Unmasking complexity: the necessity of increasing accountability in the Spanish care home sector

The proportion of individuals aged 65 and over in the European Union is projected to increase from 20% in 2020 to 30% in 2070 as the population ages (European Commission, 2020a). As a consequence of this expected evolution, there is a growing demand for care homes that cater to the needs of the elderly. Care homes focus on providing high-quality services (social rationale) and charge fees that guaranty the viability of their activity (economic rationale). One salient configuration within the landscape of care homes that has gained prominence is the emergence of large and complex care home chains. These chains have materialized through a series of mergers and acquisitions that have resulted in a significant increase in their market share. This phenomenon originated in the United States during the 1990s but has already transcended international boundaries (Campbell and Harrington, 2024) ^[1]. As Senator Warren highlights in Elizabeth (2019), there is a need to scrutinize how these new chains actually operate. In line with this concern, there have been claims that social and economic rationales might not be achieved in perfect alignment. These claims lead us to wonder whether the quality care that should prevail is compatible with the level of profit released by these companies. Preliminary evidence on this lack of compatibility based on North-American case studies has shown that large complex chains produce high profit margins to the detriment of service quality (McGregor et al., 2005; O'Neill et al., 2003; Stevenson et al., 2013). This initial evidence is, however, insufficient, as it does not consider the specificities of certain geographical contexts and is based on a methodology that does not allow for its generalization. Furthermore, large complex chains might be interested in putting their economic enrichment out of sight to protect the damaged reputation that results from a poor-quality service. More recent analyses show how large corporate chains, especially those under private equity (PE) ownership, use complex organizational structures to conceal the financial consequences of low-quality care, which is known as leakage. In these cases, the economic performance may be underreported by overcharging expenses not necessarily related to the quality of their services such as monitoring management fees, lease agreements, interest payments to owners, and purchases from related party companies (Harrington et al., 2017; Gupta et al., 2024).

Although the political debate has not been as prominent in Europe, similar tensions are present, and not only because of the growing presence of large complex chains. The European region requires more detailed studies since some European countries, especially those in the Mediterranean area, have high public coverage of care home services and this could significantly affect the results. Genet et al. (2012) and Llena-Nozal et al. (2025) suggest that Mediterranean countries combine high public funding with limited investment in publicly owned facilities and a greater reliance on outsourcing to private providers. This configuration helps to explain why, paradoxically, despite substantial public granting, a significant share of care home beds is actually provided by private companies. As a consequence, privately owned care homes manage high number of beds outsourced by public administrations. Most public funds are therefore used to pay private companies a previously negotiated price for each outsourced bed. As it is common that such companies demand increases in these “publicly assigned” prices, a new income-decreasing signaling incentive arises for them. Harrington et al. (2011, 2015, 2017) argued that private companies that manage public beds signal the high (presumably inflated) cost structures they face to achieve socially-acceptable levels of quality. They do this to justify an increase in the prices received for these beds.

Whereas in general the trimming of obligatory (necessary) costs that reduce quality service might be joined with a simultaneous leakage to avoid the bad reputation of abnormally high profits, in contexts with high public funding, leakage practices could also be related with the demand for an increase in the prices received from the public authorities. Which of these two alternatives is the cause for the lack of alignment is a relevant issue that has not been explicitly answered in the literature. The objective of our study is, then, to examine if the behavior of private care homes depends on the receipt of funds from the public authorities. We pay especial attention to the large complex chains that have recently concentrated a high proportion of the market share as these are the main receivers of these funds.

Our main research questions are:

Although some studies can be found that separately analyzes the effect of large complex chains or that of public funding in either the social or the economic performance of care homes (Grabowski et al., 2013; Harrington et al., 2015; McGregor et al., 2005; O'Neill et al., 2003; Stevenson et al., 2013), to the best of our knowledge, none of these studies has jointly analyzed the effect of these two pairs of circumstances (large complex chains, public funding) and performance outputs (economic, social). Filling this research gap is necessary in countries with a high dependence of public funds. For this purpose, we use data from a sample of 1,658 companies operating in the Spanish care home sector during the period 2014–2021. The Spanish case is perfectly suitable for this objective as this country has experienced an intense concentration process in large complex chains and around 77% of public beds are outsourced to privately owned care homes. Consequently, a large percentage of private beds (37%) receive public grants (IMSERSO, 2024). Furthermore, large complex chains are the main beneficiaries of this public funding in Spain. The two largest operators (DomusVi and Orpea) report that around 56% and 35% of their beds, respectively, are supported by public funds (El País, 2020). Although the 2006 reform introduced significant improvements ^[2], the Spanish long-term care system still faces major challenges, mainly due to persistent underfunding, divergent priorities, and coordination problems between national and regional authorities, all of which continue to hinder effective implementation. Finally, the period of study includes the COVID-19 pandemic and for these months data on mortality in care homes was exceptionally released. This provides a unique opportunity to include a measure of social performance in conjunction with the economic ones.

Our results provide clear evidence that large complex chains show lower economic performance than the other organizations in the sector and that this difference is enhanced when they receive public funds. Large complex chains also exhibit a lower social performance, but in this case the public funding interaction does not show any significant effect. These results reflect a lower alignment in these chains that would be explained more by signaling intentions to justify the demand for increasing funds from the public administration than by concerns about a socially damaged reputation. We strengthen the results by separately studying the trade-off between obligatory (necessary) and discretionary (unnecessary) costs and find that large complex chains cut the former and overstate the latter. We thus confirm that these chains dedicate fewer resources to the well-being of their patients regardless of the source of their funds, but those receiving public funds show a higher need to cover their actual economic status.

