Homeswest

CHAPTER 9

INTRODUCTION

Public housing authorities such as Homeswest now serve two main purposes:  they manage very large assets, with a view to obtaining commercial returns for the taxpayers who own those assets;  and they provide welfare assistance to those deemed unable to provide for their own housing needs.  Although in principle the two objectives are distinguishable, the relation between them is inherently confused.  In the case of Homeswest, evaluation of how well these objectives are being achieved is made all the more difficult by the recent growth of elaborate forms of cross-subsidisation, particularly from Homeswest land development am to its welfare renters.  The position is further confused by Homeswest's semi-autonomous status, which has permitted it to operate with very little public accountability.  In this discussion it will be impossible to make sense of the whole of this labyrinth.  The main points of the argument will be to demonstrate that it is a labyrinth, to show that such complexity is contrary to the public's interest in accountability, efficiency and fairness, and to suggest ways in which State housing policy can be simplified and streamlined.

Homeswest is a large operation.  It is, as one of its recent publications says, Western Australia's largest land developer, builder and property manager.  Its house and land assets are estimated to be around $3 billion. ⁽¹²⁹⁾  It operates about 35,000 rental properties, or roughly five per cent of Western Australia's housing stock.  Its annual expenditure from all sources exceeds $600 million a year.  As mentioned in Chapter 3, Homeswest's debt, now standing at $892 million, is a significant constraint over the government's ability to manage its fiscal policies with appropriate flexibility.  Although its rental stock has grown only moderately (up from 27,000 in 1980), in recent times its land and loan activities have grown rapidly.  In 1980, Homeswest developed only about 200 residential lots for sale or rental;  in 1989 it developed an exceptionally high 4,000. ⁽¹³⁰⁾  In 1989-90, government agencies, of which Homeswest was the largest, developed half of all serviced residential lots in the Perth metropolitan area.  In 1991, it had 10,000 subsidised home loans on its books, making it a significant player in the home finance market. ⁽¹³¹⁾

The size and importance of Homeswest is obvious enough;  its success or failure is another matter.  To judge this, we need both good information about its operation and some agreement about its objectives.  Much has been done by its executive in the past five years to improve the management and administration of the organisation.  But at a fundamental level its goals and strategy remain unclear.

According to its 1991 Annual Report, "Homeswest operates a series of housing services for people who would otherwise have difficulty gaining access to or remaining in suitable housing.  These include lower-income home buyers, disadvantaged groups, people in the private rental market and borrowers who experience financial difficulties". ⁽¹³²⁾  We can accept (at least for the sake of the argument) that the government has a role in helping people to satisfy their housing needs.  The obvious question which follows is:  why should the state own and manage a large housing stock in order to achieve this goal?  Why not use more direct means to assist people financially, whether to purchase their own housing, or to rent in the private sector?

It would be easier to evaluate Homeswest's performance if we could put aside these basic questions.  Why should a government agency determine the housing choices for the poorer members of society?  Why should a government agency have control of such a massive asset, to use however its executive -- or its minister -- chooses?  Public enterprises, with unlimited government backing, tend towards inefficiency and are open to political influence, and, sometimes, to corruption.  Private choice tends to maximise efficiency, liberty and even fairness.  Those who today disagree with those tenets are likely to emphasise the necessity for government action to ensure fairness;  but that claim only shows the need for government assistance, not government ownership.

The question of fairness in housing -- particularly the eligibility conditions for housing assistance -- is a matter which is properly to be answered by open and informed political decision.  Nevertheless, in practice, eligibility criteria for housing assistance have been vague and loose, and Homeswest has had remarkable discretionary power in the application of these criteria.  This too is a reason for questioning its usefulness.  But the issue does not stop there.

After 1945, the then State Housing Commission embarked on a large programme of housing expansion, designed to provide cheap but adequate housing for young families.  Today, Homeswest housing is more than ever before "welfare" housing.  The transition from "public" housing for young families to welfare housing for those deemed to be poor began slowly in the late 1960s and accelerated in the 1980s, until today 83 per cent of public renters are welfare recipients.  This process has had two direct effects:  it has reduced returns to the State, as an increasing proportion of tenants has become eligible for rental rebates;  and it has heightened the role of Homeswest as a welfare agency, a branch office of the welfare state.  (Much is now being done by Homeswest management -- quite justifiably -- to try to reduce the "stigma" which, they say, is associated with this welfare role.  But this is only cosmetic;  Homeswest rental housing remains mainly welfare housing, however stylish it may be.)

