Advantage and Choice: Social Relationships and Staff Assistance in Assisted Living

Abstract

Objectives.

To understand how “cumulative inequality” (CI), expressed as individual advantage and choice, and “external social supports” contribute to the quality of social relationships and perceptions of staff assistance for older individuals in different assisted living (AL) settings.

Methods.

Data are from 429 cognitively intact AL residents aged 60 years and older interviewed for the Florida Study of Assisted Living. Bivariate and multivariate statistical analyses show how individual advantage and choice and external social networks influence respondents’ social relationships and staff assistance in AL.

Results.

Controlling for resident and facility characteristics, being able to pay privately enhances resident satisfaction with staff assistance and having control over the move to AL is positively associated with perceptions of staff relationships and assistance. Maintaining contact with pre-AL friends predicts quality of coresident relationships, as does family contact. Regular contact with family buffers some of the disadvantages associated with CI for perceptions of staff relationships but not perceptions of staff assistance.

Discussion.

Individual advantage and choice influence the quality of staff relationships and assistance for AL residents but matter little for coresident relationships. External social relationships buffer some of the risks associated with CI for perceptions of staff relationships but not perceived quality of staff assistance. Findings highlight outcomes associated with CI, including predictable risks that disadvantaged elders face in particular types of AL settings, differential advantages others enjoy that influence positive perceptions of staff relationships and staff assistance, and the enduring importance of supportive social relationships.

AS the nationwide assisted living (AL) industry becomes more complex, specialized, and important as a site of residence, older individuals’ later life experiences are more often bounded within such settings. Successful aging in AL depends, in part, on finding an appropriate AL in the first place. That means getting the right package of services to meet care needs and developing meaningful social bonds with staff and other residents (Sikorska-Simmons, 2001; Street, Burge, Quadagno, & Barrett, 2007). For individuals with enough resources, the “right fit” can involve choosing among several attractive options before deciding on an assisted living facility (ALF) that meets preferences and needs (Ball et al., 2004). For others, AL choices are more constrained, with few meaningful choices in an industry dominated by high monthly costs and private pay arrangements (Street, Quadagno, & Burge, 2005). Low-income individuals may have few real alternatives to ALFs that are under-resourced, disproportionately dependent on public funds, and where residents have higher risks of suboptimal social and service experiences (Mor, Zinn, Angelelli, Teno, & Miller, 2004; Street, Burge, & Quadagno, 2009). However, external sources of social support may ameliorate outcomes associated with disadvantage that limits the range of AL choices some individuals experience.

LITERATURE REVIEW

The transition to AL represents a turning point, given substantial changes involved in moving from a home in the community to residential long-term care (LTC). Adjustment involves finding a suitable AL, downsizing to a new AL home, settling in, and establishing new social relationships. This residential transition triggers unfolding trajectories, with some individuals having sufficient resources to navigate successfully into a new setting, whereas others face less satisfactory outcomes. Important antecedents to successful aging in AL are the economic resources residents can use to their advantage as they choose their AL setting and the enduring social resources individuals can rely on as they acclimate to new situations.

Advantage, Choice, and Cumulative Inequalities

Within life course literature, researchers have emphasized how the accumulation of advantage or disadvantage can explain why patterns of heterogeneity exist among individuals’ economic standing or health status and why these inequalities tend to increase with age (Dannefer, 1987, 2003). The “cumulative advantage/disadvantage” (CAD) framework emphasizes the role of early life advantage, often operationalized as social location (e.g., socioeconomic status), in conferring an accumulation of later life advantage to individuals via better education, increased occupational opportunities, and access to health care (Crystal & Shea, 1990; Dannefer, 2003; Hungerford, 2007; O’Rand & Hamil-Luker, 2005; Shuey & Willson, 2008). Ferraro, Shippee, and Schafer (2009) advocate a somewhat different approach, considering instead how “cumulative inequality” (CI) conditions life chances. In contrast to CAD, CI emphasizes how disadvantage heightens exposure to risk and how advantage raises exposure to opportunities. A key insight of this theoretical framework is that, “while disadvantages may cluster,” they do not necessarily constitute an “omnibus disadvantage” (Ferraro et al., p. 422). This theoretical proposition, that disadvantage and advantage can be experienced simultaneously, albeit in different life domains, permits assessment of outcomes associated with economic disadvantage without presuming disadvantage in terms of social support. AL residents move from their homes based on different constellations of need, facility proximity, available resources, and existing social networks that influence the subsequent experiences of their AL residency. Compared with lower income counterparts, elders with higher incomes can select among a larger array of ALFs and exercise more control over whether and when to move to AL. This structured advantage, we expect, improves chances of finding an appropriate AL match for their particular care needs, preferences, and tastes. Furthermore, because residents’ ability to develop meaningful social relationships within ALFs is influenced by the fit between resident needs and what the AL has to offer (Ball et al., 2000, 2004), individuals’ initial level of advantage and choice of ALF likely impacts their social relationships in AL and perceptions of adequate staff assistance over the longer term.

Individuals needing AL services, but with meager financial resources, have many fewer options. The only AL option may be lower cost ALFs that admit poor residents whose care is subsidized under public programs like Medicaid, alongside some private pay residents. Facilities that cater to low-income residents typically have fewer resources for discretionary resident amenities beyond statutory requirements for AL care. They represent the most affordable ALFs in an industry dominated by private pay arrangements (Street et al., 2005). There may be little or no choice because, in many communities, affordable facilities are relatively few and ALFs with spaces for lower income residents may not be locally available (Golant & Salmon, 2004). Consequently, stratification processes in residents’ AL experiences can be partly understood within a CI framework that highlights inequalities AL residents confront as they exercise differential economic advantages and varying capacity to make meaningful choices about AL settings.

