Table of Contents Table of Contents
Previous Page  151 / 274 Next Page
Information
Show Menu
Previous Page 151 / 274 Next Page
Page Background

臺大管理論叢

27

卷第

3

151

As shown in Table 4, Model 3 included interaction among industry age, TMT external

ties, and TMT intra-industry experience addition to test Hypothesis 2, which predicts a

positive moderating effect of TMT external ties on the relationship between intra-industry

experiences and that firm intra-industry performance decreases with industry age. The results

of Model 3 reveal that the coefficient for the three-way interaction is significantly negative

for ties to the CTIA (

γ

= -0.45,

p

< 0.01) and non-significant for ties to other cell-phone

service providers and firms outside the focal industry. In addition, the coefficient for the two-

way interaction between a TMT’s ties to the CTIA and intra-industry experience addition is

significantly positive (

γ

= 4.49,

p

< 0.01), supporting Hypothesis 1 at the emerging stage of

the industry. To check the robustness of the observed effect of industry age in Model 3, we

estimated the specification in Model 2 again by splitting the sample into two groups, one

with an industry age lower than the mean industry age of the full sample and the other with a

higher industry age than that of the full sample. Although not reported here, the results of the

subsample analysis are consistent with the results of Model 3. Thus, Hypothesis 2 is

supported for a TMT’s ties to the CTIA, and Hypothesis 1 is further supported for a TMT’s

ties to CTIA at the emerging stage of the industry of cell-phone services.

In summary, Hypothesis 1 is generally supported but Hypothesis 2 is not for a TMT’s

ties to other cell-phone service providers. Graph 1 shows the marginal benefit of obtaining

one addition tie to the intra-industry competitors. Hypothesis 1 is supported at the emerging

stage of the focal industry for a TMT’s ties to the CTIA. However, this positive moderating

effect of ties to the CTIA gradually decreased with industry age and finally constrained firm

performance, supporting Hypothesis 2. Hypotheses 1 and 2 are not supported for ties to

extra-industry firms. Hence, support for Hypothesis 1 and 2 depends on the type of external

ties that the TMT uses to scan the environment. The results from the fully specified model

are presented in Model 4 in Table 4.