Introduction

Research has typically found economic wellbeing of women declines more than that of men after separation (Mortelmans, 2020). This is especially true for mothers since children often continue to live with them post-separation, increasing mothers’ risk of poverty and single-parenthood (Leopold & Kalmijn, 2016). As such, the disproportionate share of parenthood responsibilities that mothers (compared to fathers) must bear after separation make mothers particularly worse off (Leopold, 2018; Leopold & Kalmijn, 2016); the presence of children increases the resources needed by mothers’ households, which might not be offset by financial support from the father in the form of child support or public transfers (Andreß et al., 2006). Further, children in the household may limit mothers’ potential to participate in the labor market, making it harder to combine care work and paid labor activities (Mortelmans, 2020) which in turn impedes their ability to respond to the increased economic needs of their households.

Existing literature has predominantly considered post-separation economic wellbeing of mothers assuming children are solely living with the mother, paying less attention to arrangements where the child’s time is actually shared between the parents (see however Augustijn, 2022; Bonnet et al., 2022). Post-separation living arrangements, where children live with both their parents for a substantial amount of time, are however becoming increasingly common in Western countries (Hakovirta et al., 2023; Smyth, 2017). These arrangements, called shared careFootnote 1 in this study, refer to arrangements where a child lives with the other parent at least 25–50 per cent of their time.

Children’s post-separation living arrangements, e.g. whether children live mostly or solely with their mother (sole care) or whether children live with both of the parents (shared care), are a potentially important determinant of separated mothers’ economic wellbeing. This is because the parental responsibilities to care and financially support children may be allocated more evenly between mothers and fathers in shared care arrangements compared to sole care arrangements. Nevertheless, research about the impact such arrangements have on mothers’ economic wellbeing has been limited. From prior studies it is known that mothers practicing shared care are more affluent in terms of their income, education, and overall wellbeing than mothers are in sole care arrangements (for a recent review see Steinbach, 2019). Although the differences in economic wellbeing are at least partly due to the positive selection into shared care, shared care may have an independent effect on the economic wellbeing of mothers. Existing empirical literature suggests that shared care is economically advantageous for separated mothers (Augustijn, 2022), and some of the studies shed light on the mechanisms behind how shared care influences mothers’ economic circumstances (Bonnet et al., 2022). For instance, shared care reduces mothers’ time pressure compared to mothers in sole care arrangements (van der Heijden et al., 2016), which may make it easier for them to participate in the labor market (Bonnet et al., 2022).

The current study contributes to this literature by investigating not just how shared care is associated with mothers’ economic wellbeing but also whether child support and allocation of a child’s direct expenses between parents are mediators (mechanisms) through which shared care impacts mothers’ economic wellbeing. In shared care, the other parent of the child not only shares the responsibility to care for the child, but also bears part of the costs related to child-rearing, potentially reducing direct costs borne by mothers. At the same time, there is less income redistribution between parents since child support orders are less common and the amounts are lower in shared care arrangements compared to similar parents with sole care (Meyer et al., 2015). Child support offsets mothers’ post-separation child-related costs, and therefore it is seen as an important post-separation income source for mothers that helps to reduce poverty (Bartfeld, 2000; Hakovirta & Jokela, 2019). Thus, shared care may lead to a reduction in support that could reduce economic wellbeing, but also an offsetting reduction in mothers’ direct costs that could increase economic wellbeing. The interplay among care arrangements, child support, cost-sharing, and economic wellbeing can help in better understanding the economic implications of shared care and the relevance of income support policies for mothers in different care arrangements.

In this study, we examine the economic implications of shared care for separated mothers in Finland and the state of Wisconsin, US. Both Finland and Wisconsin have witnessed an increase in shared care arrangements among separated families (Meyer et al., 2022; THL, 2022), although shared care remains more common among parents with more advantageous backgrounds (Meyer et al., 2022; Miettinen et al., 2020)). In Wisconsin, shared care is the norm and has been the most common arrangement for at least a decade (Meyer et al., 2017). Moreover, compared to the whole of the United States, Wisconsin stands out as having the highest prevalence of shared care with over 50% of divorce cases opting to pursue shared over sole care arrangements (Meyer et al., 2022). In Finland shared care has only recently become an official living arrangement, although parents could have practiced shared care through visitation and access rights before the year 2019 (Tolonen et al., 2019). Over the last decade the fraction of shared care agreements of all agreements have gone up from 10 to 33% (THL, 2022), and in 2019 (the time period for this study), around 30% of separated families in Finland had shared care arrangements (Miettinen et al., 2020).

