Informal help in a local setting: The Dutch Social Support Act in practice

The Dutch Social Support Act provides municipal social support for people that experience limitations in daily life and cannot rely on informal help from their social network to compensate these limitations. In this paper, we study whether the probability and intensity (number of services) of receiving municipal social support for daily limitations is affected by informal help by the social network (i.e. family, friends and neighbors). This study took place in Breda, a middle large city in the South of the Netherlands. We combined data from the Municipal Personal Records Database, the registration containing information on demographics and municipal social support receipt of all inhabitants, with data from the Municipality Policy Monitor, a survey containing information on daily limitations and informal help (n = 5256). We find that people experiencing daily limitations are more likely to receive municipal social support Informal help Social network and also receive a higher number of support services (intensity). However, the perceived help from family and friends does not decrease either the probability or intensity of receiving municipal social support. Informal help from neighbors decreased the likelihood of receiving of municipal social support, but not the intensity. This implies that the overall relation between daily limitations and municipal social support is not different for people who indicate that they can or cannot rely on their informal network for help. © 2020 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license


Introductions
Like many OECD countries, the Netherlands witnessed a profound shift towards the decentralization of social policy in recent years. Decentralization refers to the shift of authority and control from the central government to the local level. More autonomy is provided to the local government to achieve higher quality and efficiency in services in better response to the local context [1]. In this context, the Social Support Act (in Dutch: Wet Maatschappelijke Ondersteuning | henceforward referred to as WMO), first introduced in the Netherlands in 2007, made municipalities responsible for providing care and services to support community dwelling and social participation [2,3]. To achieve this, municipalities can provide citizens social support at home, including personal (professional) homecare, assistance with daily household activities (like housecleaning, meal service), adaptations to the house (like step-in bathrooms and stair lifts), mobility aids (wheelchair, mobility scooter or sports equipment) and additional means of transportation (taxi service). As of 2015, the WMO also provides adult day-care services, accompaniment for daily activities (e.g. help with administration, going on outdoor visits, leisure activities) and sheltered housing for psychiatric patients, victims of domestic violence and the homeless.
To access municipal social support, the municipality not only assesses the individuals' abilities to live independently and participate in society, but another central premise of the WMO is to first rely on people's own responsibility to take care of themselves and others by relying on help from their informal social network [4]. Only when such help is not available or insufficient, citizens can turn to the municipality for support under the WMO-law. Though the WMO-law forms the basis for municipal social support policies in all municipalities, the decentralization allows each municipality to decide themselves on the implementation of the WMO, including the exact types and extent of the assistance and on the need assessment procedure [5] (a detailed description of the assessment process follows below). Because of this policy discretion, there is still little insight into the outcomes of the WMO, despite evaluations [3,4,6,7]. Currently, little evidence exists linking daily limitations and informal help to the actual take-up of municipal social support, despite strong policy assumptions that this relation exists [4]. Given that municipalities have adopted the tasks to assess and provide social support from the central government fairly recently (since 2015), we first assess whether a clear relationship between daily limitations and municipal social support exists, followed by the role of the presence or absence of informal help according to (potential) recipients own perspective on this relation. Therefore, our first research question is whether a direct relation exists between the perceived extent of limitations in daily living and receiving municipal social support. The second question addresses whether this relationship is affected by informal help by the social network (i.e. family, friends and neighbors). We examine the probability of receiving municipal social support and the intensity (number of services) by using a combination of two individual-level datasets from the municipality of Breda, a middle large city (population of 183,438 in 2018) in the South of the Netherlands.

