Market characteristics and entry strategy decision making: the market perspective of Croatian elderly care homes

Background : Socio-demographic changes increase the need for long-term elderly care. Consequently, providing formal institutional service in elderly care homes is an interesting opportunity for entrepreneurs. However, the entry strategy decision is influenced by numerous external variables. Purpose: The main goal is to answer what determines market concentration as one of the most important market entry determinants. Study design/methodology/approach : A linear regression model has been formed and tested on the Croatian elderly care home market, observed on a county level, using data for 2021. Further, a cluster analysis, as a decision-support tool, has been made to assess market characteristics that are more likely to attract new entrants to the elderly care home market. Findings/conclusions: Results indicate that demand for long-term care services plays a significant role, and the market with more elderly will attract more competitors. When the level of GDP per capita and the unemployment rate are observed together, markets with stronger economies tend to attract entrepreneurs. In other words, it is more likely that someone will open an elderly care home in a densely populated county with individuals that can afford formal institutional long-term care for themselves or family members. Limitations/future research: The shortcomings are mainly related to the lack of data on prices and quality measures. Further, information on the number of beds in each elderly care home would enable an alternative calculation of the Herfindahl-Hirschman index, while data on service prices and structure of employees as a proxy for quality (medical and non-medical staff) would enable a more reliable comparative analysis of obtained results. Future studies on this subject include variables related to the portion of unemployed females in the market since female family members more often provide informal care, and at the same time, they are more likely to be employed in formal long-term care institutions.


Introduction
Ageing society, changes in family structure in the sense that the portion of multigenerational families decreases, and expanded life expectancy provide numerous business opportunities in the context of the silver economy. When we observe their nature, many elderly-adjusted products and services belong to the group of normal, i.e. luxury goods such as various hospitality industry products and services. However, in some cases, elderly consumers are compelled to use certain products/services, such as in the long-term care (LTC) industry. Therefore, the aforementioned socio-demographic changes and the nature of the LTC increase demand for LTC products/services (both formal/informal and institutional/noninstitutional). Consequently, providing formal institutional LTC services in elderly care homes might be an interesting opportunity for investors.

O N L I N E F I R S T
However, even though existing demand is a necessary precondition, numerous internal and external factors determine the entrepreneurial decision to enter a specific market.
When it comes to the formal institutional LTC sector as an entrepreneurial opportunity, external determinants are rather specific. Users of the service are elderly individuals, often with physical and mental health issues, which makes this business both labor-intensive and specific in requiring more or less trained medical staff. Further, since the nature of the service requires residential premises, high capital investments are necessary if the company needs to rent or build them. Also, in many cases, along with consumers, family members and/or government subsidies cover the cost of the service. If they cannot cover it, families often provide informal noninstitutional care or pay for formal noninstitutional care for their elderly family members. Therefore, the decision on whether and where to open an elderly care home is under the influence of several complex external and internal determinants, ranging from level of demand, consumers' ability to pay for the service, choice consumers have in terms of substitutes for the service, to availability of suitable employees.
In that sense, this paper is focused on determinants of the formal institutional LTC market structure, i.e. market concentration as an external choice element, to provide a decision support tool while deciding whether to enter the respective market in the context described in Figure 1. A market concentration determining market structure is considered one of the most important elements of the decision whether to enter the market. Namely, it is reasonable to expect that market structure, i.e. market concentration, is correlated to the company's profitability. However, the nature of their correlation and causality is a source of debate. On the one hand, according to the Harvard school of economic thought, a company will strive to increase its market share to increase its market power manifested via more control over prices resulting in higher profits (for detailed insight on market concentration as a determinant of profitability cf. Pervan, Pervan, & Ćurak, 2019). In that setting, a higher market concentration would imply a lower level of competition, which might attract new entrants. On the other hand, lower market concentration resulting from higher competition might indicate demand sufficient to accommodate additional entrants to the market. Therefore, it is reasonable to expect market concentration to influence the new entrant's decision. Still, the outcome will result from a combined analysis of the entrant's internal and external factors.
Here presented analysis is an additional tool because the information entrepreneurs can derive from their internal factors should be complemented by market concentration. More precisely, their business outcome will be, among other factors, influenced by the fact whether they own appropriate real estate, the amount and cost of the available capital, specific knowledge about the LTC sector in terms of optimizing work processes, and the possibility of using innovations to increase efficiency.
Another important element of formal institutional LTC markets, besides overlapping of their social care and healthcare elements, is their diversity depending on the respective country. In that sense, LTC systems in countries that spent a relatively low portion of their GDP on LTC are rarely the focus of scientific interest. This is even more emphasized if we focus on their market concentration. Croatia belongs to those insufficiently analyzed, growing LTC markets. Consequently, here presented analysis will, on the sample of the Croatian formal institutional LTC market, try to answer two following research questions: 1. What determines the market concentration on market for formal institutional LTC? 2. What market characteristics are more likely to attract new entrants to the elderly care home market?
This paper adds to the field in several aspects. To the authors' best knowledge, it is the only paper that focuses on determinants of market concentration in LTC markets in countries that spend a rather modest portion of GDP for LTC and do not have highly developed quality control systems in formal institutional elderly care homes. Meaning that this paper might serve as a decision-O N L I N E F I R S T support tool for both possible new entrants to the formal institutional LTC market and for policymakers in the respective country since obtained results can serve as an indicator of probable market concentration. The rationale behind it is that it is easier to make entrepreneurial decisions if there are available detailed data on existing competitors, i.e. characteristics of their elderly care homes, consumers satisfaction etc., but there is room for improvements when it comes to making these decisions on under-analyzed areas. Therefore, since empirical analysis was made in Croatia, the presented analysis can be applied in countries with similar LTC systems and a lack of publicly available data on this sector. Further, policymakers might find presented findings useful in a situation when it is necessary to anticipate the level of market concentration. In other words, when it is necessary to predict which areas are more likely to be underserved by privately owned elderly care homes and consequently to plan for public action or development of substitutional LTC products.
The following chapter provides a brief literature overview. Chapter 2 provides data and a sample description. The methodology is explained in the third chapter, and it is followed by results presented in the fourth chapter. Discussion is given in Chapter 5 while concluding remarks are presented in the last chapter.

