ConfliCt of interests among shareholders – does it refer to dividend deCisions?

research background: Agency theory refers to the conflict of interests between different groups involved in the business. One aspect of the agency theory is conflict among shareholders. One of the areas of potential conflict is dividend payment. Purpose: The goal of our paper was to identify the impact of voting behaviour of shareholders during Annual General Meetings on dividend decisions. We make a comparison between big and small companies. research methodology: We included the 20 biggest and the 20 smallest companies from the Warsaw Stock Exchange (WSE) in our sample over the 2010–2022 period. We collected 413 observations (shareholders’ presence, voting results from annual general meetings, and financial data). We use statistical tests, a correlation analysis, and a regression analysis. results: We found that in small companies there is conflict among large shareholders on dividend payout decisions, while in big companies small shareholders show their disappointment by voting against or

or cannot be solved (Kazanský, Trifunovic, 2021). Black, Gardner, Pierce, and Steers (2019) suggest the following four types of conflicts: goal conflict (a clash over whose goals are going to be pursued), cognitive conflict (which is evident in political debates), affective conflict (when one person's or a group's feelings or emotions are incompatible with those of others), behavioral conflict. Conflicts in organizations depending on the direction are divided into vertical conflicts, horizontal conflicts, linear and personnel conflicts.
In the process of studying the nature of conflict in organizations, and corporations, it is appropriate to consider the agency theory which focuses on interest conflict between principals and agents. Corporate conflict in finance is described by the agency theory as interests conflicts between the principal (shareholders) and agents (managers). However, the later development of the agency theory allows us to identify also another dimension of the relationship: between shareholders and debtholders and between shareholders.
One of the main areas of conflict is the dividend payout decisions. In the relationship between shareholders and managers, the generous dividend is the tool for mitigating managers. However, in the relationship between majority and minority shareholders, majority shareholders are to minimize the dividend payment (Gugler, Yurtoglu, 2003). Instead, the interests of the minority shareholder are completely opposite. The short-term perspective is important to them -to receive a guaranteed income from their participation in the corporation. Additionally, minority shareholders are aware of majority shareholders' private benefits extraction and deem higher dividends as compensation.
The goal of our paper is to identify the shareholders' presence at Annual General Meetings, and their voting behaviour, especially on the dividend decision. We compare big and small companies from the Warsaw Stock Exchange. We also wonder how the presence of large shareholders affects dividend decisions.
We included the 20 biggest and 20 smallest listed companies from the Warsaw Stock Exchange (WSE) in our sample. The data on large shareholders' presence and voting results come from Annual General Meetings that were held within the 2010-2022 period (13 years).
Data were hand-collected from current reports of the companies included in the research. We implemented a statistical test to compare the results, a correlation analysis to find the relation between variables, and a regression analysis to find the impact of large shareholders on dividend decisions.
We found that big and small companies differ significantly. There is a higher shareholders' presence, higher 1 st largest shareholder presence, and lower large shareholders presence ratio at the AGM's in big companies. There is a higher consensus among shareholders in big companies for dividend payouts but lower for financial statement acceptance. In small companies, there is higher consensus on financial statement acceptance but large shareholders are involved in voting against or abstaining from voting for dividend payout resolutions. We found that the higher the large shareholders' presence the lower was the dividend payout. This finding is in line with a previous notion that large shareholders due to wealth expropriation are not interested in high dividend payouts.
We have contributed to the existing research in several ways. Our paper contributes to the research on conflict and participation in elections. Our paper contributes also to the research on corporate governance and agency theory. Existing research refers mostly to ownership structure and dividend decisions while we think that only owners present at AGMs have an impact on dividend decisions. We found several regularities in shareholders' behaviour during Annual General Meetings. We dealt with a unique set of data as they were hand-collected. We also found a new area of conflict -minority shareholders showing their disappointment by voting against or abstaining from voting for financial statement acceptance (but not dividend payout decisions).
The rest of the paper is organized in the following way. The first section includes a literature review on conflicts and agency theory achievement on interests conflict between shareholders. The next section describes the methodology: sample, data source, and statistical methods used to verify the hypotheses. The following sections includes the main findings and discussion. The last section contains the research summary and conclusion.

