Page 150 - 臺大管理論叢第32卷第2期

The More, the Merrier? The Bystander Effect on Crowdfunding Platforms

projects to raise funds from the crowd. Given the possible conflicting interpretations, prior
studies regarding the effect of project creators also provide mixed evidence (Bukhari, Us-
man, Usman, and Hussain, 2019; Josefy, Dean, Albert, and Fitza, 2017).
Consistent with prior studies (Chan et al., 2020), we found that the coefficient of

project duration is negative (B = -0.624, p < 0.01). This finding seems to show that
campaigns with longer project duration indicate the project creator’s low self-confidence,
thus influencing investors’ confidence in the focal project. However, the coefficient of

cumulative pledge amount is positive (B = 0.055, p < 0.01). Finally, we find that perceived
project quality is not significantly related to the daily pledge amount (B = 0.281, p > 0.1).

3.4.1 Robustness Test
We also check the robustness of our findings at the project level. Table 4 shows the

results of the robustness tests. We used a logistic regression since our dependent variable,
project success, is a binary variable. Project success is equal to 1 if the project achieves
the targeted funding goal at the end of the funding period, and 0 otherwise.

Model 1 examines the effect of the bystander ratio on project success. Model 1 of
Table 4 shows that the coefficient of bystander ratio is negative (B = -2.773, p < 0.05).
Models 2 and 3 show the moderating effect of project legitimacy and project duration,
respectively (B = -4.119, p > 0.1; B = 1.131, p > 0.1). We do not find the moderating
effects for our project-level analysis. Overall, these results suggest that our project-level

results are partially consistent with our daily-level findings.

3.4.2 Controlling for Endogeneity Concerns

While we treat the bystander ratio variable as an exogenous variable in our main
analysis, the negative relationship between the bystander ratio and the daily pledge
amount may be driven by some omitted variables. For example, perceived project quali-
ty likely motivates investors to become passive (i.e., bystanders) and reduce their pledge
amount concurrently. To address potential endogeneity, we use a two-stage regression

1
model approach. Specifically, in the first stage, we regress the bystander ratio variable on

1 We are thankful to the anonymous reviewer for suggesting this additional analysis.

142