CAIRN-INT.INFO : International Edition
Volume
76
format_quote Cite share Share Switch to French

Full-text

A quarter of the population in France is now aged 60 or over, and this proportion will continue to increase. Population ageing brings with it a growing demand for care for disabled older adults, to which government authorities must provide appropriate responses to help them receive in-home or residential care. Are these responses the same across the country? The responsibility of départements for the care of these adults is likely to lead to differences, or even inequalities, depending on the place of residence. The author provides an original analysis of geographical disparities of the institutionalization rate of disabled older adults by département and identifies the main determinants.

1In France, since the local freedoms and responsibilities act of 2004, each département (hereafter: department, a mid-level administrative unit) defines and implements its social welfare policy, part of which is old-age policy. This decentralization is justified in that citizens’ needs are better understood and addressed at the local level. However, the financial, human, and material resources of departments, as well as their social welfare priorities, differ from one department to another. Institutional responses thus vary widely across France and not always in direct relation to the actual needs of older adults. This leeway given to departments may affect levels of in-home care service provision (Hégé et al., 2014) and residential care capacity [1] (Ramos-Gorand, 2020). [2] It also produces differences in the out-of-pocket expenses incurred by older adults above the amounts covered by the French health insurance scheme and by the allocation personnalisée d’autonomie (APA), the French allowance targeted to disabled older adults. It is the departments that decide the number of allowances awarded to older adults and the amounts granted (Billaud et al., 2013). They also have a say in the prices charged for in-home care services (Hégé et al., 2014) and nursing-home accommodation (Nirello, 2015). These differences linked to department-level autonomy in care provision may perpetuate older people’s decisions regarding support arrangements (Trabut and Gaymu, 2016) that do not necessarily match their actual care needs or preferences. Does care-home demand vary across departments? Are these differences explained by geographical and financial inequality of access, or rather by differences in needs and resources? If people are sensitive to price and constrained by availability, the heterogeneity of local policies may produce variations in take-up of certain types of care across individuals with identical characteristics (except their place of residence).

2It is useful to explore the constraints and factors contributing to disabled older adults’ choice of care arrangement in France. Geographical disparities may stem from needs that vary according to the degree or type of disability (Brunel and Carrère, 2019), unequal availability of formal or informal care (Ramos-Gorand, 2020), income differences, or stronger preferences for certain types of care.

3This article provides an overall analysis of the determinants of residential or home care for disabled older adults. As levels of care provision may change with demand (rising population of disabled older adults, strong demand for residential care places, etc.), it is important to consider the determinants of both demand and supply, taking account of the territorial dimensions to identify the factors underlying decisions of older people on one or other care arrangement. We examine the extent to which this decision is linked to cost and availability of supply. We also measure the scale of geographical disparities in residential care for disabled older adults. A multilevel analysis is then conducted to quantify the contributions of different individual and geographical factors to the disparities observed. The two surveys used cover the entire French territory and provide results representative at the department level.

I – Theoretical framework and literature review

1 – Theoretical framework

4This article is based on a theoretical model in which older adults compare care in their own home with residential care. They express a demand for the care arrangement that gives them the most satisfaction. The older adult is denoted s. The level of satisfaction is defined by a utility function Us that depends on his/her consumption of goods C, on the well-being he/she draws from the care arrangement CA, and his/her state of health H:

Box 1. Long-term care system in France: a policy at the department level

According to the Caisse nationale de solidarité pour l’autonomie (French fund for long-term care), public long-term care spending totalled €21.8 billion in 2017. Responsibility for formal care provision is shared by the medical, social, and medico-social sectors. These sectors are involved in several types of in-home care provision (Le Bihan and Martin, 2018): nursing care by community nurses either self-employed or working in home-nursing services, home helpers, home-care assistants, domestic helpers, and cleaners either self-employed or working for public or private home support services. The health cost in the home and in nursing homes is covered almost entirely by the national health insurance scheme; [a] 90% of other costs are publicly funded via the APA. [b] The departments manage large numbers of beneficiaries and decide the APA amounts awarded to each. Field surveys by Billaud et al. (2013) revealed major disparities across departments in the treatment of APA applications. Additionally, as the prices of the in-home support services financed as part of the APA are fixed at the department level, large variations in the quantity and price of the services provided occur across departments (Hégé et al. 2014). [c] The cost of residential care is fixed jointly by the region, the department, and individual care homes. It is financed by the health insurance scheme, the department council via the APA and/or the residential welfare allowance (aide sociale à l’hébergement, [ASH]), the state via tax deductions and housing benefits, and the disabled individuals themselves.
  • [a]
    Out-of-pocket costs to patients are between €0.50 and €2.00 per medical act; 60% of the fees charged by self-employed home-care nurses and 100% of the cost of service de soins infirmiers à domicile (SSIAD) interventions are covered by the health insurance scheme.
  • [b]
    Sources: DREES, social welfare surveys 2011 to 2015, CNAF, CCMSA, DARES.
  • [c]
    If the set prices are too low to cover their average costs, certain services may be obliged to cease their activities.

6The well-being function linked to the care arrangement is determined by the older adult’s place of residence: Is = 1 if in residential care, Is = 0 if living in his or her own home. In the latter case, well-being depends on the quantities of informal care AIs and formal care AFs received. Well-being linked to the care arrangement may also be related to the person’s state of health Hs (sense of security in a care home for people with cognitive impairment, for example) and his or her preferences (Xs).

7An individual’s demand also depends on budget constraints, which mainly include the cost of care, personal income (Rs), and any welfare payments received to subsidize formal assistance. [3] It takes the following form for in-home care:

9The amount received depends directly on the beneficiary’s functional health status Hs. Disabled persons can claim the APA for in-home care but not the welfare payments offered by pension schemes; the reverse is the case for non-disabled persons. It also depends indirectly on their income via a co-payment [4] rate thome(Rs). Under French law, the amount is based on the smaller of two values: hours of formal support consumed (AFs) or theoretical number of hours (AFtheors(Hs)). A socio-medical team assesses an older adult’s needs and attributes a GIR (groupe iso-ressources), which is a disability score (Appendix A). This GIR entitles them to a theoretical number of hours of formal support (AFtheors(Hs)), but the person may actually consume more or fewer hours (AFs). The product of the theoretical number of hours of support (AFtheors(Hs)) and the theoretical hourly rate (ptheor) defines the total euro amount the older adult needs (his or her support plan). The pricing system used by the department to determine the value in euros of the need for support is ptheor. This theoretical price is not necessarily the one charged by the formal carer (pAF). The amount (AFtheors(Hs) ptheor(1 – thome(Rs)) corresponds to the allowance to which the older adult is entitled to cover the cost of disability. All expenses above this amount are out of pocket. All resources (Rs and S) are used to pay for the consumption of goods Cs and formal human support (pAFAFs).

10If the person chooses to enter residential care, the budget constraint is as follows:

12The allowance entitlement depends on the person’s state of health (Hs) [5] and indirectly on his or her income via a co-payment rate tres(Rs). [6] The individual’s resources (income and allowances) serve to cover the cost of residential care (CTres) which depends on his or her health status [7] and consumption of goods.

13In short, the individual expresses a demand for a care arrangement that provides him or her the greatest utility while respecting budget constraints in a non-cooperative environment. The individual’s demand is not necessarily met if formal or informal supply is insufficient. This non-cooperative framework appears more consistent with reality than a cooperative framework since family carers may not only have work or family concerns unlinked to those of the people they support, but also diverging interests (gifts, childcare, etc.). It has been empirically validated by Checkovich and Stern (2002) and by Knoef and Kooreman (2011) in Europe. Furthermore, public supply is highly regulated; adjustment through price changes and the creation of additional residential capacity or services is limited and takes time to implement.