Coresident and Staff Relationships in AL

Internal social relationships are especially critical to residents’ adjustment, quality of life, and sense that the AL setting is home like (Port et al., 2005; Sikorska-Simmons, 2001; Street et al., 2007; van den Hoonard, 2002). Because many AL residents’ lives are bounded primarily by their contemporaneous experiences within the AL setting, daily interactions with coresidents and staff are central to residents’ quality of life. Relationships formed in AL are a new social resource for many residents, an expansion of their pre-AL social support network (Pitts, Krieger, & Nussbaum, 2005). AL internal social relationships are novel ones that are entirely contingent on other individuals present in particular AL settings—coresidents and their visitors and the people who work in AL. For some residents, these new ties are a sizeable expansion in their social network, at a time in life when social connections are typically dwindling (Ajrouch, Blandon, & Antonucci, 2005).

Residents’ individual characteristics influence relationships with other residents and staff. For example, a study of friendship networks in a Florida retirement community by van den Hoonard (2002) demonstrates that married residents have broader social networks, whereas nonmarried residents (especially widows) risk marginalization within such communities. Furthermore, physical health and cognitive status shape capacities for meaningful social relationships, with more physically adept and cognitively intact residents having more opportunities for positive social relationships than residents with greater impairment (Cummings, 2002; Mitchell & Kemp, 2000).

ALFs have organizational characteristics that may influence residents’ abilities to form friendships. For example, individuals may meet coresidents in pleasant surroundings or activities or perceive coresidents to be similar in tastes and interests, having selected the same place to live. Differences in an ALF’s ambiance may include food quality and mealtime experiences and other discretionary amenities (Kane & Kane, 2001; Perkinson & Rockemann, 1996; Sikorska, 1999; Sikorska-Simmons, 2001). Such organizational amenities are more available in well-resourced private pay facilities compared with ALFs that cater mostly to low-income residents.

ALF resident population characteristics may also influence residents’ potential to befriend coresidents because common interests and shared experiences often spur friendship. Consequently, the age distribution, manifesting conditions, and ability levels of residents in a given ALF may influence whether there are opportunities to pursue fulfilling, interesting activities with other residents. Activities and other preferences vary for different age groups, particularly when comparing nonelderly residents with frail elders (Cohen, Bearison, & Muller, 1987; Heumann, 1996; Kellam, 1992). Similarly, the incidence of mental illness and/or organic age-related brain diseases (e.g., Alzheimer’s disease and dementia) varies across different types of AL settings and may also affect the quality of coresident relationships. For example, despite being home to nonmentally ill frail elders, behavioral ALF resident populations are generally younger and more physically intact, with a higher proportion of mentally ill residents than other types of ALFs (Street et al., 2009). Aged residents have reported feeling physically threatened in mixed AL environments, which leads to their social isolation (Heumann, 1996). In contrast, ALFs that cater to the physical care needs of very frail individuals may offer fewer opportunities for more robust residents to develop meaningful relationships centered around activities within such ALFs.

Research on resident–staff relationships demonstrates that organizational characteristics of LTC settings affect resident-staff interaction. Both in nursing homes and in AL, residents report more positive staff interactions in facilities that promote better working conditions for staff (Ball et al., 2000; Barry, Brannon, & Mor, 2005; Bowers, Esmond, & Jacobson, 2000; Eaton, 2000; Gurnik & Hollis-Sawyer, 2003; Sikorska-Simmons, 2006). Because AL residents’ relationships with staff often grow closer and more intimate over time (Ball et al., 2000), low rates of staff turnover encourage stronger bonds between residents and staff. ALFs with greater resources are likely better equipped to promote staff-resident interaction because residents are more satisfied with their AL care when staff enjoy their work environment and experience job satisfaction and fulfillment (Sikorska-Simmons, 2006).

Because the need for more help is often the catalyst for elders’ transition to AL, staff assistance is another important facet of daily AL life. Finding an ALF that closely matches evolving care needs increases opportunities to age in place, to experience more positive relationships with staff, and to have more positive perceptions of staff assistance (Ball et al., 2000, 2004; Chapin & Dobbs-Kepper, 2001). However, the capacity to meet particular care needs varies across facility types. For example, ALFs that specialize in behavioral care (addressing specialized mental health needs) provide significantly less physical assistance than those providing for more traditional care needs (e.g., activities of daily living [ADLs] support). Thus, behavioral AL settings offer fewer opportunities to age in place (Street et al., 2009). Mismatches between care needs and the extent or type of ALF services offered may negatively influence residents’ perceptions of staff assistance.

Individuals with sufficient private funds to cover AL expenses in a place they want to live can more easily find ALFs with extra amenities and, perhaps, that are more equipped to employ staff with better training and skills, which promotes positive resident-staff interactions. For example, residents with dementia report better relationships with AL staff with specialized dementia training compared with only “standard” nursing assistant training (Winzelberg, Preisser, Zimmerman, & Sloane, 2005; Zimmerman, Sloane, et al., 2005) because specialized training provides staff with better skills to meet resident needs (Mead, Eckert, Zimmerman, & Schumacher, 2005). Appropriately trained staff report lower stress and greater job satisfaction, which in turn enhances staff–resident relationships and residents’ satisfaction with their care (Zimmerman, Williams, et al., 2005). In contrast, lower income elders’ choices may be limited to ALFs that accept Medicaid or subsidies. As Mor and colleagues (2004) document, LTC facilities with large proportions of Medicaid-dependent elders tend to be under-resourced. Lack of facility resources is expressed through lower staffing levels and higher staff turnover, which may compromise resident care (Mor et al.) and lessen individual residents’ perceptions of adequate staff assistance.