Three research questions guide our study. First, how does subjective economic wellbeing differ for separated and divorced mothers with shared as compared to sole care, net of income and other demographic characteristics? Second, how are child support receipts and sharing of direct child costs between parents associated with mothers’ subjective economic wellbeing? Lastly, do fathers’ contributions in the form of child support and sharing of direct costs mediate the relationship between shared care and economic wellbeing? The study is designed to examine how these relationships play out within the two welfare state and family policy contexts of Finland and WisconsinFootnote 2 and to compare the results across the two countries. In this vein, the study sheds light on how the increasingly common shared care arrangement and its potential effect on mothers’ economic wellbeing plays out in different family policy contexts.

Impact of Shared Care on Mothers’ Economic Wellbeing

There are two primary ways in which shared care may impact mothers’ economic circumstances differently from sole care; the first is by how much shared care increases costs, and the second is by how the costs of raising the child are distributed across parents' households (Melli & Brown, 1994). Melli and Brown (1994) argue that shared care leads to duplication of costs as both parents must have households with bedrooms, furniture, and other utilities to accommodate their children. As a result, children's costs do not decline in proportion to their reduced time in the household. For example, fixed costs such as housing costs exist even if the child is not in the household, while variable costs such as food and groceries are dependent on the time the child stays at the household. In addition, not all direct costs are automatically split proportionally in shared care since costs on items such as clothing and toys can still be borne by one parent or shared (Melli & Brown, 1994).

Overall, it is not clear from existing work the extent to which economic needs of mothers’ households change even if children spend considerable amounts of time away from the household. Paul Henman (2005) investigated the costs of a child in post-separation families with standard budgets in Australia, making several assumptions of the costs borne by each parent. The study found half-time children’s costs are 72 per cent of the costs of full-time children. The costs were even higher (82 per cent) for an unemployed parent because of the higher proportion of fixed costs in low-income families, implying that shared care is not necessarily associated with substantial cost savings especially for low-income parents. Nonetheless, recent work in Wisconsin confirms that in shared as compared to mother-sole care arrangements, fathers pay a substantially larger share of variable child costs across a variety of domains, consistent with nontrivial savings in direct costs when children divide their time between homes (Bartfeld et al., 2022).

In addition to impacting mothers’ costs, the time a child spends with each parent also affects child support payments; child support orders in shared care either are more likely to be absent or are much lower compared to orders in case of sole care arrangements. This, however, may depend on the parent's relative economic resources, since parents’ resources are typically accounted for in child support determination (Hakovirta & Skinner, 2021).

The mechanism through which all of this may impact mothers’ economic wellbeing outcomes in shared care is that lower child support payments or no payments at all may potentially decrease the economic wellbeing of shared care mothers. At the same time however, greater extent of direct cost sharing between parents could increase shared care mothers’ economic wellbeing. As (author citation) note, the economic wellbeing impact of shared care depends therefore on whether the decrease in mothers’ income due to lower child support payments are offset by the declining child-related costs in mothers’ households that affect the economic needs of the household.

Different Contexts, Different Outcomes?

Family Policies in Finland and Wisconsin

Policies that reduce care and financial dependency between women and men in welfare states mediate the economic consequences of separation for women. Countries with higher levels of public support available for post-separation families may help mothers to offset the income losses related to separation (Andreß et al., 2006; De Vaus et al., 2017; Uunk, 2004), and to be less dependent on the economic support of the other parent of the child (Andreß et al., 2006). Finland’s family policies can be characterized as having high levels of public support for families to share the care of the child by supporting women’s employment (Eydal et al., 2018). Compared to the US, Finland has higher levels of public spending both on child-related services and cash transfers to families with children (Hakovirta et al., 2022a, 2022b). In contrast, the US is depicted as a residual welfare state where there is less state redistribution, and the obligation to care and financially support children resides primarily with families (Berger & Carlson, 2020). Wisconsin, the US state being studied in this paper, represents an average state in the United States in terms of its generosity and coverage of family policies (Daiger von Gleichen & Parolin, 2020). Specific to the post-separation context, Wisconsin does not have a public guaranteed child support scheme like Finland, which guarantees a minimum payment of child support regardless of whether the other parent pays the owed child support or has the financial means to it (Hakovirta et al., 2022a). Based on the differences in the extent of family policies, mothers’ economic wellbeing post-separation in Wisconsin may be more dependent on their labor market income and the financial support provided by the former spouse, and in Finland less dependent on how parents share the financial responsibility of the child.