14This analysis framework reveals a range of factors underpinning the choice of residential care: the cost of the two types of care, formal and informal care availability, income, health status, and personal preferences. If individuals are price-sensitive, a high residential care cost should limit demand for this arrangement; conversely, a high cost of in-home care should increase demand for residential care. Choice should also be influenced by the level of availability of residential care or formal or informal in-home care. The effect of income is indeterminate as a higher income relaxes the budget constraint but also reduces the level of subsidies received. Consequently, demand for residential care should increase with the level of functional disability via a direct effect of the risk of entering a nursing home and via an indirect effect of subsidies and preferences.

2 – Literature review

15The factors revealed by the theoretical model are confirmed in empirical analyses. Regarding financial constraints, Roquebert and Tenand (2017) estimated a price elasticity of demand for formal in-home care of –0.4. A 10% increase in the hourly out-of-pocket price decreases the number of hours consumed by 4%. This estimate lies between that found by Bourreau-Dubois et al. (2014) (–0.55) and that found by Hégé (2016) (–0.15). These French studies focused on in-home care provision and did not consider the alternative option of residential care, for which out-of-pocket costs are much higher. Hoerger et al. (1996) showed that demand for residential care in the United States, unlike formal in-home care, is not affected by the generosity of public subsidies, even though they reduce out-of-pocket costs for individuals. Reschovsky (1998) found that demand is price-sensitive for married people only. This confirms the importance of including informal care availability in the analysis. In the United States, Muramatsu et al. (2007) identified disparities across different states. Among childless individuals, demand for residential care is lower in states favouring formal in-home care. Van Houtven and Norton (2004) and Charles and Sevak (2005) found a negative effect of informal care on entry into residential care, while Ettner (1994) found none. Similarly, risk is increased by the absence of a spouse (Metzger et al., 1997; Billaud and Gramain, 2006). These studies did not consider the potential shortage of formal care which, while affordable, may not be chosen because it is not available. Using Norwegian data, Theisen (2017) showed that in municipalities with abundant formal in-home care supply, this care arrangement is more frequently chosen than in those where supply is limited. Conversely, in municipalities with high residential care capacity, demand for this type of care is stronger, and in-home care is less frequent. Jette et al. (1995) showed a negative effect of receiving formal in-home care on entry into residential care for people with cognitive limitations.

16Few studies have established a link between income and residential care for the reasons given above. Börsch-Supan et al. (1990) in the United States found that the richest individuals are less likely to enter residential care. They can afford higher out-of-pocket costs for in-home care and can adapt their homes to their needs (Diepstraten et al., 2020, in the Netherlands). In the United States, Lockwood (2018) showed that people aiming to bequeath a large inheritance increase their savings and reduce their long-term care spending (including insurance). The income effect tends to disappear if other variables are factored in, notably property ownership (Garber and MaCurdy, 1990, in the United States) or subsidies (Hoerger et al., 1996, in the United States). Reschovsky (1996) found an effect of income on demand for care in the United States.

17An abundant literature has explored the effects of health status and disability on the probability of entering residential care (Luppa et al., 2010). Some studies show that this probability increases when functional limitations, cognitive limitations especially (Hoerger et al., 1996, in the United States; Nihtilä et al., 2008, in Finland), are combined with activity restrictions (Gaugler et al., 2007, in the United States) or when onset is sudden (Laferrère et al., 2013, in Europe).

18Approaches for managing disability differ substantially across the world. In Europe, while formal care is the norm in northern countries, family support and parent–child coresidence are more frequent in the south (Tomassini et al., 2004; Fontaine et al., 2007; Peyrache and Ogg, 2017). France is in an intermediate position, with variations across regions. Residential care is more frequent in the Brittany–Pays de la Loire region, for example (Trabut and Gaymu, 2016). These subnational differences have been observed across Europe (Mönkediek and Bras, 2014). Davey et al. (2006) showed that geographical disparities in Sweden are largely explained by differences in needs and that these disparities tend to disappear when municipality-level variations in needs are taken into account. Sundström et al. (2006) also identify very variable needs across municipalities in Sweden and explain the differences in terms of family-based and publicly provided care provision.

II – Method and data

1 – Method

19This article aims to model the care arrangement choice for disabled older adults. As publicly provided care is managed at the department level, the factors that come into play vary across departments. To identify these factors, multilevel models are used, with individual and departmental levels of analysis. They identify the individual and department-level variables that contribute to variations in residential care use for disabled older adults and estimate department-level random effects. We posit that the utility functions are linear:

22Parameters with the superscript 1 (respectively 2) represent preferences for residential care (respectively in-home care), RACesd is the estimated out-of-pocket cost for an individual in residential care, [8] RACdsd is that of in-home care. DISAFsd is the availability of formal in-home and residential care, DISAIsd that of informal care, Rsd is the older adult’s income, Hsd his or her functional health status and Xsd the sociodemographic characteristics liable to modify personal preferences. These variables are presented in Subsection 3. αd is the random department effect d. The probability of being in residential care is thus defined by:

24The model estimates the parameters βi= β1i – β2i and αd = α1d – α2d, assuming that ε2sd – ε1sd follows a logistic distribution. We assume that the department-level random effects follow a normal distribution and that they are independent of the other explanatory variables. They represent an above- or below-average risk of living in residential care stemming from the department of residence, all other things being equal. They are interpreted as department-level preferences for residential care, excluding control variables. Several specifications are tested to determine which variables are pertinent and whether their inclusion modifies the estimated parameters, including the department-level random effect. This analysis does not aim to measure a causal effect of supply on the probability of living in a care home, but rather to compare the supply and choice of care arrangement for disabled older adults across the departments of France.

2 – Data

25This article uses two complementary databases: the 2014 daily life and health survey (Vie quotidienne et santé [VQS]) representative of the population living in ordinary households (i.e. not in collective dwellings) and the 2015 survey on residential care for older adults (Établissements d’hébergement pour personnes âgées [EHPA]) representative of care-home residents.

26The VQS survey covered a representative sample of all people aged 60 and older living in their own homes in France (excluding Mayotte). A total of 166,800 over-60s responded to the survey by mail, Internet, or phone. The household response rate was 57%. [9] Using the weightings issued by DREES, the statistical department of the Ministry of Social Affairs and producer of these data, representative indicators can be provided at the department level.

27The EHPA survey was exhaustive, covering all care facilities in France (mainland France and overseas departments, excluding Mayotte). They provided information on their facility and on all its residents, so the dataset includes individual-level data. For our study, we selected people aged 60 and over, living permanently in either a privately run for-profit or non-profit care home, in a publicly run hospital or non-hospital facility or in a long-term care facility. Included were 402,304 residents. These types of dwellings are not covered by the VQS survey. In all, 73% of the care facilities answered all or part of the online survey. The weightings issued by DREES provide representative estimates at the regional level (Muller, 2018). Department-level weightings were difficult to calculate due to the low response rate of adult day care centres. As this category is not counted among residential facilities, it is excluded from the scope. The categories included in the sample nonetheless contained a sufficient number of facilities for each department. [10] The weightings were recalibrated using the total residential capacity by facility type and by department to obtain an approximate department-level weighting. [11] Partial non-response was imputed for the variables: having a partner, [12] difficulty washing, [13] mobility limitations, [14] cognitive limitations. [15] Results were compared with and without imputations to check that these imputations did not bias the results. [16]

28Assembling these two surveys with the recalibrated weightings gives a national estimate of the proportion of people living in a residential facility (also called institutionalization rate). At 3.9%, this rate is very close to that obtained via the population census (4%). Institutionalization rates were also compared at the department level via the combined VQS and EHPA survey data and the 2015 population census. A difference of up to 2.8 percentage points was observed. The proportion calculated with census data may be higher than that obtained with the two datasets used here because the census makes no distinction between care homes for older adults and other types of collective dwellings (workers’ hostels, religious communities, etc.). Most of the departments where the difference is large are young departments with large numbers of workers’ hostels (Paris region departments in particular). For the other departments with large differences (Lozère and Corse-du-Sud), the ranking by institutionalization rate is similar for both estimates.

