NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Institute of Medicine (US) and National Research Council (US) Committee on an Aging Society. The Social and Built Environment in an Older Society. Washington (DC): National Academies Press (US); 1988.

Cover of The Social and Built Environment in an Older Society

The Social and Built Environment in an Older Society.

Show details

Current and Emerging Issues in Housing Environments for the Elderly

Raymond J. Struyk

The dominant issue in the debate surrounding the housing environment of the elderly is how to construct public assistance to support those housing transitions that are necessary to allow community-based housing to become an active and integral element in the overall long-term care system.

Two attributes in particular should affect the way we think about housing interventions for the elderly population. First, as people reach retirement age, they experience numerous changes, sometimes in rapid succession: incomes fall from preretirement levels, children leave home, health problems and activity limitations emerge, a spouse requires institutionalization or dies. Such a variety of life events means that any public policy to help with housing problems of such individuals must be flexibly designed; it should range from rent supplements to counseling homeowners about various housing options to the provision of support services to compensate for the inability to perform key activities of daily living.

Second, the elderly cannot be viewed as a monolithic group. As has just been suggested, they have widely different housing-related problems. What is at least as important is that individuals in the elderly population differ in three fundamental ways that must be taken into account in designing public policy: (1) health status (including activity limitations), (2) economic resources, and (3) householding status—whether they are homeowners or renters. Again, the resultant emphasis is on policies that are flexible enough to accommodate the elderly in these various circumstances.

The balance of this essay is in six main sections. It begins with a general inventory of the housing problems of the elderly that considers dwelling-specific items as well as the need for supportive services. The second section reviews current federal housing assistance programs for the elderly; the third presents a framework for thinking about housing policy that goes considerably beyond the present programmatic structure. The fourth section describes three specific federal housing interventions that could meet the requirement of being an element in a long-term care system. The next section gives some idea of the possible consequences of adopting such interventions, and the final section provides conclusions.

Housing and Housing-Related Needs

This section, which draws heavily on an earlier paper (Struyk and Katsura, 1985), inventories the current housing needs of the elderly to provide an idea of the extent of the need for various types of housing assistance. The inventory is not extrapolated into the future because making such estimates is a highly complicated task, the outcomes of which are dependent on broad economic and housing market developments, trends in pension payments, and changes in the incidence and severity of health problems and activity limitations.

In considering the housing needs of the elderly, it is useful to make a distinction between traditional housing "problems" and the particular needs that arise from the health problems and activity limitations that are frequently the lot of elder citizens. The housing problems of the elderly (called dwelling-specific problems hereafter) include deficiencies in the dwelling, the high price of housing relative to income, and overcrowded conditions. All of these are problems that can be measured in fairly straightforward ways and whose definition does not generally have a special dimension for the elderly. Other, more general aspects of the housing environment—for example, conditions in the neighborhood, the convenience and quality of local shopping and medical services—are also relevant. These latter aspects are not as well measured in the available data, however, and are not treated further here.1

Housing problems associated with activity limitations or so-called dwelling-use problems are much less precisely defined. Indeed, activity limitations, which are used here as a shorthand label for the larger set of health-related problems, are better thought of as an indicator of a potential housing problem. Limitations on the activities of the elderly can mean that they are unable to use their dwelling fully. For example, they may be unable to use the kitchen and bathroom without assistance (possibly because the rooms are inconveniently located in relation to living and sleeping areas), clean and maintain their home properly, or go shopping without help. On the other hand, these limitations may be effectively offset by the assistance provided by other family members or neighbors or by modifications to the dwelling itself. Unfortunately, the only general measures of housing needs arising from activity limitations focus on the limitations themselves and not on the services the household must do without because of them.

Thus, to determine the number of elderly persons with housing-related problems, we must combine reasonably rigorous estimates of traditional dwelling-specific problems with less direct estimates of dwelling-use problems. One major difficulty—this time because of data limitations—is calculating the joint occurrence of dwelling-specific and supportive service needs. The apparent needs for support services are sharply reduced by the assistance (intervention) of family and friends.

Dwelling-specific Needs

Our focus in this section is on the incidence of physical deficiencies and excessive housing expenditure burdens in 19792 (Table 1). Moreover, we limit the population considered to those households that in 1979 were not participating in federal or state housing programs—some 14 million elder-headed households.

Table 1. Incidence (percentage) of Housing Deficiencies and Excess Expenditures by Population Group, 1979.

Table 1

Incidence (percentage) of Housing Deficiencies and Excess Expenditures by Population Group, 1979.

Among the elderly, a fairly clear ranking emerges, running from those with the worst housing situation to those with the best. Impoverished renters and impoverished owners with mortgages are at the low end of the ranking, and nonpoverty owners with mortgages and without mortgages are at the higher end. Differences by location exhibit a familiar pattern; the incidence of deficiencies rises steadily as we examine successively more rural locations. In addition, this pattern holds across all tenure groups. The incidence of excessive expenditures is more varied but generally tends to be lower in rural areas.

The relative disadvantages of black households are strikingly clear. Their units continue to exhibit extremely high levels of deficiencies, and the incidence of excessive expenditures is also higher for black than for other households, especially among renters. The differences among black households are generally small, however, in comparison to the divergence in dwelling deficiency rates between the races. Finally, although it is not shown in the table, it is worth noting that there is little difference in the rate of deficiencies among the elderly aged 65–74 and those 75 years of age and older (Struyk and Soldo, 1980, Table 3–6).

To summarize, in 1979 there were about 1.61 million elder-headed households in dwellings that would be characterized as physically deficient and about 2.58 million households spending an excessive share of income on housing.3 Because only about 340,000 of the households have these problems in common—meaning that many are spending a large fraction of their incomes to live in decent housing—a total of about 3.85 million, or 28 percent of all elder-headed households, have a dwelling-specific housing problem. The incidence among those below the poverty line is much greater: of the 2.66 million elder-headed households in this group, 61 percent have at least one of these problems.

Dwelling-use Problems

Although it has long been recognized that those whose activities are limited by health problems or disabilities are less able to function effectively in their homes without assistance, national housing policy has accepted this fact only to a limited extent. In considering policy options for helping those persons with such problems, clearly, we must know the size of the population that needs assistance. A key point to note at the outset of this discussion is that dwelling-use problems can be alleviated by supportive services, by modifications made to the unit that facilitate its use, or by both of these methods.

