The predictive power of subjective probabilities

Probabilistic and deterministic polling in the Dutch 2017 election

Jochem de Bresser*, Arthur van Soest

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

Abstract

The paper evaluates the predictive validity of stated intentions for actual behaviour. In the context of the 2017 Dutch parliamentary election, we compare how well polls based on probabilistic and deterministic questions line up with subsequent votes. Our empirical strategy is built around a randomized experiment in a representative panel. Respondents were either asked which party they will vote for or were asked to allocate probabilities of voting for each party. The results show that probabilities predict individual behaviour better than deterministic statements for a large majority of respondents. There is, however, substantial heterogeneity in the predictive power of subjective probabilities. We find evidence that they work better for those with higher probability numeracy, even though probability numeracy was measured 8 years earlier.

Original languageEnglish
Pages (from-to)443-466
JournalJournal of the Royal Statistical Society, Series A
Volume182
Issue number2
DOIs
Publication statusPublished - Feb 2019

Keywords

  • Elections
  • Predictive validity
  • Probabilistic polling
  • Subjective probabilities

Cite this

@article{fc7a1253959848a5ad294afb3a7c9ab6,
title = "The predictive power of subjective probabilities: Probabilistic and deterministic polling in the Dutch 2017 election",
abstract = "The paper evaluates the predictive validity of stated intentions for actual behaviour. In the context of the 2017 Dutch parliamentary election, we compare how well polls based on probabilistic and deterministic questions line up with subsequent votes. Our empirical strategy is built around a randomized experiment in a representative panel. Respondents were either asked which party they will vote for or were asked to allocate probabilities of voting for each party. The results show that probabilities predict individual behaviour better than deterministic statements for a large majority of respondents. There is, however, substantial heterogeneity in the predictive power of subjective probabilities. We find evidence that they work better for those with higher probability numeracy, even though probability numeracy was measured 8 years earlier.",
keywords = "Elections, Predictive validity, Probabilistic polling, Subjective probabilities",
author = "{de Bresser}, Jochem and {van Soest}, Arthur",
year = "2019",
month = "2",
doi = "10.1111/rssa.12409",
language = "English",
volume = "182",
pages = "443--466",
journal = "Journal of the Royal Statistical Society, Series A",
issn = "0964-1998",
publisher = "Wiley-Blackwell",
number = "2",

}

TY - JOUR

T1 - The predictive power of subjective probabilities

T2 - Probabilistic and deterministic polling in the Dutch 2017 election

AU - de Bresser, Jochem

AU - van Soest, Arthur

PY - 2019/2

Y1 - 2019/2

N2 - The paper evaluates the predictive validity of stated intentions for actual behaviour. In the context of the 2017 Dutch parliamentary election, we compare how well polls based on probabilistic and deterministic questions line up with subsequent votes. Our empirical strategy is built around a randomized experiment in a representative panel. Respondents were either asked which party they will vote for or were asked to allocate probabilities of voting for each party. The results show that probabilities predict individual behaviour better than deterministic statements for a large majority of respondents. There is, however, substantial heterogeneity in the predictive power of subjective probabilities. We find evidence that they work better for those with higher probability numeracy, even though probability numeracy was measured 8 years earlier.

AB - The paper evaluates the predictive validity of stated intentions for actual behaviour. In the context of the 2017 Dutch parliamentary election, we compare how well polls based on probabilistic and deterministic questions line up with subsequent votes. Our empirical strategy is built around a randomized experiment in a representative panel. Respondents were either asked which party they will vote for or were asked to allocate probabilities of voting for each party. The results show that probabilities predict individual behaviour better than deterministic statements for a large majority of respondents. There is, however, substantial heterogeneity in the predictive power of subjective probabilities. We find evidence that they work better for those with higher probability numeracy, even though probability numeracy was measured 8 years earlier.

KW - Elections

KW - Predictive validity

KW - Probabilistic polling

KW - Subjective probabilities

U2 - 10.1111/rssa.12409

DO - 10.1111/rssa.12409

M3 - Article

VL - 182

SP - 443

EP - 466

JO - Journal of the Royal Statistical Society, Series A

JF - Journal of the Royal Statistical Society, Series A

SN - 0964-1998

IS - 2

ER -