Emil Reinikainen Major Shareholders and ESG Performance Evidence from the Nordic countries 2018-2023 Vaasa 2024 School of Accounting and Finance Master’s Thesis in Finance Master’s Degree Programme in Finance 2 Vaasan yliopisto Laskentatoimen ja rahoituksen yksikkö Tekijä: Emil Reinikainen Tutkielman nimi: Major Shareholders and ESG Performance: Evidence from the Nordic countries 2018-2023 Tutkinto: Kauppatieteiden maisteri, rahoitus Oppiaine: Rahoituksen maisteriohjelma Työn ohjaaja: Sami Vähämaa Valmistumisvuosi: 2024 Sivumäärä: 80 TIIVISTELMÄ: Tämän pro gradu -tutkielman innoittajana on vastuullisuuden ajankohtaisuus. Alati kasvava re- gulaatio sääntelee yritysvastuuta, ja esimerkiksi ensimmäiset EU:n yritysvastuuraportointidirek- tiivin (CSRD) mukaiset kestävyystiedot raportoidaan tilikaudelta 2024. Moni aikaisempi tutkimus ESG-aiheista keskittyy tutkimaan ESG-tekijöiden ja pörssiyhtiöiden taloudellisten mittarien vä- lillä olevia suhteita, ja samalla ESG-tekijöiden ja pörssiyhtiöiden omistajien välistä suhdetta on tutkittu huomattavasti vähemmän. Tästä syystä tämä pro gradu -tutkielma keskittyy pohjois- maalaisten pörssiyhtiöiden pääomistajien ja ESG-pisteytyksen välisen yhteyden tutkimiseen. Tutkimus vastuullisuusaiheista johtaa juurensa 1900-luvun jälkipuoliskolle, jolloin etenkin yh- dysvaltalaiset ja pohjoismaalaiset yhtiöt alkoivat kiinnittää huomiota työntekijöidensä hyvin- vointiin niin toimeentulon kuin yhteiskunnallisten kysymysten kautta. Vaikka ESG-tutkimuksia on tänä päivänä laajalti saatavilla, vain harva tutkimus 2000- tai 2010-luvulla on keskittynyt poh- joismaalaisiin yhtiöihin, joten tämä tutkimus keskittyy tutkimaan Norjassa, Tanskassa, Suomessa ja Ruotsissa listattuihin pörssiyhtiöihin laajentaakseen aikaisempaa tutkimusta vähemmän tut- kituilla markkina-alueilla. Pääomistajat ovat tämän tutkielman keskiössä, sillä etenkin Pohjois- maissa aktiivisilla pääomistajilla on merkittävä rooli pohjoismaalaisissa talouksissa. Tämän pro gradu -tutkielman tarkoituksena on selvittää omistuspohjan vaikutusta suurten poh- joismaisten pörssiyhtiöiden vastuullisuussuoriutumiseen. Tutkimuksen empiirinen osuus koos- tuu 93:sta Europe Stoxx 600 -osakeindeksiin kuuluvasta yhtiöistä usealta eri toimialalta. Yhdessä yhtiöt edustavat noin kahta kolmasosaa pohjoismaisten pörssien yhteenlasketusta markkina-ar- vosta. Tutkimuksessa selittävänä muuttujana käytetään eri pääomistajatyyppejä kuvaavia muut- tujia ja riippuvana muuttujana vastuullisuutta kuvaavaa ESG-pisteytystä ja pisteytystä kolmessa eri pääkategoriassa, jotka ovat ympäristö (E), yhteiskuntavastuu (S) ja hyvä hallintotapa (G). Tutkimuksessa saadut tulokset osoittavat tilastollisesti merkittäviä negatiivisia yhteyksiä omis- tuspohjan keskittyneisyyden ja ESG-pisteytyksen välillä. Havainto on erityisen suuri perheomis- teisissä pörssiyhtiöissä. Syy tämän havainnon takana ajatellaan olevan pääomistajien halutto- muus uhrata taloudellista hyvinvointiaan toteuttamalla kalliita vastuullisuusinvestointeja. Toi- saalta valtio-omisteiset yhtiöt menestyvät keskimääräistä paremmin hallintotapaa mittaavissa pisteissä. Lisäksi tulokset osoittavat suhteellisen velkaantumisen kasvun olevan merkittävä ne- gatiivinen asia ESG-näkökulmasta. Merkittävin positiivinen yhteys huomataan olevan yhtiön ta- seen koon ja ESG-pisteytyksen välillä. AVAINSANAT: ESG, Active ownership, Family companies, Institutional investors, State-owned en- terprises, Ownership concentration 3 ABSTRACT: The motivation for this master’s thesis is the topicality of sustainability themes. Ever-growing regulation regulates corporate responsibility and, for example, the first sustainability disclosures in accordance with the EU Corporate Responsibility Reporting Directive (CSRD) will be reported for the financial year 2024. Much of previous research on ESG topics focuses on the relationships between ESG factors and the financial metrics of listed companies, while much less research has been done on the relationship between ESG factors and shareholders of listed companies. For this reason, this master’s thesis focuses on the relationship between the controlling sharehold- ers of Nordic publicly listed companies and ESG scores. Research on ESG topics date to the second half of the 1900s when companies in the United States and Nordic countries began to pay attention to the well-being of their employees through both livelihood and social issues. Although ESG studies are widely available today, only a few studies in the 2000s and 2010s have focused on Nordic companies. Therefore, this thesis focuses on companies in Denmark, Finland, Norway, and Sweden to expand previous research in less studied markets. This thesis focuses on the impact of companies' main shareholders on sustain- ability indicators, as active owners play a significant role in companies' decision-making, espe- cially in the Nordic countries. The purpose of this master’s thesis is to examine the impact of ownership on the sustainability performance of large Nordic listed companies. The empirical part of the thesis consists of 93 companies included in the Europe Stoxx 600 stock index in various industries. Together, the companies represent approximately two-thirds of the market capitalization of the Nordic stock exchanges. The explanatory variables in this thesis are variables describing the type of main shareholder, and the dependent variable is ESG scoring, which describes responsibility perfor- mance and scoring in three main categories: environment (E), society (S), and governance (G). The variables used in several studies in the field, such as the company's age, balance sheet total, return on capital, and indebtedness, are used as contrast variables in the study. The results of this thesis show statistically significant negative associations between ownership concentration, particularly in family companies, and sustainability performance. The cause be- hind this observation is thought to be the reluctance of main shareholders to sacrifice their fi- nancial well-being by undergoing expensive sustainability investments. On the other hand, state-owned companies perform better than average in corporate governance scores. Addition- ally, the results show that the level of debt to assets is a significant negative factor in ESG per- formance. The most significant positive link is between the size of a company's balance sheet and ESG scoring. UNIVERSITY OF VAASA School of Accounting and Finance Author: Emil Reinikainen Title of the Thesis: Major Shareholders and ESG Performance: Evidence from the Nordic countries 2018-2023 Degree: Master’s Degree in Finance Programme: Master’s Programme in Finance Supervisor: Sami Vähämaa Year: 2024 Pages: 80 KEYWORDS: ESG, Active ownership, Family companies, Institutional investors, State-owned enterprises, Ownership concentration 4 Contents 1 Introduction 8 1.1 Purpose of the study and hypothesis 10 1.2 Structure of the thesis 12 2 The Nordic landscape 14 2.1 The Nordic governance model 15 2.2 Regulatory framework 18 3 Theoretical framework 21 3.1 Stakeholder theory 22 3.2 Legitimacy theory 24 3.3 Upper Echelons theory 26 3.4 Agency theory 27 3.5 Industry and country influence 29 4 Ownership concentration 31 4.1 Family ownership 32 4.2 Institutional ownership 36 4.3 Government ownership 38 5 Data and methodology 41 5.1 Research design and data sample 41 5.2 Regression variables 43 5.2.1 Dependent variable 43 5.2.2 Independent Variables 44 5.2.3 Control Variables 47 5.3 Regression models 50 5.4 Correlation coefficients 55 5.5 Hausman test and limitations of the data sample 57 6 Regression results 59 5 6.1 Results of concentrated ownership 59 6.2 Results of family-controlled companies 62 6.3 Results of institution-controlled companies 64 6.4 Results of government-controlled companies 65 7 Conclusions 67 References 69 6 Figures Figure 1. Percentage of publicly listed companies with one or more shareholders with more than 20% voting rights (grey bar) or over 50% (dotted bar) Lekvall (2014). 16 Figure 2. Illustration of different governance models by Lekvall (2018). 18 Figure 3. Stakeholder Map Rhenman, E. (1968) 24 Figure 4. Categories of ESG Pillars in the Eikon database. 44 Figure 5. Pie graph of the distribution of ownership in sample companies. 47 Figure 6. Industry breakdown and distribution by NACE codes. 52 Tables Table 1. Description of the sample selection. 43 Table 2. Descriptive statistics of regression variables. 50 Table 3. Recap of variables used. 55 Table 4. Pearson's correlation matrix with *** denoting a 95% confidence level. 56 Table 5. Regression results of concentrated ownership. 60 Table 6. Regression results of family-controlled companies. 63 Table 7. Regression results of Institution-controlled companies. 65 Table 8. Results on government-controlled companies 66 Abbreviations AGM – Annual General Meeting CSR – Corporate Social Responsibility CSRD – Corporate Sustainability Reporting Directive D/E – Debt – To – Equity – Ratio EFTA – The European Free Trade Association ESG – Environmental, Social and Governance EU – The European Union 7 NFRD – Non-Financial Reporting Directive ROA – Return on Assets SDGs – Sustainable Development Goals SOE – State-Owned Enterprise SRI – Socially Responsible investing UET – Upper Echelons Theory 8 1 Introduction During the last decades, sustainable business practices, Socially Responsible investing (SRI), and Environmental, Social, and Governance (ESG) regulation and disclosures have become an important part of corporate decision-making and academic studies (Chen et al., 2020; Lin & Nguyen, 2022; Maquieira et al., 2024; Pedersen et al., 2021; Flammer, 2021). To respond to the upbeat demand for responsibility, more and more companies have started to engage in socially responsible practices and transparency regarding their actions (Alsayegh et al., 2020). One of the earliest definitions of corporate responsibility toward society is presented in the 1953 book "Social Responsibility of the Businessman" by Bowen H. (1953). This book defines corporate social responsibility as “The obligation of businessmen to pursue those policies, to make those decisions, or to follow those lines of action that are desir- able in terms of the objectives and values of our society.” Although this definition is over seventy years old, it remains relevant today, as seen in the increased regulations and the willingness of corporations to communicate their responsible actions. More recently, re- searchers such as Gillian et al. (2021) have highlighted that interest in socially beneficial behavior and financing has grown significantly. Furthermore, according to data from UNCTAD (2022), the funds invested in sustainable assets reached approximately $5.8 trillion U.S. dollars by the end of 2022, reflecting a fivefold increase over the past five years. While regulations surrounding the ESG landscape change, the extent to which compa- nies engage in sustainability issues and integrate them into their day-to-day operations remains mainly in the company’s own hands. Defreitas Netto et al. (2020) argue that ESG regulation may not lead to actual performance but instead result in increased reporting without much real behavior change. Consequently, the interests of owners, extending beyond mere financial returns, can greatly influence the effectiveness of ESG regulations (Wang et al., 2023). This raises an important question: “Do shareholder interests 9 primarily drive ESG activities or are they a response to societal pressure and agency prob- lems.” Despite the growing global interest in sustainable business practices, Lekvall (2014) high- lights that the Nordic countries have been early adopters of sustainable development within their corporate landscape. A notable example is the Norwegian company Norsk Hydro, which was the first in the world to publish a standalone environmental perfor- mance report in 1989—well before the concepts of ESG and CSR (Corporate Social Re- sponsibility) came into existence (Strand, 2024). The ESG framework, defined by the United