HOMESWEST AS DEVELOPER

The squeeze on rental returns caused by the gradual shift to the welfare function has had profound secondary effects on the shape and purpose of Homeswest.  Reduced rental income has induced a search for new income sources.  Through the 1980s, Federal funding under the Commonwealth-State Housing Agreements remained more or less static.  A large input of State government funding in the mid-1980s has now dried up to nothing, as the excesses of WA Inc have compelled tighter budgeting.  Today, however, Homeswest claims that at least 80 per cent of its expenditure is "self-funded". ⁽¹³³⁾  How can a State agency which is not a trading enterprise be self-funding?  Homeswest's income is (roughly) one-fifth from rents, one fifth from Commonwealth grants, one-fifth from land and property sales, one-fifth from home loan repayments, and one-fifth from Key-start loan repayments. ⁽¹³⁴⁾

Both rentals and loans are taxpayer-subsidised programmes.  The loan subsidy ($7 million in 1990-91, $3 million in 1991-92) is much smaller than that to rentals ($56 million in 1990-91, $63 million in 1991-92).  In reality, as will be shown later, the subsidy to rentals is much larger than this suggests -- perhaps twice as large.  In practice, then, the loans subsidy is relatively insignificant;  it is rentals which matter.  Since Commonwealth grants have been insufficient to cover the rising cost of the rental programme, the shortfall has been met by expansion in the land and property sales department -- which, as we saw above, has grown dramatically since 1980.  In 1990, land purchase and land development made up 12 per cent of Homeswest's capital works expenditure;  land and property sales produced 19 per cent of Homeswest's income.  Since 1986, sales of land and houses have brought in an average of $32 million per year, most of it from land sales.  This money is being used to make up the rental shortfall

The rationale for this cross-subsidisation is weak at best.  To the extent that Homeswest sells its assets at prices higher than would otherwise be the case, the profit is generated at the expense of new home buyers;  to the extent that the land is purchased by Homeswest at less than market prices, the profit is at the expense of the taxpayer.  It is not obvious that this cross-subsidisation in any way serves the public interest.  If it has any purpose, it can only be to disguise the size of the subsidies to welfare rentals.  Open and accountable government requires that such transfers be made explicit.  Off-budget welfare transfers conducted by semi-autonomous agencies are undemocratic and unfair, procedurally if not substantively.  In a democracy, the final say on public spending rests with all citizens, who must be able to know how and for what purposes public moneys are being used.

HOMESWEST AS ASSET MANAGER

The 1991 Annual Report presents a chart which purports to show that Homeswest's return on its rental properties is around 14 per cent before subsidies, and around 8 per cent after subsidies.  These figures are an accurate deduction from the data presented in the Annual Report ($142 million from 35,000 properties valued at $850 million).  This is, however, an obvious undervaluation of Homeswest's rental assets, for it implies that the average value of the properties is $24,000. ⁽¹³⁵⁾  A more plausible figure is perhaps $60-70,000. ⁽¹³⁶⁾  On that basis, Homeswest's before-subsidy return on rental assets worth about $2,275 million is 6.2 per cent;  after subsidies, this is reduced to 3.8 per cent.

These figures are gross returns.  After costs are taken into consideration, the position looks much worse.  Administration costs alone amount to $40 million, at least two-thirds of which will be spent on the rental programme.  Maintenance is around $20 million and rates are around $25 million.  These expenditures reduce the gross return of $142 million to a mere $70 million, or a return of 3.1 per cent.  Other expenses, such as depreciation and interest on loans from the Commonwealth government, are too complex to be factored in here, but they can be assumed to be significant, eating up much of that $70 million.  The true net return is likely to be very small.  After the $56 million subsidy to rentals is counted in, the net figure is substantially in the negative.

In effect, then, Homeswest's considerable rental assets are virtually non-productive in an economic sense.  This does not, of course, assign a value to the welfare role they perform.  It does, however, show that the rationale that some of the rental properties are productive is specious -- all of the properties and their tenants are being subsidised by the taxpayer, in the sense that the assets are earning less than their market value would permit.  This must cast doubt on Homeswest's financial accountability -- or perhaps rather the ability of its managers to manage free from political influence.  It suggests that we need ways of separating the commercial and the welfare functions, so that both can be made to perform more efficiently, and ways of ensuring that the appropriate commercial and welfare policies operate in a context where political influence is clarified.