External Social Supports

Individuals whose family members remain active caregivers and intervene on their behalf may experience more positive and satisfactory relationships with staff and other coresidents compared with residents more isolated from or lacking family (Port et al., 2005). This underscores the enduring importance of preexisting convoys of social support as important resources that accompany at least some elderly individuals into AL. These external social connections provide emotional and instrumental support through family relationships and important friendship networks (Ajrouch, Antonucci, & Janevic, 2001; Antonucci & Akiyama, 1995; Crosnoe & Elder, 2002; Kahn & Antonucci, 1981) representing a social resource that may buffer some of the disadvantages arising due to cumulative economic inequality. Convoys of social support preexist entry to AL and these external connections with family and friends undoubtedly influence the experiences within it, as they did pre-AL transition. Data from Florida Study of Assisted Living (FSAL) provide an opportunity to analyze the relative influences of CI and social networks on AL residents’ perspectives on coresident and staff relationships and the quality of staff assistance they receive in particular types of AL.

The Florida Context

Florida has among the largest AL industries in the United States, and along with California and Pennsylvania, collectively accounts for 33% of all ALF units nationwide (Mollica, Johnson-Lamarche, & O’Keeffe, 2005). In Florida, AL is defined as a residential environment that provides a bundle of services, including meals, supportive services, medication, and 24-hr on-site supervision at a particular location. Florida has four separate licensure categories in state regulations. “Traditional” ALFs have only standard ALF licenses and provide housing, meals, and personal care services, including supervision of ADLs and self-administered medication. Facilities we define as “high-frailty” ALFs have either one or both of two state specialty licenses. Limited Nursing Services (LNS) licenses permit provision of additional services, such as changing routine dressings, passive range of motion exercises, applying heat and ice caps, and cutting toenails. Extended Congregate Care (ECC) licenses allow facilities to offer extra physical care services beyond LNS. Having the LNS and/or ECC licenses permits ALFS to provide extra services that give more options to help residents age in place, despite having physical needs that might disqualify residency in nonspecialty standard-licensed ALFs. Finally, “behavioral” facilities with Limited Mental Health (LMH) licenses provide standard AL services plus specialized behavioral care for individuals diagnosed with a mental illness (such as schizophrenia or bipolar disorder). Although differences in states’ ALF licensure determine policies and practices that shape the contours of resident populations in particular facility types and residents’ ability to age in place, all types of Florida ALFs and most types of ALFs in other states serve elderly residents. Although states certainly differ in the particulars of licensing regulations, Florida’s similarities to other states across a range of AL policies, organizational contexts, and resident populations (Street et al., 2009) make Florida an appropriate site for comparing how CI and external social supports shape older residents’ experiences in different types of ALFs.

Hypotheses

CI is expressed, in part, by the varying advantages and choice individuals can deploy when selecting an ALF. The ALF a person chooses (or settles for) determines the organizational characteristics that residents confront in their AL settings and the cast of characters available therein for potential social relationships. We expect such individual and organizational factors to be associated with resident perceptions of the quality social relationships they form with coresidents and staff and their perceptions of the quality of staff assistance they receive in AL.

Although data available to measure the amenities associated with different resource bases are limited in FSAL, ALFs that appeal to more affluent clients likely have more attractive physical surroundings, private rooms, and extra amenities (such as more varied social and recreational opportunities) and more, better trained and/or better paid staff. Thus, we expect individuals who pay privately to have opportunities for higher quality social relationships and to perceive higher quality assistance in the enhanced environments their economic advantages may permit. In contrast, individuals who depend on subsidies beyond their own resources may only be able to enter ALFs that routinely accept low-income residents, a risk that limits their capacity to choose among residential options. At the extreme, this can result in placements poorly matched to individual preferences and needs. We expect residents who reported little control over the decision to move to AL to report less positive perceptions of social relationships with coresidents and staff than those with greater control. We expect older residents in behavioral ALFs to have less positive perceptions of staff assistance compared with traditional and high-frailty ALFs, where services are more consistent with physical care needs. The diverse clientele of behavioral ALFs may also inhibit formation of satisfying social relationships, creating less positive perceptions of staff and coresident relationships. Despite expectations that indicators of CI (measured as private pay and control over move) will contribute to systematically different outcomes for coresident and staff relationships and staff assistance, external convoys of support may buffer some of the adverse effects of any mismatch arising from older residents’ relative advantages or disadvantages and the types of facilities they inhabit.

METHODS

Data are from face-to-face structured interviews with AL residents from the FSAL (N = 681). The FSAL resident survey respondent sample was broadly comparable with the gender, race/ethnicity, marital status, and age composition of the resident sample in a national AL study (Hawes, Phillips, & Rose, 2000). For this project, the first stage of the sampling frame used State of Florida administrative data (N = 1,886) to classify the population of ALFs by size (small, medium, or large facilities), payer type (exclusively private pay facilities vs. all others), and spatial distribution (by ALF size and payer source) throughout the 11 planning and service areas in Florida. We selected a purposive sample of facilities to visit (N = 148) that conformed to the facility size, payer type, and spatial distribution of the ALF population to conduct resident interviews. The analytic sample in this research is restricted to individuals aged 60 years and older, who scored 6 or higher (of 10) on the Short Portable Mental Status Questionnaire (used to assess cognitive function; Pfeiffer, 1974) and who provided valid data for all variables of interest. This analytic sample includes data from a sample of 429 residents living in 123 different Florida AL facilities. Approximately 85% of residents we asked were willing to be interviewed for the FSAL resident survey. We present descriptive statistics and a correlation matrix to depict the bivariate associations among key variables of interest. Ordinary least squares models then show how facility characteristics, advantage and choice, and external social support were associated with residents’ social relationships and perceptions of quality of staff assistance. Because ALFs were the primary sampling unit and residents’ responses within facilities may be correlated, we report robust standard errors adjusted for clustering of cases within facilities using STATA’s cluster subcommand.

