Paid informal care within Swedish eldercare: prevalence, change and variation among Swedish municipalities 2006–2016

ABSTRACT

This study analysed the prevalence of, change and variation in paid informal care as a policy strategy within Swedish municipal eldercare (n = 165) in 2006 and 2016. The results showed that the share of municipalities with this policy has declined since 2006, with just over half of the municipalities offering paid informal care in 2016. Among these municipalities, the practical use of this policy was limited in comparison to ordinary home-care services. Efforts to explain why some municipalities were more likely to have a system of paid informal care than others, were only partially successful, only identifyinga few significant municipal factors that together explained a limited share of the variation among the municipalities. The results raise questions relating to the way these policies are understood/used by municipalities and how older people and their informal caregivers are affected by the way in which the question of paid informal care is handled at the local level.

ABSTRAKT

Den här studien analyserar förekomsten, förändringen och variationen av betald anhörigomsorg som en policy strategi inom kommunal äldreomsorg (n = 165) 2006 och 2016. Resultaten visade att andelen kommuner med denna typ av policy minskade efter 2006 och att betald anhörigomsorg endast erbjöds i lite mer än hälften av kommunerna 2016. Bland dessa kommuner var den praktiska användningen av betald anhörigvård oftast begränsad i förhållande till användningen av traditionella hemtjänst. Försök att identifiera vilka kommunala faktorer som gör kommuner mer benägna att att erbjuda denna typ av policy lyckades endast delvis och större delen av variansen mellan kommer förblev oförklarad. Sammantaget väcker resultaten flera frågor som relaterar till hur kommuner förstår och använder sig av betald anhörigvård, samt hur hjälpbehövande äldre och deras anhöriga påverkas av hur frågan om betald anhörigvård hanteras på den kommunala nivån.

KEYWORDS:

• Cash-for care
• paid informal care
• older people
• informal caregivers
• Sweden

NYCKELORD:

• Cash-for-care
• betald anhörigomsorg
• äldre personer
• anhöriga
• Sverige

Introduction

In Western countries, the question of how to meet the care needs of an aging population has attracted increased attention over the last couple of decades. Many countries have experienced similar demographic changes, with traditional sources of care having declined due to falling birth rates, changing family structures and an increase in women’s labour-market participation, while the care needs of the older population have grown due to an increase in longevity (Yeandle & Ungerson, 2007). Since the 1990s, using so called cash-for-care (i.e. cash schemes to pay for either formal or informal care) to meet increseing care needs has become one of the most popular policy strategies among Western countries (Da Roit & Le Bihan, 2010). This article focuses on cash-for-care schemes in Swedish municipalities, examining the prevalence of, and changes and variations in such schemes within publicly funded eldercare.

In Sweden, the organisation and provision of services for older people is the responsibility of the local authorities in 290 self-governing municipalities (Dunér et al., 2017). Although local authorities are required to offer home-care and institutional care, each municipality is free to decide whether to deploy cash-for-care schemes or not. In such schemes, municipalities tend to pay informal rather than formal caregivers. Therefore, the term paid informal care will be used henceforth. Here, informal care refers to the help and support provided by family members or others such as friends and neighbours. Paid informal care is defined as any situation in which such a person is reimbursed financially through public funds for home-care services they provide to an adult 65 years or older.

Sweden is an interesting case because although it faces similar challenges to many other Western countries, the use of paid informal care differs from the more general trend in cash-for-care in two ways. Firstly, such schemes are not a recent phenomenon; they wereintroduced in the 1950s by local authorities to tackle issues related to the expansion of the welfare state (Johansson & Sundström, 2006). Secondly, their popularity has steadily declined since the 1970s and1980s, with a decrease in both the number of municipalities with such a policy and the total number of older people receiving it (SCB, 1991; Government Bill, 1987/88:176; NBHW, 2001, 2007).

While cash-for-care schemes has attracted much attention in the international literature (for reviews see Arksey & Kemp, 2016), research on paid informal care in Sweden is very limited(see Mossberg Sand, 2000; Forsell, 2004; Forsell et al., 2014; Brodin, 2018). Moreover, the municipal use of paid informal care was excluded from the national statistics in 2007 and 2008. Thus, current knowledge about paid informal care is extremely limited.

Questions relating to the prevalence of and changes in paid informal care are important, as the overall organisation of publicly funded eldercare greatly affects the options that are made available to older people and their families in terms of both giving and receiving high quality care (Daly, 2002).

A related question is why these schemes exist in some municipalities and not in others. The question of local variation has attracted little attention in the research on cash-for-care. For example, in the UK where the implementation, uptake and levels of direct payments have varied locally, studies exploring local patterns have been limited (Fernández et al., 2007). However, previous research has acknowledged and addressed local variation in the provision of formal services to older adults. In this research, political, structural and financial factors have been found to influence municipal spending on or the distribution of services. However, these factors do not fully explain the variation that exists between municipalities (e.g. Trydegård & Thorslund, 2001; Jenson & Lolle, 2013), thus raising questions about geographical inequality, i.e. whether citizens’ access to services is determined by their place of residence.

Aim and research questions

The aim of this study is to analyse the prevalence of, and changes and variation in paid informal care as a policy strategy within Swedish municipal eldercare. By analysing data from 2006 and 2016, the following research questions are addressed:

• How common were paid informal care schemes among municipalities in 2016? To what extent were these schemes used in practice?

• How did the prevalence of paid informal care schemes in Swedish municipalities changed between 2006 and 2016?

• To what extent does the prevalence of paid informal care schemes vary in relation to structural, political and financial factors at the municipal level?

The Swedish context

Sweden is often described as a welfare state with universal and defamilising services (Ulmanen & Szebehely, 2015). Thus, state-provided services should be used by all socioeconomic groups and make citizens less financially and socially dependent on their families (Sipilä, 1997; Leitner, 2003; Rauch, 2007). This is important from the perspective of older people’s independence but also essential facilitating the Swedish dual caregiver/earner principle in which men and women are expected to share caregiving and breadwinning responsibilities (Ulmanen, 2015).

In Sweden, services provided to older people are governed at several levels. At the national level, policy goals are formulated and then implemented through legislation and financial incentives. At the local level, by law municipalities are responsible for the provision of publicly funded institutional care and home-care services and they have great autonomy in setting the local guidelines for local social services and deciding on the budget and how to organise the work (Johansson et al., 2011).

Formal home-care has traditionally been delivered by public providers. However, over time, the influence of neo-liberal ideas with an emphasis on market solutions has been one of the driving forces behind a number of reforms, making it easier for municipalities to implement consumer-choice models and thereby allowing users to choose between public and private providers (Moberg, 2017).

In contrast to formal services, there is no obligation to deploy paid informal care schemes and each of the 290 municipalities are free to make decisions on whether and how to do so. Although this means that there are differences in how these systems are set up and used, paid informal care mainly comes in two forms: as a non-taxable allowance or as a salary. There are three subgroups in the latter category. Traditionally, informal caregivers could receive a salary by being either employed or contracted by the municipality. With the introduction of consumer-choice, informal caregivers can now also be employed by private providers, if permitted by the local authority.

Paid informal care has existed in Swedish municipalities since the 1950s. At its peak in 1973, more than 23,000 people received care from an informal caregiver with a salary (SCB, 1991), and by the mid-1980s approximately 21,000 people received an allowance (Government Bill, 1987/88:176). By the beginning of the 2000s, both forms of paid informal care had become less common and continued to decline. In 2006, allowances were deployed in less than half of the municipalities and provided to just over 5000 older people. The use of salaries was reported in about two-thirds of municipalities but covered fewer than 2000 older people (NBHW, 2001, 2007).

This decline indicates that paid informal care schemes are not seen as particularly legitimate by muncipalities. While it is not clear why municipalities have become more reluctant to use paid informal care, Johansson and Sundström (2006) note that such policies go against the Swedish welfare state’s core principle of providing services rather than cash.

Previous studies suggest that municipalities may view paid informal care as an outdated way of providing care services (Mossberg Sand, 2000), or as something that gives rise to certain problems (Forsell et al., 2014; Brodin, 2018).