The assessment process of municipal social support
The theoretical elaboration of this study relies on understanding the specificity of the WMO, i.e., supporting community dwelling and (social) participation by providing tailored packages of municipal support services [2]. The first step in the assessment of municipal social support is to determine whether applicants qualify for assistance i.e., the experienced daily limitations that relate to the WMO aims (living independently and social participation). Furthermore, the presence of a social network (e.g. family, friends and neighbors) and other (self)help options that can potentially meet the applicants' needs are discussed. The WMO-law does not specify exactly what those other options are or how they should be taken into consideration. Examples range from the ability to 'make own decisions' to more practical solutions (e.g. buying equipment or meal, laundry or ironing services). Even though at the time of the study recipients of municipal social support had to pay a contribution dependent on their income and financial assets, prohibiting access based on the income or other financial assets is not allowed and means-testing is forbidden [5]. If informal help or other 'self-help' options are insufficient or if the applicant experiences complicated limitations, individuals will receive a personalized set of support services.
Specific for the municipality of Breda is that the external network organization 'Zorg voor elkaar Breda' initially assesses all applications for municipal social support, assessing not only the possibility for informal help, but also if volunteering work or generally available services can provide a solution [8]. Only if this is not the case, a decision will be made about turning to a personalized solution and involvement of the municipality. Because of policy discretion, it is not readily clear how this differs from other municipalities. Yet a large scale study shows that most municipalities have organized such a 'central point of access' and seek to increase the role of informal help during the assessment with citizens requesting support [6].
To sum up, daily limitations on the one hand and the availability of informal help on the other hand should predict whether the applicant will receive municipal social support as well as the intensity of the support. The more severe the limitations of applicants are, and the more limited their social network is, the more likely they are to receive governmental support. In turn, individuals facing more or less the same limitations may sometimes be provided with different types of municipal social support and at different intensities, depending on the help from their social network and the available (self)help options.
In this study, we equate daily limitations with self-reported limitations that impair daily living. This approach relates to the concepts and measurements of the Activities of Daily Living (ADL) and Instrumental Activities of Daily Living (IADL) scales [9][10][11]. ADL and IADL cover daily tasks that are essential to independent living, referring to physical autonomy and cognitive function (e.g., personal hygiene, dressing, mobility, ability to manage finances, etc.) and are thus related to the WMO aims (i.e. supporting community dwelling and social participation) [9,10,12]. Subsequently, we expect that higher levels of daily limitations will increase the probability of receiving municipal social support (Hypothesis 1a) and also to receive more types (i.e. higher intensity) of municipal social support (Hypothesis 1b).
Turning to informal help by the social network, previous studies found that socially isolated individuals are more likely to seek health care for two main reasons: socially isolated individuals are less healthy compared their less isolated counterparts and they perceive health care use as a way to maintain their need for social interaction [13,14]. We thus expect that individuals that cannot rely on informal help from family, friends or neighbors have a higher probability of receiving municipal social support (Hypothesis 2a) and also to receive more types (i.e. higher intensity) of municipal social support (Hypothesis 2b).
However, in the assessment process of the WMO these two factors are not taken into account separately but in relation to each other. Informal help is considered a source for providing help with daily limitations. Individuals can mobilize either their close ties (i.e., family and friends) but also the weaker ties (i.e., neighbors) [15,16]. Although these strong and weak ties may take different roles in caretaking (individuals might be reluctant to ask neighbors or not very close friends to assist with personal needs such as hygiene or dressing), overall their presence indicates the ability of individuals to address the needs arisen by daily limitations experienced without, or with less, municipal social support.
In this study, we rely on the perceived availability of informal help. We argue that there is a strong link between the perceived and actual availability of and request for informal help. Research suggests that implicit considerations of solidarity and emotional connection are considered before practical barriers such as time or proximity [15]. To illustrate this: if individuals feel they cannot rely on their family, friends or neighbors because of a weak social bond, they will not consider them to provide informal help in the first place. In addition, the assessor of the municipality cannot assess in a more formal way if and who in the social network can be asked for informal help. Relying on the perceived availability of informal help thus in our opinion resembles most how the assessment is done in the real-life practice of the WMO.
In general, municipal social support is provided when there is no or only a limited informal network available. This implies that individuals with more or less the same levels of daily limitations will have a lower probability of receiving municipal social support (Hypothesis 3a) and to receive less types (i.e. lower intensity) of municipal social support (Hypothesis 3b) when they can mobilize more support from their family, friends, and neighbors. Furthermore, given that most limitations that impair the ability of individuals to live independently refer to the personal sphere, a follow-up expectation is that the help from family and friends will have a stronger moderator role for the relationship between the daily limitations and the probability of receiving municipal social support / intensity of such support compared to help from neighbors (Hypothesis 3c).