Nations Principles for Responsible Investment, has become a crucial standard in responsible investing (Puttonen & Puttonen, 2021). ESG has three main dimensions: Environment, Social, and Governance. The environmen- tal dimension emphasizes corporate strategies addressing climate change, adherence to environmental standards, implementation of energy efficiency measures, commitment to biodiversity, and promoting a sustainable economy (Finland’s Sustainable Investment Forum, 2021). The social dimension evaluates aspects of employer-employee dynamics, such as personnel policies, human rights considerations, children's rights, labor rights, and product liability concerns (Finland’s Sustainable Investment Forum, 2021). The gov- ernance dimension includes corporate administrative practices that emphasize the im- portance of board independence, remuneration structures, incentive schemes, tax com- pliance, and measures to combat corruption and bribery (Finland’s Sustainable Invest- ment Forum, 2021). This holistic approach through three dimensions is essential for eval- uating companies' sustainability and ethical implications through multiple points of view. Since adopting the United Nations Sustainable Development Goals (SDGs) in 2015, the Nordic countries—Denmark, Finland, Iceland, Norway, and Sweden—have consistently performed better than most in adherence to the SDGs. As Strand (2024) notes, the cor- porate landscape in the Nordics demonstrates a strong commitment to sustainability. Between 2005 and 2024, these nations achieved the top three rankings over 30% of the 10 time in the World Economic Forum's Most Sustainable Corporation in the World ranking. A Nordic company is more than 20 times as likely to be included among the top three than a U.S. company (Strand, 2024). Furthermore, the Nordic countries are geographically close, situated in the Nordic Hem- isphere, share similar welfare states, and are all part of the European Union or the Euro- pean Free Trade Association (EFTA). These countries also have a high proportion of active business ownership (Lekvall, 2014). Considering these factors, the Nordic nations pre- sent an ideal context for examining the relationship between ownership and ESG perfor- mance in a rather homogeneous group of countries. In this master’s thesis, the terms sustainability, ESG, and CSR will be utilized as overarching concepts to encompass a broad spectrum of related themes. As noted in other studies analyzing Nordic stock market data, such as Grobys and Huhta- Halkola (2019), Iceland will be excluded from this study due to its small size and relatively low data availability. 1.1 Purpose of the study and hypothesis Increased demand for sustainability raises the need to study the potential relationship between shareholder concentration and ESG performance and to understand whether the presence of a specific dominant owner type could strengthen or weaken ESG perfor- mance. The main goal of this thesis is to examine possible connections between ESG performance and dominant owners in a Nordic context. Another point of interest is the potential explanations of differing ESG performances between ownership types. This thesis will try to answer the following question: Do publicly listed companies in the Nordics with a dominant owner perform better or worse in terms of ESG than those more broadly owned? 11 This study aims to enhance our understanding of the relationship between dominant ownership and corporate ESG performance. While previous research has extensively ex- plored the effects of ESG performance on financial outcomes, stock market returns, and financing costs (Albuquerque et al., 2019; Singh, 2022; Yoo & Managi, 2022; Zhou et al., 2022), the relationship between ownership concentration and ESG has been less exam- ined. Most available studies have primarily focused on the U.S. stock market or emerging markets (e.g., Doshi et al., 2024; Jensen & Meckling, 1976; Maquieira et al., 2024; Vil- lalonga & Amit, 2006). Therefore, the Nordic business environment presents a promising and underexplored research opportunity. Additionally, Nordic countries are noteworthy for their unique approach to influential owners, early adoption of CSR practices, and dis- tinct corporate governance structures, all of which will be further analyzed in Chapter 2.1. Furthermore, there is a relative scarcity of research examining the relationship between ownership concentration and ESG performance. This gap is notable, considering that shareholders with greater voting power can profoundly influence a company's opera- tional strategies and its approach to ESG matters (Abeysekera & Fernando, 2020). To gain insight into this issue, it is crucial to investigate the relationships between various types of ownership and ownership concentration and their impact on a company's ESG perfor- mance. This is particularly relevant in countries with high ownership concentration and prevalent dual-class share structures, such as the four Nordic countries highlighted in this thesis. Following the research questions, three hypotheses based on the existing literature have been formulated: H1: The presence of a controlling shareholder affects ESG performance H2: Nordic family-owned companies exhibit a higher level of social awareness leading to greater investments in ESG activities that enhance overall ESG performance in family- controlled companies. 12 H3: Publicly listed companies with large institutional shareholding perform better in ESG metrics as societies closely monitor institutions. These hypotheses are based on a few key findings in past literature. First, existing re- search has found that controlling shareholders affect multiple aspects of corporate sus- tainability performance regardless of a company’s originating country or jurisdiction (Gil- lan et al., 2021). Second, past research has shown that when companies announce sus- tainability initiatives, it boosts their legitimacy and investor belief in stable business prac- tices, attracting more shareholders (Flammer, 2021; Tang & Zhang, 2020). As Tang and Zhang (2020) find, institutional shareholders, who are often norm-constrained, are in- terested in companies that perform well in terms of sustainability. Third, adopting sus- tainable business practices may increase the attention of investors with a longer invest- ment horizon (Flammer, 2021). Among those investors perceived to have a longer invest- ment horizon are families, governments, and certain institutions such as pension funds. Furthermore, research has shown that transgenerational persistence and socioeconomic status are two significant factors for family-owned companies, which requires avoiding potential sustainability concerns (Le Breton–Miller & Miller, 2006; Berrone et al., 2010; Cruz et al., 2014). 1.2 Structure of the thesis The thesis is structured into seven chapters, each addressing a specific topic. After an introduction in Chapter one, Chapter two will explore the development of the Nordic corporate landscape, highlighting its unique features such as welfare state principles and the distinct governance model known as the "Nordic Model”. Following the discussion on Nordic local characteristics, Chapter two will also examine the most recent regula- tions surrounding sustainability issues and disclosures from a broader European per- spective. After these foundational chapters, Chapter three will delve into widely ac- cepted theories and arguments that inform companies’ and managers' considerations 13 and developments regarding Corporate Social Responsibility. Building on this theoretical framework, the subsequent section will review both recent and earlier studies focused on different types of shareholders, their motivations, and their actual behaviors related to sustainability performance. Finally, chapter five and six will present the research find- ings, and a discussion based on those findings. The thesis will conclude with chapter seven, summarizing the research and discussing implications and directions for future studies on similar questions. 14 2 The Nordic landscape As stated in the previous chapter, the Nordic countries are home to some of the world's most sustainable and transparent companies. This chapter aims to discuss the potential reasons that have led to the development of world-leading stakeholder management practices and a culture of strong stakeholder management. The leading argument of scholars is that the Nordic societies developed from a shared cultural background from the aftermaths of World War I and II, which led to country-wide and rapid development as well as the formation of the Scandinavian Welfare model (Andersen, 2004; Simon, 2017). First, Simon (2017) shows that the success of Nordic countries lies in the balance of co- operation and competition: “The Nordic Model works because they scale up the social control mechanisms that operate spontaneously.” Second, in all Nordic countries, most pensions and welfare aid systems are managed by the public sector, which creates long- term planning and cooperation among various stakeholders. As Andersen (2004) notes: "The defining aspect of the universal welfare model is that access to welfare arrange- ments is an individual right. “This mindset cultivates attitudes and norms that encourage risk-taking behavior as there is little stigma associated with collecting social aid, and the risk of financial ruin is minimal (Andersen, 2004). Additionally, because the financing of Nordic welfare systems is collective, there is little to no direct link between an individual's social aid entitlements and their financing (An- dersen, 2004). Finally, a unique characteristic of the Nordic countries is the presence of strong trade unions and employer organizations, which are deeply rooted in the cultural norms of Nordic societies, The presence of strong employee representation can lead to pressure on Nordic companies to consider stakeholder interests in their operations (Mid- ttun & Witoszek, 2020). 15 2.1 The Nordic governance model The Nordic countries—Denmark, Finland, Iceland, Norway, and Sweden—share a com- mon heritage rooted in the medieval empires of Denmark and Sweden (Mähönen & Johnsen, 2019). With similar histories and cultures, these nations exhibit common traits in their legislation, political views, and a strong emphasis on humanitarian aspects, such as extensive welfare systems and collaboration between employers and employees through a largely unionized workforce (Mähönen & Johnsen, 2019). Denmark, Sweden, and Finland are members of the European Union (EU), while Norway and Iceland are part of the European Free Trade Association (EFTA) and are consequently connected to the European Economic Area. These countries have also been the front-runners in CSR reporting since it gained attention in the 1990s (Khatri & Kjærland, 2023; Lekvall, 2014). Lekvall (2018) shows that the Nordic countries have also produced the most companies in the Forbes 2000 list when accounting for population. The Nordic countries are unique in their distinct governance model and ownership base (Hogfeldt, 2005; Lekvall, 2014; Mähönen & Johnsen, 2019). According to Lekvall (2014), over two-thirds of publicly listed Nordic companies had at least one shareholder with over a fifth of the voting rights, and one in five of the companies had an owner with over half of all votes. For example, in the United Kingdom, the values are 27% and 5%, respec- tively (Lekvall, 2014). The use of dual-class shares is also quite common in the Nordic countries, with nearly one-fifth of companies having two or more classes of shares. This means that direct ownership does not always translate to voting rights for shareholders. Dual-class shares are primarily utilized in family companies that have gone public. In these cases, families issue shares with limited or no voting rights to the public to raise capital while retaining decision-making control through a family-controlled foundation or an investment company (Hogfeldt, 2005; Lekvall, 2014). These