Fake Polls as Fake News:

Similar documents
From Straw Polls to Scientific Sampling: The Evolution of Opinion Polling

Info Pack Mexico s Elections

WHAT IS PUBLIC OPINION? PUBLIC OPINION IS THOSE ATTITUDES HELD BY A SIGNIFICANT NUMBER OF PEOPLE ON MATTERS OF GOVERNMENT AND POLITICS

AMLO, the PRI, and the Frente: A Look at Mexico s 2018 Election

What is Public Opinion?

Mexico s 2018 Congressional elections

American political campaigns

National Survey Report. May, 2018

Public Opinion and Political Socialization. Chapter 7

Data manipulation in the Mexican Election? by Jorge A. López, Ph.D.

EDW Chapter 9 Campaigns and Voting Behavior: Nominations, Caucuses

Mexico s Evolving Democracy. A Comparative Study of the 2012 Elections. Edited by Jorge I. Domínguez. Kenneth F. Greene.

1. Amendments impacting Voting. 15th - No Racial Discrimination. 17th - Direct election of senators by citizens, not state legislature appointment

2006 CAMPAIGN POLITICAL AND ELECTORAL CONTEXT

Myths and facts of the Venezuelan election system

NATIONAL: FAKE NEWS THREAT TO MEDIA; EDITORIAL DECISIONS, OUTSIDE ACTORS AT FAULT

PEW RESEARCH CENTER FOR THE PEOPLE & THE PRESS CAMPAIGN CONSULTANTS SURVEY FINAL TOPLINE November 1997 March 1998 N=200

Election 2015: Liberals edge Conservatives as volatile electorate mulls final choice before last campaign weekend

Chapter 9 Content Statement

Politicians as Media Producers

Nonvoters in America 2012

Chapter 8: Mass Media and Public Opinion Section 1 Objectives Key Terms public affairs: public opinion: mass media: peer group: opinion leader:

Online Appendix for Partisan Losers Effects: Perceptions of Electoral Integrity in Mexico

Video Notes Unit 2 Political Beliefs & Behaviors

Clinton Maintains 3% Lead in Michigan (Clinton 47% - Trump 44% - Johnson 4% - Stein 1%)

Chapter 9 Campaigns and Voting Behavior (Elections) AP Government

The 2014 Election in Aiken County: The Sales Tax Proposal for Public Schools

SCIENCE IN THE DIGITAL AGE: MAPPING OPPORTUNITIES, PERILS AND UNCERTAINTIES

Chapter 9: The Political Process

Political Economy of Structural Reform: reforms among resurgent populism

Clinton Lead Cut in Half from August (Clinton 47% - Trump 42% in 2-way and Clinton 45% - Trump 39% in 4-way)

CHAPTER 9: THE POLITICAL PROCESS. Section 1: Public Opinion Section 2: Interest Groups Section 3: Political Parties Section 4: The Electoral Process

New Americans in. By Walter A. Ewing, Ph.D. and Guillermo Cantor, Ph.D.

AP US GOVERNMENT & POLITICS UNIT 2 REVIEW

Swing Voters in Swing States Troubled By Iraq, Economy; Unimpressed With Bush and Kerry, Annenberg Data Show

Vote Preference in Jefferson Parish Sheriff Election by Gender

Minnesota Public Radio News and Humphrey Institute Poll. Coleman Lead Neutralized by Financial Crisis and Polarizing Presidential Politics

2018 MEXICAN ELECTION SPECIAL

Repeat Voting: Two-Vote May Lead More People To Vote

Why 100% of the Polls Were Wrong

NATIONAL: POTUS LESS TRUSTED THAN MEDIA, FAKE NEWS COMES FROM ALL SOURCES

State of the Facts 2018

Clinton Leads by 13% in Michigan before Last Debate (Clinton 51% - Trump 38%- Johnson 6% - Stein 2%)

Explaining the Spread of Misinformation on Social Media: Evidence from the 2016 U.S. Presidential Election.

