Toward a Comprehensive Affordable Housing Strategy for …€¦ · · 2003-10-06Creating a level playing field for rental development 13 ... a household paying more than 30 percent

Toward a Comprehensive AffordableHousing Strategy for Canada

by

Steve Pomeroy

Steve Pomeroy is President of Focus Consulting Inc.

October 2001ISBN – 1-894598-94-6

Table of ContentsIntroduction 1

Defining The Problem – The Nature of Housing Need 2

Understanding the affordability issue 4

Shifting factors underlying affordability problems 5

Low levels of new construction 6

Lack of new rental construction worsens problem of affordability 7

Constrained access to affordable units 8

Erosion of the existing affordable housing stock 8

Consolidation – The context for housing policy 8

Responding to the need – Possible policy options 9

Supply Measures 10

Supporting public/nonprofit production 10

Incentives for private rental development 11

Creating a level playing field for rental development 13

Explanation of pooling provisions 13

Reducing development costs 14

Encouraging lower cost forms of development 15

Shifting patterns of ownership and facilitating not-for-profit ownership 16

Demand Measures 18

Rent supplement 18

Shelter allowance 19

Reform of welfare shelter benefits – A transitional shelter allowance 21

Some Concluding Observations – Toward a Comprehensive Housing Strategy 23

Glossary of Terms 28

Introduction

The issue of affordable housing andhomelessness has become increasingly promi-nent in recent years, highlighted by deathsof homeless individuals on the streets of majorCanadian cities and efforts by municipalities andhousing advocates for action by the federal andprovincial/territorial governments.

Earlier this year, the federal governmentannounced a proposal for a new AffordableHousing Program. Since then, discussions havebeen under way with provinces and territoriesabout program design and cost-sharing. The pro-posal involves a capital grant of up to $12,500per unit, with the expectation that the provincesand territories will match this level. The grantswould be provided to both private and nonprofitcorporations to help cover the cost of new rentaldevelopment. This investment is expected tostimulate construction of 12,000 to 15,000 newunits annually.

Following a meeting of housing minis-ters on August 16, 2001, the federal HousingMinister stated that the Canada Mortgage andHousing Corporation (CMHC) is prepared toincrease the maximum federal grant level, butwithin an overall budget cap of $680 million overfour years, and to allow the provinces and terri-tories some flexibility in program design.

This Affordable Housing Program isproposed as a short-term strategy, aimed atincreasing new rental construction to address theproblem of low vacancy rates. It is not a social

housing program, but provides a platform forprovinces/territories to add funding in order toassist households in need. The Minister statedthat he will continue to work with his provincialand territorial counterparts toward longer-termsolutions.1

The purpose of this paper is to presentan overview of the issues underlying the afford-able housing problem and to set out a series ofoptions that should be considered as part of acomprehensive national housing strategy.

This paper is structured into four sections:

• In Section 1, the nature of housing needand contributing causes are described.

• Section 2 describes one set of measuresintended to increase the supply of newhousing and related ways to ensure that thisnew supply creates units that are afford-able.

• Section 3 presents a similar discussion of‘demand-side’ approaches – measuresintended to help households that pay toomuch relative to their income better affordhousing in which they already live.

• Section 4 offers some concluding obser-vations and emphasizes the need for a com-prehensive policy, drawing from both thesupply and demand options to create aneffective response to housing need and thepredominant problem of affordability.

Caledon Institute of Social Policy 1

Defining the Problem – The Nature of HousingNeed

The primary basis for social housingpolicy is the presence of circumstances (housingneed) in which individuals or families are notable to meet their shelter needs through their ownresources. The appropriate choice of policyinstruments depends on the nature of the hous-ing problem being addressed.

Housing need in Canada has been definedin terms of adequacy, suitability and afford-ability. These three problems are the basis forthe definition of housing need established by thefederal housing agency, the Canada Mortgageand Housing Corporation, and are used in defin-ing and measuring need by all provinces and ter-ritories. This need measure is called core hous-ing need.

‘Core housing need’ means that a house-hold is experiencing at least one of three hous-ing problems, based on established housingstandards. Furthermore, the household hasinsufficient income to resolve this problem with-out assistance.

The three norm standards used in the coreneed assessment are:

• Suitability: This standard uses nationaloccupancy standards to determine if house-holds have a sufficient number of bedroomsbased on the family composition (effectivelya crowding measure).

• Adequacy: A standard that measures hous-ing condition to determine if the dwellingis safe, has basic plumbing and is in a rea-sonable habitable state of repair.

• Affordability: A standard based on a ratioof housing expenditures to total householdincome; a household paying more than 30percent of its income for housing is consi-dered in need.

If a household is found to be below oneor more of these standards, a second test isapplied to determine if its income is sufficient toafford a suitable and adequate dwelling in itscommunity within 30 percent of the household’sincome. This measure uses the median rent ofan appropriately sized private rental unit andconverts the rent to an annual income requiredto afford this unit based on spending 30 percentof income for rent.2 A household with grossincome below this level, and living below anyof the three housing standards, is defined asbeing in core housing need.

This formal housing need measure isbased on data collected by Statistics Canadathrough the Census and during the intercensal


TotalRentersOwners

UnattachedIndividuals

Single femaleSingle maleLiving with others

Families

Two parents withchildrenCouples, no childrenFemale lone parentMale lone parentMultiple family

Table 1Core Housing Need 1996

Source: CMHC; 1996 Census

1,725,655 1,172,270 553,385

878,415

245,950 149,330 52,705

847,235

144,030 110,140 144,030 16,945 8,475

68%32

51%

2817 6

49%

171317 2 1

period through another Statistics Canada survey.Prior to 1996, the Household Income, Facilitiesand Equipment database was used; since 1996,the Survey of Household Spending has been usedto collect data, but CMHC has not publishedupdated housing need data since the 1996 Cen-sus. Thus 1996 Census data remain the mostrecent official basis for housing need statistics,and are used here to indicate housing need.

CMHC reports that a total of 1.7 millionhouseholds or 7.6 percent of the total were incore housing need in 1996. This figure com-pares to 1.2 million households or 12.2 percentof the total in 1991.

These core need problems are dividedalmost evenly between unattached individuals(51 percent) and families (49 percent). Singlefemales make up more than one-quarter (28 per-cent) of all households in need.

Seniors comprise just over one-fifth ofhouseholds in need, distributed between unat-tached individuals and family couples. Amongfamilies, lone parents are the largest category inneed.

More than two-thirds of households inneed are renters (68 percent), even though renterscomprise only 35 percent of the population. Mostoften, renters experience an affordability pro-blem; owners in need tend more often to exper-ience adequacy problems (e.g., poor conditionof dwelling).

Affordability is, by far, the most seriousproblem: 95 percent of core need among rentersrelates to affordability. Not surprisingly, lowincome is a key determinant of core need: 85percent of households with incomes below$10,000 and 61 percent of those with incomesbelow $20,000 are in core need.

In the 1990s, heightened awareness abouthomelessness added a new dimension to thesehousing need categories – absolute homelessnessand households at risk. To date, there are noofficial statistics measuring the extent of theseproblems.

Absolute homelessness means livingwithout four walls and a roof, and perhaps moreimportant, without a door that one can lock tocreate private living space and security. Thenumber of chronically homeless persons is notaccurately known, as there is no easy way tocount this population. Proxy counts of shelterand soup kitchen use as well as efforts by streetworkers suggest, however, that this group isgrowing and becoming more diverse. Home-lessness is acknowledged as a complex socialissue – related to mental and physical illness, butoften exacerbated by a lack of access to afford-able housing and appropriate support services.

‘At risk’ has emerged as a term in the1990s to identify families and individuals thathave formal shelter, but whose circumstances areprecarious. They are deemed at risk for one of arange of reasons – the cost of shelter consumessuch a large part of their income that they arevulnerable to rent arrears and eviction; they aretemporarily living with a friend or relative buthave no permanent place of residence; they arepersonally at risk of physical or mental abuse;or they have disabilities that may cause them tolose their shelter.

One measure used to help determine thenumber of households at risk is a very high shel-ter cost burden – paying more than 50 percent ofincome for shelter. In the United States, the 50percent benchmark is used as an indicator of‘worst case need,’ also labelled ‘severely bur-dened households.’ These terms have not tradi-tionally been used in Canada, but recent efforts


by advocates to highlight the severity of theaffordable housing issue have begun to cite thenumber of households paying greater than 50percent. In the 1996 Census, some 830,000renters were identified above this benchmark –22 percent of all renters.3

Understanding the affordability problem

Clearly, affordability problems are cre-ated by a combination of two key factors – lowincomes and high relative rents. With a deepand lingering recession and slow economicrecovery, the 1990s were characterized by stag-nant income, especially among renters. Between1991 and 1996, the average real income of rentersdeclined by 12.4 percent.4

Incomes have grown since 1996,although most renters have not shared in thegains. As higher-income renters exercise theoption of moving into ownership, the renter pro-file is dominated by the remaining lower-incomehouseholds.