29The selected population consists exclusively of adults aged 60 and older who have difficulty washing (Appendix A). Washing is an activity of daily living (ADL), a measure used widely in the international literature to gauge levels of need of assistance. This sample restriction means that the characteristics of individuals living in their own homes are similar to those of people in residential care. All individuals may thus be potential candidates for a place in residential care. Additionally, while some people in residential care have few disabilities, they are isolated cases; given the shortage of available places, care homes tend to give priority to people with the greatest needs. The final sample comprises 383,953 individuals aged 60 and older who have difficulty washing, of whom 20,991 live in their own home and 362,962 in residential care.

3 – The indicators

30A set of variables is selected describing the cost of the two types of care, formal and informal care availability, income, health status, and older adults’ preferences, taking account of the context at the department level. We estimate the theoretical monthly out-of-pocket cost of in-home care (RACdsd) and residential care (RACesd) for each individual s living in the department d. To do so, we attribute a disability score to each individual based on his or her cognitive and washing difficulties (Appendix A). [17] The out-of-pocket cost is estimated as the difference between the total monthly care cost and the public subsidies received. We use a set of departmental or regional level administrative data on the pricing and funding of care. The detailed data and the calculation method are described in Appendix B. The potential availability of informal care (DISPAIsd) is proxied by whether one has a partner. A cohabiting partner is a human resource for providing care to a disabled spouse. [18] We also use women’s inactivity and unemployment rates at the department level to gauge the availability of informal care, given that care is more often provided by women. High rates should be associated with more in-home elder care. For quantity of supply of formal care, we construct a categorical variable at the department level (DISPAF) (Appendix B), which takes the following values:

  1. Limited supply of in-home and residential care, but more in-home supply
  2. Limited supply of in-home and residential care, but more residential supply
  3. Abundant supply of in-home care only
  4. Abundant supply of in-home and residential care, but more in-home supply
  5. Abundant supply of in-home and residential care, but more residential supply
  6. Abundant supply of residential care only
  7. Limited supply of in-home and residential care, no specialization
  8. Abundant supply of in-home and residential care, no specialization This variable is constructed from data on the density of residential care places and of formal in-home carers. We expect limited residential care capacity to reduce demand for this type of care.

31As we have no individual-level information on personal resources, we use the proportion of beneficiaries of the minimum old-age pension (minimum vieillesse) in the department (Rd). The percentage of beneficiaries in the department reflects the socio-economic situation of older adults. The direction of this variable is uncertain. If we observe that the proportion of minimum old-age pension beneficiaries in the department raises demand for residential care, this means that the departments give priority to the most disadvantaged older adults for residential care places (which are more expensive), or that the richest older adults can afford to adapt their home to their needs or prefer to save their money in order to bequeath a larger amount to their heirs (Lockwood, 2018). If the effect is negative, then it is an income effect: it is easier for the richest older adults to cover the out-of-pocket costs of residential care than for the poorest.

32Our sample includes only people with need of assistance. The literature shows the extent of functional limitations and need of assistance associated with entry into residential care. To measure the type of need of assistance (Hsd), two forms of functional limitations are distinguished: mobility limitations and cognitive limitations (Appendix A). Because a combination of several functional limitations increases the risk of dependence (Cambois et al., 2005), a variable with four categories is used: no limitation, mobility limitations only, cognitive limitations only, mobility and cognitive limitations. The variables used to construct this indicator are not all based on a single definition because they are drawn from two surveys in which the questions asked were not necessarily identical.

33Preferences (Xsd) are influenced by the person’s age and sex. With increasing age, people’s perceived capacity to remain in their own home changes (sense of insecurity). We test a non-linear effect of this changing perception by including age, age-squared, and age-cubed in the model. The literature shows that women are more attached to their homes than men. We can therefore expect lower demand for residential care among women. Last, the department effect captures differences in preferences at the department and individual levels.

III – Results

1 – Abundant department-level supply and heterogeneous out-of-pocket costs

34The unequal distribution of supply across the departments of France is confirmed by the contextual data on supply of in-home and residential care grouped in the typology shown on the map (Figure 1A). Most departments are specialized, either for in-home care (Categories 1, 3, and 4) or residential care (Categories 2, 5, and 6). In-home care supply is greater in departments of northern France (Nord and Pas-de-Calais, Seine-Maritime and Eure), along the Mediterranean coast, in the south-west, and in the overseas departments of Réunion and Guadeloupe (Category 3). Central departments of Centre-Val de Loire and Pays de la Loire have a dense supply (Category 6) of residential care. Conversely, in the east, in the Paris region, and the departments of Haute-Corse, Martinique, and French Guiana, supply is quite limited for in-home care (Category 1) or residential care (Category 2) or both (Category 7).

35In parallel, we observe that in-home care is more costly in the Alpine and Paris regions (Figure 1B, dark grey zones). Conversely, average out-of-pocket costs for in-home care are lower in the Nord and Pas-de-Calais departments, and in the centre (very light green-grey zones). They are highest in Maine-et-Loire, at €358 per month, and lowest in French Guiana, at €145 per month. For residential care, they are higher in the overseas departments, along the Mediterranean coast, and in the Paris region (dark grey zones) and lower in Brittany, Centre-Val de Loire, and Pays de la Loire. The estimated out-of-pocket cost is highest in Paris, with an average of €3,160 per month, and lowest in Meuse, at €1,580.

2 – Descriptive analysis of the disabled population by care arrangement

36Slightly less than a quarter of the study population (persons aged over 60 and having difficulty washing) are in residential care. [19] This proportion varies considerably across departments, ranging from 7% in Réunion to 35% in Maine-et-Loire. The very low levels are relatively isolated. They are found in certain overseas departments, in some parts of the Paris region, a part of southern France, and notably Corsica and the Alpine region. The departments of Brittany and Pays de la Loire all have proportions above 30%. These differences are consistent with those identified by Trabut et al. (2021).

Figure 1

Cost and quantity of formal care supply at the department level

A

Cost and quantity of formal care supply at the department level

37Women represent 67% of the study population, accounting for 65% of the people living in their own homes and 74% of those in residential care (Table 1). Women less often have a partner when they are in residential care (15%) than when they live at home (41%). The people in residential care are 6 years older, on average, than those living at home and more frequently combine both physical and cognitive limitations (74% in residential care and 38% at home). Hence, the population in residential care is more preponderantly female, older, and in poorer health.

38Older adults in residential care live more frequently in departments where the proportion of recipients of the minimum old-age pension is lower than that of older adults at home (–0.5 percentage points). The departmental female inactivity or unemployment rate is slightly lower among the population living in residential care (38.80%) than among those living at home (39.19%). People in residential care live in departments where the residential care capacities are higher than for people living at home. The average theoretical out-of-pocket cost of in-home care among people living in residential-care is higher than the one estimated among people living at home (€271 per month vs. €267 per month). By symmetry, the average theoretical out-of-pocket cost of residential care among people living at home is higher than the one estimated among people living in residential care (€2,049 per month vs. €1,989 per month). Hence, by comparison with people living at home, care-home residents live in departments with a low proportion of beneficiaries of the minimum old-age pension, a low female inactivity rate (hence less informal care availability), a high residential care capacity, a high out-of-pocket in-home care cost, and a low out-of-pocket residential care cost.

3 – Determinants of residential care

39The results of the multilevel model of the probability of being in residential care for people with difficulty washing are shown in Table 2.

The type of care chosen is the one for which out-of-pocket costs are the lowest

40As expected, a high out-of-pocket cost for in-home care increases the probability of being in residential care, while a low cost decreases it. The effect of the out-of-pocket cost is much stronger for residential care than for in-home care. Even when the out-of-pocket cost of in-home care is high, the price effect takes precedence.