We contribute to the academic literature on the care home sector in several ways. First, we extend the literature on the complexity of these entities by introducing the effect of public funding. This is especially relevant in European countries like Spain, where there is a high dependency on this type of finance, especially by large complex chains. Second, to the best of our knowledge, no previous study has analyzed the level of social performance in conjunction with economic performance. Finally, the fact that we separately study two types of expenses, obligatory (necessary) and discretionary (unnecessary), allow us to better establish the interrelationship between economic and social performance of these chains in the same study. This separate analysis constitutes a novelty in this field.

Our findings call into question the alignment that large complex chains claim to achieve between economic and social rationales in carrying out their activities. Such alignment should avoid the inferior levels of service quality that we have observed for these chains. Furthermore, and contrary to the initial evidence of abnormally higher profits (McGregor et al., 2005; O'Neill et al., 2003; Stevenson et al., 2013), we have also observed the practice of leakage that intends to signal the opposite. Our results show the need to enforce two types of inspections, a quality control of the service and an economic audit of the accounts, especially for those care homes that are receiving public funds. It is also necessary to review and clarify the legal framework so that “everyone has the right to affordable long-term care services of good quality, in particular home-care and community-based services” (European Pillar of Social Rights, 2023). In the specific case of Spain, the European Commission (2018, 2020a) has highlighted some shortfalls in the development of institutional coordination, unequal implementation and coverage in different regions resulting from the lack of a common regulatory framework ^[3].

The rest of this paper is organized as follows. The next section revises the literature on complex ownership structures and presents the first two hypotheses. The impact of public funding and the third and fourth hypotheses are examined in Section 3. Section 4 describes the main characteristics of the Spanish context and Section 5 describes the sample, main variables, and statistical techniques employed to test the hypotheses. The results are presented in Section 6, followed by a comprehensive discussion of the findings and the main conclusions in Section 7.

Traditionally, much of the focus of the ownership literature has been on differences between for-profit and not-for-profit facilities. Stevenson et al. (2013) highlighted that the simple distinction between these two categories is insufficient, because ownership structures vary significantly within the for-profit group. One of their key findings is the increasing complexity of corporate structures in large chains–defined as groups of private-sector entities that operate multiple care home facilities under unified brand or corporate control. The complexity is associated with multi-layered corporate configurations involving several legal entities, often spread across different sectors and jurisdictions (Phan and Zurbruegg, 2024). In the North American case, Kingsley and Harrington (2022) note that some US for-profit care home companies have developed corporate ownership structures with seven or eight layers of companies. For example, the second-largest care home chain in the United States—The Ensign Group Inc.—has developed a complex corporate structure with 22 portfolio companies that control a multilevel intricate network of 409 separate companies to manage its 228 care homes.

These complex structures extend beyond organizational design, raising serious concerns regarding transparency and accountability. According to Paligorova and Xu (2012), these kinds of complex ownership structures are associated with increased opacity and information asymmetry, reinforced when subsidiaries are controlled from abroad (Gill-de-Albornoz and Rusanescu, 2022). A higher number of subsidiaries facilitate intragroup transactions, which makes it harder to detect tunnelling activities ^[4]. This type of structure is usually developed by a singular type of shareholder, which represents another characteristic of large complex chains. Harrington et al. (2017) note that the largest complex chains in the United States are primarily owned by PE and institutional investors. Gupta et al. (2024) analyzed the effect of PE ownership and showed that PE-backed buyouts tend to provide services with lower quality of care and have no effect on operational profitability. This is possible using leakage practices. Leakage refers, particularly in the Anglo-Saxon context, to the diversion of resources through abnormal or excessive expenses such as management fees, lease rents, interest payments, consultancy, or other services related to the operation of care homes. These transactions often involve fees charged above market rates, which can reduce the resources available for direct care.

In the European context, most of the evidence on the impact of large complex care home chains comes from the United Kingdom (Burns et al., 2016; Corlet Walker et al., 2021; Kotecha, 2019). Such research has focused on the role of the biggest companies, which are typically owned by PE funds, and revealed that they take advantage of complex organization structures to engage in leakage practices that hide profit extraction and reduce service quality, as observed in US large complex chains. Harrington et al. (2017) and Bourgeron et al. (2021) have shown that these practices are also common among the largest PE firms that own care homes in Norway, Sweden, France, and Germany. Despite the efforts of previous studies, their results have two limitations. First, they do not compare the behavior of complex organizations with other types of entities in this sector, and second, they access financial information but do not include any social outcome in their studies. To clarify the distinction between large complex chains and other organizations operating in the care home sector, Table 1 presents a summary of the main differences.

Based on this comparison, we can observe that while other organizations in the sector have much simpler structures, typically based on direct ownership and local operations, large complex chains are characterized by multi-layered ownership, cross-border presence, and a higher likelihood of engaging in financial practices that reduce transparency and accountability.

In the specific case of Spain, we have not found any study that analyses the relationship between the complexity of the ownership structure and the levels of economic and social performance in the care home sector. Economic performance refers to the organization's capacity to generate financial returns and maintain operational viability through indicators such as return on assets (ROA), profit margins, and financial sustainability. Social performance is defined as the organization's contribution to meeting stakeholder expectations in the provision of socially valuable outcomes, such as quality of care, user well-being, and safety (López-Arceiz et al., 2018). The only signal of a potential negative interaction between complexity and both types of performance was published by Spanish newspapers during the COVID-19 pandemic ^[5]. They criticized the role of PE funds in the improper management of the care homes owned by these funds. In this context, Rico (2021) emphasized that major care home chains support a rationale that places profit at the forefront, although ECHO (2023) ^[6] defends that care homes prioritize people over profit by reinvesting most of their profits in the provision of their services and thus creating social and economic value. There is, then, a controversy on the actual role of large complex chains in this context. International evidence provides useful insights, Harrington et al. (2017) analyze the rise of investor-owned care chains in the US, while Burns et al. (2016) examine the governance challenges arising from fragmented and multi-entity structures in European health and social care services. These studies support the claim that large complex chains can have implications for care quality, public accountability, and financial transparency—especially in sectors heavily reliant on public funding. However, no quantitative analysis has been carried out to shed light on this debate. To address the effect of the complex organizational structure on the economic and social performance, we propose the following working hypotheses:

As we have previously explained, European countries, especially those in the Mediterranean area, have a high public coverage of care home services. Governments in these countries dedicate part of the public budget to financing access for vulnerable people to care home services. As most of these subsidized beds are in fact managed by private companies, these resources also contribute to the financial sustainability of private companies in this sector. Public funding thus has the potential to influence the behavior of such firms and affect both social and economic outcomes in the care home sector. This dependence on public funding could interact with the complexity of these chains. As they attract considerable public attention, these large complex chains could have more incentives to distort their actual economic performance by avoiding abnormal profits regarding other companies in the same sector or even by reporting losses to claim a higher public subsidy.

Although not specifically in this sector, there is extensive research that supports the suggestion that firms attempt to mitigate the risk of adverse political actions through downward earnings management to conceal their real economic performance (Cahan, 1992; Chen et al., 2011; Han and Wang, 1998; Milne, 2002) ^[7]. Closer to the care home sector, Papas et al. (2023) extended the previous evidence to companies in the healthcare sector for women affected by domestic violence. They also concluded that subsidized companies may try to avoid political backlash by engaging in management practices to avoid reporting abnormal profits. Not empirically but directly related to the care home sector, Harrington et al. (2017) suggested that large complex chains heavily dependent on public subsidies are expected to underreport profits by overstating management fees, lease agreements, interest payments to owners, and purchases from related party companies to a greater extent. An illustrative case in Spain is that of Vitalia Plus SA. According to the 2020 annual report, Geralia Home SL, a related-party company, provided maintenance and rental services to Vitalia Plus SA totalling €2.3 million. At the group level, the total amount invoiced exceeded €5 million, which represents the total revenue in Geralia Home SL's income statement. This could indicate a potential diversion of financial resources within the group. In sum, while empirical research specifically addressing this economic phenomenon in the care home sector is almost non-existent, it has been suggested that large complex chains in this sector are receiving public grants that could affect the level of the economic performance that they release in annual reports. Moreover, the dependence on public funds opens a new possible explanation for the leakage practice different from reputational effects that links to the prices paid for public beds.

An additional empirical question is whether the level of public funding received by these complex organizations conditions the quality of their service. Arrow (1978), Hansmann (1980), and Hart et al. (1997) suggest that moral hazard issues could weaken the natural ability of the market to align firm incentives with service quality. In the North American care home sector, Grabowski and Town (2011) and McGregor et al. (2005) found that public funding could be used for purposes other than providing real care service quality. The rationale behind these complex organizations is that public resources would be diverted to alternatives not directly related to the quality of their services. Gupta et al. (2024) provide empirical evidence for how care home entities owned by PE funds exhibit discernibly lower quality ratings when they receive more public subsidies from the Medicare system ^[8]. This substantiates the idea that an increase in public funds, rather than concurrently enhancing the tangible quality of care, is paradoxically associated with a decline in the quality of the services provided. On the other hand, it could be posited that the inherently aggressive nature of these large complex chains prevails regardless of the source of their funds, in which case a public funding interaction would not show any significant difference in the quality of their service.

The abovementioned studies limit their analysis to large complex chains, so they are unable to test, in the way we are, whether these organizations act differently from the other organizations in this sector when they receive equivalent public subsidies. To test the interaction effect of public funding on the economic and social performance of large complex chains, we postulate two additional hypotheses:

The Spanish care home sector constitutes a suitable case of analysis due not only to the presence of large complex chains, but also to the dependence on public funding. The Spanish welfare system regarding care homes is integrated in the European care strategy of the European Pillar of Social Rights (European Commission, 2022b). This strategy focuses on the person as the main user. This person-centered service can be provided by different initiatives (European Commission, 2022b), and care homes are highlighted as the main service provider among those initiatives, covering approximately two-thirds of the employment created in this sector (Barslund et al., 2021).

As an exemplar of how care homes are commonly organized in European countries in the Mediterranean model (Genet et al., 2012) ^[9], the Spanish care home sector is characterized by a high degree of privatization, strong dependency on public funding, and increasing penetration by large complex chains. These features make Spain not only representative of a broader Southern European trend but also particularly relevant for understanding the implications of large complex chains in publicly funded and privately delivered elder care. According to Abellán et al. (2021), about 73% of the total beds in Spain are provided by private entities, which adopt diverse forms ranging from nonprofit organizations to cooperatives and commercial companies. The European Commission (2022a) asserted that 58% of the total beds are offered by private, for-profit providers in this country. In recent years, and similar to other countries, a reduced group of large complex chains have emerged in this specific context, which has concentrated an important percentage of the Spanish care home sector (CBRE Group, 2018). These large complex chains account for approximately 20% of the total beds and 34% of the sector's total revenue. However, what actually defines the Spanish sector is not only the existence of large complex chains but also its interaction with a singular scheme of funding structures (Fourati, 2021; Iannotta et al., 2007).