Following are two estimates of the number of households with dwelling-use problems; these estimates are intended to bracket the actual number of those needing help. The more generous definition is one that counts as needing assistance all of those who have a functional impairment as a result of disability or health problems. Applying this type of criterion to data from the 1979 National Health Interview Survey, we find that about 12 percent of persons age 65 and older have a need for some form of supportive services in their homes; the figure is 7 percent of those aged 65–74 and 21 percent of those aged 75 and older. If we apply the same rate to elder-headed households, about 2 million households are in this category.4

The incidence of need defined in this way is greater for women than for men (in both age groups), greater for blacks than for other ethnic groups, and apparently (because it is possible that many of those with low incomes in 1979 had spent their way down to this level through expenditures for medical and supportive care) greater for those with lower incomes.

A more conservative (and possibly more accurate) estimate of the number needing supportive services can be obtained if we look at the share of those who have a functional limitation and who are receiving formal care services—that is, services provided by an agency, whether they are paid for by the recipient or not. This type of calculation has the advantage of deleting those who receive essential services only from family members, neighbors, and friends. Nationally, about 25 percent of the elderly who report a functional limitation are receiving formal services. Applying this rate to the 2 million households noted in the earlier paragraph yields about 500,000 households who require support services provided by a formal agency.

Yet this figure is probably too low—for two reasons. First, it is virtually certain that not all of those who need such services are receiving them. Second, some persons are now in long-term care institutions who would not be there if such services had been available to them. The evidence for this last statement appears in the analysis of the determinants of institutionalization. Those elderly persons who live alone are institutionalized at higher rates than the elderly in multiperson households, even after controlling for health status and activity limitations (Weissert and Scanlon, 1983).

All in all, we might take as fairly accurate an estimate on the order of 750,000 elder-headed households that need formal supportive services. Additionally, half again that number now need and receive informal services, either from sources within the households or from outside. Public policy should be so structured as to complement informal services rather than replace them.

It is also important to note that, when we examine the determinants of the likelihood of a person receiving formal supportive services, the dominant factors are the extent of the person's disability and the absence of informal services. After controlling for these conditions, income by itself is not an important factor, which suggests that, over some range, public programs and informal assistance are reaching many of those in the greatest need of supportive services (for details, see Soldo, 1983). Thus, service recipiency seems to be largely determined by incapacity, a lack of informal assistance, the availability of formal services, and, in some cases, the ability to pay for them.

In general, then, the patterns of dwelling-specific and dwelling-use needs are quite different. Whereas dwelling-specific problems are strongly related to income and little associated with age, dwelling-use problems are related to age and physical impairments but not particularly to income.

Dwelling Modifications

The need for some types of supportive services can be eliminated by various changes to an elderly person's dwelling, changes that can compensate for particular functional impairments. Such modifications range from the installation of grab bars and easy-to-grasp doorknobs and other hardware to specially equipped telephones to bathrooms and kitchens that have been remodeled to accommodate wheelchair use. In other cases, these changes can reduce the need for supportive services and thus complement their provision. The best estimate of the probable need for modified dwellings—beyond those already occupied by some 700,000 elder-headed households—is on the order of 1 million units (Struyk, 1982). As indicated earlier, these households are not in addition to the number of those needing some type of support services. If we assume that the needed modifications are concentrated among those with the greatest impairments (who are also most likely to be receiving formal supportive services), then approximately 250,000 households that are not also receiving formal supportive services need to occupy units with some special features. (Unit modifications are discussed further in a later section.)

Overlap Between Dwelling Problems

Newman (1985) has used 1978 data taken from a supplement to the annual housing survey to estimate that about 17 percent of the elder-headed households with a person having an activity limitation reside in a unit that is physically deficient. (Newman uses the same definition of dwelling deficiencies used earlier in this paper.) Note that this rate is substantially higher than the 10 percent rate for elder-headed households with no members with such limitations, suggesting that households with an impaired member have greater difficulty maintaining or affording decent housing.5 This rate implies that in 1979 there were some 340,000 households in the group with both dwelling deficiencies and dwelling-use problems.

Similar calculations can be performed for the overlap between those households with excessive housing expenditures and those with a member with an activity limitation. This calculation yields an estimate of 540,000 households with the combined problems.6

Summary

The figures in Table 2 summarize the information compiled on the number of elder-headed households with various housing-related needs. As implied earlier in this section, these are order-of-magnitude estimates designed to give a general picture of the current situation. The first point is that those households with dwelling-specific needs far outnumber those with dwelling-use needs, which points to the necessity of continued action in this area. An encouraging point is that probably less than a million households are characterized as having both dwelling-specific and dwelling-use needs. This figure is only about 6 percent of all elder-headed households in 1981, suggesting that it is a group for which assistance should be possible.

Table 2. Summary of Housing Needs of Elder-headed Households, 1979.

Table 2

Summary of Housing Needs of Elder-headed Households, 1979.

On the other hand, on the order of 5 percent to 14 percent of elderly households do have dwelling-use needs (or up to 9 percent of the population without dwelling-specific needs). These households are prime candidates for the shifting of elderly persons into institutions.

Finally, a cautionary note: these figures are for 1979. The sharp increases in the number of elderly that will occur in the years ahead, as well as the greater share of the older and more frail in this population, is well known and should be kept in mind when considering possible policy interventions.

Current Federal Housing Policies

How has the federal government organized its available resources to assist the elderly with their housing? This section presents an inventory of current federal programs. An overview of these programs from the perspective of dwelling-use and dwelling-specific problems provides a useful initial orientation.

Figure 1 is a simplified depiction of the arrangement of current federal policies for meeting the housing-related needs of the elderly. The central point is the essentially independent administration of programs dealing with housing problems and those providing support services. The joint provision of services is largely ''unexplored territory," with the exceptions being the fledgling congregate housing program and some local efforts in which federal resources are effectively coordinated. Conspicuous gaps in coverage are evident—such as the absence of dwelling-specific aid for homeowners (except for home purchasing assistance provided by the Farmers Home Administration) and a lack of programs to help with dwelling modifications related to activity limitations. Likewise, the targeting of resources to lower income groups is mixed; it is probably good in the housing area and much weaker in the area of support services.7 In short, the present system is a patchwork and one that only infrequently provides the right aid to persons who need both housing assistance and supportive services.