Another important factor underlyingtrends in income and core housing need is thehigh incidence of households that receive socialassistance. In most provinces and territories,social assistance benefits include an explicit shel-ter component, which is always more than 30percent (and often greater than 50 percent) of thetotal benefit cheque. Thus these households arecaptured under the measure of core need, even ifthey have been able to obtain housing within theshelter component provided to them.5

Estimates suggest that more than half ofcore need households are recipients of socialassistance (commonly known as ‘welfare’). Thepercentage varies by household type and region,in part reflecting wide variations in the level ofwelfare benefits.

Recently, analysts and advocates havebegun to use a shelter-cost-to-income ratio of 50percent as a more extreme indicator of afford-ability problems. As the shelter component ofwelfare often makes up as much as 50 percent ofthe total welfare payments, social assistance


Table 2Social Assistance Families as Percentage

of All Families in Core Housing Need

AtlanticQuebecOntarioPrairieBCCanada

Family%

656446403749

Lone Parent%

878075546574

Non-family%

556842392749

Total Non-senior%

687052443856

Source: Statistics Canada, HIFE 1996 data file.

caseloads affect the number of households withserious affordability problems.

Between 1990 and 1995, the number ofhouseholds in core housing need increased by47 percent and the number of households pay-ing greater than 50 percent of income for rentgrew by 43 percent. During the same period,there was a marked increase in welfare caseloads,associated in part with the 1990 recession andslow recovery – especially in central Canada. InMarch 1990, welfare cases totalled 1,056,000.The caseload peaked at 1,675,900 in 1994 andin 1995 stood at 1,659,200, a rise of 57 percentsince 1990 – and a larger increase than thenumber of households in core housing need.6

So the rising welfare caseload appears to havean important impact on housing need statistics,especially in assessing households with a highshelter cost burden.

Since 1995, welfare cases have followeda steady decline, in part due to reduced benefits(e.g., Ontario), more restrictive eligibility(especially for single people and youth) and theimproving economy. Total cases have fallenback below their 1991 levels – from 1,659,200in 1995 to 1,198,200 by March 2000, a reduc-tion of 461,000 (28 percent).7

Reflecting the decline in welfarecaseloads, data from the Statistics Canada 1998Survey of Household Spending indicate that thenumber of renter households paying more than50 percent of income for rent has fallen signifi-cantly from 830,000 in 1995 to 464,000 house-holds in 1998. At the same time, the number ofrenters paying between 30 and 50 percent ofincome on shelter increased by 250,000 house-holds – suggesting that while incomes may haveimproved marginally, possibly as householdsmoved into work, most are still paying well over30 percent for shelter and continue to experienceaffordability problems.8

The broad measure of renter householdspaying more than 30 percent also appears to havedeclined by 1998, but only marginally comparedto the 50 percent benchmark. This suggeststhat core housing need may too have declined,although CMHC has not undertaken an analysisof core need using the more current data files.

A more refined measure and data sourceis required to track these important trends –ideally, one that separately tracks the circum-stances of social assistance recipients and theworking poor (although these are not distinctgroups due to transitions back and forth). Therelatively high incidence of welfare householdsin the core housing need population also sug-gests that solutions should include an exami-nation of ways to improve this component ofsocial assistance programs.

Shifting factors underlying affordabilityproblems

As noted in the first half of the 1990s,the primary causes of rising core need weregrowing welfare rolls and stagnant incomeamong lower-income Canadians. Between 1991and 1995 real rents declined, although quitemarginally (-2.5 percent), so the primary causeof increased affordability problems was weakincome.

The cause of core housing need appearsto have shifted through the second half of the1990s and into the new millennium. Since 1996,increasing rents have become the more importantfactor – rising by as much as 25 percent between1996 and 2000 in some centres. This percent-age varies across the country, with the largestincreases evident in Alberta, Saskatchewan andOntario. Rents in Alberta jumped more than 20percent in only three years, while those in Ontariogrew by more than 10 percent.


While rising rents in Ontario have beenidentified as a consequence of removing rentcontrols, the more important factor in Ontarioand other provinces has been the persisting lowlevels of new supply and the consequent declinein rental vacancy rates. As the supply of avail-able units declines (and fails to keep pace withpopulation and household growth), the naturalresult is upward pressure on rents. This pro-blem is especially apparent in the lower levelsof the rental sector, where low-income house-holds seek accommodation.

Low levels of new construction

The vast majority of housing in Canadahas been produced and operated in the privatemarket. The social housing stock – i.e., unitsbuilt and operated by public agencies, nonprofitcorporations and co-operatives under varioussocial housing programs – accounts for roughly700,000 dwelling units in a total stock of justover 10 million dwelling units. Rental housingaccounts for just under four million units, sosocial housing constitutes less than one-fifth (18percent) of the rental housing stock.

Some of these social housing units are inmixed-income projects and provided at marketrent, although most have subsidized rents. The1996 Census reported a total of 1.6 million unitsrenting below $500 a month (affordable at 30percent of income for households with incomeof $20,000; this $500 rent/$20,000 incomebenchmark is used here as a crude benchmarkof affordability to illustrate the relative size ofexisting social housing stock and to highlightrecent changes). So social housing comprisesroughly one-third of the existing affordable stock.By comparison, two-thirds of affordable units arein the private rental sector and include olderapartments (many smaller units) and apartmentsuites in homes.

These statistics reveal the important roleplayed by the private sector in contributing to astock of housing that, over time, has remainedor become affordable. But the stock in this sec-tor is also susceptible to erosion. As rents risewith inflation, the number of private units rent-ing below affordable benchmarks declines. Inother cases, units are demolished or the propertyis converted to ownership (condominiums) andthe units may no longer be available at moderaterents.

Between 1991 and 1996, the total numberof renter households in Canada increased by180,000. This figure reflects net growth in rentalhouseholds, since some moved to ownership, butthis growth was not matched by new production.During these five years, a total of 111,000 newrental units were constructed with just over halfof these (54 percent) built as social housing. Theother half, produced for the private sector, weretypically at higher rents, far above affordablelevels. The shortfall in new production (roughly70,000 units) was taken up by vacancies in theexisting stock, which were relatively higher(nationally almost 5 percent) through the early1990s but declined substantially to a nationaloverall average of only 1.6 percent (much lowerin some centres).

Following a building boom in the late1980s, the level of private rental developmentalready had slowed by the early 1990s, and pri-vately initiated development has continued to fall.New private rental housing construction fell froman average of well over 30,000 units annually inthe 1980s to only 13,000 between 1990 and 1995and subsequently to only 6,000 annually from1995 to 1999.

The 1991 to 1996 period brackets the erain which new social housing programs werealmost entirely eliminated, although completionof projects approved in 1994 and 1995 (Ontario,


BC and Quebec) stretched into 1996.9 As aresult, the decline in private rental constructionwas exacerbated by the cessation of social hous-ing construction (except a little in BC and Que-bec).

Lack of new rental construction worsensproblem of affordability

This decline in new supply – both pri-vate and social – has been a critical factordriving the affordable housing crisis. As rentalconstruction has failed to keep up with newhousehold growth and demand, the rentalmarket has become increasingly constrained.Although lower rental apartment vacancy ratesand increasing rents are a prerequisite to newrental investment, recent analysis of rentalmarkets has found limited interest in new con-struction from developers and investors. Risksto investors and developers remain high while

the rate of return on investment is poor and un-competitive. Notwithstanding recent increasesin rent levels, new production of rental units isnot expected to increase in any significant way.10

Low supply and persisting demandinevitably applied pressure to rents, directlyexacerbating the affordability problem. Thishas become especially acute at the lower end ofthe market where there are already far morehouseholds seeking lower priced units than thereare units. As noted, the 1996 Census reporteda total of 1.57 million units renting below $500.In comparison, there were 1.68 million house-holds with incomes below $20,000 that cantheoretically can afford no more than $500 at 30percent of income – indicating an absolute short-fall of almost 110,000 lower rent units.

In addition, of the 1.57 million unitsrenting below $500 per month, just fewer than600,000 of these are occupied by households


Rental construction and vacancy rates declined through 1990's

0.0

5.0

10.0

15.0

20.0

25.0

30.0

89 90 91 92 93 94 95 96 97 98 99 0

Com

plet

ions

(000

's un

its)

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

Vaca

ncy R

ate

(%)

Private Social Vacancy

with incomes above $20,000. This means that600,000 lower-income households in need ofaffordable units do not have access to the lowerpriced stock – and must occupy higher pricedunits and pay a larger proportion of their limitedincome for housing.

Constrained access to affordable units

The low availability of modest rentunits underpins one of the core problems inthe area of affordable housing. The problem isnot solely an issue of absolute shortage. A rela-ted concern is limited access for lower-incomerenters.

As a rational entrepreneur, a landlordseeking to rent out a vacant unit typically willtry to minimize risk – he will prefer tenantsperceived to have greater probability of payingthe rent. Lower-income households, especiallythose living beyond their means, have a greaterrisk of not paying the rent. Inevitably, they willface a greater challenge in securing housing –especially if due to their inability to pay highrents, they previously have been evicted or arein arrears and do not have a sound reference.