However, research has also shown that in some situations paid informal care is viewed as a suitable solution for dealing with certain local issues. Specifically, in some municipalities it functions as a way of offering ethnically adapted services or services in a foreign language (Forsell et al., 2014; SALAR 2004). Motivated by financial reasons, it has also been offered in order to allow older people to remain in their own homes or to deliver services in sparsely populated areas (Mossberg Sand, 2000). Past research also suggests that it is sometimes used as income strategy by foreign-born carers with a weak labour-market attachment (Forsell et al., 2014; Brodin, 2018).

Since the 1990s, Swedish eldercare has undergone changes that in other Western countries have motivated the introduction of cash-for-care schemes (Da Roit & Le Bihan, 2010). In essence, the extent of informal care and carers’ needs for support has attracted increased attention in eldercare politics, resulting in legislative changes that obligate municipalities to provide various support services (Johansson et al., 2011). Since cash schemes have been described in national policy as a form of financial support, it is possible that this has made municipalities more inclined to implement paid informal care. Another trend is the policy shift towards providing older people with more choice and control over their care arrangements (Dunér et al., 2017). Internationally, cash-for-care has functioned as a way of achieving such personalised care (Netten et al., 2012) and it is therefore possible that Swedish municipalities have also become more inclined to use paid informal care as a strategy to achieve this.

Material and method

Data

This study uses primary and secondary data. Primary data were collected through an online survey sent out to all Swedish municipalities (N = 290) in October 2016. The survey contained many questions, but for this study data relating to de prevalence and use of paid informal care in 2016 and the prevalence of consumer-choice systems was collected. One hundred and sixty-five municipalities responded to the survey, representing a response rate of approximately 57 per cent. Secondary data relating to municipal factors and the municipalities’ previous use of paid informal care schemes in 2006 were retrieved from official databases and reports (for specific references see Table 1).

Table 1. Characteristics of the study sample in 2016.

	n*	%	Mean (SD)
Structural variables:
Population 80+ years^a	165		5.9 (1.3)
Foreign-born population (excl. EU/EFTA)^b	165		8.7 (4.2)
Unemployed 16–64 years^b	165		7.5 (3.0)
Foreign-born population 80+ years^c
Low proportion (0–7%)	71	43.0	5.9 (1.3)
Medium proportion (7–12%)	57	34.5
High proportion (>12%)	37	22.4
Foreign-born unemployed (16–64 years)^d	165		27.5 (12.6)
Foreign-born working in municipal social care services^b
Underrepresented (0–0.8)	69	42.1	0.9 (0.4)
Represented (0.81–1.20)	69	42.1
Overrepresented (>1.21)	26	15.9
Municipality type^e
Rural	71	43.0
Other	94	57.0
Political variables:
Political majority in municipal council (2006–2018)^f
Mainly left-wing majority	53	32.1
Mainly mixed majority	37	22.4
Mainly right-wing majority	75	45.5
Use of consumer-choice system^g
With	71	43.3
Without	93	56.7
Female politicians on municipal council^a	165		43.3 (5.1)
Left-wing politicians on municipal council^a	165		45.6 (10.2)
Politicians 65+ years on municipal council^a	165		19.9 (7.1)
Foreign-born politicians on municipal council^a	164		7.1 (5.1)
Previous use of paid informal care in 2006^h	164
With paid informal care	135	82.3
Without paid informal care	29	17.7
Financial variables:
Average profit/loss (2006–2016)^a
Negative-low positive result	50	30.3	1005 (958)
Positive result	84	50.9
Strong positive result	31	18.8
Share of expenditure for services for people 65+ years^b	165		27.3 (5.2)
Home-care in ordinary housing for persons 65+ years^b	164		7.5 (1.9)
Residential care for persons 65+ years^b	164		3.9 (1.1)
Average hours of home-care per recipient^b	161		31.3 (10.1)

Sources: ^aSCB, 2016, ^bKOLADA, 2016, ^cSCB, 2016, ^dSwedish Public Employment Service, 2016, ^eSAERG, 2016, ^fSALAR, 2015, ^gcalculation based on primary data, ^hNBHW, 2007. Comments: *Numbers may not add up to 165 due to missing values.

