We are testing a new beta website for gov.scot go to new site

Evaluation of the National Mortgage to Rent Scheme

Listen

CHAPTER NINE COSTS AND VALUE FOR MONEY

Costs

9.1 We start this chapter by considering the costs of the scheme overall and then look at the unit costs of cases and how these vary. We consider various possible benchmarks for comparison. Later sections of the chapter consider the wider achievements of the scheme, including some consideration of the issue of sustainability of solutions, and go on to discuss cost-effectiveness and value for money.

Budget and Spend

9.2 Table 9.1 presents simple data on the budgeted level of total spending on the scheme and compares the actual spending outturn. The spend comprises the major item of actual subsidies (capital grants and allowances) paid to social landlords, together with associated professional fees and charges. We understand that the cost of professional fees and charges amounted to £238k pa; most of the budget therefore goes towards subsidies.

Table 9.1: Mortgage To Rent Scheme Budget and Spend

Financial Year

Level of Funding (£)

Actual Spend (£)

2003-04

3.25m

1.555m

2004-05

4.00m

3.989m

2005-06

7.00m

6.878m

2006-07

10.23m

8.176m

2007-08

10.00m

9.433m

2008-09

10.00m

Source: Scottish Government

9.3 To provide some sort of benchmarks for the overall scale of this budget in a Scottish context, one may compare this programme of £10m with a number of other programmes

  • Affordable housing investment programme £584m (2007/08, including MTR within it)
  • Central heating initiative/warm deal £46m
  • Wider role £12m
  • Modernising private sector housing £10m

9.4 One may also compare it with wider relevant programmes of financial support to housing costs

  • Income Support for Mortgage Interest ( ISMI) Scotland £31m (in 2005)
  • Housing Benefit ( HB) Scotland £1,132m (in 2005)

(Wilcox 2007)

9.5 This shows that MTR is only 1.7% of the overall affordable housing investment programme. However, it is of the same order of size as such programmes as 'Wider role' or 'modernising private sector housing'. Perhaps more to the point, the cost of MTR is about one-third of the cost of ISMI paid out in Scotland, although only 0.9% of the total Housing Benefit bill.

9.6 It appears that the MTR scheme has somewhat underspent in each year of its operation, although the extent of this underspend was not massive in most years. Like most new programmes, the initial buildup of spending may have been slower than expected. In general this tendency to underspend rather supports the views of stakeholders reported in Chapter 3, that the scheme was not rationed on the basis of available money (as opposed to on the basis of eligibility rules and knowledge of the scheme). While takeup has not had to be rationed for financial reasons hitherto, a combination of Credit Crunch/market downturn and more effective promotion of the scheme could create a pressure of overspend or a greater need to ration.

Financial parameters and unit costs

9.7 We have undertaken some analysis of the financial data within the databases, although as noted in the introduction there are some uncertainties concerning the completeness or up-to-datedness of some of the data fields, particularly on 'Equity Returned' and 'Shortfall'. We have tried to replicate the calculation of subsidy and the equity/shortfall numbers, but we cannot precisely match those in the MTR database.

9.8 The transfer price for dwellings under MTR is the valuation assuming the condition of the house is made good to social landlord letting standard, less the repair costs estimated to be required to achieve that (up to a maximum of £6,000. That is not quite the same as actual current value plus total repair costs (the first two columns of Table 9.2) because the impact of repairs on value may not be commensurate with their cost. On average, value is enhanced by 67% of the cost figure, a difference of £1,801.

9.9 Subsidy is designed to make things neutral for the social landlord, so they can break even on the deal. It is basically calculated as the transfer price, less the capitalised value of rental income (net of management, maintenance and major repairs allowances), plus an administration allowance of £900. This is similar in principle to the calculation of HAG for a new or acquired RSL dwelling, although the values of typical MTR purchases would generally be below full price or cost procurement of a brand new social rented home. A subsidy of up to £6,000 is also available to cover repair costs.

9.10 Over the life of the scheme the average value (post repair) was £77,677 and the average subsidy (including repairs, net of equity contributions) was £49,607 making an average subsidy rate of 64%. The absolute figures have been rising over time and by 2008 they had reached £97,526 and £58,570 (60%) 19. The increase is not a surprise, given the logic of the scheme and what has happened to house prices over the last 4 years. However, subsidy increase has been modified by the greater availability of equity contributions in the last period. The absolute size of the subsidies clearly is an issue and this will be discussed further under the heading of value for money.