The role of formal care availability becomes negligible for an equivalent out-of-pocket cost

41When differences in theoretical out-of-pocket costs are not controlled for (Table 2, Model 5), all the situations where residential care supply is abundant, or at least more abundant than in-home care supply, increase the probability of living in a care home. All these effects disappear for equivalent out-of-pocket costs, except in departments with limited formal supply and slightly greater supply of residential care. The effect is even reversed for departments with abundant in-home and residential care supply, but more in-home supply. People living in these zones with greater in-home care supply more frequently stay in their own home. This disappearance of the effect of formal care availability (residential especially) can be explained by the link between supply density and cost: areas with abundant supply also provide services with a lower inhome and residential out-of-pocket cost than other areas (price competition or economies of scale). If supply of one or other option is abundant but very expensive, however, people have little incentive to choose it.

Table 1

Descriptive statistics of the sample by place of residence

Table 1
Persons aged 60+ who have difficulty washing All Own home Residential care Significance of difference % of people in residential care(a) 22.60 -- Out-of-pocket cost (in euros) of formal care In-home (RACdsd) 268 267 271 *** Residential (RACesd) 2,035 2,049 1,989 *** Formal care availability in the department (DISPAFd) 1. Limited supply of in-home and residential care, more in-home supply 6.38 7.05 4.08 *** 2. Limited supply of in-home and residential care, more residential supply 11.41 11.30 11.81 *** 3. Abundant supply of in-home care only 33.97 35.28 29.47 *** 4. Abundant supply of in-home and residential care, more in-home supply 8.90 8.59 9.94 *** 5. Abundant supply of in-home and residential care, more residential supply 3.31 3.08 4.08 *** 6. Abundant supply of residential care only 23.82 22.34 28.89 *** 7. Limited supply of in-home and residential care, no specialization 7.60 7.96 6.40 *** 8. Abundant supply of in-home and residential care, no specialization 4.61 4.40 5.33 *** Informal care availability (DISPAIsd) With a partner (%) 36.27 41.23 14.84 *** Female inactivity and unemployment rate in the department (%) 39.11 39.19 38.80 *** Resources (department level) (Rd) Beneficiaries of old-age minimum pension (%) 3.43 3.55 3.05 *** Functional limitations (Hd) (b) Physical and cognitive 45.94 37.72 74.12 *** Physical only 43.76 55.25 4.41 *** Cognitive only 4.74 1.99 14.15 *** No functional limitation 5.56 5.04 7.33 *** Preferences (Xsd) Women (%) 66.68 64.50 74.14 *** Age in years 81.60 80.17 86.49 *** Unweighted numbers 383,953 20,991 362,962 Weighted numbers 2,497,091 1,932,752 564,339

Descriptive statistics of the sample by place of residence

(a) Proportion of people in the EHPA database.
(b) cf. Appendix A.
*p < .05. **p < .01. ***p < .001.
Note: Percentages are weighted.
Interpretation: Among people with difficulty washing and who are living at home, 64.5% are women; in residential care, 74.1% are women.
Coverage: Individuals aged 60+ with difficulty washing, who are living at home or in residential care in France (excluding Mayotte).
Sources: VQS 2014 and EHPA 2015 surveys, DREES.
Table 2

Probability of living in a care home for disabled people (difficulty washing), multilevel regression

Table 2
Model 1 Coef. Model 2 Coef. Model 3 Coef. Model 4 Coef. Model 5 Coef. Model 6 Coef. Out-of-pocket cost for formal care In-home care (logarithm) 0.099*** 0.103*** 0.102*** Residential care (logarithm) –6.626*** –5.595*** –5.551*** Formal care availability in department (Ref. = 7. Limited supply of in-home and residential care, no specialization) 1. Limited supply of in-home and residential care, more in-home supply –0.001 –0.052 –0.152 2. Limited supply of in-home and residential care, more residential 0.677*** supply 0.602* 0.384** 3. Abundant supply of in-home care only –0.381 –0.265 0.200 4. Abundant supply of in-home and residential care, more in-home supply –0.739* –0.579* 0.276* 5. Abundant supply of in-home and residential care, more residential supply –0.426 –0.340 0.591*** 6. Abundant supply of residential care only –0.281 –0.171 0.611*** 8. Abundant supply of in-home and residential care, no specialization –0.711 –0.513 0.643*** Marital status (Ref. = No partner) With a partner –1.324*** –1.294*** –1.324*** –1.294*** Female inactivity or unemployment rate in the department –0.002 –0.015 0.012 –0.033* Share of beneficiaries of minimum old-age pension in the department –0.119*** –0.132*** 0.011 –0.017 –0.108*** –0.020 Functional limitations (Ref. = No limitation) Physical and cognitive 0.389*** 0.457*** 0.528*** 0.532*** 0.458*** 0.531*** Physical only –2.934*** –2.904*** –3.058*** –3.018*** –2.904*** –3.017*** Cognitive only 2.127*** 2.025*** 2.121*** 2.007*** 2.025*** 2.007*** Age –0.941*** –0.812*** –0.882*** –0.773*** –0.812*** –0.773*** Age² 0.011*** 0.010*** 0.010*** 0.009*** 0.010*** 0.009*** Age3 –0.000*** –0.000*** –0.000*** –0.000*** –0.000*** –0.000*** Sex (Ref. = Male) Female 0.313*** 0.001 0.319*** 0.011 0.001 0.011 Constant 27.853*** 24.260*** 75.797*** 65.388*** 23.274*** 65.601*** Number of observations 383,893 383,893 383,893 383,893 383,893 383,893

Probability of living in a care home for disabled people (difficulty washing), multilevel regression

*p < .05. **p < .01. ***p < .001.
Note: Standardized weighting used.
Interpretation: A combination of physical and cognitive limitations increases the probability of living in a care home. In Model 1, the coefficient is 0.389 and is significant at the 1% level.
Coverage: Individuals aged 60+ who have difficulty washing and who are living at home or in residential care in France (excluding Mayotte).
Sources: VQS 2014 and EHPA 2015 surveys, DREES.

Having a partner protects against entering residential care

42Having a partner lowers the probability of living in a care home, as found by Freedman (1996). Beyond the partner’s role as a potential carer, there may be financial constraints linked to the cost of two separate places of residence if one spouse enters a care home while the other is still alive. The proportion of inactive or unemployed women does not have any effect. However, when formal care availability is not controlled for, the higher the number of inactive or unemployed women, the lower the proportion of older adults in residential care. This indicates that the areas where in-home care supply is abundant or residential care is limited are also areas with a high proportion of inactive or unemployed women at the department level, providing a large pool of potential carers. In-home care is often provided by women with few qualifications, and turnover may be very high.

Demand for residential care is held back by income differences

43Not controlling for the out-of-pocket cost of formal care, the proportion of beneficiaries of the minimum old-age pension reduces the probability of being in residential care. Demand for residential care is lower in the poorest departments (simple income effect). The effect becomes non-significant, however, after controlling for out-of-pocket cost. This means that the departments where out-of-pocket costs are low (low costs or high subsidies), for residential care especially, also have a high proportion of minimum pension beneficiaries. The disappearance of the income effect when out-of-pocket cost is included shows that departments with a high proportion of minimum pension beneficiaries provide more generous support to enable older adults on low incomes to pay for residential care. This suggests that higher levels of support reduce inequalities of access to this type of care.

More frequent residential care for older adults with severe functional limitations, but also for those with the simplest care needs

44Like that of Arnault (2015), this analysis shows that having several functional limitations or having cognitive limitations only (beyond difficulties washing, common to all respondents) increases the probability of being in residential care. Conversely, among disabled people, those with mobility limitations only are less likely to live in residential care than those with no functional limitations, all other things being equal. This may be because individuals with neither mobility nor cognitive limitations but who have difficulty washing have other disorders not covered by these two limitations. Given that care homes operate at full capacity and the funding they receive depends on the average degree of disability of their residents, they may more readily accept individuals who have difficulty washing but no functional limitations as they have less expensive care needs.