Analysis

Data were analysed using IBM’s software SPSS. While the first research question was answered using only primary data, primary and secondary data were analysed to answer the second research question. Specifically, the results from the municipalities participating in the survey were compared with the data on the same municipalities’ use of paid informal care in the form of allowances and salaries gathered by the NBHW in 2006. To answer the third research question, the association between the municipal use of paid informal care as a policy strategy and different municipal factors was investigated through statistical interference methods. For each analysis, paid informal care (defined as municipalities offering allowances and/or salaries) was treated as the dependent variable, while the different municipal factors were treated as independent variables. First, bivariate analyses were conducted, using Pearson’s Chi-Squared test or Student’s T-test (Bryman & Cramer, 2012). Next, and in order to controlfor type I and type II errors, multivariate logistic regression analysis was carried out (Pallant, 2005). Here, the association between the dependent and independent variables was tested further while controlling for the political majority in the municipal council between 2006 and 2018; the proportion of people 80 years or older; the proportion of foreign-born people (excl. EU/EFTA); and the proportion of unemployed people aged 16–64 years. In all, three models were used. The first model included structural and control variables. In the second model, the political variables were added. The final and third model included all three categories of variables – structural, political and financial – as well as the control variables.

Selection of independent variables

It was anticipated that the municipal population structure would be associated with the municipal use of paid informal care schemes. Based on previous research (Forsell et al., 2014; SALAR 2004), it was assumed that paid informal care would be more common in municipalities with a large population of foreign-born older people and in those with fewer of foreign-born people working within municipal social care services. These two characteristics were operationalised using data measuring the percentage of foreign-born people 80 years or older and data measuring the ratio between the proportion of foreign-born people (18–64 years) employed within municipal social care services and the proportion of foreign-born people (18–64 years) living in the municipality as a whole. It was also assumed that municipalities with a higher proportion of foreign-born people among their unemployed would be more likely to use paid informal care(Forsell et al., 2014; Brodin, 2018). Data measuring the proportion of foreign-born people among the unemployed aged 16–64 years were used. Population density was thought to influence the prevalence of paid informal care (Mossberg Sand, 2000). This factor was operationalised using a categorisation of municipality types based on information about the municipaly’s population size and density. As cash-for-care schemes in other countries have coincided with a political ambition to provide more personalised care, the presence of consumer-choice systems was seen as an indicator of this, as municipalities have implemented such systems to achieve more personalised care (Dunér et al., 2017).

To a higher extent than men and right-wing voters, women and left-wing voters support the notion that informal caregivers should be able to get paid (Mellgren, 2017). Consequently, the composition of the politicians on the municipal council responsible for outlining local guidelines was assumed to influence willingness to use paid informal care. Specifically, a higher percentage of female, left-wing, foreign-born and older politicians on the municipal council was thought to increase the likelihood of the municipality having paid informal care schemes. Data on the composition of politicians on the municipal council after the 2014 election was used.

Previous research has shown that municipal policy is greatly influenced by local tradition and historical continuity, i.e. it is path-dependent (Trydegård & Thorslund, 2001; Jenson & Lolle, 2013). Municipalities with a history of paid informal care were therefore assumed to be more likely to continue their use of such policies. Here, information on the existence of paid informal care schemes in 2006 was used.

The implementation of cash-for-care in other countries has been driven by financial reasons (Da Roit & Le Bihan, 2010). It was therefore hypothesised that municipalities with a more strained financial situation and lower ambition for or prioritisation of eldercare would be more inclined to use paid informal care. The financial situation fo each municipality was estimated by calculating a mean score of their annual net profit and loss accounts (before extraordinary costs) between 2006 and 2016 measured in SEK/person. An index was created to capture the municipal prioritisation of/ambition for eldercare (Pallant, 2005). Concretely, four indicators were used: the municipal care services’ share of expenditure in relation to the total expenditure for key welfare areas; the coverage rate for home-care in ordinary housing; the coverage rate for residential-care; and the average number of home-care hours/person. Here, a higher value represents a higher municipal prioritisation of and ambition for care services.

