMI will result in debt, but also a principled belief amongst craft workers that the pay they receive should recognise their skills. Amongst this group, qualitative studies indicate a proportion who are prepared to work ‘on the side’ (Ford, et al 1996). An important question raised by this 32
research is the extent to which there are distinct groups of mortgagors with different responses to the work/benefit relationship, and the key influences on those groupings. Not dissimilar findings are suggested by the early results from Gray, funded by the European Commission, who is examining social assistance programmes on a comparative basis. In England, the experiences of 96 claimants, drawn from unemployed people’s centres and training projects, have been considered via focus groups and self-completion questionnaires. The results point to ISMI being perceived by many mortgagors as a work disincentive, reinforced by the difficulties they face in gaining work at the level of their previous wage and their concerns about subsequent debt. Overall, the qualitative studies cannot provide evidence on enough cases for robust analysis, although these studies provide an important guide to the range of information that is pertinent to any future survey-based work. However, the qualitative data does highlight the extent to which the response to the ISMI disincentive is contingent and cannot be assumed. To date, there is no purposively designed quantitative study of work disincentives that considers the impact of housing assistance for mortgagors. However, there are a number of large surveys, designed either to consider the effectiveness of particular labour market interventions and/or the attitudes and behaviour of unemployed people, (including mortgagors) and which are centrally interested in the work disincentive issue (for example, McKay et al, 1999; Marsh et al, 1999). These studies might, in passing, provide evidence about the housing cost/work disincentive issue. In most instances, and not unexpectedly, the emphasis in these studies is on the impact of the labour market intervention for key groups of unemployed people or low-income workers. Issues of tenure and housing costs are not central. Where tenure is considered it tends to be in the form of a comparison between the speed of return to work by owner-occupiers and tenants. A number of surveys indicate that mortgagors are more likely to leave benefit ahead of tenants, particularly social tenants, but that these differences reflect the socio-economic characteristics of these two groups not tenure per se. Thus, there is only limited analysis of the extent to which the provision of housing assistance for unemployed mortgagors constitutes a work disincentive. This work tends to be qualitative. One consequence of this lack of research is the clear gap in the available knowledge concerning the impact of the changing the nature of ISMI on the decision about accepting work amongst unemployed post-October 1995 borrowers. The theoretical arguments suggest that, in the absence of MPPI, the curtailment of ISMI is likely to encourage those with a nine month qualifying period to return to work rapidly. However, any willingness or pressure to reduce the reservation wage will potentially be tempered by a conflict between the likelihood of persistent, on-going, in-work debt (a concern noted in both qualitative and quantitative studies) and an even more difficult short term situation during the ISMI wait period but ultimate access to support with housing costs. This issue is also returned to in Section 6.1.6.
33
5.
INTERNATIONAL COMPARISONS
There are no systematic, comparative studies on most of the substantive issues discussed in this report. Indeed, for a number of countries (for example, France, Portugal, Sweden) primary data on key issues such as the number of households with arrears, in possession, with different forms of private safety-net are not available or are unreliable, being drawn from small scale surveys undertaken by advice agencies or consumer groups serving particular groups of people One exception is information on state support with housing costs for mortgagors where some comparative data have been assembled (Hirsch, 1994, Kemp, 1997). Work by Gray on social assistance in 12 European countries is on-going. These reports indicate the extent to which the specific schemes and benefits have to be situated in the wider context of the nature of the housing finance and social security system in each country, but they do provide important descriptive data and the basis for a comparative analysis. The emphasis, however, is often more strongly focussed on the rented sectors than owner occupation. The impact of support with housing costs on work disincentives sometimes forms a part of these discussions. Hirsch (1995) for example noted that due to the structure of the benefit system, work discentives for mortgagors were only perceived as an issue in England and Sweden. More recent research on housing allowance schemes in European countries is ongoing (Ditch and Wilcox) but is not yet reported. A number of studies do, however, draw on international experience. A study of mortgage default in Australia reviews and draws on research conducted in the UK and America, setting out the range of theoretical issues addressed and methodological approaches adopted (Berry et al, 1999). Studies concerned with safety-net provision have examined some or all of the systems for supporting mortgagors that operate in Sweden (Wilcox and Sutherland, 1997) or France (Holmans, 1997; Ford and Wilcox, 1998), but there is a sense of these being ad hoc or selected comparisons rather than any comprehensive comparative review. A more detailed comparative analysis of the forms of protection for mortgage borrowers in some European countries may result from a forthcoming seminar series funded by the European Commission. Comparative data on private sector social protection following unemployment should also result from a study funded by the Anglo-German Foundation (Cebulla, 1999). There is a much larger comparative literature on labour market risk, drawing on comparable data from the European Labour Force Survey and the work of OECD which has the potential to help identify comparisons, but little of this has been used in relation to mortgagors and safety-net provision.
34
6.
ISSUES FOR RESEARCH
Following the review of data sources and literature on state provision and private insurance for mortgage interest, it is possible to identify a number of significant gaps in the research base and emerging issues all relevant to the analysis of support for sustainable owner occupation. Seven areas are set out below. The aim is to identify the broad issues rather than the detailed research questions to which they give rise. 6.1.1 The risks to mortgagors A necessary underpinning for discussion of safety-net provision is a better understanding of the nature, extent and distribution of risks amongst mortgagors. The current concern about safety-net provision is taking place in a different socio-economic context to the one that pertained in the 1980s and into the mid 1990s. By the end of the 1990s there is, arguably, more recognition of the extent to which structural change has resulted in mortgagors facing a wider range of risks, for example, from the labour market and relationship breakdown. The growing risks to mortgagors raise at least two separate issues for research. These are: • • the extent, nature and distribution of risks amongst mortgagors; and the extent to which mortgagors perceive the risks they face and attempt to minimise their effects.
There is completed and on-going research addressing the second of these two issues (see Section 2.2.3). However, research assessing the distribution of risk and the risk dynamic amongst mortgagors is much less well developed. Further, such research as there is into the risks mortgagors face has typically assessed each risk separately and in isolation. Less frequently has there been analysis of the accumulation of one or more risks (including relationship breakdown, ill health, temporary work, underemployment or unemployment, financial pressures) across a period of time for households and their housing consequences. Research on this issue would identify at a fundamental level the extent and impact of structural change on home owners (and which home owners), and which changes (or combinations of changes) are most associated with payment difficulties. On the basis of this knowledge it would be possible to assess the extent to which current forms of safety-net provision are likely to be adequate and effective and the areas for further policy initiatives. 6.1.2 Issues relating to mortgage arrears Research on the proportion of borrowers in arrears and their characteristics is now well established. However, there is no recent research on the incidence of arrears and in particular whether arrears are highly concentrated by household or widely dispersed. The most recent evidence comes from the period 1990-1994 when one in five mortgagors had at least one experience of payment difficulties. 35
The very particular circumstances that pertained in the early and mid 1990s indicate that we should be cautious about assuming similar patterns now. There have been suggestions that following the recovery in the housing market, arrears are increasingly a ‘ revolving door’ problem for marginal home owners indicating that the problem of unsustainable home ownership is narrowly focussed. This view is, however, countered by the suggestion that risks, and hence payment difficulties and arrears, are more widely distributed amongst mortgagors. New research should develop an understanding of the balance between single or limited episode arrears and repeat default and, in addition, focus on what it is that facilitates sustained recovery or exacerbates ‘revolving door’ default from the perspective of the mortgagor. ‘Revolving door’ refers both to a situation of rising and falling arrears (but no full clearance) and where arrears are cleared but recur. Research on these issues is clearly distinct from research to identify the risk factors that predispose payment difficulties (6.1.1), but of related interest. 6.1.3 Lender forbearance New research is required on lender forbearance for several reasons. First, the circumstances in which lenders have to consider forbearance are changing. There is, for example, a higher proportion of mortgagors experiencing ISMI shortfalls than was the case in the mid 1990s. Only a minority of borrowers have MPPI but these claimants can also face shortfalls. Secondly, how lenders respond to these new situations is critical in understanding the effectiveness of the two safety-nets and their capacity to prevent households being taken into possession. Third, most information on forbearance pre-dates the 1995 changes to ISMI, the significance of MPPI and the introduction of the Mortgage Code (which, arguably, is at its weakest with respect to forbearance), and as a result, existing studies cannot be assumed to offer a good guide to current policy and practice. Although there is limited evidence that mortgage lenders have become more stringent following the changes to the ISMI eligibility in October 1995, there is a lack of robust, contemporary information on lenders’ policies and practices towards forbearance and the implications for borrowers. 6.1.4 MPPI Further research should address both the take-up and the effectiveness of MPPI. As already noted, survey-based measures of take-up are problematic and there is a need for some work developing and testing a more appropriate set of questions for use on the large national surveys. Equally, the large surveys provide almost nothing on claims and claiming. The only major study to date (Kempson et al. 1999), predates the introduction of the CML/ABI benchmark product and the suggestions that it will result in fewer exclusions and so form a more effective safety-net. It also spans the period of the introduction of the Mortgage Code which is designed, in part, to improve the selling of MPPI. Four other separate, but related, areas require further research. First, there is indicative evidence from existing research of a disjuncture between the requirements and practices of a 36
flexible labour market and the initiatives to secure greater clarity in the terms and conditions of MPPI policies. For example, the absence of contracts of employment, excessive casualisation, the implementation of annualised hours and zero hours contracts result in a lack of clarity as to the circumstances in which, for example, someone is permanently or temporarily employed, when they are ‘laid-off’ temporarily, or when they are redundant. Research shows that these issues emerge when claimants are asked to provide evidence as to whether, and why, their contract was terminated. Some identification of the range of labour market practices that are proving problematic is required in order to evaluate the inclusiveness of the terms and conditions of MPPI. Second, research is needed on the selling of MPPI by brokers and by direct MPPI providers. In both instances the research questions are about the selling process, the cost, terms and conditions of policies and the claims record vis-à-vis block-sold policies. There are also important questions about the ways in which individuals and households select and access brokers, the expectations they have of them, the ways in which they assess the advice given and their understanding of the regulatory processes and means of redress. Research might also identify the compliance of brokers with the Mortgage Code (research which should also be undertaken for all lenders and insurers) in order to contribute to policy discussions on regulation. Third, and closely related to above, Maclennan et al (1997) noted the need for research that captured the decisions of borrowers and lenders/sellers with respect to MPPI at the time of taking out the mortgage. Key questions concern what it is borrowers believe they are purchasing, how well does the Mortgage Code work for the different parties, how significant are cost considerations in understanding the reluctance of low income borrowers to take MPPI and to what extent their labour market circumstances are ones that exclude them etc. Fourth, further work is needed on the reasons for cancellation of MPPI. 6.1.5 ISMI Two issues relating to ISMI require further research. First, shortfall payments currently challenge the ability of ISMI to provide an effective safety-net. The substantial increase in the proportion of claimants with a shortfall indicates a potential long term problem, and an increased risk of possession. A study that identified the extent and value of shortfalls, both the ways in which borrowers and lenders manage shortfalls and the outcomes of shortfalls would be valuable. These issues should be compared for pre and post-October 1995 borrowers. A focus on longer-term ISMI recipients where the twin risks of possession and growing poverty may occur is also needed. Second, systematic research on the ISMI /MPPI interface is necessary. In particular, research is required on the nature and effectiveness of the administrative processes that surround the transition.
37
6.1.6 Safety-nets and work disincentives Current policy concerns often focus on the extent to which ISMI and MPPI constitute work disincentives or moral hazards. As noted in Section 3, several qualitative studies examining the behaviour of mortgagor job seekers have found a varied response to leaving ISMI (with little known about MPPI). One key issue for research is the impact of the pre and post-October 1995 ISMI regimes on the return to work, where a central focus should be the decision making process for the two different groups. As noted in Section 4, such a study would provide the basis for extending understanding of the work disincentive associated with ISMI. There is no comparable research exploring the extent to which MPPI constitutes a work disincentive, although insurance theory predicts the likelihood of moral hazard, with individuals able to reject particular jobs in favour of waiting for something better. Compared to unemployed people with similar characteristics, the theory predicts that the duration of unemployment will be longer where there is a successful MPPI claim. This is a key area for further research focussing on the duration of unemployment and the behavioural responses of those claiming MPPI to job opportunities. A further concern for policy makers with respect to owner occupation is likely to be the impact of the introduction of the Working Families Tax Credit in October 1999. This more generous benefit may have a number of consequences including the prevention of arrears amongst those in low-paid work and a reduction in the depth of the unemployment trap, and so any work disincentive, for mortgagors. An evaluation of this initiative would be appropriate after a period of time. 6.1.7 Comparative research The absence of comparative research on most of the issues discussed in this report suggests a number of possible areas for further comparative research. Particularly productive areas are likely to be those where research seeks to learn from the policies adopted in other countries to address similar problems. Comparative study of the public-private safety-net balance is one such area, patterns and outcomes of lender forbearance is another and associated with that the use of the judicial possession procedure or alternatives to it. 6.1.8 Summary and conclusions This section has drawn attention to a number of potential studies where research is needed and which would provide data, analyses and evaluations that currently constitute critical gaps in our understanding of public and private safety-net provision, and the prior issues of the risks to mortgage borrowers in contemporary society. Thirteen main areas for further work have been outlined. They are: • • the distribution and interconnection of the risks faced by mortgagors the incidence of arrears and the extent of ‘revolving door’ arrears 38
• • • • • • • • • • •
lender forbearance developmental work on an improved module of questions on MPPI take-up a study of the effectiveness of MPPI claims the role of brokers in the MPPI market the ‘in principle’ suitability of MPPI in a flexible labour market and the key discontinuities the decision making process at the point of MPPI take-up the reasons for the cancellation of MPPI the extent and management of ISMI shortfall payments the effective management of the ISMI /MPPI interface the work disincentive effects of different ISMI eligibility regimes and of MPPI a range of comparative research.
This list is illustrative rather than exhaustive. Discussion of each area has been undertaken separately. In practice, a number could usefully be linked together. This report has indicated that the need to focus research and policy discussion on private and public safety-net provision remains critically important. With the housing market increasingly buoyant, there were still in excess of 30,000 households who lost their property in 1998, and a further 16,410 in the first half of 1999. At the end of June 1999, 219,470 households owed three or more months mortgage payments, and the 1997/8 SEH indicated 14 per cent of mortgagors were finding it difficult to meet their payments. MPPI is developing only slowly and it’s effectiveness is not yet proven, either as a product per se or over the economic cycle. There will be an inevitable but gradual shift to a higher proportion of postOctober 1995 borrowers who face the nine month ISMI wait. ISMI shortfalls are occurring and can be substantial, creating significant payment problems for some claimants. At the same time, home ownership remains the aspiration and preferred tenure of the majority of households and, in reality, the only housing option for many. Thus, the sustainability of home ownership remains a key policy concern.