Figure 1. Traditional federal policies for housing-related needs of the elderly.

Figure 1

Traditional federal policies for housing-related needs of the elderly.

Current federal housing programs that assist the elderly can be divided into broad categories: (1) those that facilitate the operation of the private market and (2) those that provide some housing assistance, sometimes accompanied by support services but generally without them. The private market-facilitating programs that are relevant to this discussion are a set of insurance programs operated by the Federal Housing Administration (FHA). By agreeing to insure certain mortgages, the FHA reduces the riskiness of the loan to the lender, a consequence that encourages both more lending and lending at interest rates lower than otherwise would have been charged. Insurance also makes the mortgages marketable to secondary facilities (e.g., Government National Mortgage Association, Federal National Mortgage Association), which in turn increases their attractiveness to loan originators. FHA traditionally has insured market-rate housing projects designed for elderly occupancy and also nursing homes and intermediate care facilities. Legislation enacted in 1983 also permits FHA to insure congregate housing facilities, board and care facilities, and "life care centers," which offer services beyond those provided in congregate facilities. (On the other hand, rather surprisingly, FHA has not yet been willing to perform its traditional innovative role by insuring home equity conversion mortgages.)

In the second program area, the federal government has been most active among the lower income elderly in addressing traditional dwelling-specific housing needs, mainly through a variety of subsidy programs that are largely directed toward renter households. Although most of the rental subsidy programs have long required that occupants be lower income persons, in 1981 the income targeting of such assistance was considerably tightened. As a result, assistance is now quite sharply limited to households with income levels that are no higher than 50 percent of an area's median family income, adjusted for family size. (For more on the changes enacted in 1981, see Struyk et al., 1983.) Help is provided through a wide range of programmatic vehicles including those under which the development of housing projects have been sponsored, such as public housing, the Section 202 program, and Section 8, New Construction. Aid has also been provided to households through rental assistance programs, most notably Section 8, Existing, and the new housing voucher program.

Federal housing programs per se have been directed toward renters; yet elderly homeowners have received substantial one-time aid in repairing and rehabilitating their homes through locally operated programs funded under the federal community development block program (CDBP). Rough calculations suggest that, in the early 1980s, this form of aid was equivalent in expenditures to about 15 percent of the housing assistance payments to the elderly participating in the rental program. CDBP has also been the funding source for many of the "house matching" programs and other innovative local efforts. In addition, elderly homeowners have been substantial beneficiaries of weatherization and fuel payment assistance programs (Struyk, 1984a, pp. 59–62).

The figures in Table 3 provide an accounting of the number of households currently eligible to receive housing assistance and the number that are actually receiving such payments.8 The figures exclude those households receiving the kind of one-time help described in the preceding paragraph. The federal emphasis on renters is amply evident in these figures. Currently, some 4.9 million households participate in these programs, all but .6 million of which are renters. Based on participation rates from the experimental housing allowance program, we estimate that an additional 2.6 million renters would participate if they had the opportunity to do so. The figures illustrate that, compared with the nonelderly, the elderly have been very well served—indeed, overserved—by the federal programs. Among income eligible renters, two-thirds of the elderly participate, but only about one-quarter of the nonelderly participate.

Table 3. Estimates of Households (in millions) Currently Eligible for and Receiving Federal Housing Assistance and Expected Participation in an Open-Enrollment Housing Voucher Program.

Table 3

Estimates of Households (in millions) Currently Eligible for and Receiving Federal Housing Assistance and Expected Participation in an Open-Enrollment Housing Voucher Program.

The final area of federal activity is that combining assistance for housing with assistance for supportive services. These arrangements come in three forms; what they have in common is that they all involve subsidized housing projects.9 The first type consists solely of projects that have been specially designed for use by the physically impaired. Even if no additional services are provided, such living environments may be instrumental in helping their occupants remain in the community, although there is no "hard" evidence on this point. To the knowledge of this author, there is no accurate count of the number of such specially designed units in the subsidized inventory; a reasonable guess might be 400,000. There is also no information on the number of these units that are, in fact, occupied by persons who take advantage of their special features; the absence of any requirements that such units be occupied by the impaired suggests that their use by those needing such environments could be improved.

The second type of program provides supportive services for those living in assisted housing, whether it has been specially designed or not. In this case the supportive services are not specifically part of the overall housing package; rather, local agencies—housing, social services, and health—identify the elderly who need these services and try to arrange to provide them. Yet the housing projects play an important role. Because they are occupied exclusively by the elderly, they are attractive to those who provide supportive and limited health services because they facilitate the identification of those requiring services and increase the efficiency with which the services are delivered. There is no systematic information on the extent to which arrangements of this kind have been made. However, the data that have been gathered on the services available at a large sample of elderly-only public housing and Section 202 projects10 as part of the evaluation of the demonstration congregate housing services program would suggest that such arrangements are fairly common. The problem with them, of course, is the uncertainty about continued funding for these services and the pattern of managers of housing projects accepting the services that are available as opposed to being able to provide those that are most needed.

Finally, there is the small congregate housing services program (CHSP), which thus far has existed rather precariously as a demonstration program, although continued congressional support seems likely. CHSP operates in 60 public housing and Section 202 projects, with about 1,800 persons receiving supportive services under the program (Nachison, 1985, p. 34). Participation is limited to those genuinely needing the services. The service bundle consists of a mandatory component of twice-a-day meals and options under which services are tailored to the needs of the individual. Possibly the most distinguishing characteristic of the program is that the funding for both housing assistance and supportive services comes from HUD, thus solving the often difficult problem of patching together funding for supportive services at the local level. (CHSP is discussed further in a later section.)

Overall, the earlier characterization of federal policy as being concentrated on dwelling-specific problems seems fair. Most service provision in federal projects comes through local initiatives.