controlling owners of- ten take an active role in governance, participating in nomination and other committees. In the Nordic Corporate Governance model, large shareholders are essential as they of- ten take responsibility for managing a company on behalf of smaller shareholders 16 (Thomsen, 2016). They also have the necessary voting rights to appoint the Board of Directors, influence the decisions made by operative executives, and participate in deci- sion-making at annual general meetings. Furthermore, large owners are shown to be encouraged to monitor and control companies (Ekholm & Maury, 2014; Shleifer & Vishny, 1986). Ekholm & Maury (2014) suggest that there is a positive relationship between a large owner’s portfolio concentration and a company’s positive abnormal stock returns due to large shareholders having the ability to concentrate and drive changes aimed at enhanc- ing a company’s operations. However, the clear downside of the Nordic Corporate Gov- ernance model is that larger shareholders may disregard minor ones and extract re- sources from the company for their own benefit (Cronqvist & Nilsson, 2003; Jensen & Meckling, 1976; Lekvall, 2018). Figure 1. Percentage of publicly listed companies with one or more shareholders with more than 20% voting rights (grey bar) or over 50% (dotted bar) Lekvall (2014). In 2016, Thomsen highlighted the key difference between Nordic and Anglo-American corporate governance as the "pecking order" of oversight (Lekvall, 2018). In Nordic com- panies, a strict separation between operational and oversight roles is maintained, and external directors have distinct responsibilities in the decision-making process (Lekvall, 17 2018). Most Nordic companies keep the board of directors separate from executives, except in Sweden, where about 40% of public companies have the CEO also serving as a board member (Lekvall, 2018). On the other hand, the Anglo-American model features a unified board composed of executives and non-executive directors. The Nordic model aims to disperse potential conflicts of interest and agency problems between sharehold- ers and executive directors and gives ultimate control to shareholders, with a clear dis- tinction of duties between board members and executive directors. Nordic countries protect minority shareholders from potential abuse by large block holders by implement- ing laws and regulations that prevent large shareholders from making decisions that could compromise equal treatment of all shareholders (Lekvall, 2018). Unlike other Western countries, Nordic company law gives majority shareholders significant voting rights to control companies in annual general meetings (AGMs) (Lekvall, 2014; Thomsen, 2016). Large shareholders with enough voting rights can appoint and replace members of the board who, in turn, can hire and replace executive managers (Lekvall, 2014; Thom- sen, 2016). In addition to holding significant voting rights, these major shareholders of- ten serve on different nomination boards or committees, making proposals for the board of directors to AGMS, and thus enhancing their influence further. 18 Figure 2. Illustration of different governance models by Lekvall (2018). 2.2 Regulatory framework As members of the European Union or the European Economic Area, Denmark, Finland, Sweden, and Norway all encounter regulatory mandates and goals regarding environ- mentally friendly corporate behavior. This sub-chapter discusses the development of re- cent regulations. The European Union and all Nordic countries have pledged to lower poverty, find environmentally friendly solutions, and ensure everyone’s human rights in the immediate future (European Commission, 2022). Reporting on sustainability matters remained voluntary until the last decade when the EU introduced the Non-Financial Reporting Directive (NFRD 2014/95/EU), which began regulating ESG reporting in the EU (Hummel & Jobst, 2021). Initially, the NFRD only re- quired large companies and financial institutions to disclose certain non-financial infor- mation yearly. Later, the NFRD was adopted to all public companies with further legisla- tion, which will be briefly introduced below. 19 All member states of the United Nations signed the Paris Agreement in 2015, which aims to enhance living conditions through 17 Sustainable Development Goals (SDGs) (United Nations, 2015). These goals include aims such as reducing hunger and poverty and com- bating gashouse emissions globally (United Nations, 2015). According to Hummel and Jobst (2021), the Paris Agreement marked a distinct turning point in societal awareness toward combating poor living conditions and climate change. Since the adoption, more distinct efforts to include all stakeholders globally have been made (Dolan & Zalles, 2021, p. 16). Since the signing of the Paris Agreement, significant amendments have been made to the Non-Financial Reporting Directive (NFRD), which was updated in 2018. This revised directive requires companies based in EU member states to disclose how sustainability issues impact their operations and how their activities affect society and the environ- ment (NFRD 214/95/EU). To further enhance the credibility of corporate non-financial reporting, the European Union has implemented a set of financial regulations that guide organizations in developing effective sustainability strategies (Cremasco & Boni, 2022). These regulations are outlined in the Sustainable Finance Disclosure Regulation (SFRD 2019/2088/EU), which mandates financial actors to formally disclose their compliance with sustainability issues and reporting obligations, aiming to reduce greenwashing—a practice where companies publicly commit to sustainability goals without demonstrating genuinely sustainable practices (Cadman, 2011; De Freitas Netto et al., 2020). During the COVID-19 pandemic, the European Commission aimed to make Europe the first climate-neutral continent by 2050. They launched the European Green Deal and the European Climate Law (European Climate Law 2021/1119; European Commission, 2019). The goal is to reduce net greenhouse gas emissions by 55% by 2030 compared to 1990 levels. In the business world, this deal requires companies to provide clearer information about their practices and funding for sustainable technologies like solar and wind power (European Commission, 2024). For instance, in 2023, the Commission's Recovery and 20 Resilience Facility program, InvestEU, and the Innovation Fund offered a total of 662 bil- lion euros for funding (European Commission, 2023). One of the latest EU-wide regulations is the Corporate Sustainability Reporting Directive (CSRD), which begins to take place in 2024 for the largest EU companies and in 2026 for all other large and publicly listed companies (The European Commission, 2024). This di- rective will create uniform standards for how large corporations and publicly listed com- panies disclose their non-financial information regarding CSR. According to the European Commission (2024), this new directive will help stakeholders to access important infor- mation for evaluating a company's environmental and social impact. The European Com- mission hopes that this regulation will encourage businesses to work towards a greener future (Velte, 2023). 21 3 Theoretical framework This chapter offers a theoretical framework for understanding the main economic theo- ries related to ownership and management within the context of CSR considerations. This chapter mainly focuses on the societal impacts faced by companies and their share- holders when making sustainability decisions, as these choices affect not only the com- pany itself but also the reputation of its managers and owners, the environment, and the communities in which they operate. A major concern in CSR is that company management might prioritize their interests over those of its shareholders (Jensen & Mecklin, 1976). To understand the issue, this chapter lays out the theoretical foundation for understanding how ownership structure and man- agement impact corporate performance. Examining the motivations and influencing fac- tors behind the decision-making process is important as shareholders actively engage in supervisory roles, and significant shareholders can exert substantial control. Milton Friedman argued in 1970 that a company's primary responsibility to society is to increase its profits (Friedman, 1970). He claimed that individuals are free to use their private funds however they see fit, but a manager must focus on projects that maximize shareholder value (Friedman, 1970). According to Friedman, the rationality of corporate social responsibility is dependent on profitability. Even today there are diversified views on whether there are monetary benefits in ESG-related investments (Nguyen et al., 2023). One of the most significant studies in this area is by Kempf and Osthoff (2007), which examined ESG-constrained investments. Their findings revealed that positive and best- in-class ESG portfolios generated considerably better returns than portfolios made of other stocks during a time period from 1992 to 2004. They argued that the market fails to fully recognize the value of stakeholder relations, as evidenced by lower stock prices and superior investment performance during that period. Recent academic research has also shown that CSR investments are positively correlated with shareholder value. This 22 relationship is attributed to factors such as increased product differentiation, improved stakeholder trust, access to cheaper financing, and reduced stock price volatility (Albu- querque et al., 2019; Guenster et al., 2011; Kempf & Osthoff, 2007; Lins et al., 2017). However, Borges et al. (2013) present a viewpoint that is contrasting to that of Kempf and Osthoff (2007). They highlight that the end of Kempf and Osthoff's timeframe in 2004 marked a significant turning point in how markets value responsible companies. By employing Carhart's (1997) four-factor model, Borges et al. (2013) demonstrate that, from August 2004 to 2009, the returns of ESG portfolios have been underperforming their stock market peers, suggesting that these portfolios may have become overvalued. Possible explanations behind these findings include issues stemming from overinvest- ment by sustainability-constrained investors, such as mutual funds, and excessive invest- ment in CSR initiatives that do not translate into revenue growth or increased profitabil- ity in the future (Di Giuli & Kostovetsky, 2014; Hong & Kacperczyk, 2009). Furthermore, several studies in this field, such as Di Giuli and Kostovetsky (2014) and Hong & Kacper- czyk (2009), indicate that CSR investments can have adverse or negligible effects on a company's financial and stock market performance. 