Results

Paid informal care in 2016

As shown in Table 2, just over half of the municipalities reported the use of paid informal care schemes in their eldercare services in 2016. Salaries were used in just over 40 per cent of municipalities, while allowances were used in less than 25 per cent of municipalities. For municipalities offering salaries, employment by the municipality was the most common form, and was used in 85 per cent of municipalities. In contrast, employment by private providers was only permitted in 25 per cent of the municipalities and just over 10 per cent of municipalities contracted informal caregivers.

Table 2. Prevalence of paid informal care schemes among municipalities (n = 165) in 2016.

	n	%
Paid informal care	90	54.5
Care allowance	39	23.6
Salary:	68	41.2
Employment by municipalities	58	85
Employment by private providers	17	25
Contracting by municipalities	8	12

Sources: Calculations based on primary data.

Looking at the practical use of paid informal care (Table 3), the results showed that the number and proportion of older people with a paid informal caregiver varied considerably among the municipalities. Despite these variations, most municipalities tended to use these schemes moderately in relation to ordinary home-care services. In practice, of the two forms of paid informal care, allowances were used more extensively, with a significantly higher number and proportions of older persons in municipalities with allowances than those with salaries.

Table 3. Number and proportion of older people in municipalities with a care allowance (n = 37) and salary (n = 49) in 2016.

	Number of older people receiving paid informal care		Percentage of older people with home-care receiving paid informal care
	Mean	Min – Max	Mean	Min – Max
Care allowance	41.3	1–200	6.6	0.7–40.1
Salary	8.7	0–128	1.9	0–14.6

Sources: Calculations based on primary data. Comments: The information on the number of older people with paid informal care in the form of a salary does not include municipalities that permitted the employment of informal caregivers by private providers. Only 5.1 per cent of municipalities that used the care allowance option did not provide information on the number of older people with a care allowance. As for salaries, the non-response rate was 26 per cent.

Changes in the prevalence of paid informal care schemes

In comparing the prevalance of paid informal care schemes in 2006 and 2016, the general development can be described as a downward trend. As illustrated in Table 4, more than 80 per cent of municipalities in 2006 used some kind of paid informal care scheme. In 2016, this number had decreased to just over half of the municipalities. The proportion of municipalities that used both allowances and salaries also decreased during this period.

Table 4. Changes in the prevalence of paid informal care schemes among municipalities between 2006 and 2016.

	2006		2016
	n = 164	%	n = 162	%
Paid informal care:
With both care allowance and salary	50	30.5	17	10.5
With either care allowance or salary	85	51.8	71	43.8
Without any form of paid informal care	29	17.7	74	45.7
Care allowance	81	50.0	39	23.9
Salary	104	63.8	68	41.5

Sources: Calculations based on the NBHW (2007), Table 5 and primary data.

The development at the local level follows similar patterns. As shown in Table 5, it was significantly more common for municipalities that had previously used care allowances or salaries to report their use in 2016, compared with municipalities that did not use these in 2006. However, about half of the municipalities that provided either allowances or salaries in 2006 had discontinued their use in 2016.

Table 5. Changes in the prevalence of care allowances (n = 162) and salaries (n = 163) at the local level between 2006 and 2016.

	Care allowance		Salary
	% with in 2016	% without in 2016	% with in 2016	% without in 2016
% with in 2006	22.8	27.2	32.5	31.3
% without in 2006	0.6	49.4	8.6	27.6

Sources: Calculations based on the National Board of Health and Welfare (2007), Table 5 and primary data.

Municipal factors and variation in of paid informal care schemes in 2016

Finally, this study seeks to provide insights into whether the prevalence of paid informal care schemes varies in relation to different municipal factors. From the analysis, the third multivariate regression model had the highest explanatory value (Table 6, Goodness of fit).

Table 6. Test of association between municipal factors and paid informal care schemes in 2016.