39
PART 2 SOURCE BOOK
Contents
Introduction ..............................................................................................................................42 Review of articles.....................................................................................................................44 Review of data sources...........................................................................................................114 Index by key themes ...............................................................................................................126 Characteristics of mortgagors and mortgage lending.....................................................126 Mortgage arrears ............................................................................................................126 Mortgage possessions.....................................................................................................127 Means of managing arrears (mortgagors and lenders) ...................................................128 Information and advice...................................................................................................129 ISMI ...............................................................................................................................129 MPPI ..............................................................................................................................130 State and private interface ..............................................................................................132 Safety-nets and work incentives.....................................................................................132 International comparisons ..............................................................................................132 Bibliography...........................................................................................................................134
INTRODUCTION
This source book contains abstracts of studies that address issues germane to the analysis of both state and private safety-net provision for mortgagors. In addition, information is provided on the major surveys that potentially assist in this analysis. It has been necessary to put some boundaries around the literature that is reviewed. Key issues are those of mortgage arrears and possessions, public and private safety-nets, and some consequential issues such as work disincentives. While these concerns have to be placed in the context of recent trends in the housing market, the literature and studies relating to recent housing market developments per se are not reviewed. Similarly, the very substantial literature on work disincentives is not reviewed in full, but studies that focus more directly on the work disincentives issue for mortgagors are included. Thus, the source book is not designed to be exhaustive of everything published, but to cover a number of key areas in considerable detail. A number of major studies have also resulted in multiple publications, typically for professional or practitioner audiences and these publications are generally excluded. Each abstract is followed by the identification of a number of key themes. They are drawn from 10 that are used throughout the report: • • • • • • • • • • Characteristics of mortgagors and mortgage lending Mortgage arrears Mortgage possessions Means of managing arrears (mortgagors and lenders) Information and advice ISMI MPPI State and private interface Safety-nets and work incentives International comparisons
Finally, for each entry there is a brief summary chart giving some broad methodological details of the study, the form of analysis used and the author(s)’ own identification of any major limitations with the data and/or approach. A more detailed evaluation of the utility of the data available is provided in the accompanying Evaluative Report. The extent to which each of the surveys provide information relevant to the analysis of safety-net issues for mortgagors is also indicated by the use of the key words.
42
The source book is structured as follows. First, abstracts of studies are presented alphabetically by author. These are followed by brief summaries and discussion of a number of major surveys, again alphabetically by survey name. There is then an index with entries listed under the key themes and finally an alphabetical bibliography.
43
REVIEW OF ARTICLES
Association of British Insurers (1998) ‘Costing of state/private welfare products for hypothetical families’, Insurance Trends, July 1998, pp 25-33.
Objectives This article by the Association of British Insurers (ABI) forms part of an on-going stream of research on the potential of private insurance to replace in part or in whole welfare state provision. The article considers the cost of a range of products including ‘short term income replacement for accident, sickness and unemployment’ under which the authors include MPPI. They do not, however, model MPPI (replacement mortgage costs) but rather replacement income costs. The article is referenced here as short-term income replacement policies may be used as an alternative safety-net to MPPI (although clearly more expensive). Findings Seven hypothetical family types are identified and the costs of a range of insurance products are calculated. The bulk of the discussion concerns pension provision but with respect to short-term income replacement, the costs for each of the hypothetical families per month range from £5-£75 for 50 per cent of annual earnings for 12 months. In every instance this constitutes two per cent of gross income. The article lists all other ABI work that forms part of their evaluation of the role of insurance in replacing the welfare state. Key themes: MPPI; state/private interface
Type of research Source of data Sample Geography Date of study Stated limitations
Modelling Insurance companies offering particular products N/A National 1998 Hypothetical families with circumstances clearly detailed. Insurance policy assumptions made clear. Recognition that long-tem income replacement policy is also hypothetical. Statistical Association of British Insurers
Method of data analysis Sponsor
44
Burchardt, T. and Hills, J. (1997) Private Welfare Insurance and Social Security: pushing at the boundaries, York: Joseph Rowntree Foundation
Objectives This report examines three areas where private insurance has taken over part of the role of state social security (or might do in the future). It aims to make an empirical contribution to the debate on the boundaries between public and private welfare sectors. The three areas covered are: mortgage payment protection insurance, permanent health insurance and long term care insurance. It considers protection against unemployment only. The report uses existing data from various sources including the British Household Panel Survey and the Labour Force Survey (data from the early and mid 1990s). Findings The report discusses the issues and problems surrounding private insurance for the three areas noted above and surveys the policy conditions and costs of products available in 1996. It then uses evidence from the large-scale surveys to compare the claims people might make if insured with premiums paid, and the distributive effects of moving from tax-finance to riskrelated premiums. The discussion on mortgage payments protection insurance starts by noting that mortgagors as a group were less likely to be unemployed than renters, and that those with mortgage payments protection policies were even less likely to be unemployed than those without policies. The report found that the actuarial cost of premiums (the cost based on an assessment of the policy holder’s characteristics and the probability of pay-out) was around half of that actually charged (based on 1996 figures). However, there was no evidence to suggest this difference was due to the insured population having higher unemployment levels or risk than the uninsured population (as would result from adverse selection or moral hazard). Mortgagors would lose if there were a switch from general tax funding to flat-rate premium funding for mortgage protection, while other tenures would gain. By income group, the effect is regressive with flat rate premiums; if premiums are differentiated by previous employment history, the regressive effect is even stronger as those with lower incomes tend not to meet the employment history condition and hence pay higher premiums or go without cover. The report concludes that although the evidence suggests that mortgage payments protection is relatively straightforward, and the group to be insured has already been screened for risk of unemployment during the process of obtaining the mortgage, there are still a number of problems. These include: incomplete cover as those without good employment records and 45
mortgages starting after 1995 are excluded from private and public welfare provision; and expensive premiums as, for the cover offered, the actual premiums were twice as much as the actuarial cost. They also point again to the regressive effect and note that there are information problems: the report identified information costs for consumers, such as the potential for mis-selling or misunderstanding of the conditions for claiming, and a resulting high level of unsuccessful claims. Two possible solutions are suggested. First, the introduction of compulsory mortgage payment protection insurance, with no mortgagor refused cover so that low-risk mortgagors cross-subsidise those at higher risk. Alternatively, the introduction of a single national scheme (in view of the high cost of commercial insurance), financed, for instance, through a reduction in the rate of Mortgage Interest Relief.
Key themes: MPPI
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Modelling Labour Force Survey, British Household Panel Survey N/A UK 1997 Modelling for MPPI restricted to unemployment cover Statistical Joseph Rowntree Foundation
46
Burrows, R. (1997) ‘Mortgage indebtedness in England: an “epidemiology”’, Housing Studies, 13, 1, pp 5-22.
Objectives • To identify the socio-economic determinants of mortgage arrears of three and more months through secondary analysis of data from the 1993/94 Survey of English Housing. • To identify the relative importance of potential influences on arrears by logistic regression analysis and to identify the best fitting model predicting the odds of arrears.
Findings The following variables were found to provide the best prediction of an increase in the odds of arrears; age, employment status, social class, household structure, region, loan: value and date of mortgage. Those aged 18-24 were significantly more likely to develop arrears and those aged 55 and more were significantly less likely to develop arrears. Unemployed people were almost eight times as likely to develop arrears as those in full-time employment, selfemployed people also had higher odds of arrears, as did skilled manual and semi-skilled manual workers. Divorced and separated households had higher odds of arrears as had couples without dependent children and those who lived in the South-East. There was no increase in the odds of arrears consequent on being a first-time buyer or right-to-buy borrower.
Key themes: Mortgage arrears; characteristics of mortgagors and mortgage lending
Type of Research Source of data Sample Geography Timing Stated limitations
Secondary analysis 1993/4 Survey of English Housing Random sample of 20,000 heads of households of whom 8,462 were the basis of the analysis of mortgage arrears England 1993-1994 Absence of a potentially key variable - the ratio of mortgage costs to income at the time of purchase. Cross sectional data whereas the modelling infers some temporality between the variables which is not always sustainable. Respondents are heads of households and the results exclude any consideration of the impact of other household members on the odds of arrears. Logistic regression None
Method of data analysis Sponsor
47
Burrows, R. (1998) The Dynamics of The Owner Occupied Market, London: Council of Mortgage Lenders
Objectives The research examined the medium to long term dynamics of the owner occupied housing market by examining the flows into, out-of, and within the sector between 1981 and 1997. A major concern was to identify the likely structural changes over the 1990s within the housing market. The study was based on a secondary analysis of large scale data sources, principally the Survey of English Housing (SEH) for the years 1993/, 1994/5, 1995/6 and 1996/7 but also the British Household Panel Survey Waves 1-5 (1990-1995). Findings As part of the discussion of the nature of structural change impinging on the housing market, evidence of emerging patterns of vulnerability was reviewed. The socio-economic characteristics of those in mortgage arrears or facing payment difficulties were examined using the SEH data. Logistic regression analysis for all four years was used to identify the variables that best predict the risk of arrears. Key results were that the odds of arrears had declined over the four year period but that a greater risk of arrears was associated with being: younger, divorced or separated, a household with children, not in full-time employment, selfemployed, a skilled, semi-skilled or unskilled manual worker, living in London or the SouthEast, with a 100 per cent mortgage, and buying between 1987 and 1989. The report concludes by discussing the likely future impact on owner occupation of the changing circumstances of young people, changing patterns of household dissolution and changing patterns of employment. Key Themes: Mortgage arrears
Type of research Source of data Sample Quantitative Secondary Analysis Survey of English Housing (1993/4-1996/7), British Household Panel Survey Waves 1-5 (1990-1995) Random sample of approximately 8,462 heads of mortgagor households from the SEH; approximately 2,000 heads of mortgagor households from the BHPS England (SEH), Britain (BHPS) 1998
Geography Date of study Stated limitations Method of data analysis Sponsor
Statistical Council of Mortgage Lenders
48
Cebulla, A. (1999) ‘Government plans and individuals’ intentions: the case of unemployment insurance’, Benefits. Objectives This paper examines the extent to which individuals have taken forms of income replacement insurance and the factors that are associated with the willingness to take such products. Four products are considered: income replacement for unemployment; for illness and accident; mortgage payment protection insurance and other creditor insurance (for loans). Future intentions to insure were also examined. The research was funded by the Economic and Social Research Council as part of the Risk and Human Behaviour Programme entitled Families Confronting Unemployment (Walker et al, 1997-2000). The data come from commissioned questions forming part of an Omnibus survey of 1000 employed adults in Great Britain; from six focus groups and qualitative interviews with 45 families with at least one partner in employment. Findings The study found that the take-up of MPPI was 31 per cent; creditor insurance was 21 per cent, income replacement for unemployment was 6 per cent and for illness and accident, 18 per cent. Comparison figures are provided for take-up on other forms of insurance. The characteristics of those taking these insurances show that take-up increases with income. There was a gradient by socio-economic group with manual workers the least likely to take insurance of this type. Those at the greatest risk of unemployment were the least likely to insure. Self-perception of risk was also important with higher socio-economic groups likely to over estimate their degree of risk. The experience of unemployment was not a significant influence on take-up. Thirty per cent indicated a future intention to insure for unemployment income replacement; cost and insurance averse attitudes were significant reasons for not seeking insurance. In addition, many respondents attached greater significance to planning for retirement than for unemployment which was seen as an increasingly routine occurrence within society. Key themes: MPPI
Type of Research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor Quantitative Employed Individuals Random stratified sample of approx 1000 households (Omnibus). Quota based selection for focus group and qualitative interviewees Great Britain (Omnibus); Selby and Leicester for qualitative respondents November 1997 Omnibus considered only employed people Statistical Economic and Social Research Council
49
Davis, R. and Dhooge, Y. (1993) Living with Mortgage Arrears, London: HMSO Objectives This study is part of a broader research project aimed at identifying ways for lenders, local authorities and other agencies to help households with mortgage arrears and reduce the incidence of repossessions. The first part of the research focussed on the role of local authorities in tackling mortgage default; this part focussed on households with mortgage arrears. It was concerned with: identifying the reasons for mortgage arrears; outlining the various strategies adopted by households to deal with the problem (including relations with lenders); and reviewing the role of advice (whether sought, from whom and when and the quality of that received). The research was qualitative, took place in 1992 and 1993 and involved in-depth interviews with 46 households. Findings The report starts by discussing the occurrence of mortgage default and the overall economic and structural reasons which led to the rapid increase in the number of households in mortgage arrears in the early 1990s. It then uses case studies from eight households to illustrate the diversity of situations found by the researchers. They are also used to demonstrate the difficulties faced by lenders (and others) when trying to develop standardised procedures to deal with arrears. The study found that arrears were not caused by a single event, but were usually the result of a combination of factors and events (for example, disruption to employment, failure of Income Support, relationship breakdown) which lead to households ‘sliding into arrears’. The research then considered reactions to the mortgage arrears and the factors influencing the degree of stress or anxiety. It found that the scale of arrears to income and the perceived timescale of the problem (long or short term) were particularly important. However, four other factors were also identified: views on ‘who is to blame’ for the arrears; attitudes to home ownership; family responsibilities and perceptions of the lender and housing options. These are all discussed in detail. The study also reviews the different solutions adopted by households to deal with mortgage arrears, for example, reduction in standard of living, drawing on resources from inside or outside the household, prioritising debts. The report closes by reviewing the lessons that can be drawn from the findings and suggests ways in which lenders can better deal with borrowers who fall into arrears. Key themes: Characteristics of mortgagors and mortgage lending, mortgage arrears; means of managing arrears; ISMI
50
Type of Research Source of data Sample
Qualitative Depth interviews with borrowers Various methods of recruitment were used: return to people interviewed for a survey on housing needs which took place in 1992 (16) recruitment from two current surveys (18) announcement on Greater London Radio (5) two Housing Associations running Mortgage Rescue Schemes (6) informal (1) London and surrounding areas Fieldwork took place between December 1992 and May 1993 Recruitment for the interviews was not easy. The research team had intended to involve 60 households to ensure a wide variety of situations were explored but in the event 46 took part. The researchers found that asking people to take part in person improved the chances of co-operation, as did offering a small financial gift. Additionally, recruiting was easier from the contemporary surveys. The researchers felt that, although they did not achieve their desired number of interviews, they did meet their aim of achieving variety in terms of household composition and circumstances. Thematic analysis of qualitative data. Joseph Rowntree Foundation and Building Societies Trust
Geography Date of study Stated limitations
Method of data analysis Sponsor
51
Evans, A. (1998) Responding to Mortgage Arrears: a review of local mortgage arrears fora, London: Council of Mortgage Lenders.
Objectives The project aimed to examine the role, functioning and likely future for mortgage arrears fora. Mortgage arrears fora are local initiatives that bring together key public, private and voluntary sector organisations to improve local responses to mortgage arrears cases. Findings Based on a telephone survey and interviews with fora co-ordinators, the research indicated that the initiative for mortgage arrears fora had come from the local authorities. The majority had not been set up until 1995 although arrears and possessions peaked earlier. Fora operate through regular meetings aimed at improving information flow. The reported benefits included better communication, joint training, development of good practice guides and information leaflets. Reported limitations were the growing regionalisation or national centralisation of arrears management that did not fit well with a local initiative, the nonparticipation of secondary lenders and finance companies, a lack of direction and some advisors’ antipathy to lenders. Of 14 local fora, one had closed and nine were reviewing their future. One was operating under a new, wider remit as a financial services forum. Only half of all members indicated that they thought their fora had a long-term future. Options being assessed included widening the geographical area covered and widening the subjects covered to other aspects of marginal home ownership. The report also notes that some of the issues involving multiagency liaison at local level need to be reflected in national structures. Some of the issues might be better tackled at national than local level. Key themes: Means of managing arrears; mortgage possessions
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Qualitative Mortgage arrears fora co-ordinators All identified co-ordinators Great Britain 1997/8
Thematic analysis of qualitative data Council of Mortgage Lenders/Building Societies Trust
52
Ford, J. (1989) ‘Casual work and owner occupation’, Work, Employment and Society, 3,1, pp 29-48.
