Greenville County 2022 Election Turnout

There are numerous ways to split up the geography of a state or area--for example, by county, city, or congressional district. Precincts are units whose boundaries are determined by elections, and they may be split or change over a given year. Precincts may also be misaligned with other forms of geographical division, like neighborhoods, census tracts, or city limits. The main purpose of a precinct is to provide a polling place for its voters during elections. The statistics that come from precinct-level election data simply represent an administrative boundary rather than a particular group of people or any social geography.

Number of 2022 Election Precincts= 151

Median Number of Registered Voters per Precinct= 2,178

The median (the middle value in a set of numbers) is less affected by outliers and skewed data than the mean (over average). For normally or symmetrically distributed data, the median and the mean are essentially the same. See the graph on the left below. For the middle graph (positive skew), think of a hypothetical case where Bill Gates moves into your neighborhood. If you are looking at the distribution of incomes in your neighborhood, his very high income pulls the distribution to the right. The mean will be pulled in a positive direction. That value won't necessarily be representative of the typical household in the neighborhood. It will be on the high side because Bill's income pulled it that way. The median is more resistant to that positive tug and is, therefore, more representative of the central tendency of the data.

In the visual for registered voters below, the data for the precincts is skewed slightly to the right, so the mean (average) is slightly to the right of the median.

The scatterplot on the right (above) shows a moderately strong relationship between the % of Black residents and voter turnout in a negative direction. In other words, as the % Black population goes up in a precinct, there is a moderately strong tendency that turnout is goes down in that precinct. This does not imply causation--just that the variables tend to trend together in the opposite direction.

Another way to look at it is to break the data into two groups: 1) Precincts with % Black greater than the median (10.7%), and 2) Precincts with % Black below the median. The probability that a person turned out to vote, given that they live in a precinct with a higher percentage of Black residents, is 36.3%. However, the probability that a person turned out to vote, given that they live in a precinct with a lower percentage of Black residents, is 52.2%.

When you look at voter registration rates on the the left (above), the correlation (.17) between % Black population and voter registration is quite weak.

Voter turnout for precincts with a higher percentage of Hispanic / LatinX population shows a similar correlation to those with a higher percentage of Black population. There is a moderately strong correlation in the negative direction.

The relationship with regards to registration rates and race / ethnicity is different for higher percentage Hispanic precincts than higher percentage Black precincts. For the latter, the relationship was quite weak. However, precincts with a higher percentage of Hispanic residents show a moderately strong negative relationship with % registration. This means that as percent Hispanic population goes up, voter registration rates do tend to go down.

Another way to look at it is to break the registration data into two groups: 1) Precincts with % Hispanic greater than the median (6.9%), and 2) Precincts with % Hispanic below the median. The probability that a person is registered to vote is only 78.9% when they live in a precinct with a higher percentage of Hispanic residents, but this jumps to a 90.3% probability when they live in a precinct with a lower percentage of Hispanic residents.

• The White Horse Corridor (WHC) shows much lower voter turnout and voter registration rates when compared with precincts outside the WHC. Maps show a clear pattern.
• When data are segmented into and outside the WHC, the patterns become even more apparent.

• The probability of a person being registered to vote when they live in a precinct inside the WHC is 65.4%, whereas the probability is 87% in precincts outside the corridor.
• The probability of a person turning out to vote when they live in a precinct inside the WHC is 28.2%, whereas the probability is 46.8% in precincts outside the corridor.
• Of the 10 lowest ranking precincts for voting registration and turnout, 8 were within the WHC. The WHC is comprised of 23 out of the 151 voting precincts in Greenville County--15% of Greenville County Precincts.
• For higher Hispanic / LatinX precincts there is a moderately strong correlation with low registration and turnout rates.
• For higher Black percentage precincts there is a moderately strong correlation with low turnout rates--NOT with registration rates. This suggests that turnout initiatives may be more applicable in these precincts. However, each precinct should be evaluated individually.
• Early voting rates did show some moderate spatial patterns with higher early voting rates to the south of downtown ( and north of 1-85) along Augusta St., in areas near Travelers Rest, and in some areas of Mauldin and Simpsonville. We did not find a strong correlation between race or ethnicity and early voting. 

• Analyze the following in relationship to voter turnout, voter registration, and voting method:
  • Income Levels
  • Age Breakdowns
  • Housing Status (renter, owner, etc.)
  • Education
• Acquire voting data at the individual household level (age, election participation history, race / ethnicity). We are currently analyzing relationships at the precinct level. Precincts are our unit of measure, rather than individuals.