That map from the New York Times about Bernie’s Fundraising is Going Viral Again

This Map is the Bane of my Existence

This piece is a part of my “post this to win the argument series” - There are a number of dumb repetitive internet debates that play out on Twitter seemingly word for word, over and over again. Instead of leaving the same reply that I have left a dozen times, I have instead decided to place all of my thoughts into one “copy and paste-able” article. Feel free to post this piece in response to anyone that is still sharing the New York Times 2020 Primary fundraising map.

I see that you have posted this map online as supporting evidence in your argument that Bernie Sanders was an overwhelmingly popular candidate in the 2020 Democratic Primary. There is a good chance that, in posting this map, you have argued that “the Democrats” or “The DNC” rejected Bernie despite his clear popularity with average people. In my experience, about 60% of posts that use this map also include a whiff of conspiratorial thinking. If someone is sharing this article with you, it’s to let you know that this map does not argue what you think it argues.

First and foremost, this map shows which candidate got the greatest proportion of individual small dollar donations in specific geographies. Most detractors in your replies will tell you that “fundraising” is not the same as “voting.” Bernie may have been popular with small dollar donors but that clearly did not translate into actual votes. Bernie Sanders won fewer states than Joe Biden. And in Iowa, specifically, Bernie lost to the no-name Mayor of South Bend, Indiana. It’s plainly clear that this map is not a measure of popularity because, if it was, Bernie Sanders would likely be President of the United States right now.

Here’s the map that matters! This shows the results of the 2020 Democratic Primary. Dark Blue indicates States that Joe Biden won. Green indicates States that Bernie Sanders won. As you can see, Bernie lost most of the States in which he had a fundraising edge.

One reason why the New York Times Map is such a uniformly awful tool is that it represents a very specific snapshot in time. The data in that map comes from the first fundraising disclosure in the 2020 primary race, which happened at the end of July in 2019. The first actual electoral contest, the Iowa Caucus, happened seven months after this data came out.

The map you posted simply does not accurately portray the state of the race when actual voting took place. Instead, it seems to suggest that Bernie Sanders was more popular with donors than the other candidates half a year before anyone started voting.

That’s honestly in-line with my expectations, given Bernie had high name recognition from his 2016 campaign. He was the only candidate in the race that had run for President four years earlier and, at this point in the race, he had several huge advantages over the other candidates.

For instance, In 2016, Bernie was the only serious alternative to Hillary Clinton, and he built a powerful fundraising apparatus to pay for his campaign. By 2019, he had a four year head-start in building an email list, cultivating small dollar donors, and raising money from every corner of the country.

Bernie, knowing he wanted to run for President again in 2020, refused to share his donor list with the DNC after the 2016 election. That data could have helped downballot Democrats to beat their Republicans opponents. But Bernie chose to keep hold of his list as proprietary information because he knew that it would give him a huge advantage over other candidates in 2020, who would have to start from scratch.

It’s important to remember what this map is actually measuring: not support, not enthusiasm, but existing fundraising infrastructure in July of 2019.

You will notice that the other candidates that appear on the map are localized within their own states. Amy Klobuchar does very well in Minnesota. Steve Bullock does great in Montana. Elizabeth Warren shows up in Boston and its suburbs. This is because each of those candidates previously ran for office in their respective states, and they transferred their existing fundraising lists from their Senate and Gubernatorial campaigns to their Presidential campaign.

Bernie Sanders, having run a nationwide campaign once before, had a fleshed-out list of donors from Utah, Missouri, and Kentucky (along with many other places). Nobody else had that kind of data, so he easily outraised all the other candidates in those geographies.

Critically, in the places where the most Democrats live, Bernie’s fundraising advantage did not hold. When you zoom in on the map, you will notice that America’s major cities, like New York, DC, and Chicago, are not Bernie-blue. This brings us to another flaw in the NYT Map.

The map you posted is not weighted for population density. This is a common problem when data is placed on maps; as you might expect, there are far more people in America’s major cities than there are out in the countryside. Painting the countryside in Bernie’s colors makes him appear to have more support than he actually does.

The population density problem is especially vexing because this dataset measures the kind of people that donate to Democratic Campaigns. Democratic donors are overwhelmingly city-dwellers. Meaning the population density effects that you would expect with any data mapping project are exacerbated to an extreme degree; as a result, painting any major city on this map your color corresponds to far more supporters, and a lot more money, than painting most States. If you take the time to read the New York Times article that this map comes from, you’ll see that they deliberately highlight these population density disparities. The authors of the article point out that:

“Ms. Warren has more donors in a single ZIP code in Brooklyn than any candidate has in the entire state of Mississippi. Mr. Buttigieg has more donors in a single ZIP code in Washington, D.C., than any candidate has in South Dakota.”

Bernie has a lot of geographic coverage in rural areas, and that makes it look like he has a lot of support. But land doesn’t vote. It also doesn’t donate to Democratic Primary Campaigns.

Those that get caught up on this donor map are often making the same mistake that Republicans do when they post memes like this one:

Obviously, there are Blue States. Population Density makes America look less Blue than it actually is.

The truth is, sometimes maps are a bad way of conveying data. Looking at the NYT map, it’s not immediately clear how large of a lead Sanders has in fundraising. And it’s difficult to tell where the other candidates stand in relation to him and each other. A better of way of showing which candidates attracted the most donors would be a simple bar graph, like this one:

As you can see, Bernie Sanders had a large lead in the number of donors in July of 2019. But it’s not nearly as overwhelming as the map would make it seem.

It’s also worth noting that the number of donors is not equivalent to the amount of money raised. The New York Times makes a point of explaining that the average donation to the Sanders Campaign was only about $46 per donation. In contrast, each donation to Biden or Harris averaged about $80. I’m sure you will have feelings about that statistic; I include it only to point out that this dumb, stupid map does an extremely poor job of explaining where the candidates stand money-wise. For reference, the topline amounts looked like this:

As you can see, Pete Buttigieg raised almost as much money from small donors as Bernie Sanders did! This is from the same dataset - just presented in a different way!

All this to say, the map that you have shared on Twitter.com (or wherever) does not say what you think it says. At best, it is a measurement of existing fundraising infrastructure in July of 2019 – a skewed metric that is more rooted in electoral history than any actual ideological support. It simply does not provide an accurate measurement of the candidates’ popularity. And we know that because Bernie Sanders lost the election!

At best, this map is a misleading way of presenting information. It butchers the data that it’s trying to convey and its own creators include extensive caveats alongside it, trying to convince their readers not to use it in the way you have used it. It’s also an extremely outdated snapshot of the campaign landscape; it was outdated when the first caucusgoer arrived at her polling place in Iowa, it was outdated on Super Tuesday of 2020, it remains outdated as I write this piece in 2025, and it will continue to be outdated when you eventually read this paragraph.

This map is not a good tool for understanding the world. It also does not provide a strong argument for supporting Bernie Sanders. If you look at this map and see Bernie as an overwhelming frontrunner, you’re then forced to reckon with his underwhelming electoral performance. The fact that Bernie raised more money than any other candidate and still managed to lose the primary suggests that he is uniquely bad at running for President! If I supported Bernie Sanders, I would not want to highlight this fact!

Your efforts are better spent elsewhere; please spend your time pitching people on the merits of Medicare for All, or rambling about how great the Green New Deal would be. This map does not support the electability argument for Bernie Sanders! Please stop feeding this awful image to the algorithm!