A National Study of Assisted Living for the Frail Elderly: Results of a National Survey of Facilities


Catherine Hawes, Ph.D., Miriam Rose, M.Ed., and Charles D. Phillips, Ph.D., M.P.H.

Myers Research Institute

December 14, 1999

This report was prepared under #HHS-100-94-0024 and #HHS-100-98-0013 between HHS's ASPE/DALTCP and the Research Triangle Institute. Additional funding was provided by American Association of Retired Persons, the Administration on Aging, the National Institute on Aging, and the Alzheimer's Association. For additional information about this subject, you can visit the DALTCP home page at http://aspe.hhs.gov/_/office_specific/daltcp.cfm or contact the ASPE Project Officer, Gavin Kennedy, at HHS/ASPE/DALTCP, Room 424E, H.H. Humphrey Building, 200 Independence Avenue, S.W., Washington, D.C. 20201. His e-mail address is: Gavin.Kennedy@hhs.gov.

The opinions and views expressed in this report are those of the authors. They do not necessarily reflect the views of the Department of Health and Human Services, the contractor or any other funding organization.


This is the first in a series of planned reports based on data collected from surveys of a national probability sample of assisted living facilities. These data were collected as part of a study, "A National Study of Assisted Living for the Frail Elderly," funded by the U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation (ASPE), with additional support provided by the AARP, the Administration on Aging (AoA), the National Institute on Aging (NIA), and the Alzheimer's Association.

This report presents the results of a telephone survey of a nationally representative sample of 2,945 places identified as assisted living facilities.

This report presents data from a telephone survey of the administrators of assisted living facilities across the country. These facilities were selected from a national probability sample of all facilities that met the criteria for inclusion in the study. Thus, the findings are representative of the industry as a whole. As such, they represent the first empirical data on the characteristics of the assisted living industry nationwide based on a representative national sample of facilities.



The overall purpose of the study was to learn about the role assisted living facilities play in providing a residential setting and supportive long-term care services to the elderly. The specific objectives of this telephone survey were to:

  • Determine the size and nature of the supply of assisted living facilities

  • Describe the basic characteristics of the assisted living industry particularly in terms of the services, accommodations and basic price

  • Begin examining the extent and way in which the current supply of facilities embodies the key philosophical tenets of assisted living, and

  • Identify facilities for subsequent, more extensive data collection.

Assisted living represents a promising new model of residential long-term care, one that blurs the sharp and invidious distinction between receiving long-term care in one's own home and in an "institution."



In order to conduct the survey and obtain generalizable results, the project staff implemented a complex, multi-stage sampling design. At the first stage, project staff selected a random sample of 60 geographic areas, known as first stage sampling units (FSUs). These 60 FSUs were comprised of 1,086 counties in 34 states. In these geographic areas, project staff created a comprehensive listing of places thought to be assisted living facilities. Staff used a combination of sources to create this list, including state licensure agencies, industry trade associations, local and national retirement directories, telephone book "yellow" page advertisements, and Internet listings.

Assisted living is still "new enough that the businesses offering it and the states that license it do not agree on a precise definition."

Tamara Hodlewsky
National Center for Assisted Living, 1998

From this list of potential candidate facilities, project staff selected a stratified, random sample of 2,945 places. These places were then surveyed by telephone to determine their eligibility for the study and, if eligible, to secure information about the facility's size, services, price and accommodations.

Eligibility Criteria. The study's three basic eligibility criteria were that a facility had to:

1. Have more than ten beds1

2. Serve a primarily elderly population

In addition, the facility either had to:

3a. Represent itself as an assisted living facility


3b. Offer at least a basic level of services, which were:

  • 24-hour staff oversight
  • Housekeeping
  • At least 2 meals a day, and
  • Personal assistance, defined as help with at least two of the following: medications, bathing, or dressing.

The administrators of a sample of 2,945 candidate facilities were then surveyed by telephone during 1998. If the candidate facility met the study eligibility criteria specified in a set of screening questions, then the administrator was asked to respond to questions about the facility, its size, occupancy, accommodations, services, price and basic admission and discharge criteria. A total of 1,251 facilities were contacted, found to be eligible, and interviewed.

The results of this screening activity and of the more extended telephone survey are the topic of this report. The results reported here are statistical estimates about the universe of assisted living facilities, based on responses from the nationally representative probability sample of facilities that were surveyed.




