Working Paper: NBER ID: w14206
Authors: Alan Gerber; Daniel Kessler; Marc Meredith
Abstract: During the contest for Kansas attorney general in 2006, an organization sent out 6 pieces of mail criticizing the incumbent's conduct in office. We exploit a discontinuity in the rule used to select which households received the mailings to identify the causal effect of mail on vote choice and voter turnout. We find these mailings had both a statistically and politically significant effect on the challenger's vote share. Our estimates suggest that a ten percentage point increase in the amount of mail sent to a precinct increased the challenger's vote share by approximately three percentage points. Furthermore, our results suggest that the mechanism for this increase was persuasion rather than mobilization.
Keywords: Direct mail; Voter turnout; Vote choice; Regression discontinuity
JEL Codes: D70
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Difference-in-difference-in-difference specification (C32) | Control for pretreatment differences (C90) |
| Direct mail campaign criticizing the incumbent Republican Attorney General (K16) | Increase in vote share of the Democratic challenger (D79) |
| Proportion of voters receiving mail (K16) | Increase in vote share of the Democratic challenger (D79) |
| Direct mail campaign (L87) | Increase in vote share of the Democratic challenger (D79) |
| Direct mail campaign (L87) | Persuasion (D91) |
| Direct mail campaign (L87) | No significant effect on overall voter turnout (K16) |
| Fuzzy regression discontinuity design (C24) | Comparison of precincts with different mail receipt levels (L87) |