*4.1. Descriptive Statistics*

Tables 3 and 4 provide descriptive statistics and mean comparison tests of the variables used in the study, respectively. As is observed in Table 3, family firms are found to perform better than the non-family firms in terms of Tobin's Q and ROA. The mean values of family firms' ROA (net income/total assets\*100) and Tobin's Q (market capitalization/total assets\*100) are 5.092 and 0.774, respectively, as compared to 5.045 and 0.646 mean values of non-family firms, as shown in Table 3. Similarly, the median value of ROA (5.100) and Tobin's Q (0.503) for family firms is higher than that of non-family firms, as shown in Table 4. However, the mean and median comparison tests (*t*-test and *z*-test) yield a significant difference between family and non-family firms in terms of Tobin's Q, as shown in Table 4. The above results are consistent with previous literature, which points out that family firms tend to perform better than non-family firms (Chen et al. 2005; Saito 2008; Morikawa 2013; Dazai et al. 2016; Chen and Yu 2017).

Regarding the test variables, family firms have higher family ownership concentration than non-family firms because they are owned by founders or controlled by founding family members. On the other hand, family firms have a lower level of institution, government, and foreign ownership than non-family firms. The presence of institutional investors in family firms is around 15%, while it is about 17% in non-family firms. The government owns approximately a 1% share in family firms as opposed to nearly a 0.8% share in non-family firms. Besides, foreign owners tend to invest less in family firms with an average of a 2% equity stake compared to non-family firms with an average of a 3% share, as shown in Table 3.



**Table 4.** Mean and median comparison between family and non-family firms.


Note: \*\* meaning *p*-value is less than 0.01; *t*-value is the result from t-student test comparing the mean of two groups with unequal variances at confidence level of 95%; *z*-value is the result from two-samples Wilcoxon rank-sum (or Mann–Whitney U) test comparing the median of two groups. The null hypothesis is the two groups are equal versus the alternative hypothesis that the two groups are not equal.

As for board structure, family firms are found to have smaller board size, fewer board meetings, and fewer independent directors on the board than those of non-family firms. On average, board members in family firms consist of five persons, as compared to seven persons in non-family firms. For board meetings, family firms conduct about 7 sessions in a year, while it is 10 for their counterparts. For board independence, independent directors are found to be fewer in family firms with an average of 10 people against 15 people in non-family firms.

Concerning firm characteristics (control variables), non-family firms show higher market capitalization, a higher longevity level, better cash flow over operating revenue, and higher debt to equity ratio than family firms. The lower leverage ratio for family firms indicates fewer financial risks for them as compared to non-family firms. However, lower cash flow over operating revenue ratio for family firms suggests that they may encounter financial difficulties in expanding businesses. Overall, the univariate analysis presented in Tables 3 and 4 indicates that there is a significant performance difference between Japanese family and non-family firms in terms of firm-specific characteristics, ownership structure, and board composition.
