Gwinnett MARTA voting analysis: 5 takeaways you need to know

The failed Gwinnett County MARTA referendum was driven by many factors, but one of the dominant themes that seems to have determined how residents voted came down to where they live and what new services the proposal offered their area, according to an analysis by The Atlanta Journal-Constitution.

The AJC compared voting precinct results with corresponding U.S. Census tract data and found those in closer proximity to high-capacity transit options proposed by the referendum were more likely to support it. For example, the I-85 corridor and the Norcross area, which was slated to receive a rail station, voted heavily in favor of the proposal, the analysis showed.

“Those people who would be most inclined to make use of extended MARTA voted yes, and those who would probably never use it voted no,” said University of Georgia political scientist Charles Bullock.

Unsurprisingly, he said, “many people were voting their self interest.”

Interviews throughout the weeks leading up to Tuesday's election and its aftermath reveal that many factors played into the referendum's failure, including a longheld distrust in MARTA as an entity and a special election that kept turnout much lower than it would've been during a general election.

>> PHOTOS: Gwinnett votes on the MARTA referendum

>> VIDEO: 5 things to know about MARTA

>> MAP: How did each precinct vote in Gwinnett's MARTA referendum?

The AJC analysis of the data also found age, race and income were factors that seemed to influence voters’ propensity to either support or oppose the measure.

Outliers exist and, as the analysis showed, many of the trends overlap.

Wanda Jackson casts her vote at the Best Friend Park precinct at 6224 Jimmy Carter Blvd in Norcross to vote on Gwinnett's MARTA referendum early Tuesday, March 19, 2019. JOHN SPINK/JSPINK@AJC.COM


icon to expand image


• Commute, shmommute: One of the more interesting data points showed that people with longer commute times didn't seem to the support MARTA proposal.

The AJC’s analysis found that areas with the highest percentage of people spending an hour or more traveling to and from work were more likely to oppose expansion.

That could be explained, though, by the fact that those areas also tended to overlap with those with higher median incomes and those on the far eastern and northern sides of the county.

“They’re not going to take MARTA,” Bullock suggested. “They’re going to be able to pay for the Peach Pass.”

• The income effect: On that note, areas that had median household incomes greater than $70,000 (according to the Census Bureau's 2017 American Community Survey) tended to oppose MARTA expansion into Gwinnett.

Areas like Peachtree Corners, Suwanee, Dacula and Braselton have high median household incomes and most precincts in those areas voted against Tuesday’s referendum.

• Younger = transit hunger: Younger voters were targeted by pro-transit forces in the build-up to Tuesday's election, and analysis shows that was generally justified.

Areas with median ages below 35 tended to vote yes, the AJC found, while older communities more often said no.

“Says to me that what our program did worked,” said Fred Hicks, campaign manager for the New Georgia Project Action Fund’s pro-MARTA efforts in Gwinnett. “We just needed more.”

• Minority supportFurther analysis showed that areas with larger minority populations — which were also targeted by transit advocates — tended to support the referendum as well. Those areas includes Norcross, Lawrenceville and Duluth.

That was not true, thought, in the relatively diverse areas around Grayson.

• What's in it for me? That brings us back to the idea of self-interest voting.

A sign indicating the MARTA and Gwinnett County transit referendum voting calendar is displayed at the Doraville MARTA Transit Station in Doraville.

Credit: Alyssa Pointer,

icon to expand image

Credit: Alyssa Pointer,

Generally speaking, the AJC found that voters in areas where smaller-scale expansions of bus service were among the only new transit options proposed were more likely to vote against the referendum. Those areas include most of the eastern half of the county, Snellville and the aforementioned Grayson area.

Precincts in areas where Gwinnett’s would-be transit plan included more direct benefit to residents tended to vote yes. These areas included along I-85 (where the county’s first MARTA rail station was proposed), and along the Ga. 316 and Peachtree Industrial Boulevard corridors (where a variety of transit options, including bus rapid transit) were proposed.

The biggest outlier was the southern tip of the county near Centerville, where voters supported the referendum. Little was proposed transit-wise for the immediate area.


Nearly 92,000 Gwinnett residents cast ballots in Tuesday’s failed MARTA referendum — less than one-fifth of the county’s total registered voters but a sizable turnout for a one-issue special election.

The referendum was rejected, with the no votes winning by a roughly 8-point margin.

It was Gwinnett’s first referendum on joining MARTA since 1990. Supporters have vowed to hold another vote.