THE SIZE OF THE SUBSIDIES

The median rent for a three-bedroom house in the private rental market in 1990-91 was $130 per week. ⁽¹³⁷⁾  For two-bedroom flats it might have been around $85 per week.  For a mixture of housing stock such as Homeswest's, the median might have been $115 per week.  Homeswest's stock is probably somewhat below average market value -- its median market rent might be $105.  Before subsidies, Homeswest receives $142 million, or $80 per week per property.  This suggests that all tenants are receiving a "hidden" subsidy of $25 per week.  Rental rebates are officially stated as $56 million in 1990-91, amounting to an additional subsidy to 27,000 rebated tenants of $40 per tenant per week

If we go further and ask what return might be earned on an asset worth (at a very rough guess) $2,275 million, then a larger figure is suggested.  At present that asset appears to be earning little or nothing.  In effect, then, all of the asset's earnings potential is being sacrificed to serve a welfare function.  On that way of reckoning the matter, many Homeswest renters will be receiving a subsidy worth perhaps $100 per week.

HOUSING SUPPORT AS WELFARE

Homeswest's programmes are designed, as we have seen, "for people who would otherwise have difficulty gaining access to or remaining in suitable housing".  Who are these people?  The question opens up some well-known difficulties.  One answer might be, most families with young children -- very broad cross-section of the population.  Another answer might be, only those who are chronically destitute -- a very small percentage of the population.  Between these two extremes a variety of positions is possible.  At present, almost all of Homeswest's tenants are welfare beneficiaries or pensioners.  In effect, then, present policy is to tie housing eligibility to welfare recipiency.  What is the rationale for this policy?

Homeswest does not report its finances with much clarity.  It is even less open about its tenant population.  The 1991 Annual Report shows that of the 7,705 new rentals allocated, 74 per cent went to "families", 16 per cent to "seniors", and 10 per cent to "singles".  No information is given about the composition of the overall current population.  No such information has been made public since the 1987 Annual Report.  That report showed that of the then 21,342 rebated tenants, 37 per cent were sole parents and widows, 30 per cent were age pensioners, 14 per cent sickness and invalid beneficiaries, and 14 per cent unemployed.  At that time, 66 per cent of all tenancies were rebated;  since then that figure has risen to 83 per cent.  Between 1975 and 1986, the proportion of elderly tenants doubled and the proportion of sole parents and widows almost tripled. ⁽¹³⁸⁾  Much of the remainder is likely then to have been young couples with children.  Today, however, very little Homeswest rental housing goes to two-parent families with children (unless -- improbably -- much of the 17 per cent non-rebated population falls into this category).

What one thinks about this trend depends upon some complex calculations, The conventional view, which has dominated both welfare discussion and policy since the mid-1970s, is that the elderly and sole parents are particularly disadvantaged sections of our society.  This consensus is quite mistaken.  In general, there is no difference between the economic status of sole-parent families and two-parent families, despite the radical difference in their incomes and their dependency upon welfare.  In general, also, the elderly are very much better off than families with children (whether sole-parent or two-parent), despite the much higher incomes of families.  The full proof of these claims is complex and will have to be presented elsewhere, but it is possible here to refute a conventional defence of present policy criteria.

The current National Housing Strategy, in a paper entitled The Affordability of Australian Housing, has laid down a "benchmark" for public housing assistance. ⁽¹³⁹⁾  It has termed this eligibility criterion "financial housing stress", and defined it as covering all income units in the lowest 40 per cent of the income distribution range which are spending more than 30 per cent of their income on housing.  (It also regards long-term, low-income renters who spend more than 25 per cent of income on housing as likely to be suffering "housing stress";  but the 30 per cent criterion is the primary definition.)

This criterion is faulty in at least four ways:

1. "Income" here is gross or pre-tax income.  But obviously a household's real standard of living is a function not of income it never receives, but of its disposable income.  Most households will have a disposable income about 25 per cent less than gross income.
2. Similarly, real living standards are a function not just of income but also of household size.  A household of one adult and one child will have a very different capacity to meet its housing costs than will a household of two adults and four children.  We need to use not a simple monetary measure, but an "equivalent income" measure, which takes into account the number of people in the household and the economies of scale made possible by shared living.  In what follows, we will use the widely-accepted OECD equivalence scales, which count the first adult in a household as 1.0, the second as 0.7, and each child as 0.5. ⁽¹⁴⁰⁾
3. Non-monetary income also needs to be counted, including the fringe benefits that go with pensions and benefits.  Non-monetary government support can be quite substantial, sometimes amounting to about $3,000 a year.
4. More controversially, leisure should be counted as a form of non-monetary "income".  In the case of the involuntarily unemployed, spare time is not to be counted as leisure because it is not chosen;  in the case of sole parents who could work more than they do, it may be a factor to be counted.

Even if we put aside (4) as too contentious, counting in (1) and (2) makes a radical difference to the incidence and prevalence of what the National Housing Strategy calls "housing stress".