Materials and methods
For this study, we combined two data sources from the municipality of Breda: the Municipal Personal Records Database  (Basisregistratie Personen -BRP) and the Municipality Policy Monitor (Gemeentebeleidsmonitor -GBM). The BRP contains demographic information, including date of birth, gender and marital status, as well as information about municipal social support for all inhabitants of Breda (N = 183.438 on January 1, 2018). This data was combined with the 2017 GBM data (n = 5.825), which is a representative survey among inhabitants of Breda aged 16 and older. Data was collected between September and November in 2017 via a web, paper and telephone survey with a net response rate of 26.5 percent. Covered topics were among others informal help, financial well-being and daily limitations. Although we realize there is a limited time gap between the two data sources, we note that municipal social support also covers short-term care (e.g. because of limitations as a result of an operation or fall). Additionally, we are unable to verify what the amount of short-term care and the length of informal care relationships is. We argue that this is not problematic, since even for individuals with long-term care informal help relationships appear to be very stable over time (on average they last 4-6 years, [17]). Moreover, in practice individuals approach the municipality with different levels of daily limitations and informal help. In this way, our design matches the real-life situation of the municipality. Respondents in the GBM that could not be linked to the BRP (n = 88), or were younger than 16 (n = 4) or had too many missing values on relevant variables (n = 477) were removed from the dataset, resulting in a final merged sample of 5.256 respondents. Table 1 shows which variables were retrieved from which dataset.

Dependent variable -municipal social support
The receipt of municipal social support is measured in two ways. The first variable measures whether the respondent receives municipal social support or not (yes = 1, no = 0). The second variable covers the intensity/the number of services of municipal social support: it is a count variable ranging between zero and five services, namely 1) domestic help 2) physical aids (e.g. wheelchair or stair lift), 3) taxi, 4) a (protected) living arrangement or 5) an accompaniment for daily activities (e.g. adult day-care, help with administration, going on outdoor visits).

Independent variables -daily limitations
For daily limitations, we used a scale closely related to the IADL and ADL scale [9,11]. Factor analysis was done on eight items measuring whether a respondent manages with: a) 'doing groceries', 'b) household chores', c) 'walking, sitting, standing', d) 'washing, getting dressed, etcetera', e) 'administration and finances', f) 'social contacts', g) 'your physical health' and h) 'your mental health'. An item about raising children was not included in the analysis because this does not match the WMO aims. Possible answers for all items were: 0) 'yes, (almost) no issues, 1) 'reasonable, some issues', 2) 'bad, many issues' and 3) 'not applicable' (counted as a missing value). The scale had a good internal reliability (Cronbach's alpha = 0.85). Like previous studies, sum scores of daily limitations are used [18][19][20]. We allow respondents to have two missing values, calculating the sum score based on a minimum of 6 out of 8 items.

Informal help by the social network
Respondents reported if they, according to their own perception, can rely on their family in case they would need help or care. The same question was asked for friends and neighbors. Possible answers were: a) 'no', b) 'probably yes', c) 'yes, for sure' and d) 'don't know / not applicable'. Options b and c (yes, for sure and probably) together form one category Yes (1) and option a (no) is the reference category (0). Option d) 'don't know/not applicable' is also included (separately) in the analysis, because the percentage of respondents with this answer ranged between 3.2 (n = 167 for family) and 17.6 percent (n = 927 for neighbors) between the informal network items and is relevant to take into account when analyzing the influence of the presence of informal help. To examine the link between self-reported informal help by family and marital status (i.e. having a partner), we performed a crosstabulation between the two variables. Married individuals reported slightly higher levels of perceived help compared to divorced and widowed (85.4 % versus 83.9 % and 82.5 % respectively) and slightly lower than unmarried (85.7 %), but the differences were not significant.

Control variables
In order to control for the options that people have to arrange and pay for their own services, the variable 'perceived financial well-being' is included by taking into account the question 'how do you get by?' Answer options are: 'very easily', 'easily' (taken together as: 'easy'), 'just getting by', 'difficult' or 'very difficult' (taken together as: difficult, ref.cat). The highest educational level completed is also included, ranging from no education to university education. Three dummies were created: low (no education, primary or lower secondary), middle (upper secondary and postsecondary non-tertiary) and high (tertiary education, ref. cat). Age, gender, marital status and migration background are also included as control variables. Age is a continuous variable, ranging from 16 to 80. Age squared is included to account for a potential curvilinear effect of age (i.e. the effect of age could be stronger for younger and older people than for those in between). For gender, men (0) are the reference category. Marital status consists of four categories: widowed (0, ref.cat), unmarried (including cohabiters and singleparent households) (1), married or registered partnership (2), and divorced (3). For migration background, native Dutch (1), western immigrants (2) and non-western immigrants (0, ref.cat) are distinguished.
Finally, we take into account dummies for the different neighborhoods in Breda (ref.cat city center), given that socioeconomically disadvantaged neighborhoods generally have fewer social resources, and therefore, potentially have lower levels of informal help available [21][22][23]. We have chosen for neighborhood dummies instead of a multilevel model because we do not focus on neighborhood differences in our research question. By using dummies, we control for contextual differences. During the sampling of the GBM, the municipality applied stratification at the neighborhood level, meaning that the municipality estimated the number of respondents that they needed from each neighborhood based on response rates of earlier surveys. Originally, the dataset had 52 neighborhoods, but due to a lower number of respondents in some neighborhoods, we merged a few neighborhoods, finally including 43 neighborhoods, ranging between 8 and 220 respondents.