There were an estimated 11,459 assisted living facilities (ALFs) nationwide, with approximately 611,300 beds and 521,500 residents, as of the beginning of 1998.


The average bed-size was 53 beds; 67 percent of the ALFs had 11-50 beds; 21 percent had 51-100 beds; and 12 percent had more than 100 beds. Facility occupancy averaged 84 percent. The average length of time the ALFs had been in business was 15 years, but slightly more than half (58%) of the ALFs had been in business for 10 years or less. About one-third (32%) had been in business no more than five years.

EXHIBIT ES1. Distribution of Resident Units Between Rooms and Apartments2
Pie Chart: Apartments (43%); Rooms (57%).

Unit Type. A room was the dominant type of resident unit (57%) in ALFs; 43 percent of the units were apartments.2 The most common type of room was a private room with a full bathroom (42% of all single rooms). The most common type of apartment was a one-bedroom, single occupancy apartment (41%).

Privacy. Most assisted living facilities offered consumers a range of options in terms of private or shared accommodations. Only 27 percent of the facilities had all-private accommodations. A plurality of ALFs (45%) had a mix of private and shared units. However, slightly more than one-fourth of the ALF administrators (28%) reported that the facility had at least one bedroom shared by three or more residents.

Although ALFs offered residents a range of options, 73 percent of all resident units were private. Twenty-five percent of the units were semi-private, that is, shared by two unrelated persons. Two percent of resident units were in "ward-type" rooms that housed three or more unrelated persons.

EXHIBIT ES2. Distribution of Units by Privacy
Pie Chart: Private Units (73%); Semi-Private Units (25%; Ward-Type Rooms  		(2%).

Bathrooms. While nearly three-quarters (73%) of the rooms or apartments were private, slightly less than two-thirds (62%) of the units offered a private full bathroom (i.e., toilet, sink and shower or tub). An additional six percent of the units had a private "half" bath (i.e. toilet and sink) but no bathing facilities except communal facilities shared with other residents. Thus, more than one-third (38%) of all ALF units required the resident to share a bathroom.


General Services. Nearly all facilities provided or arranged 24-hour staff, three meals a day, and housekeeping. More than 90 percent of the ALF administrators also reported that the facility provided medication reminders and assistance with bathing and dressing; 88 percent of the ALFs provided or arranged central storage of drugs or assistance with administration of medications.

Nurse Staffing. Almost three-quarters of the ALFS (71%) had a licensed nurse, either a registered nurse (RN) or licensed vocational nurse (LVN), working on staff full- or part-time. Slightly more than half of the ALFs (55%) reported having an RN on staff either full or part- time. Forty percent of the ALFs reported having a full-time RN on staff.

Providing or Arranging Services. Administrators were also asked whether the facility provided services with their own staff or arranged with an outside agency for the provision of the service. With the exception of therapies, if an ALF offered a service, such as help with bathing, dressing, and managing medications, most provided it with their own staff. About half (52%) of the facilities provided some care or monitoring by a licensed nurse (RN or LPN) with their own staff, and one-quarter (25%) arranged for nursing care with an agency. However, one in five ALF administrators (21%) reported that the facility did not arrange or provide any care or monitoring by a licensed nurse.

EXHIBIT ES3. Availability of Services by RN and LPN
Pie Chart: Not Offered (21%); Provided (52%); Arranged (25%); Arranged & Provided (2%).

Most ALFs reported a willingness to admit residents with moderate physical limitations, such as using a wheelchair (71%) or needing help with locomotion (62%) (i.e., walking or using a wheelchair or cart). But fewer than half the ALFs (44%) were willing to admit residents who needed assistance with transfers (i.e., in or out of bed, a chair or wheelchair). Administrators also reported that fewer than half the ALFs would admit a resident with moderate to severe cognitive impairment (47%).3

Facilities also had criteria about the retention of residents with certain types of conditions or problems, although, as with admission policies, many facilities had idiosyncratic policies (see footnote #3). Nearly one-third of the administrators (31%) reported that the ALF would not retain a resident who used a wheelchair (or that "it depends"), and 38 percent would discharge a resident who needed assistance with locomotion. Fewer than half (45%) of the ALFs would definitely retain a resident with moderate to severe cognitive impairment, and 76 percent would not retain residents with behavioral symptoms (e.g., wandering). Seventy-two percent of the ALFs would not retain a resident who needed nursing care for more than 14 days. Nearly three-quarters of the ALFs (72%) reported that one of more of their residents had been discharged within the last six months because the resident needed skilled nursing care.


