Want to know who's going to win the Tulsa mayoral race? Flip a coin.

The mayor's race in Tulsa is in the closing week. The results are close and a lot can happen in the next seven days. Here's my take on the poll results, which a full topline and crosstab report can be downloaded here:

• The race, while showing Bartlett with a three point lead, is within the margin of error and effectively a dead heat.
• This race is going to come down, essentially, to Democrat turnout. Democrat turnout, particularly in Tulsa, has been falling for some time now.  Obama has not helped and Hillary Clinton is not expected to help any more. One of polling's biggest misses was in 2010 when Tulsa County turnout among Democrats was the fourth lowest in the state.
• Democrats are being asked to vote or pick between two Republicans in this race, a choice that will not drive turnout. Democrats already unhappy with Bartlett were going to vote for Bynum anyway. The question is, can Bynum increase Democrat turnout, something that Democrat candidates in Tulsa themselves have had a tough time doing? Plus, there is not much else on the ballot to drive Democrat turnout either.
• There is for the Republicans, however, and the competitive nature of the congressional race may actually help Bartlett, whose voters are much more align with Bridenstine’s.
• Bartlett, as the incumbent, is weak at just 41% and does not look good, considering that the challenger Bynum is expected to receive a larger percentage of the undecideds, in effect, putting him as the winner of the race on Election Day -- IF everything holds the same and true for another week (which is highly doubtful given the activity that typically occurs in the final week of most campaigns).
• However, this does not completely assure a Bynum win and here is why:
• He is effectively the “Democrat” in this race and not the place any lifelong Republican would want to be when the majority of voters on Election Day in Tulsa are Republican.  His largest support, 46.6% comes from Democrats and he is having to talk Democrats into voting for him, a Republican, over another Republican.
• Bynum is getting 31% of the Republican vote and this could spell success for him, but he’ll have to hold on to it for another week which might be difficult.  Bynum is trying to assemble a coalition of Democrats, moderates and unhappy Republicans to win this election and it might just actually work.
• The problem is, his favorables have fallen since the last poll, showing that Bartlett’s messaging is working. Bartlett is trying to break up that coalition by peeling off the Republicans with a Coburn endorsement.  We see this in the liberal to conservative scale where the more conservative you are, the more you support Bartlett. This also shows that Bartlett is viewed as the most conservative in the race as well.
• Bartlett, however, is not helping himself as much as he should.
• His favorables, while a majority, are “soft” meaning more consider him “somewhat favorable” than “very favorable,” and it’s been that way since we’ve been polling him. Voters we talk to tell us, “he’s okay” and this is probably why his re-elect numbers are so low. What is helping him, however, is how Tulsans feel about the direction of the city, which a plurality say is going in the right direction (opposite of what they might say about the nation).
• Bynum, to win, has to say that Tulsa is not going in the right direction, its Bartlett’s fault, and we need a new mayor, but he hasn’t been able to effectively say it and it be believable to the electorate.
• There is no great issue that is driving this race and this may explain why we’re seeing a tight race as well. This race is being driven by personality or who you like the most, which helps Bynum, but voters still need to be talked into (since most Tulsans are happy) “changing horses in the middle of the race” which is where Bynum’s messaging has not been effective. For Bynum, “like me, I’m a nice guy, I'm Mr. Tulsa,” may not be enough.
• We see this in the favorables. 48% or nearly half of Republicans have a favorable impression of Bynum, while 68% of Republicans have one of Bartlett.
• 32% of Bartlett voters have a favorable impression of Bynum and 28% of Bynum voters for Bartlett. These may not seem high but indicate the messaging is not driving a wedge or difference in the minds of voters.
• Again as to why Bynum is effectively the Democrat in this race, 65% of Democrats have a favorable impression of Bynum. We can also see that Bynum has driven down Bartlett’s favorable among Democrats where 55% view Bartlett as UNFAVORABLE, but this wasn’t going to be hard to do.
• The gender vote.
• Bynum's messaging and advertising is, no doubt, designed and directed toward women voters, and he is winning among women, 40 to 37 percent for Bartlett, but this 3-point differential, which should be greater, is proof the approach is either wrong or ineffective.
• You have to ask why he took this strategy considering he got the endorsement of TWO former female Democratic mayors and used a current female county commissioner in his ads.
• Bartlett, meanwhile, has a 10 lead among men, 45 to 35 percent. In all honesty, as the honorary Democrat in this race, Bynum should have been appealing more to men than women.

• Another element that may place a greater role is that of endorsements. Typically, endorsements play a greater role in races where voters do not know much about the candidates, and in this race, both of these candidates are very well known, 98% for Bartlett and 85% for Bynum, so you’d think endorsements would not.
• But here is why I believe it will play a role: Bynum has been running with endorsements of other councilors and county commissioners, which half of voters do not believe makes a difference and 28% are less likely to support a candidate with them.
• Almost half of voters are LESS likely to vote for a candidates with former Democratic Mayor Kathy Taylor’s endorsement, who has endorsed Bynum, and a plurality are MORE likely to vote for a candidate with former U.S. Senator Coburn’s endorsement, who has endorsed Bartlett. Keep in mind, Coburn has perhaps the highest favorables of any former Oklahoma elected official. These might be the sharpest contrast we have in this race.
• If you’re undecided, this might play a role in your decision making when it otherwise might have not.
• Lastly, let’s talk geography.
• Midtowners favor Bynum 50 to 35% and this hurts Bartlett who shares the same base, but Bartlett is beating Bynum in South Tulsa 46 to 39%. For Bynum to be successful, he needs to get a greater share of this South Tulsa vote, which comprises nearly four in ten votes.
• The question is, did Bynum mortgage his Republican stripes to win over Democrats only to lose among South Tulsa conservatives? It would appear that Bynum spent political capital courting Democrat endorsements to win over Democrat voters that he probably would have won in the first place.
• At this point, the undecideds in South Tulsa are breaking near evenly among Bartlett and Bynum, so the jury is still out on this point.
• Here is something to consider: 50% of South Tulsans are LESS likely to vote for a Taylor-endorsed candidate and a plurality 43% are MORE likely to vote for a Coburn-endorsed candidate. Will this be what we see in the coming week?

The poll was commissioned by the Tulsa World, Fox 23 and KRMG radio. Five hundred, ninety-eight (598) likely primary voting Tulsans participated by phone using an IVR (Interactive Voice Response) data collection method.  The results were weighted by age, party and gender, and stratified to a profile of likely primary voters in the city Tulsa. The margin of error was ±4%