A brief literature review
Studies dealing with a market concentration in LTC systems are often directed to a connection between market concentration on one side and price and/or quality on the other. As expected, more studies deal with the US market (Fulton, 2017;Nyman, 1994;Zinn, 1994 etc.), and there are studies dealing with elderly care homes in European countries with developed LTC systems, such as Sweden (e.g. Broms, Dahlström & Nistotskaya, 2020), England (e.g. Forder & Allan, 2014), Netherlands (e.g. Mosca, Pomp & Shestalova, 2010). However, market competition is also analyzed from different perspectives. In that sense, Gandhi, Song and Upadrashta (2020) focused on the impact of private equity acquisitions on the quality of care in US nursing homes. Grant, Kesternich and Van Biesebroeck (2020) analyzed competition between non-profit and for-profit LTC homes in Germany, while Yeh, Tsay, Wang, Lo and Shi (2021) used market concentration as one of the variables that determine nursing home accreditation in Taiwan. Grabowski (2008), on the other hand, deals with several issues related to LTC research and, along with emphasizing unanswered policy questions, raises the question of the construction of standard measures of LTC market boundaries. Although his research is focused on the US market, his critical standpoints are widely applicable. Scourfield (2007) also criticized marketization and the privatization of the LTC elderly care system. Even though his focus was on the UK, concerns expressed in the term "caretalization" might be applied to all countries with developed or developing LTC systems. In other words, this criticism of the commodification of long-term care for the elderly can serve as a corrective to both researchers and policymakers while advocating for efficiency increase and marketization without taking into account possible adverse effects on users of these services.
Institutional elderly care is a substitute for home care, and the financial repercussions of these options are interesting from an individual as well as from a state perspective, where it is necessary to project the demand for these services and analyses the development and sustainability of the social protection system (Manojlović, 2020;Bađun, 2017;Mihić, Todorović, & Obradović, 2014). Closely related to the issue of sustainability of this sector are products such as LTC insurance that are designed to satisfy customers' needs and, at the same time, relieve financial pressure on the government (e.g. Fuino, Ugarte Montero, & Wagner, 2022;Karagiannidou & Wittenberg, 2022;Bađun & Krišto, 2021;Wang, Abiiro, Yang, Li, & De Allegri, 2021;Eling & Ghavibazoo, 2019;Courbage, 2010, Zhou-Richter, Browne, & Gründl, 2010. Besides financial-type innovations such as LTC insurance, the very nature of long-term care requires continuous improvements in the social aspect of the provided service, and there is O N L I N E F I R S T growing literature focused on social innovations (e.g. Mali (2019) provides an excellent overview of innovations mainly related to improving quality life in elderly nursing homes). Closely related, residents' quality of life in elderly care homes is a focus of many studies (e.g. Orr, et al., 2023;Xu, et al., 2023;Harouni, Tabrizi, Fallahi-Khoshknab, Fadayevatan, & Maddah, 2022;Olumekor, Stojić, Kehler, & Polo, 2022;Sjölund, Mamhidir, & Engström, 2021;Lovreković & Leutar, 2010, etc.) Further, there is rising interest in technological innovations related to social robots (cf. Lavin, et al., 2022;Blindheim, Solberg, Hameed, & Alnes, 2022, etc.) while Lapp, et al., (2022) for example provide an overview of studies focused on clinical decision support tools in elderly LTC homes. Similar, Chiang, Hsieh and Wu (2022) analyses the implementation and acceptance of information and communication technology (ICT) in LTC institutions in Taiwan.
In countries with highly developed LTC systems, such as Sweden, more specific themes gather researchers' attention, such as the question of inequality in the use of this care in diverse ageing societies with sizable migrant populations (cf. Innes, Walsh, & Österberg, 2021).

Data and sample description
Empirical analysis has been made on Croatian companies operating in activities under code 87.3 Residential care activities for the elderly and disabled in NACE Rev. 2 classification of economic activities in the European Community. Data on each business subject registered for this formal institutional LTC service were retrieved in January 2023 from Croatian Financial Agency (FINA) database named info.BIZ. The sample covers 233 companies that operate in the respective sector with total revenue for 2021 higher than 0 euros.
The elderly prefer to stay in nursing homes relatively close to their family, friends and primary healthcare providers. Therefore, following the rationale presented in Zwanziger, Mukamel and Indridason (2002) and Grant et al. (2020), counties are selected as separate markets for elderly care homes. Hence, 198 micro and 35 small-sized companies have been grouped according to the 21 counties they are registered.
Herfindahl-Hirschman index (HHI) has been selected as a standard measure of market concentration. It is calculated by summing the squared market shares of each competing company in the industry, and its value lies between 0 and 10 000 (the higher the value, the more concentrated market; in a monopoly, it is 100 2 =10 000). There are other several widely used market concentration measures, such as concentration ratio and Gini coefficient, but following Yeh et al. (2021), Fulton (2017), Forder and Allan (2014), Mosca et al. (2010), Zwanziger et al. (2002), Zinn (1994) and Nyman (1994) we used HHI. However, the aforementioned studies mainly used the number of beds to measure the size of an elderly care home. Since the aim of this paper is rather different, the market share of each company has been calculated based on the respective total revenue so it can better relate with the entrepreneurial segment of decision that is based on the level of market concentration.
Since the main purpose is to find what determines the level of market concentration in the formal institutional LTC market, it was necessary to select explanatory variables. Demand for products/services is positively correlated with the population, yet when it comes to the LTC market for the elderly, it is necessary to slightly adjust this variable. Namely, it is expected that the higher number of elderly will increase demand for elderly care homes, consequently increasing market competition, i.e. lowering market concentration. Therefore, the analysis includes the number of individuals that are 65 and older in each county.
Further, the cost of formal institutional LTC in the majority of cases is covered by users of this care or by their family members. Hence, it is expected that higher purchasing power will have the same effect on market concentration as higher demand. Or in other words, higher GDP per capita will more likely mean that those in need can afford to be admitted to elderly care homes, which will indirectly attract more entrants to the market and decrease market concentration. These two variables have often been used in studies related to market competition (Yeh et al., 2021;Nyman, 1994;Fulton, 2017, Forder & Allan, 2014. The unemployment rate is another important variable that seems to have a relevant direct impact on market competition in the respective market. However, its influence is twofold. On one side, a higher unemployment rate indicates a higher number of individuals potentially available to take care of the elderly, whether in an informal non-institutional way, i.e. taking care of their family members, or taking care of non-related elderly that are not admitted in elderly care homes. In this sense, a higher unemployment rate O N L I N E F I R S T would imply a lower demand for elderly care homes, i.e. higher unemployment rate would mean lower market competition and higher market concentration. However, LTC is laborintensive, and as in many other (health-related) industries, companies struggle to find employees. From this perspective, a higher unemployment rate might be perceived as an opportunity for new entrants to the LTC market since they represent a possible source of necessary inputs (employees) that might attract new entrants. Or in other words, there is a possibility that this variable might have an adverse effect than in the previous case depending on which effect prevails.
All data on these three variables have been retrieved from the Croatian Bureau of Statistics ("Census of population, households and dwellings in 2021" and for data on GDP "Gross domestic product for Republic of Croatia, HR_NUTS 2021. -HR NUTS 2 AND COUNTIES, 2020"). IBM Statistics SPSS 23 has been used for the analysis.

