*3.3. Independent Variables*

Based on the motivation in the literature review section, we use three dimensions, further splitting them into six independent variables portraying corporate governance characteristics of a firm (see Table 1). The independent variables were calculated mostly based on their formulas in previous studies.

For capturing the ownership concentration, variable MAJORITY is used, which indicates in a binary form, whether there is a majority owner (i.e., having more than 50% of the shares) present. According to the Estonian regulation, an owner having more than 50% has the power to decide upon most of the actions in a firm, thus the usage of that threshold is well-motivated with legislation. Another variable for the concentration dimension is managerial ownership. To portray managerial ownership, the variable BOARDOWNER is used, which is a ratio of shares owned by the board members to the total shares. Thus, this variable directly portrays the overlap between the two levels of corporate governance (i.e., owners and board members). It must be emphasized, that the Estonian SMEs are subject to a two-level corporate governance system, in which the board is subordinate to owners directly, while the board members are legally responsible for all firm's activities.

For the demographic diversity dimension, the manager's age is portrayed with MANAGERAGE, which is calculated as the biological age of the oldest board member. Although in previous studies the mean age of board members has been used as well, it does not suit herewith, as we intend to capture the life experience available on the board, not the average experience. Furthermore, as a large proportion of firms have single-person boards, the usage of mean age would not be a suitable option. The context of gender is captured with the presence of a woman on the board (reflected with a binary variable WOMAN obtaining 1 on that occasion and 0 otherwise). In studies focusing on larger firms, a gender proportion has been used, but that option is not suitable in the case of SMEs, of which the overwhelming majority have only one or two individuals on the board.


#### **Table 1.** Variables in the analysis.

Source: own elaboration. Note: for robustness tests, BINARYDELAY is recoded to account for mild and severe violators (see also Sections 3.2 and 3.4).

For the experience dimension, business ties are portrayed with the variable TIES, which reflects the number of board memberships in other firms the board members of the firms under question hold. Thus, this variable reflects the scope of ongoing business experience outside the firm under question. Managerial tenure is captured with the variable TENURE, which reflects the time in years the longest-serving board member has been on their position. TENURE could also be used as a ratio of the time the longest-serving board member has been on their position to the firm's age. Still, such a ratio would easily lead to overestimating firm-specific experience in the case of (very) young firms. Finally, the control variable reflecting board size is captured by BOARDSIZE, which reflects the number of board members in the firm.