In a tight rental market in which there arefew units available and landlords can be selec-tive, tenants with good jobs will tend to be pre-ferred. While there is a fine line between pru-dent landlord behaviour and discrimination, thebottom line is constrained access for low-incometenants, especially those stereotyped as ‘poortenants’ – which often includes social assistancerecipients and lone parents with children. Oneway to address this issue is by facilitating non-profit acquisition and increased ownership of thisexisting stock, as nonprofit landlords will tendto be more sympathetic to needy households.

Erosion of the existing affordable housing stock

While the level of new production is notkeeping up with demand and need, and con-strained access contributes to persisting afford-ability problems, this situation is worsened by aprocess of ongoing erosion of the existing pri-vately owned affordable stock.

Between 1991 and 1996, the number ofrental units renting below $500 declined by310,000 (roughly 10 percent of the total rentalstock). At a time when new rental constructionwas falling toward an all-time low level, the ero-sion of lower priced stock is perhaps the mostserious issue facing the affordable housing sec-tor. Policy options need to examine ways to pre-serve this asset.

Consolidation – The context for housing policy

The preceding section has provided anoverview of the key issues that underpin theaffordable housing problem and help to framethe context for which solutions must be sought.In summary:

• Low levels of new rental investment andproduction (and limited investor interest).CMHC projections of household growthindicate a requirement of 45,000-50,000new rental units annually through the nextdecade to meet new demand. This com-pares with current production of less than9,000 annually.

• Erosion of the existing lower priced rentalstock. Between 1991 and 1996, in excessof 300,000 units renting below $500 werelost due to rent inflation, conversion anddemolition.


• A ‘crowding out’ effect in which house-holds that are employed and otherwise per-ceived by landlords as less risky tenantsare occupying modestly priced units, lead-ing to constrained access to the lower rentpart of the market for working poor andwelfare households.

• Very limited funding to support newsocial housing initiatives. Most provincesfollowed the federal cut in 1993 and therehas been no funding for new developmentoutside of small programs in BC and Que-bec.

• Predominance of affordability problemsdue to both low and stagnant income andrising rents. Low income was the maincause in first half of the 1990s; rising rentshave become the more significant factorin the second half of the 1990s.

• A high proportion of welfare householdsamong core need households affects boththe total count of core need households andpossible solutions.

Responding to the need – Possible policyoptions

The goal of social housing policy is toassist low- and moderate-income households toobtain adequate and suitable housing at a pricethey can reasonably afford. This goal can beachieved through two general approaches:

1. By reducing or subsidizing the constructioncost of housing so that it is inherently moreaffordable. This approach typically involvesactions related to production and accordinglyare labelled as ‘supply measures’ (i.e., theprimary focus is on influencing the cost ofbuilding so that the rent charged can be

lower). These measures include direct pub-lic or nonprofit production, measures to lowerthe costs of production and stimulative meas-ures to induce private production with con-ditions on the level of rents.

2. By increasing a household’s ability to pay(i.e., increasing or supplementing its income).These are referred to as ‘demand-side meas-ures’ as they affect the household’s effectivedemand. Such measures include rent sup-plements or allowances and income assist-ance.

3. A third approach involves measures to influ-ence the price of existing rental housing. Itincludes controlling rent levels to protectaffordability, diverting demand (e.g., by help-ing households buy a home) to reduce pres-sure on a limited rental sector and capturingthe existing affordable stock by encouraginga transfer in the ownership from private-for-profit owners to not-for-profit owners. Theseapproaches have a more limited scope andare not discussed in any detail, except wherethey relate to the two main categories.11

The following potential approaches arediscussed in the next two sections. Each meas-ure is first outlined and key terms are explained.The advantages and disadvantages of eachmeasure are then reviewed. The concludingsection of the paper considers the potentialintegration of these measures into a compre-hensive strategy.

1. Supply measures

• Direct support for public/nonprofit pro-duction

• Incentives for private rental unit devel-opment

• Creating a level playing field for rentaldevelopment


• Reducing development costs• Encouraging lower cost forms of devel-

opment – single-room occupancy, sec-ondary suites

• Shifting patterns of ownership (facilitat-ing nonprofit ownership).

2. Demand measures

• Rent supplement• Shelter allowance• Reform of welfare shelter benefits.

Each general set of approaches, and theindividual measures in each set, are not mutu-ally exclusive. There are many cases in whichsupply and demand measures can effectivelycomplement each other. This complementarityis examined within each approach.

Supply Measures

Supporting public/nonprofit production

Direct public supply of housing has beenthe predominant program response in Canadathrough the postwar period. Although variousprogram designs have been employed, the essen-tial elements of this approach include:

• A public or not-for-profit owner/operatorwith a specific mandate to operate hous-ing for low-income households.12

• Some form of subsidy, either as a capitalgrant, favourable mortgage rate or ongo-ing subsidy, so that the rents charged totenants are affordable. Generally, theproject is governed by an operating agree-ment that specifies the obligations of theoperator and requires the provider to con-tinue serving lower-income eligible house-holds.

• Typically, operating agreements and anyongoing subsidy match the amortizationperiod of the mortgage. Once the mort-gage is fully paid off, the project rents arerequired only to cover operating costs andit is assumed that even at low rents, theoperator will be able to continue servinglower-income households (in very poor100 percent targeted projects, this may notbe feasible with some renewal of subsidy).

• In most nonprofit approaches, ‘affordablerents’ are established on a rent-geared-to-income (‘rgi’) basis. Although earlierprograms used a 25 percent ratio, most cur-rently employ a ratio of 30 percent againstgross income.

• Assisted projects may be fully targeted –meaning that 100 percent of tenants havelower incomes and pay rent on a rent-geared-to-income basis – or mixed income(where some portion of units are rented atmarket rent levels, rather than at subsidizedrents based on a percentage of income). Amixed income approach generally is pref-erable as it avoids the concentration of verypoor households that was characteristic ofearlier public housing and contributed tothe stereotyping of projects.

advantages

Nonprofit production helps create a per-manent stock of units specifically to serve lower-income households – a long-term investment ina permanent asset. This form of ownership alsoprovides some assurance that rents will remainaffordable over the long term.

Even after operating agreements and sub-sidies cease, the public investment remains inplace as the nonprofit charter restricts ongoing


use to provide affordable housing and mostnonprofit providers retain this initial motive.

Nonprofit programs address both supplyissues and affordability within a single program,which is one reason why they appear to be quitecostly, compared to alternatives that seek toaddress only one issue (either supply or afford-ability).

It is possible, however, to separate thesupply and affordability objectives. One suchapproach involves using a nonprofit organi-zation to develop housing at market rent level,then stacking rent supplement assistance(described below) as a separate mechanism tospecifically address affordability.

Another variant is to assist nonprofithousing corporations to purchase existing pri-vately owned rental properties. This approachdoes not add to new supply, but it does ensurethat long-term affordability is preserved andhelps stem the ongoing erosion of the relativelymore affordable existing private rental stock.

Supply programs respond directly to lowlevels of production (as prevail currently) and,in so doing, have a beneficial effect in moderat-ing the inflationary impact of low supply (i.e., inthe absence of new construction, there is pres-sure on rents in the existing stock).

Through the development process, unitscan be designed specifically to meet particularneeds – e.g., accessible or supportive housing.New supply funding can be geographically tar-geted to markets with acute supply problems.

weaknesses

New construction costs and the associatedsubsidy costs tend to be quite high on a per unitbasis. Most recent versions of nonprofit hous-

ing involve long-term (35-year) subsidy commit-ments. As new commitments are layered on ineach year, the subsidy cost mounts exponen-tially. This problem raises the concerns offinance officials and makes these programs a vul-nerable target in times of fiscal restraint.

The alternative is to ‘front-end’ thecost with a capital grant. Again, however, thisoption tends to be expensive due to the high costof new development compared with relativelylow revenues at rent-geared-to-income rents (thelatter limit capacity to cover operating and mort-gage payments).

Income mixing improves the viability ofthe projects and lowers the average capital grantrequirement. But it increases the grant cost perrent-geared-to-income household assisted.

Relative to the volume of housing need –more than one million renters in core need andmore than 450,000 paying more than 50 percentof their income for housing – incremental newsupply programs can assist only a very smallnumber of households in need (at their peak,nonprofit programs produced only 25,000 to30,000 units annually).

Incentives for private rental development

In building new rental housing, a devel-oper is faced with a range of costs and, in return,creates an asset that generates a cash flow in theform of rents. The value of the property to apotential investor is determined not by its costbut by the level of revenue it can generate. Themarket value of the new property is based on thevalue of future income flows.

In the same way that an investor in gov-ernment bonds is prepared to pay a price for abond based on an expected rate of return, the


rental investor converts future income to a cur-rent price based on a similar expectation of anannual return on his equity. The main differ-ence is that a government bond is a secureinvestment, easily traded for cash and the annualinterest payments are guaranteed. For the rentalinvestor, the asset is more difficult to trade forcash; it must be listed for sale and await a will-ing buyer. In addition, the annual cash flow isnot guaranteed. The property may experiencevacancies with no revenues produced or regu-lations (such as rent controls) may be introducedor modified and thus constrain the ability toincrease rental income and achieve anticipatedcash flows.