Analytical strategy
To estimate the probability of receiving municipal social support, we use logistic regression. In the first model, we examine the influence of daily limitations on the likelihood of receiving municipal social support. In model 2, we study the relationship between informal help and receiving municipal social support. We end in the following models with an inspection of the interactions separately by type of perceived informal help and limitations in order to test whether individuals with similar limitations but different levels of perceived informal help have a different likelihood of receiving municipal social support.
For the intensity of municipal social support (i.e. the number of services) we applied negative binomial regression, a type of Poisson regression. Poisson regression is recommended for count variables with a skewed distribution, as is the case for our data: most respondents do not receive municipal social support and among those who do the majority receives one service. Using a negative binomial model, allows to take into account the full amount of services used (ranging 0-4) for the entire sample. The same setup is employed for the intensity of municipal social support (the number of services): first the direct effect of limitations on the number of services is examined (model 1), followed by the direct effects of perceived informal help (model 2). Finally, the interactions between limitations and informal help are estimated. A significance level of p < 0.05 was set for all analyses.
In our additional analyses, we removed the recipients with the protected living arrangement, because this is a relatively intense type of municipal social support compared to the other services (e.g. domestic help or physical aids). Therefore, we verified the robustness of the results after leaving out recipients of this type of municipal social support. In addition, because neighbors are more likely to provide 'lighter' informal help tasks, we tested if using a scale with only the 'lighter' limitations (e.g. doing groceries and household tasks) compared to the more personal tasks such as dressing, would provide different results. Finally, we examined more closely the role of the partner in informal help by assessing the differences in municipal social support among different marital statuses. Table 1 shows the descriptive statistics of the variables in this study. Regarding gender, the sample is comparable to the population of Breda, but our sample is slightly older (average age 49.5 versus 47.1 in the population) and has an underrepresentation of non-western immigrants (9.3 % versus 11.1 % in the population). In our sample, 6.9 % of the respondents receives municipal social support, which is lower than the percentage of recipients in the population (8.5 %) but comparable to the national number of 6.2 % [24]. Overall, our sample differs slightly from the population, but can be considered comparable.

Results
Considering daily limitations (top part of Table 2), we find that recipients of municipal social support experience more daily limitations (average daily limitations of recipients: 1.63 versus average daily limitations of non-recipients: 0.98, p < 0.05) Moving to logistic regression, we find in model 1 that more limitations result in a higher likelihood of receiving municipal social support (OR = 1.10, p < 0.05, Table 3). When comparing the -2ll of this model to the model with only the control variables, this model was a better fit to the data (-2ll diff. = 15.31, p < 0.05). This supports   Hypothesis 1a: higher levels of daily limitations are associated with a higher probability of receiving municipal social support.
In model 2, we test the direct effect of perceived informal help on the probability of receiving municipal social support. We find no significant effects for help from family and friends: those who report they can rely on their family or friends do not have a lower likelihood to receive municipal social support. However, individuals who perceive they can rely on help from their neighbors are less likely to receive municipal social support compared to those who cannot rely on such help (OR = 0.67, p < 0.05). This partially supports Hypothesis 2b: not being able to rely on neighbors increases the chance of receiving municipal social support.
Next, we turn to the interaction effects between daily limitations and the three types of informal help (family, friends and neighbors) on the probability of receiving municipal social support (see Table  5 in the Appendix A). None of these interactions yield significant effects, indicating that individuals with the same levels of limitations, but different levels of informal help have a similar likelihood to receive municipal social support. These results are not in line with Hypothesis 3a-c: the effect of daily limitations on receiving municipal social support is not different for people that can rely on different kinds of help from their social network.