3.1 Stakeholder theory In the Nordic countries, it is believed that a company is more than just its employees, managers, and owners (Strand & Freeman, 2015; Strand, 2024). Instead, a company’s success or failure also depends on the unwritten rules and the social context where it operates (Rhenman, 1966; Freeman, 1984; Strand & Freeman, 2015; Strand, 2024). The term “stakeholder” was first used in management literature by Erik Rheman in a 1968 publication, Industrial Democracy and Industrial Management. The book includes the world’s first stakeholder map as portrayed in Figure 3. Stakeholder theory explains that companies affect different groups and individuals inside and outside the organization and vice versa (Freeman, 1984). A stakeholder can be 23 anyone who impacts or is impacted by the company’s actions, including suppliers, cus- tomers or government bodies (Freeman, 1999). This theory contrasts with Porter’s Five Forces model (1979), which views the business world as a competitive battleground. Stakeholder theory focuses on collaboration instead. Stakeholder theory focuses on how engaging with different stakeholders can create value for everyone involved by introducing new perspectives (Freeman, 1984; Strand, 2024). According to Freeman (1984) and Rhenman (1966), maximizing of long-term value requires enhancing overall societal well-being. Today, stakeholder engagement re- mains important in the European Commission's original definition of CSR (2001). Re- search by Sarkis et al. (2010) also indicates that stakeholders pressure companies to re- duce negative impacts like greenhouse gas emissions and adopt sustainable practices. To meet the demands and expectations of stakeholders, it is important for a company to identify all parties affected by their operations (De Camargo Fiorini et al., 2018). Nordic societies have a strong tradition of stakeholder engagement (Strand & Freeman, 2015). Freeman and Strand introduce the concept of the “Nordic Cooperative Advantage” to characterize the tendency of these societies to adopt cooperative strategies that en- hance value creation for both their communities and businesses. As discussed in Chapter 2.1, Nordic societies have long-standing traditions and legislation aimed at protecting minority shareholders and prioritizing the welfare of society as a whole (Lekvall, 2018; Thomsen, 2016). Contributing factors to this cooperative engagement are the strong in- stitutional frameworks, welfare states, and the Nordic Corporate Governance model, which all developed largely after World War II. Strand (2024) notes, that referring to in- dustry counterparts as peers instead of competitors can induce cross-industry collabo- rations in non-competitive areas, which is commonly seen in the Nordics. As a practical example, the Swedish clothing retailer H&M describes its collaborative ap- proach by stating: “Many challenges are best addressed collectively; we work with in- dustry peers and even companies in other sectors to establish industry standards and 24 common responses to shared challenges” (H&M Stakeholder Engagement Guidelines, n.d.). Figure 3. Stakeholder Map Rhenman, E. (1968) 3.2 Legitimacy theory Hong and Kaceperczyk (2009) argue that social norms play a significant role in influenc- ing economic behavior and market outcomes, often outweighing the motives for profit- ability. An early example of this concept can be found in the 1957 model of discrimina- tion, where employers with discriminatory beliefs conform to societal norms by avoiding interactions with certain groups (Arrow, 1972). Arrow (1972) notes, that this behavior may help explain why hiring inexpensive labor in unfavorable conditions may not be prof- itable in the long run for a company’s overall profits. One of the most notable theories regarding corporate behavior and social norms is the Legitimacy theory (Dowling & Pfeffer, 1975). This theory states that companies maintain their legitimacy by aligning their actions with societal values and norms, particularly re- garding environmental practices, to meet public expectations. Legitimacy theory and its 25 applications emphasize the study of voluntary sustainability disclosures made by com- panies. According to this theory, several factors influence a company's disclosure prac- tices (Dowling & Pfeffer, 1975; Zhang et al., 2024). First, external stakeholders, including customers and regulatory bodies like the European Commission, have increasingly high expectations for companies to operate sustainably and report their efforts. Second, com- panies believe that demonstrating strong sustainability performance and transparent disclosures can enhance their market value and competitive positioning (Albuquerque et al, 2019). Hummel and Jobst (2021) and Albuquerque et al. (2019) highlight that companies are motivated to disclose their sustainability efforts to be viewed more favorably. This posi- tive perception can ultimately lead to increased sales and improved customer retention. The discussion of environmental actions is intended to justify and legitimize a company's operations, thereby enhancing the appeal of its products or services (Hummel & Jobst, 2021). Furthermore, De Freitas et al. (2021) suggest that companies with poor sustaina- bility performances may engage in voluntary sustainability disclosures as a form of greenwashing, which serves to conceal their actual practices and protect their public image. According to legitimacy theory, it is possible that companies involved in green- washing tend to disclose only vague financial information, while sustainability leaders disclose non-financial information that more accurately reflects their actual perfor- mance. Hummel and Jobst (2021) define high-quality sustainability disclosure as the complete provision of relevant and comparable numerical data that meets or exceeds clearly de- fined quality standards. In summary, legitimacy theory explains why both poorly and well-performing companies will attempt to voluntarily disclose their environmental and social initiatives to legitimize their business, and by doing so, it can provide material ben- efits such as increased sales or reduced loan interest rates (Hummel & Jobst, 2021; De Freitas et al., 2021; Albuquerque et al., 2019). However, in some cases poorly performing 26 companies may opt to conceal their true actions with low-quality disclosures and infor- mation. 3.3 Upper Echelons theory The Upper Echelons theory (UET) was created in the 1970s (Hambrick & Mason, 1984). The theory examines how top executives' personal traits and backgrounds affect their actions and the results of their companies (Hambrick & Mason, 1984). Hambrick and Mason (1984) suggest that the demographic factors of top managers can shape their beliefs, thoughts, and values. The Upper Echelons theory has impacted research in psy- chology and mainstream economics, as the characteristics of executives are important for understanding various business practices (Bertrand & Schoar, 2003). The figure below captures how any situation faced by executives is filtered through their personal charac- teristics, which ultimately shape how strategic choices are made and lead to different outcomes for companies (Carpenter et al., 2004). Figure 4. Model of UET perspective. Carpenter et. al. (2004) In contrast to agency theory, which views managers as interchangeable, Upper echelons theory (UET) states that individual differences between managers can lead to a variety of person-dependent outcomes (Hambrick & Mason, 1984). Recent research, such as 27 that by Harrison et al. (2019), has demonstrated that the personality traits of top execu- tives, as assessed by standard Big Five personality tests, have a significant impact on companies’ strategic decisions. For example, the diversity traits of top managers, includ- ing gender, socioeconomic status, ethnicity, and age, influence CEOs' risk-taking behavior and the long-term financial performance of the companies that they manage. Moreover, recent research has shown that the personal attributes of top management play a crucial role in shaping CSR initiatives and disclosures (Kish-Gephart & Campbell, 2015; Kutzsch- bach et al., 2021; Ortiz‐de‐Mandojana et al., 2019; Smith et al., 2019). However, understanding the cognitive processes that affect decision-making in the con- text of UET is complex. Neely et al. (2020) express concerns about inconsistencies in em- pirical findings across UET studies, particularly in moderation and replication of past studies in the area. 3.4 Agency theory The agency theory is based on managers (agents) making decisions on behalf of share- holders (principals) (Jensen & Meckling, 1976; Spence & Zeckhauser, 1971). During the last decades, it has become the dominant theoretical framework for finance research papers on executive behavior and compensation (Pepper & Gore, 2015). According to the works of Jensen and Meckling (1976), conflicts arise when managers prioritize their own interests over those of shareholders, leading to decisions that may harm principals. This theory describes that agency problems emerge from a complex array of both written and unwritten contracts among various parties, including owners, different company functions, and employees. These problems typically surface due to the differing interests and arrangements between investors and top management (Jensen & Meckling, 1976; Fama & Jensen, 1983). The Agency theory is focused on resolving two problems that can occur in these rela- tionships (Eisenhardt, 1989). The first is the agency problem of differing incentives or 28 goals of managers and owners and related costs, known as agency costs, that occur when a principal has to verify and monitor an agent (Eisenhardt, 1989; Fama & Jensen, 1983). The second is the risk-sharing problem that arises when principals and agents have dif- ferent attitudes toward risk. According to Jensen and Meckling (1976) and Carney (1999), principals are considered risk-neutral due to the possibility of portfolio diversification, while agents are considered risk-averse due to the meaningful economic benefits of sus- tained employment. Because of these assumptions, problems may arise when manage- ment prefers different actions to principals (Eisenhardt, 1989). Agency costs are particularly significant in the Nordic countries, where dual-class shares are common. In this structure, owners with voting rights often have fewer rights to a company's cash flow than their voting rights (Cronqvist and Nilsson, 2003). Cronqvist and Nilsson (2003) note that managers who hold significant voting rights can make decisions that negatively impact minority shareholders' returns by as much as 25%, as measured by Tobin’s Q. This type of agency problem where managers and shareholders’ interests do not meet is called a principal-agent problem or type I agency problem (Abeysekera & Fernando, 2020). The other primarily studied agency problem is called type II agency problem, or principal–principal problem (Abeysekera & Fernando, 2020; Villalonga & Amit, 2006). In the Nordic context, the principal–principal problem can arise when large shareholders also derive non-economic benefits from their companies (Abeysekara & Fernando, 2020). The non-economic benefits may include viewing one's position as an extension of oneself or deriving a sense of identity from one's ownership. If the socio- emotional status of majority shareholders is threatened, they may vote for strategic de- cisions to protect their socioeconomic wealth rather than prioritizing the wealth of all shareholders (Berrone et al., 2010). The Agency theory identifies two main issues: the apparent agency problem related to the conflicting incentives between managers and owners, and the risk-sharing problem that arises from their differing attitudes towards risk. Eisenhardt (1989) and later re- search, such as that by Cronqvist and Nilsson (2003) and Abeysekera & Fernando (2020), 29 highlight the economic implications of these agency problems, particularly emphasizing the agency costs and the effects of these conflicts on minority shareholders. This is no- tably evident in the Nordic corporate scene, where the differences of interests between different shareholders can lead to significant disparities (Lekvall, 2014, 2018). Research on executive managers highlights the importance of effective corporate governance and the need for mechanisms to align managers' interests with shareholders' interests to mitigate potential negative impacts on company value. According to Courteau et al. (2017) and Vural (2018), non-executive shareholders are more likely to monitor exec- utives when their ownership increases, thus reducing the related agency costs and po- tentially increasing value for all shareholders. Controlling shareholders, who hold a sig- nificant portion of shares, are believed to be motivated to oversee and influence deci- sion-making to ensure the company's long-term development. Since they have more at stake, they are also more exposed to risks. Therefore, academics suggest that a certain level of ownership concentration is necessary to prevent opportunistic behavior by ex- ecutive management. (Courteau et al., 2017; Vural, 2018). 