Compared to alternatives, rental invest-ment is relatively high risk and, as such, demandsa much higher expectation on rate of return. Thecritical issue in the rental market is that the ratesof return remain far below those expected bypotential investors. Costs are too high relativeto potential rent revenue, mainly because thecosts of production are set outside of the rentalsector. Ownership housing and condominiumsestablish land values while construction labourcosts and materials are established in the widermarket that includes commercial development.

In the past, stimulus programs have beenused to encourage private development. Theseinclude programs that provide a grant or inter-est-free loan to a developer in return for mod-estly designed units (which should commandlower rents than luxury development) and tem-porary tax measures that provide favourable taxtreatment to investors (namely, creating eligiblecosts that can be used to reduce taxable income).

Effectively, these programs replace someportion of the investor’s own equity with gov-ernment funds without affecting the rental rev-enue. So the investor’s rate of return increasesto a level that can make development moreattractive.

Currently, potential rates of return onmodest rental housing have been estimated to beless than 5 percent, compared with a minimalexpectation of 12 to 15 percent. Increasing a pri-vate investor’s rate of return to this level likelywill require a capital grant or interest-free deferredloan in excess of $15,000 to $20,000 per unit;this amount will vary significantly between cit-ies and depend on the type of unit being devel-oped.

advantages

Any type of new supply can have a bene-ficial effect in releasing the pressure of low sup-ply. Even new development at the higher end ofthe rental market will draw some households outof lower priced stock (a so-called ‘filtering’ or‘trickle down’ effect).

Household growth and new demand covera spectrum of incomes, including higher-incomehouseholds. Without new development, thesehouseholds occupy lower priced units and crowdout lower-income households.

Depending on the development econom-ics in a particular city, stimulus measures lever-age private investment and will cost less thanassisting nonprofit development (since, unlikeprivate developers, nonprofits typically have lit-tle, if any, equity to invest).

weaknesses

Stimulus measures have been controver-sial. Even the housing industry has argued thatthis type of intervention creates an artificial stimu-lus and disrupts market equilibrium, taking manyyears to rebalance.

The expenditure does not create perm-anent affordable housing, although it can be


designed to do so through certain conditions inexchange for the assistance.

Once operating agreements enforcingthe conditions and targeting expire, the units areno longer available as affordable housing (theso-called ‘expiring use problem’).

No long term public asset is created inreturn for the government investment.

Creating a level playing field for rentaldevelopment

The poor economics of rental investmentare attributable, in part, to the current tax treat-ment of rental housing, which is seen as unfaircompared with other sectors and investments.

Current inequities in the tax systeminclude:

• GST on rental housing. Rental landlordscannot charge GST on rents, yet unlikeother goods, rental housing is not zero-rated. So landlords pay GST on suppliesand services but cannot claim input taxcredits as they do not collect GST and haveno GST remittance against which to claimcredits.

• Small rental investors are not currentlyconsidered as ‘small businesses’ and there-fore are not eligible for the lower smallbusiness tax rate on the first $200,000 ofincome.

• Elimination of ‘pooling’ provisions thatpreviously encouraged reinvestment innew rental projects.

Explanation of pooling provisions

When a rental project is sold, investorsmust pay tax on the difference between the saleprice and the depreciated value of the project –i.e., the original cost of the project less the Capi-tal Cost Allowance (CCA) deductions. ‘Recap-tured depreciation’ (the difference between thedepreciated value – which applies only to thebuilding, not to the land – and the original cost)is taxed at full income tax rates. Capital gainstaxes apply to the increase in value above theoriginal cost.

Prior to 1972, rental investors coulddefer paying income taxes on recaptured depre-ciation on buildings sold by ‘pooling’ the recap-tured amount with CCA from other buildings.For example, if they acquired another rentalproject in the same year, they could avoidrecaptured depreciation by transferring theexcess CCA to reduce the depreciable value ofthe newly acquired project.

The restoration of pooling (also called‘roll-over’) does not eliminate tax liability. Itsimply postpones the tax penalty on recaptureddepreciation and capital gains for owners ofrental properties who invest the proceeds inanother rental property. Thus restoring this pro-vision creates opportunity for existing investors/owners of fully depreciated properties to lever-age the existing assets without facing large taximpacts.

Ironically, in some cases it is more bene-ficial from a tax perspective to demolish the pro-perty and thereby eliminate the recapture ofdepreciation. Under this approach the existingunits, which may be relatively affordable, arelost.


advantages

These three possible measures could cor-rect inefficiencies in tax treatment and, whileimposing an expenditure on government, do notconstitute a temporary disruptive measure –unlike short-term stimulus measures.

The measures may improve after-tax fea-sibility of new production (at least for higher rentproperties) and can stimulate new constructionwith a positive moderating impact on the mar-ket, thus easing rent pressures.

Most tax measures would apply only toprivate developers. However, any reduction inGST also would benefit nonprofit providers.

weaknesses

Federal finance officials have stronglyresisted expanding tax expenditures and arereluctant to implement tax changes unless a clearcase can be made on equity grounds.

These measures have a direct impact uponthe production of affordable housing but maystimulate market rent development at higher rentranges.

Reducing development costs

One of the most critical issues confront-ing both private and not-for-profit developmentis the relatively high cost of producing new hous-ing, which is a function of land costs, labour andmaterial costs. Total development costs also havebeen increasing as a result of layers of taxes andfees on development.

While costs vary across the country anddepend on the type of development, new rental

housing in a typical urban center will cost in therange of $65,000 to $105,000 for one-bedroomapartments and from $90,000 to $160,000 forthree-bedroom family units.

Land costs generally will account for 15to 30 percent of the cost. In most markets, landvalues are established by the more dominant con-dominium market (ownership apartment) andrental developers typically are unable to com-pete against condominium development for land.Land is zoned based on use (e.g., residential) anddensity (units per hectare or as a ratio of the totalsite area); land cannot be zoned for tenure. Pro-vincial enabling legislation to empower munici-palities to bonus densities specifically for rentaldevelopment could help to address this situation.Such a provision could increase density for rental(compared to condominium) use so that the netresidual value of the land, per unit, would equateto the values created by lower density condo-minium development.

Labour and material costs are establishedin a competitive market place. Currently, abooming construction industry in many parts ofthe country has resulted in shortages of bothlabour and materials with an inevitable upwardpressure on prices. An affordable housing stra-tegy with a long-term vision might seek to directassistance on a countercyclical basis, supportingdevelopment in markets that are in a downturnto take advantage of lower input costs and to helpoffset the costs of an underutilized labour force(e.g., lower income tax revenues and higherEmployment Insurance payments).

Finally, the trend over the past twodecades has been toward a pay-as-you-goapproach to funding the cost of municipalinfrastructure related to new development ratherthan covering this expense from general pro-perty tax revenues spread across all existingdevelopment. This practice has resulted in theimposition, by municipalities, of development


cost levies on new construction. Concurrently,the GST added another new cost to developmentand has carried provincial sales taxes with it inthe Atlantic region where federal and provincialconsumption taxes are harmonized.

advantages

Measures to lower land costs for rentalhousing and development charges can have acritical impact in reducing the cost of newdevelopment. A number of municipalities alreadyhave waived development fees for certain hous-ing types and, in certain areas such as the innercities, as a mechanism to encourage residentialdevelopment in the core and minimize suburbansprawl.

Reduced fees and charges will help toencourage private development and new supplywhile also marginally narrowing the developmentcost gap faced by nonprofit organizations seek-ing to build.

weaknesses

Without other complementary measures,these cost-reducing approaches will not lowercosts sufficiently to enable development to beviable at affordable rents without some subsidy.Even with free municipal land, nonprofit devel-opment for lower-income households is notviable without additional subsidy.

Requiring municipalities to lower orremove development costs eliminates municipalrevenue and requires the municipalities alone toshoulder the burden of new development, whilethe province and federal government gain incometax revenues from the construction labour andongoing operations (if for-profit).

Encouraging lower cost forms of development

Another way to address the relativelyhigh cost of new development is to encouragealternate, lower cost building forms.

Two such options are secondary suitesin single detached homes (which are alreadybeing created but often in contravention of localbuilding regulations) and small unit or ‘singleroom occupancy’ (SRO) accommodation.

The formal approval and explicit encour-agement of secondary suites is a controversialoption in many municipalities. Such suites havebeen created in basements or by subdividing theupper floors of existing homes. Distinct from arooming house, secondary units typically arecreated in a home in which an owner occupantis also resident so there is close management oftenants (unlike an absentee landlord). The mostbothersome issue from neighbours’ perspectivesis the absence of on-site parking and the result-ing use of limited street parking space. The con-cern from municipalities is one of health andsafety. Often suites are self-built and have notbenefited from a permit process that requiresinspections to ensure that the units are safe(especially regarding electrical wiring).

Some municipalities have sought tolegalize secondary suites, but few have expli-citly promoted this use either through encourag-ing statements in official plan policies or throughsmall grants to existing owners. A demonstra-tion project under Homegrown Solutions (a fed-erally funded seed grant program administeredby the Canadian Housing and Renewal Associa-tion) did provide funding to the town of Sidney,BC, for a small homeowner grant of $5,000 tohelp offset installation costs of an adaptable suitefor a person with physical disabilities.