The NHS paper goes on to present a set of income statistics designed to determine which sections of Australian society are most in need of housing assistance.  Comparisons are presented between the housing expenditures of various standard types of low-income household unit.  These are taken to show that couples with dependent children and sole parents have a far higher proportion of income unit types facing "housing stress" than any other type of income unit.

The NHS figures suggest that there is little overall difference between the stress levels of these two groups -- sole parents come out as slightly more "stressed" than couples with dependants.  But the comparison makes no sense, because the criterion is absurd.  Sole parents, on average, pay very little income tax, whereas the couples will, on average, pay at least twenty per cent of gross income.  And the sole-parent income unit will be on average one adult and half a child smaller than the couple family.

When the effect of taxation is included in the equivalence calculations implied by the OECD scale, it will be seen that the gross income needs of the two-parent unit will be very substantially higher than those of the sole-parent unit.  The net effect of these two considerations will be that the "housing stress" of the sole-parent group will fall away to less than half the proportion represented in the NHS paper, while that of the couples with dependants might rise by fifty per cent.  If we now include the third point (above), the value of the fringe benefits which go with pensions and benefits, the comparison is even further tipped in favour of sole parents and against couples with children.  Without access to all the relevant data, no exact measurement of these changes is possible here, but it is obvious that the NHS figures bear only a very rough relation to real housing needs.

Similar points need to be made about the other conclusions drawn from income unit data by the NHS.  The paper claims that private renters, recipients of social security payments and – surprisingly -- single-person income units contain the highest proportion of all units facing housing stress (relative to their proportion of the overall population).

For instance, we are told that "... single income units comprise one-third of all income units, but 53 per cent of those in housing stress" (defined by the 30 per cent benchmark). ⁽¹⁴¹⁾  These single income units turn out to be mainly older women, and women aged between 35 and 64 renting privately.  But, again, such older women have lesser needs than other income unit types.  Most over-60s pay no tax, and they have no dependants to support;  taking those considerations into account would give us a very different picture.  Only by the very artificial criterion of the NHS could this group be counted as "stressed".  The younger (under-60) single women are equally unlikely to be stressed.

Pre-tax income figures are therefore very unreliable guides to housing "needs".  Yet it is just such income figures which Homeswest uses to allocate its housing assistance.  In 1992, income eligibility limits (for metropolitan and rural applicants) were as follows:

Size of household	Single income	Double income
One person	$335	-
Two persons	$441	$507
Three persons	$528	$607
Four persons	$571	$657

(Additional persons:  Add $43 per child)

These figures heavily favour small households and disadvantage large households.  They thus favour singles, most sole parents and the elderly, and disadvantage couples with two or more children.  Because they are based on pre-tax income and do not count non-monetary welfare benefits, they are even further biased against working families (whether couples or sole-parent families) and favour pensioners and beneficiaries.  Specifically,

1. A sole parent with one child is eligible on an income of less than $441 per week, whereas a single-income "standard family" of two adults with two children is eligible up to $571 per week.  The OECD equivalence scales would count the sole-parent family as 1.5 (1.0 for the parent and 0.5 for one child) and the two-parent family as 2.7 (1.0 for the first adult, 0.7 for the second adult, and 1.0 for two children).  This gives a ratio of 1 to 1.8 for this comparison.  The Homeswest eligibility ratio is 1 to 1.3 (441:571) which favours the first family over the second,
2. Elderly pensioner couples are given similarly favourable treatment in comparison with the "standard family" (1 to 1.3);  though the OECD scale rates them as 1.7 as against 2.7, or a ratio of 1 to 1.6.
3. The OECD ratio for young singles and single age-pensioners to a family of four should be 1 to 2.7.  The Homeswest eligibility criterion is $335 compared with $571, or a ratio of 1 to 1.7, which heavily favours single people.

These disparities are magnified considerably when we take taxation into account.  A family of four with a disposable (or post-tax) income of $528 per week would have an earned (or pre-tax) income of about $33,000 per annum.  This would make many working families with children eligible for Homeswest rentals.  Their present ineligibility is a form of unjustified discrimination against these working families, and in favour of pensioners and welfare recipients.

If the eligibility limits are a guide to its allocation practices, then Homeswest is quite inequitable in the allocation of its rental resources.  The published evidence supports this suspicion.  In the period 1988 to 1991, singles were allocated 310 rentals, seniors 1,668, and families 2,099. ⁽¹⁴²⁾  Sixteen per cent of the waiting list for seniors have obtained Homeswest accommodation, as against only seven per cent of families and less than four per cent of singles.  It is quite likely that much of the "families" category would be sole parent families, though no figures are published on this subject.