Methodology
The empirical segment of this paper is divided into two parts. The first part is focused on the determinants of market concentration on the market for formal institutional elderly long-term care measured by HHI. Therefore, to test whether selected explanatory variables affect HHI as a measure of market concentration and what is the nature and strength of their impact, the following linear regression has been used: HHIit=β0+β1 GDPPC+ β2POP65 +β3UNEMP+εit (1) i=1, 2,…, 21; t=2021 HHI -Herfindahl-Hirschman index, county level (units) POP65 -the number of individuals in a county that is 65 and older GDPPC -gross domestic product per capita, county level (units, EUR) UNEMP -registered unemployment rate, county level (percentage).
In this manner, by analyzing data for a single year, it will be possible to answer the first research question.
The aim of the second empirical part is to determine what characteristics of the market are more likely to attract new entrants to the elderly care home market. Hence, to answer the second research question and detect how objects (counties) are classified into groups, a cluster analysis has been performed following the same logic Morrison and Bryan (2010) used to find spatial clusters of elderly consumers in the U.S.A. Three previously mentioned explanatory variables (POP65, GDPPC, UNEMP) have been used in kmeans clustering. In order to perform the analysis, this method requires selecting the number of clusters prior to the analysis. Consequently, taking into consideration studies with similar logic (Kočanová, Kováč, Serzhanov, & Buleca, 2023;Zhang, et al., 2022;Tobis, et al., 2021) and the size of Croatia, the analysis is set to be performed with three cluster groups.

Results
As previously stated, the empirical analysis consists of two segments. Consequently, obtained results for regression and cluster analysis are presented separately.

Regression analysis
The model summary is presented in Table 1, and according to the Durbin-Watson test, there is no autocorrelation in this model since the value of 2.429 is within the acceptable range (1.5-2.5). Results of the Breusch-Pagan test for heteroscedasticity are available at request, and they indicate that heteroscedasticity is not present (p=0.09; p>0.05). With a p-value of 0.013, the estimated regression model is statistically significant ( Table 2).

Cluster analysis
As previously stated, a cluster analysis has been used to answer the second research question, and three previously used explanatory variables, along with the value of HHI for each county have been included in the analysis. Their descriptive analysis is given in Table 4. Sixteen out of twenty-one counties belong to cluster 1, City of Zagreb is the only member of cluster 2, while the remaining four counties belong to cluster 3 (Table 7).  The presented Euclidean distance between each case, i.e. county (Table 5), as measures of similarity, and final cluster centers (Table 6) reveal that clusters are not uniform. It is evident that Cluster 2, with intense competition (the lowest level of HHI), also has the highest GDP per capita and the number of individuals 65 and older, while the unemployment level is the lowest. The difference between "mid-range competition" Cluster 3 and "low competition" Cluster 1 is most pronounced in the population segment, while concerning GDP per capita and unemployment level differences are not as drastic.

Discussion
Considering both regression and cluster analysis results, it is evident that when it comes to GDP per capita, its effect is stronger in terms of increased demand for elderly care homes. In other words, in areas with higher GDP per capita, individuals and their family members will be able to afford formal institutional LTC, which will attract new entrants to the market and decrease market concentration. This logic is supported by results for the unemployment rate indicating that in the areas with a higher unemployment rate, it is more likely that unemployed individuals will provide non-institutional care for the elderly, consequently leading to lower competition and higher levels of HHI. Li, Li and Huang (2022) indirectly confirmed results presented here since their research indicates that rural areas, usually with lower GDP per capita and higher unemployment levels than urban areas, have underdeveloped elderly care institutions. However, managing elderly care homes in areas with low unemployment levels might result in problems of labor shortages. In that sense, Vogt and König (2021) propose increased use of robotic devices and ICT to reduce employees' physical hardships and mental stress. Even though their study is focused on Japan, their findings are beneficial to both managers and policymakers in all areas with high levels of emigration of medical and nursing staff (Cabanda, 2023;Byrne, et al., 2021;Lanati & Thiele, 2021;Żuk, Żuk, & Lisiewicz-Jakubaszko, 2019, etc.).
The importance of a population aged 65 and older is confirmed by both methodologies. Namely, more individuals that are potential residents in elderly care homes increase demand for this service, attract new entrants, decrease market concentration and increase market competition. June, et al., (2022) obtained similar results regarding population density, and even though their research deals with factors that are associated with the closure of assisted living facilities, respective results take into consideration that an oversaturated market increases the risk of being unprofitable. However, a new entrant in the formal institutional LTC market should, along with the level and portion of the population older than 65, take into consideration a growing trend of ageing-in-place, i.e. deinstitutionalization of ageing individuals (Szweda-Lewandowska, 2022;Salime, Clesse, Jeffredo, & Batt, 2022, etc.). Namely, future demand for this type of service will increasingly consist of high-need elderly (Alders & Schut, 2019), and in that sense, market entrants should adopt their supply of institutional LTC care to residents that have severe physical and mental health issues. In other words, demand for elderly care homes is likely to be delayed, observed from the age of a resident entering the LTC institution, because of aforementioned efforts invested in deinstitutionalization on the government's level and substitutes developed on the free market, such as assisted living facilities (Bogataj, Bogataj, & Drobne, 2023;Kim, et al., 2022;Bowblis, 2014, etc.).
Additional comparative analysis of here obtained results shows that the variable related to O N L I N E F I R S T