Objectives This article discusses the implications of the then recent labour market changes (the growth of flexible employment) for the maintenance of owner occupation in contrast to the prevalent emphasis on the repercussions of unemployment for mortgagors. It brings to the fore the ways in which the changing forms of employment (i.e. the growth in casual and contract work, and self-employment) increased the risks of mortgage default. The article uses data from a study of 40 households in mortgage default. This study was principally concerned with the borrowers’ perception of their route into mortgage arrears and their experience and management of default. It took place in 1985 and the participants were interviewed up to three times over a nine month period. The secondary analysis for this article explored three questions: • • • To what extent are unemployed home owners re-employed in the secondary sector, and is there any evidence that this is an increasing trend amongst home owners? To what extent is re-employment associated with a lower wage and/or with intermittent employment? Is there any evidence that re-employment of those in the secondary sector is associated with difficulties in the maintenance of owner occupation?
Findings The participants in the research, when unemployed, showed a marked trend to be re-employed in more casual and/or insecure employment. Such employment reduced income and decreased security, and increased vulnerability to default as re-employment meant the participants were excluded from any possibility of state support toward mortgage interest payments. In the event of arrears, and possession, a wider chain of events could result: for example, possession might mean a change of tenure, which in turn could result in a change of location; this change potentially made it more difficult to maintain previous work-based or social ties, in turn making it more difficult to find further work. Case studies are used to illustrate the various consequences. The article concludes that ‘the combination of owner occupation and casual work are a prescription not for guaranteeing continuation as an owner-occupier and worker, but for setting in motion increasing vulnerability and disadvantage’. Key themes: Mortgage arrears, safety nets and work incentives
53
Type of Research Source of data Sample Geography Date of study Stated limitations
Qualitative Owner-occupiers in mortgage default Drawn from four building societies. Participants owed about three months back payments when first contacted. East and West Midlands 1985 Participants formed a cohort and were monitored over a nine month period. Each household was interviewed up to three times, and were dropped from the study if – they cleared their arrears; left the tenure; or, failed to respond to requests for interview. Thematic analysis of qualitative data Economic and Social Research Council
Method of data analysis Sponsor
54
Ford, J. (1993) ‘Mortgage possessions’, Housing Studies, 8, 4, pp 227-240
Objectives The objectives of the article were to: • • • • examine the conceptualisation and measurement of possession; identify the characteristics of those at risk of possession; identify the factors associated with the different routes out of possession; and consider the impact of possession on borrowers’ finances and attitudes to home ownership.
The study involved primary data collection on 307 mortgagors in possession with one large national lender. Data were drawn from the mortgage files and followed by 22 in-depth interviews with borrowers in or on the point of possession. Interviewees were drawn from the lender, a local authority and National Debtline. Findings The paper suggests that administrative statistics mask the extent to which voluntary possession occurs. It also draws attention to the phenomenon of voluntary possession with and without arrears. Voluntary possession was most likely amongst those who were young, with recent (1991) mortgages, 100 per cent mortgages, high monthly payments and commitments to second loans. These borrowers tended to be concentrated in the South-East. Depth interviews identified the routes into arrears. The decision to give voluntary possession was influenced by personal stress, a belief that the loss of the property was inevitable and a (mis)perception that voluntary possession conferred advantages, particularly a removal of financial liabilities. The absence of access to alternative accommodation (other than following compulsory possession) was a key constraint on voluntary action as was the belief that the lender would sell for less than the borrower would. Compared to those in arrears, experience of the possession process increased the critique of home ownership. The majority of those giving possession retained debts after the sale of their property. The paper concludes that together, the critique of home ownership and the retention of debts, will increase both the short and medium term demand for rental housing. Key themes: Mortgage possessions; characteristics of mortgagors and mortgage lending; means of managing arrears
55
Type of research Source of data Sample
Quantitative and qualitative Mortgage records of borrowers in possession, mortgagors facing possession or in possession Random sample of 307 records of mortgagors in possession from one national lender. Mortgagors in or facing possession purposively accessed via lender, local authority and National Debtline. National 1991-1992 Sample from one lender, but as a national lender likely to be a good guide to national picture Descriptive and cross-tabular statistics of administrative data; thematic analysis of qualitative data. Major building society
Geography Date of study Stated limitations Method of data analysis Sponsor
56
Ford, J. (1994), Problematic Home Ownership: the management, experience and consequences of arrears and possessions in a depressed housing market, Loughborough: Loughborough University.
Objectives Following the growth in arrears in the early 1990s, this study examined two major consequences of that growth. First, the nature and outcomes of the administrative and judicial management of arrears for households and second, the social, economic and housing consequences of possession for borrowers, local housing providers and for attitudes to home ownership. The study used three sources of data; households in arrears (drawn from a survey of owner-occupiers in Bristol and Luton), households in Luton and Bristol who had experienced possession (accessed through the local authorities, advice agencies and respondent contacts) and county court records on actions for possession in Luton and Bristol. Data were collected in 1993. Findings An analysis of the administrative arrears management process indicated that approximately a third had access to some safety-net provision, a third were making concessionary payments while a third had failed to reach repayment agreements with their lenders. Agreements were constrained by poor communication, and lenders’ attitudes to borrowers. A minority of arrears cases were treated judicially but the proportion of cases granted a suspended order increased in the early 1990s. The changing balance between outright and suspended orders was influenced by the defendants’ representations to the courts; the size of arrears; the length of arrears; the previous payment pattern; and the exercise of judicial discretion. Overall, the evidence suggested that the judicial processes prevented possession in some cases, but more frequently only delayed possession. Qualitative analysis of 16 post-possession cases showed the extensive degree of housing mobility that characterised the search for suitable, permanent re-housing. There was also evidence of persistent financial disadvantage. The social consequences of possession included a loss of personal identity and status, concerns about the impact of possession on children and social isolation. Key themes: Mortgage arrears; mortgage possession; means of managing arrears
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor Qualitative and Quantitative County Court records, Households in possession Census of Court records Luton and Bristol 1994 Non-representative interviews Descriptive statistics, thematic analysis of qualitative data Joseph Rowntree Foundation
57
Ford, J. (1995) Which Way Out? Borrowers with long-term mortgage arrears, London: Shelter
Objectives This study looked at households with long term mortgage arrears. It considered the social and material consequences of long term mortgage arrears; the costs (to borrowers) of attempts to remain as home owners; the likelihood of their recovering their arrears and the options available to them and their chances of success. The study involved detailed analysis of five case studies (all of households in arrears for two or more years) and used existing data. Findings The report starts by outlining the recent trends in mortgage arrears and possessions (going up to 1994). It then outlines the five case studies in detail to highlight the experience of living with arrears, the strategies adopted by the borrowers to stay in home ownership and the complexity of the relationship between employment and recovery from arrears. The second part of the report uses the case study material and other published data to review the likely outcomes for those with long term arrears. In particular it evaluates the role of employment in recovering arrears and the wages needed to clear arrears. It found that, for borrowers in serious arrears, there was little prospect of them finding employment at a wage that would enable them to pay their regular mortgage payment and something toward the arrears. In fact, the report notes that borrowers with arrears may have to find as much as a third extra for arrears, thus increasing their reservation wage. The report also noted that: the growth of the flexible labour market and the absence of safety net provision for mortgagors in low paid work increased the likelihood of arrears; and that long term mortgage arrears impoverished individuals and families (materially and socially). It proposed the following policy initiatives: preventative measures such as the introduction of a mortgage benefit scheme; policies to aid arrears recovery like more flexible repayment schemes; more effective responses to borrowers who lose their homes including the reconsideration of mortgage rescue schemes and the provision of more effective advice and support for homeless households. Key themes: Mortgage arrears; means of managing arrears; information and advice; safety-nets and work disincentives
58
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Qualitative Households in arrears Clients of advice agencies Midlands 1995 Non-representative respondents Thematic analysis of qualitative data Shelter
59
Ford, J. (1998) Risks: home ownership and job insecurity, London: Shelter.
Objectives The main aims of this report were to: • • • • update the existing account of the nature and extent of risks to home owners at the end of the 1990s (particularly the labour market); understand how individuals and households respond to such risks; identify the range of housing consequences stemming from labour market risks; identify a policy agenda.
The report draws together published data; carries out some new analysis of available national surveys and a small number of case studies. The work took place in 1998. Findings The report discusses the growth of unstable employment and presents information which indicates that approaching one in four mortgagor households are headed by someone in a ‘flexible’ labour market position and so potentially at risk. Available evidence suggests that individuals themselves are more risk aware, particularly those subject to such risks. The report identifies potential strategies available to mortgagors for dealing with the growing risks in the labour market. These include private insurance protection against the loss of income through accident, sickness or unemployment; making additional mortgage payments against the possibility of future loss in earnings; and adapting personal circumstances, for instance, a reluctance to move, or start a family, or trading down or leaving the housing market altogether. The report considers MPPI in detail. It points out that there has been a lively debate on the reasons for the low take-up of MPPI. It does not cover those issues in detail but looks at whether policies do respond to the labour market risks associated with temporary work or self-employment, low pay or reduction in income. It notes that initially access to this type of insurance was restricted, but is now more widely available. At the time the report was prepared, obstacles to greater coverage were concentrated in the stringent claiming criteria. The section concludes that there is a poor match between those who take out such insurance and those that need it, and that take-up is not related to exposure to labour market risk. The report includes detailed case studies which discuss attitudes to labour market risk and home ownership. They also include experience of claiming on an MPPI; the process of decision-making around claiming/remaining on ISMI and problems encountered during the administration of the benefit (including the impact of a shortfall and lack of knowledge about what was and was not covered by the benefit); the impact of contract work. The case studies are used as illustrations of how the strategies adopted for dealing with risk vary from household to household.
60
The review concludes by discussing the continuing concerns and potential directions for MPPI, including the arguments for and against compulsion, and the implications of the current operation of the state safety-net. Finally, Shelter outlines some policy recommendations on: tenure diversification; effective safety-net provision; flexible tenure; knowledgeable consumers/regulated providers. Key themes: ISMI; safety-nets and work disincentives; mortgage arrears; mortgage possessions
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor Secondary analysis; qualitative Labour Force Survey, Survey of English Housing, British Household Panel Survey N/A England (SEH) Britain (LFS and BHPS) 1997 Non-representative qualitative interviews Descriptive statistical analysis; thematic analysis of qualitative data Shelter
61
Ford, J., and Burrows, R. (1999) ‘The costs of unsustainable home ownership in Britain’, Journal of Social Policy, 28, 2, pp 305-330
Objectives This paper details the mechanisms by which the growth in mortgage arrears and possessions came about and develops a framework for examining the overall costs of this growth for society as a whole. It then considers the ways in which a focus on arrears and possessions can contribute to a number of social policy debates.
Findings The whole is illustrated by drawing together the evidence on one set of costs (financial ) for one set of actors (borrowers). The paper concludes that, with just this one example, the financial losses incurred in possessions have a considerable impact in terms of the number of people affected, the sums of money involved and the resulting (and continuing) social disadvantages that follow possession. Not only are immediate life chances affected, but expectations of an asset on which to draw in later life or to pass on to future generations have been disappointed. The paper considers whether arrears and possessions are ‘yesterday’s problem’ or a permanent feature of contemporary life. The authors conclude that the evidence points to the latter, though the precise incidence will rise and fall in line with the economic cycle. The paper discusses the implications of the issues raised and divides them into two broad, interrelated, streams. First, the question of whether policy should continue to support the current tenure balance or place emphasis on the development of, for example, an adequate private rented sector. Second, the ways in which policy could support mortgagors in home ownership and/or cushion those who have lost their homes from the impact of repossession. The paper draws on data available from the Survey of English Housing and various other research studies on owner occupation, arrears and possessions. Key themes: Mortgage arrears; mortgage possessions
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Secondary Quantitative and Qualitative Review Survey of English Housing (1994/5); Ford et al 1995 N/A England 1998 Statistical and thematic review None
62
Ford, J. and Griffith, A. (1994) ‘Preventing mortgage arrears: A shrinking safety-net’, Benefits, April
Objectives In the absence of official data on the number of ISMI claimants in arrears, four existing studies were used to draw together the evidence, for the early 1990s, on the effectiveness of ISMI in preventing mortgage arrears and possession. The four studies were the Social Security Policy Inspectorate (1984); Ford and Wilcox, (1992), Ford, (1993) and Griffith (1993-unpublished M.A. Dissertation). Findings Three studies offered estimates of the number of ISMI claimants in arrears at 13 per cent, 20 per cent and 17 per cent respectively. Together, they indicate that ‘for those who fall within its remit, the safety-net operates effectively in preventing arrears in around four-fifths of cases’. Arrears amongst the fifth of claimants are primarily the consequence of the structure of ISMI and a range of direct and indirect processes. Amongst those in arrears, shortfall payments affected up to a third, primarily consequent on deductions for additional loans. Arrears were also associated with the payment of interest on arrears, communication and administration problems between DSS the lender and the mortgagor, the rules for repeat claims and more indirectly, the growth in endowment mortgages. Key themes: Mortgage arrears; ISMI
Type of research Source of data Sample
Secondary analysis Four studies, 1984, 1992, 1993a, 1993b. Examination of 300 live claims (1984); random sample of 302 mortgagors four or more months in arrears (1992); sample survey of mortgage lenders (1993a), qualitative assessment of 42 arrears cases brought to a law centre (1993b) National survey; local case studies 1994 Necessity to estimate national level data from small sample sizes or more local data Collation and assessment of existing findings to address a small number of questions. None
Geography Date of study Stated limitations Method of data analysis Sponsor
63
Ford, J. and Kempson, E. (1997) Bridging the Gap? Safety-nets for mortgage borrowers, York: University of York, Centre for Housing Policy Objectives This research was designed to assess the impact of the changes to the social security safety net for mortgagors and explore the main factors influencing the take-up and development of MPPI and MPPI policies. Information was collected on the perspectives and responses of borrowers, lenders and insurers. Both quantitative and qualitative research methods were involved. The practical working of mortgage payments protections policies were not explored in any detail as it was felt that such an investigation should wait until more time had elapsed from the ISMI changes implemented in October 1995. The research was carried out in 1996. Findings The research found that borrowers were equivocal about the State help that is/should be available for unemployed mortgagors for a number of reasons. These included: negative feelings about claiming benefits and the process of claiming; low expectations of the provision; poor levels of knowledge and understanding of ISMI. Lenders responded to the ISMI changes by tightening lending policies. Lenders had expected the ISMI changes to lead to increased levels of arrears and possessions but at the time of the research this had not happened. The research discussed the development of MPPI products and noted that there was little innovation in response to the ISMI changes. More were able to obtain MPPI, but coverage was still not universal - those most in need of the protection were most likely to face difficulties with respect to eligibility. About one in five of the borrowers who took part in the research had an MPPI policy (data from lenders gave a proportion of one in six). Take-up was higher among borrowers who had taken out their mortgage after October 1995. However, the research found that a third of ALL borrowers had no obvious way of covering their mortgage payments should they lose all income (i.e. no insurance, no savings, no partner in paid employment). Of those likely to be eligible for ISMI, three out of four had no private insurance, so could find themselves with no assistance toward their mortgage payments for as long as nine months. The research discussed the implications of this and suggested that owner occupation has to be seen as more risky after the 1995 changes. A number of characteristics correlated significantly with possession of an MPPI (male, single, under 35; mortgage after October 1995 with 20 or more years outstanding; owing £25,000 £50,000; with other insurance cover and a positive attitude toward insurance). Two main factors explained take-up: the date the mortgage was taken out and the borrower’s attitude toward insurance (risk averse, insurance averse or pragmatic), irrespective of actual risk. Borrowers had poor knowledge of the cover provided by policies. Reasons for claiming split evenly between unemployment and sickness/accident. About 80 per cent of claims were accepted, and payments lasted for an average of six months. Reasons for rejection tended to
64
be ineligibility through a pre-existing medical condition or job insecurity or because the claim had been submitted too soon. The research concluded that the October 1995 changes to ISMI had, initially, had less effect than anticipated. Borrowers’ ignorance of the ISMI provisions was thought to be a major reason for this lack of impact - they were neither more likely to take out MPPI nor more cautious about taking out new mortgages or re-mortgaging. Contrary to expectations, arrears and possessions had not increased as a result of the changes but lenders attributed this to a number of factors (economic recession, more cautious lending, borrowers with low levels of risk). Some lenders described the changes as a ‘time bomb ticking away’. The research concluded that the private market for MPPI had only developed slowly, and that, as a result the risk gap that had to be closed was still substantial. Key themes: ISMI; MPPI
Type of Research Source of data Quantitative and Qualitative Lenders: depth interviews and postal survey Borrowers: depth interviews and postal survey Insurers: depth interviews Lenders: depth interviews: proposed ten key lenders (purposively selected range by number of borrowers); achieved nine (often two or three interviews per lender); postal survey: 42 issued, 28 replied. 66% response rate (sample was all large lenders and a random selection of smaller borrowers) Borrowers: depth interviews: 31. Purposive sample of borrowers obtained from one national lender and one regional lender (selected so gave mix of with and without MPPI and date of mortgage) postal survey: random sample from three large lenders. 1,940 issued, 933 achieved. 48% response rate Insurers: five depth interviews, with insurers covering 75% of the market National Summer and autumn 1996 Borrowers postal survey had a low response rate. Attributed to: out-of-date information (included some without current mortgages) • inability to target reminders and use of only one reminder, rather than two (both prevented by • costs and logistics) Post October 1995 mortgagors were over-sampled to ensure the survey included a significant proportion with MPPI, and the data then weighted to restore the balance of pre and post October 1995 mortgages. Method of data analysis Sponsor Statistical including multivariate analysis. Thematic grid analysis of qualitative data. Joseph Rowntree Foundation
Sample
Geography Date of study Stated limitations
65
Ford , J. and Wilcox, S. (1992) Evaluating the Initiatives: mortgage arrears and possessions, York: Joseph Rowntree Foundation
Objectives The aims of this research were to: • • • identify and analyse the range of circumstances of households in mortgage arrears and facing potential possession; evaluate the potential impact of a number of initiatives introduced in response to the then rapid rise in arrears and possessions; to assess the potential impact of a mortgage benefit.