The SRO approach involves developmentof very small bed-sitting rooms, which gener-ally range in size from 150 to 300 square feet (atypical bachelor/studio suite is usually 350 to 400square feet). Well-designed small suites can pro-vide cost-effective options for urban singles andhave become popular in a number of US cities.Given that almost half of core need householdsare low-income single individuals, this buildingform potentially could respond to significantdemand.

SRO type development can be cost-effective since the total floor area per suite ismuch less than that required in a typical bach-elor or one-bedroom apartment. Few residentscan afford to own a car, so parking requirementsand costs can be reduced when located in areasserved by public transit if municipalities agreeto this regulatory relaxation. A study for theOntario Ministry of Housing in 1999 found thatsuites could be developed in large urban centersat a cost of 40 to 50 percent of the typical newone-bedroom unit.

advantages

Alternate building forms such as second-ary suites (apartments within homes) and SROunits have significant cost advantages and caneffectively augment more traditional buildingforms.

These approaches could stretch any lim-ited subsidy funds further since per unit costsare lower.

weaknesses

Regulatory barriers and neighbourhoodresistance (‘Not In My Backyard’ or NIMBY)may limit opportunities.

There are few proponents pushing thisapproach. It is a new concept and, as such, evenmore risky than conventional rental, for whichthere remains limited investor appetite.

Shifting patterns of ownership and facilitatingnot-for-profit ownership

Much of the effort of the nonprofit hous-ing sector is focussed on building new housing.Without subsidies, the costs are prohibitivelyhigh and, even with subsidies, the subsidyrequirements per unit remain significant.

In other sectors, where poverty similarlylimits access to necessities, the solution is to seeklower cost options. Food banks provide food atno cost, and used clothing stores make availableclothing at affordable prices. If lower-incomehouseholds can afford a vehicle, or must haveone for transport to work, most lower-incomehouseholds also drive older used cars. Whilemany households seek housing in the olderexisting rental stock, the practice of drawing uponolder depreciated assets is not frequently utilizedby nonprofit providers, even though a large hous-ing stock already exists and properties often areprovided for sale – typically at values far belowthe cost of building new.

This dismissal of property acquisition asan option is, in part, a legacy of former nonprofitprograms. A number of nonprofit housing pro-viders did pursue acquisition/rehabilitation, andtypically expended more than the cost of build-ing new because the properties were in seriousstate of disrepair. Few providers went out to finda reasonably sound 20- to 30-year-old apartmentthat needed merely minor repair and painting.The second factor that acted to preclude thisoption was that the social housing funding pro-grams required all tenants to be in core housing


need, with incomes below specified thresholds.It was problematic to acquire buildings thatalready were occupied with some tenants not incore need, as this would require evictions. Soproviders avoided this option.

Far more multiple unit rental propertiesare sold each year than the number of nonprofitunits that were constructed by the programs ofthe early 1990s. Although not all properties willbe appropriate, due to poor condition or highprice, a significant number of existing proper-ties are regularly transacted (and this could beincreased if rollover provisions of the tax sys-tem were revised as discussed under a separateoption above). Tax reform would encourageexisting owners to sell to nonprofit buyers with-out incurring significant income tax impacts andwould provide a source for selective acquisitionby nonprofits.

Currently, as these properties are offeredfor sale, new private investors (including insti-tutional investors and Real Estate InvestmentTrusts or REITs) are the purchasers. Typically,new investors upgrade the property with theintent of increasing the rental cash flow andthese acquired properties move beyond afford-able levels. The alternative – purchase by non-profit organizations – effectively can preserveand potentially expand the remaining stock ofhousing that is relatively affordable to lower-income households. With nonprofit providers,access to lower-income households would beimproved as they no longer would be competingagainst ‘better’ higher-income tenants.

Any such nonprofit acquisition willrequire some subsidy, ideally a capital grant tofacilitate down payment. But because the totalcost per unit typically is much lower than build-ing new (40 percent to 50 percent lower), thegrants can be more effectively used and stretchedto secure more units.

In lower cost markets – such as Prairiecities – existing detached homes can be acquiredfor as little as $20,000 to $30,000. Some com-munities are exploring options for purchasing aportfolio of existing homes as a rental portfolioinstead of constructing new buildings.13

This acquisition approach also can pro-vide opportunities for lower-income householdsgradually to become homeowners. A criticalelement of an assisted ownership program iscounselling and ongoing mentoring as well aslinkage to employment and human resource train-ing to enhance employability. Despite low housecosts, these lower-income households generallywould be unable to qualify for a mortgage, eventhough the ongoing ownership carrying costsmay be less than paying rent on an apartment.With modest assistance – such as nonprofitmentoring, a loan guarantee and a modest downpayment grant of up to $7,500 – low-incomehouseholds can be assisted in obtaining a homeof their own with no ongoing subsidies.

advantages

The primary advantage of an acquisitionapproach is the relative cost effectiveness of theproperties, compared with new construction.

Because the properties already exist, pur-chasers do not have to contend with NIMBY typechallenges.

Acquisition permits the retention of anexisting mix of market tenants while slowlyintroducing lower-income tenants on unit turn-over. Acquisition with existing tenants remain-ing can effectively facilitate a mixing of incomewithout added cost (since market rents coverbreakeven rent).


Homeownership and scattered rental port-folios offer options to access existing rehabili-tation programs to upgrade dwellings.

weaknesses

There appears to be reluctance on the partof many providers in the nonprofit sector to pur-sue acquisition. Most prefer new construction,as they are concerned with adding to the supply.

The acquisition option is limited by theavailability of properties offered for sale on themarket and requires careful selection and assess-ment of potential properties.

The option is well suited for market down-turns. The best window for an acquisitionapproach in many cities already may have passed,as currently the rental market is under high pres-sure from low vacancies and rising rents (whichtend to raise the value of existing properties forsale).

Demand Measures

The brief overview of housing needhighlighted the predominance of affordabilityproblems. Low-income households cannot findhousing at low enough costs (rent) to match theirlow income and 90 percent of core need problemsare related to affordability. In 1998, more than1.5 million renter households paid more than 30percent of their income for housing and roughly500,000 paid more than 50 percent [StatisticsCanada Survey of Household Spending].

While supply approaches can play animportant role in moderating the market pressuresthat exacerbate affordability problems, supplyinitiatives cannot be implemented on a sufficient

scale to tackle the large backlog of these pro-blems. Some form of rental assistance is neces-sary to address the affordability gap.

This section describes two alternativeprogram approaches to assist households facingaffordability problems – rent supplements andshelter allowances – as well as a variant on shel-ter allowances applied to existing welfare bene-fits.

Rent supplement

A rent supplement approach involves anagreement between a public funding agency anda landlord in which the landlord agrees to enterinto contracts to provide rental units for low-income tenants on specified terms. Usually, thetenant’s out-of-pocket rent will be based on arent-geared-to-income basis. The contract willmake up the difference between the tenant-paidrgi rent and the actual market rent on the units.During the term of the agreement, there may bean inflationary index to allow the market rent toincrease annually, increasing the landlord’s rentalincome but leaving the tenant protected at thergi rent. As assisted tenants leave, during theterm of the agreement, the funding agency hasthe right to place a new low-income tenant intothe unit.

Rent supplements are dependent onsecuring the interest of private landlords. How-ever, there is a poor history of rent supplementsuccess in most provinces, despite efforts adver-tising for landlords to participate. Even for pro-perties for which rent supplements have beennegotiated, many landlords have elected not torenew these contracts when terms expire. Initialterms for agreements in the 1970s were gener-ally for 15 years with renewals now usually on a3- to 5-year term.


In cities with low vacancy levels, whichcurrently predominate in much of the country,landlords generally can fill their buildings withprivate market tenants, without the difficultiesrelated to having their tenants selected by a thirdparty. In addition, administrative requirementsassociated with this approach generally are seenby landlords as a deterrent. Thus, there may notbe significant opportunities to use and retain rentsupplements as a means of providing assistanceto low-income tenants living in privately oper-ated rental housing.

While there are difficulties in applying arent supplement program in the private market,there has been a long tradition of stacking rentsupplements on nonprofit projects.

A number of existing older nonprofitprojects are not rgi based, but include unitswith rents slightly below market levels. Thisapproach facilitated either by a capital grant ora favourable mortgage rate that helped to lowerthe breakeven rent below market levels. Allo-cating rent supplements to these units to allowrgi assistance to tenants with very low incomescould provide a low-cost rent supplement option;over time, breakeven rents will not rise asquickly as market rents.14

A further option is to utilize a rent sup-plement program in conjunction with nonprofitacquisition of existing properties. This option isdiscussed in the next section on preserving andexpanding the affordable housing stock.

advantages

This option addresses affordability pro-blems.

A rent supplement agreement is specificto contracted units, so thier condition and qual-ity can be verified.

It can be stacked on nonprofit supply butfocussed only on the affordability issue (sepa-rate from supply).