Data

Dependent variables.—

We modeled effects on three outcomes. Residents can exercise the most choice about interpersonal relationships on the dimension of coresident friendships. Perception of relationship quality (“coresident relationships”) is measured by an index (Cronbach’s α .66) that averages responses to six dichotomous items (Do you regard any of the people who live here as your friends? Have you met other residents here with similar interests to yours? Do you feel like a member of the family? Do you attend most social events? Do other residents respect your privacy? Do other residents bother you? [reversed]). “Staff relationships,” unavoidable in ALFs, are measured by an index (Cronbach’s α .76) that averages responses to five dichotomous items (Do you feel that … you have friends among the staff? … the staff listens to you? … the staff shows affection and caring for you? … the staff shows you respect? … you feel comfortable discussing health concerns with staff?). Perceptions of the quality of “staff assistance” likely reflect both AL organizational characteristics and the extent of individual need for assistance. The perceived staff assistance index (Cronbach’s α .71) averages responses to five dichotomous items (Would you say that … staff are dependable? … you are satisfied with the assistance you receive? … staff are slow? [reversed] … staff are well-trained? … your complaints are taken seriously?) that assess whether residents receive the kinds and quality of assistance they expected from AL staff. For ease of interpretation, the three-scaled dependent variables were multiplied by 100, so that each scale ranges from 0 to 100 (original coding ranged from 0 to 1).

Independent variables.—

We control for “resident characteristics” including gender (female = 1 and male = 0), marital status (never married or divorced, widowed, and married [reference category]), and physical function (a standardized z-score index, Cronbach’s α .81, that averages residents’ scores on five ADLs [eating, dressing, transferring, bathing, and toileting] original coding ranged from not very hard at all =3 to very hard = 1). Because only 4.75% of the sample is non-White, we do not include a race variable in the models. “Facility characteristics” include facility size (measured continuously as the number of licensed beds) and facility type categorized by licensure category—traditional (standard license only = reference category), high frailty (LNS or ECC license), and behavioral (LMH license). Measures of “advantage and choice” associated with CI include whether residents paid privately for AL (private pay and/or LTC insurance = 1 and all other forms of payment = 0) and perceived control over the move to AL (How much control did you have over the decision to move to AL? some or complete control = 1 and little or no control = 0). Measures of “external social support” include family contact (an index, Cronbach’s α = .78, that averages two items: “How often does a family member visit you?” and “How often do you speak on the phone with family?” coded from never or almost never = 0 to daily or almost daily = 4) and external friend contact (“Do you have regular contact with friends that do not live here?” any contact = 1 and none = 0).

RESULTS

Table 1 provides descriptive characteristics of the analytic sample. Women predominate (72%) and nearly two thirds of the sample is widowed. Just over half reside in traditional facilities. Fifty four percent are private pay AL residents and just more than 71% of respondents said that they had control over the move to AL.

Table 1.

Descriptive Statistics for Analytic Sample

Variables (range for total sample)	% or M	n or SD
Social relationships
Resident relationship index (0–100)	70.1	29.5
Staff relationship index (0–100)	89.3	21.8
Staff assistance index (0–100)	85.8	24.0
Resident characteristics
Female (0, 1)	72.0	309
Never married or divorced (0, 1)	22.1	95
Widowed (0, 1)	64.6	277
Married (0, 1)	13.3	57
Physical function (z-score, −2.5–1.0)	0.02	0.9
Facility characteristics
Facility size (number of licensed beds)	80.3	66.5
Traditional facility	51.8	222
Behavioral facility	7.2	31
High-frailty facility	41.0	176
Advantage and choice
Private pay (0, 1)	53.9	231
Control over move (0, 1)	71.4	307
External social support
Family contact (0–4)	2.65	1.27
Friend contact (0, 1)	53.4	229
Observations	429

In descriptive analyses (data not shown), we distinguished among resident populations in different facility types. Behavioral facilities’ residents were younger; more likely to be men, to be single (never married or divorced), and to be physically healthier; less likely to be private pay; and have lower levels of family and friend contacts. High-frailty and traditional facilities have resident profiles that are similar in terms of age and gender; high-frailty ALFs are distinguished by poorer levels of physical function and fewer private pay residents.

The correlation matrix presented in Table 2 shows bivariate associations among facility type, key indicators of older residents’ CI (private pay and control over move), external social support, and social relationships with coresidents and staff. Advantages associated with CI measures have positive and significant associations with coresident relationships, staff relationships, and satisfaction with staff assistance. For external social relationships, regular family contact is positively associated with coresident and staff relationships and staff assistance, whereas friend contact is positively associated with the quality of coresident and staff relationships but not staff assistance. Behavioral ALF residents have less positive relationships with staff and perceptions of staff assistance in AL compared with residents of traditional and high-frailty ALFs. Greater proportions of both traditional and high-frailty facility residents are private pay compared with behavioral ALFs (Street et al., 2009). At the bivariate level, residents of behavioral facilities have significantly less external social support (family and friend contact) compared with residents of other types of ALFs.

