Correspondence from Engstrom to Hodgkiss Re: Gantt Analysis

    University of 
Department of Political Science 3 New Orleans 

New Orleans, LA 70148 

  

  

February 7, 1957 

Ms. Anita S. Hcdgkiss 
Ferguson, Stein, Wallace, Adkins, Gresham, and Sumter 
Suite 300 
741 Renilworth Ave. 
Charlotta, NC 28204 

Dear Ms. Hodgkiss: 

I have perfcrmed, at your recuest, an analysis of the vcte for 
Mr. Harvey B. Gantt in the November 1236 general electior for a 
United States Senate seat in North Carolina. This werk supplements 
the analyses of WNcrth Carolina elections that 1 performed 
previously for the State of Nor=: Carolina in the Shaw v. Hun- 
litigation. The methcdolcgies emploved in the analysis of this 
election, regression ard homcgensous precinct analyses, are 
identical to those employed in my previous reports for the state. 
The voter registraticn data utilized tec analyze this electicn are 
for Ccrtober 11, #1%%96, and nersiore reflec: the registesrad 
electorate at the time of this electi 

The homogeneous precinct analysis concerns the votes cast in 
precincts in which over 90% c¢I the registered voters was African 
American and in which less than 10% was African American. Mr. 
Gantt received $7.9% of the vctes cast in the homogenecus African 
American precincts acycsgs the state, "but only 38.1% in the 
homogencus non-African American precincts. The voter participation 
rate in this eleckicn in tI homegenecus African American 
precincts, expressed as a percent of the registered voters, was 
49.6%, while the participation rate within the non-Alrican American 
precincts was 59.C%. 

The estimated support for Mr. Gantt among the African American 
voters in this eiecticn produced by the regression analysis, which 
is based on the votes cast in al of the precincts in the state, is 
100%. His suppor: among the non-African American voters is 
estimated by regression to have been 35.7%. The ccrrelation 
coefficient for the relationship betwean the racial composition of 
the precincts and the vette for Gantt is a statistically significan 
.777.. The regression estimate cf the participation rate in thls 
election among state’s African American registered voters 1s 45.3%, 
while that for non-Alrizan Americans is 58.9%. 

You alsc requested the results of these analyses fcr the 
northeast regicn of the state, an area that you informed me is 
comprised cf the following 18 ccunties; Beaufort, Bertie, Craven, 
Edgecomb, Gates, Granvil.e, Greene, Halifax, Hertford, Lenoir, 

A Member of the Louisiana State University System Committed to Equal Cpportunity Employment  



Martin, Northadfffon, Pitt, Vance, Warren, @rington, Wayne, and 

Wilson. The vote in this area is more racially divided than in 
state as a whole. Mr. Gantt received 96.6% of the wvctces cas: in 
homcgeneous African American precincts in this region, but only 
29.0% of those cast in the homogeneous non-African American 
precincts. Voter participation in these African American 
precincts, again expressed as a percentage of registered voters, 
was 50.2%, compared toc 61.9% in the non-African American pracinct 

The regression ana’ysis of the votes cast in all of the 
precincts in these 18 counties places Mr. Gantt'’s support among the 
African American voters at 100% and his suppor: amcng the non- 
African American voters at 24.9%. The correlation coefficient fcr 
the relationship between the racial composition of the precincts 
and the vote for Gantt 1s .930, higher than that for the state as 
a whole. This is also a statistically significant correlation. 
The regression estimates of the voter participation rates in the 
northeast region are 47.0% for African Americans and 61.56% for nen- 
African Americans. 

I hope you £ind this information useful. If you require any 
additional analysis, please let me know. 

  

Sinceraly, 

Trl Co 
Rickard L. Engstrom 

ok Protfesscr of 

Politi cal Science