Case Study 15.1: The Obama Ground Game — Field Organizing as a Multiplier

Background

In the spring of 2007, a junior senator from Illinois named Barack Obama was running for president against Hillary Clinton and a large Democratic primary field. His campaign was outspent and out-structured by Clinton in most states. But the Obama campaign was doing something unusual: rather than concentrating on a few large primary states and building centralized get-out-the-vote operations in each, it was attempting to build a distributed network of volunteer "neighborhood teams" in caucus and small primary states that conventional wisdom had written off.

By the time of the general election against John McCain, the Obama campaign had developed this neighborhood team model into one of the most extensively analyzed ground game operations in American political history. The campaign deployed approximately 700 paid field organizers in battleground states, supported by 2.2 million volunteers, organized into geographic units of 20–30 households. Each neighborhood team leader was responsible for recruiting, training, and deploying their neighbors as volunteers.

The Evidence for Ground Game Effects

Political scientists David Plouffe, Jeremy Bird, and academic researchers who later studied the 2008 operation documented several specific measurements:

Voter registration: The Obama campaign registered approximately 2.2 million new voters in battleground states. Post-election analysis by Catalist found that newly registered Obama supporters voted at significantly higher rates in 2008 than newly registered voters in previous cycles — a finding attributed to the campaign's intensive follow-up contact with its own registrants.

Contact rates: The Obama field program achieved contact rates (actual conversations with voters, not just door knocks) that substantially exceeded industry benchmarks. The neighborhood team model, which put local volunteers on their own blocks, produced contact rates 30–40 percent higher than traditional phone-bank-and-canvass operations.

State-level analysis: Political scientist Joshua Darr and Matthew Levendusky conducted a systematic analysis of Obama's field office presence and county-level vote share, finding that counties with Obama field offices outperformed expectations by approximately 1 percentage point on average, controlling for a comprehensive set of predictors. The effect was concentrated in competitive counties where the field program was most active.

The Multiplier Mechanism

Researchers identified several mechanisms by which the neighborhood team model generated larger effects than conventional canvassing:

Social network activation: When volunteers canvassed their own neighbors, they were operating within existing social networks. A conversation with a neighbor who knows you is qualitatively different from a conversation with a stranger. The commitment obtained from a neighbor may be more durable because it activates social accountability.

Volunteer recruitment cascade: Volunteers recruited by local team leaders brought in their own social networks. Each volunteer recruited (on average) 2.3 additional contacts who participated in campaign activities. This network multiplication effect meant that each paid field organizer's recruitment work had a much larger downstream impact than a model assuming no cascade would predict.

Local knowledge: Neighborhood teams had information about their blocks that centralized operations lacked: which households had participated in prior years, which residents had recently moved, which local issues were most salient. This information allowed adaptive targeting that modestly improved contact quality.

The Academic Debate

Not all researchers agree that the Obama ground game was as decisive as the campaign's partisans claim. Several critical perspectives:

Selection bias in field office placement: Field offices were opened in competitive counties, which are also the counties where external factors (national environment, candidate quality, advertising) are most active. Separating the field office effect from these confounding factors is methodologically challenging. Darr and Levendusky's county-level analysis, while careful, cannot fully rule out selection effects.

Counterfactual issues: McCain's campaign was underperforming its structural baseline in 2008 partly because of the national environment (an unpopular Republican president, a collapsing economy). Some of what looks like Obama ground game effect may simply be favorable national tailwinds being especially strong in counties where Obama was organizing.

Scale and replication: The Obama 2008 operation was unique in its scale, resources, and the enthusiasm of its volunteer base (a first-time candidate with enormous grassroots excitement). The model has not consistently replicated to the same effect in subsequent cycles, raising questions about whether the effect was generalizable or specific to the 2008 moment.

Lessons for the Garza Campaign

The Obama 2008 case offers several lessons for Maria Garza's operation:

The multiplier matters. A field program that generates volunteer cascade and social network activation can have effects substantially larger than a simple headcount of volunteer-hours would predict. Nadia's field deployment model, which treats each canvassing contact as equally valuable regardless of who makes the contact, may systematically underestimate the value of contacts by neighbors in tight-knit communities.

Contact quality vs. contact volume. The Obama campaign's insight — that a smaller number of genuinely high-quality conversations in voters' own neighborhoods was more valuable than a large volume of scripted phone contacts — echoes the experimental evidence on deep canvassing. The operational challenge is that quality is harder to monitor and enforce at scale than volume.

Infrastructure investment pays back. The Obama campaign's early investment in field organizer training and neighborhood team infrastructure had effects that compounded over time. The Garza campaign, running in a single-state Senate race rather than a 50-state presidential campaign, has less opportunity for this compounding. But the principle applies: early investment in volunteer infrastructure produces returns in the final push.

Discussion Questions

1. How would you design a study to cleanly identify the causal effect of a ground game program in a single Senate race? What are the key obstacles?

2. The Obama neighborhood team model requires highly motivated volunteers willing to canvass their own neighbors. In what political contexts is this motivation likely to be highest, and does the Garza race have those conditions?

3. If campaign ground game effects can be as large as 1–2 percentage points in competitive states, why don't all campaigns maximize their field investment at the expense of television advertising?

4. What ethical considerations arise when campaigns build neighborhood team structures that rely on the personal relationships and social trust of community members? Is leveraging social accountability for political mobilization different from other forms of community organizing?