VCS uses three different methods to indicate the probable race or ethnic origin of voters in their national voter file. Each method is useful in some circumstances for locating people of common heritage. The three methods are Probable Ethnic Origin Based on Last name; match of 2000 Census Data at the ZIP+4 level; and direct coding of race by election officials in the “voting rights” states. The first two are available nationwide on our voter file, the third only in those areas so designated by the Department of Justice.
Taken together this coding provides an extremely effective tool for identifying such groups of voters.
Probable Ethnic Coding Based on Last Name – The VCS dictionary for coding probable ethnic origin of voters is the best available. It has been developed over the last 30 years through the cooperative efforts of political consultants, sociologists and specialists in ethnic origin. It contains over 288,000 last names that code voters into 35 different ethnic groups. Counts for the most common groups can be found in the National Political DataBase. We know examples that would produce misleading results, but overall it produces very effective statistical selections.
This coding is most often used for locating voters of Hispanic, Asian, Jewish, Irish or Italian origin. Under certain circumstances it can be useful for locating any of the groups below.
ARMENIAN
FRENCH
JEWISH
RUMANIAN
CAMBODIAN
GEORGIAN
KOREAN
RUSSIAN
CHINESE
GERMAN
LAOTIAN
SLAVIC
CZECH
GREEK
LATVIAN
HISPANIC
DUTCH
HUNGARIAN
LITHUANIAN
SWEDISH
EAST INDIAN
IRANIAN
MID-EASTERN
THAI
ESKIMO/INDIAN
IRISH
NORWEGIAN
UKRANIAN
ESTONIAN
ITALIAN
POLISH
VIETNAMESE
FINNISH
JAPANESE
POLYNESIAN
You can run counts or selections based on ethnic coding but the coding for individual voters is never delivered on voter files. Under some circumstances a limited ethnic set is available for delivery on voter files at extra charge.
Census Coding Based on ZIP+4 – VCS has coded almost all individual voters on its nationwide voter file with ten different census counts based on the ZIP+4 of residence of that voter. Four of these variables – Black, White, Asian and Hispanic – are related to race or ethnic origin. This information is contained in the “PL94-171” Census records, given to every household. That means that most voters will be coded in accordance with the percentage of racial breakdown in their block on their side of the street. See complete details.
This coding is very useful for locating black voters in areas that are not covered by the voting rights act. In some cases VCS has used this census coding to determine sets of precincts that cover predominantly black areas and coded them as “jurisdictions” on the voter file. This makes selection and counting much easier. Generally the last name based ethnic coding is more effective at locating voters of Hispanic or Asian origin, but the census data can sometimes be useful in augmenting those selections.
You can run counts or selections based on census coding but if the specific census coding for individual voters is delivered on voter files there is an extra charge.
Direct Coding of Race – The Department of Justice requires registrars in certain states to reflect the race of each individual voter in their voter registration data. This includes North Carolina, Kentucky, Tennessee, South Carolina, Georgia, Florida, Alabama, Louisiana, and the City of Philadelphia. In areas where this data is present you will be able to see the summaries reflected in the counts in the National Political DataBase. In these areas this is definitely the most accurate method for locating black voters.