Over the long run, it is more cost-effec-tive to stack rent supplements on nonprofit hous-ing as the annual rate of increase in nonprofitbreak-even costs has been found to inflate moreslowly than market rents.

weaknesses

A rent supplement does not address lackof supply.

The options is dependent on a willinglandlord and available units – but few landlordsare interested.

Lack of renewal at term can result in dif-ficulty for existing tenants.

Historically, there has been low landlordtake-up of rent supplements due to administra-tive burden and to restricting use and rent on thecontracted units.

Shelter allowance

Unlike a rent supplement in which thelandlord is implicated directly in a formal agree-ment, shelter allowances are direct payments tothe tenant and, as such, overcome the necessityto negotiate agreements with landlords.

A feature of shelter allowances, asopposed to both rent supplements and socialhousing supply programs (in which the numberof units is limited by the size of the existing stockand any new production), is that all tenants whoare in need and eligible for assistance potentiallycan receive it. There would be no waiting lists(and unmet need) with shelter allowances. This


option also implies a potential expansion of bene-ficiaries and higher levels of total expenditure.

With a shelter allowance, the assistanceis formula-based; it takes into account bothincome and market rent for the unit.15 Maximumlevels of assistance (i.e., up to a maximum rentlevel) can limit overconsumption (i.e., the riskthat a household will choose a unit that is largerand more costly than necessary) but also are usedto constrain program benefits and manage theoverall budget.

The structure of the assistance formulais the primary means of public cost control. Byencouraging tenants to select units with lowmarket rents, costs are minimized for both thegovernment and the tenant. By contrast, rentsupplements and social housing provide a sub-sidy to reduce the tenant’s rent to 30 percent ofincome, regardless of the quality or amenity ofthe unit occupied or the associated market rent.One very low-income household may live in a40-year-old apartment while a similar householdmay receive a new townhouse.

An immediate concern with respect toadoption of a potentially major new program(such as a universal shelter allowance) is thescope and cost of the program: Potentially asmany as 1.5 million households might be eligi-ble. A number of approaches are available tomanage this problem.

The earlier analysis highlighted thesignificant number of renter households iden-tified as being in core need that receive socialassistance (including a shelter component).Since social assistance already includes a shel-ter allowance, it would be most practical to sepa-rate welfare recipients in this analysis. Theoptions for welfare households are either toreform the existing shelter benefits within wel-fare (which are already conditionally linked to

actual shelter costs up to a maximum) or to limitsocial assistance to income support and addressshelter needs under a separate but linked shelterassistance program.

A review of the 1996 HIFE databasereveals that, among 1,172,000 core need house-holds, seniors account for about 240,000 house-holds, welfare cases number 530,000 and theworking poor account for the remaining 400,000.Excluding welfare recipients, the target group fora new shelter allowance program would beroughly 640,000 households (55 percent of the1996 core housing need population).

A recent costing analysis of a shelterallowance initiative (undertaken for the Federa-tion of Canadian Municipalities National Afford-able Housing Strategy) estimated the cost of anincremental shelter allowance program, assist-ing 40,000 households annually. The first-yearcost for 40,000 households was determined tobe between $30 and $60 million. To address theneeds of 400,000 working poor householdswould therefore cost in the region of $300 to $600million annually. Adding the 240,000 seniors tothis estimate would raise costs to between $500and $950 million annually.

The wide range of the estimate reflectstwo differenct benefit formulas. The lower costestimate reflects a program targeted only tohouseholds in severe need, paying more than 50percent of income on shelter with benefits gearedto lower this rent burden to roughly 37 percentof income. The more expensive option reflectsa benefit formula that lowers household sheltercosts to 30 percent of income, equivalent to theassistance levels in social housing.

The actual costs of a shelter allowanceprogram could be reduced by a variety of means,including restricting potential client groups(e.g., welfare recipients), phasing in the benefits


or using lower levels of benefit. Initially, eligi-bility criteria could be relatively restrictive andbenefits could be established at minimum levelsto address very high shelter cost burdens (e.g.,paying less than 50 percent). Gradually, withexperience in program takeup, the criteria couldbe relaxed and benefits enhanced to levels thatallow tenants to obtain adequate accommodationat a maximum of 30 percent of income, whilestill ensuring that public program costs are man-aged and contained.

Although shelter allowances typically areseen as entitlement programs, with broad eli-gibility, it is possible to limit enrollment. Forexample, the approach used in the US involvesusing certificates or vouchers, which are limitedin number. Like nonprofit housing, there wouldbe a waiting list. Households are required toreapply annually to ascertain continued eligibil-ity. When households currently served by vouch-ers are no longer eligible (due to improvementsin income or success in finding lower cost hous-ing), the voucher can be reallocated to a house-hold on the waiting list.

advantages

The option directly addresses the afford-ability issue and helps to ease very high shelterburdens.

The allowance can be broad based orrationed.

The benefit formula allows the mechan-ism to be targeted and to provide varying levelsof assistance to differing target groups and house-holds facing more severe affordability problems.

Used in combination with nonprofitsupply, a stacked shelter allowance reduces thechallenge for new nonprofit supply, as it isnecessary only to get breakeven rents down tomarket levels.

weaknesses

The option does not address low levelsof supply, and subsidy costs are impacted byinflating rents (caused by low supply relative todemand).

It does not create a long-term asset.

The allowance may be perceived bypoverty advocates as a benefit to landlords bysubsidizing private sector rents, with no long-term retention of units.

Reform of welfare shelter benefits – A trans-itional shelter allowance

Almost half of core need households andthose facing high shelter-income ratios aresocial assistance recipients. Various data sourcessuggest that just under half of core need house-holds and over half of households spending morethan 50 percent of income for rent in 1996 werein this category. This points to a need to exam-ine the adequacy of welfare benefits – at leastthe portion linked specifically to shelter assist-ance.

The shelter component of welfare is notobjectively related to actual rental costs and isnot indexed to the cost of living. During the firsthalf of the 1990s, this practice was not a pro-blem. But with very significant levels of rentincrease in the last few years in a number of cen-tres, the calculation of the shelter component willbecome an increasingly serious issue.

Using the illustrative case of a single-parent family, the accompanying chart shows thatthe maximum allowance for shelter falls far shortof average market rents. This shortfall forcesrecipients to seek units renting well below theaverage level – a difficult challenge given the


play in this process. Other than existing socialhousing (where there is only a finite stock andlong waiting lists), there are currently no transi-tional mechanisms that enable households tomove off welfare, accept low wage work and stillbe able to afford to pay their rent.

Welfare households considering work arediscouraged because as soon as they leave wel-fare, they lose their shelter benefit. Becauseshelter costs account for such a large part ofthe household budget, welfare reforms withoutongoing housing assistance are less effective. Ashelter allowance, separate from welfare, can bea valuable complement to other initiatives toenable households to move back into the labourforce.

Providing ongoing support to assist withthe rent provides greater stability for low-incomeparents with children, helping to avoid thearrears and evictions that often cause families tomove continually and disrupt children’s devel-opment and schooling.


previously noted crowding-out effect whereinhigher-income working tenants tend to have bet-ter access to these lower rent units.

Clearly, there is a need to realign intendedshelter assistance benefit levels with true sheltercosts, and ideally to index benefits to actual mar-ket rents (e.g., the median or 40th percentile).

In addition to addressing the inadequacyof current welfare rental assistance levels, itwould be effective to link revised benefit levelsto ongoing welfare policies that seek to encour-age and facilitate the transition from welfare towork.

A shelter allowance for the working poor(as outlined in the previous section) could beadapted to create a transitional benefit for wel-fare households. While most provinces and ter-ritories have implemented reforms intended tohelp welfare recipients acquire skills and workexperience to make the transition back to work,these approaches have not recognized the criti-cal role that stable and affordable housing can

Table 4Welfare Shelter Maximum, Market Rents and Affordability Gap

Illustrative case of lone parent with one child less than 5 years old,selected provinces, 1999

BCManitobaOntario

1999 rents averaged for a 1- and 2-bedroom unitWelfare maximums from provincial ministries, 1999 rates

Max. ShelterComponent of

Welfare

$610$387$511

AverageMarket

Rent

$778$521$851

WelfareShelter Max.as % AverageMarket Rent

78%74%60%

AffordabilityGap (rent

less allow-ance)

$168$134$340

Source: CMHC rent survey.

In the case of families with children, amodest shelter allowance in combination withminimum wage work and the new Canada ChildTax Benefit can leave a household with moreincome than it would receive on welfare, at a costto government significantly lower than the costof full welfare benefits.

Typically, shelter allowances are designedto provide relief against excessive rental costswhile leaving an incentive for the household toeconomize, limiting program costs. As discussedin the previous section, this is achieved by a for-mula that covers part of the gap between a speci-fied percentage of income and actual rent, up toa rent maximum.

The precise formula can be adjusted toensure that the recipient does not pay more thana specified percentage of income for rent. Theformula can be varied by household size andcomposition, since larger households have higherfood budgets and also must pay higher rent forlarger units. A shelter subsidy of roughly $150to $250 per month, depending on the market,effectively can reduce a rent burden of 55 per-cent down to a more reasonable proportion ofincome (30 to 35 percent).