In terms of ownership structure, Rico (2021) found that the main shareholders of these large complex chains in Spain are PE funds located in tax havens, investment funds, pension funds, insurance companies, and some multimillionaires or business entrepreneurs of dubious reputations ^[10]. In all cases, this singular structure involves multiple legal entities that form part of large corporate groups operating across several regions (Investigate Europe, 2021). However, the Spanish care home sector combines the presence of large complex chains with a second characteristic, which is the high dependence on public funding as a result of the process of decentralization that has taken place in this country. According to IMSERSO (2021), 63% of the total number of beds are publicly funded (public beds plus concerted beds). In this context, Rico (2021) reported that public administrations allocated 1,895 billion euros to the private sector, an amount that increased annually. Most of these funds went to large complex chains. Unlike other countries, public funds can have different origins in Spain, ranging from national and regional public administrations to public contracts associated with public–private collaboration in the management of care homes (Costa-Font et al., 2023; European Commission, 2018). As a result, there is a complex landscape of entities co-existing in Spain. The table in  Appendix 1 show a summary of different entities working in this sector by distinguishing between provider types and structural features. This table cross-tabulates service provider types (public, private), funding sources (private, public), and the presence of large complex chains (yes, no). It makes possible to appreciate the different entities living together in the Spanish sector. In general terms, these entities are committed towards the creation of social and economic value, providing high-quality services.

Large complex chains have also been under public scrutiny and criticism in Spain regarding the quality of their services. Although public authorities generally require care home providers to report quality indicators, public access to this information remains limited (Ariño Blasco et al., 2014; Rodrigues et al., 2014). In the absence of comprehensive data, journalistic investigations have documented poor working conditions and substandard food in facilities managed by these companies (El País, 2021; infoLibre, 2023). The COVID-19 pandemic further revealed systemic weaknesses, including inadequate crisis management and structural deficiencies in care homes, as reported by the Spanish Ombudsman and Amnesty International (Defensor del Pueblo, 2020; Amnesty International, 2020). These shortcomings, along with legal proceedings for negligence and mistreatment (elDiario.es, 2022), raise serious concerns about the compatibility of for-profit care models with the high-quality, person-centered care envisioned by the European Care Strategy. In sum, the Spanish care home sector combines the presence of public and private service providers, public funding dependency and the penetration of large complex chains. These three elements ensure its adequacy to test the proposed hypotheses.

The context of our study focuses on Spain as an illustration of a European Mediterranean country. There are 5,580 care homes for the elderly in Spain, with a total of 386,004 beds in 2020. About 73% of the total beds were provided by private entities, while only 27% were in public care homes (Abellán et al., 2021). Of the total number of private beds, 65% are purely private and 35% are subsidized-concerted beds (Observatorio Sectorial, 2022).

Three organizational levels can be identified in the Spanish care home sector. The lowest level includes all care homes that are in direct contact with the elderly (Level 3). Every care home depends on a superior legal entity, which usually adopts the form of a commercial company (Level 2). Finally, in some cases, these legal entities belong to the same group, so it is possible to identify a parent company (Level 1). This top level is, in all cases, restricted to parent companies domiciled in Spain. Sometimes these three levels give rise to complex structures involving companies in related sectors and extending to networks of companies in other countries above the Spanish parent company. Many of these structures are, however, simplified to the extent that a single care home belongs to a single company at Level 2 and there is no parent company at Level 1.

We created our sample using the SABI ^[11] database in the following way. We started by identifying companies at Level 2 whose industry classification code (four-digit CNAE code ^[12]) indicates that they belong to the care home sector. Next, we analyzed their ownership and identified the Spanish parent company (Level 1). We then tried to determine the complexity of their corporate structure by validating the conditions for classification as large complex chain. First, the presence of the three levels; second, a significant number of commercial entities and care homes per parent company ^[13], and finally, the type of ultimate owner ^[14]. Using this method, we identified 13 corporate structures that align with the classification criteria of Rico (2020a). He investigated the ultimate beneficiaries of public funds for elderly care in Spain and, after analyzing the 48 largest providers, he identified 13 major groups that each manages a minimum of 15 care homes. His findings indicate that these large groups are predominantly controlled by PE funds, billionaires, and, in some cases, individuals with controversial business backgrounds. We categorized the remaining organizations within the Spanish care home sector as the general group. To include them in the final sample, we had to have access to their financial information in the SABI database. We found such information for all the Level-2 companies in the 13 large complex chains (47 companies) and for 1,611 of the Level-2 companies in the general group. The final sample thus contains 1,658 companies operating at Level 2 in Spain from 2014 to 2021. Table 2 presents the composition of the sample and the main characteristics of the organizations at the three levels.

The 13 large complex chains include 47 Level-2 companies operating in the sector by managing 528 care homes. These care homes offer more than 20% of the total beds and represent approximately 40% of the sector's total revenue and assets. We accessed a sample of 305 care homes belonging to complex structures. Both the ratio of commercial entities in the sector and care homes by the parent company are higher in the large complex chains group (3.62 and 23.46, respectively) than in the general group (1.12 and 1.56, respectively). While in the large complex chains group a parent company owns many commercial companies and each of these runs also a number of care homes, in most cases, the general group limits itself to one entity at each of the three levels. The average size of each care home in the large complex chains group (140 beds) is more than twice the corresponding size in the general group (69 beds). Finally, the percentage of concerted (subsidized) beds over total beds in the large complex chains group is 35% (15,088/42,949), whereas in the general group it is only 27%—that is, 7% points lower.

To assess social performance, COVID-19 crisis served as a powerful stress test for the Spanish care home system. During the first wave of the pandemic, public authorities were compelled to disclose mortality data that had previously been unavailable due to a lack of transparency or fragmented reporting linked to the administrative design implemented in Spain. This unprecedented release of data has enabled us to construct a robust proxy for social performance, grounded in observable and comparable outcomes across providers ^[15]. We link this mortality data (available at level 3) to the financial and structural information (level 2) to build a unique dataset. This database provides an essential baseline for evaluating the capacity and behavior of care home organizations. In this sense, the death rate, normalized by number of beds, allows us to assess differences in outcomes that may reflect structural disparities in resource allocation, staffing, and investment priorities, especially when comparing large complex chains with other organizations in the sector. This contribution is particularly relevant given the limited availability of quality-related indicators in the Spanish context. By incorporating this singular dataset, our study offers not only an original measure of social performance but also an empirical contribution that can inform both academic research and public policy in evaluating provider effectiveness in times of crisis.