VIOLENCE AND TERRORISM IN LATIN AMERICA A GLOBAL CONTEXT: AN OVERVIEW

STATEMENT OF THE NDI PRE-ELECTION DELEGATION TO MEXICO S JULY 2, 2006 PRESIDENTIAL ELECTIONS. Mexico City, April 9, 2006

Election 2015: Race narrows to Conservative-Liberal contest with NDP in third place; soft voters still cause for volatility

- Bill Bishop, The Big Sort: Why the Clustering of Like-Minded America is Tearing Us Apart, 2008.

The New Battlegrounds over Facts in Democracy

Another Billion-Dollar Blunder?

Ohio State University


The November WHO ELECTED JIM DOYLE? AND PRESERVED CONSERVATIVE CONSTITUTIONAL IDEAS JAMES H. MILLER

CHAPTER 11 PUBLIC OPINION AND POLITICAL SOCIALIZATION. Narrative Lecture Outline

COMPARATIVE ANALYSIS ON POLITICAL PARTY AND CAMPAIGN FINANCING. APPENDIX No. 1. Matrix for collection of information on normative frameworks

THE AFGHAN ELECTIONS: IS ABDULLAH RIGHT THAT HE WAS WRONGED (TWICE)? By Andrew Garfield

Change versus more of the same: On-going panel of target voting groups provides path for Democrats in 2018

POLI 201 / Chapter 10 Fall 2007

Panel: Norms, standards and good practices aimed at securing elections

I. Chapter Overview. Roots of Public Opinion Research. A. Learning Objectives

Why the United Socialist Party of Venezuela (PSUV) Won the Election. James Petras

1. A Republican edge in terms of self-described interest in the election. 2. Lower levels of self-described interest among younger and Latino

Voter turnout in today's California presidential primary election will likely set a record for the lowest ever recorded in the modern era.

Politicians who needs them? 1 of 5 10/23/2014 8:30 AM. October , 5.34am EDT. Glenn Altschuler

Addressing Danger originated by the Increasing Spread of Populism

Partisan Advantage and Competitiveness in Illinois Redistricting

Lab 3: Logistic regression models

5 Key Facts. About Online Discussion of Immigration in the New Trump Era

Map of Mexico. Civil Society in a Globalizing World: The Case of Mexico. Regime Stability. No Meaningful Opposition.

Texas JSA LoneStar. Spring State Approaching. 2 Donald Drumpf. 3 Super Tuesday Results. 3 Police Brutality

ALABAMA: TURNOUT BIG QUESTION IN SENATE RACE

CIRCLE The Center for Information & Research on Civic Learning & Engagement. Youth Voting in the 2004 Battleground States

- 1 - Second Exam American Government PSCI Fall, 2001

ELECTORAL BEHAVIOR POLS Spring, Course Outline

1 Year into the Trump Administration: Tools for the Resistance. 11:45-1:00 & 2:40-4:00, Room 320 Nathan Phillips, Nathaniel Stinnett

MATERIAL ON THE TEST Edwards Chapters 6, 9, 8, 10, 11 Sides ( Science of Trump ) chapters 4, 5, 6, 15, 24, 12 CHAPTER 6

AIM: Does the election process guarantee that the most qualified person wins the presidency?

For First Time, Majority of Mexicans Hold Unfavorable View of United States

Edging toward an earthquake Report on the WVWV March National Survey

COLORADO: CLINTON MAINTAINS DOUBLE DIGIT LEAD

MEREDITH COLLEGE POLL February 19-28, 2017

Clinton Lead Cut to 8% in Michigan (Clinton 49% - Trump 41%- Johnson 3% - Stein 1%)

Light often arises from a collision of opinions, as fire from flint and steel Benjamin Franklin, 1760