literature review
The agency theory focuses on interest conflict between principals and agents (principalagent conflict, PA conflict, and type I agency conflict). One of the representatives of this theory, Jensen (1986), indicates that principals are forced to increase monitoring costs in order to prevent managers from investing in ineffective projects. In order to diminish the conflict between owners and managers, a company should pay out free cash flow in the form of dividends. Otherwise, managers will simply follow their own self-interest instead of maximizing the welfare of a company's shareholders. The conflict of interest between owners and managers results in activities having an effect on financial decisions such as dividends debt or investments. If managers are overinvesting, an increase in the dividend or debt will reduce the overinvestment and increase the market value of the firm (Tijjani, Bello, 2019).
In particular, Rozeff (1982) presents a model of optimal dividend payments in which increased dividends reduce agency costs. But Renneboog and Szilagyi (2020) perceive dividends and shareholder control more as complementary rather than substitute mechanisms in mitigating agency concerns.
Although the main concept of the agency theory describes the conflict of interests between the principal (shareholder) and agents (managers), later development of the theory led to the identification of other conflicts of interest within corporate governance area, i.e.: between shareholders: principal-principal conflict (PP conflict, type II agency conflict). Shareholders' conflict might refer to the conflict between e.g. majority and minority shareholders (La Porta et al., 1999) or between large shareholders (Cheng et al., 2018).
Large shareholders quite often have different interests from those of minority shareholders in determining the future of a company (Rathod, 2019). The conflict of interest between large and small shareholders lies in the possibility of expropriation by large shareholders, who frequently control the decisions through their appointed directors (Chang, 2003). Forms of expropriation include below-market value asset transfers to large shareholders, corporate expenditures on non-value-creating assets for large shareholders, and corporate diversification resulting in stable cash flows to large shareholders (Su et al., 2008). That is why many countries introduce legal systems protecting minority shareholder interests.
However, the minority shareholders have some benefits of large shareholders' presence -the free-rider problem. Most of the monitoring expenses are incurred by large shareholders. Minority shareholders are usually passive and not engaged in everyday company management but they benefit from the activities of large shareholders and participate in the financial results of large shareholders' activity -dividends. Gugler and Yurtoglu (2003) believe that smaller shareholders anticipate expropriation, thus they demand higher dividends. However, the interests of majority shareholders are aimed at minimizing the payout dividends to minority shareholders. An empirical analysis reveals that firms make lower dividend payouts as the voting rights of the largest shareholder increase (Faccio et al., 2001;Mancinelli, Ozkan, 2006).
The negative relationship between ownership concentration and the dividend was also proved by Khan (2006), and Renneboog and Trojanowski (2007), and Sáez and Gutiérrez (2015).
But there are studies that indicate the dividend payout relates positively to the majority shareholder's ownership percentage in the firm (Jiang et al., 2017). This finding is inconsistent with the idea that controlling shareholders reduce dividends in order to extract private benefits. Instead, the evidence suggests that such shareholders take action to mitigate the potential conflict of interest between majority and minority shareholders. In particular, controlling shareholders seem to pay high dividends in order to build a reputation for not expropriating Conflict of interests among shareholders -does it refer to dividend decisions? 21 minority shareholders. Firms that pay high dividends benefit from this enhanced trust by being able to attract greater minority investment later on. Hence, it is in the majority shareholder's best interest to voluntarily abstain from opportunism and instead adopt a minority-friendly and less myopic dividend policy (Berzins et al., 2018). Several studies confirm that large shareholders prefer to pay high cash dividends to alleviate minority shareholder concerns about expropriation (Firth et al., 2016;Jiang et al., 2019;Anh, Tuan, 2019). However, Rodrigues, Felici and Matos (2020) revealed that none of the analyzed ownership variables was significant for dividend decisions.