The research involved a survey of borrowers with arrears and took place in 1992. Findings The report describes the main household characteristics and circumstances of those in arrears. In the majority of such households, at least one adult was in paid employment. Mortgages tended to be relatively long-standing, often extended, and those in arrears divided evenly between first time buyers and previous buyers. The report also describes the mean monthly mortgage payment, and the amounts owed in arrears. It describes levels of income, receipt of benefits, the type and level of other financial commitments and whether or not there were arrears on other items. A third of households NOT in receipt of benefit had an income level below the level of Income Support allowances; a fifth had mortgage interest costs above their net income; and 16 per cent had net incomes after mortgage costs of £100 pw or less. Additionally, arrears on commercial loans and other commitments (poll tax, water rates and so on) were widespread. Most attributed their arrears to labour market changes and the recession. Although more than half were in work when interviewed, a substantial proportion had experienced periods of unemployment and under-employment over the previous three years. The report discusses the initiatives announced in December 1991 that resulted from discussions between the Government and mortgage lenders. These were: • • • • mortgage interest direct for Income Support claimants; mortgage rescue schemes involving owners becoming tenants or shared owners; the use of repossessed properties by housing associations to rent to homeless households; improved arrears management.
On mortgage interest direct, the research found that shortfalls in the mortgage interest paid through Income Support were the reason for arrears still continuing for those in receipt of Income Support. However, borrowers were in favour of direct payment of interest, to ease administration and to assist with budgeting.
66
There was widespread misunderstanding of mortgage rental and shared ownership schemes and the evidence suggested that such schemes would only work as a ‘last resort’. On arrears management, respondents were critical of the initial management process and communication (remote, impersonal, slow). Few were given any advice and information. However, borrowers still found the lender to be helpful overall. The research concluded that improved take-up of Family Credit plus a Mortgage Benefit scheme would ensure that eight out of ten households would be able to pay their mortgage and be left with disposable income above Income Support levels, effectively reducing the employment trap for low paid working home buyers. The research discusses the costs to the Exchequer of such a scheme and concludes that it was one of the best ways to limit possession among those in low income employment. Key themes: Mortgage arrears; ISMI; means of managing arrears
Type of Research Source of data Sample
Quantitative survey, modelling Borrowers in arrears of more than four months duration Random sample drawn from one large building society, consisting of: post code districts with more than 40 borrowers in arrears 575 issued addresses; 51 were ineligible for administrative reasons; 66 were unoccupied 302 achieved interviews, 66% response rate National 1992 Authors note that a comparison between borrowers in arrears and those not in arrears would have been helpful because of the nature of some of the initiatives. While lender agreed to provide such a sample, the timing meant that it was not possible to make the comparison. Descriptive statistics and modelling Joseph Rowntree Foundation
Geography Date of study Stated limitations
Method of data analysis Sponsor
67
Ford, J., Kempson, E. and Wilson, M, (1995), Mortgage Arrears and Possessions : perspectives from borrowers, lenders and the courts, London: HMSO
Objectives The research was the first large scale study of mortgage arrears and possessions in England. It examined the characteristics of those in arrears, the reasons for arrears, the experiences of those who defaulted at different levels compared to all mortgage borrowers. The research examined the policies and practices of lenders for making loans, recovering arrears and dealing with possession and voluntary surrenders, and the potential and problems of mortgage rescue. Information was gathered from a random sample survey of 362 borrowers in arrears and 111 households who had experienced possession. A random Omnibus survey of 646 mortgagors not in arrears was undertaken. Data on lenders came from a postal survey, interviews in 15 organisations and three case studies. District judges in 15 county courts were interviewed. Findings Evidence on the incidence of arrears and movements in and out of arrears noted that one in 20 households had experienced arrears between 1990 and December 1993, some on a number of occasions. The main reasons for arrears were redundancy, loss of earnings while in work and small business failure. Financial over-commitment was not as significant a reason for arrears as it had been at the end of the 1980s. Borrowers in arrears were drawn from throughout the socio-economic spectrum. The evidence from lenders indicated that they were slow to respond to arrears only developing more appropriate policies and practices post 1992 (after possessions had peaked). Lenders showed considerable variation in policies and practices. Consequently, borrowers in similar circumstances, but with different lenders, were treated differently. There was heavy reliance on negotiating agreements to repay with borrowers. Only a minority of borrowers were summonsed in court and in the early 1990s, compulsory possession predominated. District judges exercised discretion with respect to suspended orders and longer repayment periods more frequently than earlier studies had shown. The majority of borrowers in arrears had no access to either public or private safety-nets. Take up of MPPI was around 14 per cent and only a minority of those in receipt of ISMI received full interest payments. Only eight per cent of lenders had offered a mortgage rescue scheme with less than 700 borrowers converting to full tenancies or shared ownership. Moves out of owner occupation were usually into temporary accommodation with just a quarter of those in possession housed immediately by the local authority, typically only when a possession order had been granted. Lenders lacked any consensus for managing residual debt. Key themes: Characteristics of mortgagors and mortgage lending; mortgage arrears; mortgage possessions; means of managing arrears; ISMI; MPPI.
68
Type of research Source of data
Quantitative and Qualitative Sift via omnibus surveys for mortgagors, postal survey of lenders, case studies with lenders, interviews with District Judges Random sample of borrowers with and without arrears England 1993/94 The response rates for arrears and possessions were low Statistical and thematic analysis of qualitative data Department of the Environment
Sample Geography Date of study Stated limitations Method of data analysis Sponsor
69
Ford, J., Kempson, E. and England, J. (1996) Into Work? The Impact of Housing Costs and the Benefit System on People’s Decision to Work, York: Joseph Rowntree Foundation
Objectives The objectives of this research were to: • • identify the ways in which households made decisions about employment and the range of factors influencing those decisions; assess the extent to which higher rents and Housing Benefit for tenants and ISMI for mortgagors constitute a work disincentive.
The research was qualitative and involved interviews with 44 individuals in 40 households. It took place in 1995. Findings Three descriptive models of the approach to decision making around the issue of benefits, wages and work disincentives were identified. Respondents either: • • • made a better-off calculation and acted in accordance with it; or, made a better-off calculation but then overrode it; or, did not undertake a better-off calculation at all.
The approach adopted varied according to tenure. Owner-occupiers were less likely to make a better-off calculation than renters. Where owners did make a better-off calculation, they were more likely to override it. Those who overrode their better-off calculation were more likely to be employed. The report considered the factors influencing decision-making, grouped them into three broad areas (commitment to work; financial factors; and, attitudes to, and knowledge of, benefits), and examined their interaction with the better-off calculations and the decision made on whether to work. Overall, the strength of the commitment to work provided the main impetus to employment and the response to benefits was made within the context of this commitment. For owner-occupiers, therefore, losing ISMI on taking paid employment was outweighed by their desire to work but the report suggested that the changes to ISMI might deter mortgagors from taking temporary work or jobs which did not appear to be secure. The research concluded that the traditional economic model of rational decision-making did not always prevail and stressed the importance of social and attitudinal factors in structuring and informing decisions.
70
A number of policy implications were identified; recommendations on the availability and quality of advice for those considering a return to employment, the development of take-up campaigns for in-work benefits, and an ‘Extended Payments for ISMI’ Scheme, to be administered in a similar way to that for Housing Benefit. Key themes: Safety-nets and work incentives; ISMI
Type of Research Source of data
Qualitative Adults of working age: employed two years or less and with income at or below set levels unemployed three years or less Identified via doorstep screen questionnaire Quota sample of 40 individuals (20 owner-occupiers, 20 tenants, mix of household types, low income) Achieved 40 households (16 owner-occupiers, 18 social renters, 6 private renters; mix of household types); interviewed 44 individuals Luton/Hemel Hempstead and Coventry Fieldwork July – October 1995 Gap in fieldwork (August 95) as non-contact rate for screen interviews was very high. Once eligible respondents were found, refusal rate was negligible Thematic analysis of qualitative data Joseph Rowntree Foundation
Sample
Geography Date of study Stated limitations
Method of data analysis Sponsor
71
Foster, S. (1992) Mortgage Rescue: what does it add up to? London: Shelter
Objectives This study provides an overview of the extent and causes of arrears and possessions in the period up to 1992 and discusses the potential impact of a range of measures taken in 1992 designed to stem possessions. These include: an evaluation of a number of claims that paying Income Support Mortgage Interest (ISMI) directly to lenders will reduce possessions and; consideration of ‘in-house’ mortgage rescue and mortgage rescue involving the transfer of property to housing associations. Findings Following the review of the available evidence the report concluded that the number of households likely to be assisted to avoid possession as a result of ISMI direct in 1992 was 10,000. A further 14,000 were estimated as able to be helped by mortgage rescue. The need for an in-work mortgage benefit was identified. The report is based on secondary data sources and some primary data collection on mortgage rescue schemes in 1991/2. Some simple costings are developed based on information about local rents, housing benefit levels etc. indicating the likely savings to households from mortgage rescue. Key themes: Mortgage possessions; means of managing arrears
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Secondary analysis of published reports Published sources N/A Britain 1992 Some reliance on untested assumptions Reviewing, developing simple estimates Shelter
72
Green, H., Bumpstead, R., Thomas, M. and Grove, J. (1999) ‘Trends in Tenure, 19811997/8’, Housing In England 1997/98, London: The Stationery Office
Objectives A report of the 1997/98 Survey of English Housing (SEH) undertaken for the Department of the Environment, Transport and the Regions. The SEH is an annual, continuous survey of approximately 20,000 heads of households in England undertaken between April and March of the following year. The results are presented as grossed up estimates. It aims to identify housing trends. Findings While Housing in England includes chapters on all tenures, and on a range of housing processes, Chapter 3 presents key data on owner-occupiers with respect to mortgage borrowing, mortgage arrears and possessions. In 1997/98, the module of questions on mortgage arrears indicates that 2.9 per cent of mortgagors indicated they were in arrears. The figure rose to 7.4 per cent on probing for a set of circumstances that research has shown lead people to overlook or not recognise arrears. These are situations where people are in receipt of ISMI, making (agreed) reduced payments to lenders, or are currently making full payments but have old payments outstanding. In addition the survey reports the proportion of mortgagors facing payment difficulties but not in arrears (12 per cent). Two-thirds of those in arrears reported reduced income through job loss or reductions in earnings. Other reasons included changes in household composition (25 per cent) and increased expenditure on mortgages and other payments (26 per cent). There was a decrease in the impact of unemployment compared to 1996/97. The groups most likely to be in arrears or facing difficulties were: lone parents with dependent children (42 per cent); households headed by an unemployed person (42 per cent); households with a gross joint income of under £200 per week (40 per cent) and households in London (20 per cent) and the South West (19 per cent). In 1997/98, 420, 000 householders reported that, at some time in the past, someone now in their household had given up a home because of difficulties paying the mortgage. The current characteristics of householders that had experienced possession showed that they were disproportionately younger, living in households with dependent children, likely to be unemployed or economically inactive but under retirement age. The proportion of households who had experienced possession was relatively high in the South West of the country. Overall, only a fifth of possessions had been compulsory. Most householders had left voluntarily either with arrears or moving to avoid arrears. Key themes: Mortgage arrears; mortgage possessions; characteristics of mortgagors and mortgage lending
73
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Statistical, modelling Households Representative sample of approximately 20,000 heads of household including 8,000 mortgagor heads of household England 1997/98 Discussion of sampling errors and grossing up Statistical Department of Transport, Environment and the Regions
74
Hart, P. (1998) From the Deputy Ombudsman’s Casebook, London: Office of the Insurance Ombudsman
Objectives This article focusses on payment protection insurance (sometimes referred to as creditor insurance) for personal loans, including mortgages. It identifies what the public expects from such insurance and provides a short summary of the ways in which borrowers expectations are disappointed and why. Findings Difficulties include issues of polices being offered to those excluded, the duration of the cover, complex wording of policies, limited cover etc. A number of these difficulties stem from particular features of the market. For example, the lack of a proposal form means underwriting is carried out at the point of claim; there is not a ‘true’ market and, policies are not sold by staff who are experts in insurance. While the Association of British Insurers has made steps to address a number of these problems, the article concludes that they have not been completely eradicated. Twenty-one cases from amongst those adjudicated by the Ombudsman’s Office are presented as supporting evidence of continuing difficulties. Not all refer to MPPI but nevertheless illustrate the principles and difficulties that apply to all loan protection insurance. Key themes: MPPI
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Review of administrative case load Cases adjudicated by Ombudsman N/A Britain 1998 Potential bias in those who complain Review None
75
Hogarth, T., Elias, P. and Ford, J. (1996) Mortgages, Families and Jobs, Warwick: Institute for Employment Research.
