By Sheeva Azma
Engaging voters early in the campaign is key, writes Ben McGuire. Analyzing (and optimizing) large troves of campaign data can help.
Outside of political campaigns, data is crucial to help make important decisions — using analytics to drive decision-making at major companies is something that is pretty standard. The data is often plentiful — too plentiful — and the most difficult task is to analyze it and see in what ways it can aid in decision-making.
Data is increasingly becoming a large part of Democratic presidential campaigns, and it’s in part thanks to the work of a single person!
In 2019, Ben McGuire, a UPenn grad with a background in economics and history, wrote his Master’s thesis at the Harvard Kennedy School of Government about what aspects of organizing in political campaigns are most effective based on data.
McGuire’s thesis is called “Scaling the Field Program in Modern Political Campaigns: Investigating Determinants of Capacity in Mobilizing and Organizing.” You can read his thesis here.
Impressively, both before and after leaving Harvard, McGuire used his data background to support a number of high-profile women Democratic presidential candidates. While at Harvard, McGuire did some data hacking with the Hillary Clinton campaign. Afterwards, he served as a data lead for Elizabeth Warren’s presidential campaign. From there, he jumped to the Biden campaign, helping them win in 2020.
After using data to win a presidential campaign, McGuire leveraged his data background to help support the New York City COVID vaccine supply chain. Then, he did a brief stint in the United States Digital Service (the precursor to Elon Musk’s Department of Government Efficiency or DOGE for short) and got back to his political organizing adventures as Analytics Reporting Director for the Kamala Harris campaign. With the Harris campaign, he built the first ever data asset to track all campaign outreach to improve the campaign’s interactions.
When writing our guide for scientists interested in political campaigns, I encountered McGuire’s 47-page thesis, which looks at the data on field organizing in political campaigns.
If you don’t know what field organizing is, check out the guide I wrote, where I discuss what field is:
“The field department [of a political campaign] is what one typically thinks of when they think of political campaigns. It focuses on grassroots organizing, voter outreach, and get-out-the-vote efforts. Field is the ground operations of the campaign and therefore must be closely linked to other areas such as Communications and Digital.
Key responsibilities of the Field Department include direct voter contact via:
● Canvassing
● Phone and text banking
● Digital field organizing teams on social media
● Local event organization
● Managing campaign offices
● Voter records management”
In other words, field is everything you’d typically expect from a political campaign — getting out on the campaign trail, going door-to-door, and meeting people and convincing them to vote for you (or your candidate). For smaller campaigns, a small team may be tasked with all of the different campaign functions: science PhD and former Elizabeth Warren staffer, Kristin Hook, who challenged Chip Roy (TX-21) in the U.S. House, went door-to-door herself for a grueling several months. However, presidential races these days are so large that they have several different departments with many different functions — in other words, they work as a large, temporary, company.
Knowing all of this didn’t make me any less surprised to stumble upon McGuire’s Master’s thesis; I did not realize people were studying field organizing in any sort of data-driven way. Sure, nerds from Facebook, Google, and Twitter had helped elect Barack Obama by creating digital infrastructure for donations and so on, but the STEM-politics duality felt like it had died out after that. In his thesis, McGuire argues that engaging volunteers early in the campaign cycle in specific ways — namely, in one-on-one interactions — can help presidential campaigns do more.
I spent nine months as an unpaid volunteer organizing phone banking for Kamala Harris in Arizona (remotely — and in Spanish!) — partly with the official Harris campaign, but mostly through the Arizona Democratic Party — and, while we did have some data reporting, in the end, phone banking did not feel like it was too effective. That’s apparently because, according to McGuire, in-person canvassing, such as going door-to-door or going to rallies, is the gold standard in field organizing, while phone events can be useful for other functions, such as rallying volunteers. Phone banking is not as effective for gaining people’s votes, writes McGuire, citing the data.
Another takeaway from McGuire’s report is that one-on-one interaction with volunteers, and thinking of them as crucial assets with unique skills, rather than just a herd of volunteers who will do what you want to elect a candidate, is a factor for success. To be honest, we could have used that attitude in our efforts in Arizona. We certainly had the most talented field organizers, but the resources were low and our desire to affect change very high. A better state-level strategy, aligned with what works best for federal presidential campaigns as described by McGuire, could have helped.
Lastly, I think scientists are the perfect campaign volunteers. We have the right skill sets to be analytical and methodical in canvassing, and facts-oriented and mission-driven enough to make the case for why our candidate is simply the best. Perhaps it is no surprise that I kept running into new friends in science and technical fields in my work on the Arizona campaign trail (which was both in-person and virtual).
You’ll have to read the report, written in collaboration with the national Democratic National Committee analytics team, yourself for more about ways to optimize field organizing (often just called “field” in political campaign circles).
Fancy Comma, LLC 