The reason for inconsistent findings on ownership and dividend decisions might lie in the problems associated with the high dispersion of shareholders. High dispersion leads to a situation in which none of the shareholders have the incentive to properly monitor the activities of the company (collective action problem) and conflict among large shareholders might arise (Rodrigues et al., 2020). Faccio, Lang and Young (2001) studied multiple blockholders in several East Asian countries and found that blockholders cooperate to expropriate wealth from minority shareholders by withholding dividend payouts. However, Gugler and Yurtoglu (2003) studied multiple blockholders in Germany and found that blockholders exert a monitoring function on the largest blockholder and force dividend payouts. Jiang, Cai, Jiang and Nofsinger (2019) found that firms with multiple large shareholders are more likely to pay dividends and pay large dividends. They also found that dividend payouts and large payouts are more likely when the largest shareholder needs the cooperation of other blockholders to exert control over the dividend payout policy. The agency theory suggests that the largest shareholder and dividend policy might be viewed as substitute monitoring devices. It is possible that the shareholding of the largest shareholder is negatively related to dividend payout at low levels of shareholding, yet as the shareholding increases, this relation becomes positive. Where the largest shareholder has a comparatively small interest in a company, there will be a tendency for them to monitor more actively the firm's management, and so there is less need for the use of dividends in controlling agency costs. As ownership levels of the largest shareholder increase, this creates new agency problems, resulting in the need for higher dividend levels and the external monitoring activity that accompanies them. This is largely consistent with the traditional agency theory where it is argued that ownership and dividends provide substitute monitoring devices. Truong and Heaney (2007) provide evidence to support the positive relationship between the stake of the largest shareholders and dividend payout -the greater the shareholding of the largest shareholder, the greater the dividend payout.
Although existing research focuses on the relationship between ownership structure and dividend payouts, the crucial issue is the presence of shareholders during general meetings and their participation in voting, esp. on dividend payment. The large shareholders have a small role to play if they are not present and not voting at general meetings.
Shareholders are providers of capital but they also have a set of rights (these rights vary from country to country and from corporation to corporation). The main right of any shareholder is the voting right at the shareholders' meetings (Lazarides, 2020). The most important shareholders' meeting is the Annual General Meeting. An annual general meeting (AGM) is a yearly gathering of a company's interested shareholders. At an AGM, the directors of a company present an annual report containing information for shareholders about the company's performance and strategy. Shareholders vote on current issues, such as approval of audited financial statements, financial result distribution (dividend paying, earning retaining, covering the losses), and granting discharge to the members of the board of directors and supervisory board. Voting is the main avenue to aggregate shareholder preferences. In the US average discretionary participation rate in the US corporate voting is at 73% and there is considerable variation across proposals (Cvijanovic et al., 2020).
Minority shareholders, due to little power, lack of coordination, interest misalignment, and high cost of monitoring, have few options as far as their presence at AGMs is concerned (Hirschman, 1970). According to the Hirschman EVL model (Exit, Voice, Loyalty), the first option is the exit -when shareholders disagree or they believe that their interest is not served with the current strategy made by the dominant stakeholders and the only viable choice is to exit (Lazarides, 2020). This means that such investors sell their shares and do not take part in AGMs. In the second option -voice -minority shareholders are present at an AGM to show their concerns (Achilles heels -Lazarides, 2020). Loyalty (to the dominant shareholders) -as the third option -is the most common one. As Aguilera and Jackson (2003) argue the continental Eurosystem system is characterized by a commitment by large shareholders to their investment and minority shareholders to the leadership and competence of major shareholders. Minority shareholders choose to be present to show their support to the majority shareholders (useful idiots -Lazarides, 2020) or to be absent at the AGM which is a less costly option for showing support (free riders -Lazarides, 2020). Summing up, at an AGM there are large shareholders present, but also minority shareholders showing support to the majority shareholders or wanting to show their concerns (Barros et al., 2021;Brav et al., 2020).
On the basis of the literature review, we formulate several research hypotheses: H1. Voting results on profit distribution reach a lower consensus (than on financial statement approval).
The justification for hypothesis 1 is that dividend decisions are the main issue of conflict between shareholders (both majority and minority and among large shareholders - Gugler, Yurtoglu, 2003).
H2. The higher presence of large shareholders the lower dividend payout.
The justification for hypothesis 2 is the fact that the shareholders, especially minority shareholders prefer dividends and they are more prone to show their support for such decisions (Gugler, Yurtoglu, 2003). Their support increases with the increase in the dividend payment.