Objectives The research provides an analysis of changes in social class structure, household composition, the economic activity of household members and housing tenure as recorded over the period 1984-1991. A key issue concerns the changing pattern of risk, primarily from unemployment, in the context of the changing patterns of both owner occupation and household structure. The analysis is based on data from the 1984 and 1991 Labour Force Surveys. Findings The key findings were that: between 1984 and 1991, mortgagors increased as a proportion of all social classes with an expansion of mortgages amongst households whose heads have a relatively high risk of unemployment (but over the period, there was little evidence of a greater risk of unemployment amongst ‘traditional’ owners); household structures changed with a growth of single adult households, one parent households and couples without children, with a growing proportion of these households comprised of younger people; rising female employment resulted in more dual earner households and; heads of household were more likely to be unemployed and their partner not working, regardless of social class or household structure, if they were council tenants. The researchers cautiously suggest the hypothesis that holding a mortgage provides an employment incentive due to the absence of in-work assistance to low-paid mortgagors and the partial payments for many unemployed mortgagors under ISMI. Key themes: Characteristics of mortgagors and mortgage lending; safety-nets and work incentives
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Secondary analysis of quantitative data Labour Force Survey Representative sample Great Britain 1994 Statistical Joseph Rowntree Foundation
76
Holmans, A. (1997) ‘UK housing finance: past changes, the present predicament and future sustainability’ in P. Williams (ed), Directions in Housing Policy, London: Paul Chapman Publishing, pp 175-199.
Objectives The article draws together a wide range of existing data to examine the financial consequences of providing decent, sustainable housing for households in the rented and home ownership sectors. Three issues are reviewed: the rise in households without the means to afford decent housing; the growth in public and private expenditure for housing costs and the growth in households with low or no income and/or temporary disruptions to income. A number of policy proposals for securing financially sustainable housing are developed. Findings (with respect to owner occupation only) A distinction is drawn between a single year risk for possession (under 1 per cent) and the currently one in five life time risk of possession. In response to these risks there is a need for a policy response to support mortgagors. The potential of a scheme whereby borrowers become eligible for assistance when their housing costs exceed a defined proportion of net income is outlined. A range of factors that need to be taken into account and accommodated are outlined. The similarities of the proposed scheme to the pattern of assistance available in France under the prets conventionnes is noted. The financing of such a scheme is outlined along with the proposal to provide assistance not only following income disruption or loss but following substantial interest rate rises. Key themes: Means of managing mortgage arrears; international comparisons
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Review Wide range of secondary sources N/A United Kingdom 1997 N/A Assembly and review of data None
77
Holmans, A. and Whitehead, C. (1999) ‘Who should buy MPPI?’ Housing Finance 42, pp 22-27. [The article draws on a research paper due to be published by the CML later in 1999: A Holmans and CME Whitehead, Why Mortgage Payment Insurance? Principles and Evidence].
Objectives The article concentrates on the factors which affect take-up of MPPI and the level take-up should reach in the long term. The discussion is centred around four main issues: the factors which determine whether it is appropriate for a borrower to take out MPPI; whether the price of MPPI is likely to be reduced to make the insurance more desirable; the benefits achieved from taking out the insurance; the extent to which different types of households might have access to different ways of achieving the same security. Findings The article concludes that the product is still in the early stages of development – taken up mainly by those with a predisposition to insurance and a greater understanding of the product. It notes that take-up, to date, has been unresponsive to price or chances of needing the product. Additionally, more information on individual risk assessments is required to help people make decisions on their own circumstances – the evidence suggests that even though the extent of coverage should rise, it should not get to 100 per cent as not all need it and so a significant proportion of households are better off ‘self-insuring’. Conversely, there is still concern that those most at risk are least likely to take out MPPI. However, should those most at risk start to take–up the insurance, the cost of premiums is likely to rise as the preponderance of risk averse households falls. The authors note that lenders needs to be careful not to over-sell the product to those unlikely to benefit directly, and that government needs to recognise that many gain little from MPPI when assessing the levels of take-up to be encouraged. They propose that overall costs should be brought down by further standardising the product and limiting the need for commission and that the performance indicators for the industry should not be levels of coverage but ensuring that an appropriate mix of households take out insurance. Key themes: MPPI
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Review and modelling CML Annual Housing Finance Survey and Family Resources Survey Representative sample of households Britain (FRS) UK (CML) 1998 Statistical Council of Mortgage Lenders
78
Jenkinson S. (1992) Repossessed: a fresh look at mortgage lending practice, York: Catholic Housing Aid Society
Objectives This booklet starts by noting that, in 1991, the rate of repossession had ‘never been higher’. At the date of writing, there were expected to be more than 80,000 repossessions of homes due to mortgage arrears, compared to 5,000 in 1981. However, there had been few signs of any policy initiatives from government or mortgage lenders in response and so the report explicitly set out to fill the gap. Findings It discusses the main economic and structural factors influencing the growth in arrears and possession and draws on 60 cases picked at random from the files of National Debtline. It uses these cases to identify the main factors arising from lenders’ practices which contribute to arrears and repossession. These were: excessive credit availability; inadequate consumer information on a wide range of issues (from commission payments to the impact of renegotiating loans and the dangers of voluntary possession); inadequate co-ordination of lenders’ policies and practices; inadequate monitoring of lenders; inadequate requirements upon lenders; and, market inefficiency. The report made a number of recommendations. These included: the establishment of a Home Buyers Council responsible for consumer protection and information as well as market efficiency within the mortgage industry; maximum requirements for lenders on making a loan (for example offering loans only up to 95 per cent of the purchase price); the provision of information on the costs and benefits of services (including commission payments); and advice on the desirability of the borrower obtaining independent advice. There should be a voluntary code for lenders on loan recovery procedures - the CML to be involved in the development of this code. Key themes: Mortgage possessions; means of managing arrears
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Qualitative Borrowers cases from National Debtline Random selection England 1991 Thematic analysis of qualitative data Catholic Housing Aid Society
79
Kemp. P., Oldman, C., Rugg, J. and Williams, T. (1994) The Effects of Benefit on Housing Decisions, Department of Social Security, London: HMSO
Objectives The aims of this research were to examine the extent to which benefit (Housing Benefit and ISMI) influenced two related housing decisions: the decision to move home and the choice of new accommodation. The research was qualitative and involved in-depth interviews with private tenants and homeowners in receipt of ISMI. Twenty-five home-owners took part in the research (eight of whom were recent movers). The research took place in 1992, and in two local authorities, one in north England, the other in London. Findings The report included some findings around attitudes to home ownership and ISMI. First, for owners (and renters) benefit was a constraint on moving and restricted possible housing options. Those who moved, did so because they had to (lender insistence, repossession). Benefit was, therefore, not just a constraint on moving but also a way of avoiding an unwanted or enforced move. Second, owner-occupiers were strongly attached to their homes and believed the receipt of ISMI was a temporary measure which made any long term strategy like moving unnecessary. Moving was not an easy option: respondent’s felt they would not be able to get a mortgage, or had insufficient equity to move or believed their properties would not sell. Not surprisingly, trading up was not an option while on benefit. Third, owner-occupiers tended to know little about benefit regulations. They lacked knowledge of the excessive housing costs regulations and the concept of ‘over-large’ accommodation. Owners also tended to have less contact with the benefit office than renters. Last, owner-occupiers did not consider whether or not the DSS would pay all of their mortgage interest payment; in all but four instances the ISMI payment fell short of the amount due (no cases were the result of excessive housing costs regulations). Those affected felt vulnerable; tended to be building up arrears; and were ‘looking anxiously toward’ their lenders. Key themes: ISMI
80
Type of Research Source of data Sample
Qualitative Owner-occupiers in receipt of ISMI Random samples in two locations. Those selected were sent questionnaires and given the opportunity to opt out. Of the 313 questionnaires returned only 13 had moved; so non-movers were contacted as well. 25 interviews were carried out. Northern city and London Fieldwork was carried out in 1992 Proved difficult to identify movers Thematic grid analysis Department of Social Security
Geography Date of study Stated limitations Method of data analysis Sponsor
81
Kempson, E., Bryson, A. and Rowlingson, K. (1994) Hard Times? How Poor Families Make Ends Meet, London: Policy Studies Institute
Objectives The main aim of this research was to study in detail the circumstances and experiences of families living on very low incomes. It looked at the approaches used to making ends meet, the extent to which the families had any choice of how to overcome their problems and the advantages and disadvantages of the various strategies used. The research was qualitative, involved 74 families, all with dependent children and took place in 1994. Twenty two of the families were home owners. Findings One chapter of the report considers work incentives. It discusses the interaction of help with housing costs through state benefit with views on whether working is financially worthwhile. Little evidence was found to suggest that eligibility for help with mortgage interest costs influenced an owner’s decision about whether or not to work. Owners, in fact, were found to work even though they would have been better-off not working, no matter how difficult it was to meet their mortgage payments and make ends meet. The findings did suggest that receipt of mortgage interest affected decisions made about working e.g. two of those receiving ISMI did not receive all their mortgage interest, had incurred arrears as a result and so looked to earn over their disregard (‘cash-in-hand’) to help make up the shortfall. Other families, with very low mortgages, did not claim ISMI but remortgaged or worked very long hours to maximise their income. Paying the mortgage was usually given top priority, in part because people felt it was important to maintain a roof, but also because the fear of repossession was greater than the fear of having the phone cut off. The report concluded that home owner’s desires to work and to remain in home ownership were both very strong, despite any financial implications involved. Problems managing the mortgage were caused by unstable low income employment and by the transition period between employment and unemployment (and vice versa). Arrears were avoided by overdrafts and remortgaging. Key themes: Safety-nets and work incentives
82
Type of Research Source of data Sample
Qualitative Previous national survey of low income families and electoral register 1. Potentially eligible families in four different areas were sent a self-completion questionnaire and then approached and asked if they were willing to participate (59 families took part) Addresses were selected at random from the electoral register in the catchment area for a credit and debt advice agency (15 families took part)
2.
Geography Date of study Stated limitations Method of data analysis Sponsor
Manchester, Wolverhampton, North London, East London, Birmingham 1994 Thematic grid analysis of qualitative data Joseph Rowntree Foundation
83
Kempson E., Ford, J. and Quilgars, D. (1999) Unsafe Safety-nets, York: University of York, Centre for Housing Policy Objectives This research project was designed to evaluate the two safety-nets available to mortgagors ISMI and MPPI. Its main aims were to: • • • examine the practical working of MPPI at all stages; consider the ways those without MPPI manage the ISMI gap; examine the effectiveness of ISMI once it becomes operative (for those with and without MPPI).
The research used both quantitative and qualitative methods of data collection and took place in 1998.
Findings The report starts by outlining the history of state safety-net provision and detailing the changes implemented in October 1995. On MPPI, it lists the major players in the market, the format of policies, and discusses market development over the previous five years. It outlines the number of claims made for both MPPI and ISMI and describes the characteristics of those claiming, as compared to mortgagors with and without MPPI policies. The research found that claiming both MPPI and ISMI was not easy. Each assumed potential claimants were well informed about the process of claiming and the way in which claims were paid once awarded. Both systems had stringent procedures - those for MPPI often very costly for the claimant (requiring e.g. more medical certificates). The report likened the process of claiming MPPI and ISMI to a ‘series of tests or hurdles’, designed to reduce fraud, but sometimes made more difficult by poor administrative procedures and inadequate communication. Around three out of ten claims made on MPPI policies were not successful. The research found that rejected claims were less straightforward than successful ones. Rejected claimants were more likely to live in households with no other wage earners and to have post-October 1995 mortgages. The report discusses the reasons for rejection in detail (often ignorance but sometimes over-zealous application of terms and conditions) and considers the working of complaints procedures (on the whole unsatisfactory). An unsuccessful MPPI claim did not automatically result in arrears. Many of those turned down kept up their mortgage payments - households with MPPI were more affluent than those without, tended to have someone still in full time work and/or savings. Sometimes lenders were willing to negotiate reduced payments. In the absence of such facilities, households
84
invariably fell into substantial arrears and were put under pressure to repay or face possession proceedings. Successful MPPI claims meant that mortgage arrears were avoided. Some did incur arrears during the initial wait period and/or when the MPPI claim expired. Arrears were also incurred where the amount insured did not cover the current monthly mortgage payment (including any endowment policy). When the MPPI payments ended, most managed to meet their payments. Where they were unable to, it was either due to doing lower paid work than previously, or problems relating to temporary jobs or continued unemployment. Those receiving ISMI tended to be pre-October 1995 borrowers and so subject to the shorter wait period. Most bridged that gap with savings, help from friends or family or social security benefits. However, just over four out of ten were unable to meet their payments in full and on time; a third on IS/JSA and in the wait period had already defaulted on their mortgage payments. Fewer IS/JSA and ISMI claimants were able to pay in full during the wait period than were MPPI claimants. Eight out of ten of the ISMI recipients had a shortfall on their payments. The main cause of the shortfall was the operation of the standard interest rate (introduced in October 1995). Most tried to cover the shortfall through help from friends and family and drawing on benefit payments, but slightly over half had arrears. Most were able to make an arrangement with their lender to repay. Borrowers in receipt of MPPI and eligible for ISMI found it difficult to manage the change over when MPPI ceased to apply. These problems often led to arrears. The research considered attitudes towards the safety-net provision for mortgagors. There was considerable support for existing state provision, even where people were critical of the way it operated in practice. Views on MPPI were conditioned by the process of claiming and/or the outcome of a claim - rejected or problematic claims often resulted in negative views. Most respondents felt a mixed model of provision was inevitable but supported increased state involvement (direct provision and/or running/policing insurance schemes). The research concludes by asserting that if the effectiveness of MPPI and ISMI was measured by the prevention of arrears, neither was performing very well. The research results suggest that a quarter of all borrowers have been left in a vulnerable position because of the changes to the state safety net. On policy, the research concludes that the debate needs to take a more radical approach, reassess the existing division of responsibility between private insurance and state provision and question whether the current sequencing of the two safety-nets is the most appropriate way of protecting homeowners who lose an earned income. Key themes: ISMI, MPPI, means of managing arrears
85
Type of Research Source of data Sample
Quantitative and Qualitative Borrowers claiming on an MPPI in the previous nine months IS/JSA recipients MPPI claimants: random sample obtained from two insurance companies 1000 issued: divided 70%/30% accepted to rejected claims response rate: 47% for successful claimants; 15% for rejected claimants 32 follow-up interviews IS/JSA recipients: random sample obtained from DSS 1,684 issued, sampled from February 1998 DSS quarterly statistics response rate: 41% for IS/JSA claimants; 50% for ISMI recipients follow-up interviews National 1998 Shortfall of rejected MPPI claimants – data from their questionnaires was analysed qualitatively. Low response attributed to possibility that the rejection meant they had to move but also could be because they were approached through the insurance company. Disproportionate number of borrowers with post-October 1995 mortgages responded; weighted back into their correct proportions for analysis. Statistical; thematic analysis of qualitative data Joseph Rowntree Foundation
Geography Date of study Stated limitations
Method of data analysis Sponsor
86
Keoghan, M. and Pryce, G. (1999) Private Unemployment Insurance for Mortgagors: responsiveness of take-up to changes in the benefits system in variations in the unemployment rate, Edinburgh: Scottish Homes
Objectives The purpose of this research was to test the ‘crowding out’ hypothesis specifically in the Scottish Context, and examine whether or not take-up of MPPI would increase as unemployment increased. ‘Crowding out’ assumes that the existence of ISMI inhibits the take-up of MPPI. The research used data from the Scottish House Condition Survey and the Family Resources Survey. Findings The research found that high risk groups (the most vulnerable categories of employees e.g. unskilled manual workers; own account non-professionals; one parent households; households with three children) were less likely to have insurance. It considers a number of factors such as: risk of unemployment; risk of ill-health; entitlement to state benefits; level of ISMI cover; loan to value ratio; cost of insurance, with the aim of identifying those which influenced the take-up of MPPI. Take-up was unresponsive to changes in ISMI cover, thus refuting the ‘crowding out’ hypothesis, but had some sensitivity to the probability of unemployment (a 10 per cent rise in unemployment would lead to a one per cent increase in take-up). Take-up was also unresponsive to changes in the cost of premiums, but influenced by the source of mortgage finance (banks having higher take-up than building societies) and more common among first time buyers and those buying under a Right to Buy scheme. The report goes on to draw conclusions, discuss implications and suggest alternatives. Key themes: MPPI
87
Type of Research Source of data Sample Geography Date of study Stated limitations
Quantitative, secondary analysis 1996 Scottish House Condition Survey; 1995/96 Family Resources Survey Approximately 7,400 mortgagor heads of household (SHCS); approximately 9,350 heads of mortgagor households (FRS) Scotland 1998/8 Full description given of the method of modelling, statistical analysis used, construction and description of variables Data available for variables such as employment status only available for head of household (not other household members) Statistical modelling Scottish Homes
Method of data analysis Sponsor
88
Maclennan, D., Meen, G.,Gibb, K. and Stephens, M. (1997) Fixed Commitments, Uncertain Incomes: sustainable owner-occupation and the economy, York: Joseph Rowntree Foundation
Objectives The research addresses three broad questions concerning the housing market in the late 1990s. Has the housing market behaviour of mortgagors and home owners changed following the early and mid 1990s housing market recession? Has structural change in the labour market affected the demand for home ownership and what policy actions are required to promote cyclical and long-term sustainability for the owner-occupied sector? The study drew on three surveys of home owners in Bristol and Glasgow dating from 198588, 1989-93 and 1992-1995. Data were also drawn from the Survey of Mortgage Lenders, and an ad hoc survey of mortgage lenders. Findings Since 1992, home buyers have made more considered choices in entering home ownership than in the late 1980s. More owner-occupiers now have insecure jobs and between 1992-95 a quarter of home owners were affected by a loss of a job and/or income. Debts incurred with respect to higher education are likely to increase caution and delay the point of entry for firsttime buyers. Mortgage lending has become more cautious than it was in the late 1980s. A minority of households in each place believed house purchase was riskier than previously. There was less aversion to renting. The research concludes that labour market changes best explain thee observed patterns. A third of owners had taken out MPPI. Take-up was not characterised by adverse selection. The research also provides a definition of ‘sustainable’ home ownership and models the likely level of sustainable home ownership to 2006. The research indicates average annual house price increases in the order of 6-7 per cent amid a modest rise in the home ownership share of tenure from 66.8 per cent to 69.1 per cent (excluding right-to-buy sales). National and regional forecasts are provided. The research discusses the policy implications of these findings and the potential shape of policy initiatives to secure sustainable home ownership.