A Policy Framework

In previous sections, we have seen that dwelling-specific and dwelling-use problems often occur independently. Nevertheless, in a substantial minority of cases, they occur together. It also seems probable that as dwelling-use problems become more acute, the incidence of dwelling-specific problems will increase rapidly. Activity limitations on the part of a family member (and the energy required of others in the household to provide informal care) means that dwelling upkeep is likely to be diminished. In addition, "drawing down" assets to pay medical bills or to support formal care expenses will lower incomes, possibly to the point at which housing expenditures become "excessive." We have also seen that current housing assistance deals only minimally with the nexus of housing and dwelling-use problems.

The challenge is to design a programmatic response that is flexible enough to deal with the variety of need mixes and ability-to-pay circumstances that will be encountered. If the response is properly designed—that is, if better alternatives to institutional long- and short-term care are developed—such programs may well permit savings in the total public resources going to the elderly. In a number of instances, for example, in-home services have been found to defer institutionalization and to reduce the number and length of visits to acute-care hospitals (see, e.g., Miller and Walter, 1983).

Three principles should guide the design of the general federal policy response. First, cost-effectiveness is essential. The criterion here is that the new approach should be no more costly and at least as effective as present programs. In this calculation, costs include assisting additional households beyond those to whom the assistance is actually directed; that is, serving those who otherwise would have received informal assistance is viewed as a cost. One issue in particular is the substitution of formal for informal supportive services. Also, achieving cost-effectiveness may well require a degree of coordination among services that is far beyond what is now occurring, especially between housing and support services that are currently administered independently in most cases. A further dimension worth exploring is that financial costs can be lowered by the elderly making in-kind assistance contributions to others in some types of residential arrangements.

Second, to the maximum extent feasible, the programs should be constructed to permit a range of choice to the elderly in terms of the solution adopted to meet each individual's needs: whether a recipient remains a homeowner or shifts to rental quarters; whether he or she uses community-based versus institutionally based services. Of course, the recipient may have to pay a larger share of the cost of more expensive solutions, but the choice should still be present.

Third, the options should be structured so as to foster timely adjustments in the "housing bundle" selected. As noted earlier, the housing needs of the elderly can be highly dynamic. Solutions that are offered in response to those needs ought to encourage timely changes in the elderly person's basic housing situation—for example, from living alone in a single-family home to living in an apartment in a congregate housing project. Because of the understandable resistance of the elderly to multiple moves, however, in-place changes through dwelling modifications and room use solutions must also be an integral part of the adjustment process.

The key idea behind the framework set forth here is that it is essential to tailor solutions to fit each of a range of housing-related needs. As needs change, alternative solutions should be available. To achieve this matching requires that one differentiate both among types of housing needs and among recipient populations. The various types of housing-associated needs were discussed at length earlier. Three household attributes seem to be most important: (1) economic resources (and hence the ability to pay for services); (2) mode of tenure (owner-occupancy versus renting), and (3) type of structure—single versus multiunit structures. The latter strongly affects the efficiency with which many support services can be provided, and tenure affects the range of options available for coping with dwelling deficiencies and excessive housing expenditures.

Figure 2 summarizes some of the public interventions that might be appropriate for households in differing circumstances.

Figure 2. Possible types of public intervention to improve the housing situation of the elderly.

Figure 2

Possible types of public intervention to improve the housing situation of the elderly.

The key here is explicit differentiation among types of needs and types of households requiring services. The role of government (which is listed in each box in the matrix) is also defined quite differently, depending on the household's ability to pay for services. (Note that "congregate services" is shorthand for a variety of supportive community housing environments.) For example, for households needing extensive support services, a voucher for a congregate housing program (described later) is appropriate; but only referral services to market-rate programs are necessary for middle-income households. Similarly, for those homeowners needing dwelling modifications only, grants make sense for those in the low-income groups, while referrals to contractors and possibly reverse annuity mortgages are the best form of government assistance to the more well-to-do.

In cases in which the household has both dwelling-specific and dwelling-use problems, greater coordination is essential. For renters, such needs will frequently be addressed most efficiently through a congregate services program, either subsidized or at market rates. For homeowners, the solution varies with their ability to pay and their desire to remain in their home. For lower income homeowners, in-home services provided through Medicaid (in states that have applied for this waiver) and housing assistance provided through a housing allowance may be economically feasible.

The leadership in the provision of a number of the services listed in the figure devolves on the local government, although federal support may be instrumental. The superiority of local organizations and solutions seems likely for most referral services, for grants for housing rehabilitation and modifications to owner-occupied dwellings, and for the provision of modest amounts of support services. At the same time, however, it is important to ensure that those nominally eligible for such assistance actually fall within an active service area. Spotty coverage—both between and within jurisdictions—has been an unfortunate hallmark of local initiatives for the elderly.

Selected Interventions

Since the late 1970s, discussion of housing policies for the elderly have been dominated by reports of innovative activities being undertaken first at the local level and then at the state level. Most of these initiatives grew out of the perceived absence of available alternatives to deal with the frail elderly in a community context. Among the alternatives that have been launched or at least seriously proposed as local programs have been accessory apartments, granny flats, house sharing implemented through "matching" programs, small group homes, congregate housing, revamped board-and-care facilities, and home equity conversion mortgages (also known as reverse annuity mortgages). It is probably fair to say that the perspective in starting such programs locally has been a version of "try it and see if it works." The federal government has been largely an observer of these initiatives and a passive observer at that. Consequently, we have a poorly documented record of the effectiveness of many of these local programs, while at the same time, there is a groundswell of sentiment for these programs to be replicated, often through federal legislation and appropriations. In general, however, the record is far too weak to justify proceeding on anything but a more rigorously organized demonstration basis.

Given the scope of this paper, it is simply not possible to give even cursory treatment to so many programs. Instead, three have been selected for comment based on their meeting at least one of two criteria: (1) the appearance of the potential to play a substantial role in the evolving long-term care system and (2) the ability to illustrate the range of possibilities in public housing interventions for the elderly.

Those alternatives that have been selected for discussion are congregate housing, the implementation of modifications to the dwelling to offset some of the effects of physical impairment, and small group homes. Each program is described, what is known about the program's effectiveness in reducing institutionalization is inventoried, and what appear to be the next steps in developing the approach are outlined. In particular, the following question is considered: Should the federal government now sponsor more demonstrations and evaluations, or should it begin to encourage such approaches by making them eligible for voucher recipients, by providing direct support or mortgage insurance, or through other means?