Additionally, we formulate the hypothesis including the size of the company into the research: H3. In small companies, there is a higher shareholders' presence when compared to big companies.
H4. In small companies, there is a higher voting consensus than in big companies (in all points of the AGM agenda -financial results distribution, financial statement approval).
The justification for hypotheses 3 and 4 is the fact that in small companies there are fewer shareholders present at AGMs and thus it is easier to reconcile their interests and reach a consensus.

sample and data source
We included the 20 biggest companies and 20 smallest companies from the Warsaw Stock Exchange (WSE) in our sample. The big companies come from the WIG20 index (to that index belong the biggest WSE companies) and the small companies from the sWIG80 (to that index belong the smallest WSE companies). Since 1994 (the start of the WSE) some companies entered and left the indexes (WIG20 and sWIG80) we took the set of companies that constitute those indexes at the date of 20 th June, 2022. In the further part of the paper, we will call them big companies (those coming from WIG20) and small companies (those coming from sWIG80). Elżbieta Bukalska,Myroslava Zabediuk,Nataliia Mostovenko 24 The data on voting results and shareholders' presence refer to the Annual General Meetings that were held within the 2010-2022 period (13 years). We were able to make the sample consisting of 413 firm-year observations. Data on AGM presence and voting results were hand-collected from current reports. There are two kinds of current reports: one referring to voting results on all points of the AGM agenda and the second one referring to the largest shareholders present at the AGM. Therefore, we had to collect almost 830 reports to collect the data on the AGMs for our research. To find these current reports, we had to browse almost 10,000 current reports. Additionally, we collected financial data on the sample companies. The source of the financial data was the Notoria Serwis database.

variables
We employed several variables describing: a) the shareholders' presence at the AGM, b) voting results at the Annual General Meeting, and c) variables describing dividend payout. Additionally, we included several financial ratios as control variables. The set of variables is presented in Table 1. As for dividend payout we included in our research several variables to provide a reliability (robustness) check.
We define large shareholders as those with more than 5% of votes during the AGM. This is in line with La Porta, Lopez-de Silanes and Shleifer (1999) and Barroso Casado, Burkert, Dávila and Oyon (2016) and De Cesari (2012). The level of 5% is also in line with Polish legal rules. These rules assume that every shareholder whose stake in a company exceeds 5% of all votes is obliged to report to the Polish Financial Supervision Authority and is the subject of specific disclosure obligations.  (2019) Source: authors' own elaboration.

methods and models
To verify hypothesis H1 we compared the voting result on dividend decisions and financial statement approval. We compared the voting results separately within each subsample (big and small companies). To compare the variables within the subsamples of big and small companies we used a statistical test for dependent samples.
To verify hypothesis H2, we calculated the relation between dividend payout and large shareholders' presence at the AGM. To show the relationship between variables, we used a correlation coefficient analysis. Additionally, to provide the reliability (robustness) of our findings, we applied the regression analysis to find the impact of large shareholders on dividend decisions. The regression model is the following:   presence at the AGM between big and small companies are statistically significant. Although we found statistically significant differences between subsamples as for shareholders' presence our findings contradict hypothesis 3 (assuming that in small companies, there is a higher shareholders presence when compared to big companies). Our finding indicates that there is still a big number of shareholders with their votes who refrain from attending the AGM both in big and small companies.

findings and discussion
As for the 1 st largest shareholders present at the AGM, we find that in big companies they hold a bigger part of the votes during the AGM -more than 61%, while in small companies more than 52%. These differences are statistically significant. While the presence of large shareholders (holding more than 5% of votes present at the AGM) is higher in small companies -they hold 92% of the votes present at the AGM, in big companies it is only 80%. These differences are statistically significant. Additionally, there are larger shareholders in small companies (median is 3.5) than in big companies (3.0). These differences are statistically significant. This means that in small companies the ownership is more dispersed. In small companies, a lower presence ratio with a lower ratio of 1 st shareholders and a higher presence ratio and a number of large shareholders show higher diversity among large shareholders. This means that minority shareholders in small companies choose more often the exit option (they do not appear at annual general meetings).