Intensity of municipal social support (number of services)
The bottom part of Table 2 displays the average levels of daily limitations among groups with more or less municipal social support services: people receiving three municipal social support services report the highest level of limitations (1.78) (the n = 7 people receiving four services is too small for reliable conclusions).
In Table 4, we find like in the previous model, that daily limitations have a significant, positive effect on the intensity of municipal social support (using negative binomial regression). This supports Hypothesis 1b: individuals with higher levels of daily limitations receive more types of municipal social support. Furthermore, in Table 4 Negative binomial regression: influence of limitations and informal help by the social network on the intensity of municipal social support (n = 5.256). model 2 the effect of support from family, friends and neighbors is tested. Unlike the previous model, we do not find a significant effect of either of the dummies for informal help, thereby not confirming that help from the informal network is associated with less intense use of municipal social support (Hypothesis 2b).
Turning to the interaction effects between daily limitations and the types of informal help from the social network for the intensity of municipal social support, we find no significant interaction effect for any of the combinations of limitations and informal help. This indicates, contrary to Hypotheses 3b and 3c that people with similar levels of limitations, but a different informal network they can rely on do not differ in the number of municipal social support services they receive (see Table 6 in the Appendix A).

Additional analysis
In order to test the robustness of our results, we first eliminated individuals with a protected living arrangement (a relatively intensive type of municipal social support, n = 19) from the analysis and our results were unchanged. In addition, because neighbors might be more likely to help with lighter daily limitations (doing groceries and household chores) compared to more personal tasks (e.g. dressing or administration), we added an analysis using a scale with only these two lighter limitations. This did however not change the findings.
Finally, we looked further into the role of a partner for informal help by assessing the relationship between marital status and municipal social support and leaving out perceived family support. We found that when compared to widowed, married individuals have a significantly lower probability to receive municipal social support and to receive a higher amount of municipal social support services, but there is no significant difference with unmarried or divorced. In accordance with the literature, we found that a partner plays a role in providing support for daily limitations [25].

Discussion
This study is one of the first to examine the effects of the decentralization of social support from the national government to the municipalities by studying the relationship between daily limitations, informal help and recipiency of municipal social support in the Dutch municipality of Breda. Using a combination of administrative and survey data we found that daily limitations are related to a higher likelihood of receiving (more) municipal social support, a finding that supports our theoretical depiction of the WMO allocation process. Next, we found that being able to rely on help from neighbors decreases the likelihood of receiving municipal social support (but not the intensity), whereas help from family and friends did not have a significant direct effect on receiving such support. In addition, the subjective appraisal of informal help did not play a moderating role, i.e., recipients who reported to be able to rely more on informal help from various sources did not have a different likelihood of receiving (more) municipal social support.
Our findings are in line with (the limited number of) recent studies that found limited support for the idea that informal help alters municipal social support under the WMO-law. For instance, a previous study found no difference in the use of municipal household support between municipalities with and without a strong policy focus on informal help [25]. Furthermore, another study found no relationship between social capital and the provision of services and amenities in municipalities [21]. Reasons for this might be that individuals turn to the municipality after they have already exhausted their informal social network. Two third of the informal caregivers of applicants indicates that they cannot provide more care than they already do [4].
In this study, we rely on the perceived presence of informal help, because we argued that this is a strong indicator of actual levels of informal help available (e.g. through emotional connection and solidarity) [15] and reflects the real-life assessment process of the municipality. A limitation of this approach is the lack of more objective measures capturing who would be able to provide informal help and how much, especially since this could be related to WMOrecipiency. Future research along these lines is thus warranted.
In addition, we provide a broad overview into the general practice of the WMO-law by combining all municipal social support services, but due to the limited information on objective measures of informal help available and the limited sample size of respondents receiving municipal social support (n = 363), we cannot verify if the relationship between daily limitations and informal help is different between the different types of services. For example, some services might be more easily compensated by informal help than others (e.g. domestic help or taxi). Further research could give more insight into these specific relationships. Finally, though the WMO-law forms the basis for developing municipal social support policies in all municipalities (and the assessment is based on the same principles), due to the policy discretion municipalities in the Netherlands differ in their implementation of social support policies, which means that our findings cannot be readily generalized to other municipalities.

Conclusion
Our study focused on municipal social support in Breda, a city in the South of the Netherlands. We linked two datasets to analyze the relationship between daily limitations, informal help by the social network and receiving municipal social support under the WMOlaw. We conclude that the recipiency of municipal social support is only marginally influenced by individuals' perceived informal help. This fits in with the limited evidence in the literature. Reasons for this, as others pointed out, might be that individuals often apply for municipal social support after they have already exhausted their informal social network. This is in line with the idea that 'informal care is already provided at a large scale, and that the potential of 'unexplored' informal care might be overestimated' [5].

Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of Competing Interest
None.