Shared Care and Child Support in Finland and Wisconsin

Even if parents share the care of the child, both in Finland and Wisconsin there may still be child support agreements between the parents. When child support orders are determined in shared care cases, the economic resources of both parents are considered and the assumed shift in costs from the mother’s household to the father’s household reduces the expected child support. Comparing the child support policy outcomes in Finland and Wisconsin for parents with gender specific median income, Hakovirta et al., (2022a, 2022b) found that the expected child support amounts are higher in Wisconsin when a child has a sole care arrangement. However, the drop in the expected amount was substantially larger in Wisconsin than in Finland when parents had a shared versus sole care arrangement. Child support determinations are therefore more strongly attached to the living arrangements in Wisconsin than in Finland. From the perspective of mothers' household income, this implies that the trade-off between sharing children’s costs and receiving child support may be more dramatic for Wisconsin mothers in shared care than for shared care mothers in Finland. Thus, shared care may more negatively affect mothers’ economic wellbeing if the income loss is not offset by the other parent sharing the child’s expenses directly, or in the lack of public support offsetting the child-related costs.

Both in Finland and Wisconsin it is, however, less common among shared care parents to have child support orders (Meyer et al., 2015; Miettinen et al., 2020), which ultimately makes the locus of responsibility to divide the child’s costs rest on parents. Moreover, parents in Finland have the discretion to enter into the formal child support system and exercise discretion within the formal system when determining child support amounts (Skinner & Davidson, 2009), which might lead to heterogeneous child support outcomes for shared care families. In Wisconsin, courts must approve child support arrangements for divorcing parents regardless of placement but do have the authority to issue or approve orders that differ from guidelines; in practice this has resulted in fewer orders and orders that are likely lower than guidelines dictate, compared to sole care (Cook & Brown, 2013).

Measuring Economic Wellbeing

The study of economic or financial wellbeing through subjective measures of economic wellbeing is becoming increasingly popular in household finance (Collins & Urban, 2020). Subjective wellbeing measures usually capture “features of individuals’ perceptions of their experiences” rather than their utility in economic terms through objective measures such as income, wealth, or income poverty measures (Kahneman & Krueger, 2006). Such measures, like the Financial Wellbeing (FWB) Scale developed by the United States’ Consumer Financial Protection Bureau (CFPB), have been found to reliably track individuals’ wealth, income, and financial health, and shown to be different from measures of overall subjective wellbeing (Collins & Urban, 2020). The CFPB defines financial wellbeing as “a state of being wherein a person can fully meet current and ongoing financial obligations, can feel secure in their financial future, and is able to make choices that allow them to enjoy life” (CFPB, 2017). A particular advantage of such a measure in the current study is the fact that it may be responsive to fluidity in household membership where children are only part-time members. It allows us to assess how mothers’ perceptions of wellbeing differ with different care arrangements net of earnings and other relevant factors. Objective measures of characterizing economic wellbeing such as income poverty are limited in more complex living arrangements as it is difficult to ascertain who should be counted as a part of the household while computing these measures.

While the CFPB’s comprehensive FWB scale operationalizes financial wellbeing using multiple dimensions, we restrict the scope of this study to singularly focus on one of those dimensions, namely, an individual’s “control over day-to-day, month-to-month finances” (CFPB, 2017). Our measure of economic hardship, which asks respondents how difficult it is for them to meet regular expenses in a typical month, is a modified version of the survey item included in the Survey of Household Economics and Decision-making (SHED) by the United States’ Federal Reserve (Board of Governors of the Federal Reserve System (U.S.), 2021).

Methodology

The aim of the study is to understand the mechanisms through which shared care impacts mothers’ economic wellbeing in two welfare state contexts with differing forms of family policies (Finland and Wisconsin). We examine mothers’ economic wellbeing in this study through mothers’ self-reports of how difficult it is to cover their household’s expenses, which we operationalize as a measure of subjective economic wellbeing. Based on the literature discussed above, the direction of the association between shared care and subjective economic wellbeing is ambiguous and depends on the extent to which lower child support is offset by greater sharing of mothers’ direct child costs. Because child support policy implies larger declines in child support with shared versus sole care in Wisconsin than in Finland, our first expectation is that the economic advantages of shared care would be smaller in Wisconsin. This is because, compared to Finland, in Wisconsin fathers have to contribute towards the costs of the child more through other ways than child support to offset the income loss due the lower child support mothers are receiving in shared care compared to sole care arrangement. Our second expectation is that regardless of living arrangements, child support and greater cost-sharing would both be associated with less economic hardship in both of the countries. Our third and last expectation is that in Wisconsin and Finland, child support and cost-sharing would mediate any association between living arrangements and economic wellbeing.