As for voting results, there is a higher consensus as for financial statement acceptance in small companies than in big companies. Higher voting consensus on financial statement acceptance in small companies is connected with lower shareholders' presence ratio and higher large shareholders' presence ratio. This shows cooperation among shareholders as for financial statement acceptance. However, in small companies, there is a lower consensus with a higher standard deviation as for financial results distribution than in big companies. These differences are statistically significant. This finding partially confirms hypothesis 4 (assuming that in small companies, there is a higher voting consensus than in big companies in all points of the AGM agenda -financial results distribution and financial statement approval). We were able to confirm hypothesis 4 only in the part relating to voting results on financial statement acceptance.
A lower consensus with higher standard deviation as for financial results distribution in small companies shows that dividend decisions cause conflict among shareholders, especially in small companies.
We found that the dividend per share ratio is higher for big companies, while the dividend to total assets ratio is higher for small companies. These differences are statistically significant. But we did not find statistically significant differences between big and small companies as for dividend-to-profit ratio and dividend-to-equity ratio. Our findings contradict those of DeAngelo, DeAngelo and Stultz (2006) who found that small and young companies having many investment projects will prefer to use free funds to finance them, since borrowing external capital will be more expensive and difficult for them to access. Big and mature companies do not face such a problem, but at the same time opportunities for such companies to expand their operations become less and less, and agency problems become more pronounced.
As a result, big and mature companies will make dividend payments in the first place (life-cycle theory of dividend). However, a similar level of dividend-to-profit ratio might be explained by mimicking behavior. Younger, smaller and harder-to-value companies are more likely to mimic their larger, older peers as for dividend decisions are concerned (Lee, 2020).
The value of financial ratios (total assets, leverage, profitability, and cash holdings) significantly differs between big and small companies. As for total assets the big companies have 100 times bigger total assets than small companies. Small companies have a significantly lower leverage ratio, higher profitability, and higher cash holdings.
We noticed that the voting results between financial statement acceptance and financial results distribution and between big and small companies differ. We decided to dig deeper into this issue. To verify hypothesis H1 (assuming that voting results on profit distribution reach a lower consensus than on financial statement approval) we applied a statistical test for dependent samples -as we compared the voting results within each subsample. Due to the fact that the data of our sample are not normally distributed, we applied a non-parametric Wilcoxon test. Table 3 presents the results of the Wilcoxon test. Comparing the financial results distribution acceptance (dividend decisions) ratio with financial statement acceptance we can see that this difference is statically significant in each subsample. However, in the big companies subsample, voting results on Financial Results Distribution Acceptance Ratio (dividend decision) are higher than on Financial Statement Acceptance, while in the small companies subsample there is quite the opposite situation. These findings partially confirm hypothesis H1 (assuming that voting results on profit distribution reach a lower consensus than on financial statement approval). Our assumptions turned out to be confirmed but only in a subsample of small companies.
The voting results show again that dividend decisions cause conflict among shareholders, especially in small companies. But conflict among shareholders on dividend decisions is well recognized in existing research. However, our findings show that in big companies financial results distribution (dividend decisions) voting reaches a higher consensus than financial statement acceptance. This finding raises questions about a new area of conflict among shareholders. It seems that shareholders cannot agree on financial statement acceptance while they agree on financial results distribution (dividend decisions). Thus, we applied a more thorough analysis of voting results distribution on the financial statement acceptance ratio and financial results distribution acceptance (dividend decisions) ratio. Table 4 presents the results of this analysis. We identified three types of situations: a) the set of companies in which voting results on Financial Statement Acceptance and Financial Results Distribution are the same, b) the set of companies with higher approval for financial statements than for financial results distribution, and c) the set of companies with lower approval for financial statement than for financial results distribution.
We found that in 17% of cases of big companies and 75% of cases of small companies, the voting results on financial statement acceptance are the same as on financial results distribution. This means that shareholders vote in the same way (for, against or abstaining) on financial statement acceptance and financial results distribution. However, in the subsample of big companies, the voting results are lower as there are some shareholders voting against or abstaining from voting (the average is lower than 100%). This means that the attitude of shareholders in small companies is more consistent and more supportive.