	Bivariate analysis		Multivariate analysis
	% with paid informal care	p-value	Model 1 (n = 164)		Model 2 (n = 162)		Model 3 (n = 159)
Independent variables:			OR	p-value	OR	p-value	OR	p-value
Structural variables:
Population 80+ years	5.86	0.394	1.162	0.444	1.454	0.088	1.283	0.360
Foreign-born population (excl. EU/EFTA)	8.84	0.646	1.141	0.048	1.080	0.261	1.081	0.263
Unemployed 16–64 years	7.06	0.043	0.851	0.268	0.853	0.304	0.817	0.211
Foreign-born 80+ years
Low prop.	58.6		1.584	0.291	1.425	0.448	1.296	0.587
Medium prop.	50.9	0.682	1.000		1.000		1.000
High prop.	54.1		1.039	0.942	1.530	0.483	1.505	0.536
Foreign-born unemployed 16–64 years	26.1	0.112	1.034	0.285	1.025	0.481	1.042	0.276
Foreign-born working in municipal social care services
Underrepresented	42.0		1.000		1.000		1.000
Represented	60.3	0.005	3.073	0.009	2.552	0.052	2.756	0.044
Overrepresented	76.9		7.470	0.003	7.664	0.007	8.710	0.005
Municipal type
Rural	50.7	0.348	1.000		1.000		1.000
Others	58.1		1.431	0.392	1.065	0.891	1.113	0.825
Political variables:
Political majority municipal council
Mainly left wing	48.1		1.377	0.519	1.072	0.902	1.086	0.895
Mainly mixed	37.8	0.005	1.000		1.000		1.000
Mainly right wing	68.0		3.932	0.004	4.098	0.012	5.224	0.006
Use of consumer-choice system
With	69.0	0.001			2.184	0.067	2.270	0.065
Without	44.1				1.000		1.000
Female politicians on muncipal council	43.3	0.992			0.960	0.314	0.960	0.320
Left-wing politicians on muncipal council	45.1	0.435			1.042	0.155	1.053	0.101
Politicians 65+ years on municipal council	19.1	0.113			0.986	0.627	0.988	0.686
Foreign-born politicians on municipal council	7.4	0.308			1.033	0.481	1.036	0.456
Previous use of paid informal care 2006
With	61.9	<.0001			5.730	0.002	5.228	0.005
Without	20.7				1.000		1.000
Financial variables:
Average profit and losses 2006–2016
Negative-low positive result	44.0						1.000
Positive result	56.6	0.102					2.196	0.093
Strong positive result	67.7						1.293	0.707
Ambition and prioritisation of services to people 65+ years (index)	0.1	0.390					1.062	0.649
Goodness-of-fit:
Models predictability rate (%)			69.5		73.5		74.2
Cox & Snell R Square			0.116		0.256		0.270
Nagelkerke R Square			0.222		0.342		0.362

However, in this model only three factors were significantly associated with the presence of a paid informal carescheme. In terms of structural factors, a positive association was observed between municipalities with either a representative or an over-representative share of foreign-born people working in municipal social care services and paid informal care in 2016. The probability of a municipality having a paid informal care scheme was almost three times as high (odds ratio = 2.76) for municipalities with a representative share of foreign-born people working in municipal social care services and almost nine times as high (odds ratio = 8.71) for municipalities with an over-representative share compared with municipalities with an underrepresentation of this group. Two positive associations were detected for the political factors. Firstly, the analysis showed that the probability of having paid informal care schemes in 2016 was more than five times as high (odds ratio = 5.22) for municipalities that had mainly been governed by a right-wing majority during the last three terms of office, compared with municipalities that mainly had a mixed majority. Secondly, municipalities with paid informal care in 2006 were more than five times as likely (odds ratio 5.228) to also have paid informal care schemes in 2016.

Discussion

This study analysed the prevalence of, change and variation in paid informal care as a policy strategy within Swedish municipal eldercare. As for the prevalence of and change in paid informal care schemes, the results showed that this policy strategy has become less common among municipalities since 2006. In 2016, just over half of municipalities reported having policies that allowed the use of either allowance or salary based schemes. Here, salary-based schemes was more common than anallowance-based scheme. Municipalities with allowance-based schemes had on average a higher number and proportion of older people receiving this form of paid informal care, compared with municipalities with salary-based schemes. However, most municipalities with either salary or allowance based scheme used these schemes to only a moderate extent.