Sex and age seem to explain differences in residential care preferences

45Age increases the probability of being in residential care, but not linearly. Among the study population, the effect of age increases with an increasing rate, all other things being equal, a finding consistent with the meta-analysis of Gaugler et al. (2007). This variable captures both care needs not included in the disability variables used and individual preferences. The capacity of older adults, either real or perceived (by themselves or their family), to stay in their own home evolves as they advance in age. Being a woman increases the likelihood of being in residential care only if marital status is not controlled for (Table 2, Models 1 and 3). After controlling for marital status, sex has no effect, a finding that contrasts with that of Gaymu et al. (2006) but is consistent with that of Bouvier et al. (2011) for adults aged 80 and older. When the department effect is not controlled for, this negative effect is verified, indicating that women are better able to remain in their own homes in certain geographical areas.

A strong geographical pattern of residential care use

46The maps in Figure 2 show the random departmental effects estimated from the models. The dark grey departments have a much higher institutionalization rate than would be expected according to the variables used. They reflect the differences in probability of living in residential care across departments with equivalent characteristics. The distribution of Model 2 is used as reference and the random department effect obtained is separated into quartiles. [20] In Model 2, the departments in Brittany, Pays de la Loire, and Picardie have higher proportions of older adults in residential care for equivalent levels of needs, income, and informal care supply, while departments close to the Spanish border, along the Mediterranean coast, and in the east have lower proportions.

47When supply variables are added (quantity and out-of-pocket costs) (Table 2, Model 4), the north-west–south-east diagonal is less marked. Proportions in residential care remain high in Brittany, but the Mediterranean departments (excepting Corsica) and the Paris region become those with the highest random effects. The reverse is observed for Pays de la Loire and Centre-Val de Loire. This shift is linked to the two dimensions of supply included in the model. When only the quantity of supply is added (Table 2, Model 5), proportions are lower in Pays de la Loire, central France, and Picardie. The Doubs department is the only one where more abundant supply produces a notable increase in demand for residential care. The inclusion of out-of-pocket costs (Table 2, Model 6) modifies the map more substantially. If out-of-pocket expenses were identical everywhere, the Paris regions, Hauts-de-France, and the Mediterranean coast would have significantly higher demand, while Centre and Pays de la Loire would have much weaker demand.

Figure 2

Disparities across departments of the random department effect on the probability of living in residential care in France

Figure 2

Disparities across departments of the random department effect on the probability of living in residential care in France

Note: Standardized weighting in models.
Interpretation: The dark grey departments have a much higher institutionalization rate (random effect) than would be expected according to the control variables used in the model. Light green ones have a lower rate than would be expected. The department-level random effect ranges from – 0.980 to 0.606 for Model 2; from –1.599 to 1.851 for Model 4; from –1.600 to 0.611 for Model 5; from –1.680 to 2.319 for Model 6.
Coverage: Individuals aged 60+ with difficulty washing, who are living at home or in residential care in France (excluding Mayotte).
Sources: VQS 2014 and EHPA 2015 surveys, DREES.

V – Limitations and conclusion

48This study’s results are based largely on the data and measures used. Out-of-pocket costs are estimated from pricing data that are not exhaustive, and we do not have enough data concerning older adults (notably GIR and income) to make accurate estimates. The old-age allowance (for independent living or residential care) is estimated from the mean allowance per disability score (in-home or residential) in the department. For the out-of-pocket cost of residential care, this is acceptable because the sum of old-age allowance plus tax credits gives a relatively stable amount across different income levels. For inhome care, on the other hand, this method underestimates the amount for the poorest individuals and overestimates it for the richest. We tested several measures to estimate the disability score and different in-home care prices. The results remain similar (mainly due to the limited number of variables that can be used to estimate it). Overall, the measure can be interpreted more as a direct cost than as a true out-of-pocket cost.

49Drawing on more than one survey means using data based on different definitions and data collection methods. These differences may generate biases that affect situations of in-home care and residential care in dissimilar ways. Data were harmonized (reweighting, imputation of missing values, etc.) to obtain institutionalization rates and distributions of individual characteristics that were consistent with those of the census. The EHPA survey only provides information on individual situations at the time of the survey, but this information is liable to change as duration of residential care increases. To address this problem, we ran an analysis limited exclusively to people in residential care for less than 1 year. The results are similar to those presented. To control for potential movement between departments and a cultural effect linked to greater acceptance of residential care, the analysis was conducted on individuals in care homes in the same department as their last home address. Department-level random effects are weaker; people in a care home outside their department of origin may have an intrinsic preference for residential care. By removing these individuals, only people whose preferences and habits are similar because they live in the same department are considered.

50Despite these limitations, this study estimates and explains the interdepartmental variations in the management of the long-term care system in France. It reveals significant geographical disparities in the probability for a disabled older adult of living in residential care. These disparities are linked to demand (needs, preferences, and income constraints) and supply, and to unexplained department-level factors. The creation of long-term care facilities, the funding of care, and the prices charged by care homes are overseen by department councils. Their effect on care choices suggests that levels of availability and out-of-pocket costs are two important public policy levers. This article shows that most of these differences are explained by supply quantity, as they are less pronounced for equivalent density of supply. Pricing, on the other hand, accentuates differences in care choices; departments where care is most expensive are also those where the population is most willing to pay. We also found that the protective effect of being a woman with respect to institutional care disappears when department-level differences are taken into account. That more women stay in their own homes appears to be linked to department-level markers. If formal in-home care provision is insufficient, women cannot stay in their own home, even if they would prefer to do so or would be more able to cope. Managing the quantity of care available appears to be a key factor for public elder care policy. Greater availability broadens the range of care options open to individuals. Increased supply would doubtless also improve the working conditions of personnel working in the understaffed medico-social sector.

Acknowledgements:

This article draws upon research conducted for a PhD thesis that received financial support from the Mission recherche of the Direction de la recherche, des études, de l’évaluation et des statistiques (MiRe-DREES) and the Caisse nationale de solidarité pour l’autonomie (CNSA) as part of the ‘Handicap et perte d’autonomie - session 6’ call for projects.

Appendices

Appendix A

Table A.1

Method used to construct the GIR score and the measures of functional limitations

Table A.1
Constructed variable Value of the variable VQS: use of questionnaire responses EHPA: use of the rating (A, B, C, with A = few limitations and C = severe limitations) on the axes of the AGGIR administrative grid GIR score 1 or 2 Total incapacity for one of the following limitations: • Understand others or be understood by others • Concentrate for more than 10 minutes • Remember important things • Solve problems of daily life (like following directions or counting money) C on one of the following axes: • Orientation in time • Orientation in space • Behavioural consistency • Orientation in communication 3 Totally incapable of washing him/herself or at least severe difficult for one of the following limitations: • Understand others or be understood by others • Concentrate for more than 10 minutes • Remember important things • Solve problems of daily life (like following directions or counting money) C on the Washing axis or at least B for one of the following axes: • Orientation in time • Orientation in space • Behavioural consistency • Orientation in communication 4 Severe difficulties washing B on the Washing axis 5–6 All other situations All other situations Disabled Yes ‘Does she have difficulty washing?’: Yes, slight difficulty; Yes, severe difficulty; Cannot wash herself B or C on the Washing axis No ‘Does she have difficulty washing?’: No A on the Washing axis Cognitive limitations Yes ‘Does she have difficulty…’: Yes, severe difficulty; Total incapacity; difficulty for at least one of the following items • Understand others or be understood by others • Concentrate for more than 10 minutes • Remember important things • Solve problems of daily life (like following directions or counting money) B or C one at least one of the axes: • Orientation in time • Orientation in space • Behavioural consistency • Orientation in communication No ‘Does she have difficulty…’: No; Yes, slight difficulty; difficulty for all the following items: • Understand others or be understood by others • Concentrate for more than 10 minutes • Remember important things • Solve problems of daily life (like following directions or counting money) A for all the following axes: • Orientation in time • Orientation in space • Behavioural consistency • Orientation in communication Mobility limitations Yes ‘Does she have difficulty climbing stairs to the next floor or walking 500 metres?’: Yes, slight difficulty; Yes, severe difficulty; Cannot wash herself B or C on the Indoor mobility axis No ‘Does she have difficulty climbing stairs to the next floor or walking 500 metres?’: No 2A on the Indoor mobility