We found that in almost 20% of cases of big companies and almost 23% of cases of small companies, the voting results were higher in favour of financial statement acceptance. This means that dividend decision is the area of conflict among the shareholders but only in 20% of companies. This type of conflict among shareholders is already recognized and well-researched. And this conflict refers to both big and small companies but to a higher extent is present in small companies. Additionally, in the big companies subsample, we found 27 cases (out of 42) in which big shareholders (and one 1 st biggest shareholder) were involved in voting against or abstaining from voting on financial result distribution. While, in the small companies subsample, we found 31 cases (out of 45) in which big shareholders (and six of the 1 st biggest shareholders) were involved in voting against or abstaining from voting on financial result distribution. The participation of large (and 1 st largest) shareholders in voting against or abstaining from voting on financial results distribution is a clear sign of conflict among shareholders but also with managers.
We were able to identify that there are more than 63% of cases among big companies and only 2% of cases among small companies in which some shareholders vote against or abstain from voting on financial statement acceptance and at the same time they are in favour of financial results distribution decisions. This is quite surprising as the financial statements of all listed companies are of high quality as they are audited by the external auditor, and verified by the supervisory board, and publicly published. But, by not accepting financial statements, shareholders show their dissatisfaction. But their dissatisfaction disappears when they vote for the dividend payout. Within this set of companies, we were able to find 65 big companies (out of 136) and 4 small companies (out of 4) that some shareholders vote against or abstain from voting on financial statement acceptance and fully support (100% of votes for financial results distribution) the decisions on financial result distribution. Among shareholders that vote in this way, there were only minority shareholders (holding less than 5% of votes at the AGM). These results support the findings of Gugler and Yurtoglu (2003) who found that minority shareholders prefer dividends and they are more prone to show their support for such decisions. But our findings are consistent with those of Gugler and Yurtoglu (2003) only among big companies. In this way, we were able to identify a new type of conflict among shareholders that is not fully recognized and not researched. This is in line with the Hirschman EVL model and the second option -V -in which minority shareholders are present at the AGM to show their disappointment (on the financial statement) without hurting their own interest (dividend payment).
Additionally, for the shareholders' presence and voting results, we present correlation coefficients. Due to the fact that our data are not normally distributed, we implemented the Spearman correlation analysis. Table 5 presents the correlation coefficient for shareholders' presence and voting results. Our results show that shareholders' presence ratio is positively related to the size of the company (the bigger company the more votes and shareholders are present at the AGM). The Size of the company is also positively related to the presence and votes of the 1st biggest shareholder.
However, the more votes of the 1 st largest shareholders the lower profitability (which indirectly indicate the wealth expropriation by large shareholders). Thus, the more profitable the company is the higher the financial statement acceptance ratio is.
However, the bigger company the lower the voting results are both on financial statement acceptance and financial results distribution. This might indicate that the bigger company the more shareholders are present but some are present to show their concern.
The higher the ratio of large shareholders' presence is then the higher the voting results on accepting the financial statement and financial results distribution is. This means that large shareholders cooperate in voting and they achieve a higher consensus in voting.
The next step of our analysis is the correlation analysis between dividend payout and shareholders' presence. Due to the fact that our data are not normally distributed, we implemented the Spearman correlation analysis. Table 6 presents the correlation coefficients for dividend payout and shareholders' presence.  The higher shareholders' presence at the AGM the higher the dividend payout. However, with the higher 1 st largest shareholder and the large shareholder presence ratio the dividend payout is lower (a negative sign of a relation).
For a more thorough analysis of the relation between dividend payout and shareholders' presence, we applied the OLS regression analysis. Table 7 presents the regression analysis results.
In line with our expectations and assumption of hypothesis 2, we found that the higher presence of large shareholders the lower the dividend payout. Additionally, we found that the higher shareholders' presence ratio (including minority shareholders) the higher the dividend payout.