Two-parent families with children are likely to be getting very little Homeswest assistance, even though their needs are as great as, or greater than, the needs of those who are supported.  If sole parents occupy one quarter of Homeswest's housing (a reasonable guess based on the 1986 proportions), then about one quarter of all sole-parent families are Homeswest tenants.  If two-parent families with children occupy, say, 40 per cent (and the true figure may be lower), then only 7 or 8 per cent of all two-parent families are Homeswest tenants.

The rental subsidy is, as we have seen, substantial -- not less than $25 per week for non-rebated tenants and probably much higher even for them;  possibly more than $100 per week for some tenants.  This makes public housing a very "lumpy" welfare good.  Those who get public housing do much better than those who do not.  The main alternative funded by Homeswest is rental support in the private market, which in 1990-91 it provided to over 11,000 customers at an average value of $15 per week.  Homeswest is also helping with the financing of 10,000 home loans, at a value of around $400 million, but the level of assistance to each home buyer seems to be low -- a subsidy of $7 million in total in 1990-91, worth $13 per buyer per week.  It is easy to see from this comparison why there is a long waiting list for public housing.

Homeswest has 35,000 rental properties.  At present Western Australia has about 90,000 unemployed, about 50,000 sole parents, about 46,000 invalid pensioners, and about 118,000 age pensioners.  But there are also many low-income employed two-parent families who are at least as deserving and needy as any of the above groups.  What possible yardstick could be used to distribute public housing to some of these people but not to others?  Homeswest does have a role in helping to house those who are chronically destitute and unemployable, but they are only a small fraction of its present rental population.  Its remaining renters are indistinguishable from numerous others who get little or no housing assistance.  (The waiting lists are one indicator of the size of this population:  18,000 for rental housing, 17,000 for home loans, with 9,000 registered for both loans and rentals, making a total of 26,000 applicants.)

Until demonstrably fair criteria are devised there can be no satisfactory welfare housing policy.  Fairness involves overcoming the "lumpiness" of the housing assets.  But to do this would be to transform Homeswest, downgrading its role as a provider of housing and re-designing its role as a welfare agency.  It is no reply to this argument to say that governments cannot afford to support the housing requirements of so many families.  The argument being presented here does not require increased housing support.  What is required is a restructuring of the present eligibility and allocations criteria.  And to do that it is probably necessary to eliminate the "lumpiness" problem by shifting to a system of housing vouchers or allowances.

The lumpiness' problem is exacerbated by the current upgrading of Homeswest's rental properties.  Homeswest talks of being a "market leader" and prides itself on constructing "an appealing variety of homes matching and often exceeding the standards of private developers". ⁽¹⁴³⁾  Figures given in the 1991 Annual Report suggest that at present the construction costs for completed rental units are averaging $77,000 per unit ⁽¹⁴⁴⁾ -- counting in the cost of land would often double the value of these properties.  When it is taken into consideration that some of these are flats and many are small units or small houses, this seems to be a very high figure.  Casual observation shows that these new properties are generally of a high standard and sometimes (as is the case of the Bennett Street apartments) almost in the luxury class.

The value of subsidies to tenants calculated above is, of course, an average, which may conceal wide variations.  There will be a massive difference between the real value to the tenant of an old Homeswest house in Hall's Creek and a new Homeswest townhouse on Buckland Hill.  The Buckland Hill project cost $4.2 million for construction alone of 40 units for seniors and 33 townhouses for families, or $57,500 per dwelling.  The townhouses are likely to have cost about $70,000 each.  The block value of the land could well be about the same.  Whatever the figure, the value of the subsidy to such tenants is very considerable.  Homeswest notes that, despite its very long waiting lists, it still has some vacant housing "... mostly in remote country areas or on the outskirts of Perth where demand is minimal". ⁽¹⁴⁵⁾  Demand for townhouses on Buckland Hill is not likely to be a problem,

All of this shows that Homeswest has little concern to spread the benefits of housing assistance to the maximum number of tenants.  A commitment to that principle would involve greatly reduced construction activities, more spot purchasing of existing houses, and more leasing of properties from the private market.  As Walsh points out, "If, say, $60,000 might be spent on purchasing or constructing a single dwelling, that amount might be invested instead at, say, 10 per cent per annum.  The resulting annual income of $6,000 would then provide subsidies of almost $60 a week to two households, instead of to only one". ⁽¹⁴⁶⁾  Extending the logic of this argument leads away from public housing authorities altogether, towards a voucher scheme.