9.11 By way of benchmarks, in 2007-08 a new build housing association rental unit approval had an average cost/value of £129,740, an average grant amount of £87,180 and a grant rate of 67.2% (unpublished data from SG). For the LCHO programme, the figures were £139,470, £55,710 and 39.9%. Thus the MTR subsidy cost is more similar to that for the New Supply Shared Equity scheme than for new social rent, partly because of the lower property values.

Table 9.2: Key Financial Values for All Settled Cases by Year 20, Region and House Type (Mean, £ at current prices)

Value (before repair)

Total Repair Cost

Net Subsidy

All O/S Secured Loans

Unallowable Loan Amount

Year Settled

2003

45342

3266

26491

39237

0

2004

53727

3985

33884

42129

3514

2005

66043

5486

46410

48120

4356

2006

78222

6148

54040

60357

5944

2007

87094

5908

57336

67246

4624

2008 (qtr 1)

91826

5023

58570

68550

5357

Region

Clydeside/West C B

67247

5378

45172

52884

3972

Lothian

95150

5848

65768

75282

7234

Central-Fife-Tayside

68040

5611

47528

52176

4700

North East

68954

5257

42830

47940

4682

Rural South

72286

4938

36901

41732

1495

Highlands & Islands

69813

4137

44255

47175

926

House Type

Flat/Mais

62638

5380

42754

47550

3760

Terrace

77365

5273

51858

60804

5527

Semi

85061

5810

55396

61967

4517

Detached/Bung

75333

4668

48280

58786

4095

Total

74069

5465

49607

56538

4675

Source: Authors' analysis of MTR and Allocation of Cases databases.

9.12 The average level of debt secured on the properties is £56,538 overall (£68,550 in 2008). This implies an average equity of £19,531 (26%), but of course individual circumstances will vary greatly. Also, many households have additional unsecured debts. Within the total secured debt there is, on average, £4,675 worth of unallowable borrowing items (not for housing, essential transport or care purposes). Clearly this will be distributed very unevenly between individuals.

9.13 All the financial values have tended to increase over time, but this is particularly true of the values of properties, which increased by 92% from 2003 to 2007 (compared with 71% for all sales in Scotland - CLG mix-adjusted data). Between 2004 and 2006 MTR prices went up by 29% compared with 18.5% in Scotland (Sasines median data). Subsidy has gone up even more, by 120%, basically because social rents have only increased modestly relative to house values. Repair costs have gone up by 80% to 2007, although falling slightly again in 2008. Loans have gone up by 75%, less than the price rise indicating an increase in average equity margins.

9.14 Looking at the shaded areas in the Table 9.2, it can be seen that all of the financial values (except repairs) are highest in Lothian - this clearly reflects housing market conditions, with these numbers primarily driven by house prices. Although prices are lower in Clydeside, subsidies are lowest in the Rural South. Prices, subsidies and debt are highest for semi-detached houses. However, debt values are also almost as high for terraced homes.

9.15 Subsidy costs have been increasingly defrayed by capital contributions, mainly from equity left after sale. The amounts involved are calculated taking account of limits on amounts of capital which households can retain. No specific evidence was forthcoming on the appropriateness or otherwise of these limits, so we assume that they are reasonable.

Relative values

9.16 We have examined average MTR values and compared them with lower quartiles and medians for all house sales in 2004, 2005 and 2006 by local authority. This exercise generally confirms that MTR sales are consistent with values in the lower part of the open market. Overall, MTR prices equated to 0.75 times the median or 1.06 times the lower quartile of all sales. There was no clear trend over time and most local authority areas conformed broadly to this pattern. A few areas had relatively higher values for MTR cases, but these were typically low value areas such as Dundee and Renfrewshire, although not invariably - e.g. Midlothian, 0.88 of median, 1.17 times lower quartile. Some areas had relatively lower still values for MTR, but these were typically high value areas such as Edinburgh, East Renfrewshire and Stirling. For example, East Renfrewshire values were 0.61 times the median or 0.87 times the lower quartile.

Stakeholder views on Values

9.17 Chapter 3 identified value limits (related to local average values) as one of the more problematic elements within the design of the scheme. Advisers saw value limits as barriers to potential applicants in real need, whereas landlords did not have strong views on the property values being approved. Some perceived that the scheme was set up only to capture those in former Right to Buy properties, although clearly that is not formally the case. There was some concern amongst advice agencies that in Edinburgh and Lothians there were a number of ineligible cases because of high property values and/or recent increases. However, as Table 9.2 shows MTR values were significantly higher in this region, suggesting that the scheme is able to respond to such pressured markets.

