Robustness Check Clause Samples
Robustness Check. Change in Termination Probability The table reports logit regressions of termination on fund manager and fund characteristics. I define termina- tion at a fund level, which is similar to specification in ▇▇▇▇▇▇▇▇▇ and ▇▇▇▇▇▇▇ (1999). This dummy variable equals 1 if a fund manager is no longer a fund manager at the fund in year t+1. The first two specifications follow the specifications of Table 3 and compare the earlier period to the recent period. The next four spec- ifications follow the specifications of Table 8 and compare between junior and seasoned managers in two sub-periods. Standard errors are double clustered at the manager and year levels. Resulting standard errors are reported in brackets. A detailed description of each variable is included in Appendix A.
Robustness Check a skill-biased shock from the skill premium The failure of the identified IST shock to explain the data in its entirety implies that there might be a distinction between a sector-biased technology shock and a skill-biased technology shock, especially in the short term. It leaves open the task to identify a perfectly skill-biased technology shock. In pursuit of this goal, I revisit the growth model in ▇▇▇▇ and Lim (1998), in which they separately identifies efficiency parameters for each input factor as follows
Robustness Check. Deconstructing State Democracy As described in section 4.2.2, the democracy score aggregates several different factors to determine the level of democracy in state election systems. To ensure the validity of the index and to determine whether the variables that comprise it are significant predictors of NPVC membership, I conduct a separate analysis without states’ democracy scores, instead using its component parts as covariates. Table 4.4 shows the results of the second analysis. Variable Estimate HR 95% CI p-value Democratic Governor 1.642 5.164 (0.271, 98.532) 0.275 Population -0.065 0.937 (0.286, 3.074) 0.915 Population Density 1.971 7.1811 (1.719, 29.998) 0.00688∗∗∗ Voted for ▇▇▇▇ (2000) -2.011 0.134 (0.004, 4.641) 0.266 Unemployment Rate 0.312 1.367 (0.590, 3.166) 0.466 Recall Provision 0.243 1.276 (0.134, 12.109) 0.832 Ballot Initiative Provision 0.100 1.107 (0.418, 2.907) 0.845 Voter Turnout 0.101 1.106 (0.572, 2.142) 0.764 Voter ID Law 1.216 3.375 (1.621, 7.024) 0.00115∗∗∗ Voter ID Law (No Photo) -1.070 0.343 (0.121, 0.971) 0.04386∗∗ Observations 118 Number of Events 15 R2 0.080 Wald Test 15.7 (df = 10) LR Test 21.71 (df = 10) Score (Logrank) Test 32.75 (df = 10) Note: ∗p<0.1; ∗∗p<0.05; ∗∗∗p<0.01 Table 4.4: Predictors of NPVC Membership, Survival Model Robustness Check In the second analysis, population density still returns significant with p < 0.01. Notably, some of the component parts of the democracy score also return statistically significant: the adoption of voter ID laws that do and do not require photo verification are also significant predictors of NPVC adoption with p < 0.01 and p < 0.05, respectively.