Political journalism needs fewer "takes" and more analytical legwork
Why I think what you’re getting with this newsletter is different, and why it matters right now
Today felt like a good moment to give readers a behind-the-scenes update on my strategic thinking for this newsletter. What follows is a short essay on what makes Strength In Numbers different from the rest of the political data journalism landscape, what my goals are for this business in 2026, and some of what I’m changing in the months ahead to accomplish those goals.
There’s something broken in political journalism and political data analysis today. Too many thinkers and strategists start with a conclusion in mind — Democrats need to move right, Democrats need to fight, moderates decide election outcomes, voters are irrational, etc. — and then go hunting for numbers to back it up. Often, the data you see is just decoration for an argument that was already finished before anyone looked at the numbers.
I think we can do better. Data-driven journalism should start with the data, not the conclusions.
As a trained pollster, empirical journalist, and author of a book about polls and democracy, I started Strength In Numbers in part as an antidote to this tired way of covering politics and voter behavior. I want to be your “numbers guy” for American politics — and that is really what you get here. Here at SIN, no editorial board or investor collective is deciding what poll questions to ask so they can gin up ideologically useful results. There is no institutional pressure to only publish what’s viral or worth the space in print. I’m not working on a broader partisan ideological project or punching against certain groups to distinguish my brand from theirs. SIN is just me, working with some excellent partners, following the evidence where it goes. I want people who read SIN to have a better understanding of politics — that’s it.
That might sound like table stakes, but it’s actually fairly rare nowadays. FiveThirtyEight is gone, and most of the remaining data operations at major outlets are constrained by the same editorial incentives that shape everything else in media now: spend little resources per story, optimize X for clicks, fit Y to “the narrative,” don’t alienate subscribers. Even the independent analysts on Substack and social media tend to fall into predictable ideological lanes. In my assessment, there is a real opening for people who want to cover election results and public opinion simply for the sake of covering them well, from the point of view that democracy and the public matter, full stop.
But letting data lead doesn’t just mean making charts — it means taking public opinion seriously as an object of study. That means reading the academic work on voter psychology, looking carefully at what voters actually say in surveys and what might drive response patterns, and interrogating the limits of the data itself. I’m confident I’ve read more social science, written more statistical models, and conducted more polls than any other journalist/pollster, e.g., participating in public debates about how to win voters in the 2026 and 2028 elections. That expertise is what this newsletter is built on.
I should be clear, though: letting the data lead doesn’t mean I don’t have values. This newsletter gets its name from my book, which contains two chapters on why/how democracy works and why we should trust The People — and thus incorporate the polls into the democratic process. I think the American public is smarter and more coherent than most of the political class gives it credit for, and I think a functioning democracy requires that the people who govern actually listen to the people they represent. That’s the driving force behind every article here at SIN. When I criticize the Trump administration (which, given the data, I do pretty often), it’s not because I’m playing for a team, but because the evidence points in a clear direction and I’m not going to pretend otherwise. If the centrist pundit class is wrong about something that I know something about (again, something that is happening a lot lately), I’ll say that too.
Strength in numbers
So, my pitch to you has been that I’ll send original, differentiated analysis and commentary from someone with a commitment to showing his work, who also has a point of view about how public opinion should be measured and operates. And most of the time, all you have to give me in return is your attention. A small fraction of you are paying subscribers, for which I am very grateful; your support makes it possible for me to do this as a full-time job. And it is definitely a full-time job now — lately, I’m working close to 45-50 hours a week on analysis for this Substack. That’s way beyond my targets and not sustainable long term, but I’m figuring out a rhythm, and for the first time in my life, I feel like the hard work is actually worth it.
At any rate, I’m very happy that so many people have found it worthwhile to support me in these first nine months of SIN. The site just passed 58,000 total readers here, and the publication is on the verge of being a top-50 newsletter on Substack (currently #55 — I think we’ll hit 50 next month).
More strength, more numbers
So now, a few words about what’s working here at SIN and some changes I’m going to make to bring me closer to my goal.