Table 2.

Correlation Matrix Between Key Variables in Regression Models

Variable	1	2	3	4	5	6	7	8	9
Behavioral ALF	—
High-frailty ALF	−.23**	—
Private pay	−.23**	−.31**	—
Control over move	−.08	−.13**	.16**	—
Family contact	−.33**	.02	.10*	.19**	—
Friend contact	−.14**	−.06	.10*	.15**	.24**	—
Coresident relationships	−.05	−.05	.06	.12*	.24**	.19**	—
Staff relationships	−.19**	.02	.10*	.16**	.23**	.12*	.43**	—
Staff assistance	−.14**	−.01	.16**	.15**	.16**	.08	.29**	.57**	—

Note: ALF = assisted living facility.

**p < .01; *p < .05.

Tables 3, 4, and 5 show series of regression models that assess the relative influences of differential CI (private pay and control over move) and external social supports on the perceived quality of coresident and staff relationships and staff assistance. In each table, Model 1 includes resident and facility controls and measures of residents’ relative advantage and choice to assess how CI influences social relationships and perceptions of staff assistance. Model 2 in each table replaces the CI variables with measures of residents’ external social support to assess the influences of convoys of social support on the dependent variables. The third model includes both CI and external social support measures. Model 4 in each table adds interaction terms between the CI variables and family contact (external social support) to the variables of Model 3 in order to assess whether external social support mitigates disadvantages for AL residents. Because contact with family has consistent statistically significant effects in regression models with controls across all three dependent variables, but contact with friends does not, we test for an interaction between CI variables and family contact but not CI variables and friend contact in the final models.

Table 3.

Resident Perceptions of Relationships with Coresidents

	Model 1	Model 2	Model 3	Model 4
Resident characteristics
Female	10.759** (3.334)	8.890** (3.368)	8.807** (3.346)	8.238* (3.346)
Never married or divorceda	−4.563 (5.102)	−2.860 (5.471)	−2.390 (5.387)	−1.718 (5.122)
Widoweda	4.050 (4.407)	3.857 (4.755)	4.397 (4.650)	4.539 (4.496)
Physical function	4.963** (1.411)	4.452** (1.325)	4.486** (1.352)	4.359** (1.378)
Facility characteristics
Facility size	−0.049* (0.022)	−0.050* (0.020)	−0.051* (0.021)	−0.053* (0.021)
Behavioral ALFb	−3.496 (6.843)	3.154 (6.568)	3.780 (7.021)	6.383 (7.207)
High-frailty ALFb	−2.398 (3.458)	−2.345 (2.976)	−1.592 (3.127)	−1.457 (3.187)
Advantage and choice
Private payc	0.307 (3.714)		0.492 (3.387)	7.209 (7.654)
Control over moved	7.459* (3.591)		4.719 (3.250)	11.496† (6.761)
External social support
Family contact		4.057** (1.432)	3.806** (1.406)	7.071** (2.316)
Friend contact		8.333* (3.189)	7.889* (3.099)	8.146* (3.142)
Private Pay × Family				−2.477 (2.516)
Control × Family				−2.892 (2.404)
Constant	60.297** (6.440)	51.252** (6.543)	47.882** (7.386)	40.168** (8.322)
Observations	429	429	429	429
R2	0.10	0.14	0.14	0.15

Notes: ALF = assisted living facility.

^aReference category is married.

^bReference category is traditional ALF.

^cReference category is all other sources of payment besides private funds and long-term care insurance.

^dReference category is little or no control over the move to AL.

**p < .01; *p < .05; †p < .10.

Table 4.

Resident Perceptions of Staff Relationships

	Model 1	Model 2	Model 3	Model 4
Resident characteristics
Female	4.911† (2.545)	4.128 (2.522)	4.046 (2.510)	3.674 (2.489)
Never married or divorceda	−6.956 (4.428)	−6.262 (4.674)	−5.684 (4.447)	−5.536 (4.147)
Widoweda	0.744 (3.453)	0.252 (3.464)	0.896 (3.403)	0.803 (3.226)
Physical function	2.365† (1.210)	2.146† (1.179)	2.193† (1.175)	2.016† (1.146)
Facility characteristics
Facility size	−0.014 (0.019)	−0.013 (0.018)	−0.015 (0.019)	−0.017 (0.019)
Behavioral ALFb	−10.792† (5.673)	−8.303 (5.911)	−7.411 (5.830)	−5.573 (5.675)
High-frailty ALFb	0.701 (2.199)	0.054 (2.380)	1.030 (2.200)	1.436 (2.243)
Advantage and choice
Private payc	0.590 (1.784)		0.785 (1.674)	−2.835 (6.073)
Control over moved	7.184* (3.064)		6.016* (2.862)	21.729** (7.873)
External social support
Family contact		2.359* (1.100)	2.057* (1.017)	5.713** (2.106)
Friend contact		2.777 (2.145)	2.187 (1.989)	2.286 (1.936)
Private Pay × Family				1.258 (1.935)
Control × Family				−6.286** (2.323)
Constant	82.995** (5.417)	81.464** (6.170)	77.049** (6.647)	68.908** (8.236)
Observations	422	422	422	422
R2	0.10	0.10	0.12	0.14

Notes: ALF = assisted living facility.

^aReference category is married.

^bReference category is traditional ALF.

^cReference category is all other sources of payment besides private funds and long-term care insurance.

^dReference category is little or no control over the move to AL.

**p < .01; *p < .05; †p < .10.

Table 5.