These findings (correlation analysis and regression analysis) support hypothesis 2 (assuming that the higher presence of large shareholders the lower the dividend payout). This means that the higher the large shareholders' presence (and the lower the minority shareholders' presence) the lower the dividend payout. This finding is in line with the findings of Gugler and Yurtoglu (2003), Faccio, Lang, and Young (2001), Mancinelli and Ozkan (2006), Khan (2006), and Renneboog and Trojanowski (2007) and Sáez and Gutiérrez (2015). However, this finding is consistent with the idea that controlling shareholders reduce dividends in order to extract private benefits in other forms. The majority shareholders act in their own best interest and refrain from adopting a minority-friendly dividend policy.
Due to the fact that most companies have several large shareholders present at the AGM, our findings prove that they cooperate. While the role of the 1 st largest shareholder is not significant, probably due to the dispersed shareholdings and their relatively low level of stake in the company.
Additionally, we found that profitability has a positive impact on dividend payout. But the Size is negatively related to dividend payment (the lower the dividend the bigger the company).
Cash holdings are negatively related to dividend decisions (the bigger the dividend the lower the cash holdings). This might mean that dividend is used as a tool for diminishing available cash to prevent overinvestment and reduce agency problems between owners and managers. Moreover, we find a negative relation between leverage and dividend (the higher dividend the lower the leverage). This might mean that dividends and debt are alternative tools for mitigating the conflict between owners and managers. Easterbrook (1984) adds that regular dividend payments may force management to raise external capital for new projects, inflicting market discipline on the firm.

Conclusions
The goal of our paper was to identify the shareholders' presence at an Annual General Meeting, the voting results, and dividend payout decisions. The analysis included big and small companies listed on the Warsaw Stock Exchange.
We found that in big companies there are more votes present at the AGM than in small companies. But there are fewer large shareholders' votes present in big companies than in small companies.
We also found that the higher presence of large shareholders the lower the dividend payout. This finding is in line with the findings of Gugler and Yurtoglu (2003), Faccio, Lang, and Young (2001), Mancinelli and Ozkan (2006), Khan (2006), and Renneboog and Trojanowski (2007) and Sáez and Gutiérrez (2015). However, this finding is consistent with the idea that controlling shareholders reduce dividends in order to extract private benefits in other forms. Majority shareholders act only in their own best interest and refrain from adopting a minority-friendly dividend policy.
We found a higher voting consensus on financial statement acceptance in small companies than in big companies. Higher voting consensus on financial statement acceptance in small companies is connected with fewer votes present at the AGM, and higher large shareholders' votes present. Additionally, we found that in small companies there is a higher voting consensus on financial statement acceptance than on financial results distribution (net profit distribution/ losses compensation). This shows that dividend decisions cause conflict among shareholders in small companies. This conflict involves mainly large shareholders.
We found that in big companies there is a lower voting consensus on financial statement acceptance than on financial results distribution (dividend decisions). This shows the new area of conflict among shareholders -voting "against" or "abstaining from voting" on financial statement acceptance and voting "for" financial results distribution (dividend decisions). This conflict involves mainly minority shareholders. We believe that this reflects the second optionvoice -of the EVL model. The minority shareholders are present at the AGM to show their concerns by voting against or abstaining from voting for financial statements. But at the same time, they do not want to act against their own interests and thus they vote "for" financial results distribution (dividend payment).
We also found a positive relation between dividend payout and shareholders presence ratio, but dividend payout is negatively related to large and 1 st largest shareholders being present. This finding is consistent with the idea that controlling shareholders reduces dividends in order to extract private benefits in other forms.
Summing up, at the AGM, there are not only large shareholders present but also minority shareholders. Minority shareholders show their support to the majority shareholders but also show their concerns (Barros et al., 2021;Brav et al., 2020). Additionally, we found a conflict among large shareholders on dividend decisions, especially in small companies. Moreover, we found that minority shareholders that are present at AGMs in big companies show their concern when voting on financial statement acceptance is concerned (but not dividend decisions).
Our research is not free of limitations. We tackled only public companies listed on the WSE. Additionally, we conducted our research in one country -Poland. These limitations give good grounds for the future direction of research. We might expand our research on private companies and other countries. However, it is difficult to find proper data as many companies do not reveal detailed information on voting results during the AGM as the Polish companies did.