Congregate Housing

We may define a congregate housing facility as a rental housing complex in which households have separate units that include kitchens and bathrooms; the complex includes public spaces that can be used for social functions and for serving meals to residents on a regular basis (at least one meal per day). Housekeeping, chore performance, and personal services are available to the residents who need such assistance; the residents are frail, elderly persons who are capable of living quite independent lives but need limited assistance. The congregate facilities would be carefully targeted to this particular segment of the frail elderly who would otherwise probably be institutionalized. As defined here, congregate housing would be an integral part of the long-term care system.

Congregate housing is seen as a vehicle of great potential for the elderly and as a cost-effective alternative to institutionalization; that is, it is viewed as a way to permit persons who would otherwise be in institutions to remain in a more stimulating and less expensive community setting. If such an approach can be proven to be cost effective, federal assistance for residences (in lieu of Medicare payments for institutional care) would certainly make sense.11 Early evidence on the cost of congregate housing versus long-term care facilities certainly looks promising.12

Several states already have operating programs that incorporate at least some of the elements listed above, although they appear to differ significantly among themselves in regard to structure and the services they provide. These programs in general are expanding rapidly, and other states are seriously considering similar programs (see Nenno et al., 1985, p. 13; and Anthony, 1984).

The only information on the effectiveness of the congregate approach comes from CHSP, the recently concluded HUD demonstration project that served a population somewhat different from that just described. The key question behind the demonstration was the effects on institutionalization of providing supportive services to households living in HUD-assisted elderly projects.

In the CHSP setup, participants were limited to about 20 percent of the residents of a project, with assistance concentrated on those most vulnerable to institutionalization. (The limitation was imposed to maintain an atmosphere of independent living.) A professional assessment committee was formed at each project to screen applicants regarding their need for the services that would be provided.

Twice-a-day meal service was seen as the core of the supportive services. Additional nonmedical services such as housekeeping/chore, personal assistance, transportation, escort, and social services could also be provided to fill gaps in a project's service delivery system.

Although 48 projects participated in CHSP, the part of the evaluation dealing with its impact on institutionalization drew on 17 projects. Each of these projects was ''matched" with a control project of the same type (public housing or Section 202) in the same area. Baseline data (i.e., prior to CHSP implementation) on samples of residents were obtained at the experimental and control projects; data were also obtained at two later times—on average, at 14 months and 26 months after the baseline samples.

The results regarding institutionalization can be briefly summarized: at the end of the period, there was no significant difference in the rate at which participants and controls were permanently placed in institutions. On the other hand, there was a difference in the rate at which members of the two groups were temporarily institutionalized over the period: 15 percent of the experimentals versus 23 percent of the controls (Sherwood et al., 1985, Table IV.2). On the basis of these findings, it would certainly be difficult to argue for the enactment of a major congregate housing program.

Still, there were several problems with the program that was demonstrated and with the evaluation that need to be considered in reaching a balanced judgment about congregate housing. Four points in particular are important:

1.

The population served by the program was not restricted to those that were vulnerable to institutionalization, according to the definition used in the evaluation; also, the share of participants who were vulnerable varied considerably from project to project.

2.

CHSP was structured around the provision of twice-a-day meals, a service that many of the participants neither needed nor especially wanted. Hence, a great many of the services provided would not have been expected to affect institutionalization rates.

3.

Many of the control projects were fairly rich service environments themselves. Thus, the comparison was not between a case of no services available to residents and CHSP but rather between some services (which varied sharply among projects) and more and possibly better tailored services.

4.

The observation period may have been too short, given the kind of program and control projects employed in the demonstration. Even with all of the problems just listed, some significant effects may have been evident over another year or two. (Alternatively, if a short observation period was to be used, there should have been better measures of health status in addition to the relatively crude institutionalization rate measure.)

My own conclusion after reviewing the CHSP experience is that we do not know nearly enough to proceed with a larger congregate services program at this time. On the other hand, we have learned a great deal from the demonstration, and HUD should launch another demonstration as soon as possible, a demonstration designed to overcome the problems of the initial one. In particular, the service bundle should be adjusted away from meals, and the control population should be drawn from clearly service-poor environments, even if this means having to draw on those of the elderly who are not living in subsidized or even elders-only housing projects. An additional such demonstration will permit the design and testing of a program that may well be ready for use as a key element in the long-term care system.

Dwelling Modifications

The alteration of the features of a dwelling to make it easier for a person with a physical impairment to use is a logical response to the onset of activity limitations. Some have thought that the installation of appropriate modifications to dwellings occupied by elderly homeowning households that have an impaired member, through public programs if necessary, could be cost effective in terms of helping sustain the person in the community. The modifications could be directly effective by permitting the person to do more for himself around the unit; they could be indirectly effective by relieving those providing help to the impaired person of some of the burden of assistance. The case for the cost-effectiveness of these modifications is bolstered by their modest one-time charge, compared with the costs of continuing service provision. Yet, while such cost-effectiveness may indeed be the case, as we shall see, there is really no evidence currently available to support this contention.

It may be worthwhile to note that modifications are only one of a series of adjustments that households can and do make to their housing without relocating. Three other adjustments include changing the use of rooms (often in response to the activity limitations of a household member); taking in a roomer or boarder to generate additional income or for the companionship and help they might provide; and changing (typically, lowering) the amount of repairs and improvements to the dwelling because of changes in the ability to make such repairs.

The figures in Table 4 show the rate at which a sample of households made these adjustments over a 2-year period. The sample consists of households in selected neighborhoods from seven large central cities that did not relocate over a 5-year observation period. Only a very small percentage of households took in a roomer or boarder or changed the use of rooms within their homes over this period. A rather surprisingly high 10 percent of elder-headed and 5 percent of nonelder-headed households undertook modifications to their homes to facilitate their use by a physically impaired family member. Taken together, these figures indicate that about 8 percent of elder-headed and 6 percent of nonelder-headed households make one of these three types of housing adjustment each year. These rates can be put into perspective by noting that about 3.3 percent of elderly homeowners and 13 percent of nonelderly homeowners adjust their housing circumstances by changing residence each year. As a result, the rates of in-place adjustments for the elderly are at least double those achieved by relocating.13

Table 4. Incidence of Housing Adjustments (percentage) by Age of Household Head for Selected Households in Seven Large Central Cities.