Resident Perceptions of Quality of Staff Assistance

	Model 1	Model 2	Model 3	Model 4
Resident characteristics
Female	2.953 (2.602)	2.620 (2.536)	2.468 (2.501)	2.339 (2.522)
Never married or divorceda	−11.636** (4.204)	−12.655** (4.257)	−10.939** (4.049)	−10.923** (4.053)
Widoweda	−1.740 (3.147)	−2.880 (3.257)	−1.658 (3.182)	−1.723 (3.237)
Physical function	3.358** (1.216)	3.254** (1.224)	3.258** (1.216)	3.198* (1.224)
Facility characteristics
Facility size	−0.009 (0.019)	−0.008 (0.018)	−0.009 (0.019)	−0.010 (0.019)
Behavioral ALFb	−4.812 (5.533)	−5.927 (5.753)	−2.946 (5.863)	−2.259 (5.928)
High-frailty ALFb	1.860 (2.866)	−0.340 (3.042)	2.035 (2.888)	2.202 (2.978)
Advantage and choice
Private payc	4.537† (2.300)		4.640* (2.260)	3.180 (7.012)
Control over moved	6.628* (2.936)		5.978* (2.841)	11.953 (8.682)
External social support
Family contact		1.374 (1.021)	1.131 (0.970)	2.506 (2.043)
Friend contact		1.973 (2.829)	1.241 (2.649)	1.268 (2.620)
Private Pay × Family				0.507 (2.199)
Control × Family				−2.387 (2.747)
Constant	80.389** (5.507)	84.983** (5.395)	77.114** (6.379)	74.078** (8.333)
Observations	424	424	424	424
R2	0.10	0.08	0.10	0.10

Notes: ALF = assisted living facility.

^aReference category is married.

^bReference category is traditional ALF.

^cReference category is all other sources of payment besides private funds and long-term care insurance.

^dReference category is little or no control over the move to AL.

**p < .01; *p < .05; †p < .10.

Coresident Relationships

Residents who had control over the move to AL reported significantly better relationships with coresidents compared with individuals who did not; there were no significant differences by whether or not an individual paid privately (Table 3, Model 1). Family contact and friend contact were associated with more positive coresident relationships (Table 3, Model 2). When CI variables are entered into the model alongside external social supports (Table 3, Model 3), having control over the move to AL no longer significantly affects residents’ perceptions of their coresident relationships. Individuals with regular external sources of social support are still significantly more likely to report positive relationships with coresidents, net of the effect of CI. Consequently, in terms of their relative weight in shaping coresident relationships, having external social support is more important for developing coresident relationships than CI. However, the ability to pay privately or choose and external social support may be compensatory or buffering resources. That is, residents who lack external social supports or who cannot pay privately or exercise much choice about entering AL may be buffered from negative outcomes (like low-quality relationships with coresidents) by substituting one resource for the other. To test this, we include interaction terms between CI measures and family contact (Table 3, Model 4). Because the interaction term is not significant, neither resource appears to buffer a lack of the other for predicting the quality of coresident relationships.

Staff Relationships

Controlling for resident and facility characteristics, residents with greater control over the move to AL had more positive perceptions of staff relationships compared with residents with little or no control over the move (Table 4, Model 1). Compared with older residents in traditional and high-frailty facilities, those living in behavioral facilities reported less positive relationships with staff, underscoring the unique challenges frail elderly residents may encounter in behavioral ALFs compared with more conventional AL environments. Individuals with regular family contact were significantly more likely to report positive perceptions of staff relationships than those without family contact, but friend contact had no significant effect (Table 4, Model 2). After accounting for the influence of family contacts, residents of behavioral ALFs were not statistically different in perceptions of staff relationships compared with residents of traditional or high-frailty ALFs, although the direction of the coefficient remained negative. When CI and external social support measures are included together (Table 4, Model 3), both control over move and family contact remain significant predictors of better staff relationships. The final model (Table 4, Model 4) includes an interaction term for both CI measures and family contact. The negative and significant interaction between control over move and family contact indicates that, among older residents with little or no family contact, those with control over the move scored nearly 22 points higher on the index of staff relationships compared with those who had little or no control. However, among residents with regular family contact, having control over the move did not predict differences in staff relationships. Consequently, having social support in the form of family contact buffers some of the negative effects that disadvantages associated with CI contribute to staff relationship outcomes.

Staff Assistance

Even controlling for level of physical function, residents with more advantages perceived staff assistance more positively. Both private pay and control over move had significant positive effects on perceptions of the quality of staff assistance (Table 5, Model 1). External social supports had no significant effect on residents’ perceptions that staff assistance was adequate, after controlling for resident and facility characteristics (Table 5, Model 2). Residents with regular family and friend contacts were no more satisfied with staff assistance than were residents with fewer external social connections. When the effects of CI and external social support are considered together (Table 5, Model 3), the influence of CI on perceptions of staff assistance is still apparent. Residents who had control over the move and who paid privately were significantly more satisfied with staff assistance than those who did not. The final model predicting quality of staff assistance shows interactions between both CI variables and family contact (Table 5, Model 4). The lack of statistically significant interaction terms indicates that external social supports do not mitigate disadvantages associated with CI on resident satisfaction with staff assistance. Consequently, residents who lacked control over the move and who depended on public subsidies to pay for ALF are at greater risk of being dissatisfied with the staff assistance they receive and cannot substitute resources associated with external social support (e.g., regular family contact) to compensate for their lack of advantage.