Economic performance involves engaging in stable and continuous economic activities (European Commission, 2020c). To approach this type of performance in care homes, the literature and annual reviews use financial indicators such as net income margin, EBITDA margin, and operating margin (Harrington et al., 2024; Knight Frank, 2022; O'Neill et al., 2003). We chose two profitability variables, economic return and net income margin. Return on assets (ROA) compares the profit before interest and taxes to its assets (Ross et al., 2010). The net income margin quantifies the percentage of sales that the company converts into net income (Leahy, 2011). As Ortas and Moneva (2011) and Cochran and Wood (2017) mention, these ratios provide accounting-based indicators that reflect internal efficiency and, consequently, the stability of the economic activity.

Social performance considers the identification of an explicit social objective with the aim of benefiting specific interest groups or society as a whole (European Commission, 2020b). It implies the creation of a mission-related impact for broader stakeholders (Nicholls, 2006). The primary objective of care homes is to provide high-quality services to elderly individuals and improve their life expectancy. However, measures of social outcome are difficult to obtain (Bauer and Johnston, 2020; Van Slyke, 2007). Comondore et al. (2009), Winblad et al. (2017), and Broms et al. (2020) adopted Donabedian's (1988) input-process-outcome framework, which considers five indicators: staff density, nurse density, staff education, share of residents with an updated action plan, and resident satisfaction. Similarly, Patwardhan et al. (2022) employed a rating that summarizes quality across five domains: safe, effective, caring, responsive, and well led. Figueroa et al. (2020) and Gupta et al. (2024) linked care home quality with resident deaths. Unfortunately, most of these data are not publicly available in the Spanish context. In our study, we used the care home death rate as a proxy for social performance. Baker et al. (2020) highlighted that during the COVID-19 pandemic, deficiencies in care homes could potentially lead to an increase in the number of deaths. Traditionally, Spanish public authorities have been reluctant to disclose information about deaths in this sector; however, the extraordinary situation caused by the pandemic forced authorities to make this information available, which allows us to use it as a proxy for the social outcome. We have used the available data corresponding to the first wave of COVID-19, the period with the greatest impact (Barrera-Algarín et al., 2021). To facilitate comparison between the analyzed groups and mitigate the impact of size differences (Marcuello and Salas, 2001), the number of deaths was normalized based on the number of beds per care home. This information was obtained from various sources, such as media reports and the transparency portal ^[16] specific to each regional government.

As mentioned, some entities in the care home sector are large complex chains, characterized multi-layered ownership networks with a singular type of ultimate owner, involving cross-border entities, financial intermediation, and potentially aggressive tax and profit-shifting strategies. A good illustration of these large complex chains is the case of Vitalia Home, one of the four largest chains in Spain (Figure 1). Vitalia Plus SA is the parent company that consolidates financial statements in Spain (Level 1). Vitalia Home is dominated by CVC Capital Partners, a British PE fund. Above the Spanish parent company, there are 15 companies spread across different layers and locations (Spain, the Netherlands, Luxembourg, and Jersey, a tax haven). Below Vitalia Plus SA, there are four subsidiaries in Spain (Level 2). One of them operates solely as a holding company that has two additional companies under its control. All these subsidiaries collectively manage 66 care homes (Level 3). In contrast, many other organizations in the Spanish care home sector have much simpler structures, typically based on direct ownership, usually structured around a single legal entity or a straightforward holding structure, often operating within a single jurisdiction. For example, Residencias Asistenciales Niño Jesús SL is a single legal company headquartered in Spain (Level 2) with no parent company (Level 1), directly managing nine care homes (Level 3).

Large complex chains also rely more heavily on public subsidies and have an incentive to divert resources through intra-group transactions (leakage) to conceal their real economic performance (Corlet Walker et al., 2021; Harrington et al., 2015; Kotecha, 2019). In the case of Vitalia Plus SA, this diversion of resources can be observed in the company Geralia Home SL, which is incorporated in Spain, but Vitalia does not include it in the consolidation of the group ^[17]. In contrast, no such intercompany arrangements were identified in the case of Residencias Asistenciales Niño Jesús SL, suggesting a more transparent ownership structure and internal financial flows.

The dummy variable COMP is used to represent ownership structure; a value of 1 is given to the 13 large complex chains and 0 to the general group ^[18]. The variable PUBF is used to represent public funding, which is measured by the proportion of subsidized beds out of the total number of beds. To explore the joint effect of public funding and ownership structure, the interaction variable COMPxPUBF is incorporated into the analysis.

Five control variables that affect the performance of care homes were considered: the COVID-19 effect, financial leverage, solvency level, penalties per care home, and location.

The COVID-19 effect was introduced as a binary variable to differentiate between pre-pandemic years and the years during which the pandemic occurred. The variable was assigned a value of 1 tor the years 2020–2021, and 0 for the other years. According to the Spanish National Research Council, among developed countries, Spain was severely affected by the virus (Pino et al., 2020). We only introduce this variable to control the economic performance dimension.

Financial leverage signifies the proportion of debt a company holds in relation to its net equity. Excessive financial leverage could negatively affect economic performance and could also be a symptom of difficulties in meeting debt payments. In this context, managers in highly leveraged firms could have an incentive to reduce service quality. In line with this, Al-Shattarat (2024) demonstrates that leverage is positively correlated with accrual-based and real earnings management. This highlights how debt pressure can encourage opportunistic behaviour. On the other hand, a significant percentage of debt could increase profitability ratios through financial leverage gains (Rico, 2021).

The solvency level represents the extent to which total assets would cover the total debt of the company. This ratio measures the capacity of a firm to face its liabilities in the case of a hypothetical bankruptcy. Companies with high solvency levels are expected to have better economic and social performance.

