According to MS NOW’s official race ratings, control of the Senate could come down to four states: Alaska, Maine, Michigan and Ohio.
As of Tuesday morning, all four Senate races were rated toss-ups, meaning that either party could plausibly win them.
Democrats hold the edge in Georgia, Minnesota, New Hampshire and North Carolina, while Republicans are favored in Florida, Iowa, Nebraska and Texas.
If those ratings hold, Democrats would need to sweep the four toss-up states to reach a narrow 51-seat majority. Republicans, meanwhile, would need just 50 seats, because Vice President JD Vance would break any tie votes.
Project 47 spoke earlier this week with MS NOW director of elections and analytics Kabir Khanna about the ratings.
Race ratings for U.S. campaigns began in the 1970s and really took off in the 1980s. How have they changed since then?
Race ratings indeed have a rich history in American political analysis, and there are several organizations with an impressive track record. Independent evaluations have found them to be quite accurate, and they are often used as inputs to quantitative election forecasts. The ratings are super helpful for identifying the most competitive races for U.S. Senate, U.S. House and governor. The action is in races classified as toss-ups or merely leaning one party’s way.
Over time, it seems that these ratings have leaned more on quantitative analysis, such as polls and statistical models. That also means that when horse race polling is systematically off, like it was in 2020, race ratings may suffer too. Helping guard against that possibility are the inclusion of political “fundamentals,” such as the president’s approval rating and a state or district’s partisan lean, which are also important ingredients these days.
At the same time, they are perhaps a little less reliant on more subjective factors, like assessments of candidate quality, though those more qualitative assessments still play a role. However, as voting behavior has become increasingly partisan in recent years, candidate-specific things, like experience in office or a scandal, don’t seem to matter as much as they once did. That’s not to say that stuff doesn’t matter at all — see Graham Platner in Maine — but it’s more likely to be decisive in races that are close for other reasons.
Why is MS NOW doing its own race ratings now?
Our mission is to deliver smart, data-driven political analysis, and MS NOW’s race ratings are part of that larger goal. We’re letting you know where key races currently stand, based on our reporting and a variety of data sources, including polling and recent election results. It’s difficult these days to separate real signals from noise. As students of campaigns, polls and analytics, we’re in a good position to parse through everything and focus on the most informative signals.
A related goal is transparency, or showing our work. When we change a race rating, we aim to explain why. Sometimes that exercise of walking through various factors, including some that point in opposite directions, is more important than the final race rating itself! And hopefully you will walk away with an idea of which races are most consequential for the balance of power in D.C., as well as a sense of what’s possible on election night.
And again, these ratings characterize how we view key races today. We don’t have a crystal ball that perfectly predicts the future. Ratings will likely evolve as we approach Election Day, based on shifts in voter sentiment and important campaign developments, among other factors. A race that looks fairly safe today may be recharacterized as more competitive down the stretch, and a race that’s leaning one way today may change to a pure toss-up if it looks like neither side has a real edge any longer. In fact, tune in to MS NOW this evening for some big news on that front …
What is your background? How did you learn how to do this?
My training is in quantitative social science, and I’ve spent the past decade in data journalism and election data science. I earned a Ph.D. in political science at Princeton, where I learned the ins and outs of polling, voter registration data and predictive modeling. I went on to manage the decision desk at CBS News and oversaw the methodology underlying our polls and statistical models. That involved race ratings and eventually calling races on election night, in consultation with a team of analysts I hired, including leading experts from Sabato’s Crystal Ball at UVA and the Cook Political Report.
I joined MS NOW this past spring as director, elections and analytics. I manage our election data operation, including the exit polls, vote results and race projections. And I oversee our new race ratings, which have been a fun collaboration between our elections team and political unit.
When you are looking at a race, what is the first thing you look at?
I typically start with polling data, but certainly don’t stop there. Looking at polls of a race in the aggregate gives you a rough idea of how voters are feeling at the moment. Pollsters often say that polls are a “snapshot,” and I like to add that they’re a blurry snapshot, because they come with uncertainty baked in. They are rough measures of where voters stand today. Voters can obviously change their minds, and aside from that, polls clearly do not have pinpoint accuracy. So horse race polling offers signals about which races are close and which side has an edge, but they are hardly the final word. And importantly, it helps to go beyond the toplines and look at demographic and regional breaks, as well as what issues are motivating voters … [P]olls often tell us what the election is about. In 2024 all the polling finding widespread dissatisfaction with the economy and prices was a real sign of trouble for Democrats — perhaps underappreciated at the time.
What else do you consider?
There are plenty of other data points to consider, including recent electoral results, demographic trends and campaign spending. For example, a sizable blue shift in special elections since the 2024 presidential elections portends a strong Democratic turnout this year. Another interesting trend is relatively strong turnout in Democratic primaries. Back in March, it was quite striking to see Democratic primary votes outpace Republican votes in Texas. We’ve seen a similar pattern across several important states, including North Carolina, Georgia and Michigan. Democratic votes have also been outnumbering Republican ones in places with blanket primaries, such as Alaska and California’s 22nd District, in the Central Valley. All signs point to an energized, angry Democratic base.
Aside from the datasets I just ran through, we’re looking at which candidates are successfully raising money and how resources are being deployed across the electoral battleground. That includes ad buys by campaigns and outside groups. Lopsided spending can make a particularly big difference in downballot races. And our deep reporting on campaign events and issues also influences how we see key races.
At MS NOW, we benefit from a large newsgathering operation that puts reporters on the campaign trail, talking with supporters, voters and even people with no intention of casting a ballot. Crowd sizes and campaign-trail moods are only a small part of our analysis, but they can offer an early signal that something may be shifting in a state — and that it’s time to take another look at the data. Polls aren’t a substitute for these kinds of on-the-ground conversations — we see them as complementary sources of information.
Looking at November, what stands out to you the most among the races you’re tracking?
Big picture, it’s pretty remarkable that Democrats have a real shot at taking back the Senate. The Senate is favorable for Republicans, because small states count the same as larger ones, and the seats up this year (class 2) are in states that lean even more Republican than average. But the deteriorating national environment for Republicans, driven by worsening views of [President Donald] Trump and the GOP, are putting several red seats into play. What really stands out is that Democrats have multiple paths to get to 51 seats. States like Ohio and Iowa, which have trended rightward in the Trump era, look genuinely competitive.
And it’s notable that James Talarico has led Ken Paxton in most public polls of Texas’ Senate race. That doesn’t necessarily mean that Talarico is going to win. Those leads tend to be small, and it’s possible that many Republican leaners who are currently undecided ultimately decide that Paxton is the lesser of two evils. But it does look like this could be a very tight race. Texas hasn’t elected a Democrat to statewide office for over 30 years. The closest they have come in recent history was in 2018, when Beto O’Rourke lost to Sen. Ted Cruz by under 3 points. Talarico is polling better than O’Rourke did against Cruz, and Paxton is viewed more negatively.
What are the big questions you’re looking at over the next nine weeks?
One big question is what most voters will feel the midterm election is about, and while there are some clues, the issue environment could change by Nov. 3. Right now inflation and cost-of-living concerns remain front and center. And Trump’s numbers are underwater on these key issues, along with his tariff policy and the war with Iran. His numbers have really tanked among independents, who are fairly consistently breaking toward Democratic candidates in vote choice.
Democrats want to make the midterms a referendum on Trump’s second term and make sure his sagging numbers drag GOP candidates down with him. Republicans are trying to make the election a choice between themselves and a Democratic Party they portray as extreme and out of touch with ordinary Americans. The issues that are the most salient this fall will influence how successful each side is in making its case.
A related question I have is whether Trump’s 2024 voters turn out in big numbers. Some of his electoral gains two years ago were among lower-turnout groups, like young voters. Republicans are trying to ensure their base shows up, because Democrats are currently enjoying a sizable enthusiasm advantage. Meager Republican turnout could be the difference between the typical midterm losses for the party in power and an electoral tsunami.
This is a preview of MS NOW’s Project 47 Newsletter. As President Trump continues implementing his ambitious agenda, get expert analysis on the administration’s latest actions and how others are pushing back sent straight to your inbox every Tuesday. Sign up now.
The post Inside MS NOW’s new Senate race ratings appeared first on MS NOW.
Author: Source Author
From MS Now.
Leave a Reply