These results indicate that a nationally driven policy of improving the support for informal caregivers or efforts to personalise care services has not significantly increased the share of municipalities with a paid informal care scheme. The limited use of these schemes in practice further suggests that this policy strategy, more often than not, is viewed as an arrangement that is to be used sparingly. These results are not very surprising considering that previous research shows that local politicians and civil servants associate paid informal care with a number of risks. Specifically, it has been viewed as problematic in relation to the quality of the care services, since most informal caregivers lack formal education and/or because it is difficult for the municipality to monitor the care provided to older people by their paid informal caregivers. Paid informal care is also believed to limit the possibilities for of already marginalised groups such as foreign-born female caregivers, to enter the regular labour market (Forsell et al., 2014; Brodin, 2018). Long-term implications, such as weak labour-market attachment and low pensions, have also been addressed by scholars within the national and international literature (see, for example, Brodin, 2018; Ungerson, 2004). However, in Sweden, only two studies have investigated paid informal care from the perspective of older service users and their informal caregivers (Mossberg Sand, 2000; Forsell, 2004). Further research is therefore needed to better understand the implications of paid informal care within Swedish eldercare.

In relation to the question of variation in paid informal care schemes among the municipalities, this study identified three factors that increased the probability of a municipality having paid informal care. Municipalities with a representative or an over-representative share of foreign-born people working in municipal social care services were more likely to have paid informal careschemes. This was contrary to what was initially assumed, i.e. that municipalities with an unrepresentative share of foreign-born workers would be more inclined to use paid informal care in order to provide services adapted to the needs of their population of older foreign-born people. Nevertheless, it is still possible that the results support the proposition that municipalities with a larger population of older foreign-born adults who require adapted services are more likely to use paid informal care schemes, i.e. the share of foreign-born workers in municipal social care services is more likely to be representative or over-representative in municipalities with paid informal care schemes simply because these municipalities are more likely to employ this group as paid informal caregivers.

The probability of having paid informal care schemes was also higher for municipalities that had mainly been governed by a right-wing majority. One explanation is that the liberal and/or conservative parties, making up a right-wing majority, are more familialist and/or market-oriented than left-wing parties. Municipalities with a right-wing majority have more positive attitudes towards the marketisation of services (Suzuki 2001), something that in itself has been promoted as a way of delivering more personalised care. The result therefore support the hypothesis that municipalities with a more positive attitude towards strengthening older peoples’ choice and control by enabling them to choose the providers of their services would also be more likely to use paid informal care schemes. However, it is worth noting that this hypothesis was tested by investigating the association between the variable use of a consumer-choice system and the prevalence of paid informal care schemes and that although the result indicated that there was a tendency (p-value = 0.065) among municipalities with consumer-choice systems to use paid informal care schemes (odds ratio = 2.27), the result was not significant. This does perhaps indicate that familialist-driven politics have a more significant impact on the likelihood of municipalities with a right-wing majority having paid informal care schemes.

Finally, municipalities with paid informal care in 2006 were more likely to have such systems in 2016. This is not surprising, as the analysis presented in Table 5 shows that the great majority of municipalities with paid informal care schemes in 2016 also had it in 2006. These results are also in line with previous research that has proved that a substantial portion of the measures taken by local authorities in relation to the provision of care is path dependent (Trydegård & Thorslund, 2001; Jenson & Lolle, 2013). Although these results lend support to some of the previously stated hypotheses, the majority of municipal factors tested had no association with the occurrence of paid informal care schemes, and much of the variation remained unexplained. Such difficulties in identifying associations could stem from difficulties in identifying relevant municipal factors or how such factors were operationalised. It is also possible that the use of aggregated municipal data in general makes it difficult to explain more of the variation. Previous studies with a similar setup investigating local variation in direct payments in the UK (Fernández et al., 2007) or services and expenditure to older people more generally (e.g. Trydegård & Thorslund, 2001; Jenson & Lolle, 2013), have struggled with similar issues. However, in a study using individual level data, local variation in home-care coverage almost vanished (Davey et al., 2006). For this study, relevant micro-level data have not been available. Yet another interpretation is that it is difficult to explain the variation with a quantitative research strategy because paid informal care as a policy is given diverse meanings at the local level, making the level of consensus low among municipalities in the way paid informal care is viewed and used. This is partly supported by the fact that only half of the municipalities with older people in need of ethnically adapted services reported using paid informal care as a strategy to meet the special needs of this group(SALAR 2004). Hence, to better understand the variation that exists among municipalities further qualitative research investigating how paid informal care is understood and used within the local context is required.