Penalties per care home reflect the average number of sanctions imposed on care homes by region. These penalties are established by the Dependency Law, which enforces disciplinary measures in instances of non-compliance with the regulations that govern the operation of care homes, with the aim of safeguarding the well-being and safety of elderly residents. Nevertheless, as highlighted by Rico (2021), the enforcement of these prescribed sanctions lacks consistency, and it is possible that the cost savings of non-compliance could be higher than the penalties imposed. The direction of the impact on economic performance is thus not clear. For social performance, a clear negative relationship is expected.

Finally, the location variable has been included to indicate the geographical placement of the care home. Different types of care homes exhibit variations in location, with private facilities being more common in affluent urban areas where demand is higher (Molinuevo et al., 2017). Managing larger care homes could positively affect economic performance because it could be easier to implement techniques related to the standardization of services and knowledge transfer. However, as shown in most official statistics, care homes located in urban areas suffer a higher mortality rate than those located in rural ones (IMSERSO, 2020). These data show that some urban areas, particularly densely populated regions such as Madrid, recorded significantly higher mortality rates in residential care settings than their rural counterparts. We codified urban areas with a value of 1, while rural areas have been assigned a value of 0.

 Appendix 2 provides more details about the main variables, indicators, sources, and measurement scales used.

First, we analyzed the descriptive statistics for the studied variables. This analysis distinguished between the groups formed by complex structures and public funding. Subsequently, we performed a Student's t-test to assess the equality of means between the complex structure and the general groups. The variables considered in this analysis included the economic return (ROA), net income margin (net income margin), and the number of deceased users (death rate).

To test the proposed working hypotheses, we specified a set of linear regression models. These models examine the relationship between the endogenous variables (economic and social performance) and the exogenous variables related to the complex structure and public funding. The linear models were specified using the following equations and the control variables [1–6]:

The dependent variables ROA_it and net income margin_it represent the economic performance while the death rate_it reflects the social performance. The independent variables COMP_i, PUBF_it, and COMP_i*PUBF_it represent the complex structure, the public funding, and the mixed effect, respectively. The following control variables were used: the COVID-19 pandemic, financial leverage, solvency level, penalties, and location. β_i represents the regression parameters, and ε_it is the error term. The coefficients that relate the complex structure to the economic and social performance variables test Hypotheses 1 and 2, respectively. The coefficient for the COMP_i*PUBF_it interaction shows the moderating effect of the public funding represented by Hypotheses 3 and 4. The different models were estimated using the ordinary least squares method. This analysis enabled us to test the proposed hypotheses.

Table 3 presents the descriptive statistics of the sample organized in two panels. Each panel considers two groups with their mean values and standard deviations for financial statement variables and ratios. Panel A considers the complex structure group (COMP, G1) and the general group (GRAL, G2). Panel B presents the data from the public funding group (PUBF, G3) and the one formed by non-PUBF (G4). Panel C shows the one formed by the interaction variable (COMP*PUBF, G5) and the GRAL*PUBF interaction (G6). The last column of all panels indicates the parametric t-mean tests. Panel A shows the mean contrasts between COMP and GRAL, Panel B between PUBF and non-PUBF and Panel C between COMP*PUBF and GRAL*PUBF.

The two columns in Panels A (G1 and G2) represent COMP and GRAL. The results show that companies of the COMP group are generally larger than those in the general group in terms of assets and employees. Sales levels also reflect this difference. Despite investing more in assets (248.73 million euros vs 17.13 million euros), the COMP group has a lower proportion of tangible assets (44% vs 48%), which suggests the potential presence of rent of its infrastructure. They also have a higher proportion of intangible assets (19% vs 9%), possibly due to acquisitions and their complex ownership structures. Both groups show similar levels of financial leverage (4.65 vs 4.54), indicating no substantial difference in debt reliance. Interestingly, despite their larger size, companies in the COMP group report lower average salaries per employee (€20,437 vs €28,422). This suggests that large complex chains allocate fewer resources per care worker, which may negatively affect the quality of care, given the central role of professional staff in delivering this kind of human-dependent services. On average, the COMP group has lower economic returns and net income margins than the GRAL group (ROA: 6.55% vs 7.46%; net income margin: 4.74% vs 5.07%). The mortality rate (death ratio) is, however, higher in the COMP group (7.54% vs 5.67%).

Panel B includes PUBF and non-PUBF groups (G3 and G4). In terms of assets, employees, and revenue levels, companies that receive public funding (G3) are generally larger than those without such funding (G4). The PUBF group shows lower financial leverage and lower average salaries per employee. The economic return and the mortality rate are higher in the PUBF group, while the net income margin is lower. Analyzing groups by the interaction variables (Panel C), companies in the COMP*PUBF group (G5) are generally larger than those in the GRAL*PUBF one (G6) in terms of assets, salaries, and revenue. The COMP*PUBF group shows higher financial leverage but lower average salaries per employees. It also has a lower economic return and net income margin (6.28% and 4.06%, respectively) and a higher mortality rate (7.93%) than GRAL*PUBF (8.16%, 5.62% and 6.52%, respectively). These findings suggest statistically significant differences exist for all variables. However, it is important to interpret them with caution due to the observed standard deviations in most of the studied variables.

Table 4 presents the results of the linear regression models used to analyze the relationships between the variables. Two regressions were estimated for each of the two economic performance proxies, economic return (ROA, columns 1,2) and net income margin (columns 3,4) and for the social performance proxy, death rate (columns 5,6). The table includes information on the main variables, control variables, and goodness-of-fit measures.

Columns 1 and 3 present regressions for the two proxies of economic performance. Both are significantly lower for the COMP group (β: −0.0223; p < 0.01) and (β: −0.0134; p < 0.01) for ROA and net income margin, respectively. This suggests that the COMP group shows significantly lower economic performance than the GRAL group which means that we cannot reject Hypothesis 1. Column 5 show results for social performance. Consistent with Hypothesis 2, the death rate in the COMP group is significantly higher than in the GRAL group (β: 0.0148; p < 0.01).