Table 4

Incidence of Housing Adjustments (percentage) by Age of Household Head for Selected Households in Seven Large Central Cities.

A better idea of the national incidence of dwelling modifications is available from data gathered by a special supplement to the 1978 annual housing survey (AHS). These figures, which are shown in Table 5, indicate that only about 10 percent of elderly households with at least one member with self-reported health or mobility problems had made a modification to their unit at the time of the survey. These rates appear to be considerably lower than those cited earlier for the seven-city sample; because the AHS data are from a national representative sample of dwellings, the lower rates presumably are more generally valid.

Table 5. Dwellings Occupied by Elder-headed Households with at east One Member with Health or Mobility Problems.

Table 5

Dwellings Occupied by Elder-headed Households with at east One Member with Health or Mobility Problems.

The available research strongly suggests that the receipt of assistance from inside or outside the home, including such services as meals programs, significantly reduces the likelihood that a household will undertake a dwelling modification. Such changes do not seem to be very sensitive to the household's economic position. The likelihood does increase, however, when the person with activity limitations or the person who must use assistance (e.g., a cane or wheelchair) in getting around is the spouse in a husband-wife household (Struyk and Katsura, 1985).

Although this information is helpful in understanding why households make unit modifications, it tells us nothing about the effectiveness of such changes in delaying institutionalization. Some research on this topic using the 1982 long-term care and annual housing survey data sets is just now getting under way. Until more information on effectiveness is forthcoming, it is very difficult to argue for any type of broad-based public intervention to provide dwelling modifications.

All of the information generated to date, however, points to the fact that a decision to make such changes to a dwelling is complex, and it is critically dependent on the types of supportive services available to the impaired person. Under these conditions, the provision of modifications may make little difference over a wide range of cases and, hence, may not be cost effective. It is to be hoped that the work now beginning will help to isolate those instances in which such modifications will be the most effective strategy for delaying institutionalization.

Group Homes

The terms "house sharing" or "group home" usually refer to a single structure in which a number of unrelated people live. Many group homes are single-family homes that have been converted to provide private rooms as well as common space for residents. In these small group home arrangements, outside health or social services can be provided to individuals who need them, although they are seldom provided to all persons in the structure.

Arrangements of this type have been organized by individuals (sometimes an elderly person who wants to share his or her home in a particularly expansive way) and by local agencies. Our attention is on those that have been developed by local agencies. The central premise of such arrangements is that the occupants help each other and together take care of common tasks such as meal preparation. The occupants thereby substitute their own labor for that of others, which could lower considerably the cost of living in a relatively "service-rich" environment. (See Morgan, in this volume, for a more detailed discussion of the possibilities for mutual support without cash transactions.)

There is little systematic information about small group homes, largely because the development of these homes has been so dominantly one of local initiative. One survey in 1982 of 21 shared residences found the "typical" resident to be a woman in her early seventies who had resided in the group household for at least a year. The majority of residents had at least weekly contact with their families. As might be imagined, monthly charges varied widely as did the service package provided by the residence. All resident households used community services such as a senior center or visiting nurse. According to a study by the Shared Housing Resource Center reported in Gold (1985, pp. 49–50), two-thirds of the organizations expected the residents to participate at least minimally in housekeeping chores. Another survey of such living arrangements (van Dyke and Breslow of the Jewish Council for Aging of Greater Washington, reported in National Policy Center on Housing and Living Arrangements for Older Americans, 1983, pp. 26–27) found that residents tended to be somewhat old and in declining health; the shared living arrangement was clearly viewed by occupants as the only real alternative to a nursing home. Perhaps the most significant aspect of these arrangements is their diversity, as recently documented in 15 case studies by Streib et al. (1984, chap. 6).

With so little material of even a descriptive nature about shared living, the absence of any real evaluation on the impact of such arrangements on institutionalization is not surprising. In this case, three distinct types of issues should be addressed in any demonstrations and accompanying evaluations that might be mounted.14 First, the effectiveness of such arrangements in forestalling institutionalization should be documented. Second, the potential for shared housing to allow the elderly to assist each other actively, thereby extending their own period of productive activity and lowering the cost of this alternative compared with others, should also be studied. Third, it will be important to understand the types of people who do well in this environment. The limited information we do have is clear about the fact that such arrangements do not work for everyone; it would be helpful to have more accurate profiles of those who seem especially suited to shared housing.

The Impact of Such Arrangements

What would be the effect of the widespread adoption of the kind of living arrangements just described, assuming that they were found to be effective in delaying the institutionalization of the frail elderly? To begin with, it seems quite evident that the life satisfaction of those who would have been placed in institutions otherwise will be significantly improved. Although the value of such improvements is difficult to quantify, it is nevertheless clearly important.

The possible savings to society in the resources that must be used could be substantial. Let us assume that 20 percent fewer persons would be institutionalized in the future because of the existence of congregate and other housing arrangements. We will also use Heumann's (1985) estimate that equivalent services for those needing less than nursing home care can be provided in a congregate setting for about one-third less cost than in the long-term care facilities. Together, these figures imply a reduction in long-term care costs of about 7 percent. The overall resource savings would presumably be greater because the incidence of couples continuing to live together would rise, thereby cutting down the need for two separate living arrangements (one of which would have been in an institution).15 Such effects might raise the savings from these alternatives to the equivalent of 8 or 9 percent of long-term care costs.

This is clearly the upper limit of such savings, however, because we have tacitly assumed that only those persons that would have been institutionalized would be served in congregate facilities. From Heumann's study (1985, Table 10), we know that the services portion of the congregate package constitutes about half of the total cost. If we assume that a household would have been paying for the shelter component in any event (or would have received a subsidy to help pay for it), then for a system of congregate facilities and long-term care facilities to operate with the same resources as the present arrangement would require that no more than half of the households in congregate facilities be without a member who would otherwise have been in an institution. Of course, if the households would have purchased some of the supportive services in other environments, then a higher share of congregate housing occupants could be those who would have been in community housing. If congregate "slots" were targeted in such a way that one-third of the occupant households were living in such facilities "inappropriately," then the savings figures cited above would be cut in half.16

The foregoing suggests two things. First, the savings potential is large—in the billions of dollars—from the use of congregate and other arrangements that provide appropriate levels of supportive services as a substitute for those provided in long-term care facilities. Second, the realization of such savings depends critically on the effectiveness of such arrangements in preventing institutionalization and on the degree to which services can be targeted toward those genuinely at risk of being institutionalized.