There is a third kind of awkward and inequitable "lumpiness" in public housing, that to do with security and duration of tenure.  New tenants get much more than a substantial housing subsidy;  they get semi-permanent possession of at least part of that subsidy.  Yet within a few years many, perhaps almost all, tenants will grow out of the need for those subsidies.  Homeswest can adjust to such changing circumstances by requiring increased rental payments in line with increasing incomes.  Yet this is only a partial adjustment, for the maximum rent chargeable (the so-called "cost" rent) is still well below market rents.  Homeswest does not publish figures on the median duration of tenancies.  The turnover rate in 1990-91 was 24 per cent, but some of these movements were transfers from one rental dwelling to another.  On this basis, median durations must be more than five years and might well be ten or higher.  Ten-year tenants will be getting very favourable treatment, compared with private renters or home buyers.

It is common to argue that assistance to public housing is justified because home buyers and owners are themselves heavily subsidised by the taxpayer, and public rental assistance only counterbalances these subsidies.  In their well-known 1987 analysis of this subject, Flood and Yates concluded that public renters received subsidies worth $1,795 per annum, home owner/ buyers $1,450, and private renters only $295 (in 1984 dollars). ⁽¹⁴⁷⁾  Since that time, assistance to private renters has been increased, but the general picture is probably little changed.  As we have seen, subsidies to public renters in Western Australia seem to average around $50-$80 per week, or $2,500-$4,000 per annum.  Flood and Yates' derivation of the subsidy to home buyers is debatable, resting as it does on the assumptions that imputed rental income should be subject to income tax, and that capital gains from home ownership should be subject to capital gains tax.  In reply to these claims, it can be argued that council rates and water rates are in varying degrees taxes levied on home ownership;  and that home ownership is in part a form of savings, which should not be double taxed.  Be this as it may, there are two further considerations which are not usually brought out in discussions of "tenure neutrality".

Suppose that we were to abolish the "inequitable" tax expenditures which favour home ownership.  Let justice be done, though the (political) heavens fall.  Government revenues would rise considerably, and since our intention was not to increase government expenditure but simply to achieve tenure neutrality, taxation would need to be lowered.  If personal income tax were lowered, most of the reductions would benefit higher tax payers, almost all of whom are home-owners or -buyers.  Very little would go to the welfare-poor, who pay little or no income tax.  Even if all taxes, and not simply personal income taxes, were lowered, the share of benefits going to the poor would be not very considerable, since the net effect of all taxation is progressive.  "Subsidies" and tax expenditures apparently favouring home ownership are in reality often a "transfer" from A to A, and therefore not an equity matter.

Indeed, if the argument were carried right through, public renters would be much worse off, for removing the "inequitable" tax expenditures which favour home ownership removes the "equity" argument for public housing.  Only private renters would gain.  The net effect of redistributing equally all housing assistance would be to improve considerably the position of private renters;  to lower the position of most public renters;  and to lower very slightly the position of home owners.

The tenure neutrality case for increasing support for private renters is a reasonable one but it too is weak at a point which is often overlooked.  Most private renting is renting by young people before they move on to marriage, mortgages and children.  Seventy per cent of 15-24 year olds are private renters;  after that stage of life private renting falls away rapidly, to 32 per cent for 25-34 year olds, 18 per cent for 35-44 year olds and 10 per cent for 45-54 year olds. ⁽¹⁴⁸⁾  (Most private renting is of flats, not of houses.)  The question of fairness to private renters is largely a matter of life-cycle fairness.  Before we can pass any useful judgements on this question, we need something like the life-cycle analysis pioneered in New Zealand by David Thomson, ⁽¹⁴⁹⁾ which documents a multiplying of housing costs for younger families at a time when an older generation of long-term home-owners has done relatively well.  This has, however, little to do with State government housing policy, and is thus for the most part outside the scope of this discussion.

INCENTIVE EFFECTS

Homeswest is, as this chapter has argued, essentially a welfare agency.  (In terms of sheer size, it rates about equal with the other State welfare agency, the Department of Community Development.)  Its welfare functions are, however, being "managed" by little more than guesswork.  No publicly-argued criteria or standards are being applied, no evaluations conducted, no research commissioned.  This failure ought to be obvious to anyone in the organisation, but it is not;  mainly, it would seem, because Homeswest's attention is fully occupied by its quasi-commercial component, and by its desire to improve its "image" and its internal staff relations.  The failure is also political.

