(p.207) Appendix C Ancillary Statistical Models
(p.207) Appendix C Ancillary Statistical Models
Table C.5.1 Explaining StateLevel GOP Party Growth in the South, 1962–2008
Model 1 
Model 2 


Political: 

Republican Strength_{t1} 
.7469^{**} (.0336) 
.7628^{**} (.0358) 
SubState Party Competition_{t1} 
.1129^{**} (.0161) 
.0822^{**} (.0182) 
Presidential Vote 
.0002 (.0001) 
— 
Black Electoral Strength 
.1431^{**} (.0374) 
.1468^{**} (.0353) 
% Black 
.2814 (.2069) 
.2174 (.2082) 
Goldwater 1964 
— 
–.0217 (.0145) 
Nixon 1968 
— 
–.0103 (.0105) 
Nixon 1972 
— 
.0054 (.0103) 
Ford 1976 
— 
.0196^{*} (.0087) 
Reagan 1980 
— 
.0060 (.0062) 
Reagan 1984 
— 
— 
Bush 1988 
— 
.0061 (.0065) 
Bush 1992 
— 
.0001 (.0084) 
Dole 1996 
— 
.0185 (.0095) 
Bush 2000 
— 
.0418^{**} (.0130) 
Bush 2004 
— 
.0447^{**} (.0159) 
Bush 2008 
— 
.0370 (.0196) 
Demographic: 

InMigration 
.0055 (.0450) 
–.1180^{*} (.0511) 
Evangelical Protestants 
.4728^{**} (.1383) 
.3565^{**} (.1354) 
Economic: 

Per Capita Income ($1,000) 
.0001 (.0004) 
–.0012 (.0008) 
Agricultural Sector Employment 
–.1906 (.1128) 
–.1595 (.0945) 
R^{2} 
.973 
.977 
N 
253 
253 
Notes: GMMIV Coefficients with fixed effects and robust standard errors in parentheses.
In Model 2, the 1984 presidential election is used as the excluded category.
(*) p 〈 .05 (twotailed test); ^{**} p 〈 .01 (twotailed test)
Table C.6.1 Panel UnitRoot Tests for GOP Registrants, North Carolina
Test 
Test statistics 

Proportion: 

Levin, Lin, and Chu 
–4.870^{***} 
Im, Pesaran, and Shin 
.097 
Logged: 

Levin, Lin, and Chu 
–11.086^{***} 
Im, Pesaran, and Shin 
–14.809^{***} 
Notes: For both tests, a significant test statistic allows for the rejection of the null hypothesis that the series is nonstationary.
^{*} p 〈 .10; ^{**} p 〈 .05; ^{***} p 〈 .01
Table C.6.2 Panel UnitRoot Tests for GOP Registrants, Louisiana
Test 
Test statistics 

Proportion: 

Levin, Lin, and Chu 
22.182 
Im, Pesaran, and Shin 
14.023 
Logged: 

Levin, Lin, and Chu 
–15.451^{***} 
Im, Pesaran, and Shin 
–5.405^{***} 
Notes: For both tests, a significant test statistic allows for the rejection of the null hypothesis that the series is nonstationary.
^{*} p 〈 .10; ^{**} p 〈 .05; ^{***} p 〈 .01
Table C.6.3 TSCS Granger Tests, North Carolina
Black Mobilization Grangercauses GOP Registrants 
F_{1} 

Lags (Years) 

t–1 
3.404^{**} 
t–2 
4.013^{**} 
t–3 
1.503^{**} 
Black Mobilization Grangercauses GOP Registrants 
F_{2} 
Lags (Years) 

t–1 
3.378^{**} 
t–2 
3.905^{**} 
t–3 
1.510^{**} 
Notes: Critical Values: ^{*} p 〈 .05; ^{**} p 〈. 01
GOP Registrants is logged.
A significant F_{1} test statistic indicates the presence of a causal relationship.
A significant F_{2} test statistic indicates the causal process is heterogeneous or does not occur across all counties, while an insignificant test statistic indicates the causal process is homogeneous or uniform across all counties.
Table C.6.4 TSCS Granger Tests, Louisiana
Black Mobilization Grangercauses GOP Registrants 
F_{1} 