3.5 Industry and country influence Different factors influence a company’s ESG performance beyond managers and owners. Earnhart and Lizal (2006) point out that companies in different industries often show differing levels of ESG and CSR. For example, some environmental concerns and differing regulations can be related to a specific industry, significantly influencing a company’s ability to affect its CSR performance (Ioannou & Serafeim, 2012). For example, in the financial industry, companies often have more reasons to invest in socially and environ- mentally responsible actions and to report on them compared to companies in other sectors (Bose et al., 2017). Similarly, companies in the manufacturing sector tend to have lower environmental performance across the sector (Liu et al., 2019). Tang et al. (2018) and Ioannou and Serafeim (2012) argue that a company's country of residence and operations significantly influence its CSR performance. This is primarily 30 due to political, labor market policies, and education system variations. For instance, global data indicates that Finnish companies consistently rank higher in sustainability than their U.S. counterparts, as evidenced by their performance in the World Economic Forum's “Most Sustainable Corporations in the World” ranking (Strand, 2024). Addition- ally, companies with government involvement tend to exhibit stronger CSR engagement compared to their non-government-owned counterparts (Faller & Zu Knyphausen- Aufseß, 2018). Similar associations between different ownership structures and CSR performance relate to the extent of non-financial reporting by companies and specific industry obligations (Bose & Biswas, 2017; Dan & Scholtens, 2012). Dan and Scholtens (2012) highlight a con- nection between non-financial reporting and successful CSR outcomes, indicating that companies that engage in more extensive reporting are more likely to improve their fu- ture performance in these areas. The following chapter will examine various ownership types and existing literature on how the behavior and motivations of various owners may influence ESG/CSR performance. This thesis controls industries using eight proxies based on NACE codes. Potential country influences are not accounted for due to the small sam- ple size of individual companies within the selected countries. Additionally, the legisla- tion and cultures in these four countries are similar, as discussed in Chapter 2. 31 4 Ownership concentration Research indicates that individuals associated with an organization greatly emphasize its reputation and image because reputation is closely linked to individual reputation (Zell- weger et al., 2013). Therefore, a positive relationship between ownership concentration and ESG performance could be expected. However, ESG engagement comes at a cost, and it is unclear whether equity concentration increases the discrepancy between in- curred costs and benefits, particularly from a societal perspective (Dam & Scholtens, 2013). Earnhart and Lizal (2006) find that companies with more concentrated ownership are associated with better sustainability performance, as owners can create pressure to- wards more sustainable activities. Dam and Scholtens (2013) also support that owner- ship concentration in public companies notably affects a company’s strategic decision- making, including ESG initiatives. On the other hand, dominant shareholders may have a lower willingness to drive socially responsible initiatives and may, therefore, show worse performance in governance and social initiatives (Faller & Zu Knyphausen-Aufseß, 2018). Previous literature has shown mixed evidence regarding the relationship between eq- uity ownership concentration and CSR performance. For instance, Faller and Knyphau- sen-Aufseß (2018) observe that different levels of ownership concentration were asso- ciated with varying degrees of engagement in sustainability-related aspects for different companies. These findings also affected CSR decision-making (Faller & Knyphausen- Aufseß, 2018). Similar observations were reported earlier by Dam and Scholtens (2012), who emphasized that different ownership structures significantly affect a company's strategic decision-making, especially regarding sustainability matters. In the literature, a concentration of ownership is generally defined as holding around 20% of total voting power (Cruz et al., 2014). This thesis will adopt the same threshold, implying that when a specific ownership type is referenced, it indicates a voting power of 20% or more. 32 In previous research, the focus has mainly been on for-profit companies when examining how ownership structures relate to ESG performance, leaving out non-profit organiza- tions (Faller & Zu Knyphausen-Aufseß, 2018; Gillan et al., 2021). Like previous research, this thesis will concentrate on three for-profit ownership types commonly mentioned in the literature: family, institutional, and state ownership. Drawing on past literature, this chapter will review these different ownership types and research on how different own- ership types influence ESG and CSR initiatives and performance. 4.1 Family ownership Family-owned businesses have distinct characteristics that can influence their behavior (Abeysekera & Fernando, 2020). First, family owners often have concentrated portfolios due to focus on companies they have founded (R. C. Anderson & Reeb, 2003; Cheng, 2014). As a result, much of their wealth is tied to the success of their company. Second, family owners are usually perceived to have a longer investment horizon compared to other investors (Le Breton–Miller & Miller, 2006). Another potential issue for family owners is agency problems related to perceived socioeconomic benefits and the exten- sion of the owner's personality to their company (Berrone et al., 2010). Family compa- nies are also a global phenomenon; it is documented that a third of S&P 500 companies, more than half of East Asian corporations, 44% of European companies, and more than a half of Nordic listed companies have a family owner as the single largest shareholder (R. C. Anderson & Reeb, 2003; Claessens et al., 2000; Faccio & Lang, 2002; Lekvall, 2018). Decisions in ESG matters in family-owned companies are strongly influenced by owners’ values, especially as the owners tend to be top executives (Cronqvist & Nilsson, 2003; Cruz et al., 2014). The relation is expected to be caused by either risk appetite (Abey- sekera & Fernando, 2020) or generational thinking (Cruz et al., 2014). Transgenerational reasons are suggested to be the single most significant reason for family owners to invest in CSR initiatives, which pairs well with legitimacy theory and the implied view of a long- 33 term investment horizon (Lamb & Butler, 2016; Hummel & Jobst, 2021; Le Breton–Miller & Miller, 2006). Multiple previous research papers have focused on studying ESG performance and fam- ily ownership. Most previous papers have focused on samples from a specific geograph- ical market, such as Asia (e.g., Cheng, 2014), the United States (e.g., Lamb & Butler, 2018), or Europe (e.g., Cruz et al., 2014). Some have also conducted studies with a global per- spective (e.g., Abeysekera & Fernando, 2020). Studies regarding family equity ownership can be roughly divided into three categories: positive relationship, negative relationship, or inconclusive findings (Faller & Zu Knyphausen-Aufseß, 2018). The following three par- agraphs will discuss these findings thoroughly. Multiple studies on a U.S sample find that family ownership greatly enhances environ- mental performance due to concerns about the legitimacy and preservation of the own- ing family’s reputation and social circles (Dyer & Whetten, 2006; Lamb & Butler, 2018; McGuire et al., 2012). Dyer and Whetten (2006) used S&P 500 company data and found non-significant data surrounding CSR strength or initiatives in family-owned companies. However, they find statistically significantly lower CSR concerns in these companies, sug- gesting that family companies are not eager to undertake projects that may cause them reputational harm. Another point they make is the special connection between family company managers; although they can quit or sell their shares, they cannot exit their families. Later studies by Lamb and Butler (2018) also support this finding. On the other hand, Berrone et al. (2010) found better environmental performance in family compa- nies, which is also supported by later studies, such as Lamb and Butler (2018) and Abey- sekera et al., 2020). Cruz et al. (2014) analyze European public stock market data and suggest that CSR con- tributions of family-owned companies are complex. They argue that while European family companies tend to be cautious in addressing the claims of internal stakeholders, such as employees, they are more responsive to external pressures and concerns, 34 particularly environmental issues. Cruz et al. (2014) explain that internal stakeholders are often neglected as family companies prefer to retain internal control of their opera- tions. However, research by Litz and Stewart (2000) indicates that family companies are actively involved in addressing community issues at the local level. Lamb and Butler (2018) find a negative correlation between CSR performance in family companies and the growth of institutional ownership. This issue arises from the differing goals of share- holders: family owners prioritize the long-term socioeconomic value of their businesses, while institutional investors focus on maximizing shareholder returns. CSR investments are thought to be particularly valuable for managers if the CEO is a member of the found- ing family (Lamb & Butler, 2018). Considering the potential agency issues involved in ESG/CSR choices and whether these choices benefit shareholders equally, Abeysekera and Fernando (2020) hypothesize that family and non-family companies differ in matters regarding sustainability choices. They expect that the managers of family companies would likely align their decision-making with shareholder return maximization due to the lack of diversification by controlling families. Similar findings have been reported earlier by Anderson & Reeb (2003) and Cheng (2014). However, researchers also point out that an alternative hypothesis exists as conflicts can arise between shareholders; in this case, the family companies could make decisions that do not maximize wealth for all shareholders. Abeysekera and Fer- nando (2020) examine this hypothesis by measuring numerical ESG/CSR performance using the KLD -database environmental rankings. They conclude that family companies are more responsible and take stakeholders more into account in their decisions about environmental issues than non-family companies. Consistent with this view, Gillan et al. (2020) present evidence suggesting that Swedish family companies cater to investor de- mand for environmental investment but lack the ability to better their social perfor- mance. Contrasting the positive findings, McGuire et al. (2012) found that family companies tend to have poorer social performance than non-family companies, although their https://www-sciencedirect-com.proxy.uwasa.fi/topics/economics-econometrics-and-finance/wealth 35 sustainability performance is comparable to that of other companies. Similarly, El Ghoul et al. (2016) find evidence that suggests that family-owned companies exhibit lower ESG performance than other companies by studying companies in nine East Asian countries. The authors argue that this discrepancy is consistent with the principal-agent problem, which highlights the differing interests between family owners and other stakeholders. Additionally, they provide evidence indicating that family-controlled companies with poorer ESG performance face more significant agency problems and operate in countries with weaker institutional frameworks. Outright negative relationships between ESG/CSR ratings and family-owned businesses are found by Mackenzie et al. (2013) and Rees and Rodionova (2013). Both papers find a negative relationship between family ownership and ESG ratings using global data across over twenty countries. Both conclude that in place of CSR engagement, family companies are more focused on financial results as their portfolios are less diversified when compared to other investor types. It is speculated that family shareholders may benefit from sustainability investments as they may adversely affect short-term financial performance (Mackenzie et al., 2013; Rees & Rodionova, 2013). Faller and Zu Knyphau- sen-Aufseß (2018) argue that conflicting results suggest a problem with data availability when analyzing family businesses from a global perspective. For instance, data from smaller companies often rely on sources like interviews or local surveys, which makes it challenging to compare results and variables across different studies. Also, differentia- tion between ownership types in studies can sometimes be challenging. Overall, most studies support the view that family companies have stronger engage- ment with CSR/ESG issues, especially when it is related to the legitimacy or reputation of family-owned companies, which, given the long-term or transgenerational time hori- zon of families, are often expected in the hypothesis of studies in the area. In the ana- lyzed studies, such as El Ghoul et al. (2016) and Abeysekera et al. (2020), it is noted that global studies are difficult as country-specific factors are harder to examine. 