After analyzing the effect of being part of the COMP group, Hypotheses 3 and 4 test the combined effect of the use of public funding within this COMP group. The combined effect of COMP*PUBF on the economic performance is presented in columns 2 and 4. The interaction effects on both ROA (β: −0.0712; p < 0.01) and net income margin (β: −0.1463; p < 0.01) are significantly negative. Consequently, we cannot reject hypothesis 3, that is, public funding implies a significant decrease in the economic performance of large complex chains. A different pattern is observed for the interaction effect on social performance, that is, the number of deaths for complex companies receiving public funds is not significantly different from the other organizations in the sector (β: −0.0208; p > 0.10). In this case, we do reject hypotheses 4. Most of the control variables show the expected signs.

To further explain the above results, we explored the extent to which they could be related to leakage practices, as suggested in the literature (Bourgeron et al., 2021; Corlet Walker et al., 2021). The main objective behind these practices is to overstate certain expenses to signal a lower benefit, thus explaining their lower profitability. This lower benefit could be the signal exploited to justify or demand an increase in the public funding received. Following previous studies, companies commonly overstate their reported fee, rental and financial expenses, among others, to hide their profits. However, not all expenses are expected to be overstated. In fact, we consider it necessary to analyze two groups of expenses separately: those that are strictly necessary to provide a minimum level of social performance (e.g. personnel expenses) and those less relevant for that purpose (e.g. rent expenses). We first assumed that companies in the COMP group would have an incentive to reduce the first group ^[19]. On the other hand, actual leakage would make them artificially increase items in the second group that, although not essential, will be recognized for amounts clearly above the standard levels in the sector. We took personnel expenses as a proxy for the first group, while other operating and financial expenses represent the second group. All these variables are calculated as a percentage of sales.

Table 5 presents the regressions using each of the three proxies as dependent variables. Similar to the regressions in the previous section, we introduced COMP and PUBF, and their interaction as independent variables and the necessary controls.

Column 1 shows that large complex chains dedicate significantly fewer resources to personnel (−0.0728, p < 0.01) but, at the same time, they show higher levels of financial expenses (0.0055, p < 0.01 in column 5). This reveals both cost-cutting practices and the existence of leakage. The sign for other operating expenses is unexpectedly negative (−0.0123, p < 0.01 in column 3), which initially implies that leakage does not exist. Other operating expenses are, in fact, a controversial item, as they might combine expenses with opposite expected signs. Some of them would be necessary for the quality of care home services (utilities, insurance, catering services), and companies in the COMP group would tend to minimize them where possible. Some others are more prone to leakage (rent, fees, management services) with an expected positive sign. As the net effect is negative in column 3, the understatement of obligatory (necessary) expenses would be having a higher impact than the practice of leakage in this case. These patterns remain consistent after including the interaction with public funding (columns 2, 4, and 6). Large complex chains that receive public funding dedicate fewer resources to personnel and operating needs, while maintaining higher financial expenses. This reinforces the interpretation that these providers follow a cost-minimization strategy, regardless of their funding source.

A robustness analysis was conducted by limiting the variable COMP to the eight structures referred to as the giants of the for-profit sector (Rico, 2021). Following this author, each of these companies manages more than 40 care homes, while no other entities in the COMP group manage more than 25 homes. In 2020, the eight large complex chains managed 44,112 beds (17.66% of the total). These features mark a significant difference between these eight firms and the rest. Focusing on this group provides confirmatory evidence of the different behaviour of complex structures. Table 6 presents the results of the robustness regressions considering the COMP-8 group.

The findings are consistent with the initial results. In terms of economic performance, column 1 confirms a significantly lower effect on economic return (ROA) for the COMP-8 group. For the net income margin, the results are similar, as the effect is also significantly lower for COMP-8 group (column 3). For social performance (column 5), the COMP-8 group maintains a significantly higher mortality rate, again consistent with the previous results. The interaction effects (columns 2, 4 and 6) also maintain the evidence found in Table 4, significant for economic performance and a lack of it for social performance, providing robustness to our results.

As a second analysis, we defined new proxy variables for both the economic and social dimensions. In the first case we tested the operating margin instead of the net income margin. This new proxy, calculated as operating income/sales, measures the efficiency of care home companies while disregarding the effects of both total assets and debt levels, which might be overstated in COMP cases. As for the social dimension, we divided the number of deaths per occupied instead of per available bed. Some users abandoned care homes during COVID-19, which justifies this new proxy. This assessment acknowledges the potential impact of prioritizing occupancy, especially during the early months of the COVID-19 pandemic.

Table 7 shows the robustness analysis with these new proxies. In the case of the operating margin (column 1) we observe that it maintains the lower sign for the COMP group. This result is consistent with the one obtained in the original analysis. For social performance (column 3), the COMP group confirms the previous results when the number of occupied beds is considered instead of the number of available beds. Columns 2 and 4 also confirm the conclusions regarding the COMP*PUBF interaction term.

As a final robustness analysis, Table 8 replicates the same regressions as in Table 3 using two additional estimators (robust OLS and cluster ID). Robust OLS uses White-corrected standard errors to address heteroskedasticity, while cluster estimator accounts for potential autocorrelation. In both cases, the parameters maintain the same signs and levels of significance as in the initial OLS regression in Table 4, which gives robustness to the analysis.

In general, the different robustness analyses confirmed our previous conclusions about our proposed hypotheses. This supports the notion that both economic and social performance are influenced by the complexity of the corporate structure and that the use of public funding moderates the economic performance but does not affect the social one.

