The Unpredictable Predictive Power of Polling

By Tharon Johnson: Guest Op-Ed

Polls are a funny thing. I have been active in politics for most of my life – from the time my mom took me out canvassing for the first time when I was four, I cannot remember a time when campaigns were not a part of my life.  In that time, I have worked on campaigns for state representatives, state senators, mayors, members of Congress, and a president. In just about every one of those campaigns, we have looked to polls for insight on how the race was going and where it should go next.

Polls are valuable tools. They can help prove viability, determine key issues for a constituency, and garner attention from the media. All polls differ in some ways, but the vast majority of them share at least one trait – they poll likely voters. “Likely voters” are those who turn out to vote in most elections. They form the backbone of most electoral victories, from the smallest city council seat all the way up through the White House.

From a financial standpoint, it makes sense – you do not want to waste time polling people who never show up to vote, nor do you want to waste money targeting those people with mailers and ads. 99 times out of 100, focusing on likely voters is a prudent decision, one I have made many times in my career.

That brings us to Stacey Abrams. She is pursuing a different strategy, one that is not about likely, unlikely, or inactive voters, but one that encourages every single man and woman to show up to the ballot box. She is registering voters and bringing new people into the electoral fold, reminding hundreds of thousands of people that their voice is worth hearing. Her campaign is that one in a hundred, if not one in a million.

That kind of novel strategy makes it difficult for pollsters to capture the reality of Georgia’s gubernatorial election. When those pollsters make their calls, they are not reaching the voter that registered last week or the voter that has been unenthusiastic about the candidates for the last three or four election cycles. They are only reaching those that turn out in your usual election.

This is not your usual election, nor is this your usual election year. We have seen upsets across the country – not a single poll had Andrew Gillum winning the nomination for governor of Florida. The best finish any poll had for him was third place. The result? Gillum beat the third-place finisher by over 200,000 votes. We saw a similar result in Massachusetts recently – not a single poll had Ayanna Pressley ousting Mike Capuano in the primary for the 7th congressional district. She won by 18 points.

Abrams is working from similar playbook as Gillum and Pressley, one which motivates as many people as possible to get engaged with her campaign, regardless of their vote history. Democrats, progressives, and forward-thinking moderates are voting in droves this year, even in Georgia. One need not look any further than the record turnout in our Democratic primary to see that. She is giving everyone a reason to vote, both the typical and atypical voters.

Times are changing. Georgia is changing. We have someone who is giving disenfranchised and infrequent voters a reason to do whatever it takes to cast their ballots. Someone who is working hard to overcome our state’s sad history of voter suppression to make her own history. And people are going to overcome voter suppression to make that someone, Stacey Abrams, the next governor of Georgia.


Tharon Johnson is an accomplished public, political, and business strategist with more than a decade of experience creating successful legislative solutions and political campaigns. Johnson has fifteen years experience leading key Presidential, Congressional, state, and local campaigns and government teams to craft innovative approaches to complex public affairs challenges for a diverse array of corporate, nonprofit, and public clients.  He is the Founder & CEO at Paramount Consulting Firm.

Leave a Reply

Please Login to comment
2 Comment threads
2 Thread replies
Most reacted comment
Hottest comment thread
4 Comment authors
CarolineWill DurantBenevolusbethebalance Recent comment authors
newest oldest
Notify of

I asked on another thread the other day what pollsters had done to update their sampling of likely voters. I have to think that there is room for improvement in modeling the likely voters. Besides obviously incorporating 2016 turnout, is there room to use figures on new registrants, or do some weighting for enthusiasm (if that’s not done already), or at some point, including early voting turnout?
Hopefully a true expert can chime in.

Will Durant
Will Durant

Paging Mark Rountree.


The likely voter model is probably the key. If they just polled registered voters it might be more indicative if there are going to be a lot of new voters.


Kemp should resign. He has no business running this election.
Just do it.