One kind of welfare failure has already been documented -- the inequities in the rental eligibility criteria and allocation practices.  Homeswest's failure to maximise the spread of its assistance is partly a function of the inherent "lumpiness" of its merchandise, but it has been exacerbated by management decisions.  A less obvious problem, but perhaps a more serious one, concerns the incentive effects generated by welfare benefits.  There is very good reason to question the social effects of government cash benefits, leaving aside housing provisions.  Many low-skilled workers will be little or not at all worse off on benefits than they would be after working a full week.  Many sole parents will be no worse off on benefits than they would be if married to a low-skilled worker. ⁽¹⁵⁰⁾

To add housing assistance selectively to one side of this set of scales is to unbalance radically a very delicately balanced situation.  As we have seen, public rental assistance can be worth between around $60 and $100 per week or more.  Over the past fifteen years, Homeswest has given preference to welfare recipients and age pensioners ahead of the claims of working -- two-parent and sole-parent -- families.  The effect must have been to make many public renters better off on benefits than they would have been in the workforce.  To suppose that such incentives have little effect on actual behaviour is to make a leap of faith, for no research supports such a conclusion, and much suggests the opposite.  The tendency must have been to encourage voluntary welfare subsistence and family breakdown, and to discourage workforce participation and family stability.

These propositions could be debated more fully if Homeswest had conducted and published research into the characteristics of its clientele, or, more broadly, "performance indicators" of its own social impact.  Because it has not done so, we can only discuss the issue in the abstract.  We do know, however, that welfare dependency has expanded and deepened considerably in the past two decades;  and we do know that in the same period Homeswest rental housing has become almost entirely housing for welfare recipients and pensioners.  And it is at least as plausible to call Homeswest the cause, and welfare dependency the consequence, as it is to describe it the other way round.

Homeswest management could hardly fail to be aware of issues of this sort.  (Its maintenance costs alone -- $550 per property annually -- would seem to testify to the social difficulties associated with long-term welfare dependency.)  It has also had to cope with a legacy from the 1960s of unsuitable, now run-down, flats in areas noted for a variety of social problems.  Some of these flats are being redeveloped or sold, and Homeswest's current construction programme is intended to avoid similar future difficulties.  It claims that "Public housing estates that dominated entire suburbs have gone, to be replaced today by small, discreet pockets of housing scattered in suburbs throughout the metropolitan area". ⁽¹⁵¹⁾  This is only half true:  the "discreet pockets" are still pockets, the product of Homeswest's desire to play the part of developer and market leader.  The more important question is about the effects.  No doubt Homeswest hopes that upgraded housing will ameliorate (albeit very slowly) the associated social problems.  But to hope this is to make debatable assumptions about the origins of those social problems.  If the real origins are welfare dependency and family breakdown, then we can expect that much of this housing will end up as shabby and as trouble-ridden as some of the older stock.  (The infamous Pruitt-Igoe housing project in St Louis lasted only seventeen years before it had to be demolished;  in its youth it won an award for its architectural excellence.)

CONCLUSIONS

A number of points stand out from this discussion.  Homeswest is unaccountable in any real sense;  and its workings are difficult to understand.  It offends against a basic principle in that its welfare function is not openly funded.  The help it does give does not necessarily go to the most needy or the most deserving;  and the form that help takes is far from ideal.  It does not make the best use of its available resources.  Reform is, however, unlikely to be swift.  Homeswest has an extensive client base who are wedded to its existence, some of whom may, indeed, have reasonably fair claims to their entitlements (entitlements in some cases closely resembling property rights).

We will, therefore, suggest here some of the basic steps toward reform, rather than an instant and comprehensive reform package;  recognising that reform may well take at least one term and perhaps two terms of government.

The first step, important regardless of whether other steps are taken, is to recognise that Homeswest, having in essence only a welfare role, is not structured appropriately as a PTE, any more so, say, than Community Development would be.  It should be operated as a Ministry for Housing;  so that its Minister, its policies and its finances are completely open and accountable to people and Parliament.  Next, its functions need to be disentangled.  Welfare must be distinct from property management, and both from loans and property development,

The loans function is probably not run efficiently.  The Commonwealth government discovered this to be the case some years ago with its extensive Veterans' Affairs loans portfolio, which was sold to a commercial bank and openly subsidised.  This is likely to be the best policy in the present case, and will improve simplicity and transparency while still providing low-income borrowers with a degree of support and security;  it should not be too politically contentious.

The property development function is one which is hard to defend.  In general, government should not be a player in activities where it is also the umpire:  the temptation to fiddle the rules is usually too great.  And once again, there are considerations of efficiency:  other arms of government perform the same function, and it is unlikely that any of them performs the service as efficiently as the private sector.  On the other hand, the cross-subsidy to the welfare function is almost certainly too great to be absorbed in the general government budget all at once, at least without a considerable reduction in the welfare expenditure.  We recommend, therefore, that the government consolidate its property development activities under one head;  that the cross-subsidy continue, in suitably transparent form;  that the land development activities be progressively scaled down to zero over the next six years;  and that the general government sector gradually pick up the corresponding shortfall over the same period.

