4. Research Results
4.4 Additional analysis
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
58
4.4 Additional analysis
To further support the first hypothesis in this study, I perform additional tests by using different measures of conference calls and interlocked independent directors.
4.4.1 Alternative measures of conference calls- The number of call divided by the maximum number of calls in the industry
The original sample is consisted of 8,669 firm-year observations that come from 15 industries during a 10-year period from 2000 to 2009. I find out the maximum number of conference calls in each industry every year and measure the dependent variable by using number of conference calls divided by the industry maximum in each year. For example, Taiwan Semiconductor Manufacturing Company (TSMC) held 4 conference calls in year 2000 and the maximum number of conference call for 2000-electronic industry group is 4, then the dependent variable measurement for TSMC in 2000 is 1 (4/4). I exclude cases when the industry maximum is zero and firms, which leads to 7,977 firm-year observations. As this alternative measure changes the distribution of conference calls to a continuous distribution between 0 and 1, I use Ordinary Least Squares (OLS) regression to test the relation between the frequency of conference calls and board interlocks. The empirical results of conference calls (PCALL) and whether the focal firm is connected to other call firms (DLINK), conference calls (PCALL) and the number of other call firms connected to other call firms (LINK) and conference call (PCALL) and whether the focal firm is connected to other call firms through independent directors (INDLINK) are presented in table 19-21, respectively. The significantly positive coefficient in table 19-21 illustrate that the relationship between board interlocks and conference calls is not affected by using different measurement of conference calls.
‧
Table 19. An alternative measure of conference call and the focal firm’s link to other call firms calls in the industry. Other variable definitions are in Table 1.
‧
Table 20.An alternative measure of conference call and the number of other call firms linked to the focal firm each year. Other variable definitions are in Table 1.
‧
Table 21. An alternative measure of conference calls and independent directors links PCALL = α + 𝛽1𝐼𝑁𝐷𝐿𝐼𝑁𝐾 + 𝛽2𝑅𝑂𝐴 + 𝛽3𝑆𝐼𝑍𝐸 + 𝛽4𝐿𝐸𝑉 + 𝛽5𝑀𝐵 + 𝛽6𝐷𝑆𝐻𝐴𝑅𝐸 each year. Other variable definitions are in Table 1.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
62
4.4.2 An alternative measurement of interlocked independent director- Percentage of interlocked independent directors of the focal firms
To test Hypothesis 2, I use the dummy variable that equals to 1 if the focal firm is connected to other call firms through independent directors and 0 otherwise. Now I further use the percentage of independent directors among interlocked directors in the focal firm to test Hypothesis 2. This measurement represents a different perspective of interlocking independent directors. The significantly positive coefficient in table 22 and 23 shows that when the focal firms have higher portion of independent directors among interlocked directors, they are more likely to hold conference calls and hold conference calls more frequently. In other word, if firms have more independent directors sit on other boards, they have better chances to perceive the importance of conference calls and would facilitate the decision to hold conference calls in the focal firms.
‧
Table 22. An alternative measure of interlocked independent directors and the decision to hold conference calls focal firms. Other variable definitions are in Table 1.
‧
Table 23. An alternative measure of interlocked independent directors and the frequency of conference call
Log-likelihood -4547.114 -4476.179 -4381.505 -4290.782
Pearson χ2 1245.399 1387.268 1576.616 1758.063
Note:
1. *** indicates significant at 1% level, ** indicates significant at 5% level and * indicates significant at 10% level.
2. PIND is the percentage of independent directors among interlocked directors in focal firms. Other variable definitions are in Table 1.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
65
4.4.3 The content of conference call
I further search the call content of interlocked firms and examine whether these interlocked firms provide similar information. For example, WahLee Industrial Corp.
is interlocked with Wah Hong Industrial Corp and both firms held conference calls in year 2005. According to the content published on Market Observation Post System, both firms mentioned the expected sales and market reaction of their new products in the conference calls. Another example is that YAGEO is interlocked with ProMOS Technologies and both firms held conference calls in year 2006. In their conference calls at the beginning of the year, both firms mentioned the current capacity utilization about their major plants. This analysis implies the shared directors by interlocked firms may further influence the information content disclosed in conference calls.
‧
This study investigates whether firms with board interlock are more likely to hold conference calls. By using the dummy variable of whether the focal firm is connected to other call firms and the number of other call firms connected to the focal firm to measure board interlocks, the empirical evidence supports the argument that firms connected to other call firms through board interlock are more likely to hold conference calls and hold conference calls more frequently. The findings hold after using alternative measures of conference calls. In other words, when the focal firm has directors sit on boards of other firms that hold conference calls, the interlocked directors would have chances to observe the benefits of holding conference calls and to facilitate the decision to hold conference calls in the focal firm.
I also find that firms linked to other call firms through independent directors are more likely hold conference calls and hold conference calls more frequently.
Moreover, the results in additional analysis also indicate that firms have higher percentage of interlocked independent directors are more likely to hold conference calls and hold conference calls more frequently. These results support the hypothesis that interlocked independent directors shared by focal firms and other call firms contribute to a firm’s voluntary disclosure policy. The evidences in this study also support the argument that independent directors play important roles in monitoring firms’ disclosure policies. The above findings further imply that firms connected to other call firms through non-independent directors may not affect the likelihood of holding conference calls.
Overall, my empirical results support that board interlocks, specifically, independent board interlocks, is positively associated with the decision to hold conference calls. In terms of interorganization connection, a firm’s behavior