Method used to construct the GIR score and the measures of functional limitations

Appendix B

Table B.1

Variables drawn from department-level data

Table B.1
Indicator Calculation method Income variable Proportion of minimum pension beneficiaries Number of minimum old-age pension beneficiaries divided by the population aged 60+ on 31 December 2014. Sources: CNAF, MSA, Pôle Emploi; ASSEDIC national database; FSV and DREES survey of CNAV, RSI, MSA, CDC, ENIM, SNCF, CAVIMAC for metropolitan France; INSEE, population estimates on 1 January 2015. Informal supply variable Female inactivity and unemployment rate Sum of the female unemployment rate and the ratio between the number of inactive women (whole population aged 15–64 minus women in employment and unemployed women looking for work) and the whole female population aged 15–64. Sources: INSEE, 2014 population census primary analysis and local unemployment rates, overseas departments labour force survey. Formal in-home care variables Community nurses Number of community nurses per 1,000 population of all ages. Sources: DREES / ASIP-Santé, Adeli 2015 directory. SSIAD Number of SSIAD places per 100 population aged 75+. Sources: DREES–DRJSCS, Panorama Statistique Jeunesse Sports Cohésion Sociale and INSEE, provisional population estimates on 01/01/2015. Home helpers employed by private individuals Number of hours per 1,000 inhabitants, all ages. INSEE, DADS grand format 2015 - Fichiers détail. Occupational categories 563b and 563c. Activities OQ and RU. Domestic helpers working in home support and care services Number of hours of assistance provided to older adults per 100 inhabitants aged 75+. Sources: DGE, NOVA 2015. Prices charged by authorized providers Prices charged by authorized providers (mean rates set by the department council and the welfare services to bill and refund hours of care via the APA personalized care plan). Sources: DREES, SolvAPA database, database of APA operations and tariffs in 2015. Hourly rate of private employers Hourly rate of private employers. The data concern private individuals who employ domestic helpers and pay them directly. The hourly rate is calculated as the total net payroll divided by the total declared volume of hours. This information is available at the regional level, distinguishing two possible exemptions: for APA beneficiaries or for persons aged over 70 (we average the two). Sources: ACOSS, Stat 2014. Mean amount of in-home APA Mean amount of in-home APA paid by the department council. Sources: DREES, social welfare surveys 2014. Formal residential care supply variables Residential care capacity Number of places in care homes or long-term care facilities per 100 people aged 75+. Sources: DREES, EHPA 2015. Median care-home price Median monthly single room charge for permanent residents (including for residents with GIR 5-6 dependence score). Sources: CNSA portal on 31 December 2016 (https://www.pour-les-personnesagees.gouv.fr). Median care-home price for GIR score 1–2 Median monthly single room charge for permanent care-home residents, GIR score 1–2. Sources: CNSA portal on 31 December 2016. Median care-home price for GIR score 3–4 Median monthly single room charge for permanent care-home residents, GIR score 3–4. Sources: CNSA portal on 31 December 2016. Mean APA amount for residential care Mean amount of residential care APA paid by the department council. Sources: DREES, social welfare surveys 2014.

Variables drawn from department-level data

Formal care typology calculation method

51One of the problems with the quantitative formal care variables at our disposal is their denomination in different units. To compare them, we rank them in increasing order (for example, the department with the fewest community nurses takes a value of 1 for the density of community nurses variable). A composite variable of in-home care supply is calculated at the department level, corresponding to the mean ranking of each in-home care supply variable. This composite variable can be matched against the distribution of residential care supply to identify both the degree of specialization in in-home or residential care in each department and the resources it allocates for each type. A typology combining these two dimensions can then be constructed. A department is considered to have high in-home and/or residential care supply if supply levels are above the national median. A department is not specialized in either residential or in-home care if the difference between its rankings for in-home care and for residential care is below 9 in absolute value. [21] If its ranking is higher for residential care than that for in-home care (modulo 9), the department is specialized in in-home care. The last category corresponds to specialization in residential care. We combine these two dimensions to obtain a categorical variable that takes the following values:

  1. Limited supply of in-home and residential care, but more in-home supply
  2. Limited supply of in-home and residential care, but more residential supply
  3. Abundant supply of in-home care only
  4. Abundant supply of in-home and residential care, but more in-home supply
  5. Abundant supply of in-home and residential care, but more residential supply
  6. Abundant supply of residential care only
  7. Limited supply of in-home and residential care, no specialization
  8. Abundant supply of in-home and residential care, no specialization

Method for calculating out-of-pocket costs of in-home and residential care

52The total cost of residential care is determined on the basis of the median monthly cost of a permanent care-home place (including the dependence charge for a person with a GIR 5-6 dependence score) in the department (TAR_HEBd) plus the median additional dependence charge in the department: ΔGIR12d if the individual has a GIR score of 1 or 2; ΔGIR34d if the individual has a GIR score of 3–4. Covered entirely by the social security system, the fixed nursing cost is not taken into account.

53The total in-home care cost is determined via the median monthly number of hours of care provided for each GIR score as estimated by Soullier (2012) [22] and calculated as the average of two costs: the theoretical hourly rate (ptheor) set by the department [23] and the hourly rate paid by private employers (P_PEMd). [24] The cost of community nurses and SSIAD home nursing services are not included because they are mostly covered by the health insurance system.

54The funding of residential care is calculated as the mean amount of the residential care APA in the department (APA_Ed) if the person has a GIR score of 1 to 4, and is 0 otherwise. As very few people receive the ASH, and because we have no information on their income, this allowance is not included. The funding of in-home care is based on the mean amount of the in-home APA in the department if the person has a GIR score of 1–4 (APA_Dd), and is 0 otherwise.

55The out-of-pocket cost is calculated as the difference between the total cost and the amount of funding received for each type of care. Hence, for an individual s living in a department d, the out-of-pocket cost (RACesd) of residential care is as follows:

57For an individual s living in a department d (RACdsd), the out-of-pocket cost of in-home care is as follows: [25]