36 4.2 Institutional ownership Institutional investors are the largest group of owners in securities worldwide (Gillan et al., 2021). An institutional investor is an organization or a company that invests money on behalf of others. Some of the largest institutional investors in the world are mutual funds, insurance companies, pension funds, or hedge funds. A distinct trait of institu- tional investors is that they often sell or buy substantial amounts of diversified securities (Dam & Scholtens, 2013). An institutional investor is distinguished from other types of investors by being considered more finance savvy and facing less protection and more regulation from authorities. Due to their significant presence in global securities, studying institutional investors can help understand shareholder preferences and actions toward sustainability initiatives (Chen et al., 2020; Gillan et al., 2021). While sustainability is a growing trend, institu- tional investors have differing attitudes toward CSR policies (Chen et al., 2020). In aca- demic studies, institutional investors are often treated as a homogenous group, which is argued not to be the case, as different institutions have differing goals (Gillan & Starks, 2000). For example, norm-constrained institutions such as pension funds are found to negatively screen their investments and avoid “sinful” industries such as gambling, to- bacco, or alcohol (Hong & Kacperczyk, 2009)Another example from the same authors is that mutual fund managers’ personal beliefs and political stance affect their portfolios— the more the manager supports politicians on the left, the less they invest in socially irresponsible equities. This is consistent with Di Giuli and Kostovektsy's (2014) findings, which found that managers in the United States voting for the Democratic party are more likely to prefer undertaking ESG/CSR investments. However, Fernando et al. (2010) demonstrate that institutions are less likely to invest in green and toxic companies and prefer environmentally neutral companies. Similarly, Nofsinger et al. (2019) and Fernando et al. (2017) point out that institutional investors avoid stocks with sustainability concerns but do not overly invest in ESG leaders, 37 indicating that a negative screening approach is more likely used than positive screening strategies. It might also be that ESG leaders attract more long-term institutional investors (Starks et al., 2017). According to Fernando et al. (2017), these findings lead to conse- quences where corporate policies that mitigate environmental risk create shareholder value. However, research has shown that environmentally friendly approaches do not create additional shareholder value on their own (Fernando et al., 2017). Similar findings were made with global data across 28 countries; Tang and Zhang (2021) discovered that only the first announcement of green bond issuance is statistically significant with an increase in institutional ownership percentage and future abnormal shareholder returns, while subsequent issuances have no significant effect. Moreover, the first announce- ment of green bond issuances significantly increases stock liquidity globally, allowing for more institutional investments (Tang & Zhang, 2021). While institutional investors might have differing tastes in CSR, they are not silent inves- tors (Chen et al., 2020). Studying the 3000 largest publicly traded companies in the United States, Chen et al. (2020) find that institutional investors can pressure managers towards investing more in social goodness and that the higher the level of institutional investors, the better the overall ESG rating in the future. Furthermore, Chen et al. (2020) and Gollier and Pouget (2014) show that these active institutions drive sustainability in- itiatives with a positive financial effect. Specifically, institutions can reduce the risk of making decisions that could lead to regulatory penalties or lawsuits infringing CSR re- quirements (Gollier & Pouget, 2014). To support these arguments, Aksoy et al. (2020) note that it is in the interest of institutional investors to acquire stakes in companies that perform well on social issues. Overall, results on institutional ownership in terms of ESG matters are mixed, and there- fore further examination of the preferences and influences of institutional decision-mak- ing regarding ESG/CSR activities is needed. Particularly, studying the relationship be- tween different types of institutional owners and ESG is needed. Based on the studies analyzed, it can be concluded that the relationship between ESG/CSR initiatives and 38 institutional owners is dynamic and highly dependent on sustainability concerns rather than outstanding performance. 4.3 Government ownership Past research has often concluded that governmental agencies and institutional inves- tors share similar characteristics and are often grouped together (Earnhart & Lizal, 2006; Fan et al., 2021; Motta & Uchida, 2018). As multiple past studies have included both in the same category, segregation between these ownership types can be difficult. How- ever, in this thesis, governmental agencies are identified as a separate group of investors as they have broader interests and often politics involved in their decision-making (Earn- hart & Lizal, 2006; Faller & Zu Knyphausen-Aufseß, 2018; Gillan et al., 2021). An example of broader interest besides profit-seeking is pressuring portfolio companies to adopt governmental agendas, which can differ between governments and years (Faller & Knyphausen-Aufseß, 2018). Scholars have been especially concerned that man- agers in state-owned enterprises are often poised to be politically aligned, have low in- centives, and are poorly monitored by boards, which are also packed with politicians (Shleifer & Vishny, 1986). Even today, state-owned enterprises are often seen as policy tools to achieve political objectives (Boubakri et al., 2019). On the other hand, Gillan et al. (2021) argue that governments have a superior ability to handle environmental or societal issues. These companies may be better positioned to deal with market turmoil and externalities caused by market-wide issues. For example, governments have supe- rior access to capital via taxation and loans. Additionally, state-owned businesses can be better positioned to address climate change issues by implementing and forcing stronger policies on companies under their control (Hsu et al., 2023;Hart & Zingales, 2017). These policies can be essential in emerging markets, such as China, Russia, or Brazil, where state-owned companies represent many publicly listed companies. 39 Even though multiple studies have previously grouped government owners and other institutions, some studies focus on government investments and CSR/ESG issues globally. For example, mainly studying state-owned listed companies in 45 countries, it is found that state-owned companies are more responsive towards environmental issues (Hsu et al., 2023). Government-owned companies in emerging economies are especially respon- sive, often lacking robust regulatory frameworks or secure access to energy. Hsu et al. (2023) find that government-controlled companies can engage in environmental agen- das without adversely affecting shareholder value. While these companies often engage more in social issues, they cannot perform better in governance performance (Hsu et al., 2023). These effects are found only in companies where the local government owns a direct or indirect stake, not in companies owned by sovereign wealth funds or public pensions. Consistent with Hsu et al., it is found that after privatization, state-owned companies had higher ESG scores than their peer group (Boubakri et al., 2019). Using a sample of almost 19,000 companies across 41 countries, Boubakri et al. (2019) conclude that CSR intensity is much higher in previously state-owned companies even after changing own- ers to a private investor. Furthermore, state-owned companies are found to invest more in CSR activities, especially in developing economies with developing regulations. Even after privatization, these companies are, in aggregate, more aligned with sustainable business practices. However, Schleifer and Vishny (1986) argued that state-owned enter- prises have distorted objectives and are geared toward political goals, which can hurt smaller shareholders. The findings from various studies, including those by Hsu et al. (2023) and Boubkari et al. (2019), provide evidence that state-owned enterprises play a crucial role in driving environmental and social agendas globally. This is particularly present in emerging mar- kets, where these companies respond to environmental challenges and contribute to the broader goals of sustainable development without necessarily compromising share- holder value. However, the challenges associated with political alignment and 40 governance issues within SOEs cannot be overlooked. Looking forward, the evolving role of state-owned enterprises in addressing global challenges such as climate change and social inequality calls for further research. Specifically, understanding how these enter- prises can navigate the broader demand for corporate disclosure and transparency (McGuinness et al., 2017). 41 5 Data and methodology The research design for this thesis employs a quantitative method. This approach is con- sistent with previous research on similar topics that also primarily use quantitative meth- ods (Abeysekera & Fernando, 2020; Chen et al., 2020; Fernando et al., 2017). In line with other studies in the field, this method will be utilized to effectively investigate how own- ership concentration may influence a company's ESG performance. 