9.18 The property value limit in the eligibility criteria was an issue when advisers were generating applications and some potential applicants were considered not eligible at this stage. There was evidence that the MTR team had exercised some small flexibility over fluctuations or small changes in property prices at the ceiling or margins of scheme eligibility, and also some successful appeals over value limits. Nevertheless, many owners with financial problems but in higher value properties are likely to be deterred from even applying.

9.19 There is a clear need to review scheme value limits and for a commitment to update them periodically in line with local market changes. Indeed, interviewees involved in the administration of MTR recognised that use of comparisons backdating to 2003 was an anomaly which should be reviewed. There is currently no consistent definition of what constitutes the "local area" for this comparison. This has been beneficial in the early years of the scheme in that it allowed a good deal of flexibility which could act in the applicant's favour if they could argue at appeal that the correct comparison was over a differently defined area than in the assessment. However, as the scheme has grown and if it is to be a stable element of policy it would be more transparent and accountable if a consistent definition of local area could be adopted (discussed further in chapter 10).

9.20 Any move to raise value limits would have wider implications. For example, it was suggested that higher eligible property values may be a disincentive to association involvement if it puts pressure on the loan value which would have to be serviced by the property rent. The implication of this is either one or other or a combination of higher rents or higher grant levels. For the scheme as a whole, higher value limits would raise total demand and also raise the average grant cost, although there could be some offset in higher levels of equity contributions. As we go on to discuss in Chapter 10, such a move might make more sense as part of a widening of the scheme to embrace shared equity solutions. This in turn might entail widening the eligibility criteria to incorporate more formal income assessments.

Value and Sustainability of Solutions

MTR household outcomes

9.21 Chapters 5 and 8 provided detailed evidence on the outcomes as experienced by households using the scheme. The predominant picture is a positive one of relief at resolving the arrears problem and the uncertainty and satisfaction with their situation as social tenants. Insofar as there were remaining problems in some cases, these typically related to either continuing financial difficulties or to dissatisfaction over repairs (including their financial impact). Many new MTR tenants are still repaying debts after the sale of their property as they had insufficient equity to repay all of their debts. As noted in Chapter 8, some landlords identified higher than average problems with paying rent amongst MTR tenants. Repairs was an aspect which split respondents, with some very dissatisfied with this area in particular, due to costs, delays or the quality of works, as discussed in the previous Chapter.

9.22 The other main concerns expressed were about the level of remaining equity or the property valuation. Around 1 in 6 of those who had settled through MTR had felt that the outcome was unfair because of the level of equity they had lost.. One of the factors contributing to this area of dissatisfaction was the financial effect of delays in the sale process, which can have a considerable impact on arrears and equity levels. These delays need to be minimised as far as possible.

Overall Value for Money

9.23 It is difficult to do a formal 'value for money' analysis of MTR, in the sense of a cost-benefit analysis or even a full financial appraisal. While we do know the costs and the asset values involved in MTR transactions, we do not have such a 'hard' or well-defined set of measures of the benefits of the scheme. Often such benefits require reference to what might have happened in the absence of this particular scheme or opportunity, the so-called 'counterfactual', which is not something we can be very confident about. Nor at this stage have we specified what alternatives it is to be compared with.

9.24 With these important cautionary qualifications, we can make some relevant observations.

Benefits or effectiveness

9.25 The primary stated aim of MTR is to prevent homelessness through repossession, so this provides key pointers to relevant measures or estimates. Effectiveness measures would include the number of cases of homelessness prevented, while financial benefits would include the cost of dealing adequately with that number of extra cases of homelessness.

9.26 MTR hitherto has probably reduced homeowner repossessions by about 140 cases per year or 8% (Ch. 4). Overall, the number of homeless applications due to mortgage default is around half the number of repossessions (c.900pa). Whether the same proportion would have characterised MTR cases is not clear, but the interview evidence in Ch.5 and above suggests that half or more would indeed have presented as homeless. Not all homeless presentations turn into households being rehoused, but we suspect the proportion would have been relatively high for this group, although it should further be noted that only around half might have been 'priority need'. In short, this evidence suggests that between a quarter and a half of MTR settled cases might represent a homeless rehousing case averted.

9.27 Another way of looking at this is to ask the 'counterfactual' question: if there had been no MTR, what would have happened? The interview evidence suggests many of these households would have applied to the local authority or other social landlord for accommodation, whether through the homeless route or not. So this supports the view that MTR is preventing homelessness, and also reducing the demand on the rest of the social sector, by bringing additional property assets (as well as tenants) into the sector.