This study has analyzed the combined effect of complex structures and public funding on both the economic and social outcomes of care homes in Spain. We have provided empirical evidence that large complex chains show lower economic performance than the other organizations in the sector and that the difference is enhanced when they receive public funds. Social performance is also significantly lower for large complex chains but in this case, public funding does not have a moderating effect. To further explain these results, we have also examined the trade-off between obligatory (necessary) and discretionary (unnecessary) costs and found that large complex chains cut the former but overstate the latter, mainly when they receive public funds. This means that large complex chains are more concerned on justifying the demand for public funding than on their social reputation. These chains do not appear to use financial opacity to shield themselves from public criticism but to construct a narrative of financial necessity aimed at securing additional resources from the government. The fact that public funding does not moderate the relation with social performance would mean that these chains invest less than necessary in the service quality regardless of the (private/public) source of their funds. Our findings, then, call into question the alignment that these firms claim to achieve between economic and social rationales in carrying out their activities.

Previous studies that analyzed the role of ownership structure in care homes in the North American context had shown that large complex chains pursue profit maximization strategies for which they underinvest in aspects that are essential for the quality of service, such as the level of direct care and support staff (Kingsley and Harrington, 2022; Stevenson et al., 2013). As they do not want to explicitly exhibit higher than the average profits, they take advantage of the complexity of their structures to spread and hide those profits by overweighing expenses whose beneficiary is not the user (Gupta et al., 2024; Harrington et al., 2017). These leakage practices have also been followed in the European context. Burns et al. (2016), Kotecha (2019), Bourgeon et al. (2021), and Corlet Walker et al. (2021) have shown how the most typical shareholder behind these large complex chains, PE firms, generates tidy returns for itself and its investors through predatory financial practices, such as monitoring fees, interests, and lease payments. Our results have confirmed partially these ideas for the Spanish context. The ownership structure associated with care homes affects both the lack of necessary investment and the practice of leakage, with a higher impact of the latter on the lower economic returns. We have also demonstrated that the COMP group suffered from a higher number of deaths during the COVID-19 pandemic.

In countries like Spain, the care home sector is not only characterized by an increasing penetration by large complex care chains but also by the strong dependency on public funding. Therefore, a relevant question in our study has been the extent to which the rationales behind care homes are affected by the use of public funding. So far, no study had previously shown the impact of public funding on economic and social performance. Initially, public funding for the subsidized beds represents a level of assured income that should lead to greater operating margins and final economic returns. On the contrary, it seems to be reinforcing the incentives to signal low profitability, thereby justifying the need for a higher price for those beds. Our study has shown that the problem is not only the lack of public funding in this sector, but the potential misuse of these resources by the COMP group.

In the Spanish context, public funding has increased from 567 million euros (2008) to approximately 1,857 million euros (2022; Costa-Font et al., 2023). The allocation of public funds is, however, not proportional, as care homes that are part of the COMP group obtain a significantly larger amount than the rest of organizations in the sector ^[20]. This circumstance seems to be enhancing the incentives for leakage practices in large complex chains. It is important to note, however, that the results obtained in this study do not represent a general characteristic of the Spanish care home sector, as some companies demonstrate more efficient resource management and a stronger commitment to providing quality services aligned with their social objective (Lares, 2022). Our findings suggest that care homes that are not part of the COMP group can attain greater efficiency and quality in service provision. The problem is that both types of care homes co-exist in this sector but compete in unbalanced economic terms to provide services of similar quality for their users. This makes us wonder which is the actual role of the COMP group in the Spanish care home sector.

The importance of our study is that it contributes to the existing body of knowledge by shedding light on contexts where public funding is a relevant aspect. This is the first study that jointly examines how the economic and social objectives of care homes intersect and potentially conflict within the large complex chains that receive public funding, thus making the profit/quality of service alignment a challenge. Practitioners can also benefit from the study's insights by gaining a better understanding of the impact of the provision of these services on their quality. Additionally, our results could aid policymakers, investors, and stakeholders in identifying those care homes that genuinely prioritize social objectives without hindering their economic returns. Regulatory bodies and public authorities should reinforce oversight mechanisms to ensure that care homes, particularly those operated by large complex chains, adhere not only to quality standards but also to their social objectives. These chains often exploit the opacity of their ownership structures to hide their actual economic performance and the quality of care they provide. By reporting consolidated group-level accounts, they can mask the true financial situation of individual care facilities. Such accounts frequently include non-essential expenditures, such as inflated interest payments resulting from high levels of intra-group debt, which can significantly reduce reported profits. In some cases, these financial flows may ultimately benefit the owners, who often act as both shareholders and lenders. In line with this, Álvarez-Botas et al. (2024) reveal that large controlling shareholders expropriate wealth from other investors.

Such practices hinder public authorities from properly evaluating whether public funds are being used efficiently and in alignment with policy objectives. To address this, policymakers should implement robust accountability measures to ensure transparency and prevent the misuse or diversion of public funding. Without such safeguards, large chains would continue to exploit their size and organizational complexity to evade accountability, thereby undermining the ability of public administrations to enforce effective oversight.

This study is not without limitations. Obtaining data at the three levels that co-exist in this sector (care homes, commercial companies, and parent companies) entails an intricate challenge. Although we have carefully matched the information across these levels, this structure may introduce some aggregation bias. This underscores the need for greater transparency and availability of information. Authorities should support the availability of data at the care-home level, given the sector's social significance. Moreover, the reliability of the information is dependent on the quality controls implemented by public administrations. Finally, we also note the absence of non-financial information on aspects such as specific staffing numbers, the health status of the elderly as users of care homes, and the available equipment to provide the service. Future lines of research would open up if these data limitations were overcome.

We are particularly grateful to Fermín Lizarraga Dallo for his extensive comments, insights and suggestions throughout the development of this paper. We also thank Francisco Javier Husillos Carqués for his helpful comments and support.