Conclusions

This paper has argued that the central issue in federal housing policies for the elderly is the development of public assistance that will help them make necessary housing transitions in such a way as to allow community-based housing to be an active and integral element in the overall long-term care system. Current housing policy for the elderly only gives passing attention to this issue. Consequently, the programs now in operation are poorly designed to address broad housing needs.

There is no shortage of suggested alternatives for using housing-oriented programs to assist the elderly to remain in the community rather than being institutionalized. Indeed, local governments and some states are moving forward on a number of options. There has, however, been little evaluation of the effectiveness of these interventions in reducing institutionalization and little assessment of the extent to which they substitute formal for informal care.

It seems clear that the first step must be for federal policy to be reoriented toward the nexus of dwelling-specific and dwelling-use problems, as they have been defined in this essay. We are not, however, in a position to argue for any particular programmatic package to implement this policy. Rather, the next phase must be one of intense experimentation and evaluation. The role of the Department of Housing and Urban Development in carrying out the essential evaluations will be paramount. If we proceed with alacrity, it may be possible to be in a position by 1990 to make forceful recommendations for the adoption of operational programs in this area.

Appendix

Definitions

Dwelling Deficiency

See Table A-1. Specifics of the definition were dictated by the data available in the annual housing survey. This definition is the same as that employed by HUD.

Table A-1. Deficiencies that Cause a Housing Unit to be Judged Physically Inadequate (using the HUD/Simonson Definition), Based on Annual Housing Survey Items and a Revised Definition (1981).

Table A-1

Deficiencies that Cause a Housing Unit to be Judged Physically Inadequate (using the HUD/Simonson Definition), Based on Annual Housing Survey Items and a Revised Definition (1981).

Excessive Housing Expenditures

Here we follow HUD's lead so that our results will be consistent with other tabulations. Excessive burden is defined separately for renters and homeowners. For renters, a gross rent (the contract rent plus utilities paid by the tenant) above 30 percent of gross household income is considered excessive. For owner-occupants, out-of-pocket expenditures for housing (excluding expenditures for maintenance and improvements) above 40 percent of family income is considered excessive. The higher standard for homeowners is based on the tax advantages accruing to home-owners and on the capital gains-producing investment embodied in their housing expenditures. (See Feins and White, 1979, for more discussion on this point.)

Need for Supportive Services

Two definitions are used, based on data in the 1979 National Health Interview Survey. The "generous" definition, developed by Soldo (1983) includes any person with at least one of the following characteristics:

  • needed or received help with at least one of the seven activities of daily living (ADL);
  • needed or received help with at least one of the four instrumental activities of daily living (IADL);
  • was not able to perform one or more of the ADL functions;
  • stayed in bed all or most of the time; or
  • needed help with urinary or bowel devices.

The ''stringent" definition includes only those persons in the above group who receive formal home care services.

References

  • Anthony, A. S. 1984. "Prepared Statement." In Sheltering America's Aged: Options for Housing and Services, U.S. Special Committee on Aging. Senate Hearing 98-875. Washington, D.C.: Government Printing Office.
  • Bielby, W. T. 1979. Evaluating Measures of Neighborhood Quality in the Annual Housing Survey. Washington, D.C.: Government Printing Office.
  • Burke, P. 1984. Trends in Subsidized Housing, 1974–81. Washington, D.C.: U.S. Department of Housing and Urban Development, Division of Housing and Demographic Analysis.
  • Drews, R. 1982. Rural Housing Programs: Long Term Costs and Their Treatment in the Federal Budget. Washington, D.C.: Congressional Budget Office.
  • Gold, M. 1985. The Older American's Guide to Housing and Living Arrangements . Washington, D.C.: Institute for Consumer Policy Research.
  • Heumann, L. 1985. A Cost Comparison of Congregate Housing and Long-Term Care Facilities in the Mid-West, Urbana: University of Illinois, Housing Research and Development Program.
  • Irby, 1. 1984. Housing Problems in 1981: A Synopsis. Washington, D.C.: U.S. Department of Housing and Urban Development, Housing and Demographic Analysis Division.
  • Kingsley, G. T., and P. M. Schlegel. 1982. Housing Allowances and Administrative Efficiency. N-1741-HUD. Santa Monica, Calif.: The Rand Corporation.
  • Levine, M. 1985. Statement Before the Subcommittee on Housing and Consumer Interests, Select Committee on Aging, U.S. House of Representatives, March 7. Washington, D.C.: Congressional Budget Office.
  • Marans, R. W. 1979. The Determinants of Neighborhood Quality: An Analysis of the 1976 Annual Housing Survey. Washington, D.C.: Government Printing Office.
  • Miller, L. S., and L. Walter. 1983. "The Comparative Evaluation of the Multipurpose Senior Services Project—1981." Berkeley: The University of California.
  • Nachison, J. S. 1985. "Who Pays? The Congregate Housing Question." Generations Spring:34–35.
  • National Policy Center on Housing and Living Arrangements for Older Americans. 1983. Alternative Housing and Living Arrangements for Independent Living. Ann Arbor: University of Michigan.
  • Nenno, M. K., J. S. Nachison, and E. Anderson. 1985. "Support Services for Frail Elderly or Handicapped Persons Living in Government-Assisted Housing." Washington, D.C.: National Association of Housing and Urban Development Officials.
  • Newman, S. 1985. "Housing and Long-term Care: The Suitability of the Elderly's Housing to the Provision of In-Home Services." . [PubMed: 3920125]
  • Palmer, J., and B. B. Torrey. 1984. "Health Care Financing and Pension Programs." In Federal Budget Policy in the 1980s, G. Mills, editor; and J. Palmer, editor. , eds. Washington, D.C.: The Urban Institute Press.
  • Sherwood, S., J. N. Morris, C. C. Sherwood, S. Morris, E. Bernstein, and E. Gorstein. 1985. Final Report of the Evaluation of Congregate Housing Services Program. Boston: Hebrew Rehabilitation Center for the Aging.
  • Soldo, B. 1983. "A National Perspective on the Home Care Population." Working Paper in Demography CPR 83-4. Washington, D.C.: Georgetown University, Center for Population Research.
  • Streib, G., W. E. Folts, and M. A. Hilker. 1984. Old Homes—New Families: Shared Living for the Elderly. New York: Columbia University Press.
  • Struyk, R. 1982. "The Demand for Specially Adapted Housing by Elderly-Headed Households." Project Report 3014-1. Washington, D.C.: The Urban Institute.
  • Struyk, R. 1984. a. "Home Energy Costs and the Housing of the Poor and the Elderly." In Energy Costs, Urban Development and Housing, A. Downs, editor; and K. Bradbury, editor. , eds. Washington, D.C.: Brookings Institution.
  • Struyk, R. 1984. b. "Housing-Related Needs of the Elderly Americans and Possible Federal Responses." In Sheltering America's Aged: Options for Housing and Services, U.S. Senate Special Committee on Aging. Senate Hearing 98-875. Washington, D.C.: Government Printing Office.
  • Struyk, R. 1985. "Housing-Related Needs of the Elderly and Possible Federal Responses." Journal of Housing for the Elderly 2(4):3–26.
  • Struyk, R., and H. Katsura. 1985. Aging at Home: How the Elderly Adjust Their Housing Without Moving. Urban Institute Report 3166-03. Washington, D.C.: The Urban Institute.
  • Struyk, R., and B. Soldo. 1980. Improving the Elderly's Housing. Cambridge, Mass.: Ballinger.
  • Struyk, R., and M. Turner. 1984. "Changes in the Housing Situation of the Elderly." Journal of Housing for the Elderly 2(1):3–20.
  • Struyk R., and J. Zais. 1982. "Providing Special Dwelling Features for the Elderly with Health and Mobility Problems." Washington, D.C.: The Urban Institute.
  • U.S. Department of Housing and Urban Development. 1985. FY 1986 Budget . Washington, D.C.: U.S. Department of Housing and Urban Development.
  • Weissert, W., and W. Scanlon. 1983. "Determinants of Institutionalization of the Aged." Working Paper no. 1466-21 (rev.). Washington, D.C.: The Urban Institute.

Footnotes

1. The annual housing survey is the primary source of national housing data. For analyses of the neighborhood data included in the survey, see Bielby (1979) and Marans (1979).

2. Since 1979 we know, in general, from annual housing survey data that the share of households with "excessive" housing expenditures has increased, whereas the share living in dwellings with deficiencies has decreased slightly.

3. Less detailed estimates with 1981 data by Irby (1984) and using somewhat different definitions for deficiencies show about 1.2 million elderly households in deficient units and 2.55 million households with excessive expenditures.

4. This procedure seems to be reasonable, given that Newman (1985), using a similar definition of impairment, found about 13 percent of elderly households had at least one member with such a condition.

5. Together with the information on the determinants for the receipt of support services, the higher rate also implies that the services received do not have much effect on dwelling conditions.

6. In doing this calculation, the rate was applied only to those elder-headed households that were not participating in a housing program, the general assumption being that, if they were part of a housing program, they would not have excessive expenditures for housing.

7. Services provided under the Older Americans Act are not means tested. Those funded by the social services block grant program have varying income limits that are set by the states; almost universally, however, these limits are less stringent than those in the housing programs.

8. These figures correspond generally to those prepared by the Congressional Budget Office; see Levine (1985, p. 11).

9. Of course, one could approach the issue from the other direction, that is, starting with households that are receiving supportive services and then examining their housing circumstances. The housing-oriented approach has been chosen here because it is more in keeping with the overall perspective of the paper.

10. The Section 202 program is one in which specially designed housing is developed by nonprofit sponsors for occupancy by elderly households and households with a handicapped member. Federal subsidies are provided in the form of direct loans that carry interest rates below the market level and, in recent years, through rental assistance payments, available under the Section 8 program, for all occupants who are eligible to receive them.

11. For a discussion of a congregate housing voucher program for low-income households, see Struyk (1984b).

12. Heumann (1985) in the first careful analysis calculates that, for equivalent services, congregate facilities are about one-third less expensive overall than long-term care facilities.

13. Repairs and improvements are associated with the types of adjustments we have just reviewed and with longer term strategies of housing upkeep and investment. A central hypothesis considered here is that there is a cohort of elderly homeowners who, because of economic or health circumstances, decide implicitly to draw on the equity in their homes through a program of lower maintenance. Similarly, we are interested in which households persistently are investing in their homes. The last two rows of figures in Table 4 show that 20 percent of the elderly in the sample undertook little or no repair activity over 2 years and that they were somewhat more likely to do this than their nonelderly counterparts; the differences between the two groups are not large, however. Similarly, the elderly as a group are undertaking repairs, and improvements—even large improvements—to their homes at quite high rates. Still, they undertake fewer such repairs and investments, and each year they spend less, than their more youthful counterparts. The overall pattern is one in which properties are indeed largely being maintained, and disinvesting households are a definite minority and comparable in size with the proportion of such households in the nonelderly population.

14. The implied call for a demonstration may seem odd, given the existence of ongoing projects. It might be that one could evaluate the present programs, but as a group, they may be so heterogeneous as to preclude anything but a series of case studies. Although case studies may be an essential first step in designing a demonstration with an evaluation, they probably could not serve as the basis for an evaluation by themselves.

15. For a discussion and projections of future long-term care costs to the federal government, see Palmer and Torrey (1984).

16. It is assumed that supportive services constitute half of all congregate costs and that congregate costs are about 67 percent of long-term care costs. It follows that services in congregate facilities are equivalent to 33 percent of long-term care costs.

Raymond J. Struyk is with the Urban Institute.

Copyright © National Academy of Sciences.
Bookshelf ID: NBK219340

Views

  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (9.5M)

Related information

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...