9.28 Some unsuccessful MTR applicants might still end up applying for social housing as well, given their backgrounds, preferences and circumstances (i.e. continuing difficulties). So MTR does not prevent all homelessness associated with mortgage arrears and possessions. There is an additional intangible social benefit from avoiding people having to move and uproot. Some stakeholders as well as households themselves regarded this as an important benefit.

9.29 However, part of the wider benefit of the MTR scheme (for people it helps and those that are ineligible, or withdraw) may be in getting people into contact with advice agencies and acting on their advice, and thereby preventing worse outcomes including repossession and homelessness. The size of this impact is probably smaller than the direct impact of settled cases, based on the database analysis. The proportion ineligible or withdrawn is 51%, of which between 40% and 55% (about 70 households a year) have an apparently 'favourable' outcome - but some of these would have happened regardless. This part of the benefit of MTR could be reflected in lower demand on social rented housing, insofar as these cases mainly achieve resolution entailing retention of owner occupation.

9.30 Sticking with homelessness, what is the 'benefit' of a homeless case (or indeed a repossession case) averted? We would be reluctant to attempt to put a financial value on the psychological and social effects of these processes. A more practical approach might be to attempt to put a cost on the rehousing process. In simplistic terms, an extra homeless household might mean that the SG has to fund an extra new unit of social housing, which as we saw above costs about £87,000. In some more pressured areas, additional costs may be incurred for temporary housing.

9.31 However, this is still too simplistic. Not all areas of Scotland have a particular shortage of affordable housing (Bramley et al 2006), and relatively few areas use all or most of their lettings for homeless cases. There are 50-60,000 homeless applications per year but only 18-20,000 of these translate into a social letting. Meanwhile, around 7,000 new social rented or low cost home ownership units are completed per year. So you could stay that at best there is only a probability (well below one) that for any repossession/homeless case averted a new build investment is saved; this probability could be in the range 15-35%, on the above figures.

9.32 By this rather crude (or some would say 'heroic') set of assumptions, you could reach the conclusion that an MTR settled case is 'worth' between £7,600 and £20,400 in terms of costs avoided for additional social housing investment 21. These figures obviously look low compared with the unit cost of MTR (£58,500). One should make allowance for the intangible social and indirect benefits mentioned above, but it would be hard to argue that these fully account for this difference. But it is also necessary to consider MTR in a somewhat wider policy context here.

Wider benefits.

9.33 The discussion so far has focussed on a rather narrow range of costs and benefits. In reality, the MTR scheme arguably has a wider range of benefits, in a general sense or in specific cases.

9.34 In a more general sense, MTR could be regarded as a sort of state-backed insurance policy for homeownership - either on its own or more especially as a part of a package with the Mortgage Rights Act and other measures. If you divide its cost (£10m) across 1.5 million homeowners, it looks quite cheap compared with commercial MPPI premiums, and almost certainly more effective. However, it is necessary to underline that it, like other safety nets, provides only partial cover in some circumstances.

9.35 While this may not have been its main intention, MTR is playing a significant role in countering wider problems of problem debt and financial exclusion in society. Whether the design of the scheme provides a best fit for this role is questionable, but there are some wider benefits here from MTR, particularly from the role of financial advisers working with the scheme.

9.36 Coming down to a more specific level, MTR helps people remain in their home and a neighbourhood to which they are attached. There are psychological and social benefits from avoiding the disruption of a forced move and the stress of continuing insecurity, which can take a toll on mental and physical health and wellbeing, including on family relationships, on children's educational performance and potentially on problems like delinquency.

Assessing the Costs

9.37 Having discussed benefits or effectiveness, we should return to costs. Surprisingly few stakeholders mentioned costs as an issue, although one or two Government stakeholders did do so. However, as outside observers we are obliged to draw attention to the question of the costs of the MTR model. It is not so much the total costs of the scheme, and certainly not its administration costs, but its high up-front unit costs in terms of subsidy which must be critically questioned. £58,500 is a lot of public subsidy to have to pay to help someone avoid a repossession and (probably) resolve their personal debt problems, given that the asset which the social sector acquires is an occupied property, not a vacant lettable unit (and one which will probably not become a relet for 10-20 years). As argued, in summary terms, above, it is difficult to mount an argument that the benefits in terms of homeless cases averted are worth that much in terms of costs which would definitely be avoided.