In the case of a person leaving welfare toreturn to work, such an allowance, costing gov-ernment less than $3,000 per year, would leavehouseholds in a much better financial positionthan welfare benefits that may cost governmentin excess of $12,000 to $15,000 annually. Thelevel of the shelter assistance would phase outas earned income improves. The phase-out canbe designed to be gradual and thus avoid theinherent tax back disincentives that often under-mine assistance programs when participants seekto earn an income.

advantages

A transitional and separate shelter allow-ance for social assistance recipients could ensurethat they are less at risk of losing their housingwhen moving off welfare into work.

Such a mechanism would support otherwork incentive policies.

The costs of this transitional initiativewould be significantly lower than ongoing fullwelfare benefits.

The shelter allowance formula is basedon actual rent and earned income and benefitsreduce as income increases. The slope of thereduction can be controlled through the formuladesign.

weaknesses

Over time, more households may remainon assistance and costs could expand – unlikewelfare, which tends to be more intermittent andtransitional.

Implementing this approach requiressignificant commitment to welfare reform.

Some Concluding Observations – Toward aComprehensive Housing Strategy

This overview has outlined the nature ofthe affordable housing challenge and describedan array of policy/program options. While thegeneral advantages and disadvantages of eachapproach are noted, this overview was notundertaken within a formal comparative frame-


work, which would evaluate each option againstcommon criteria. Such a formal evaluationrequires more precise details on the specificdesign of the alternative programs and associ-ated costs, which depend on geographic alloca-tions and respective market conditions.

It is important to stress that, alone, noneof the identified options is sufficient to addressthe full array of housing issues. The nature andmagnitude of housing problems vary geographi-cally and depend on prevailing local market con-ditions.

That said, the vast majority of problemsrelate to affordability – households spending toolarge a proportion of an already inadequateincome for shelter, leaving little for other neces-sities.16 Even households not experiencing anaffordability problem face the indirect conse-quences of an insufficient number of lower rentunits since some make a trade-off between priceand quality, sharing a dwelling with another fam-ily (overcrowding) or accepting poor conditions(inadequate housing). Lowering the cost of pro-ducing housing is not a realistic option; mucheffort has been expended on this approach andmost feasible options have been tried – savereducing the various taxes fees and leviesimposed by all levels of government. The great-est potential lies in approaches aimed at increas-ing ability to pay.

Improving the ability to cover sheltercosts for most households would be the fastestand most desirable way to reduce their sheltercost burden. This can be addressed through aconditional income transfer, linked to actualhousing costs – such as a rent supplement or ashelter allowance or simply through additionalincome. Increasing household income, eitherthrough supplements or through increasedearning capacity (e.g., higher minimum wages,

employment training) also would serve to reducehigh shelter-to-income ratios and leave moreincome for other necessities.

This conclusion suggests that income-based options outside of the traditional housingpolicy toolbox merit serious examination. Taxcredits or other means of improving income withsome conditional link to housing consumptionmust be explored. The most fertile option islikely the reform of the shelter component ofwelfare, since this subpopulation is highly over-represented among households in core housingneed.

Although affordability problems pre-dominate, simply tackling the demand side of theequation would not be sufficient. It would notaddress the lack of new supply that is the rootcause of rising rents and worsening affordabilityand which cause cost inflation in a demand-sidesubsidy. Nor would a demand measure alonecurb the ongoing erosion of the existing limitedaffordable housing stock outside of social hous-ing or improve access for those stereotyped as‘less desirable tenants.’

To control for inflating rents in anundersupplied market, some form of supplyresponse to support new construction is a neces-sary complement to any demand-side initiative.Any form of new construction will help easeupward pressure on rents, so measures to stimu-late private rental development can play animportant role.

By the same token, an exclusive nonprofitsupply option would be equally insufficient,mainly as it costs much more to address two pro-blems (supply and affordability) within the samesolution and because a very large scale produc-tion program would be required to meet currentneed.


It would be impractical to tackle the levelof outstanding housing need (in excess of onemillion households) entirely through a new sup-ply program. However, as population and thenumber of households grows, some new supplyis required to meet new need – much of whichderives from lower-income households with lim-ited effective demand.

If new supply were to be subsidized,nonprofit supply likely would generate betterpublic investment in creating a lasting asset andaddressing issues of access than stimulativesubsidies directed to private development. Atthe same time, impediments to private rentaldevelopment, particularly the tax treatment ofrental investment income, need to be identifiedand remedied if the necessary volume of newrental production is to be achieved.

This overview has not provided anydetailed comparative cost assessment. Theissue of relative cost-effectiveness is a highlycontentious one and is not examined in detailhere. However, any assessment of policyoptions should consider carefully this aspect.17

The gradual erosion of lower priced rela-tively affordable existing rental stock is perhapsthe most serious issue contributing to afford-ability problems. Measures to mitigate this phe-nomenon are critical. Combining the longer-termbenefits of nonprofit ownership (less inflation-ary pressure on rents and long-term preservationof affordable units) with the inherent lower costof buying existing rental properties – through an

active program to support nonprofit acquisitionof existing rental properties – may be the moreeffective way to expand the affordable stock thanbuilding new. The key advantage of the nonprofitsector is not its ability to build new affordablehousing. The benefit of this sector comes in theform of ongoing ownership and management ofassets. The long-term objective is to maximizeaffordability rather than return on investment.

While an acquisition strategy can help topreserve the existing stock of affordable hous-ing, other measures are required to enhance theability of renters to meet their rental costs and tobroadly stimulate new supply.

The key point here is that demand andsupply measures are not substitutes for each other– although each is preferred and promoted byparticular interest groups for differing reasons.They are complementary measures that, whencombined in a well considered comprehensivestrategy, can help tackle the persisting afford-able housing problem.

In responding to the current crisis of lowrental vacancy rates, it is critical that the federal-provincial/territorial housing ministers and theirofficials look beyond just a short-term, relativelysmall-scale new construction program. In par-ticular, the prevailing nature of the affordabilityproblem – inadequate income and ineffectiveincome assistance benefits – must be embracedas part of the solution. This implies the need fora broader dialogue including ministries respon-sible for income assistance.


Endnotes

1. Minister Gagliano meets with P/T Ministers Responsiblefor Housing. CMHC Press Release August 16th, 2001.The Ministers plan to meet again in November to discussprogress.

2. For example, a lone parent with two children of thesame gender aged 5 and 8 would require a two-bedroomunit. If the median monthly rent of a two-bedroom apart-ment is $640, the income required to afford this appropriateunit is $25,600 ($640*12 months/.30). Similar householdswith an income above this level would not be in core need.

3. By comparison, the number of renter households payingmore than 30 percent was counted in the Census at 1.67million. This method does not exclude households withincomes above the CMHC income thresholds. Becausethe CMHC core need method is more specific in applyingan income filter, it generates a lower number. CMHCreports a similar 1.66 falling below their three standards,but only 1.17 million of these were categorized as beingin core housing need.

4. Statistics Canada, The Daily, Cat. No. 11-001E,Ottawa.

5. The key issue here is that someone who receives anincome assistance benefit specifically to help with shelter(e.g., $615) may spend more than 50 percent of total benefitincome on shelter, even though rent might be within the$615 allowance. Under current CMHC definitions, thishousehold would be defined to be in core housing need.However, if this household can secure housing within theamount of the shelter allowance, it should be excludedfrom a count of housing need (or at least included in aseparate group defined to encompass welfare recipients,as opposed to the working poor).

6. Data from the National Council of Welfare. Profiles ofWelfare: Myths and Realities. Ottawa: Minister of PublicWorks and Government Services, 1998.

7. Data from Quantitative and Information AnalysisDivision, Social Policy, HRDC.

8. Another reason for the apparent decline is a shift inStatistics Canada survey methodology. Both the Censusand former HIFE datafiles used current year rent comparedto previous year income. The SHS has improved on thismethodology by collecting data on income and rent forthe same period. So comparison across the two datasources may be inappropriate.

9. Federal funding for new nonprofit housing commit-ments ended effective December 31,1993 and sinceprograms were cost-shared with provinces and territories,almost all new social housing development ceased. Ontariomaintained a unilateral program until it, too, was cut in1995. Only BC and Quebec have maintained modest socialhousing programs (the Quebec programs are more focussedon acquisition and rehabilitation than on new construc-tion).

10. One of the critical risks for potential investors is theregulation of the rental sector. While rent regulation hasbeen relaxed in a number of provinces, investors still fearreintroduction. It is not rent controls themselves that scareinvestors away, but the uncertainty about future regulatoryregimes that increases risk.

