Relational campaigning

Lennard Metson

2024-12-11

Presentation outline


1. ⚛️ Defining relational campaigning

2. 📚 Academic work & research in the UK

3. 🧪 Results: petition sharing RCT

4. 📋 Exploratory results: survey

5. 🔜 Next steps

Presentation outline


1. ⚛️ Defining relational campaigning

2. 📚 Academic work & research in the UK

3. 🧪 Results: petition sharing RCT

4. 📋 Exploratory results: survey

5. 🔜 Next steps

⚛️ Defining relational campaigning

⚛️ Defining relational campaigning

⚛️ Defining relational campaigning

⚛️ Why relational? Effectiveness

  • Easy for campaigns to contact supporters; easy for supporters to contact friends; but hard for campaigns to contact supporters’ friends
  • Social connections more trusted than the campaign representatives → the contact is potentially more effective

⚛️ Why relational? Scale

  • Supporters can contact multiple voters they know at once without additional effort from the campaign

⚛️ Why relational? Scale

  • Supporters can contact multiple voters they know at once without additional effort from the campaign
  • Voters contacted by the suspporter might contact people they know → potential cascades

Presentation outline


1. ⚛️ Defining relational campaigning

2. 📚 Academic work & research in the UK

3. 🧪 Results: petition sharing RCT

4. 📋 Exploratory results: survey

5. 🔜 Next steps

📚 Relational campaigning in academic research

the most influential canvassers are the rest of the people in the voter’s family […] if party workers or civic organizations want the greatest return on the use of their scarce services, they should make contact with the most politicized member of a household […] and motivate that member to bring all the other members to the polls

📚 Research in the UK

  • Studies in the UK find that when you mobilise voters, you also mobilise their (uncontacted) housemates (like in 🇺🇸) and (uncontacted) neighbours (unlike in 🇺🇸) (Foos and Rooij 2017; Foos et al. 2021)
  • This shows the effect of GOTV contact “spills over” to social connections

📚 Research in the UK

  • Studies in the UK find that when you mobilise voters, you also mobilise their (uncontacted) housemates (like in 🇺🇸) and (uncontacted) neighbours (unlike in 🇺🇸) (Foos and Rooij 2017; Foos et al. 2021).
  • This shows the effect of GOTV efforts “spillsover” to social connections

📚 Research in the UK

Mechanisms:

  1. demonstration: contacted voters vote, connections see them vote which acts as a reminder
  2. recommendation: contacted voters explictly ask their connections to vote

Presentation outline


1. ⚛️ Defining relational campaigning

2. 📚 Academic work & research in the UK

3. 🧪 Results: petition sharing RCT

4. 📋 Exploratory results: survey

5. 🔜 Next steps

🧪 Relational petition recruitment

🧪 Relational petition recruitment

🧪 Relational petition recruitment

🧪 Relational petition recruitment: Who recruited?

🧪 Relational petition recruitment: Who recruited?

🧪 Why don’t (most) supporters recruit?

🧪 Why don’t (most) supporters recruit?

  • Supporters don’t attempt to recruit others (first-stage failure)
  • Supporters’ attempts to recruit others fail (second-stage failure)

🧪 Why don’t (most) supporters recruit?

  • Supporters don’t attempt to recruit others (first-stage failure)
  • Supporters’ attempts to recruit others fail (second-stage failure)

🧪 Why don’t (most) supporters recruit? First-stage

🧪 Why don’t (most) supporters recruit? First-stage

Why focus on the first-stage?

Explainations for first-stage failure: supporters…
  • …underestimate their ability to get others to do things (Flynn and Lake 2008)
  • …feel embarassed about talking about politics with their connections (Gerber et al. 2012)
  • …aren’t motivated enough to ask others to engage in political action

Presentation outline


1. ⚛️ Defining relational campaigning

2. 📚 Academic work & research in the UK

3. 🧪 Results: petition sharing RCT

4. 📋 Exploratory results: survey

5. 🔜 Next steps

📋 What do people ask others to do?

📋 What do people ask others to do?

📋 What do people ask others to do?

📋 Who asks? Random forest models

Which individual-level characteristics predict asking?

📋 Who asks? Random forest models

Which individual-level characteristics predict asking?

📋 Who asks? Political attention

📋 Who asks? Political attention

📋 Who asks? Age

📋 Who asks? Age

📋 Who asks? Party