Discussion

Individuals moving from a home in the community to an ALF may experience transitions characterized by uncertainty and anxiety. Successful transitions to AL depend, to a large extent, on developing place-specific social relationships and feelings of security in a new environment (Sikorska-Simmons, 2001; Street et al., 2007). In this research, we find evidence that elders’ AL relationships and satisfaction with staff assistance are shaped by processes associated with CI —the differential capacity to pay privately for AL and to exercise control over the move—and access to external networks of social support in AL settings. Successful adjustment to AL is a process at least partly determined by the relative advantages individuals can bring to bear on the timing and circumstances of their move to AL, the range of choices they can exercise in selecting ALFs most appropriate to their needs and tastes, and the potentially ameliorating effects of external social support.

An individual’s capacity to leverage personal resources influences the type and particular characteristics of ALFs they can select from the outset. Although the Florida AL industry had some overcapacity at the time of the FSAL, there were patches of scarcity for affordable ALFs for elders with limited means. Low-income elders are often restricted to ALFs that accept Medicaid level of reimbursement for services, and still must come up with the “hotel” or “room and board” costs themselves. This severely limits their capacity to choose from a range of ALFs with optimal service mixes to meet both care needs and preferences. In larger cities, with larger pools of affordable ALFs to choose among (Golant & Salmon, 2004), even low-income elders can exercise some choice over ALFs they regard as optimal given their circumstances. In smaller communities, or where affordable ALFs have reached capacity, older individuals who cannot live independently may have no choice but to settle for any ALF they can afford. For some, this may involve placement in a behavioral ALF with fewer physical assistance services, yet which is more likely to accept Medicaid and low-income residents compared with standard and high-frailty ALFs (Street et al., 2009). Such placements may not be ideal, but they may be unavoidable depending on the supply of affordable AL and the disadvantages a particular resident has accumulated by the time AL is needed. This raises two concerns: First, the limited ADL support in behavioral ALFs also limits the capacity for elder residents to age in place if frailty increases. This could precipitate the need for a second move or entering nursing home care at an earlier stage of decline than would be the case in a traditional or high-frailty ALF. Second, our bivariate findings indicate that older residents of behavioral ALFs report lower quality staff relationships and staff assistance compared with other ALF types. Given the importance of internal social relationships for successful adjustment to AL (Street et al., 2007), it seems likely that there are significant quality of life risks to elders whose scant resources limit options for ALFs they can afford to enter, especially if the ALF lacks some key service or staff capacity to meet their particular needs for assistance.

Factors associated with CI influence initial selection and the timing of entry into a particular ALF. More advantaged AL residents have the opportunity to move to ALFs that better match their needs if they do not find a suitable match at first pass. Residents whose range of ALF choices is severely limited by lack of resources may be unable to make adjustments to improve their quality of life if a mismatch occurs. Residents with few economic resources are more likely to be “stuck” in less optimal ALFs.

Advantaged elders can choose AL settings with more facility resources, services, and a more appealing physical setting and ambiance, and a more positive organizational setting for both residents and staff. Our findings indicate that being able to pay privately and having control over the move also influence residents’ capacity to form supportive social networks, especially with staff, and to receive good quality staff assistance. In part, our findings may reflect the fact that well-resourced ALFs tend to be better places for staff to work, perhaps because there are enough organizational resources to enhance staff training, recruitment, and retention, all of which facilitate better and resident-staff interaction and resident care (Sikorska-Simmons, 2006; Winzelberg et al., 2005; Zimmerman, Williams, et al., 2005).

Social relationships with coresidents are more a matter of choice than the obligatory and unavoidable resident–staff relationships. AL facilities with A greater resources may provide programming that provides better opportunities to meet and form meaningful coresident relationships. However, it is a testament to the resilience of elders that our findings indicate they are able to experience high-quality coresident relationships, whatever their placement situation. Contrary to expectations, there were no significant differences by facility type in terms of the reported coresident relationship quality. Elders with relative advantage and more choice did not reap additional benefits in terms of significantly more positive coresident relationships. Put another way, residents who were disadvantaged did not have more negative coresident relationships. Friendships are, indeed, priceless. What did predict differences in coresident relationships quite consistently was the influence of external social supports. Our findings suggest that maintaining prior social relationships may signal a level of individual gregariousness or sociality that is associated with a capacity for making new friends in an ALF, expanding the network of social support.

To the extent that CI influences relationships with coresidents and staff, it works mainly through having or lacking a sense of control over the circumstances that surrounded moving to AL, not just the capacity to pay privately for ALF residency. A plausible explanation is that sufficient resources work through complex mechanisms of advantage, choice, and control. That is, individuals with enough resources to enter the private ALF they prefer have more advantages of many kinds (and not just economic ones) that give them more control over a whole range of choices associated with moves to AL. With sufficient resources, an individual can make a series of choices, deciding how long and which community based services they might buy to postpone transition to AL, and have the capacity to research and select among a range of appropriate ALFs once a decision to move is made. Once in AL, sufficient resources also make necessary adjustments posttransition possible, by providing opportunities to move to another ALF if the first one does not suit an individual’s preferences or fails to meet care needs. Such capacity to exercise advantage and choice no doubt enhances many facets of individual well-being (which we do not test in this research), including those experiences in AL.

Previous studies have found that mechanisms of advantage do not automatically translate into outcomes that are more positive for AL residents because the reference may be an implicit comparison to one’s previous experiences in a community setting (Street et al., 2007). Another important consideration is how social support may ameliorate, or serve as a substitute, for CI arising from economic hardship in the AL setting. Our findings suggest that external social support, especially from family, may mitigate some of the negative consequences for otherwise disadvantaged residents in terms of staff relationships. We speculate that this is because family remains active nonresidential caregivers in elders’ lives even after the transition to AL (Port et al., 2005) and involved family may continue to advocate and contribute to their family member’s well-being in AL.

