Table 1 Likelihood-ratio tests of link distributions. The null hypothesis is that, beyond some threshold, the upper tail of a link distribution is characterized by a power law distribution. The alternative hypothesis is that the upper tail follows a lognormal beyond the same threshold. The significance level is set at 5 percent. The power law exponent associated with each link distribution is estimated using observations belonging to the corresponding power law range
From: Structure of global buyer-supplier networks and its implications for conflict minerals regulations
Power law exponent | Standard error | Power law range | |
|---|---|---|---|
Recent period (2013, 2014) | |||
Supplier linkage | −1.37 | 0.018 | \(2 \le N_{s} \le89\) |
Customer linkage | −1.20 | 0.028 | \(2 \le N_{c} \le79\) |
Licensee linkage | −1.58 | 0.027 | \(2 \le N_{\mathit{le}} \le30\) |
Licensor linkage | −1.60 | 0.030 | \(2 \le N_{\mathit{lo}} \le24\) |
Strategic alliance linkage | −1.29 | 0.032 | \(2 \le N_{\mathit{sa}} \le23\) |
Entire sample period | |||
Supplier linkage | −1.43 | 0.017 | \(2 \le N_{s} \le420\) |
Customer linkage | −1.52 | 0.042 | \(2 \le N_{c} \le516\) |
Licensee linkage | −1.31 | 0.021 | \(2 \le N_{\mathit{le}} \le30\) |
Licensor linkage | −1.55 | 0.032 | \(2 \le N_{\mathit{lo}} \le40\) |
Strategic alliance linkage | −1.30 | 0.030 | \(2 \le N_{\mathit{sa}} \le47\) |