11. Mention should be made about the omission of rentregulation/control from this discussions. Rent controlsimpose specific limits on the level and frequency withwhich rents can be increased. In the context of seriousaffordability problems, caused in large part by rising rents,some advocates have pointed to the removal of rentcontrols as the cause of growing affordability problems.A regime of tenant protection continues to exist in allprovinces, but focusses more specifically on consumerprotection – defining the rights and relationship betweenexisting tenants and landlords. Many provisions of formerregimes that also sought to control rents have been revisedand largely eliminated in favour of this more specific levelof consumer protection. Proponents of the former controlspoint to their impact in preserving affordability; opponents(largely landlords) counter that controls are discriminatoryand place undue burden of public policy on one specificsector (rental landlords). Rent controls also have beenidentified as a key deterrent to new rental investment –although, arguably, it is the uncertainty associated with aregulatory regime rather than the controls themselves thatunderpin this deterrent. Overall, rent controls are a bluntpolicy instrument with undesirable outcomes, as arguedby opponents. More specifically, while controls may limitlevels of rent and annual increases, they cannot ensurethat this controlled resource is made available to theintended beneficiaries. In fact, controls benefit manyhouseholds that do not require assistance. The previousreview identified a substantial portion of the low renthousing stock occupied by households with incomestheoretically able to pay higher rent, while those most inneed continue to face severe affordability conditions.

12. The generic term ‘nonprofit housing’ is used in thisreport. This term can also include co-operative housing –a variant that seeks to assist lower-income householdsand operates on a not-for-profit basis – although the


management structure and philosophy are different(encouraging active resident participation).

13. Such options are not necessarily restricted to lowercost markets, although the viability is obviously greater.An analysis of real estate listing for one month in 1999 inBC found more than 1,000 dwellings (including duplexand townhome units) for sale outside of the two majorcities for under $90,000 – affordable to working poorhouseholds at a payment of less than $600 a month.

14. Previous research has examined the cost trends ofprivate rent supplements over a 20-year period comparedwith the comparable cost of stacked rent supplements innonprofits and found that, over time, nonprofit breakevenrents increase at a slower rate than market rents –suggesting a cost benefit to stacking rent supplements onproperties under nonprofit ownership.

15. For example, a shelter allowance may provideassistance equal to 85 percent of the difference betweenactual rent and 35 percent of income. So if income is$1,500 and rent is $600/month (40 percent of income), ashelter allowance may provide relief of [$600-($1500x35percent)]*.85 percent = ($600-$525)*.85 = $64. Thislowers net rent to $536 or 36 percent of income. This

formula can be adjusted so the net effect lowers the shelterto income ratio to 30 percent, equivalent to social housingrents. A maximum rent, say $600, might be imposed toprevent households selecting expensive units – althoughthere is already an incentive to seek a reasonably pricedunit, as the household pays a share of the higher rent.

16. Notably, Statistics Canada has recently published aresearch study on ‘food insecurity.’ This is defined as asituation where household members are forced tocompromise on the quality of their diet because they haveinsufficient income to cover the cost of basic necessities.High housing costs may be an important factor contributingto this phenomenon. See Statistics Canada, “FoodInsecurity in Canadian Households.” The Daily, August15, 2001.

17. The key tradeoff is between paying a premium fornew supply with the benefit of nonprofit ownership, a lowerrate of subsidy inflation and long-term availability of theaffordable unit, versus low initial costs but inflating subsidycosts in a shelter allowance as market rents inflate into thefuture. The comparable present value is very sensitive tothe initial cost gap and assumptions of rent inflation anddiscount rates.


Glossary of Terms

A measure of poor physical condition of a dwelling unit. A household has anadequacy problem when the dwelling occupied is in need of major repair and isin an unsafe condition.


Not defined as a specific income or rent, affordability refers to a relativesituation in which income is deemed insufficient to pay for rent. Typically, theaffordability benchmark used in housing analysis is 30 percent of income spent on shel-ter.

Difference between the amount a household can afford at a specified percentage ofincome (e.g., 30 percent) and the actual rent paid or market rent.

Used by some municipalities as a development incentive to encourage a specificoutcome, such as affordable housing. Bonusing refers to the practice of allowing a higherdensity on a development site in exchange for providing a public or social benefit. Insteadof a three-storey building, for example, a developer may obtain a bonus of one additionalfloor (or some percentage of allowable floor area) to four storeys with the condition thata specified amount of the floor area or units produced meet certain criteria.

The rent that would have to be charged to cover all operating costs and mortgage pay-ments related to the cost of building.

Canada Mortgage and Housing Corporation is the federal housing agency that imple-ments federal housing policy and programs.

A form of ownership tenure associated with a multiple unit building (apartment ortownhome). The occupant owns the individual unit and shares common space such asfoyer, halls and outdoor space.

A form of tenure, common in social housing, in which the residents are co-op membersand participate in the management and operation of the property. Much of the co-opera-tive housing that has been built in Canada is owned on a not-for-profit basis – the occu-pants have no or limited equity share. Some equity co-ops also have been built outside ofsocial housing programs.

A measure developed by the CMHC to determine housing need. See detailed discussionin section 2 of report.

The monthly or annual payments of principal and interest associated with a mortgageloan.

An approach that is focussed on the consumer/tenant with the objective of increasingability to pay rent and thus improving effective demand.

AdequacyProblem

Affordability

Affordability Gap

Bonusing

Breakeven Rent

CMHC

Condominium

Co-operativeHousing

Core HousingNeed

Debt Service

Demand Approach

The primary unit of analysis in housing research. The term household is usedgenerically and can include unattached individuals, a group of unrelated peopleoccupying the same dwelling, a family or a combination of these groupings.

The use of a small amount of capital to support or obtain the use of a largeramount. For example, using the future rental income to secure a mortgage loan isoften referred to as leveraging the mortgage. Using the value of existing assets tosecure a mortgage loan is referred to as leveraging the asset.

The practice of accommodating a range of households with different incomelayers. Income mix is seen as a way to avoid concentration of very low-incomehouseholds that can lead to stigmatization or ghettoization of certain projects andresidents.

An incorporated association with the objective of operating without gain orprofit for members or board members. Much social housing in Canada is oper-ated by not-for-profit agencies or corporations. These may be a private corpora-tion, such as a subsidiary of a faith group or service club, or they might be a publicnonprofit corporation – an arms-length subsidiary to a municipality incorporatedspecifically to own and operate housing for low- and moderate-income house-holds.

A nonprofit corporation that owns and manages housing (a label for a nonprofitlandlord or co-operative).

Housing owned and operated by a provincial or municipal agency – formerlyused as a generic term for assisted housing, now largely replaced by the moreencompassing term ‘social housing.’ See below.

A form of assistance paid to a landlord to provide units to low-income tenantsusually over a contracted term. Payment to landlord is based on the differencebetween actual negotiated market rent and a rent geared to income (rgi) paid di-rectly by tenant (typically 30 percent of income).

Rent-geared-to-income (usually based on 25-30 percent of gross income, as veri-fied by the administering agency).

A form of assistance paid directly to tenant, based on the difference betweenactual negotiated market rent and an rgi rent paid by the tenant. Shelter allowancepayments are calculated by formula and often pay only a percentage of the gap(e.g., 75 percent of the difference between full market rent less 30 percent ofincome).

Portion of income assistance calculation that uses an explicit shelter variable.Typically, this component represents one part of the income assistance and is paidup to a maximum amount (varying by household size and composition) based onactual verified rent.


Household

Leverage

Mixed Income

Nonprofit

Provider

Public Housing

Rent Supplement

Rgi

Shelter Allowance

Shelter Component/Shelter Maximum

Popularly known as ‘welfare,’ social assistance is the income support of last resort forpeople who do not qualify for or have exhausted other income programs and other sourcesof income.

Combination of two different types of assistance or programs. For example, a rent sup-plement may be stacked with nonprofit supply assistance to ensure that rents are afford-able at a specified percentage of income.

Shelter to Income Ratio.

Term used to define an overcrowded dwelling – CMHC used very specific criteria todetermine an appropriate unit size – based on national occupancy standards. Parentsand children have separate rooms and children over age 5 of opposite gender areassigned separate rooms; those under 5 may share up to two per room.

A program response focussed on providing a dwelling – usually involves new construc-tion but may include acquisition and rehabilitation.

The difference between breakeven rent and market level rents. The supply gap reflectsthe cost or subsidy amount necessary to stimulate or support new supply (since typicallybreakeven rent on new construction exceeds potential market rent).

Specifying the maximum percentage of units that must be made available to householdsmeeting certain income criteria. Public housing is 100 percent targeted – all tenants aregenerally below income thresholds and receive subsidized rent. Some nonprofit devel-opments cap the number of targeted households at 60 percent to encourage a degree ofincome mix and avoid concentrations of very low-income households.

A measure of the availability of rental units. Distinct from a turnover rate (when onetenant moves out and a new one is ready immediately to move in), the vacancy rate (asmeasured and reported in the CMHC annual rental survey) measures the percentage ofunits that are vacant and available for rent as of a specific date. Typically, a vacancy rateof three percent is accepted as a benchmark of a healthy rental market (sufficient units tomeet demand without excess pressure on rents).

A household whose main source of income over a continuing period is social assistance.

Households in which one or both adults are employed, but family income is belowStatistics Canada low income cut-offs.


Social Assistance

Stacking

STIR

Suitability Problem

Supply Approach

Supply Gap

Targeting

Vacancy Rate

WelfareDependant

Working Poor

Toward a Comprehensive Affordable Housing Strategy for …€¦ · · 2003-10-06Creating a level playing field for rental development 13 ... a household paying more than 30 percent

Documents