Josipa Višić
Market characteristics and entry strategy decision making: the market perspective of Croatian elderly care homes STRATEGIC MANAGEMENT, Vol. xx (2022), No. x, pp. 0xx-0xx population seems to have the same effect as in Yeh et al. (2021) and Fulton (2017). Further, both papers used household income variable and obtained contradictory results, while Fulton's (2017) analysis revealed the negative impact of unemployment level on HHI. However, a comparison of here presented results should be made with caution. Namely, these two studies, even though they follow the same logic regarding the LTC sector as a segment of the health sector, are different since one is focused on determinants of accreditation scores for nursing homes (Yeh et al., 2021) and the other on the market concentration for hospitals, physician organizations and health insurers (Fulton, 2017).
Along with presented factors that affect new entrants' decisions, the elderly care homes market structure in highly developed countries has an additional element that concerns policymakers. Namely, observed increased ownership transfers might result in the emergence of local monopolies, which increases the possibility of a decline in the quality of care for residents (Holden, 2002). Even though this research does not include quality in empirical analysis, concerns related to the connection between market concentration and quality should not be neglected and should be taken into consideration by both managers and policymakers involved in formal institution LTC care (c.f. Espuny Pujol, Hancock, Hviid, Morciano, & Pudney, 2021;Hirth, et al., 2019;Castle, Engberg, & Liu, 2007, etc.).

Conclusion
The main purpose of this paper has been to analyses what determines the market concentration on the formal institutional LTC market, i.e. to detect characteristics of the market that is more likely to have higher market competition. This is important both for new market entrants and policymakers since LTC has a social and health dimension, and providing this service should not be left solely to market forces. By observing obtained results, it is evident that all analyzed variables play a significant role. As expected, demand for LTC service plays a significant role, and the market with a higher number of elderly will attract more competitors. When the level of GDP per capita and the unemployment rate are observed together, it appears that markets with stronger economies tend to attract entrepreneurs. In other words, it is more likely that someone will open an elderly care home in a densely populated county with individuals who can afford formal institutional LTC for themselves or their family members. From an entrepreneurial perspective, this is expected and logical. Still, from the perspective of necessary social and health protection, obtained results can serve policymakers to more easily detect areas/counties where the free market economy might not satisfy the needs for LTC for all elderly in need. In other words, this study helps detect areas where: a) developing non-institutional formal LTC might be needed or b) it might be beneficial to develop measures/incentives to attract entrepreneurs to the formal institutional LTC market.
However, even though empirical analysis has been made on Croatian counties, obtained results and conclusions are applicable in all countries with similar social protection and health care systems. This is especially important in areas where there is no (publicly available) data on many segments of LTC for the elderly such as internal characteristics of LTC providers and their users that might help entrepreneurs decide whether to enter the respective market.
The shortcomings of this paper are mainly related to the lack of data on prices and quality measures. Information on the number of beds in each elderly care home would enable an alternative calculation of HHI, while data on service prices and structure of employees as a proxy for quality (medical and non-medical staff) would enable a more reliable comparative analysis of obtained results. Further, as Grabowski (2008) stated, changing market boundaries might distort the results, but the analysis has been optimized according to the available data.
Therefore, future studies on this subject will be directed at minimizing the aforementioned shortcomings. Further, the analysis will be widened by including variables related to the portion of unemployed females in the market since many studies indicate that female family members more often provide informal care, which is a substitute for the observed formal institutional LTC market, and at the same time, they are more likely to be employed in formal LTC institutions. Hence, including additional variables might provide additional information to potential entrants and policymakers related to the LTC system.