Understanding the Factors that Affect the Incidence of Bellwether Counties: A Conditional Probability Model

• Author: Bernard Grofman,Haotian Chen
• DOI: http://doi.org/10.1177/10659129211057601
• Published date: 01 March 2023
• Date: 01 March 2023
• Subject Matter: Articles

Article

Political Research Quarterly

2023, Vol. 76(1) 119–126

© The Author(s) 2022

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/10659129211057601

journals.sagepub.com/home/prq

Understanding the Factors that Affect the

Incidence of Bellwether Counties: A

Conditional Probability Model

Bernard Grofman*and Haotian Chen*

Abstract

We update previous work on bellwethers in U.S. presidential elections. Comparing the most recent elections (2000-

2020) to those in earlier periods (1960-1980), we see a striking decline in the proportion of bellwe thers. We provide a

model linking this decline to conditional probability calculations that recognize that (a) a county’s predictive success

likelihood varies depending upon whether the winning candidate is going to be a Democrat or going to be a Republican;

(b) as polarization rises, the number of potential bellwethers declines; and (c) election competitiveness can matter, but

closer elections do not guarantee a greater likelihood of bellwethers. Indeed, we now have very close elections but a very

low likelihood of bellwethers.

Keywords

bellwether counties, presidential elections, polarization

Introduction

The goal of this essay is to offer a simple formal model,

involving conditional probabilities, to explain changes in

the likelihood that we will f‌ind bellwethers, that is, po-

litical subunits (such as counties) whose votes consistently

match those of the national election in some sequence of k

elections.In particular, we wish to show why the likelihood

of f‌inding bellwether counties that match presidential

outcomes has been declining.

Tufte opens his 1975 article on bellwethers (co-

authored with Sun) with this trenchant historical quote:

“Prior to the 1936 presidential election, the conventional

political wisdom had it that as Maine voted, so went the

rest of the nation. After the 46-state landslide, James

Farley, Roosevelt’s campaign manager, revised the the-

ory: ‘As goes Maine, so goes Vermont’.”And yet, the

belief in bellwether states or counties as predictors of U.S.

presidential elections is a phoenix that is constantly being

reborn. It is as if the devastating rebuttals by Tufte (1974,

chapter 3; 1975) or the equally devastating review of the

evidence in Hopkins (2017) never happened.

Zimny-Schmitt and Harris (2020), who draw on pre-

viously identif‌ied bellwethers, look for bellwether

counties in the period 1980-2016 and identify nineteen

such. In a 7 October 2020 essay in the New York Times,

Wasserman (2020) looks to ten allegedly bellwether

counties for evidence that Donald Trump is in serious

trouble. Articles after the 2020 election examining the

predictive power of previous bellwethers are found in

other sources, including the Wall Street Journal

1

and the

Epoch Times.

2

Similar stories appeared in other recent

presidential elections. Relatedly, after the 2020 election

was over, some supporters of President Trump, such as

J.B. Shurk (23 November 2020) review the evidence of

bellwether results to argue that there must have been

electoral fraud since Biden won the presidency while

losing all but a handful of bellwether states and counties.

3

Why is belief in the existence of bellwether states and

counties (especially the latter) so strong? We believe that

there are two main reasons.

University of California, Irvine, CA, USA

*Grofman is Distinguished Professor of Political Science and Jack W.

Peltason Chair of Democracy Studies at the University of California,

Irvine. Chen is a PhD student in Political Science at UCI. The participation

of the authors was supported by the Peltason Chair.

Corresponding Author:

Bernard Grofman, Department Political Science, University of California

System, School of Social Sciences UCI, SSPB 2291, Irvine, CA 92697,

USA.

Email: bgtravel@uci.edu