First, the key weekly trio of posts. I think the weekly Deep Dive on Tuesday is a great product, and really seems to resonate with people. But it would be optimal for me to produce and send it out on Wednesday, so I’ll be doing some experiments with open rates and engagement over the next few months. The free Sunday Data Roundup and Friday Chart of the Week posts will stay where/how they are, with the addition that I’ll be testing out a short sponsorship for the occasional Friday post as a way to increase website revenue and business sustainability.
On the polling front, the monthly Strength In Numbers/Verasight survey has become one of the most-cited independent polls in political journalism. I’m very proud of that, and want to grow this segment of the newsletter. The natural way to scale up is speed; monthly isn’t always fast enough to contribute usable findings to the public, so soon I’ll be supplementing the monthly tracking poll with quicker-turnaround, self-contained surveys designed to inject hard data into narrower debates and discourses as they unfold. When pundits argue about what voters think/want, I want SIN to be the publication that actually does the hard work of going out and asking them.
Beyond writing and polling, the Strength In Numbers Podcast is also now a regular feature of our editorial calendar after a promising test run, and David Nir of The Downballot and I plan to continue our live recording sessions on Thursdays at 2:00 Eastern Time. I may add special, deeper dives with sources to really drill down on subjects, or introduce guests to the main pod. We’ll see. As a reminder, you can listen after the fact as a normal podcast, wherever you get your podcasts.
Finally, I’m also (begrudgingly) investing more time into discovery tools beyond text-based social media, primarily via short-form video on YouTube. Our podcast gets uploaded to the latter as well as Substack, and in other videos, I walk through charts or break down news in real time. The newsletter will always be home base for SIN, but I want to grow the publication’s impact, and that means meeting people where they are.
Also worth mentioning is that I’ve also been doing some private survey research on the side for clients in politics and the media. A lot of that work will become public before the midterms, but I’m trying to keep a hard line between the newsletter and my consulting work. To the extent the findings are interesting to readers here, I’ll share more when there’s more to say, probably via Substack Notes.
We’re in a moment in American politics where empirical political coverage matters more than usual. Big pronouncements about the electorate are everywhere. A startling number of them are wrong, mostly because no one did the legwork of actually asking voters about the options they’re faced with, or because an analyst/strategist ignored counterarguments that didn’t fit their model of the world.
A key input into democracy is democratic knowledge: the public’s ability to know itself, and to be known by the people who govern it. That’s what polling is for, and largely what this newsletter is for. In a democracy, the political class must respond to public opinion — but it’s important that what we are calling “public opinion” is measured carefully and discussed with care and nuance.
So, thanks for being part of the team. Strength In Numbers is just getting started.
I’m not going to ask you for a paid subscription today, but instead, if you think someone in your life would resonate with my work, please send this to them. Word of mouth has always been the highest-leverage way SIN grows.
Elliott -- Thanks for the update, and for your continued good work. One issue I'd like you to address in future polling is this: Many polls indicate that almost as many people are disapproving of the Democrats as they are disapproving of Trump; why is that? What specifically do people find unsatisfactory with the Democrats? Do they like their policy positions but feel they are ineffective at advancing those positions? [probably mostly magical thinking] Or do they think the Dems are taking the wrong policy positions? Our is it just a poorly informed "pox on both your houses"?
I am a retired English professor (Shakespearean and Renaissance Drama), so I cannot always follow the intricacies of the data at Strength in Numbers, but I subscribed because Elliott's method is exactly what I taught students about writing, whether in freshman composition, lower and upper division literature courses, or graduate courses: research and write to explore a hypothesis or accepted fact, and if you find that your original idea does not fit, then change that idea accordingly. When I write, I always start with a rough outline of what I think the argument will be, but by the end of the writing process, I have usually gone in another direction based on new evidence or ideas. Too many of the political "analysts" these days want to jump from the beginning to the end instead of spending time in the process of hammering out ideas. Perhaps that is part of living in an "instant" society. If more people did process oriented tasks, such as baking bread, gardening, etc., they would develop a healthy respect for what the process can teach us and how changing course can help us learn.