Data and Sample

Data for this study come from two surveys on separated families’ living arrangements and economic indicators, designed and fielded separately in Finland and in Wisconsin. We use the Survey Among Separated Parents in Finland, conducted by the Social Insurance Institution of Finland (Kela). The Survey Among Separated Parents is a web-based survey which was conducted among residential and non-residential parents in 2019. A sample of 7,000 residential and 7,000 non-residential parents with at least one child born in 2002, 2005, 2007, 2009, 2011, 2013, 2015 or 2017 was drawn from a register-based dataset (compiled by Kela), which comprised all parents (and not just those who were married) with children below 18 years of age who were not living together with the other parent of the child. The sample was restricted to those parents whose native tongue was Finnish, Swedish or Sami, and who lived in Finland in 2019. The sample was further restricted to those for whom a valid e-mail address could be obtained from Kela’s administrative registers. The samples of the residential and non-residential parents were drawn separately, and the survey was predominantly answered by only one of the parents of the focal child. The register dataset included information on the birth years of resident/non-resident children to a parent. The focal child was randomly chosen from these children, and the parent was asked to report mainly about this child in the questionnaire. The age range of the focal children was between 1 and 17 years at the time of the survey. The response rate among residential parents was 32% (n = 2156) and among non-residential parents 20% (n = 1293). Both the residential and non-residential responses were used in this study. Residential parents were mostly mothers (90%), and non-residential parents mostly fathers (71%). We only use data from mothers in this study.

The Wisconsin data come from the Wisconsin Parents Survey fielded by researchers at the [name redacted for peer review] in 2020.Footnote 3 This survey collected data on divorced parents (excluding parents separated from a cohabiting relationship) drawn from cohorts 30 and 33 of the Wisconsin Court Records Data (CRD), which is a sample of divorce and paternity cases coming to court in 21 Wisconsin counties. The surveyed parents began divorce proceedings 6–10 years before the survey, between 2009–2010 (cohort 30) 2013 (2013) with at least one child under six at the date of the divorce petition and comprised mothers in sole care and both parents in shared care arrangements. The youngest child at the time of the divorce was termed as the focal child for the purposes of the survey, and information on the living arrangements of this child was collected and used to classify mothers as being in shared versus sole care arrangements. Data was collected through in-person interviews before the COVID-19 pandemic and subsequently through telephonic interviews conducted by trained personnel from the University of Wisconsin Survey Center. The response rate for the survey was 54.8% overall, 56% among shared care mothers, and 54.3% among sole care mothers. Information on time of divorce and duration of marriage are obtained from the Wisconsin CRD. We link this data with earnings records from the Unemployment Insurance Program from Wisconsin’s Department of Workforce Development which we use to measure of respondents’ own earnings.

We took several steps to make the data comparable for analysis: first, we restricted the Finnish and US sample to mothers with sole and shared care. To define the focal child’s living arrangement as shared or sole care, we first calculated the number of nights the child stays with each parent in a typical month, and then divided the sample into sole and shared care by the percentage of time the focal child spends with each parent separately for Finland and Wisconsin. In this study, we follow Wisconsin’s threshold for defining shared care, which categorizes mothers who spend 25–75% of overnights in a year with children as being in shared care, and mothers who live with their children for more than 75% of overnights as being in a sole care arrangement. Thus, in shared care, a child can either spend more time with one or the other parent or an equal amount of time with each parent. These definitions impact which policy guidelines are followed in determining child support and are therefore relevant for how economic wellbeing might vary by these groups at least in Wisconsin, where child support amounts are more strictly tied to time-share arrangements than Finland. Further, in Finland reduction to child support starts when a child has at least 7 overnight stays in a month with the non-resident parent equalling a little under 25% of the time. Second, we further restricted the Finnish sample to cases where the respondent had dissolved a cohabitational relationship of marriage or consensual union at least six years ago. This was to address the predominant difference in the sampling frame between the two surveys and ensured we were studying divorced mothers at the same point in time during the post-divorce period, and that focal children were similar in age in both samples. Finally, we excluded cases with missing information on study variables using listwise deletion (total of 50 cases for Finland and 10 for Wisconsin). After these procedures we ended up with 850 mothers in Finland of which 608 had sole care and 242 shared care of the focal child, and for the US with 395 mothers of which 207 had sole care and 188 had shared care of the focal child.Footnote 4

Table 1 describes the mothers in our sample for both countries. In both countries, the mothers are on average 41 years old and half of them are partnered with about two children in their households on average. The focal child is between eleven and 12 years old in Wisconsin and thirteen years old on average in Finland.Footnote 5 A majority of mothers have high levels of education in both places, and are more likely to work full time versus part-time or be unemployed. Mothers in Finland earn about $35,000 on average annually while Wisconsin mothers earn about $48,000. This difference in earnings may stem from the fact that the Wisconsin sample has only divorces, and not separated cohabitations, so that the segment of all separated parents being studied in Wisconsin is relatively more advantaged.Footnote 6 Annual spousal earnings are around $20,000 for mothers in both places, and this average is across all mothers irrespective of their partnership status (and considers spousal earnings as zero when a spouse or partner is not present). Close to 60% of mothers in Wisconsin own a home as opposed to 63% in Finland. Mothers in both places have been separated for about 8–9 years on average.

Table 1 Descriptive statistics
Full size table

Dependent Variable

Our dependent variable is operationalized as a binary variable indicating perceived hardship, and is created from a survey item which asked respondents how difficult it was to cover regular expenses. The Finnish survey asked “When you consider your household income as a whole, is covering your expenses with the income (1) Very difficult, (2) Difficult, (3) Fairly difficult, (4) Fairly easy, (5) Easy, and (6) Very easy” where respondents could choose any option between (1) and (6). The US survey asked, “Over the past year, how difficult has it been to cover your expenses and pay all your bills in a typical month?”, to which respondents could answer (1) Not at all, (2) A little, (3) Somewhat, (4) Very, (5) Extremely. The interpretation of “regular expenses" was left up to the respondents, and could potentially include being able to pay bills on time, manage any outstanding debt, and make ends meet. Both the Finnish and US measures were dichotomized to binary variables where 0 indicates no difficulties and 1 denotes difficulties to cover their expenses. For the difficulty category of the binary variable, we combined very difficult, difficult, and fairly difficult for the Finnish data and somewhat, very, and extremely for the Wisconsin data. The binary difficulty variable is therefore an indicator for mothers facing at least some difficulties in meeting their monthly expenditures and is referred to as perceived hardship throughout the paper. We report the distributions of the individual variables in the appendix and carry out sensitivity tests with alternative ways of defining this variable.

Independent Variables

Shared care or focal child’s living arrangement: The main variable of interest is living arrangements, differentiating shared care from sole care as described above. Child support: Child support is measured as a continuous variable for the annual amount of child support received for the focal child and the focal child’s siblings, scaled to thousands of US dollars. The original Finnish child support amount variable was a monthly amount for focal child which was multiplied to annual and US$PPP converted (1$ = 0.863). Extent of cost-sharing: Cost-sharing captures the way parents divide direct child costs as distinct from support via child support payments. We create this measure based on a scale for which higher values indicate greater amounts of costs shared between parents. In both the Finnish and Wisconsin data, this variable is created by averaging responses across a battery of items which ask respondents how expenses on clothing, school, extracurricular activities, insurance, medical expenses, and childcare were shared in the past 12 months. Respondents could answer these underlying items on a 5-point scale ranging from (1) the respondent paid for all the expenses to (5) the other parent paid all the expenses. The resulting measure is a continuous variable taking values between one and five, five indicating the highest extent of costs shared by the other parent.

Control Variables

We control for mothers’ own annual earnings including unemployment benefits, her partnership status (single versus cohabiting/married), and annual partner earnings in our models of subjective financial wellbeing and shared care. The original Finnish income variables were monthly income which were multiplied to annual and US$PPP converted (1$ = 0.863). Other covariates include mothers’ education, employment status and age, number of minors in mother’s household, age of the focal child, time since mother had separated from the focal child’s other parent, and an indicator for whether the mother owns a home. Following the International Standard Classification for Education (ISCED), we classified mothers by their highest education to (1) low, (2) middle and (3) high educated mothers. Based on their current employment status, mothers were classified to be (1) full-time employed, (2) part-time employed or (3) unemployed or other. We used continuous variables to account for mothers’ and focal child’s age, the number of minors in the mother’s household as well as the time since separation. We also added a dummy variable to control for mothers who had no child support owed to them.

Analytical Framework

We estimate logistic regression models with robust standard errors treating the dichotomized hardship measure as primary dependent variable. We use a sequential modeling strategy, first controlling only for shared care, and subsequently adding in control variables followed by child support, and cost-sharing into the model to disentangle how they influence the shared care coefficient. All regression analyses are done separately for the Finnish and Wisconsin samples. This allows us to carry out within-country comparisons of how care arrangements are associated with hardship and the role child support and cost-sharing play in this relationship. For all point estimates, we present average marginal effects.

Results

Descriptive Results

Table 2 shows unconditional differences in the percentage of mothers in sole and shared care on the four analytic variables of interest. In both Finland and Wisconsin, mothers in shared care are less likely than those with sole care to experience hardship (i.e., have lower economic wellbeing) (50% vs. 62% in Finland and 31% vs. 45% in Wisconsin, p < 0.01). The amount of child support received on average is also much lower than that of mothers in sole care, while the extent of cost sharing is much higher among mothers in shared versus sole care in both places (p < 0.001), all of which is consistent with policy and prior literature. The overall higher rates of hardship in Finland as compared to Wisconsin are consistent with the lower earnings of mothers in the Finnish sample (see Table 1).

Table 2 Descriptive statistics of analytical variables by care arrangements in Finland and Wisconsin
Full size table

Multivariate Results

Results from multivariate models of maternal hardship and shared care in Finland and Wisconsin are reported presented in Tables 3 and 4. We add our independent variables of interest sequentially to a model of hardship and shared care in each table, with the first column showing bivariate results, followed by a model including control variables, and subsequently adding child support and cost-sharing in column three. Column 2 thus addresses our first research question, namely the association between placement and mothers’ subjective economic wellbeing net of demographic and economic differences between placement groups. Column 3 addresses our second and third research questions, namely the relationships of child support and cost-sharing to mothers’ subjective economic wellbeing and the extent to which these moderate the role of placement. We find Finnish mothers in shared care are significantly less likely to experience hardship by 11.6 percentage points before controlling for any characteristics (Table 3). This advantage disappears when we add in our vector of control variables in column (2), with mothers’ own earnings playing the largest role in reducing the magnitude of the shared care coefficient (sequential result not shown). Other variables in the model largely have the expected relationships. Economic hardship is less common as partners’ earnings increase, and less common for homeowners compared to others, while hardship is more common with more children in the home and when a partner is in the home (both indicating more people share the available resources). This model shows that after controlling for observable socioeconomic characteristics, there are no remaining care-related differences in subjective economic wellbeing. Adding in child support and cost-sharing in Column (3) does not have any notable effect on the shared care coefficient, which remains statistically insignificant and substantively small throughout. Child support does not have any significant impact on mothers’ perceived hardship either, although the coefficient is in the expected direction and similar in magnitude to the coefficient on own and partners’ earnings. Greater cost-sharing in itself significantly reduces the odds of mothers reporting hardship by 7 percentage points for each point in the scale.

Table 3 Logit regressions of difficulty to cover monthly expenses on shared care in Finland
Full size table
Table 4 Coefficients from logit regressions of hardship on shared care in Wisconsin
Full size table

Similar to Finland, shared care mothers in Wisconsin are 14 percentage points less likely to report hardship on average (Table 4). In Wisconsin however, this effect persists even after adding individual level controls including own and partners’ earnings in column (2). As in Finland, hardship is less common as mothers’ earnings and partners’ earnings increase and with homeownership, and more common with more children in the home. The shared care coefficient indicates that even after controlling for socioeconomic characteristics, there appears to be a net reduction in hardship with shared care. The last column shows that after controlling for fathers’ contributions in child support and sharing of direct costs, there is no remaining care type related difference. The child support coefficient itself is not significant, though as in Finland it is in the expected direction and similar in magnitude to that for mothers’ earnings. In contrast to Finland, cost-sharing is not individually significant but also in the expected direction. Taken together, this result shows that fathers’ child support and cost-sharing contributions, while not statistically significant themselves, appear to account for the shared care advantage.Footnote 7 We then expand our shared care categories into equal and unequal shared to examine granularity in the relationship between shared care and hardship (Table 5). We define equal shared care as children spending exactly 50% of overnights with each parent in a typical month. Panel A in Table 5 shows that in Finland, the coefficient on the unequal shared care remains insignificant but negatively associated with hardship throughout the models, while the equal shared variable drives the effect of shared care on hardship as seen in Table 4. We find similar results in Wisconsin (Panel B),Footnote 8 with mothers in equal shared care having significantly lower likelihood of reporting hardship throughout, from being 21 percentage points less likely than sole mothers before controlling for anything else, to being 14 percentage points less likely to report hardship after accounting for child support and cost-sharing. The shared care advantage in economic wellbeing is therefore driven by mothers in equal care arrangements in both countries.

Table 5 Coefficients from logit regressions of hardship on shared care in Finland and Wisconsin, expanded care categories
Full size table

Sensitivity Tests

We check the sensitivity of our results to how we define our outcome and independent variable of interest and an alternative model specification (results shown in Appendix Table 6). We use an OLS specification with the original 6-point and 5-point scale definitions of the dependent variable and find that in both contexts, the direction of the effect remains the same as in the logit specification but the precision of the estimates is different. Specifically, for Finland we find that shared care mothers are less likely to report hardship when we do not control for anything else in the model and the magnitude of this association reduces and ceases to be significant when we add controls (Appendix Table 6, Panel A). Cost-sharing is significantly negatively associated with wellbeing in this specification as well. Overall, results remain consistent with our primary model: mothers with shared care report less hardship; this is fully explained by the control variables; child support has no significant association with hardship; and more cost-sharing is associated with less hardship. In Wisconsin, the coefficient on the shared care variable reduces in significance and magnitude when we add control variables and becomes smaller and insignificant with the addition of child support and cost-sharing in our models. The overall results—that baseline differences in economic hardship persist after economic and demographic controls but are no longer evident after controlling for child support and cost-sharing—are largely consistent with the primary model. (Appendix Table 6, Panel B).

Discussion

In this study, we investigated whether subjective economic wellbeing differs for separated and divorced mothers with shared as compared to sole care, net of income and other demographic characteristics. We further examined how child support receipts and sharing of direct child costs between parents are associated with mothers’ subjective economic wellbeing. Finally, we studied whether fathers’ contributions in the form of child support and sharing of direct costs mediate the relationship between shared care and economic wellbeing. Our findings suggest mothers in shared care arrangements have greater economic wellbeing in Finland and Wisconsin (US), and, departing from prior research (Augustijn, 2022), that this is driven by equal versus unequal time share arrangements. While child support receipt does not have a significant association with hardship in either context, there is some evidence that sharing of children’s costs are suggestively negatively associated with hardship in both. Further, the relationship between shared care and mothers’ experience of economic hardship can be fully explained by controlling for several characteristics of the mothers and the mechanisms through which this relationship manifests in both contexts; child support and sharing the costs of children between the parents. Our findings also complement those of (author citation) who find that shared care in Wisconsin is not associated with better income-to-poverty ratios up to four years after divorce, presumably due to greater cost-sharing in shared care households.

The potential reason behind the economic advantages of shared care differs between Finland and Wisconsin. While shared care mothers experience less economic hardship, this relationship holds after controlling for income and demographic characteristics only among mothers in Wisconsin—not those in Finland. In Wisconsin, the relationship becomes insignificant after accounting for child support receipts and cost-sharing. In Finland, socio-economic characteristics, particularly mothers’ own income, play a much larger role in explaining the shared care advantage in economic wellbeing. This suggests there may be economic advantages to shared care in Wisconsin that are linked to overall differences in how fathers help offset child-related costs in different care arrangements.

The economic advantages of shared care were expected to be smaller in Wisconsin compared to Finland due, primarily, the greater decline in child support amount in shared care arrangements. A potential explanation why this assumption did not hold is that in both countries child support determination may depart from the guidelines, which makes the predictability of how policies affect payments in different countries difficult to observe. Especially in Finland, how child support payments are determined might not affect a large proportion of mothers in shared care, since parents have the freedom to make private arrangements, which is a common practice among shared care parents. In fact, our results indicate that child support has no direct influence on subjective economic wellbeing, contradicting our second expectation, in either context. Child support comprises only a small portion of mothers’ post-separation income both in Finland and Wisconsin (author citation; Hakovirta & Jokela, 2019), which may explain the result.

Finally, we find that mothers benefit from sharing children’s costs with the other parent. The extent of cost-sharing between the parents reduces mothers’ economic hardship in Finland. While the relationship was insignificant in Wisconsin, child support along with cost-sharing, both of which account for father’s contributions to their children’s expenses, do, however, fully explain the shared care advantage, supporting our third expectation. Overall, our findings suggest that the relationship between mothers’ economic wellbeing and fathers' contribution to children’s expenses in Finland and Wisconsin are quite similar altogether, despite differences in the levels of public support for families in the two contexts.

Limitations

Our country case study design however has its limitations as only the differences of the factors associated with economic wellbeing within a country context can be compared. We also note the sample designs between the Finnish and Wisconsin surveys were different—while the Wisconsin sample consisted of mothers who underwent divorce at least six years prior to the administration of the survey, the Finnish survey included no such restriction. However, by limiting the Finnish sample to, keep only those mothers who had separated six years ago we were able to remove this difference. Due to the restriction to the sample, this study investigated mothers with children between six and 18 years of age. Whether the same results would hold among mothers with younger children is an avenue for future research.

Finally, the Finnish sample includes union dissolutions of married as well as cohabiting mothers, as compared to the Wisconsin sample which only includes divorcing mothers. This is because cohabitation is common in Finland in all socio-economic groups, even when there are children, though many couples tend to marry after the birth of their first common child (Jalovaara & Andersson, 2023). Because married couples in the US tend to have higher incomes than cohabiting couples, this may contribute to a more economically advantaged sample in Wisconsin. It is possible that our finding about Finnish mothers’ own income having a stronger association with economic wellbeing might be because they are overall more economically disadvantaged, in that differences in earnings may be most important in impacting hardship at lower versus higher income levels. Child support policy, however, addresses children from married and cohabiting union dissolutions in similar fashion, both in WI and Finland.Finally, unlike the Finnish sample, the Wisconsin sample only has divorced mothers and not mothers who have separated from a cohabitation which would imply they are economically more advantaged than their Finnish counterparts.

Conclusion and Implications

Despite these limitations, the study suggests child’s post-separation living arrangement, and the mechanisms through which it influences the economic wellbeing of mothers, can yield similar economic wellbeing outcomes for mothers in different family policy contexts. Although, post-separation economic outcomes for mothers have been found to be less severe in welfare states with more support to families and less dependency between the former partners (e.g. Andreß et al., 2006), the mechanism through which the other parent share the costs of children may still be relevant irrespective of the context. Moreover, while studies have investigated formal child support arrangements and their effect on mothers’ post-separation economic wellbeing (e.g. Bartfeld, 2000; Hakovirta & Jokela, 2019), we acknowledge the importance of how parents share the direct costs related to children. This is especially important when investigating shared care families, since in both countries it is less common among shared care parents to have child support orders (Meyer et al., 2015; Miettinen et al., 2020). Thus, the allocation of children’s costs rests more with parents. What impacts parents’ cost-sharing in shared care is a black-box in research. As was seen from the results of this study, how these direct costs are shared has implications for mothers’ post-separation economic wellbeing. Whether parents share the direct expenses of a child equally or whether parents take into consideration their relative ability to support the child is, for example, an important area for future studies. Furthermore, as the direct expenses are allocated between the parents in shared care, there are potential consequences for fathers’ post-separation economic wellbeing as well, which were out of the scope of this study, but should be explored in future studies.

The research has two main implications. First, it highlights the importance of shared financial responsibilities between parents in alleviating economic challenges for mothers. The study complements prior studies that have found economic benefits in shared care (Augustijn, 2022), and suggests that mothers in shared care arrangements could experience less economic hardship if both parents actively financially contribute to raising their children. Thus, as mothers usually see the decline in their economic wellbeing after divorce/separation (Mortelmans, 2020), shared care arrangements in which sharing the costs of the child go hand-in-hand with sharing the care of the child, could buffer against the financial ‘shock’ of separation. Secondly, the findings of the study call for attention of policymakers towards how parents in shared care arrangements share the costs of children. The aim should be to promote a fair distribution of financial responsibilities. Policies that promote equal forms of family not only in relation to allocating the care responsibilities but also allocating the costs of the child may help in alleviating economic hardship of mothers post-separation (possibly by incentivising and promoting maternal employment). This is so especially in contexts with less public support for families with children as exemplified by the differing outcomes observed in Wisconsin compared to Finland. Relatedly, the findings of the study are important for child poverty as well. Children in shared care arrangements are less exposed to the risk of poverty than those living mainly with their mother (Bonnet & Solaz, 2023), and therefore understanding the mechanism through which this appears is important in attempts to tackle child poverty.