5.1 Research design and data sample This study focuses on the association between ownership concentration and ESG in a Nordic context. The sample of companies used in this study consists of those included in the Europe Stoxx 600 index and listed and domiciled in the Nordics, excluding Iceland. The Europe Stoxx 600 index is a multi-country stock index comprising 600 large to me- dium-sized companies in the European continent, covering almost 90% of the market capitalization in European stock markets. The index is operated by STOXX LTD, part of Deutsche Börse Group (Stoxx Ltd, n.d.). The Nordic companies within this index are well- established companies operating in diverse industries with differing financial profiles. The oldest company by founding date in the sample is the Norwegian company Orkla ASA, founded in 1654 and operating in the consumer branding and consumer goods sec- tor. The most recently established company in the sample is a Swedish property portal operator called Hemnet AB, founded in 2016. The company sample consists of 109 companies in the Europe Stoxx 600 index. Like other studies in the field, companies operating in the financial industry, such as banks, invest- ment companies, and insurance companies, are excluded from the selection due to in- dustry-specific regulation regarding ESG and their unique business characteristics (Alshbili et al., 2019; Abeysekera & Fernando, 2020). For example, companies operating in the financial industry are often highly leveraged and rely on interest- or investment income (Alshbili et al., 2019; Abeysekera & Fernando, 2020). Excluding the financial 42 industry leaves out 13 companies. Three other companies were excluded from the sam- ple due to missing ESG information. This leaves the total company sample at 93 compa- nies and 558 company observations across six fiscal years from 2018 to 2023. The geo- graphical location is the following: Denmark, 22 companies (23% of the sample), Finland 19 companies (20% of the sample), Norway 10 companies (11% of the sample), and Swe- den, 43 (46% of the sample). The distribution of companies within the sample has a high concentration of Swedish companies, which is unsurprising because, according to DataStream, Nasdaq Stockholm currently holds the highest number of unique listed companies and the largest overall market capitalization of the four markets. The data for the selected companies is collected through primary and secondary sources, including official financial statements, the companies' official websites, and data pro- vided by the Thomson Reuters Eikon ESG database, DataStream, and Orbis database. Annual reports of public companies are considered a reliable source of information and have been used in many finance studies. Thomson Reuters Eikon ESG database, DataStream, and Orbis database are also used by previous studies examining ESG per- formance and financial data in the Nordics (Ravid & Sekerci, 2020). The selected timeframe corresponds with the implementation of the European Commis- sion’s Non-Financial Reporting Directive. In line with the adoption of the NFRD, it is evi- dent that most companies did not have accessible ESG Scoring in the Thomson Reuters Eikon ESG database prior to 2018. The year 2023 is chosen because the 2024 fiscal year is still in progress during the data collection period in September 2024, and the update of ESG scores does not happen in real time. 43 Table 1. Description of the sample selection. 5.2 Regression variables 5.2.1 Dependent variable This thesis studies the impact of differences between the largest shareholders on a com- pany’s ESG performance. The dependent variables in this analysis are the combined ESG score and scores for the different pillars of ESG (environmental pillar, social pillar, and governance pillar) collected through the Thomson Reuters Eikon ESG database. The ESG score is used as a proxy for the company’s sustainability performance and is collected separately for every company and year within the sample data. The ESG score provided by LSEG, Thomson Reuters, KLD STATS, or other competing databases is used by most of the research studying CSR performance associated with different ownership types, e.g., Ioannou and Serafeim (2010), Cruz et al. (2014), Abeysekera and Fernando (2020). The Thomson Reuters Eikon ESG database was chosen for this thesis due to its availability. The ESG score provided by the Eikon ESG database (formerly known as Refinitiv) is a nu- merical score between zero (worst) and one hundred (best). It is computed by measuring over six hundred indicators with varying weights, such as Resource use, Workforce Di- versity, Community Engagement, Management measures, and CSR initiatives across all 44 three pillars of ESG (Refinitiv, 2022). The weight of these indicators is visually presented in the figure below. According to Thomson Reuters, all data within the database is de- rived from publicly accessible company disclosures, commitments, performance, and news coverage. In addition to the combined ESG score, the separate components of (E) environmental, (S) social, and (G) governance scores are all used to understand better whether differing dominant owner types perform worse or better on a particular pillar. For example, Earnhart and Lizal (2006) identify a negative correlation between govern- ment ownership and environmental performance, and Ioannou and Serafeim (2010) find a negative association between institutional owners and social performance, while Starks (2020) argue that Swedish family companies fair best when it comes to environ- mental matters. Figure 4. Categories of ESG Pillars in the Eikon database. 5.2.2 Independent Variables The independent variables in this thesis include different variables for the four owner- ship types discussed earlier. Like previous studies, the association between a specific ownership type and ESG performance will utilize two approaches (Abeysekera & 45 Fernando, 2020; Cruz et al., 2014; Dam & Scholtens, 2013; Fan et al., 2021; Lamb & Butler, 2018; Rees & Rodionova, 2013; Vural, 2018). First, dummy variables will be used to rep- resent the presence of a particular controlling shareholder, and second, a numerical var- iable to represent the percentage of shares held by the dominant shareholder. Like Vural (2018), this study will get around the use of dual-class shares in many sample companies by calculating the ownership percentage based on voting rights and not economic rights. This is done by dividing the total number of votes in a company by the total number of votes held by the largest shareholder. While the majority of studies ownership studies focus solely on the direct shareholders (e.g., Alshbili et al., 2019; Di Giuli & Kostovetsky, 2014; Nofsinger et al., 2019), overlooking the presence of an ultimate shareholder, who indirectly controls the companies via an investment company or a complicated owner- ship structure where voting rights are given to the ultimate holder. Due to the large pres- ence of complex ownership structures in the Nordics, it is necessary to trace the ultimate owner beyond the direct largest shareholder in a company. This is done using the Orbis database and open-source resources such as news outlets and corporate disclosures. An example of a company seemingly controlled by an institutional investor but ultimately controlled by a family investor is the Swedish security solutions provider Securitas AB, whose largest shareholder is Latour AB, which the Douglas family directly controls with 80% voting rights. Similar to earlier studies such as Cruz et al. (2014) and Fan et al. (2021), this study will utilize dummy variables to express the presence of a dominant owner and their type. First, a dummy variable will express whether a company has a major shareholder with over 20 percent voting rights. This variable will take the value of 1 if a dominant share- holder is found and 0 otherwise. Similarly, this study will use “family,” “institution,” and “government” dummies to indicate whether the largest shareholder is found to be a family investor, an institution, or a government entity. For example, suppose a company in the sample has the largest shareholder with 30% voting rights and is a family investor. In that case, it will get a value of 1 for the major shareholder dummy and family dummy and a value of zero for the institution dummy and government dummy variables. 46 In addition to ownership dummies, a variable indicating the largest owner-holding per- centage will be used. The potential value for this variable ranges between 0,05 and 1, indicating the total voting rights owned by the largest shareholder with a 5% threshold as shareholders are not required to publicly inform on holdings under a 5% stake in com- panies whose stock is traded on exchanges located in ETA countries (arvopaperimarkki- nalaki, the Finnish Securities Markets Act, 746/2012, Chapter 9 Section 1). This variable is used to study the potential relationship between various levels of control by the ulti- mate owner. Motta and Uchida (2018) and Aksoy et al. (2020) use a similar variable. Due to difficulties in obtaining historical ownership data, this study will consider only the most up-to-date ownership information and will apply it to all the years in this study (2018 – 2023). Ownership data is hand-collected from the sample companies' investor or corporate governance webpages, of which 100% were updated within quarter three of 2024. After collecting the ownership data, it is concluded that 65% of the company sample has an identifiable dominant shareholder. Of these dominant shareholders, 32 (36%) are family owners, 17 (19%) are institutional investors, and 9 (10%) are govern- ment held. A visual presentation of ownership characteristics within the selected sample is presented in the below figure. 47 Figure 5. Pie graph of the distribution of ownership in sample companies. 5.2.3 Control Variables This thesis employs several control variables often used in financial and ESG research to control known influences in ESG performance. Even though the main point of this study is the ESG performance of Nordic companies, it is important to control company-related characteristics that are known to influence most company-level performance in many of the phenomena studied in finance literature. For example, a company’s profitability is a well-known factor often associated with a company’s financial health and the ability to undertake often expensive sustainability-related investment (Dam & Scholtens, 2013; Flammer, 2021; Motta & Uchida, 2018; Rees & Rodionova, 2013)This study uses Return on Assets (ROA) as the profitability metric, one of the most used profitability metrics in finance studies, and used as a control variable in almost all the ESG-related studies cited in this thesis to control company profitability. 𝑅𝑒𝑡𝑢𝑟𝑛 𝑜𝑛 𝐴𝑠𝑠𝑒𝑡𝑠 = 𝑁𝑒𝑡 𝐼𝑛𝑐𝑜𝑚𝑒 𝑇𝑜𝑡𝑎𝑙 𝐴𝑠𝑠𝑒𝑡𝑠 48 Total company size is often applied as a control variable in finance studies, as scholars believe that the size of a company is often related to their ability to take on certain in- vestments, such as expensive projects, to make their supply chains more transparent and reliable (Dam & Scholtens, 2013; Cruz et al; 2016; Lamb & Butler, 2016; Abeysekera & Fernando, 2020). Companies larger in size might also benefit more from CSR invest- ments and be more pressurized by larger stakeholder groups to invest in sustainability initiatives (Kabir & Thai, 2021; Uyar et al., 2020) According to Kabir and Thai (2021), com- pany size also influences the number of initiatives companies undertake to address var- ious sustainability concerns, which in turn results in better sustainability performance over time. On the other hand, some opposing views from the literature exist, suggesting that larger companies might also due to their size be more exposed to CSR-related con- cerns, which can accelerate discussion and publicity related to these concerns (Bose et al., 2017; Dam & Scholtens, 2013). Most previous studies have, therefore, controlled for company size using the natural logarithmic of total assets each year. In this study, all financial statement data is translated into Euro using end-of-year values from the Orbis – database. This is done to account for differences in currency buying power. For example, as of 10/18/2024, one Euro amounted to 11,42 SEK. Euros are used as all the countries have significant connections to the Euro area in terms of trade and international agree- ments. Furthermore, the Euro is the most used currency in international transactions, compared to the currencies used in these sample countries, and the second most used currency after the US dollar worldwide (Statista, 2024). 