ALF administrators estimated that about 24 percent of their residents received help with three or more activities of daily living (ADLs), such as bathing, dressing, and locomotion. They estimated that about one-third of the residents (34%) had moderate to severe cognitive impairment.


The information provided by administrators identified two significant variations among the ALFs that are worth noting. One group of ALFs identified or described themselves as assisted living facilities. Another much smaller group provided the same basic services but identified themselves by some other designation, such as adult congregate living, residential care, or community residential facility. Another significant variation was between ALFs that were free-standing and ALFs located on a campus that offered multiple levels of care. Such "multi-level" campuses typically housed an ALF and a nursing home or some other type of residential care, such as congregate apartments or independent living facilities.

Self-Described ALFs. Seven of ten (72%) of the administrators represented or described the facility as being an "assisted living facility or residence." Twenty-eight percent of the administrators did not describe the facility as assisted living; however, the facility still met study eligibility criteria. Despite the differences in how the administrators characterized the facilities, the two groups of facilities were remarkably similar in terms of their size, the services they offered, their nurse staffing, most admission and retention criteria, and the basic characteristics of their residents. Self-described ALFs, however, tended to have lower occupancy rates, had been in business for a shorter period of time, were more likely to offer apartments and private units, and were more likely to admit and retain residents who used a wheelchair or received help with locomotion. They also tended, on average, to have higher monthly prices.

Free-Standing ALFs Compared to ALFs Located on a Multi-Level Campus. The majority of ALFs (55%) were free-standing, while 45 percent were located on a campus housing multiple facilities or residential settings offering different levels of care. ALFs on a multi-level campus had higher occupancy rates and tended to have higher monthly prices than free-standing ALFs. They were also more likely to have private units and apartments and to provide or arrange more services for residents, most notably nursing care and therapies. ALFs on multi-level campuses also tended to have higher levels of nurse staffing than free-standing ALFs. In addition, they were more likely to admit and retain residents who needed nursing care and residents who used a wheelchair. Despite this, the administrators did not report having a "heavier care" resident case mix than the free-standing ALFs.

Categorization of ALFs By Combined Levels of Service and Privacy. Any attempt to understand assisted living and its role in providing long-term care to the frail elderly is hindered by the lack of a common definition of "assisted living." Currently, places known as ALFs differ widely in ownership, auspice, size, services, staffing, accommodations, and price. Thus, analyzing data on facilities and reaching conclusions about "assisted living" as a whole involves comparing "apples to oranges." As a result, project staff developed a classification that divided the universe of assisted living facilities into distinct categories or types of facilities, representing their mix of services and privacy. The four types the study identified represent reasonably homogeneous groups of facilities. Moreover, the data revealed significant differences among groups.

Definitions of high, low and minimal privacy. "High privacy," meant that at least 80 percent of the resident units were private. A total of 31 percent of the facilities met this definition of high privacy. Twenty-eight percent of the ALFs offered "minimal privacy" because they had one or more rooms that housed at least three residents. The remainder of the ALFs (41%) fell between these two types of facilities in a "low privacy" category.

Definitions of high, low and minimal services. "High services" was defined as having a full-time RN on staff and providing nursing care, as needed, with facility staff, as well as providing help with at least two ADLs, 24-hour staff, housekeeping, and at least 2 meals a day. Thirty-one percent of the ALFs met this criterion. Five percent of the ALFs did not offer help with even two ADLs and were thus defined as providing "minimal" services. The remaining ALFs (65%)4 were categorized as "low service," although some that did not provide nursing care with their own staff were willing to arrange a higher level of services through an outside provider, such as a home health agency.5

Combining the mix of services and privacy revealed four basic types of ALFs. The first type of ALF combined facilities in the "minimal" group of ALFs (i.e., the 32% with either minimal privacy or minimal services) and facilities offering low privacy and low service (i.e., 27% of the ALFs). The combined low/minimal privacy and services group was the most common type of assisted living facility, comprising 59 percent of all the ALFs. This type of assisted living facility cannot be easily distinguished from the traditional concept of board and care homes. A significant proportion of resident rooms were shared rather than private, and such facilities offered little beyond assistance with medications, bathing, or dressing. In two of five (41%) ALFs described by this model, there was at least one room shared by three or more people. Facilities of this type not only represented the majority of all ALFs nationwide, they also constituted 58 percent of all the facilities that described themselves as assisted living.

