Data Digest: Stop The Runaway Take Train
It is fashionable to say that inflation is no longer the top issue in the 2022 midterms. I think that take is, at best, premature. Here's why.
Welcome to Codebook's Weekly Data Digest! Data Digest is a great way for premium subscribers to start the week. Today's edition is out from behind the paywall in celebration of the start of football season. Hooray!
Hello, friends!
I would like to take a moment to complain about something I am seeing in The Takes as of late: the view that inflation is just not the main driving issue in the midterms anymore.
This is being driven by economic data showing inflation is not quite as grisly as it was a few months ago, plus a pair of poll findings, both of which are from high-quality public media polls conducted by professionals for whom I have much respect.
Let me be clear up front. The problem is not the polls. The problem is The Takes they have spawned.
In no particular order...
1) The "threats to democracy are the top issue" takes. I wrote about this a few weeks ago when it came out, but I'll revisit it briefly here. This data point comes from NBC's polling, where voters were asked to name their one top issue heading into the election. They were given what I think is a pretty good list of options that encompasses most of the big issues facing voters this November. And NBC's polling team does something that is totally defensible and is something even my own firm does as well, which is code inflation or cost of living as a separate matter from "jobs and the economy" on this question.
This is interesting and useful insofar as we researchers can tease out whether people are anxious about losing their job versus anxious about the cost to buy groceries. But at the end of the day I think "jobs and the economy" and "cost of living" both fall into the Carvilleian "Its The Economy, Stupid" bucket in most voters' minds.
The problem arises when voters are asked their top issue and "threats to democracy" gets 21 percent while "cost of living" gets 16 percent and suddenly this gets pundit-curdled into NO ONE CARES ABOUT INFLATION ANYMORE, SEE. I'm exaggerating a little, but this is only sort-of a straw man. Plus, another 14 percent said jobs and the economy is an issue! Is cost of living not a key part of "the economy" for most people?
2) The "abortion is now the biggest issue" takes. No question abortion has risen in the issue mix in the last few months. (Though I would note that the aforementioned NBC poll from August found only 8 percent of voters said it was their top issue, which is exactly the same as what my own firm found.)
But a new poll from the WSJ polling team (with the tantalizing URL "WSJ_Poll_Redacted_Aug_2022.pdf"...what are you hiding, Jerry Seib?!) has produced a number that is making the rounds as proof that abortion is a bigger driver than inflation.
The WSJ poll asks voters a few different questions to measure how they feel about issues. When voters can name whatever they want and the researchers code this "open-ended" answer on the back end, they can separate out different economic threads. "Economy (general)" is considered a separate item from "Inflation" which is separate from "Jobs" and so on. Again: totally defensible and useful for researchers to have that level of detail. On that question and at that level of granularity, "economy (general)" gets 16 percent while inflation gets 11 percent. Abortion gets 13 percent. Here, too, we have what I'd suggest is an artificial splitting-up of the importance of the economy, but that's not the data point driving The Takes.
Later in the poll, they ask "which of these has made you more likely to vote in the upcoming midterms" and give voters options. Abortion tops this at 34 percent, followed by inflation at 28 percent. This six point gap isn't very large, but that didn't stop it from spawning chatter that, again, inflation is no longer as much of a big deal to voters and the election is shifting to be all about Roe.
Let's pump the brakes for a moment and look at the question. The five choices the WSJ polling team offered respondents were: Inflation and rising prices, the Supreme Court overturning Roe, illegal immigration and border security, gun violence, and the FBI's search of Mar-A-Lago.
Here's my problem with using this question alone to fuel The Takes. Respondents can only pick one on the list. Of the five choices given, two of them are things that are absolutely going to pick off chunks of the Republican electorate: immigration and the FBI search. But for Democrats, there's no "passage of climate legislation" or equivalent to split up the Democratic electorate. (Gun violence is there and arguably something Democrats might be more inclined to choose, but came in relatively low overall.) So Democrats cluster around the Roe item, which we know from other polls has really been galvanizing for them, while Republicans on this question get split up a few ways.
But let's say hypothetically you added a "climate legislation" option to the mix; I doubt it would be in the top three overall, but I bet it'd peel away a few points worth of Democrats, enough to knock the Supreme Court ruling down closer to 28 percent, maybe even putting it just below inflation.
A lot of people like to criticize issue polling by saying, well, this stuff is all so subject to wording, you can make a poll say whatever you want it to say. Here, the WSJ's polling team wrote a perfectly good question! I do not believe they are trying to noodle with things by providing a certain slate of options to engineer an outcome. But this is exactly why you should not let the Take Train blast out of the station without taking a closer look at all the available data and consider how and why a question got the result it did.
One final note: that same WSJ poll does another valuable thing where they let voters rate each issue on its own and express whether that issue made them more or less likely to vote. Setting aside my own allergy toward "more/less likely to vote because of X" poll questions, using them as a comparative measure is interesting. And here, 63% of voters said that inflation and rising prices made them more likely to vote. 56% said SCOTUS overturning Roe made them more likely to vote.
