Bounding the Effect of Persuasion with Monotonicity Assumptions: Reassessing the Impact of TV Debates

Televised debates between presidential candidates are often regarded as the exemplar of persuasive communication. Yet, recent evidence from Le Pennec and Pons (2023) indicates that they may not sway voters as strongly as popular belief suggests. We revisit their findings through the lens of the persuasion rate and introduce a robust framework that does not require exogenous treatment, parallel trends, or credible instruments. Instead, we leverage plausible monotonicity assumptions to partially identify the persuasion rate and related parameters. Our results reaffirm that the sharp upper bounds on the persuasive effects of TV debates remain modest.

💡 Research Summary

This paper revisits the persuasive impact of televised presidential debates by focusing on the “persuasion rate,” an causal measure that captures the proportion of individuals who change their behavior because of a persuasive message. Unlike most of the existing literature, which relies on strong exogenous‑treatment assumptions or instrumental variables to point‑identify the persuasion rate, the authors develop a robust partial‑identification framework that requires only two monotonicity assumptions: monotone treatment response (MTR) and monotone treatment selection (MTS).

MTR posits that the potential outcome under treatment (watching the debate) is never lower than under control; in the context of the study this means a debate can only increase vote‑choice consistency, never decrease it. MTS assumes that individuals who are surveyed later (i.e., after the debate) are at least as likely to be “successful” (have consistent vote choices) as those surveyed earlier. This is a weak, behaviorally plausible condition that does not demand full control for all confounders.

Under these assumptions the population is partitioned into three latent types: never‑persuadable (NP), already‑persuadable (AP), and treatment‑persuadable (TP). The observable data provide the conditional probabilities P(Y=0|D=1) and P(Y=1|D=0), which serve as sharp lower bounds on the shares of NP and AP, respectively, while the unconditional margins give the corresponding upper bounds. Consequently, the share of TP—identical to the average treatment effect (ATE) under MTR—lies between these bounds.

The persuasion rate θ = P(TP) /