6. Should I donate to campaigns?

The answer hinges on whether or not advertising actually drives votes. And it's complicated.

^ The dock of tears (less fun than the ring of fire)

It’s been a hard three weeks. I got the news that RBG died while in a cabin in upstate Minnesota with one of my oldest childhood friends. We left our phones and walked outside to sit on the dock in the cold. The only two people on the still lake in the dark, I sat behind her and held her while we cried. 

The next day we drove through a gorgeous forest whose leaves were just turning from green to scarlet and turmeric yellow, I thought about how many years it’d take before we lose this forest. 20 years? 30? All the climate migration models show that people will move north to escape increasingly hot weather, so who knows? Soon we might be logging to build condos in Northern Minnesota. 

After RBG’s death, the train wreck that was the presidential debate, Trump utterly predictably contracting coronavirus, and what seems like the most consequential election of our lifetime, it feels more imperative than ever to throw myself into doing something, anything, to abate the helplessness. 

My friends have thrown themselves into different things to deal: one friend is writing personal cards to send to voters urging them to turnout. Another friend started a weekly email update to track and share how much his friends have donated and phone banked. 

But to be honest, I got tired of people telling me to donate and phone bank on social media. One of the primary reasons I didn’t donate or phone bank for political candidates is that I didn’t believe, in my heart of hearts, that my dollar or my call impacts the outcome of the election. I had (have?) an image of the political machine as a thing that will keep churning regardless of me, beholden mostly to hedge fund managers and hyper-rational, well-meaning 30-somethings who work at Facebook (“But Anya, I work for Facebook Groups, and we build tools to connect veteran moms who sew face masks!” Whatever. Shut up, Jonathan.)

So what’s the truth? Does my donation actually help? If so, who should I donate to?

Do donations help candidates get elected?

One way to figure out whether donating will help your preferred candidate get elected is to follow the money. What do campaigns spend your money on? And will that help them get elected? 

What will a campaign spend your donation on?

Half of the spending  by presidential campaigns in 2020 went to purchasing media, mostly TV advertising (the other half went to staff salaries and various administrative costs, so I’m focusing on advertising as the primary vote-getting spending). That means one answer to whether or not your donation dollars “matter” for the election hinges on whether or nor ads work to influence voters. 

Does advertising drive votes?

Boy, oh boy. I spent three weeks and a zillion browser tabs trying to answer this question (aside: Why don’t we say “girl, oh girl”?).

Several teams of political scientists have causally measured the impact of advertising on voting by exploiting FCC regulations that assign bordering counties to different media markets. Campaigns buy ads at the media market level, not at the county level, meaning that people in some neighboring counties end up seeing a different number of TV ads by accident of being in different media markets.  Spenkuch and Toniatti (2018)  exploit these bordering counties to measure the impact of seeing a TV ad on voting for presidential races from 2004 - 2012. They find that while seeing a TV ad doesn’t influence whether voters turnout at the polls, TV ads do influence who people vote for once they’re there. So the short answer is yes: advertising does impact voting.

But of course, that’s not the whole story (if it was, this issue of Giving to Strangers would’ve been in your inbox a week and a half ago). 

This week, Sides, Vavreck, and Warshaw released a  working paper , which uses a similar approach to study the impact of advertising on voting outcomes for 2,250 races from 2000 — 2016 (not just presidential races). Their results agree with the first study: they find that advertising matters for votes. 

But they introduce important nuance around the effectiveness of ads across different kinds of races, nuance that is echoed in other literature I’ve read:

The effect of advertising is larger for “down-ballot” (i.e., Senate, House, gubernatorial, and state) races than presidential races.

…and the effect is larger in primary races 

…and for candidates who are not well-known. 

This makes a ton of sense: people are more affected by new information when they don’t know a lot to begin with, and people are less likely to know about candidates running for smaller offices. (Cue some frat boy after taking one stat class: “Well that’s trivial, of course people update more when their priors are weaker” sips Red Bull). By this same logic, by the way, ads run by a challenger should be more impactful than ads run by an incumbent, given that people are more familiar with the incumbent. 

But back to the specifics. Just how much further will your dollar go in smaller elections? Our second set of authors find that the effect of airing on TV ad is four times bigger in gubernatorial elections than presidential elections.

Not only does advertising have a larger effect in down-ballot races, but it does so at a lower cost. For presidential races, we estimate that the cost per vote is about $285.26 A $10 million advantage in an individual state might gain a candidate 35,000 votes, or enough to tip Nevada, Maine, Michigan, Wisconsin, and New Hampshire in the 2016 election. The cost per vote is much lower in other offices: about $142 in Senate races and $88 in governors races. 

Consider this nice summary from the 538 explainer

Basically, said Darrell West, vice president and director of governance studies at the Brookings Institution, advertising is useful for making voters aware that a candidate or an issue exists at all. Once you’ve established that you’re real and that enough people are paying attention to you to give you a decent chunk of money, you reach a point of diminishing returns (i.e., Paul Ryan did not have to spend $13 million to earn his seat). But a congressperson running in a close race, with no incumbent — or someone running for small-potatoes local offices that voters often just skip on the ballot — is probably getting a lot more bang for their buck.


  1. Even if partisanship determines which candidate wins in many races, TV ads can still influence people’s voting. 

  2. A rule of thumb: the lesser known the candidate / race, the more impactful their ads will be, the more your donation will matter.   

  3. Your donation is more likely to make a candidate viable, rather than make that candidate win. So one way to make a big difference is to donate to early stage, yet-unknown donors you really like. 

Which strangers should you give to?

The obvious thing you might be wondering is whether or not it would realistically matter if you donated to Joe Biden’s campaign. My read of the research (including stuff I didn’t cite in the post) says that the presidential race and Biden’s campaign is so big, so partisan, and so well-funded that your donation will likely make little difference. This is especially true given that so much of the funding for huge campaigns like Biden’s comes from Super PACs (a topic I’ll get into next week). 

Instead, save your money for several incredibly important (and just in reach!) toss-up Senate races. In Maine, Iowa, North Carolina, and Montana, challenger Democrats (who will be lesser known than the incumbents) have a good chance at winning. And right now is exactly when the research shows TV ads make the most difference. 

The importance of flipping the Senate cannot be overstated. Consider this reasoning from my friend Vishan (the one who started an email thread to drive donations) when I asked him about his donations: 

“Even if there’s a lot of uncertainty in whether my donation will help elect someone [to the Senate], the outcome [of flipping the US Senate] is so important that I’m willing to tolerate that uncertainty.”

He’s making an expected value argument: that when you multiply even the tiny probability that your donation will help flip the Senate by the sheer importance of flipping the Senate, the [benefit of donating] becomes very large. I suppose you can take this view with the presidential campaign as well, but it seems like my dollar will go further in the (relatively) less-concentrated Senate races.

Donate here: 

Maine:  Meet Sara - Sara Gideon for Maine

Montana:  Steve Bullock for Senate

Iowa:  Theresa Greenfield | For Iowa

North Carolina:  Home | Cal Cunningham, Democrat for U.S. Senate

If you found this post valuable or at least mildly apocalyptically entertaining, it would make my day if you shared Giving to Strangers with friends so they can subscribe:

Share Giving to Strangers