So, again, your mileage may vary, but considering the totality of the data, I am not yet persuaded that inflation is becoming some second-tier issue in the midterms. (Later in this newsletter, we'll look at the WSJ poll's data on voter views on the economy. TLDR: voters think it bad and getting worse, not better!)
On to the rest of the data...
Presidential Approval
The current Real Clear Politics averages for the President's job approval are:
Overall job approval: 42.1% approve (up 0.1 points from last week, up 2.5 points since August 6)
Economic job approval: 38.3% approve (up 1.7 points from last week)
Apropos of nothing: I'm deeply amused that the poll with the highest job approval for President Biden this week is Rasmussen.
The aforementioned Wall Street Journal poll, conducted by the Impact Research/Fabrizio Lee polling duo, is one of my favorites. They ask questions that a real campaign might ask, they release complete toplines and full question wording and put past trendlines right there for analysts like me to pick at. It is glorious!
So a few things stick out in this poll pertaining specifically to President Biden and his job approval:
Biden's improving job approval numbers come despite the fact that more voters today say things are "on the wrong track" than they did in March.
Biden's improving job approval numbers come despite the fact that more voters today say they have a negative view of the economy than they did in March.
This is just more fodder for the notion that Biden's improving numbers have less to do with swing voters going, "gosh, the President is doing a much better job" and more about Democratic voters who previously said things were going OK now feeling unleashed to say things are a hot disaster mess, while also giving higher marks to Dark Brandon for his handling of things.
2022
Because Real Clear Politics and FiveThirtyEight calculate their "generic ballot" in different ways, I'll include each here each week for context:
RCP: Democrats +0.1 (GOP advantage down 0.4 from one week ago, down 0.2 from August 6)
538: Democrats +1.0 (Democratic advantage up .5 from one week ago, up .6 from August 6)
The trendline that ought to most concern Republicans regarding the generic ballot? The WSJ poll takes Republicans from +5 on the generic ballot in March to -3 on the generic ballot today. And they find this even though their poll seems to hold its partisan makeup stable, meaning the poll's weighted findings contain the same proportion of Republicans, Democrats, and Independents today as it did back in March. That shuts down some of the theory that the Democratic poll bump is coming from Democrats being newly eager to take polls; this WSJ finding does suggest some voters are changing their minds.
2024
The RCP roundup of polling on a 2020 rematch shows Trump +0.8 in a hypothetical race versus Biden if it were held today. The range of results they include is pretty wide, from Rasmussen's Trump +6 to the WSJ poll at Biden +6.
But for me, the trend within polls is useful. That WSJ poll had the race even-steven back in March. From even to Biden +6 isn't nothing. And, again, they've got party ID held stable.
COVID-19
Here's how this week's COVID-19 picture looks, via The New York Times.
COVID-19 7-day new case average: 80,230 (14 day change of -13%).
COVID-19 7-day hospitalizations average: 35,769 (14 day change of -11%).
The data about COVID that is driving news this week is less about disease and more about learning loss, where our national education data now confirms two decades worth of improvement in student learning was destroyed in two years by COVID closures. My friend Mary Katharine Ham is writing at The Daily Beast about how anyone openly worried about this outcome was shamed as a COVID denier during the 2020-2021 and 2021-2022 school years, and it is time to take a cold hard look at who did this to America's kids.
The Economy
Here's how this week's economic picture looks:
AAA's National Average Price of a Gallon of Gas: $3.78 (down seven cents from last week)
Unemployment (August 2022): 3.7%
Inflation (July 2022): 8.5% year over year (next release: September 13)
The uptick in unemployment did not seem to freak out the economists I follow on social media; to the contrary, some seem to think it is a good sign and reflective of more people trying to get into the labor force. Slightly higher unemployment will mean employers are under slightly less pressure to raise wages in a tight labor market, meaning slightly less inflation, or so the explanation goes.
But it is worth revisiting that WSJ poll that I keep bringing up: March was when gas prices were higher than they are now. Voters in that WSJ poll report experiencing the recent drop in gas prices, and there's a very slight improvement in the percentage of voters saying the economy is headed in the right direction. However. Half of the respondents say we're in a recession and another one-in-six say we're not now but will be within the next year. Only 37% say cost of living isn't really a problem for them right now (it was 41% in March). Only 30% say the economy is headed in the right direction.
These are not the numbers you'd want to see as an incumbent party heading into a midterm year.
Odds and Ends
Is IndyCar wunderkind Colton Herta actually going to go to Formula 1? Am I going to have to buy a truckload of AlphaTauri merch if he does?
THE GATORS ARE RANKED 12TH IN THE MOST IMPORTANT POLL. (The AP college football rankings poll.)
Liz Truss is the new Tory leader and therefore prime minister in the United Kingdom, and Axios notes that her cabinet has no white men in any of the top jobs.
Chris Pine is the best Chris, besides my actual husband whose name is also Chris. (I say this with affection for Chrises Evans, Pratt and Hemsworth.)
(Cover photo: Catherine McQueen/Getty)
Thanks again for reading Codebook! If you have any feedback, leave it here in the comments - I read them all and value your opinions! - Kristen