Key themes: Characteristics of mortgagors and mortgage lending; mortgage arrears; MPPI
89
Type or research Source of data Sample
Quantitative, modelling Ad hoc household surveys, Survey of Mortgage Lenders, ad hoc survey of lenders Surveys of borrowers 1457 (1984), 886 (1989), 822 (1992). Data from the 5% survey of mortgage lending; qualitative interviews with 11 mortgage lenders. 1996/7 Bristol and Glasgow Statistical, modelling Joseph Rowntree Foundation
Date of study Geography Stated limitations Method of data analysis Sponsor
90
Merrick, W. (1998) Creditor Insurance; the Ombudsman’s view, Paper given at IBC’s Annual Creditor Insurance Conference, December 1998
Objectives This paper considers the types of issues arising around loan protection insurance and dealt with by the Insurance Ombudsman. It covers all such insurance, of which MPPI is just one part, and notes that loan protection cases form a ‘significant and troublesome sector’ for the Ombudsman. The majority of the paper is couched in general terms. Findings There is only one specific reference to MPPI, on the issue of under- or over-selling. The Ombudsman expressed concern about the perception that MPPI had been under-sold to date, with the Government urging the insurance industry to sell more. He points out the consequences (negative) of similar government urging on personal pensions. The Ombudsman felt the Government was pressing the insurance industry to commit to some sort of target for MPPI take-up, while at the same time not appearing to be concerned about code compliance. ‘Mystery shopping’ tests carried out on MPPI revealed that: only half of banks and building societies complied with the code when explaining cover; and only 60 per cent drew attention to the exclusions in the policy. The types of issues raised on loan protection in general were, first, the problem of consumer ignorance, this linked to the immaturity of the market and the absence of ‘reservoirs’ of knowledge and experience. Second, the way such protection is obtained – this is not a product where people make a choice between suppliers or the products, but is one where they make a decision between whether to have a policy or not as it tends to be a secondary purchase. Third, the paper discusses questions around the definitions of, and rationale for: pre-existing conditions; exclusion periods; voluntary unemployment. For the Ombudsman, these issues mean that people should be given more, not less, advice on eligibility conditions, exclusion clauses and value for money – all in all, advice on whether the product is appropriate to their circumstances. The Ombudsman felt that it was essential to identify those for whom such insurance was unsuitable. The paper concludes that greater compliance is required with the various codes, plus a substantial improvement in the way sales and claims are handled. Key themes: MPPI
91
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Review Cases submitted to Ombudsman Selected cases England 1998 Cases not representative of all cases Review None
92
Muelbauer, J. and Cameron, G, (1997), A Regional Analysis of Mortgage Possessions: causes, trends and future prospects, London: Council of Mortgage Lenders.
Objectives The research aimed to analyse the variation in mortgage possessions by region in order to understand the causes of the high rates of possession in the 1990s. A key concern was to identify the impact of economic and administrative processes. A second concern was to form a view about future prospects for the level of possessions. The analysis examined four sources of empirical evidence: national data on arrears and possessions; regional data on suspended possession orders from English county courts; the ratio of court orders to court actions and; the rate of court actions in relation to the expected pattern given the economic context. The ‘expectations’ of possession, given the economic fundamentals in each region, was modelled. The analytical objective was to confirm a national policy shift in responding to possessions across regions, notwithstanding the variations in economic fundamentals by regions and time. Findings Using the sources of evidence outlined above, the research found that by 1995, mortgage lenders were possessing at rates below those predicted on the basis of key economic relationships (unemployment, house price change, debt service ratio etc.). However, the rate varied depending on the ‘measure’ or dimension of possessions adopted but all were associated with a shortfall in possession. The authors point to a temporary softening of policy towards possession in the county courts starting in 1991 resulting in possessions being between 17 and 8 per cent below the level expected. The shift in policy was a result of intended changes to behaviour and the lag in setting up systems to deal with arrears and possessions. By 1995, possessions had returned closer to the levels expected by the economic fundamentals suggesting that the ‘softer’ policy was ‘decaying’. Predicting forward to 1997, the research indicates significant further falls in possession in 1997/8 for England as a whole, but with regional variations in the extent of the decline, providing there was no significant rise in interest rates. Key themes: Mortgage possessions
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor Modelling County Court statistics, housing market data N/A England and Wales 1996 Statistical Council of Mortgage Lenders
93
Munro, M. (1997) ‘Perceptions of uncertainty and risk in the housing market’ Housing Finance, 36, pp 17-22
Objectives The research, which formed part of the ESRC Programme of work on Economic Beliefs and Behaviour, aimed to understand the response of owner-occupiers to the risks and uncertainties of owner occupation as revealed by the housing market downturn in the early and mid 1990s. Findings Against the expectations of economic theory, the research indicated a complex relationship between borrowers’ perceptions of risks and uncertainty and their behaviour, with the former offering a poor guide to the latter. Home owners were conscious of the financial risks associated with home ownership and of the risks from the labour market. Substantial minorities of movers, stayers and first-time buyers had experienced unemployment and a minority had employment that was not permanent. These risks were often explicitly considered, but also largely accepted as inevitable. Housing decisions were more typically related to family and life-cycle events and housing moves based on these considerations were pursued. Only at that point were other risks considered. The inevitability of risk was encompassed within continuing support for owner occupation as the preferred tenure.
Key themes: Characteristics of mortgagors and mortgage lending
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Quantitative and qualitative Mortgagor households Quota sample of first-time buyers and moving households split between movers and non-movers Bristol and Glasgow 1995 Statistical; thematic analysis of qualitative data Economic and Social Research Council
94
Nettleton, S. and Burrows, R. (1998) ‘Mortgage debt, insecure home ownership and health: an exploratory analysis’ in M. Bartley, D. Blane and G. Davey-Smith (eds) The Sociology of Health Inequalities4$+56789/$:;<=>?@;;4$ 171-192. 4$+56789/$:;<=>?@;;4$pp 4$+56789/$:;<=>?@;;4$ Objectives This chapter explores the impact, if any, that insecurity among home owners has on their health and use of primary health care services. It looks particularly at whether mortgage indebtedness impacts on mental health (measured by a commonly used subjective scale, GHQ12). The analysis provides contextual detail of the development of home ownership in the UK since 1945 and the causes and distribution of mortgage arrears. The authors drew a sample of individuals from the British Household Panel Survey (BHPS), using five waves of the survey from 1991 to 1996. Between 15 per cent and 10 per cent of the weighted sample of 3,700 individuals were in arrears or difficulties with their mortgage across the period covered. Findings The authors used two measures of change in subjective well-being for each individual and compared their scores at the two points. Both multiple and logistic regression were used to control for the various factors thought likely to impact on well-being, including: income and income changes; physical health problems and changes in physical health problems; employment and changes in employment; and age. The analysis found that the onset of mortgage indebtedness affected adversely the subjective well-being of men and women. It increased the likelihood of men visiting their GPs and suggests that housing insecurity can impact on health. However, the associations were complex and it was impossible to disentangle whether poor health was caused by mortgage indebtedness or whether those in poor health were more likely to find themselves in difficulty. Their analysis suggested that mortgage indebtedness is one of a number of biographical changes and can, in fact, be conceptualised as a ‘biographical disruption’. As such, it has consequences for an individual’s sense of self, identity, self-esteem and social status. These are coupled with the ‘privatisation of risk’ and the changing perceptions of home ownership. The conclusion was that any rise in arrears is likely to generate unease among home owners. This in turn compounds feelings of insecurity generated by the substantial economic and employment restructuring that has been taking place.
Key themes: Mortgage arrears
95
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Quantitative British Household Panel Survey Individuals in BHPS 1991-1996 Britain 1998 Statistical None
96
Nettleton S., Burrows, R., England, J. and Seavers, J. (1999) Losing the Family Home: understanding the social consequences of mortgage repossession, York: Joseph Rowntree Foundation
Objectives This research examined the social consequences of mortgage repossession for families. It focussed on the experiences and perceptions of 30 families with children, all of whom were rehoused in social rented accommodation. The research was qualitative and took place in 1998. A total of 44 adults and 17 children were interviewed. Findings The experience of mortgage repossession and subsequent rehousing was immensely stressful, and compounded by protracted administrative processes and procedures of the lenders, the courts and local authorities. The processes of repossession impacted upon the lives of families for some time after the event. People felt they lacked control over events, despite their many efforts to find solutions to their problems, and all endured long periods of uncertainty. Repossession was often a route to long-term poverty and substantial debt. Women were especially vulnerable where the repossession was the result of relationship breakdown. The experience of repossession invoked extreme emotional responses and people described the profound sense of loss they felt at losing the family home. Parents and children often felt ashamed at their move from the status of ‘owner’ to ‘renter’. The experience of repossession had consequences in six main aspects of peoples’ lives: social status and identity; personal and family relationships; health and well-being; the quality of their lives; future hopes and aspirations; and, the lives of their children. The research found that the shortfalls in the ISMI payments from DSS led to arrears. Often lenders added interest to these arrears, so the amounts grew on a compound basis, and very quickly. The changes to the regulations introduced in 1995 had caused problems, exacerbated by the fact that the DSS and lenders did not seem to inform people about any shortfall or problems with the payments. Consequently borrowers were not always aware of the build up of debt. Finally, the report includes a chapter on the childrens’ reactions to, and experience of, repossession.
Key themes: Mortgage possessions; means of managing arrears; ISMI
97
Type of Research Source of data Sample
Qualitative Families who had been repossessed and rehoused in social rented accommodation Quota sample. Families were contacted in one of three ways, via: local authority homeless families units; housing associations and advice agencies. Fifty two families completed a short screen questionnaire, and 30 were selected as mirroring the known socio-demographic characteristics of households rehoused in the social rented sector after repossession. A total of 44 adults and 17 children (aged 7 to 18) took part. South-East and northern England Fieldwork: summer and early autumn 1998. None Thematic analysis of qualitative data Joseph Rowntree Foundation
Geography Date of study Stated limitations Method of data analysis Sponsor
98
Nixon, J., Hunter, C., Smith, Y. and Wishart, B. (1996) Housing Cases in County Courts, Bristol: The Policy Press
Objectives The research examined the way in which mortgage and rent possession cases were managed by county courts and the impact of different forms of representation. The research took place from September 1994 to February 1996. It involved an examination of 490 housing cases in 10 courts, half of which had duty desk schemes; case-study interviews with defendants and district judges; a national telephone survey of county courts and a postal questionnaire to the co-ordinators of court-based schemes. 203 of the 490 housing cases were mortgage actions. Findings The reasons for arrears were very similar for tenants and home owners. Both tenants and borrowers felt they had little influence over the causes of their arrears and so felt powerless to prevent the lender or landlord from taking legal action. Participation by defendants at hearings was low, but more borrowers attended or were represented than were tenants. Those who attended hearings or were represented had a lower risk of eviction. Borrowers were twice as likely as tenants to have outright possession orders made against them and five times as likely to have had possession warrants enforced in the long term. This was due to differences in the law applicable to borrowers and tenants, and differences in the strategies adopted by lenders and landlords. Social landlords perceived that there was a greater chance of receiving arrears if tenants remained in situ. A minority of defendants obtained advice prior to the court hearing. Those who attended court valued, and benefited from, the ‘on-the-spot’ advice provided by the duty desk schemes. The research found that the number of court-based advice schemes was increasing (though still low). Twice as many provided services to borrowers as to tenants. The exercise of judicial discretion was significant in determining the type of order made. Similar cases were not treated consistently. Borrowers’ cases were more likely to be dealt with privately in chambers, while tenants’ cases were held in open court. The report concludes with a number of recommendations on the working of the courts and the judiciary; lenders’ and landlords’ arrears management practices; and, court based advice schemes.