𝑆𝑖𝑧𝑒 = 𝑁𝑎𝑡𝑢𝑟𝑎𝑙 𝑙𝑜𝑔 (𝑇𝑜𝑡𝑎𝑙 𝐴𝑠𝑠𝑒𝑡𝑠) The third control variable in this study is a company’s leverage to proxy for an individual company’s financial position, and it is used in various research papers cited in this thesis (e.g., Bose et al., 2017; Motta & Uchida, 2018; Abeysekera & Fernando, 2020 and Flam- mer, 2021). In academic papers, a company's debt level influences a company's 49 performance, such as shareholder returns, stock price volatility, and CSR performance. Additionally, high leverage is associated with the ability and willingness of top manage- ment risk-taking (Ducassy & Montandrau, 2015) In this thesis, the debt–to–equity (D/E) ratio will be used as a proxy for a company's financial health and use of leverage. The D/E ratio is calculated as the remainder of the company’s book value of all liabilities and the book value of the shareholder’s equity. The same financial metric is used in ESG stud- ies such as Flammer (2021) and Dam and Scholtens (2012). 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒 = 𝑇𝑜𝑡𝑎𝑙 𝐿𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠 𝑇𝑜𝑡𝑎𝑙 𝑆ℎ𝑎𝑟𝑒ℎ𝑜𝑙𝑑𝑒𝑟𝑠 𝐸𝑞𝑢𝑖𝑡𝑦 The last control variable in this study is company age, a common variable applied in eco- nomic research to control for differences in companies’ natural lifespan; for example, younger companies might be less established than older ones. Therefore, they may need to focus on other investments and less on sustainability efforts (Vural, 2018). Similarly, older companies may be more concerned about their image and try to be more environ- mentally friendly. On the other hand, older companies may be inclined to use older, more environmentally harmful technologies (Abeysekera & Fernando, 2020; Borghesi et al., 2014). Like other studies that control for company age, such as Borghesi et al. (2014), this study uses the natural logarithmic of years since the company’s inception or since a merger or demerger, which marks a significant turning point for some companies. 𝐶𝑜𝑚𝑝𝑎𝑛𝑦 𝑎𝑔𝑒 = 𝑁𝑎𝑡𝑢𝑟𝑎𝑙 log(𝑦𝑒𝑎𝑟𝑠 𝑠𝑖𝑛𝑐𝑒 𝑖𝑛𝑐𝑒𝑝𝑡𝑖𝑜𝑛) Table 2. shows the descriptive statistics for this study's main variables. The main points of interest are the dependent variables: the ESG score and each pillar, which have a mean and median significantly above fifty, and the mean of possible ESG scores in the Eikon database. However, as the standard deviations show, there are significant differ- ences between companies, of which the most deviating one is the Environment pillar. As discussed in Chapter 3.5, industries can significantly affect environmental prowess, while the country's influence on Nordic companies is likely relatively small due to their similar 50 historical backgrounds and political and economic systems (Andersen, 2004; Ioannou & Serafeim, 2012; Bose et al., 2017; Simon, 2017; Liu et al., 2019). When it comes to the control variables, it is evident that the sample of companies has heterogeneity in them with largely variable ages, profitability, balance sheet sizes, and use of outside financing. Also, the ownership characteristics vary between sample companies, with some having their largest block holder in the 5% ownership/vote level and the most concentrated company having an owner with 44% of all outstanding votes in their companies. Table 2. Descriptive statistics of regression variables. 5.3 Regression models This study uses a multiple regression model to estimate associations between ownership concentration and sustainability performance. Multiple regression analyses are often used in analysis where a dependent variable (ESG score and each pillar in this study) y is related to two or more independent variables (D. R. Anderson & Sweeney, 2014, p. 684). According to Anderson and Sweeny (2014), multiple regression models describe how the dependent variable y is related to both independent variables and an error term to catch all other factors that are not accounted for, which can influence the dependent variable y. In economic analysis, the error term accounts for endogeneity (Antonakis et al., 2010). 51 Endogeneity refers to the general problem caused by relative information not explained by the chosen explanatory variables. It needs to be corrected to improve the model's explanatory powers. According to Antonakis et al. (2010), the main points causing en- dogeneity include omitting important variables from regression models, simultaneity bias where the outcome variable also explains xi, and measurement errors in the regres- sors, such as using imperfect proxy variables. In this thesis, endogeneity problems may arise between ownership variables and ESG and ESG pillar performance. For example, the hypothesis in this thesis states that owners impact sustainability performance. How- ever, it can be a plausible explanation that greater or worse performance in these areas attracts different owners. The multiple regression model takes the following form: 𝑦 = 𝛽0 + 𝛽1𝑥1 + 𝛽2𝑥2 + ⋯ + 𝛽𝑛𝑥𝑛 + 𝜖 In the model, the 𝛽0, 𝛽1, 𝛽2 and 𝛽𝑛 denote the independent and control variables, while the Greek letter epsilon is the error term or the random variable (Anderson & Sweeney, 2014, p. 686). The model assumes that the dependent variable y is a linear function of the beta coefficients with known values of x plus the error term to account for factors not explained by chosen control variables. The multiple regression model is thought to describe reality accurately and potential predictors using panel data, which is why it has been employed in past research studying either ESG performance or ownership (Flam- mer, 2021; Hettler et al., 2021; McGuinness et al., 2017; Ravid & Sekerci, 2020; Vural, 2018). Like Mcguiness et al. (2017), Vular (2018), and Flammer (2021), this study em- ploys year and industry-fixed effects. Industry fixed effects are chosen instead of com- pany fixed effects, as the majority shareholders and their voting power are considered time-invariant, as discussed in Chapter 5.2.2. The company industries are categorized into eight categories based on their 4-digit NACE code. The industry categories and com- panies within categories are: 52 Figure 6. Industry breakdown and distribution by NACE codes. Regressions (1.1) to (1.4) are used to answer whether there is an association between ownership concentration and ESG or E, S, and G pillar performance. These regressions use the presence of a controlling shareholder (over 20% of votes) as a dummy variable. The difference between regression models 1.1 to 1.4 is that the dependent variable changes from ESG combined score to each pillar on its own. This is done to better un- derstand ESG performance among tightly held companies. 𝐸𝑆𝐺𝑖,𝑡 = 𝛽0 + 𝛽1 × 𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑖,𝑡 + 𝛽2 × 𝑂𝑤𝑛𝑒𝑟𝑠ℎ𝑖𝑝𝑖,𝑡 + 𝛽3 × 𝐴𝑔𝑒𝑖,𝑡 + 𝛽4 × 𝑃𝑟𝑜𝑓𝑖𝑡𝑎𝑏𝑖𝑙𝑖𝑡𝑦𝑖,𝑡 + 𝛽5 × 𝑆𝑖𝑧𝑒𝑖,𝑡 + 𝛽6 × 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 + 𝑌𝑒𝑎𝑟𝑓𝑒 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝑓𝑒 + 𝜀𝑖,𝑡 (1.1) Where, 53 ESG = the Combined ESG score of a company at time t Control = dummy variable with value of 1 if a controlling shareholder is found and 0 otherwise. Ownership = The percentage of total votes held by the largest owner in company i at time t Age = Natural logarithmic of years since a specific companies inception. Profitability = Percentage value of return on assets with year-end values. Size = Natural logarithmic of total assets with year-end values Year = Year dummy Industry = Industry dummy i,t = Company i at time t fe = fixed effects ϵ = Error term Regressions (1.2) to (1.4) aim to answer whether specific pillars (Environmental, Social or Governance) are influenced by the presence of a dominant owner. In these regres- sions 𝑌𝑡 𝑖 represent the score of company i at time t for these different components of the total ESG score. Regressions (1.2) to (1.4) follow the following formula: 𝑌𝑡 𝑖 = 𝛽0 + 𝛽1 × 𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑖,𝑡 + 𝛽2 × 𝑂𝑤𝑛𝑒𝑟𝑠ℎ𝑖𝑝𝑖,𝑡 + 𝛽3 × 𝐴𝑔𝑒𝑖,𝑡 + 𝛽4 × 𝑃𝑟𝑜𝑓𝑖𝑡𝑎𝑏𝑖𝑙𝑖𝑡𝑦𝑖,𝑡 + 𝛽5 × 𝑆𝑖𝑧𝑒𝑖,𝑡 + 𝛽6 × 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 + 𝑌𝑒𝑎𝑟𝑓𝑒 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝑓𝑒 + 𝜀𝑖,𝑡 (1.2, 1.3, 1.4) Where, 𝑌𝑡 𝑖 = ESG performance measured with different pillars of ESG scores of the company i at time t. Following the regressions on ownership concentration and ESG performance, regres- sions (2.1) to (2.4), will focus on family ownership through a family dummy, which gets the value of 1 if there is a dominant owner in company i, and the shareholder is classified 54 as a family investor and a value of zero otherwise. Regressions (3.1) to (3.4) will focus on institutional owners, with similar regressions, and finally, regressions (4.1.) to (4.4) will focus on government-held companies. 𝐸𝑆𝐺𝑖,𝑡 = 𝛽0 + 𝛽1 × 𝑂𝑤𝑛𝑒𝑟𝑡𝑦𝑝𝑒𝑖,𝑡 + 𝛽2 × 𝑂𝑤𝑛𝑒𝑟𝑠ℎ𝑖𝑝𝑖,𝑡 + 𝛽3 × 𝐴𝑔𝑒𝑖,𝑡 + 𝛽4 × 𝑃𝑟𝑜𝑓𝑖𝑡𝑎𝑏𝑖𝑙𝑖𝑡𝑦𝑖,𝑡 + 𝛽5 × 𝑆𝑖𝑧𝑒𝑖,𝑡 + 𝛽6 × 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 + 𝑌𝑒𝑎𝑟𝑓𝑒 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝑓𝑒 + 𝜀𝑖,𝑡 (2.1,3.1,4.1) Where, Owner type = Dummy variable representing each dominant shareholder type. For re- gression (2.1), it gains the value of 1 if the largest shareholder is a family investor; for (3.1), the value is 1 for institutional investors, and for (4.1) for government holdings and zero if the dominant owner is not the specified type. 𝑌𝑡 𝑖 = 𝛽0 + 𝛽1 × 𝑂𝑤𝑛𝑒𝑟𝑡𝑦𝑝𝑒𝑖,𝑡 + 𝛽2 × 𝑂𝑤𝑛𝑒𝑟𝑠ℎ𝑖𝑝𝑖,𝑡 + 𝛽3 × 𝐴𝑔𝑒𝑖,𝑡 + 𝛽4 × 𝑃𝑟𝑜𝑓𝑖𝑡𝑎𝑏𝑖𝑙𝑖𝑡𝑦𝑖,𝑡 + 𝛽5 × 𝑆𝑖𝑧𝑒𝑖,𝑡 + 𝛽6 × 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 + 𝑌𝑒𝑎𝑟𝑓𝑒 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝑓𝑒 + 𝜀𝑖,𝑡 (2.2 to 2.4, 3.2 to 3.4 and 4.2 to 4.4) Where, 𝑌𝑡 𝑖 = ESG performance is measured with different pillars of the combined ESG score of the company i at time t. 55 Table 3. Recap of variables used. 5.4 Correlation coefficients In the empirical analysis, correlation coefficients, especially Pearson correlation coeffi- cients, are the essential implications of connections between selected variables (Ander- son & Sweeney, 2014, pp. 140-149). Correlation refers to the statistical relationships be- tween variables, and it is a statistical measure that illustrates the relationship, but not necessarily causality, between a pair of variables and the unbiasedness of sample data (Anderson & Sweeney, 2014, pp. 140-149). The correlation coefficient between two var- iables ranges between -1 and 1, with a negative (positive) one corresponding to a perfect negative (positive) correlation – that is, if one variable increases by one unit, another decreases (increases) by one unit. To ensure the validity of regression data, no independ- ent variables should be constant, nor should there be any perfect linear, multicollinearity relationship between variables; generally, correlation coefficients of under 0,7 are seen to conclude no multicollinearity (Anderson & Sweeney, 2014, pp. 140-149). Thus, an im- portant observation in the correlation matrix below is that there is no collinearity be- tween independent variables. 56 Table 4. Pearsons's correlation matrix with *** denoting a 95% confidence level. 57 5.5 Hausman test and limitations of the data sample In this thesis, the fixed effects model is chosen for the regression analyses, as it is com- monly utilized in panel data studies. As previously discussed, the time-invariance of the largest owners and their voting power supports this choice. A Durbin-Wu-Hausman test is conducted to determine whether the fixed effects model is more appropriate than the random effects model for this study. The Hausman test assesses whether the error terms in a regression model are correlated with the independent variables, which can heighten the risk of omitted variable bias (Hausman, 1978). In this test, the null hypothesis is that the random effects model is the preferred approach; rejecting the null hypothesis indi- cates that the fixed effects model is more appropriate for the data analysis. The Hausman test results for all regression models produced a p-value below the 95% confidence threshold, reinforcing the decision to use fixed effects over random effects in this analy- sis. Last, before discussing the regression results, it should be acknowledged that there are some limitations in the data selection, which can cause this thesis not to give a fully comprehensive view of the sustainability performance of Nordic companies. First, the