  EXHIBIT ES4. Distribution of ALFs by Categories  
Category   National Estimate  
1a. Low Privacy & Low Service 27%
1b. Minimal Privacy or Service 32%
2. High Privacy & Low Service 18%
3. High Service & Low Privacy 12%
4. High Privacy & High Service   11%

Another ALF type offered a high degree of privacy in accommodations but low services, a sort of "cruise ship" model of assisted living. In this type of ALF, more than 80 percent of the accommodations were private. However, these facilities would have had a difficult time helping residents age in place, since they had no RN on staff and most were unwilling or unable to provide or arrange any nursing care for residents. Only 19 percent of the ALFs in this model would provide or arrange nursing care and retain a resident who needed such care. This ALF type comprised 18 percent of all ALFs nationwide.

A third type of ALF was one described as high service/low privacy. In such facilities, two-thirds of the accommodations were in single rooms rather than apartments, and fewer than 80 percent of the rooms were private. However, all such facilities had a full-time RN on staff. About half (53%) the ALFs of this type were willing to provide or arrange nursing care, as needed, and retain residents who needed such care. This was also the type of ALF that had the most expansive admission and retention criteria and the highest resident acuity. For example, such facilities were more likely to retain residents who needed assistance with transfers and to retain residents who needed nursing care. Compared to the other ALF types, the high service/low privacy type also had a much higher proportion (35%) of residents who received assistance with three or more activities of daily living (ADLs), such as help with locomotion or using the toilet, as well as bathing and dressing. An estimated 12 percent of the ALFs across the country were in this category.

A fourth type of ALF offered high service and high privacy. Only 11 percent of all ALFs fell into this category. While resident accommodations were almost evenly split between rooms and apartments, nearly all (98%) of the accommodations were private. In addition, 41 percent of the high service/high privacy ALFs offered to arrange or provide nursing care and retain residents who needed such care. All had an RN on staff.


There were many variations in pricing structure among the ALFs nationwide. Some ALFs had a single monthly price for what they defined as basic services and accommodations. Other ALFs had multiple rates, varying with either the type of accommodation or the service package provided to the resident.

The most common monthly basic price was between $1000 and $1999 for both facilities with a single rate (i.e., 45% were in this range) and facilities with multiple rates (i.e., the range covered the most common rate for 52% of the ALFs). Thus, the most common basic price was between $12,000 and $24,000 per year. However, it is important to note that the average price was depressed by the presence of a very large number of ALFs (59%) that offered minimal or low privacy and services and had relatively low monthly rates. The most common base price for facilities with multiple rates was just over $22,000 per year for the high service/low privacy ALFs and just over $21,000 for the high privacy/low service ALFs. The basic annual charge was slightly more than $23,000 for the high service/high privacy ALFs.

EXHIBIT ES5. Distribution of ALF Monthly Basic Prices
Bar Chart: Single Rate, Most Common Multiple Rate and Highest Multiple Rate for <$1000, $1000-1999, $2000-2999, and >$3000.

These rates are striking for two reasons. First, in many ALFs, they do not cover all services. Residents often pay extra for such services as medication administration, transportation, and any assistance with ADLs or nursing care above the minimum covered by the basic rate in a facility.

Second, the rates are largely out of reach for most low-income older persons and unaffordable for many moderate-income elderly, unless they supplement their income with additional funds generated by disposal of their assets. According to data from the U.S. Bureau of the Census,6 40 percent of persons aged 75 and older had incomes in 1997 of less than $10,000 per year. Eight-four percent of persons aged 75 and older had incomes of less than $25,000 per year in 1997. This would make the average high service ALF or the average high privacy ALF unaffordable for the vast majority of older persons, particularly since they must also pay for other basic needs (e.g., supplemental insurance, out-of pocket spending on health care and medications, clothing).7

Even if some facilities embody the key tenets of assisted living's philosophical model, that is, policies emphasizing autonomy, dignity, and service flexibility that facilitate maximum independence and aging-in-place, the degree to which this model predominates in the industry is unknown.




