Does it Help or Hurt Kerry if Nader is on the Ballot ? ∗

Does it Help or Hurt Kerry if Nader is on the Ballot? We report results of a pair of experiments designed to test whether Ralph Nader’s presence on the ballot in 2004 makes some voters more, not less, likely to vote for John Kerry. Intuitive spatial considerations suggest Nader should hurt Kerry, because most voters placed Nader left of Kerry and Kerry left of Bush. But a psychological theory based on the tradeoff contrast principle and the extremeness aversion principle suggests that, precisely because Nader is left of Kerry, Nader’s presence on the ballot may help Kerry. We use estimates of treatment effects supported by matching and a simple spatial model, binary and trinary logit models and GEV choice models to assess the effects Nader’s presence may have. One experiment is conducted using a sample of 566 university students, and the other experiment uses a national sample of 1,048 adults obtained through the NSF-funded Time-sharing Experiments for the Social Sciences project. We use matching methodology both to estimate average treatment effects and to construct matched samples in which to estimate the choice models. The student data support the psychological theory: when Nader is in the choice set, moderate voters see themselves as closer to Kerry and are somewhat more likely to vote for Kerry. Such effects do not occur for other voters. In the national sample the conditions for the theory to apply do not hold: most respondents do not place Nader left of Kerry. Nader hurts Kerry among liberals, but hurts Bush among conservatives. The reason for that is not captured by the intuitive spatial story, but reflects the fact that when Nader is in the choice set all voters’ choices are more sharply aligned with their spatial placements of the candidates.