LSE - Small Logo
LSE - Small Logo

Blog Admin

November 1st, 2017

Canvassers tend to seek out supporters who are like themselves, and that’s not good for political participation.

1 comment

Estimated reading time: 10 minutes

Blog Admin

November 1st, 2017

Canvassers tend to seek out supporters who are like themselves, and that’s not good for political participation.

1 comment

Estimated reading time: 10 minutes

Petition canvassers play an important role as political recruiters by introducing citizens to political issues and seeking their support. But not much is known about how these canvassers decide whom to recruit or about their methods. Research by Clayton Nall, Benjamin Schneer, and Daniel Carpenter sets forth a model of political recruiting that changes depending on canvassers’ experiences, is constrained by geography, and conditioned on the relations between canvasser and prospect. They find that canvassers seek out supporters like them when going door-to-door – sometimes even following geographically and politically “inefficient” paths.

Political campaigns and movements try their best to bring new supporters into their fold, and the process poses deep and abiding challenges to campaigns and those who labor for them. Take, for example, the experience of Eli Reiter, who recounted his efforts gathering signatures for Eliot Spitzer’s 2013 campaign in an article for Gothamist.

Due to the high pay, I felt obligated to get as many signatures as I could. The West Village has many aging Jewish women and gay couples. These are my people. I realized the trick was to start with the most potent part to draw people in. “Hi, I’m with Eliot Spitzer for Comptroller of the City of New York. Would you like to sign our petition to put him on the ballot?” has a LOT of syllables. So I tried innumerable permutations of my pitch. 

“Sign for Spitzer?”

“I’m with Eliot Spitzer. Sign our petition please?”

“Eliot Spitzer for office. Sign please?”

Then I started barking, “SPITZER! SIGN FOR SPITZER!” It worked. Many people stopped and signed. Others told me I was crazy.


“Heavens no.”

“Heck no.”

“You must be kidding.”

“You’re joking right?”

“You must be joking.”

“That schmuck!”

“What nerve!”

“I’d never vote for a hellion!”

I was also insulted with what seemed like 30 different Yiddish words. Who knew Yiddish could be so versatile? 

Eli Reiter’s experience in New York City is telling. Reiter, a free-lance journalist, found himself enticed by the $800 a day he was offered to collect signatures to place former New York Governor Eliot Spitzer on the ballot for the post of New York City Comptroller.  A non-professional canvasser, Reiter faced the simple task of maximizing signatures. Largely unguided by campaign staff, Reiter was left to determine the best strategy for finding people to sign for Spitzer. He went to a neighborhood familiar to him, where he thought he would be most likely to harvest signatures. He picked the neighborhood because people like him lived there.  Yet once he found himself in the neighborhood, his early efforts to strike up a rapport ran up against the hard realities of capturing the attention of potential signatories.  After identifying a bustling neighborhood street with a large flow of potential signers, he faced the stream of passersby and experimented with strategies for getting their attention and their signatures. He changed course midstream, shifting from polite requests to “barking.”  While he received many negative responses and a few lessons in Yiddish, he nevertheless learned, through trial and error, how to prospect for signatures.

The experimental approach to canvassing described by Reiter is suggestive of several broader lessons about how political recruiters search for support. In our new research, we develop a new model of political recruitment that draws lessons from the experiences of petition canvassers in multiple geographic and historical contexts. We view petition canvassing as a process in which a canvasser may explore different locations at which to gather signatures, either by staying in one central location (a crossroads, shopping center, or town square) or by going from door to door. Our model emphasizes that a canvasser must weigh the value of trying a new place – including places the canvasser knows less about but that might nonetheless yield many new signatures – against remaining in a current location where the rate of signatures is better known. Our model also takes into account the differential costs borne by canvassers such as travelling to an unfamiliar environment or traversing a geographic boundary.

To test our model, we used original data gathered from a 2005-2006 anti-Iraq War initiative and 2008 nominating paper signature drive in Wisconsin as well as an 1839 antislavery campaign in New York City. By gathering completed petition signatory lists (containing the names and addresses of signatories) from these campaigns, we were able to take advantage of the sequential nature of petition signatures to reconstruct a picture of each canvasser’s strategy. By examining the path taken by a canvasser in geographic space, we could learn whether they walked a petition from door to door or laid it out in a central location.

The data reveals that canvassers were significantly more likely to walk door to door in search of signatures in neighborhoods where the demographic characteristics of the residents were similar to their own. For example, the middle-class, white Wisconsin canvassers who gathered signatures to protest the Iraq War were more likely to stay in predominantly white and middle class neighborhoods when they travelled door to door. Our analysis also revealed that canvassers appeared to follow a rational process in which canvassers quit or changed approaches when facing increased costs.

On the other hand, canvassers sometimes made politically inefficient choices based on a tendency towards “friends and neighbors” politics. For example, when US Representative Gwen Moore (D-WI and the state’s first black member of Congress) prepared for reelection in 2008 by gathering signatures to put her on the ballot, her canvassers relied overwhelmingly on the majority-black neighborhoods that had comprised Moore’s old State Senate district. They left signatures on the table in many other strongly Democratic-leaning neighborhoods and passed up an opportunity for performing campaign outreach and mobilization.

Because recruitment helps determine who participates in politics, we think that it is crucial to understand the details of how political recruiters search for support. Past research has already shown some of the glaring inequalities in participation – across characteristics such as race and income level – that exist. Our work provides one reason that political recruiters may not be reaching everyone and, in the process, sheds new light on the dynamics that feed inequality in participation.


Please read our comments policy before commenting.            

Note:  This article gives the views of the authors, and not the position of USAPP– American Politics and Policy, nor of the London School of Economics.

Shortened URL for this post:


About the authors

Clayton Nall – Stanford University
Clayton Nall is an Assistant Professor of Political Science at Stanford University. His research focuses on American political geography, with an emphasis on the role of the state and public policy in the creation of place-based interests. 


Benjamin Schneer – Florida State University
Benjamin Schneer Is an Assistant Professor in the Department of Political Science at Florida State University. 


Daniel Carpenter – Harvard University
Daniel Carpenter is Allie S. Freed Professor of Government in the Faculty of Arts and Sciences, and Director of Social Sciences at the Radcliffe Institute for Advanced Study at Harvard University.

About the author

Blog Admin

Posted In: Benjamin Schneer | Clayton Nall | Daniel Carpenter | Democracy and culture


LSE Review of Books Visit our sister blog: British Politics and Policy at LSE

RSS Latest LSE Events podcasts

This work by LSE USAPP blog is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported.