The CBS News Battleground Tracker is back: It explains what’s on the minds of voters and provides regular detailed snapshots of the 2022 midterm elections in each state and district throughout the 2022 campaign. A key feature is that our model estimates how many seats each party has. currently in the US House of Representatives.
Where exactly do our Battleground Tracker numbers come from? Here are five things to know.
1. Analyzing races district by district
We take a district-by-district approach to analyze race and measure public opinion, since control of Congress is gained through hundreds of individual elections, not by national popular vote. Indeed, relying too much on national polls can be misleading, because the party with the most votes nationally wins.
The Battleground Tracker estimates individual districts and translates the current support of each major party into the measure that matters: seats in Congress. Our approach also gives you an idea of what voters in different parts of the country think about this year’s candidates and major issues.
2. More than just a poll
While polling voters across the country is an integral part of the Battleground Tracker, it’s much more than a typical poll. The core is a statistical model that is fed by big data. We combine polls, voter bases, U.S. census data, and historic election results to get a clearer picture of what’s happening in each state and district.
Here’s what the data tells us:
We learn which party supports different types of voters from our ongoing polls, which contain much larger samples – in the tens of thousands – than most surveys.
We estimate how many voters like them there are in each state and district, as well as their turnout history, from voter records and census data.
And we know the past election results of every state and district, which helps us anchor our 2022 estimates in recent history.
Our model combines all this data using. A feature of this technique is that we use trends across the country to give our view of a specific neighborhood. For example, if we find Spanish voters in the south who are shifting support, we use the information to estimate more specific districts in which Spanish voters live. The same is true for many other types of voters. The survey lets them tell us what they think, and we map that out with how many of them live in each district.
We collaborate on data collection and modeling with youGovbuilding on our successful as well as the 2020 edition which includes both the and .
3. House landscape informs the model
The landscape of the newly drawn US House districts is challenging for Democrats and giving Republicans a built-in advantage. Democrats often find their voters crammed into fewer districts, both because of their tendency to live in more concentrated areas and because of the gerrymandering that, by definition, packs voters. And despite population growth among racial minorities over the past decade, many states have not added minority districts. If the national popular vote were 50-50, Republicans were expected to win the majority of the seats.
Based on the 2020 presidential election, President Biden’s performance was enough to gain a majority of the new districts across the country. However, there are enough of those districts close enough that a small shift to Republicans would flip many seats. And even if the Democrats saw the voters shift slightly toward them, they simply have fewer pick-up options this year than the Republicans.
Moreover, we have seen Republican turnout advantage – all of which contribute to a favorable environment for GOP pickups.many Democratic members of Congress and indications of a
4. Snapshots, not predictions
We tell you where the races are today and explain why and what could change. Unlike forecasting, we estimate each candidate’s current support and include all the data we’ve collected to date. There’s nothing here to account for forward-looking uncertainty – we’re not modeling changes in economic conditions or future campaign dynamics, for example. We’re fully anticipating movement before the first vote is cast, so we’ll be updating everything regularly over the coming months.
For example, if we estimate a party at 230 seats with a margin of error of ±12 seats, we are confident that their support today translates to between 218 and 242 seats – that’s different from predicting their final score. Likewise, our qualitative assessments of individual races reflect their current status. A race that tends to party these days can be reclassified as a toss-up if it becomes more competitive.
5. Solid track record
While we take a different approach to traditional polls, the Battleground Tracker is based on rigorous methods of political science, survey research, and statistics. In addition, we have a strong track record with comparable models at CBS News in recent years.
performed particularly well, steadily following the Democrats’ improvement in key congressional races and the eventual blue wave in the House. In reality, nailed the end result, when it came to the fact that the historic rise was fueling democratic gains.
were based on a similar model. We estimate that the Democrats had built up a lead towards Election Day, but the Republicans were able to close the gap with a late turnout wave. Every state we classified as leaning or likely Democratic primary went to Mr. Biden, and every state we rated as leaning or likely Republican went to Donald Trump. And of the six states we rated as tosses, Trump won four and Mr. Biden won two.