Any attempt to understand assisted living and its role in providing long-term care to the frail elderly is hindered by the lack of a common definition of "assisted living." Places known as ALFs differed widely in ownership, auspice, size, and philosophy. Indeed, the results of this national survey identified four different types of ALFs within the industry that had very different patterns with respect to accommodations, services, staffing, policies on admission and retention of residents, and price. Some of these types, such as those offering high privacy and the high privacy/high service ALF, appeared to be consistent with the philosophy of assisted living. Other types, such as the low-minimal privacy/low service types were much closer to the traditional concept of domiciliary care or board and care, with few services and relatively little privacy.


The answer to this question is mixed. On the one hand, residents of assisted living facilities had considerably more privacy and choice than residents of most nursing homes and the majority of board and care homes. On the other hand, there was significant variability within the assisted living industry, and a substantial segment of the industry provided environments that did not appear consistent with the environmental aspects of the assisted living philosophy.


The ability of assisted living facilities to meet health-related unscheduled needs of residents is still an open question - in part because of facility policies (e.g., staffing, retention criteria or discharge policies) and in part because of potential constraints imposed by state licensure regulations.


The answer depends on one's concept of aging in place. For example, in most ALFs, a resident could move from relative independence (e.g., needing or wanting only meal preparation, housekeeping, and staff that can respond to emergencies) to a more complex stage at which the resident needed help with bathing, dressing, and managing medications and used a wheelchair to get around. If this "span" or change in needs were the definition of "aging in place," then the admission and retention policies of ALFs suggested that they were willing to allow residents to age in place.

On the other hand, if aging in place meant that the average consumer could select an assisted living facility and reasonably expect to live there to the end of his or her life, regardless of changes in health or physical and cognitive functioning, then the answer must be "no." In most ALFs, a resident whose functional limitations necessitated help with transfers or whose cognitive impairment progressed from mild to moderate or severe or who exhibited behavioral symptoms would be discharged from the facility. The same was true for a resident who needed nursing care for more than two weeks.

Thus, there was a limitation in terms of the ability of ALF residents to age in place.


Assisted living was largely not affordable for moderate and low-income persons aged 75 or older unless they disposed of their assets and spent them down to supplement their income. Further, to the degree that some assisted living facilities were affordable for low- and moderate-income older persons, they were more likely to be ALFs categorized as low-minimal service/low-minimal privacy facilities.



  1. A previous study funded by DHHS/DALTCP focused on licensed and unlicensed board and care homes. Two-thirds of those facilities had 10 or fewer beds, and the study found that none of the small homes called themselves assisted living. Few provided assistance with more than two activities of daily living (ADLs). As a result, we concluded that the vast majority of these small facilities would not provide the services generally considered a fundamental part of assisted living. Further, the small homes tended to serve a younger population of residents and a population that was more likely to have mental retardation, developmental disabilities, or persistent and serious mental illness. In addition, no states that licensed a specific category known as "assisted living" reported any facilities with fewer than 11 beds. For all of these reasons, ASPE and the project staff decided to exclude small homes from this study of assisted living for the frail elderly.

  2. These results are based on the most detailed information administrators provided about the accommodations (i.e., when they provided an exact count of the number of apartments and rooms). In another item on the survey, they were asked to estimate the distribution between rooms and apartments. The responses to this other item indicated that administrators estimated that 48% of the units were apartments and 52% were rooms.

  3. Many facilities had idiosyncratic policies about admission and retention. That is, the administrators responded "it depends" when asked about whether the facility would admit or retain residents with a specified condition. For example, one-quarter (26%) of the administrators responded "it depends" when asked whether they would admit a resident with moderate to severe cognitive impairment. One-third (33%) reported that "it depends" when asked whether they would retain a resident with moderate to severe cognitive impairment. When the "it depends" response was given, it was counted as a "no" since residents and families could not rely on either admission or retention in such instances.

  4. Numbers may not total 100% due to rounding.

  5. The differences between facilities that had a full-time RN and provided nursing care with their own staff and those that did not have a full-time RN on staff but were willing to provide or arrange nursing care are discussed at greater length in Section 7 of this report.

  6. U.S. Bureau of the Census, published data from the 1998 Current Population Survey, "Money Income in the United States, 1997" (series P60-200), Table 8, Income Distribution of Older Persons, 1997.

  7. This is based on estimates of annual income. More people could afford assisted living for some period of time by selling their assets, such as a family home, and using those funds to pay the monthly charges for assisted living.