Lags (Years) 

t–1 
2.111^{**} 
t–2 
1.254^{*} 
t–3 
.910 
Black Mobilization Grangercauses GOP Registrants 
F_{2} 
Lags (Years) 

t1 
2.109^{**} 
t2 
1.226^{*} 
t3 
.872 
Notes: Critical Values: ^{*} p 〈 .05; ^{**} p 〈 .01
GOP Registrants is logged.
A significant F_{1} test statistic indicates the presence of a causal relationship.
A significant F_{2} test statistic indicates the causal process is heterogeneous or does not occur across all parishes, while an insignificant test statistic indicates the causal process is homogeneous or uniform across all parishes.
Table C.6.5 Explaining CountyLevel GOP Registration in North Carolina, 1966–2008
Model 


Log GOP Registrants_{t1} 
.8613^{**} (.0225) 
Black Electoral Strength 
.2385^{*} (.0962) 
Percent Black 
–.8915^{**} (.1454) 
InMigration 
–.1426^{**} (.0285) 
Evangelical Protestants 
–.0710^{*} (.0477) 
Percent Military 
–.2364 (.2547) 
Percent Urban 
.1212^{**} (.0252) 
Median Household Income ($1,000) 
.0017^{**} (.0006) 
R^{2} 
.93 
N 
4,100 
Notes: GMMIV Coefficients with fixed effects and robust standard errors in parentheses.
Dependent Variable (GOP Registrants) is logged.
(*) p 〈 .05 (twotailed test); ^{**} p 〈 .01 (twotailed test)
Table C.6.6 Explaining ParishLevel GOP Registration in Louisiana, 1966–2008
Model 1 
Model 2 


Log GOP Registrants_{t1} 
.9504^{**} (.0045) 
.9500^{**} (.0047) 
Black Electoral Strength 
.3553^{**} (.1228) 
.3286^{**} (.1267) 
Percent Black 
–1.2867^{**} (.2026) 
–1.2422^{**} (.2079) 
InMigration 
.5662^{**} (.0670) 
.5716^{**} (.0721) 
Evangelical Protestants 
— 
.0951 (.1273) 
Catholic 
–.0790 (.0552) 
— 
Percent Urban 
–.0689^{**} (.0167) 
–.0707^{**} (.0167) 
Median Household Income ($1,000) 
–.0050^{**} (.0008) 
–.0048^{**} (.0008) 
R^{2} 
.99 
.99 
N 
2,624 
2,624 
Notes: OLS GMMIV Coefficients with fixed effects and robust standard errors in parentheses.
Dependent Variable (GOP Registrants) is logged.
(*) p 〈 .05 (twotailed test); ^{**} p 〈 .01 (twotailed test)
Table C.8.1 TSCS UnitRoot Tests for Black Mobilization
Panel A: Deep South 


Test 
Test statistics 
Levin, Lin, and Chu 
–6.413^{***} 
Notes: A significant test statistic allows for the rejection of the null hypothesis that the series is nonstationary.
^{*} p 〈 .10; ^{**} p 〈 .05; ^{***} p 〈 .01
Panel B: Louisiana 


Test 
Test statistics 
Levin, Lin, and Chu 
–18.071^{***} 
Notes: A significant test statistic allows for the rejection of the null hypothesis that the series is nonstationary.
^{*} p 〈 .10; ^{**} p 〈 .05; ^{***} p 〈 .01
Panel C: North Carolina 


Test 
Test statistics 
Levin, Lin, and Chu 
–30.052^{***} 
Notes: A significant test statistic allows for the rejection of the null hypothesis that the series is nonstationary.
^{*} p 〈 .10; ^{**} p 〈 .05; ^{***} p 〈 .01
Table C.8.2 Explaining CountyLevel Black Mobilization in Louisiana and North Carolina, 1966–2008
Louisiana 
North Carolina 


Republican Strength _{t1} 
.0474^{*} (.0209) 
.4118^{**} (.0452) 
Black Mobilization _{(t1)} 
.8448^{**} (.0145) 
.7383^{**} (.0293) 
% Black 
.2059^{**} (.0365) 
.0860^{**} (.0726) 
Jackson 1984 
.0155^{**} (.0033) 
.0913^{**} (.0075) 
Presidential Election Year 
.0051^{*} (.0024) 
.0493^{**} (.0027) 
Obama 2008 
.0213^{**} (.0034) 
.0816^{**} (.0063) 
Constant 
.0550^{**} (.0144) 
.0315 (.0227) 
Estimation Procedure 
FE 
FE 
R^{2} 
.90 
.68 
N 
2,688 
4,200 
Notes: OLS coefficients with fixed effects and robust standard errors in parentheses.
(*) p 〈 .05 (twotailed test); ^{**} p 〈 .01 (twotailed test)