Key themes: Mortgage possessions; information and advice
99
Source of data
County court records Co-ordinators of duty desk schemes Defendants District judges A sample of all cases listed over a set period in 10 case study locations. 752 cases were listed and detailed information collected on 490 cases. 203 were mortgage cases. Postal survey of all co-ordinators of duty desk schemes – 50% response rate Depth interviews with 26 defendants and six district judges England Case study areas were: Watford, Oxford, Nottingham, Kingston-upon-Hull, Edmonton, Colchester, Southampton, Salford, Gateshead, Romford September 1994 to February 1996 None substantial: full detail of the methodology, definitions and reasons for actions are given in the report Descriptive statistics; analysis of case study data Joseph Rowntree Foundation
Sample
Geography
Date of study Stated limitations Method of data analysis Sponsor
100
Oldman, C. and Kemp, P. (1996) Income Support for Mortgage Interest: an assessment of current issues and future prospects, London: Council of Mortgage Lenders
Objectives This report considered past trends in ISMI payments and claimants and examined the likely impact of possible changes to the system. In the course of the work, the Government introduced the 1995 changes to ISMI, so the project was able to examine the detail of the changes and assess their impact. The report also considered ways in which the provision of help for low income home-owners could be improved. The research involved a questionnaire to the top 20 mortgage lenders in the UK, in-depth interviews with three representatives of the insurance industry, and secondary analysis of the qualitative interviews with home-owners carried out by Kemp(1994) et al for the study ‘The Effect of Benefit on Housing Decisions’. Findings The report summarises the development and structure of ISMI and outlines the impact of the 1995 changes. It reviews the cost of such help – pointing out the increase since the 1980s and the changing nature of the caseload, with more claimants classed as long term sick and disabled and elderly; fewer as unemployed. Overall, most claimants were long term. The research used data from ‘The Effects of Benefits on Housing Decisions’ to explore the experiences of living on ISMI. People were generally ignorant of the workings of ISMI and saw ISMI as a way of keeping a ‘roof over their head’. Claimants found it difficult to avoid arrears or build up debts on household bills. ISMI shortfalls contributed to arrears. Claimants did not have the opportunity to move house, nor were willing to do so, to reduce their housing and living costs. The study considered work disincentives and showed that the home-owners were highly committed to working even if they would be worse off. The likely impact of the ISMI changes were, in the short term, an increase in arrears. Lenders were looking to tighten procedures because of the reduction in the safety net. The report predicted an increase in the take-up of MPPI, and improvements in the quality of the product, but that significant numbers would still be excluded (price, inadequate cover, incorrect risk assessments). The report concluded that there was little prospect of MPPI becoming a fully comprehensive substitute for ISMI. Consequently, a form of provision was needed which was universal but targeted on lower income households, including those in low-paid work. Such provision needs to minimise public expenditure and, where possible, to rely on private rather than public insurance. The report finally concluded that a mortgage benefit scheme could meet these criteria. Key themes: ISMI; safety-nets and work incentives
101
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Review Official statistics; previous study N/A Great Britain 1996 Review of official statistics, re-analysis of qualitative data Council of Mortgage Lenders
102
Pryce, G. (1998) ‘ Income support for mortgage interest and the crowding out of mortgage payment protection insurance,’ Housing Finance, 39
Objectives This paper models the decision-making process involved in taking out an MPPI policy, using data obtained from a household survey in Bristol and Glasgow. It examines the responsiveness of MPPI take-up rates to cut backs in ISMI, testing the ‘crowding out’ thesis, that a less generous state safety-net would increase the incentive for mortgagors to take out MPPI policies and encourage the development of better and cheaper MPPI products. Findings The research developed a model for the insurance decision. The variables considered were: attitude to risk; perceived utility of the cover; coverage of job loss/absence through accident and sickness as well as unemployment; entitlement to benefits; knowledge and time horizon; household decision-making, and used logistic regression to test their significance for take-up. Knowledge of the changes and time horizon had no conclusive influence. Further analysis found that take-up was most likely to be influenced by the cost of premiums, though the influence was small (a 10% reduction in premiums would create a 5% increase in take-up). Take-up was unresponsive to cuts in ISMI and the benefits system. There was some limited responsiveness to improved cover and the risk of unemployment. The article concluded that the sluggish response to the ISMI cuts could have been anticipated and was not just due to falling unemployment. ISMI was not holding back the development of private finance. At the same time, the results also suggest the converse – that the reinstatement of ISMI would not lead to a fall in the take-up of MPPI.
Key themes: MPPI
Type of research Source of data Sample Geography Date of study Stated limitations Modelling, Secondary analysis survey of data Ad hoc household surveys undertaken as part of an ESRC project on housing decisions under conditions of uncertainty (Munro et al, 1995) Survey of 822 respondents. Sample sizes for regressions varied between 240 and 290 depending on which variables used. Glasgow and Bristol Data collection 1995; analysis 1998 The data set was chosen as most of the respondents in the sample were questioned before the 1995 changes, which allowed examination of what could have been anticipated by policy makers at the time Statistical analysis None
Method of data analysis Sponsor
103
Walker, R., Shaw, A. and Hull, L. (1995) ‘Responding to the risk of unemployment’ in M. Boleat (ed.) Risk, Insurance and Welfare, London: Association of British Insurers (ABI), pp 37-51. Objectives This article forms part of a collection addressing the potential of the private insurance market to provide social protection, presented as part of the ABI’s on-going discussion of the role of private insurance in welfare reform. It aims to assess the extent to which unemployment is a risk that can be insured in the private market. Findings A historical review notes the development and then demise of state unemployment insurance in Britain, the current heavy reliance on social assistance and concludes that unemployment has generally not proved to be a commercially insurable risk. Historically, major constraints on the development of the market were the high concentration of risk and so prohibitive premiums for some groups, the absence of risk and any requirement for insurance amongst others, an unpredictable and unstable rate of unemployment and the failure of risk to be independent of the actions of the insured. The article reviews the extent to which the emergence of a flexible labour market in the 1980s changes the ‘arithmetic of private insurance’ and concludes that it does not. Data are presented on the trends in unemployment, employment opportunities, the incidence and speed of re-employment for different occupational groups, allowing the risk of unemployment to be calculated for different social class groupings and different employment statuses. The paper concludes that there is a rising risk of unemployment for all but a risk gradient from professional to unskilled workers. Increasing risk is accompanied by falling levels of social assistance and limited personal financial resources. Private insurance has grown to fill the gap. The private insurance market for unemployment cover in 1995 is reviewed. Despite some growth, the severe limitations on the cover available at affordable cost are noted and shown to be poorly matched with the opportunities in the flexible labour market. The difficulties of responding to unemployment as an insurable risk in the private market are detailed. International comparisons suggest that this position is not unique to Britain, and identifies the range of ways in which other countries respond to meeting the rising risk of unemployment.
Key themes: MPPI; safety-nets and work incentives; international comparisons
104
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Policy review; secondary data analysis Published studies N/A Great Britain 1995 Review, statistical Association of British Insurers
105
Wilcox, S. and Ford, J. (1997) ‘At your own risk’, in S. Wilcox (ed.) Housing Finance Review, 1997-1998, York: Joseph Rowntree Foundation. Objectives This article reviews the October 1995 provisions for public assistance and private insurance to provide a safety net for home owners. It discusses the ‘post-1990s recession’ challenges to the home ownership market, including: • • • the social and economic diversification of the sector; trends in the labour market and their implications for sustainable home ownership; the potential consequences for home owners of the political commitment to low inflation.
Findings In the first six months following the October 1995 changes to the ISMI safety net, the numbers in receipt of ISMI had fallen by 58,000 - half of which could be directly attributed to the new regulations. Drawing on the findings of Bridging the Gap (Ford and Kempson, 1997), the paper discusses take-up of MPPI, outlining the various factors influencing take-up. The slow rate of growth in MPPI take-up implies a gradual and continuing increase in the numbers of home buyers not covered by either state or private safety net and so an increase in the number of households entering home ownership at their own risk. The article concludes with a discussion of the rationale for, and implications of, some alternative policy directions. These include ways of increasing the take-up of MPPI including; a reduction in premiums, or the introduction of a significant tax incentive or compulsion (as for car insurance). The article also considers whether all lenders should be required to offer cover on the basis of a general, not differentiated, pattern of MPPI premiums and discusses the feasibility of the introduction of a mortgage benefit scheme to provide cover for home owners caught in the employment trap. Key themes: MPPI; ISMI; state/private interface
106
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Review Official statistics and previous studies N/A Great Britain 1997 Review None
107
Wilcox, S. and Sutherland, H. (1997) Securing Home Ownership: producing an effective safety-net for home buyers, London: Council of Mortgage Lenders Objectives This report examines ways of developing more universal security for home-buyers unable to meet their mortgage commitments as a result of unemployment or health problems. Findings The report outlines current state and private provision for home-owners in difficulty, notes that its coverage is patchy, and indicates the rationale for an alternative approach to providing help for home owners. It includes a short discussion of work incentives and in-work housing benefits. The report considers some options for reform. The proposals put forward have three main dimensions: • lender flexibility in managing debt. [Instead of private or state provision for very short periods of unemployment or sickness, lenders should provide either an automatic or negotiated element of flexibility in repayment schedules. This approach builds on and formalises current good practice and aims to prevent short term arrears providing the basis for possession]; promotion of more comprehensive private insurance schemes. [A more universal approach to MPPI could be secured either through tax incentives or by regulations requiring compulsory cover. Such a universal approach would reduce unit costs, produce lower premiums and greater stability in the home-owner market]; the introduction of a mortgage benefit scheme. [Low income tenants receive help with their rent through Housing Benefit. There is an equitable case, reinforced by the reduction in Mortgage Interest Tax Relief, for such help to be available to homeowners. The help would be means-tested and cost about £300 million per year, taking into account savings through the reduction in repossessions].
•
•
The report discusses each of these in detail and includes the estimated costs of a mortgage rescue scheme. Key themes: MPPI; ISMI; safety-nets and work incentives; international comparisons
108
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Modelling Official statistics N/A Great Britain 1996 Discussion of a number of assumptions Statistical and modelling Council of Mortgage Lenders
109
Wilcox, S. and Williams, P. (1996) Coping with Mortgage Default : Lessons from the recession, London: Council of Mortgage Lenders Objectives This study examines the lessons that can be learned from the 1990s recession in order to secure sustainable home ownership. Findings The report discusses the main features of the housing market recession, the factors likely to influence future trends in arrears and possessions and the future economic and social trends likely to influence the home ownership sector (low inflation economy, labour market and demographic change, government policies). It reviews the availability of help with housing costs and potential directions of future help (government provision and private insurance). It also considers the existence and working of flexible repayment strategies (described as reducing the number of possessions where used and influencing the slow take-off of ‘mortgage to rent’ schemes). The report notes that the level of usage of mortgage to rent schemes has been lower than envisaged and discusses the factors influencing their use. The report notes the lack of ‘systematic information available about what happens to home buyers after possession’. It also discusses the effectiveness of the ways used to ‘dispose of’ homes taken into possession. The issues associated with persisting negative equity are considered along with the working of schemes to assist such households. The report concludes by outlining lessons from the housing market recession, including increased recognition of the risks in the home-ownership market. These risks require careful management through advice, policies and procedures, pricing and insurance, and the report stresses that for government ‘the question of an adequate safety-net for home owners must be a priority’. Key themes: Characteristics of mortgagors and mortgage lending; mortgage arrears; mortgage possessions; means of managing arrears; ISMI
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor Review Published studies, official statistics N/A Great Britain 1996 Review of secondary data sources Council of Mortgage Lenders
110
Williams, P. (1995 ) ‘A safety net for a changing market’, in S. Wilcox (ed) Housing Finance Review 1995/1996, York: Joseph Rowntree Foundation, pp 32-37. Objectives This article discusses the changes made to ISMI in 1995. It was written at the time the changes were being implemented and was concerned to assess (briefly) the background to the changes and consider what their impact might be. Findings The article reviews the changes taking place in the housing market in the mid 1990s (e.g. tighter lending criteria and narrower access to home ownership; delayed entry to home ownership; lower levels of transactions and a smaller market). It outlines levels of take-up of ISMI and average amounts claimed and describes the characteristics of ISMI claimants, noting that although the unemployed were the largest single category of claimants, they were not the majority of all claimants. The article suggested that the numbers of unemployed claimants would continue to fall and the proportions of lone parent, pensioner and disabled claimants rise. Consequently, the amount of government expenditure on ISMI was unlikely to reduce. The article discusses the impact the ISMI changes might have on the housing market (affecting confidence, particularly of first time buyers; increasing costs for existing buyers). It outlines the (1995) take-up of MPPI and discusses its potential coverage and applicability.
Key themes: Characteristics of mortgagors and mortgage lending; ISMI; MPPI
Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor
Review Official statistics and survey data N/A Great Britain 1995 Reviewing of secondary sources None
111
Williams. T., Hill, M. and Davis, R. (1999) Attitudes to the Welfare State and the Responses to Reform, Social Security Research Report No. 88, London: HMSO Objectives This report draws together the findings from a number of both qualitative and quantitative studies commissioned by the Department of Social Security in order to: • • explore the attitudes of the general public (including claimants) to the welfare state and the Government’s proposals for welfare reform in general; explore attitudes to state and private provision in a number of specific areas of provision including ISMI and MPPI.
Findings (with respect to support for housing in general and ISMI and MPPI in particular) Overall, studies showed that the public believed that the State had responsibility to assist with housing costs for vulnerable people. Such assistance should be targeted with the expectation that those able to work should be willing to do so. No distinction was drawn between renters and owner-occupiers with respect to state assistance. The purpose of housing support should be to keep people in their own homes. A specific consideration of MPPI and ISMI, based on 10 focus group discussions, indicated that there was little awareness of ISMI but a general assumption that where home owners were in difficulty ‘something would be done’. Those aware of ISMI (both claimants and nonclaimants) nevertheless had a poor understanding of how the benefit worked. Key misunderstandings related to the ISMI wait period, the full period of payment, and that ISMI covered capital as well as interest payments. Respondents believed that ISMI should be available where people had lost all income but were willing to work and for those in low paid and part-time work. Respondents judged that the ISMI wait period was punitive and a better alternative would be to restrict the payment period. A suggestion that ISMI might be repaid had only equivocal support. Owners were more aware of MPPI than renters, but still poorly informed. Those aware of MPPI, but without policies, believed it was cheaper than it was. A number of respondents who had taken MPPI had found it of no use to them. There was an acceptance of a role for private insurance in meeting mortgagors’ costs but for an MPPI style of insurance to be successful it would need to be Government backed, cheaper than commercial insurance, less restrictive and compulsory. Key themes: ISMI; MPPI
112
Review Type of research Source of data Sample Geography Date of study Stated limitations Method of data analysis Sponsor Existing qualitative and quantitative studies commissioned by DSS; newly commissioned work N/A United Kingdom 1999 Reviewing existing studies; analysis of primary data Department of Social Security
113
REVIEW OF DATA SOURCES
British Household Panel Survey (BHPS) This longitudinal survey was established in September 1991. It provides data on an initial sample of 5,500 households comprising over 10,000 individuals. Individuals remain part of the panel even if they leave their original household and form a new one. To date, eight ‘waves’ have been completed with 77 per cent of wave 1 respondents participating in wave 7. The aim of the survey is to enhance the study of micro-social change. The survey has core and variable components. There is a detailed focus on the core issue of employment and work history analysis, but core data are also collected on household formation and change, housing, and income.
Sampling Geographical coverage Timing Main areas covered Commissioned/sponsored by Undertaken by No of respondents Source of data Report Key themes Representative sample of households Great Britain Annual Employment, household change, consumption, health, financial commitments Economic and Social Research Council Centre for the Study of Micro-Change, University of Essex 5,064 households (1997/8) of whom approx 2,100 were mortgagor heads of household. Interviews with head of household and adult members of household. Some interviews with young people. Continuous publications Characteristics of mortgagors (risks faced); mortgage arrears
The importance of the BHPS is in encouraging and facilitating longitudinal analysis with an emphasis on documenting and understanding the impact of social and economic change at the individual and household level. The BHPS provides some information about the mortgage and housing characteristics of home buyers. Data on the numbers with mortgage arrears during the last 12 months are collected. Arrears are defined as owing two or more months payments. Under 1 per cent of mortgagor households reported that they had experienced arrears in 1997. The data available from the BHPS on MPPI is limited and incomplete. Take-up of MPPI is identified by a question that asks if MPPI is included as part of the mortgage payment. The approximately 40 per cent of households thought to pay separately for MPPI are therefore excluded. The number responding positively to the MPPI question is also much higher than might be expected given the way the question is asked, suggesting that respondents are including as MPPI other forms of insurance such as MIG or insurance on non-endowment mortgages to cover death. British Social Attitudes Survey (BSAS)
114
The British Social Attitudes Survey is an annual survey of individuals in Great Britain. There is a mix of annual repeated core questions and new topics/questions which are not asked every year. The number of achieved interviews has ranged from around1,300 to approaching 3,600. Since 1986, the sample has been split into two or three “versions”, so the sample size for any particular question may vary from year to year. The survey is either reduced or not carried out in election years. Attitudes to home ownership have been one of the core issues addressed by the BSAS. Other core areas include attitudes towards the welfare state, public finances and the labour market. Respondents are interviewed and also answer a selfcompletion questionnaire. The first BSAS took place in 1983 and the survey is an important source of time series data on attitudes.
Sampling Geographical coverage Timing Main areas covered
Representative sample of adults aged 18 and over Great Britain Annual Core areas; housing, the economy, work and employment, neighbourhood, welfare state and taxation. Other areas include: environmental issues, volunteering, Europe, the Lottery. A Sainsbury Family Charitable Trust and various Government Departments. A wide range of other sponsors have been involved at different times. The National Centre for Social Research (formerly Social and Community Planning Research) 3,146 in 1998 (and 2,360 for the self-completion) Individuals Report of the British Social Attitudes Survey Characteristics of mortgagors and mortgage lending (attitudes to owner occupation, attitudes to labour market risk, attitudes to public welfare), means of managing arrears
Commissioned/sponsored by
Undertaken by No of respondents Source of data Report Key themes
The BSAS is not designed as a source of information on key aspects of the housing market (safety-net provision, housing careers or mortgage borrowing etc). However, it provides important information on a range of issues that interact with housing decisions and experiences and which potentially shape attitudes towards risk and safety-net provision (for example, perceptions of labour market insecurity, expectations of house price movements or of household finances). With respect to home ownership, attitudes are tracked by a question which asks respondents whether they would advise a young couple with steady jobs to buy a home as son as possible, to wait a bit and then try to buy, or not to buy at all. In addition, 10 dimensions of home ownership are considered; four positive and six negative. These attitudinal questions are often used in other surveys.
115
Council of Mortgage Lenders Annual Housing Finance Survey A survey of this kind has been undertaken by the Council of Mortgage Lenders (and previously the Building Societies Association) at intervals since 1975. Traditionally, the survey was concerned with identifying changes in tenure preferences. This focus remains, but since 1997 questions have also explored a wider range of issues including aspects of the changes faced by mortgagors throughout the course of a mortgage, and the impact of changes in products and market conditions. The data are primarily an internal resource for the CML and the research projects they commission.
Sampling Geographical coverage Timing Main areas covered Quota sample of adults (16 plus) in all tenures, in 286 enumeration districts. From 1998, the survey covers the United Kingdom Annually 1998 Demographic and household characteristics; housing tenure, respondents’ mortgage borrowing; insurance products; mortgage payment related insurance; additional borrowing; flexible mortgages; the role of lenders and intermediaries; housing market and customer charters; tenure preferences; risks factors Council of Mortgage Lenders MORI 2,329 (1998); 2,027 in GB in 1997 Heads of household or partner Annual Housing Finance Survey Report. Articles in Housing Finance Characteristics of mortgagors and mortgage lending; means of managing arrears; MPPI
Commissioned by Undertaken by No of respondents Source of data Report Key themes
This survey offers one of the more comprehensive sources of information on safety-net issues. Key variables pertaining to MPPI include: take-up; the socio-economic characteristics of those with MPPI (age, income, region, social class, household structure, economic status, type of property); through whom MPPI was bought; date of purchase; how payment for MPPI is made; take-up of other insurances; why respondents do not take out MPPI; likely strategies for managing a loss of income; recent incidence of household income change, unemployment and changes in household composition. The survey provides time series data on MPPI takeup. In 1998, 39 per cent of respondents noted that they had an MPPI policy. This figure is considerable higher than reported by a range of other surveys. It is also higher than the number of the survey respondents (28%) who indicated that they could draw on MPPI should they lose all income. This lower figure is likely to be a more reliable guide to take-up. The discrepancy in the figures in all probability indicates the uncertainty amongst borrowers about the nature and purpose of the products they have (see the Evaluative Report, Section 2.3.3). The survey does not consider the costs of MPPI or the claiming process where, in any case, the numbers would be too small for analysis. Council of Mortgage Lenders provision of statistics on mortgage arrears and possession 116
Statistics on mortgage arrears and possessions have been provided by the Council of Mortgage Lenders (and previously the Building Societies Association) since 1980. The scope of the statistics has gradually enlarged over the period. Initially, figures were provided for arrears of six and more months (as one category) and for possessions. In 1982, arrears were disaggregated into arrears of 6-12 months and those in excess of 12 months. In 1993, the CML added figures on arrears of 3-5 months and also provided statistics in the form of a percentage of balance analysis, indicating the number of mortgages 2.5-<5%; 5-<7.5%; 7.5<10% and 10% or more in arrears.
Sample Geographical coverage Timing Main areas covered Commissioned/sponsored by Undertaken by No. of respondents Source of data Report Key themes Around 17 of the UK largest lenders United Kingdom Six monthly figures reported for June 30 and Dec 31 . Restricted time series since 1980. Fuller time series since 1993. Arrears of 3 and more months, properties taken into possession, arrears as proportion of outstanding balances. Council of Mortgage Lenders Lenders Not applicable, records of all in arrears and possession Mortgage lenders’ administrative records Twice yearly press releases. Housing Finance and annual statistical publication. Mortgage arrears; mortgage possessions
th st
The sample size and composition has also varied throughout the period, initially based on building societies only but since the second half of 1990 it has been based on a sample of lenders. Typically the sample represents approximately 85 per cent of total mortgage business. Figures are grossed up to represent the industry as a whole. The data exclude cases with arrears of under three months. Voluntary and compulsory possession are not routinely differentiated in published reports. Measurement of the number of months in arrears is affected by interest rate changes. The CML statistics carry the following statement: ‘Changes in the mortgage rate have the effect of changing monthly repayments and hence changing the number of months in arrears which a given amount represents. The figures for arrears cannot be used to show precise trends from one half year to another. A given amount of arrears represents a higher number of monthly payments the lower the level of mortgage rates’.
117
Family Expenditure Survey (FES) This is a long-standing sample survey of private households in Great Britain. Data are collected from each individual in the household aged 16 and over, and interviews are also held. Data collection on the expenditure of children has been introduced recently. Approximately 6,500 households provide data.
Sample Geographical coverage Timing Main areas covered Commissioned/sponsored by Undertaken by No of respondents Source of data Report Key themes Multi-stage, stratified random sample of 11,000 households drawn from the Small Users file of the Postcode Address File Great Britain excluding the Scottish off-shore islands and the Isles of Scilly Annual 14 areas of households expenditure (housing, food, household services etc.) plus household income data Government Departments/Office of National Statistics Office of National Statistics and Northern Ireland Statistics and Research Agency 6,409 (62% response rate) Households, adult individuals within households, children Annual publication of Family Spending Characteristics of mortgagors and mortgage lending (weekly payments, mortgage related expenditure)
The FES does not collect data on mortgage arrears, although it is an important source of data on mortgage costs per se and in relation to other items of expenditure and household income.
The FES asks about MPPI using a composite question: ‘Do you have any mortgage protection policies to repay this mortgage in the event of sickness or accident, redundancy or death’. It is therefore not possible to identify what specific cover people have nor to distinguish those just reporting polices for non-endowment death cover. Weekly expenditure on insurance is collected, but in view of the question asked, the data should be treated cautiously as a source on the cost of MPPI. The FES also provides some income data and asks about the receipt of money from unemployment insurance. Again, it is not clear that this necessarily relates to MPPI, if at all. The 1994 FES Report included a particular focus on insurance expenditure, but gave little attention to MPPI.
118
Family Resources Survey (FRS) The FRS is an annual, continuous survey of householders in Great Britain undertaken for the Department of Social Security. The survey started in October 1992 although the first full year’s data relates to 1993-4. It collects data on both households and individuals and focuses primarily on the wide range of financial resources available to households (income, benefits and other financial sources).
Sampling Geographical coverage Timing Main areas covered Effective sample size of 34,502 households in 1987/8from small users Postal Address File Great Britain (excluding Scottish Islands and north of Caledonian Canal) Continuous interviewing Household structure and characteristics; tenure and accommodation; receipt and extent of benefits; employment status and earnings; household finances (all in-flows, intra household contributions); financial assets
Commissioned by Undertaken by
Department of Social Security Social Survey Division of the Office of National Statistics and the National Centre for Social Research (formerly Social and Community Planning Research) Approximately 23,500 (68% response rate) for whom over 9,000 are mortgagor heads of household. (1997/8 survey) Heads of households and adults (all aged 16 and over except 16-18 year olds still in full-time non-advanced education) Annually, Family Resources Survey, published by Corporate Document Services Characteristics of mortgagors and mortgage borrowing. ISMI (numbers receiving, how much, how long); MPPI (take-up, what is covered, last payment for MPPI, period covered, year taken out, eventualities covered, (MPPI included in mortgage payment, receipt of MPPI payments, amount and for how long)
No. of respondents Source of data Report Key themes
In 1997/8 The FRS collected data from 9,357 mortgagor households . Data are available on the mortgage loan (including re-mortgages, the date and the reasons for them). The survey does not collect information on the number of respondents with mortgage arrears. Information is collected about several aspects of MPPI (take-up, date of take-up, what is covered, last payment for MPPI, period covered, year taken out, eventualities covered). The questions on MPPI have been improved to ensure that those taking non-endowment death policies only can be excluded from the figures on MPPI take-up. In 1997/8, 2,500 mortgagor households with MPPI were recorded as having MPPI. Data are also collected on the number in receipt of an MPPI claim, at what level and for how long. In 1987/8 there were only four cases recorded of mortgagors in receipt of MPPI. There are no data collected on the number of rejected claims. The FRS also provides information on the receipt of ISMI with 208 ISMI recipients identified in 1997/8. There are, however, indications that the FRS under-represents the number on IS/JSA.
119
General Household Survey (GHS) The General Household Survey is continuous survey that has been undertaken since 1971. It interviews a sample of the general population in Great Britain. It is designed as a modular survey with some core elements and other topics introduced or repeated intermittently. The survey did not take place in 1997/98 and 1999/2000. The survey will take place in 2000/2001.
Sample A two-stage sampling procedure from post-code sectors. Sample size was 13,247 households in 1996. Interviews are held with all adult members of the selected households. Great Britain April-March Core areas: Demography, family and household formation, housing, health, employment and education. Office of National Statistics and Government Departments Office of National Statistics 17,043 individuals in 9,158 households (1996) Households and individuals Living in Britain Characteristics of mortgagors and mortgages
Geographical coverage Timing Main areas covered Commissioned/sponsored by Undertaken by No of respondents Source of data Report Key themes
The 1996 GHS provides data on 3,790 mortgagors, recording whether they have a mortgage loan, purchase price of property (not since 1993), employment and household characteristics. There are no data on mortgage arrears, ISMI or MPPI.
120
Income Support (IS) Statistics Quarterly Enquiry (QSE) Jobseeker’s Allowance (JSA) Statistics Quarterly Enquiry Statistics on the characteristics of IS and JSA claimants and claims are provided by the Department of Social Security on a quarterly basis. The data are derived from the records of a 5 per cent sample of all IS and JSA claimants taken on the last weekend of February, May, August and November. The JSA sample is taken from the middle of the same months. Two quarterly reports are published, one for IS and one for JSA claimants. Statistics on the number in receipt of ISMI can be obtained by combining data from the two reports.
Sampling Geographical coverage Timing Main areas covered Commissioned/sponsored by Undertaken by No of respondents Source of data Report Key themes
5% sample of all IS/JSA claimants Great Britain Quarterly Characteristics of claimants, duration and amount of claims, regional analyses, capital, deductions from benefit, housing costs and outcomes Department of Social Security Department of Social Security Variable with case load Administrative records Quarterly reports published by the Analytical Services Division of DSS. ISMI
Data from the sample are rated up by 20 in order to provide national estimates. The QSE records data on many more variables than it routinely uses in its reports. Relatively little is reported on ISMI other than the numbers claiming, by full or half interest, average weekly amounts, by statistical group, and as a proportion of all claimants in those statistical groups. Data on the incidence of MPPI amongst IS/JSA claimants is not recorded. There are some official reservations about the reliability of the data on mortgage loans in the IS/JSA claim period.
121
Scottish House Condition Survey (SHCS) A survey of private households in Scotland. The survey involves data collection from heads of households and a survey of properties by professional surveyors. The survey was conducted in 1991, 1996 and another is planned for 2002. .
Sampling Geographical coverage Timing Main areas covered Commissioned/ sponsored by Undertaken by No of respondents Source of data Report Key themes Scotland 1991, 1996, 2002 Household structure, employment income, tenure, neighbourhood, house conditions, maintenance and repairs Scottish Homes and the Scottish Office MORI 19,892 interviews (70% response rate) 17,918 physical dwelling inspections (81%) Households, surveyors reports Scottish House Condition Survey, Main and Summary Reports $ Characteristics of mortgagors and mortgage lenders, MPPI
Roughly 6,340 mortgagor households were interviewed as part of the 1996 survey. The SHCS does not ask any questions on the experience of mortgage arrears. The SHCS asks one question about MPPI to measure take-up. It asks all those with a mortgage ‘Do you have a mortgage protection insurance policy, that is a policy that will cover the repayments on your mortgage in the event of you being made redundant or being unable to work.’ The emerging evidence on the extent to which respondents understand insurance products suggests that the data should be treated cautiously (See the Evaluative Report, Section 2.3.3) There are no questions on the cost of MPPI and while those claiming on MPPI might be included as respondents in a question that asks if there is any benefit received from accident, sickness scheme etc. this might also include non MPPI paymentsA$#B@8@$<8@$C7$DE@FGH7CF$FI@=H6H=<;;J$9H8@=G@9$