figure im15

Notes

  • [1]
    The terms nursing home and care home are used in this article to refer to residential care facilities for disabled older adults.
  • [2]
    If supply is inadequate in the department of residence, people rarely select a care home in a different department. Only 15% of care-home residents live outside their department of origin, and, in half of these cases, it is a neighbouring department (Ramos-Gorand, 2013).
  • [3]
    An informal support allowance also exists, but is rarely awarded, so we consider that only formal support is publicly financed.
  • [4]
    The copayment rate of the APA or the pension fund welfare allowance depend on the person’s income. The person can also claim a tax deduction if he/she employs a home help directly (deduction applicable only for persons whose income is in or above the first tax bracket).
  • [5]
    The APA is only awarded to persons who are disabled under the GIR definition.
  • [6]
    The copayment rate of the residential APA depends on the persons income. He/she can also apply for a housing allowance (such as APL or ASL) or social allowance (ASH) depending on his/her income.
  • [7]
    The disablement fees depend on the person’s degree of disability.
  • [8]
    The out-of-pocket cost is the difference between the total cost and the disability allowances received.
  • [9]
    All household members were surveyed, but only the responses of those aged over 60 were used for our study.
  • [10]
    The department with the fewest responding care facilities (and residents) was French Guiana. Data were obtained for 211 residents (of whom 182 have difficulty washing) living in five nursing homes out of seven within scope.
  • [11]
    The ratio of initial weight to final weight was bounded at 7 to avoid estimation bias.
  • [12]
    The ‘has a partner’ variable was missing for 25,364 individuals (6.3%). It was imputed based on sex, age, and department.
  • [13]
    The ‘difficulty washing’ variable was missing for 89,543 individuals (22.3%). It was imputed based on sex, age, department, and GIR.
  • [14]
    The ‘physical mobility limitations’ variable was missing for 89,380 individuals (22.2%). It was imputed based on sex, age, department, GIR, and difficulty washing.
  • [15]
    The ‘cognitive limitations’ variable was missing for 90,189 individuals (22.5%). It was imputed based on sex, age, department, GIR, difficulty washing, and physical limitations.
  • [16]
    The individuals’ degree of disability is slightly higher than before imputation, but by only 2–3 percentage points. The models before and after imputation yield similar results.
  • [17]
    To justify this choice, we drew on data from the 2015 old-age disabilities, care and resources survey (Capacités, aides et ressources des seniors) to run a multinomial regression of the estimated GIR using the same limitations variables as those of VQS to identify the highly discriminant variables. We then chose a systematic ranking system so that a similar method could be applied with the EHPA and VQS data, and because the prevalences obtained at national level were consistent with the actual counts.
  • [18]
    Non-cohabiting couples cannot be identified in VQS.
  • [19]
    This is the weighted proportion of individuals in the EHPA 2015 database.
  • [20]
    The first quartile is –0,218, the median 0.060, and the last quartile 0.268.
  • [21]
    This is the threshold value above which the differences between residential care and in-home care rankings are higher.
  • [22]
    NBAF12 = 65 hours per month if the individual has a GIR 1 or 2; NBAF3 = 42.5 hours per month if the individual is GIR 3; NBAF4 = 25 hours per month if the individual is GIR 4; NBAF12 = 12.5 hours per month if the individual is GIR 5 or 6.
  • [23]
    These are the hourly amounts charged by authorized carers whose rates are set by the department.
  • [24]
    The data concern private individuals who employ domestic helpers and pay them directly. The hourly rate is calculated as the total net payroll divided by the total declared volume of hours. This information is available at the regional level, distinguishing two possible exemptions: for APA beneficiaries or for persons aged over 70 (we average the two).
  • [25]
    Negative estimated values are truncated at 0.

Abstract

English

Population ageing raises the question of long-term care arrangements for disabled older adults, be it in their own homes or in a residential facility. Analysis of trade-offs in the long-term care market is especially interesting for France, where old-age policy is managed at the département level. This decentralized system may give rise to geographical inequality in access to care because not all departments are equally endowed with resources to offer the desired levels of care provision. This article analyses the determinants of residential care access by combining data from two surveys and from administrative sources on long-term care. To identify the constraints that weigh upon individual choices, multilevel models are used to examine how needs, income, and informal and formal care shape geographical inequalities in care arrangement. This study reveals substantial differences in the probability of living in a care home, due partly to geographical disparities in the availability of formal care, further accentuated by disparities in out-of-pocket costs.

  • institution
  • care home
  • disability
  • geographical inequalities
  • multilevel
  • older adults
  • France
Français

Vivre en établissement pour personnes âgées dépendantes ou rester à domicile : le rôle du contexte territorial

Le vieillissement de la population soulève la question de l’organisation des soins de long terme pour les personnes âgées dépendantes et du lieu de leur prise en charge : à domicile ou dans un établissement. La France a une politique gérontologique décentralisée au niveau départemental ce qui questionne l’équité territoriale de l’accès à la prise en charge de la perte d’autonomie, car chaque département ne dispose pas des mêmes ressources pour mettre en place les politiques qu’il souhaite. Cet article propose un éclairage sur les déterminants de la prise en charge en établissement en combinant deux enquêtes et des données administratives sur l’offre de services de soins de longue durée. Il estime grâce à des modèles multiniveaux le rôle des besoins, des ressources, de l’aide informelle et formelle dans les inégalités territoriales de recours aux établissements, afin d’identifier les contraintes pesant sur les choix des individus. L’analyse montre de fortes différences quant à la probabilité de vivre en établissement en partie dues aux disparités de disponibilité géographique de l’offre formelle. Elles sont renforcées par les disparités de coût financier à la charge des personnes.

Español

Vivir en establecimiento para personas mayores dependientes o quedarse en su domicilio: el papel del contexto territorial

El envejecimiento de la población plantea la cuestión de la organización de los cuidados a largo plazo para las personas mayores dependientes así como la del lugar de su atención: a domicilio o en establecimiento. Francia cuenta con una política gerontológica descentralizada a nivel departamental que cuestiona la equidad territorial del acceso a los servicios que exige la pérdida de autonomía, ya que cada departamento no dispone de los mismos recursos para implementar las políticas que él desea. Este artículo propone informar sobre los factores determinantes de los servicios institucionales, combinando dos encuestas y datos administrativos sobre la prestación de servicios de atención a largo plazo. Evalúa a través de modelos multinivel el papel que juegan las necesidades, los recursos, la ayuda informal y formal en las desigualdades territoriales del acceso a los establecimientos, con el fin de identificar los límites que pesan sobre las opciones de los individuos. El análisis revela fuertes diferencias en la probabilidad de vivir en establecimiento, debidas en parte a la disparidad geográfica en las disponibilidades de la oferta formal. Estas diferencias son agravadas por las disparidades en el costo financiero que queda a cargo de la persona.

Outline

  1. I - Theoretical framework and literature review
    1. 1 - Theoretical framework
    2. 2 - Literature review
  3. II - Method and data
    1. 1 - Method
    2. 2 - Data
    3. 3 - The indicators
  5. III - Results
    1. 1 - Abundant department-level supply and heterogeneous out-of-pocket costs
    2. 2 - Descriptive analysis of the disabled population by care arrangement
    3. 3 - Determinants of residential care
      1. The type of care chosen is the one for which out-of-pocket costs are the lowest
      2. The role of formal care availability becomes negligible for an equivalent out-of-pocket cost
      3. Having a partner protects against entering residential care
      4. Demand for residential care is held back by income differences
      5. More frequent residential care for older adults with severe functional limitations, but also for those with the simplest care needs
      6. Sex and age seem to explain differences in residential care preferences
      7. A strong geographical pattern of residential care use
  7. V - Limitations and conclusion

