Ownership ties, conflict of interest, and the tone of news

• Date: 01 June 2020
• Author: Carlo Raimondo,Emanuele Bajo,Marco Bigelli
• DOI: http://doi.org/10.1111/eufm.12237
• Published date: 01 June 2020

Eur Financ Manag. 2020;26:560–578.wileyonlinelibrary.com/journal/eufm560

|

© 2019 John Wiley & Sons Ltd.

DOI: 10.1111/eufm.12237

ORIGINAL ARTICLE

Ownership ties, conflict of interest, and the

tone of news

Emanuele Bajo

1

|

Marco Bigelli

2

|

Carlo Raimondo

3

1

Department of Economics, University of

Bologna, Bologna, Italy

2

Department of Management, University

of Bologna, Bologna, Italy

3

Institute of Argumentation, Linguistics

and Semiotics, Università della Svizzera

italiana, Lugano, Switzerland

Correspondence

Emanuele Bajo, Department of

Economics, University of Bologna, Piazza

Scaravilli 2, 40126 Bologna, Italy.

Email: emanuele.bajo@unibo.it

Abstract

This paper investigates the tone newspapers use in

reporting information on a company that it is linked with

through an ownership tie. Our empirical setting is Italy, a

country characterized by dominant national industrial

groups’high ownership of newspapers. Based on a sample

of about 123,000 articles, we document that newspapers’

coverage of firms in conflict of interest is greater, with

significantly fewer negative and uncertain words. We also

document that the slant increases with ownership stakes

and decreases with the newspaper’s reputation.

KEYWORDS

conflict of interest, media, ownership structure

JEL CLASSIFICATION

G32, L26

1

|

INTRODUCTION

The media undisputedly influence readers’perceptions and have an impact on the social,

political, and economic environment. For instance, the economic literature has shown how the

media can influence election outcomes (DellaVigna & Kaplan, 2007) and how this relation

EUROPEAN

FINANCIAL MANAGEMENT

We thank John A. Doukas (the Editor) and two anonymous reviewers for their constructive and helpful comments on

this paper. We are also grateful to Massimiliano Barbi, Helen Bollaert, Eleonora Broccardo, Ettore Croci, Claudia Curi,

Giancarlo Dillena, Robert Faff, Håkan Jankensgård, Rüdiger Kiesel, Florencio Lopez de Silanes, Tim Loughran, Stefano

Mengoli, Maurizio Murgia, Malvina Nissim, Andrea Rocci, Sandro Sandri, Armin Schwienbacher, and Luigi Zingales

for insightful discussions; Federica Vecchioni for her precious assistance with data collection. We additionally thank the

participants of the IAFDS (Ljubljana 2014), FMA Europe (Lisbon 2017), DAFC (Lugano 2017), EFMA (Milan 2018)

conferences and of the seminars at the University of Luxembourg, Stigler Center for the Study of the Economy and the

State at the University of Chicago, Universitè de Paris XII, University of Duisburg‐Essen, Lund University, SKEMA

Business School, University of Bolzano, and University of Trento for useful comments and suggestions.

works in the opposite direction, as the media shift ideological exposure in favor of the dominant

party (Durante & Knight, 2012). The financial literature has documented similar important

effects on financial markets, such as the role of the media on market reactions to corporate

announcements (Dyck & Zingales, 2003; Tetlock, 2007) or to trading behaviors (Engelberg &

Parsons, 2011; Fang, Kempf, & Trapp, 2014). Interestingly, these effects are present not only

when the media report new and genuine information, but also when stale news is disclosed

(Fang & Peress, 2009; Huberman & Regev, 2001; Tetlock, 2011). Some studies have also shown

that the media deliberately slant news to provide economic benefits to related parties. For

instance, Reuter and Zitzewitz (2006) and Gurun and Butler (2012) show that past or current

advertisers are treated differently (and more favorably) by the newspapers from which they

have bought advertising. Firms also actively consider the possible bias of major media outlets.

According to Baloria and Heese (2018), companies diminish their disclosure of bad news when

they fear being portrayed in a negative light by the media.

Along this line, we argue that the media can treat firms that are connected to a newspaper or

its publishing firm through direct or indirect ownership ties in a more favorable and lenient

fashion. For example, on 18 May 2010, Fiat, Italy’s main automotive manufacturer, announced

to the market a dramatic drop in monthly sales (−26.6%) and market share (from 9.9% to 7.6%),

which caused its stock price to fall by 3%. This news was reported differently by two of the main

Italian newspapers: La Repubblica ran the headline ‘Automotive sales drop in Europe. Subsidies

ended: in April −7%, Fiat −27%,’whereas La Stampa, a newspaper controlled by Fiat, reported

‘Fiat’s sales surge in UK and Spain, where the effect of subsidies is still significant’and, then, in

the subheading, ‘With no government subsidy the car market slows down in April by 6.9%,

Germany is the worst, dragging along the Fiat Group.’

We argue that this example cannot be considered an isolated incident, and that newspapers

commonly shape the news they report to make information on related parties appear more

positive (or less negative) than it actually is. This news slant is economically motivated: the

increased visibility and/or laudatory articles on connected firms can attract investor and

customer attention and result in higher stock market quotes or more sales. We believe that this

type of media slant could be the outcome of two distinct but convergent processes: (a) the media

owner could be seeking favorable media treatment to create positive conditions for connected

firms; and (b) journalists could be trying to attract the owner’s sympathy to advance their

careers, or they could simply be too fearful or in awe to tarnish the reputation of the owner’s

connected firms. Although we are unable to disentangle these two rationales, we document that

newspapers voluntarily misrepresent information to investors, thus creating a severe

informational flaw in the market and a motivation for regulatory intervention.

Using the methodology developed by Loughran and McDonald (2011), we measure the tone

of approximately 123,000 articles published in a 5‐year time span by the top‐five Italian

newspapers. We believe that Italy presents a unique empirical setting in which to test conflict of

interest within the newspaper industry, since it is characterized by industrial families and

business groups controlling the major newspapers. By examining the ownership structure of the

publishing firms of the top newspapers in the 5‐year period from 2007 to 2011, we identify a

sample of 37 listed firms (13.1% of all Italian listed firms) with a conflict of interest due to direct

or indirect ownership ties with a newspaper or its publishing firm. After controlling for year,

firm, and newspaper fixed effects, we find that articles covering firms with a conflict of

interest have a significantly lower propensity to use negative and legal tones when reporting

information on connected companies.

BAJO ET AL.EUROPEAN

FINANCIAL MANAGEMENT

|

561