By Adam Pagnucco.
The release of a poll showing Congressman David Trone with a seven point lead over Prince George’s County Executive Angela Alsobrooks in the U.S. Senate race has set tongues wagging in the Maryland political world. What should we make of it? Here are a few observations and questions to put this poll in context.
This is an internal poll, not a public poll.
Trone has been a client of Hickman Analytics, the firm which conducted the poll, since at least 2016. Through the third quarter of 2023, Trone’s senate campaign has paid Hickman $654,159, including a $213,000 payment for polling on July 28. That does not make the poll wrong, but let’s be aware that Trone paid for it and we only see the metrics from it that he wants us to see.
Because of the poll’s sample size, the race could be essentially tied.
The poll has a sample size of 1,000 “likely Democratic primary voters” and a margin of error of 3.1 points. Let’s take its results – a 41-34 edge for Trone among the entire sample – at face value for the moment. Because the margin of error is 3 points, the race could be anywhere from 44-31 for Trone to 38-37 for Trone. Since Trone’s margin is 6 points among “those who say they are certain to vote in the primary,” the race could conceivably be close to tied.
Now this raises the whole issue of how the poll defined “likely Democratic primary voters.” The poll’s technical note said the following:
This memo is based on a random sample poll of 1,000 likely Democratic primary voters interviewed by professional interviewers November 27-30, 2023. The sample was selected in a way to ensure that every Democratic voter in Maryland had an equal chance of being included. Respondents were interviewed via land lines and mobile devices. The interviews were adjusted slightly to accurately reflect the known characteristics of the Maryland Democratic primary electorate.
From a statistical perspective, I’m not real happy with the statement that “every Democratic voter in Maryland had an equal chance of being included.” The fact is that not all voters are equally likely to vote – hence the results were “adjusted slightly.” What were those adjustments? There is not enough here to answer that question.
Defining likely voters is a controversial topic in polling and varies a lot among pollsters. Consider Gallup’s complicated model for likely voters in presidential races. I’d rather be up than down in any poll, but seven points is far from the last word on the state of this race. Let’s just say it’s competitive and move on from there.
Trone has massively outspent Alsobrooks but he is not blowing her away in his poll.
Through the third quarter of 2023, Trone’s campaign has spent $9.75 million while Alsobrooks’s campaign has spent $1.15 million. Trone has been all over TV and the mailboxes while Alsobrooks has not. The result so far is a competitive race and perhaps a statistical tie. If I were on Team Alsobrooks, I would be fine with that. After all, what happens when her TV ads go up?
Alsobrooks has a natural advantage in Democratic primary voters. In the 2022 Democratic primary, 126,993 people voted in Prince George’s County while 56,154 people voted in Trone’s Congressional District 6. Trone’s money has to overcome that advantage while introducing him to the rest of the state. That is what’s underway now. Of course, Trone has tens of millions of dollars more headed for his campaign coffers and voters aren’t paying attention quite yet. Let’s see how many mailboxes are still standing come spring.
It’s still early.
There is a loooooong list of key events that have not happened yet but will be meaningful. For example:
Trone has not yet fully exploited his endorsement by the teachers union.
Alsobrooks has not done TV or mail featuring her endorsements by Governor Wes Moore, Senator Chris Van Hollen or other heavyweights. In fact, Alsobrooks has so far engaged in almost no paid comms – yet.
The Washington Post and the Baltimore Sun have not yet endorsed.
Outside groups have not started to play yet. For example, Emily’s List has endorsed Alsobrooks and Trone has a vulnerability on abortion based on his company’s political spending. Will Emily’s List do anything about that?
No one has gone negative – yet. The race is getting a bit chippy. Will either candidate roll the dice and throw a punch?
Trone has a ton of analytical capability. Will he use it?
When I worked for David Blair’s Montgomery County Executive campaign in 2021-2022, one of my favorite tasks was parsing our poll results. We employed the great Fred Yang and when he sent over his polling memos, it was like Christmas morning for me. I was particularly delighted to go through our crosstabs but there was a problem – sample size prevented us from digging too deeply.
That’s apparently not a problem for Trone. His poll has a sample size of 1,000 respondents, a mammoth total. That should enable his campaign to examine minute demographics and geographies, especially if he is frequently using that sample size in multiple polls. They can not only run horse race results for many different groups – they can also message test among them and capitalize on that. Is there a particular message that benefits Trone among seniors? Trone’s campaign will know exactly what it is and they can put it in mail to seniors. Is there a particular message that hurts Alsobrooks in the Baltimore suburbs? Again, they will know exactly what it is and they can put it in targeted mail and digital. Alsobrooks could do some of this too but Trone’s capacity to do it will exceed hers.
We all know that Trone will outspend Alsobrooks, but if he campaigns smart as well as big, he can multiply his advantages through polling and targeting. In fact, he might need to do that to win.