N Parameters
Using data from the 2006 Michigan primary election, we seek to examine the effect of social pressure through sending postcards on whether people will vote. This is to forecast effect of social pressure on the 2026 Texas gubernatorial general election. There is concern over the assumption of representativeness because our data is from Michigan while we’re trying to forecast causal effects in Texas. Michigan and Texas have different political cultures and expectations. We used a Bayesian regression model with the formula = voted ~ age_z + sex + treatment + voter_class + treatment*voter_class to determine the causal effect of different types of postcards (independent variable) on whether a person will vote (primary_06, the dependent variable). A postcard from neighbors, regardless of past voter frequency, resulted in a positive average treatment effect, indicating that pressure from neighbors may lead people to vote. Although the average treatment effect is greater than 0% for all voter frequency classes, the variation within voters who vote rarely is big.
\[y_{i} = \beta_{0} + \beta_{1} age\_z + \beta_{2}male_i + \beta_{3}civic\_duty_i + \\ \beta_{4}hawthorne_i + \beta_{5}self_i + \beta_{6}neighbors_i + \\ \beta_{7}Sometimes\ vote_i + \beta_{8}Always\ vote_i + \\ \beta_{9}civic\_duty_i Sometimes\ vote_i + \beta_{10}hawthorne_i Sometimes\ vote_i + \\ \beta_{11}self_i Sometimes\ vote_i + \beta_{11}neighbors_i Sometimes\ vote_i + \\ \beta_{12}civic\_duty_i Always\ vote_i + \beta_{13}hawthorne_i Always\ vote_i + \\ \beta_{14}self_i Always\ vote_i + \beta_{15}neighbors_i Always\ vote_i + \epsilon_{i}\]
Characteristic |
Beta |
95% CI 1 |
|---|---|---|
| (Intercept) | 0.16 | 0.14, 0.17 |
| age_z | 0.04 | 0.03, 0.04 |
| sex | ||
| sexMale | 0.00 | -0.01, 0.01 |
| treatment | ||
| No Postcard | — | — |
| treatmentCivicDuty | 0.02 | -0.03, 0.06 |
| Hawthorne | 0.01 | -0.03, 0.05 |
| Self | -0.01 | -0.05, 0.03 |
| Neighbors | 0.08 | 0.04, 0.12 |
| voter_class | ||
| Rarely Vote | — | — |
| voter_classSometimesVote | 0.11 | 0.10, 0.13 |
| voter_classAlwaysVote | 0.30 | 0.28, 0.32 |
| treatment * voter_class | ||
| treatmentCivicDuty * voter_classSometimesVote | 0.00 | -0.05, 0.04 |
| Hawthorne * voter_classSometimesVote | 0.01 | -0.04, 0.05 |
| Self * voter_classSometimesVote | 0.06 | 0.01, 0.10 |
| Neighbors * voter_classSometimesVote | -0.01 | -0.05, 0.04 |
| treatmentCivicDuty * voter_classAlwaysVote | -0.01 | -0.06, 0.04 |
| Hawthorne * voter_classAlwaysVote | 0.03 | -0.02, 0.08 |
| Self * voter_classAlwaysVote | 0.05 | -0.01, 0.10 |
| Neighbors * voter_classAlwaysVote | 0.00 | -0.06, 0.05 |
| 1
CI = Credible Interval |
||