Ethnic Minority Analysts’ Participation in Public Earnings Conference Calls

• Published date: 01 December 2023
• Author: RACHEL W. FLAM,JEREMIAH GREEN,JOSHUA A. LEE,NATHAN Y. SHARP
• Date: 01 December 2023
• DOI: http://doi.org/10.1111/1475-679X.12504

DOI: 10.1111/1475-679X.12504

Journal of Accounting Research

Vol. 61 No. 5 December 2023

Printed in U.S.A.

Ethnic Minority Analysts’

Participation in Public Earnings

Conference Calls

RACHEL W. FLAM,∗JEREMIAH GREEN,†JOSHUA A. LEE,‡

AND NATHAN Y. SHARP†

Received 18 June 2020; accepted 20 July 2023

ABSTRACT

We investigate ethnic minority and nonminority sell-side analysts’ participa-

tion in public earnings conference calls. We f‌ind that minority analysts are un-

derrepresented in conference call Q&A sessions, and minority analysts who

do participate on the calls experience lower levels of prioritization than do

nonminority analysts. Minority analysts’ lower participation rates are partially

but not fully mediated by characteristics such as experience, work environ-

ment, and stock rating favorability. Additionally, f‌irm and conference call

f‌ixed effects mediate approximately half the magnitude of lower minority par-

ticipation rates. Extroverted minority analysts participate at higher rates, but

the negative association between minority status and conference call partici-

pation is exacerbated when calls are more time constrained, when executive

teams are less diverse, and when analysts are from less prestigious brokerage

houses. Overall, we document the underrepresentation of minority analysts

∗London Business School; †Texas A&M University; ‡Brigham YoungUniversity

Accepted by Douglas Skinner. We appreciate helpful feedback from an anonymous asso-

ciate editor, two anonymous reviewers, Michael Clement, Mike Drake, John Hand, Stephani

Mason, Dawn Matsumoto, Ken Merkley, Senyo Tse, Brady Twedt, Jim Westphal, Gwen Yu,

and participants at the 2020 Salem Center for Policy Ph.D. Symposium. We are grateful

for excellent research assistance from Christina Berger, Brigham Brau, Benjamin Harrison,

Madelyn Hill, Joshua Kercheville, Dakota Klein, Sean Wilson, and Kyle Zabadal. We acknowl-

edge generous f‌inancial support from BYU Marriott School of Business, London Business

School, and Mays Business School. An Online Appendix to this paper can be downloaded at

https://www.chicagobooth.edu/jar-online-supplements.

1591

© 2023 The Authors. Journal of Accounting Research published by Wiley Periodicals LLC on behalf of The

Chookaszian Accounting Research Center at the University of Chicago Booth School of Business.

This is an open access article under the terms of the Creative Commons

Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium,

provided the original work is properly cited, the use is non-commercial and no modif‌ications or

adaptations are made.

1592 r. w. flam, j. green, j. a. lee, and n. y. sharp

on earnings conference calls and provide evidence suggesting both analysts’

and managers’ choices inf‌luence minority analysts’ participation rates.

JEL codes: D22, G02, G20, G23, G24, G28, J15, J71, M40

Keywords: ethnic diversity; f‌inancial analysts; earnings conference calls

1. Introduction

Access to management of the companies they cover is among the most

highly valued services sell-side analysts provide to their institutional-investor

clients (Brown et al. [2016]). Sell-side analysts with management access

attract greater interest from institutional investors who depend on them

to facilitate access to management, and these analysts garner greater trad-

ing and other commissions (Jackson [2005], Juergens and Lindsay [2009]).

In our study, we investigate whether access to management differs for eth-

nic minority versus nonminority sell-side analysts. Specif‌ically, we examine

whether ethnicity is associated with analysts’ participation on public earn-

ings conference calls and, if so, what factors drive any disparities in partici-

pation rates for ethnic minority versus nonminority analysts.

Our study is important because ethnic minorities continue to be under-

represented in many corporate settings. For example, recent studies con-

clude that the C-Suites of the Fortune 100 ref‌lect a “dismal state of diver-

sity” (Larcker and Tayan [2020]), and corporate leadership opportunities

continue to elude even highly qualif‌ied ethnic minorities (Field, Souther,

and Yore [2020]). We focus on analysts’ participation in conference calls

because analysts play a signif‌icant role in information and price discovery

in capital markets (e.g., Brav and Lehavy [2003], Asquith , Mikhail, and Au

[2005], Beyer et al. [2010], Matsumoto, Pronk, and Roelofsen [2011], Der-

rien and Kecskés [2013]), and because access to management is a key input

to their production function (Mayew, Sharp, and Venkatachalam [2013],

Green et al. [2014], Brown et al. [2015]).

We gather a sample of 8,112 sell-side analysts and 94,582 quarterly earn-

ings conference calls between 2002 and 2017. To determine the ethnicity

of each analyst in our sample, we follow prior research and use the ex-

pertise of List Service Direct in matching analysts’ names to distinct ethnic

groups (Brochet et al. [2019]).1We classify each analyst as an ethnic minor-

ity or nonminority using a classif‌ication based on the U.S. Census Bureau’s

def‌initions.

We then investigate whether ethnic minority and nonminority analysts

have similar rates of participation (i.e., asking a question) in the Q&A

1Our measurement of ethnicity is similar to the way other papers use names to classify f‌i-

nancial professionals (Kumar, Niessen-Ruenzi, and Spalt [2015], Brochet et al. [2019]). How-

ever,recognizing that name-based ethnicity classif‌ication can be prone to error, we also collect

LinkedIn prof‌ile pictures for 4,233 analysts (52% of the analysts in our sample) and manually

examine whether List Service Direct’s categorizations appear to be accurate. Our extensive

manual verif‌ication suggests an error rate of less than 3.5% (see section 2.2).

minority analysts’ participation1593

session of public earnings conference calls. In this setting, which is often

carefully scripted by the management team (Lee [2016], Brown et al.

[2019], Cen et al. [2021]), we test whether analyst ethnicity is associated

with the likelihood that an analyst asks a question on the call. We also

consider whether analyst ethnicity is associated with different levels of

interactions with management in the form of the ordering on the call,

asking follow-up questions, or speaking more words.

We f‌irst examine whether differences exist in the conference call partic-

ipation rates between minority and nonminority analysts, because dispari-

ties based on ethnicity are important to identify regardless of the reason

the disparities exist. The null hypothesis is that there are no differences

in the conference call participation rates of the two groups. This null hy-

pothesis relies on several critical assumptions: that minority analysts are not

less qualif‌ied or less willing to participate in conference calls; that minor-

ity analysts do not differ in their experience, reputation, or performance

from nonminority analysts; and that managers do not favor nonminority

analysts over minority analysts in their conference call decisions. We later

test these assumptions to provide evidence on potential drivers of observed

differences in conference call participation. Our primary result is that mi-

nority analysts do not participate on conference calls at the same rate as

nonminority analysts. Specif‌ically, for minority analysts who cover a com-

pany, the percentage who participate on a call, ask the f‌irst question, or ask

a follow-up question is lower than the same percentage for nonminorities.2

Minority analysts also speak later and speak less. When we consider only

underrepresented minorities (Black, Latino, and Indigenous), we f‌ind that

underrepresented minority analysts also participate at lower rates than non-

minority White analysts.

For the call participation variables we measure, the difference in partici-

pation rates between minority and nonminority analysts may appear small

at f‌irst. For example, of analysts that cover a f‌irm, 42.4% of minority ana-

lysts participate on a call versus 44.1% of nonminority analysts. This 1.7%

difference is statistically signif‌icant, but interpreting what this magnitude

means in terms of its economic effects on analysts, f‌irms, or society is diff‌i-

cult. Prior research argues that even when inequalities in outcomes might

be small for a single event, it is the cumulative effect of these small oc-

currences that better ref‌lects the impact of any difference (Blank [2005],

Stolzenberg, D’Alessio, and Eitle [2013], Greenwald et al. [2015], Wallace,

2For discrete participation variables, we test the rate differences between minorities and

nonminorities. In the test of means, this is a test of the differences in the percent of minori-

ties versus the percent of nonminorities. Similarly, in regressions, we test probability models.

The estimated coeff‌icients in these models on the indicator variable for whether an analyst is

a minority compare the probability that the dependent variable is equal to one for minority

analysts with the probability that the dependent variable is equal to one for nonminority an-

alysts. By comparing probabilities rather than the number of minority and nonminority analysts,

our analyses control for the mechanical differences in each subpopulation of analysts.