Crunching the numbers: W&J student examines Washington County voter data

“I like to be the first person who votes,” said Jesse Crammer, a Washington & Jefferson College economics major who has delved into an independent study and crunched voter registration numbers for Washington County for the past two decades.
“Since the 1980 presidential election, it’s been a national trend than women have been turning out at greater rates than men,” he said.
That year, voter turnout among women was 0.04 percent higher.
Under the 19th Constitutional Amendment, women were granted the right to vote in each state in 1920, so it took six decades for women, who comprise 51 percent of the population, to overtake men among the ranks of voters. The trend continued through the 2016 presidential election, according to Rutgers.edu, when 63.3 percent of female voters turned out compared with 59.3 percent of males.
Crammer’s independent study was done as an economics project because it was based on statistical models. He compared voter registration data with that of census tracts.
As part of it, Crammer has perused Washington County voter turnout available on a database for the past 20 years, but he ran into a roadblock of sorts when he found “a lot of blanks for gender classification.”
There were 141,828 voters in Washington County as of Feb. 5.
But only about half – 71,030 – are identified on the voter registration rolls as either male or female.
The categories were female, male, blank or “U” for unknown.
With so many unknowns, Crammer said he wasn’t as confident in the demographics of gender as he was about his data in another category.
The W&J student, who was scheduled to graduate yesterday, also examined turnout by party registration and found that in Washington County, Democrats are much more likely to vote in local elections that Republicans. He found members of the Democratic Party have a 13 percent higher turnout than members of the GOP.
Why is this significant?
“Previous election history tells you how likely someone is to vote,” Crammer explained. “Nationally, turnout has been declining for quite some time. It’s interesting to see how it works.”
Crammer also said voters tend to cast ballots more often as they age.
“Every additional year of age increases turnout by 0.3 percent,” he noted. “The older you are, the higher chance is you’re going to vote.”
Perhaps this is because retirees have more time or flexibility during election day to show up at the polls? The data doesn’t explain it, but it might be a matter of common sense.
Crammer also questioned how there could be 150 voters on the rolls listed as having been born Jan. 1, 1800.
It was the practice many years ago to use that date as a default when a voter’s birth date was unknown, and 1800 was chosen rather than 1900 because, at the time, there was a possibility that voters might have been born in 1900.
After combing through the data, Crammer saw a listing for a 109-year-old voter who was born in 1909 and voted in 2018. Could this be the correct date of birth for what’s known as a “supercentenarian,” or was it a data entry error?
According to a 2016 Smithsonian magazine article, there were at that time 72,197 Americans aged 100 or older, according to the federal Centers for Disease Control and Prevention.
Crammer also examined data on Allegheny County voters, with more than 900,000 registered voters, but he was somewhat surprised to find that in Washington County, for its size, “there are a ton of parties.”
Crammer reformatted his statistical analysis to include Democrats, Republicans, Libertarians, Greens and a catchall category of “other.”
Recent lists have included the Halloween Party, Jedi, Rainbow, National Socialist, Communist, Christian, Conservative, Federalist, Whig and Socialist parties, just to name a few.
He said of the major-party members, “Their voting behavior would be different with the massive statewide structures footing the bills for direct mail, robocalls and advertisements.”
“Democrats and Republicans will have more contact with their party in state and mid-level races,” Crammer said.
Crammer said he was “thoroughly satisfied with the elections office” and the amount of cooperation he received in gathering data for his independent study.
Elections Director Melanie Ostrander, for example, let him know when her office, with data from the geographic information system department, compiled an interactive precinct map for the county website.
Crammer said no candidate should ever be dissuaded by demographic information.
“Knock doors, talk to people – for every single door that candidate knocked, that result would swing in favor of the candidate more than historical results,” Crammer said.
“It didn’t mean that candidate would win the precinct, just that the precinct would tend to sway toward that candidate more than it would historically.”
Crammer said he’d like to find employment “modeling data for voters” while perhaps pursuing a master’s degree in statistics.