Implications of study

The decrease in municipalities with paid informal care means that this care option has become less available to the population over time. How this affects older people and family members is unknown. Considering that public support for the formal care system in Sweden is high (Daatland et al., 2011) and most older people prefer to receivi long-term services from outside their informal networks (Szebehely & Trydegård, 2007), these changes seem rather unproblematic. If, however, it is also taken into account that these changes have occurred during a time when places in residential-care have decreased and not been fully compensated by an increase in home-care services;and when the share of informal caregiving has gradually increased (Ulmanen & Szebehely, 2015), then the decline in paid informal care schemes becomes more problematic, potentially limiting the options that are made available to older people and their families in terms of giving and receiving quality care.

In particular, this trend is likely to have implications for families for whom ordinary services, for various reasons, are not a viable option. For these older adults, the abolishment of paid informal care schemes means that they can no longer access the only source of publicly funded services that could adequately meet their individual care needs. In turn, and because family members tend to continue their caregiving activities even after they stop receiving pay (Brodin, 2005), this forces informal carers to take the responsibility for providing care from local authorities. This poses a challenge for those caregivers (often women) who balance the demands of caregiving and working life, since this balancing act can negatively affect their working life and life situation as a whole (Ulmanen, 2015; Lilly et al., 2007). Considering that previous research (SALAR 2004), as well as this study suggest that paid informal care could be used to deliver home-care to foreign-born older people, the decline in these types of schemes may put this particular group of older people and their families at a higher risk of being excluded from both publicly funded servuces and the labour-market.

In light of this, the results of this study therefore calls on municipalities to consider how they handle the question of paid informal care. Concretely, municipalities without paid informal care schemes or those in the process of eliminating such schemes should consider whether their ordinary services adequately cater to all older adults in need of care services. If this is not the case, municipalities might consider using paid informal care schemes, at least until ordinary services can be developed in such a way that they can be utilised by the population as a whole. However, and due to the risks associated with these types of policies, municipalities with paid informal care also need to consider what types of schemes they use and how they use them. For example, salary provides carers with better financial security, than a non-taxable allowance. Furthermore, and as argued by Brodin (2018), municipalities can also make efforts to deal with risks such as weak labour-market attachment and issues relating to the quality of care by how they choose to organise their work surrounding paid informal care schemes.

Limitations of the study

The study has some limitations. The analysis is based on just under 60 per cent of all municipalities, thus raising questions about generalisability. However, a Pearson Chi-square test showed that the municipalities in the study were representative in terms of political majority, type of municipality and previous use of paid informal care. To analyse how the prevalence of paid informal care has changed over time, primary and secondary data were compared. However, the two sources of data were not fully comparable, thus potentially affecting the level of accuracy. Firstly, the data from 2006, labelled as salary, only include information on the municipalities’ employment of informal caregivers, while the data from 2016 also pertains to situations in which informal caregivers were contracted by the municipality or employed by private home-care providers. However, it is unlikely that the contracting of informal caregivers or their employment by private providers was common in 2006. Thus, these discrepancies do not significantly impact the possibility of making a viable comparison. Secondly, the data from 2006 only report the number of older people with either salary or care allowance and therefore do not tell us specifically whether the municipality formally used/permitted any form of paid informal care as a policy strategy. In order to analyse changes in the prevalence of paid informal care schemes between 2006 and 2016, municipalities that did not report having any older people receving paid informal care were interpreted as not having such schemes, while municipalities with one or more older persons were categorised as having this kind of scheme. This may of course mean that the prevalence of paid informal care among municipalities in 2006 is underestimated.

Ethical considerations

This study relies on municipal level data and no ethical approval was therefore necessary.

Declaration of funding

The author received no financial support for the research, authorship or publication of this article.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Satsuki Murofushi

Satsuki Murofushi is a PhD student at the Department of Social Work at the University of Gothenburg, Sweden. This article is part of her dissertation that concerns paid informal care within Swedish eldercare and its implications for older adults and their informal carers.