The property management division of the new ministry will need a very clear charter.  It should, in the first place, have a primarily management role.  That is, it should undertake as little actual work as possible, and contract out as much as possible.  It should, second, have a very clear mandate to provide as much assistance as possible within its given allocation.  That will mean a rather different kind of asset management from that currently practised.  In particular, the best use of scarce resources will require that as old housing stock comes up for renewal, its unimproved land value be impartially assessed and a clear decision made as to whether sale is not better than redevelopment.  The ministry should also cease competing in the aesthetics stakes:  while architectural services should be contracted out, it is not necessary to use some of Australia's most eminent architects as designers.  (There may be some public good, however, in cooperating with those architects interested in high-quality, low-cost housing initiatives.)

The welfare function may be the most difficult to sort out.  It cannot be undertaken without a review having access to a statistical base -- precisely the sort of data not now available to the public.  That review will be more easily undertaken if the ministry announces in advance that any changes to its tenant base will exclude certain classes or individuals:  there is no point in causing needless anxiety to eighty-year-old pensioners, for example.  The review should concentrate on developing new eligibility criteria for assistance which take into account the subtle but real problems of income definition outlined earlier in this chapter.  It should also take into account the life-cycle variations in individual and family wealth which, if properly codified, should produce eligibility criteria which would make public housing assistance a transitional phase for most tenants.  The ministry should then look closely at the question of appropriateness, and in the context of the best use of scarce resources.  The logic of the arguments presented above leads us to believe that the public ownership of a large stock of housing is an inappropriate and inefficient way of delivering housing assistance, and that a cash housing allowance would be superior.  (It would have the additional advantage, not canvassed above, of providing competition in the welfare rental sector.)  Ideally, such an allowance should be introduced early on as a substitute for expansion of the existing stock.  In time, beyond the scope of this study, it would enable the State to dispose of large parts of its housing, and enable it to achieve, far more than it now does, far greater equity, fairness and efficiency in this important welfare function

---

ENDNOTES

129.  Homeswest, Homeswest Housing in the 1990s, Perth, n.d., page 1.

130.  Ibid., page 2.

131.  Homeswest, Annual Report 1991, Perth, page 10.

132.  Ibid., page 1.

133.  Homeswest, Homeswest Housing in the 1990s, page 1.

134.  Ibid., page 3.

135.  That this is an undervaluation seems to be admitted at page 42 of the Report, where it is noted that "The State Housing Commission in conjunction with the Department of Land Administration is to be involved in an ongoing process to verify its rental asset base".  A reliable valuation would involve input from private-sector valuers.

136.  This is conservative.  In 1992, Homeswest sold 474 houses to tenants at an average price of $55,000.  These are likely to have been the older homes of long-term tenants, and so well below the median value of all Homeswest housing -- except to the extent that some older housing stock will be in inner suburban areas where unimproved land values will be relatively high.

137.  Homeswest, Annual Report 1991, page 13.

138.  Homeswest, Annual Report 1987, Perth, page 17.

139.  The National Housing Strategy, The Affordability of Australian Housing, Issues Paper 2, AGPS, 1991, Chapter 3.

140.  Organisation for Economic Co-operation and Development, The OECD List of Social Indicators, Paris, 1982, page 37.

141.  NHS, op. cit., page 28.

142.  Homeswest, Annual Report 1991, page 22.

143.  Homeswest, Homeswest Housing in the 1990s, page 4.

144.  Homeswest, Annual Report 1991, page 7.

145.  Ibid., page 12.

146.  Cliff Walsh, "Housing", in Richard Blandy and Cliff Walsh (eds), Budgetary Stress:  The South Australian Experience, Allen & Unwin, Sydney, pages 240-265, page 262.

147.  J. Flood and J. Yates, Housing Subsidies Study, Australian Housing Research Council Project No. 160, Canberra, 1987.

148.  National Housing Strategy, Australian Housing:  The Demographic, Economic and Social Environment, Issues Paper 1, AGPS, Canberra, 1991, Figure 2.2, page 9.

149.  See D. Thomson, Selfish Generations?  The Ageing of New Zealand's Welfare State, Bridget Williams Books, Wellington, 1991, pages 135-147.

150.  The detailed evidence for these claims is easily obtained.  See, for example, Alan Tapper, The Family in the Welfare State, Allen and Unwin, Sydney, 1990, pages 63 ff.

151.  Homeswest, Homeswest Housing in the 1990s, page 4.