References

References

  • Arnault L., 2015, La prise en charge des personnes âgées dépendantes: analyse microéconométrique de l’aide familiale (Doctoral dissertation), Université Paris-Dauphine.
  • Billaud S., Gramain A., 2006, Les déterminants de l’entrée en institution des personnes âgées, Actualités et dossiers en santé publique, 56, 43–44.
  • Billaud S., Bourreau-Dubois C., Gramain A., Lim H., Weber F., Xing J., 2013, La prise en charge de la dépendance des personnes âgées: les dimensions territoriales de l’action publique, Paris, MiRe/DREES.
  • OnlineBörsch-Supan A., Hajivassiliou V., Kotlikoff L., Morris J., 1990, Health, children, and elderly living arrangements: A multiperiod-multinomial probit model with unobserved heterogeneity and autocorrelated errors (Working Paper w3343), Cambridge, MA, National Bureau of Economic Research.
  • Bourreau-Dubois C., Gramain A., Lim H., Xing J., 2014, Impact du reste à charge sur le volume d’heures d’aide à domicile utilisé par les bénéficiaires de l’APA, Centre d’économie de la Sorbonne (Working Paper, halshs-00973938), Paris, Maison des Sciences Économiques.
  • Bouvier G., Lincot L., Rebiscoul C., 2011, Vivre à domicile ou en institution: effets d’âge, de santé mais aussi d’entourage, in France, portrait social 2011, 125–134.
  • BruneL M., Carrère A., 2019, La perte d’autonomie des personnes âgées à domicile. Quelles disparités entre départements? Les Dossiers de la Drees, 34, 1–43.
  • OnlineCambois E., Robine J.-M., Romieu I., 2005, The influence of functional limitations and various demographic factors on self-reported activity restriction at older ages, Disability and Rehabilitation, 27(15), 871–883.
  • OnlineCharles K., Sevak P., 2005, Can family caregiving substitute for nursing home care? Journal of Health Economics, 24(6), 1174–1190.
  • OnlineCheckovich T. J., Stern S., 2002, Shared caregiving responsibilities of adult siblings with elderly parents, Journal of Human Resources 37(3), 441–478.
  • OnlineDavey A., Johansson L., Malmberg B., Sundström G., 2006, Unequal but equitable: An analysis of variations in old-age care in Sweden, European Journal of Ageing, 3(1), 34–40.
  • OnlineDiepstraten M., Douven R., Wouterse B., 2020, Can your house keep you out of a nursing home? Health Economics, 29(5), 540–553.
  • OnlineEttner S., 1994, The effect of the Medicaid home care benefit on long-term care choices of the elderly, Economic Inquiry, 32(1), 103–127.
  • OnlineFontaine R., Gramain A., Wittwer J., 2007, Les configurations d’aide familiales mobilisées autour des personnes âgées dépendantes en Europe, Économie et Statistique, 403–404, 97–115.
  • OnlineFreedman V., 1996, Family structure and the risk of nursing home admission, The Journal of Gerontology Series B: Psychological Sciences and Social Sciences, 51(2), S61–S69.
  • Gaugler J., Duval S., Anderson K., 2007, Predicting nursing home admission in the US: A meta analysis, BMC Geriatrics, 7(13).
  • OnlineGarber A. M., MaCurdy T., 1990, Predicting nursing home utilization among the high-risk elderly, in Wise D. (ed.), Issues in the Economics of Aging, Chicago, University of Chicago Press, 173–204.
  • OnlineGaymu J., Delbès C., Springer S., Binet A., Désesquelles A., Kaligirou S., Ziegler U., 2006, Determinants of the living arrangements of older people in Europe, European Journal of Population, 22(3), 241–262.
  • Hégé R., Roquebert Q., Tenand M., Gramain A., 2014, La tarification des services à domicile: un outil au service des politiques départementales? Notes MODAPA, 2, December.
  • Hégé R., 2016, La demande d’aide à domicile est-elle sensible au reste-à-charge: une analyse multi-niveaux sur données française (CES Working Paper), Paris, Centre d’économie de la Sorbonne.
  • OnlineHoerger T. J., Picone G. A., Sloan F. A., 1996, Public subsidies, private provision of care and living arrangements of the elderly, The Review of Economics and Statistics, 78(3), 428–440.
  • OnlineJette A., Tennstedt S., Crawford S., 1995, How does formal and informal community care affect nursing home use? The Journal of Gerontology Series B: Psychological Sciences and Social Sciences, 50(1), S4–S12.
  • OnlineKnoef M., Kooreman P., 2011, The effects of cooperation: A structural model of siblings’ caregiving interactions (IZA Discussion Paper No. 5733), Bonn, Institute for the Study of Labor.
  • OnlineLaferrère A., Van den Heede A., Van den Bosch K., Geerts J., 2013, Entry into institutional care: Predictors and alternatives, in Börsch-Supan A., Brandt M., Litwin H., Weber G. (eds.), Active ageing and solidarity between generations in Europe: First results from SHARE after the economic crisis, Berlin, Boston, De Gruyter, 253–264.
  • Le Bihan B., Martin C., 2018, French social and long-term care system, Global Social Security Review, 7, 5–15.
  • Luppa M., Luck T., Weyerer S., König H.-H., Brähler E., Riedel-Heller S., 2010, Prediction of institutionalization in the elderly: A systematic review, Age and Ageing, 39(1), 31–38.
  • OnlineLockwood L. M., 2018, Incidental bequests and the choice to self-insure late-life risks, American Economic Review, 108(9), 2513–2550.
  • Metzger M. H., Barberger-Gateau P., Dartigues J. F., Letenneur L., Commenges D., 1997, Facteurs prédictifs d’entrée en institution dans le cadre du plan gérontologique du département de Gironde, Revue d’épidémiologie et de santé publique, 45(3), 203–213.
  • OnlineMönkediek B., Bras H., 2014, Strong and weak family ties revisited: Reconsidering European family structures from a network perspective, The History of the Family, 19(2), 235–259.
  • Muller M., 2018, L’enquête EHPA 2015. Les différentes phases de l’enquête (Working Paper Series ‘Sources et méthodes’ No. 67), Paris, DREES.
  • OnlineMuramatsu N., Yin H., Campbell R., Hoyem R., Jacob M., Ross C., 2007, Risk of nursing home admission among older Americans: Does states’ spending on home and community-based services matter? The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 62(3), S169–S178.
  • OnlineNihtilä E. K., Martikainen P. T., Koskinen S. V., Reunanen A. R., Noro A. M., Häkkinen U. T., 2008, Chronic conditions and the risk of long-term institutionalization among older people, European Journal of Public Health, 18(1), 77–84.
  • Nirello L., 2015, La construction problématique de la relation d’emploi dans l’ESS. Les EHPAD: entre régulations publiques et régulations d’entreprise (Doctoral dissertation), Université Nantes Angers Le Mans.
  • OnlinePeyrache M., Ogg J., 2017, Concilier vie professionnelle et aide informelle à un parent âgé, Revue européenne des sciences sociales, 55(1), 97–125.
  • Ramos-Gorand M., 2013, Accessibilité de l’offre en établissements d’hébergement pour personnes âgées: enjeux territoriaux, Dossier Solidarité Santé, 36.
  • Ramos-Gorand M., 2020, Accompagnement professionnel de la dépendance des personnes âgées, Dossier de la Drees, Série Méthodologie, 51.
  • Reschovsky J. D., 1996, Demand for and access to institutional long-term care: The role of Medicaid in nursing home markets, Inquiry, 33(1), 15–29.
  • Reschovsky J. D., 1998, The roles of Medicaid and economic factors in the demand for nursing home care, Health Services Research, 33(4 Part 1), 787–813.
  • OnlineRoquebert Q., Tenand M., 2017, Pay less, consume more? The price elasticity of home care for the disabled elderly in France, Health Economics, 26(9), 1162–1174.
  • Soullier N., 2012, L’aide humaine auprès des adultes à domicile: l’implication des proches et des professionnels, Études et résultats, 827, December.
  • OnlineSundström G., Malmberg B., Johansson L., 2006, Balancing family and state care: Neither, either or both? The case of Sweden, Ageing & Society, 26(5), 767–782.
  • OnlineTheisen T., 2017, What makes people nursing home residents: Individual need or municipalities’ supply? In Schofield N., Caballero G. (eds.), State, institutions and democracy: Contributions of political economy, Springer, 251–270.
  • Tomassini C., Glaser K., Wolf D., Broese van Groenou M., Grundy E., 2004, Living arrangements among older people: An overview of trends in Europe and the U.S.A., Population Trends, 115(Spring), 24–34.
  • OnlineTrabut L., Gaymu J., 2016, Living alone or with family beyond age 85: Major differences across French départements, Population & Societies, 539, December.
  • OnlineTrabut L, Garabige A, Gaymu J., 2021, Family care for the dependent elderly: How does it fit in with professional care at the French sub-national level?, Revue européenne des sciences sociales/ European Journal of Social Sciences, 59(2).
  • OnlineVan Houtven C., Norton E., 2004, Informal care and health care use of older adults, Journal of Health Economics, 23(6), 1159–1180.

Authors

Amélie Carrère
Institut des politiques publiques (IPP), Institut national d’études démographiques (INED).
amelie.carrere@ipp.eu
Translated by
Catriona Dutreuilh
This is the latest publication of the author on cairn.
Latest publication on cairn or another partner portal
Abstract Full-textfile_download Download
Uploaded on Cairn-int.info on 16/11/2021
Cite
Distribution électronique Cairn.info pour I.N.E.D © I.N.E.D. Tous droits réservés pour tous pays. Il est interdit, sauf accord préalable et écrit de l’éditeur, de reproduire (notamment par photocopie) partiellement ou totalement le présent article, de le stocker dans une banque de données ou de le communiquer au public sous quelque forme et de quelque manière que ce soit.
keyboard_arrow_up
Chargement
Loading... Please wait