9.38 A further consideration here is that, according to much of the evidence from this study, the factors which are giving rise to many MTR applications are less to do with malfunction in the housing markets and much more to do with consumer credit - reckless lending linked to reckless borrowing. This suggests that the right solution to these problems lies in a different albeit overlapping area, financial regulation and preventative work in terms of financial education and advice.

9.39 Broadly this suggests that we should be looking at other policy mechanisms which might effectively target these problems, as well as the ostensible problem addressed by MTR, and do so in ways which cost less per unit in up front government cash. A range of options are discussed in the next chapter.

9.40 However, the 'insurance' perspective mentioned above paints MTR in a more favourable light. Instead of seeming expensive because its cost is divided over very few units, it looks cheap because it is pooled across everybody. This assumes that the scheme does very effectively target those 'last resort'/'unavoidable'/'bad luck' cases which insurance is good for, does cover most or all of them, and does reassure all parties that these problems have been covered. The MTR scheme is currently targeted as a 'last resort' but it is not in practice able to help all potential cases. It is not very widely-known among the general public. Some of the problems it deals with might have been avoided by better awareness and more responsible lending, particularly in the secondary sector.

9.41 The points just made suggest that the alternatives to be explored should include insurance type arrangements. The problems most frequently identified with 'insurance' type solutions to social problems are generally known as 'moral hazard' and 'adverse selection' ( or more generally, 'asymmetric information') problems. Moral hazard arises where parties have some disincentive to fully reveal their problems and thereby offload costs onto the insurer and avoid the costs of their own mistakes or misbehaviour. 'Adverse selection' is when people or businesses who are bad risks are the ones who tend to take out the insurance. We suspect that these problems are quite significant in the mortgage safety net field, which is partly why MPPI has not been very successful. Our worry in this instance is that billing MTR as the insurance against mortgage failure will encourage borrowers to behave fecklessly and while encouraging lenders to lend recklessly and not attempt to exercise forebearance or flexibility.

Conclusions

9.42 MTR currently costs £10m per year, comparable with some other smaller programmes in the housing and regeneration field. Widening the scheme's scope and responding to the current market crash could both increase costs. Although property values are modest and correspond with local market conditions, these have increased. The average subsidy cost per settled case is relatively high, similar to the cost of new LCHO units although below the cost of new social units. This raises questions about the value of benefits, but we can only make crude estimates for some of these items.

9.43 The evidence from this study (drawing on the earlier chapters) clearly shows that in many respects MTR has been relatively successful. It was implemented quickly and has spent within its allocated budget without having to ration help arbitrarily, while maintaining close targeting on 'last resort' cases. Most stakeholders welcome the scheme and generally praise its operation. Most households going through the scheme benefit significantly and regard the outcomes for them as favourable and appropriate. The administration arrangements adopted have worked well in many respects. MTR has prevented significant numbers of repossessions in Scotland and many of these cases have averted the disruption and stress of homelessness. The scheme may also have directed other households into taking and acting on financial advice which has alleviated their problems.

9.44 Most settled MTR cases represent good outcomes for households, with only minorities expressing regret or displaying signs of unsustainability. About half of these cases may represent a homeless case averted, while some of the remaining applicants may benefit from involvement with the scheme. These outcomes need to be weighed alongside the costs of MTR in an overall assessment.

9.45 Although it was not raised as an issue by many stakeholders, we are obliged as independent researchers to raise the issue of unit subsidy cost as a cause for concern. The average cost per case is similar to the unit cost of new low cost home ownership provision, yet what the social sector acquires are occupied units. It is difficult to claim a commensurate saving to balance this. The international review in Chapter 2 also found the Scottish scheme expensive compared with the admittedly limited range of comparators. However, against these costs, it is also important to allow for the social and wider benefits of the scheme, particularly its potential role as a form of 'insurance', taken in conjunction with other mechanisms.

9.46 In view of the cost consideration there is a case for looking at a range of other possible options to meet the same objectives, possibly entailing less up-front cost. Some possibilities are considered in Chapter 10.

9.47 Some of the adverse outcomes perceived by some households (and other stakeholders) are undoubtedly exacerbated by delays. One way of reducing delays is to give clear guidance about how applicants can limit delays(e.g. you need to do X, Y and Z to make the application run smoothly). At the same time, guidance should ensure that applicants do not have unrealistic expectations. Other ways of tackling delay may involve changing the way landlords are involved and the treatment of repairs, as discussed further in the next chapter.