Our findings show several ways that advantage and choice shape social relationships and perceptions of staff assistance among older AL residents, and the important role that preexisting external social supports play in buffering some of the risks of economic disadvantage. Although CI exerts less influence on residents’ relationships with coresidents (which are a matter of choice), processes associated with CI are influential in determining residents’ staff relationships and their satisfaction with staff assistance. Our findings suggest that external social support can ameliorate some risks of disadvantage in terms of staff relationships but not the risks of poorer quality staff assistance. Consequently, when low-income residents can only choose suboptimal AL arrangements, family support is insufficient to make up for the disadvantage.

Some obvious caveats apply when considering the implications of these findings. First, restricting the analytic subsample to cognitively intact individuals aged 60 years and older limits generalizeability. Not included are the perspectives of a large proportion of elderly AL residents with moderate or severe cognitive impairment, neither did we assess the circumstances of younger ALF residents living alongside older residents. Both of these AL subpopulations may have experiences very different from older AL residents who are cognitively intact, suggesting important avenues for future research. Readers should consider these findings in light of cognitively intact older AL residents only.

Second, our capacity to assess how CI may be experienced in ALFs was limited to available measures in the FSAL resident survey, none of which were particularly refined indicators for the complexities that accompany the accumulation of inequality over the life course. For example, the capacity to assess relative individual economic advantage was very limited, especially in the absence of reliable income data. Instead, we used the private pay and/or LTC insurance variable as a proxy for individual economic advantage compared with those who either depended on subsidies for AL residency or whose incomes were so low they had few alternatives to ALFs that accepted mostly low-income residents. The measure of resident control over the move to AL was used to capture the idea that more economic resources expand elders’ array of LTC options and choices. However, choice of timing and place to enter AL is a complex process, no doubt influenced by other factors having less to do with lifelong advantage than some other, unmeasured processes (like proximity of an ALF to the original community or family members). Unfortunately, we cannot determine the relative influence of such processes with the data available in FSAL, nor can we offer a finely grained analysis of how or which accumulation of inequalities is most critical for positive and negative AL outcomes. Such research would require both more refined measures of factors associated with CI and a longitudinal research design to take the lifetime accumulation of inequalities into account.

Finally, despite many broad similarities, states’ licensure regimes also vary. Some states lack entirely the distinctive types of ALFs (traditional, high frailty, and behavioral) found in Florida. However, many states do have regulatory systems that permit specialized behavioral ALFs and mixing elderly and younger residents with mental illness into general resident populations (National Center for Assisted Living 2008). To the extent that other states are similar, our findings related to the impact of facility type on the quality of older residents’ social relationships and quality of staff assistance are informative.

These findings have several implications for this segment of the LTC industry. Our results identify another way that processes of inequality shape the entire life course, even in its later phases as individuals adjust to AL. As other research on the processes of cumulative inequalities has shown, we find that residents with greater advantage and choice generally have better prospects for successfully navigating the ALF milieu and settling into lives imbued with friendship and adequate supportive assistance. A key finding in this study is the way external social supports, such as long-standing relationships with family and friends, counter some of the obvious disadvantages of having low income and few choices. This research also underscores some of the risks experienced by disadvantaged elders who lack sufficient resources to enter ALFs that are a good fit for their needs. Individuals who are disadvantaged are at greater risk for less satisfying social relationships and inadequate assistance in AL. Moreover, given that our indicators of CI focus on the ability to leverage private economic resources for AL, these findings suggest that in the current context of economic uncertainty, even elders who previously could anticipate contemplating ALF options from economically advantaged positions may face such increased risks.

Finally, earlier research shows that behavioral ALFs offer less opportunity for frail residents to age in place due to restrictive admission and discharge policies and fewer physical support services (Street et al., 2009). Despite those organizational shortcomings, older residents do age in place in behavioral ALFs. Although the number of older behavioral ALF residents in our analytic sample was too small to generate statistically significant findings in most of the multivariate models, bivariate differences in the experiences of residents of behavioral ALFs and other types raise concerns. Reports of lower quality of staff relationships and staff assistance for older residents of such ALFs are troubling from both quality of life and a quality of care perspectives. Although their capacity to form friendships with other residents in behavioral ALFs may be an indicator of resilience among elders in such facilities, the negative outcomes on both staff variables signals an organizational incapacity within behavioral ALFs to adequately provide either the type or the quantity of care and assistance that low income cognitively intact elders need most. Given their mission to serve the needs of residents with mental illness (a mission very different from serving frail elders), it may be that behavioral ALFs are simply inappropriate placements for cognitively intact elder residents who are not mentally ill. This is not to suggest that behavioral ALFs would be appropriate for elders with cognition problems, either. The wisdom of placing any elder residents in behavioral ALFs should be a matter of further study for social gerontologists and debate for policymakers. Future research should attend the social–psychological experiences of elder residents in behavioral AL settings as well as potential organizational and policy interventions that might ameliorate these and other negative outcomes for disadvantaged (whether economically or relationally) elders, particularly those with little access to external social supports, in AL settings.

Funding

An earlier version of this article was presented at the Southern Sociological Society 2008 Annual Meeting. Support for this study was provided by Grant 1R03AG021770-01 from the National Institute on Aging, Contract M0330 from the State of Florida Agency for Health Care Administration and a Florida State University Research Foundation Program Enhancement Grant.