The 2004 Election Aiken County Exit Poll: A Descriptive Analysis

Media and Elections in Asia: The Changing Role in Coverage and Control

Polling and the Ballot: The Venezuelan Referendum By David Rosnick 1

Public Opinion and Political Participation

PARTISANSHIP AND WINNER-TAKE-ALL ELECTIONS

Latinos and the 2008 Presidential Elections: a Visual Data Base

November 2018 Hidden Tribes: Midterms Report

THE PUBLIC AND THE CRITICAL ISSUES BEFORE CONGRESS IN THE SUMMER AND FALL OF 2017

The Media and Public Opinion

California Politics: A Primer, 4 th Edition. Chapter 10

Searching for Meaning in Presidential Elections

What the 2016 Election Means to My Millennial Generation Destiny Goede

Patterns of Poll Movement *

How The Mountain West States Voted In 2016: A Post-Election Analysis of Trends, Demographics, and Politics in America s New Swing Region

Illinois Redistricting Collaborative Talking Points Feb. Update

Political Beliefs and Behaviors

EMBARGOED NOT FOR RELEASE UNTIL: SUNDAY, OCTOBER 17, 1993 FLORIO MAINTAINS LEAD OVER WHITMAN; UNFAVORABLE IMPRESSIONS OF BOTH CANDIDATES INCREASE

Transcription:

Fake Polls as Fake News: The Challenge for Mexico s Elections By Jorge Buendía Global Fellow, Mexico Institute April 2018

Fake Polls as Fake News: The Challenge for Mexico s Elections By Jorge Buendía Global Fellow, Mexico Institute Voter manipulation through misinformation is a long and established practice in many polities regardless of whether they are democracies or authoritarian regimes. Parties and candidates alike will frame and even distort issues to their advantage. Misinformed voters can be a potential threat to democracy if they get to know the issues through fake news: poor decision-making will follow. For many analysts, this is exactly what led to Brexit and to Trump s victory in the United States. It can be argued that the spread of fake news during a campaign is as old as elections themselves, but the speed at which they are produced and disseminated nowadays is certainly new. Through social media, fake news can reach a global audience in a matter of minutes. As a pollster in Mexico, I have long observed all kinds of fake news being disseminated during electoral campaigns. A peculiarity of fake news in Mexico, however, is that polls themselves have become the fake news. For years, many media outlets and political columns have reported results of public opinion polls that are of dubious quality. Likewise, established polling firms have often had to denounce that their reports have been doctored or that they did not conduct the poll being reported as theirs in the media. In recent elections, on election day, fake exit poll results have circulated through social media. While the dissemination of fake news through social media is easy, it is striking how receptive the Mexican mainstream media is to fake news. In fact, many fake news and fake polls are first reported in established media outlets and then widely circulated on social media. What is particularly troublesome is not only that fake polls have been attributed to established firms but also that many fake polls are attributed to non-existent firms. Polls are fake but so are the firms behind them. Moreover, the stories about these fake polls have run in major media outlets, so it is not a marginal issue. Further, the practice of publishing fake polls predates the appearance of Facebook or Twitter in our lives. 1

When Did It all Start? It is difficult to identify a single moment when fake polls began to appear in the Mexican media. It is easier to say that polls began to matter in Mexico once the political system opened up and contested elections became a norm rather than the exception. This happened in the decade of the 1990s, especially after the severe economic crisis of 1994-1995. The economic crisis led to the defeat of the incumbent party, the Institutional Revolutionary Party (PRI by its acronym in Spanish), in a number of gubernatorial elections. In the 1997 mid-term legislative elections, the PRI lost control of the Chamber of Deputies for the first time. The sea change in the relevance of public opinion data in this period is best reflected in the amount of published polls during the presidential campaigns of 1994 and 2000. Throughout the entire 1994 presidential campaign, only 17 national polls were circulated. In contrast, 48 national polls were published from January to June 2000. In 2000, when the presidential election that led to the end of 71 years of uninterrupted PRI rule in Mexico took place, the publication in established media outlets of fake polls was a widespread practice. During the 2000 presidential campaign, several newspapers (El Sol de México, Excélsior, Unomásuno) reproduced the results of shadowy firms such as Ceprocepp, Technomanagement, and Fishers with widely favorable numbers to Francisco Labastida, the PRI candidate. These firms that no one knew existed (and that no one has even heard of since) circulated 10 polls throughout the campaign, which represent 21 percent of the total number of public polls. Citizens being exposed to questionable data and disinformation was a strategy to influence and sway public opinion. The strategy, of course, did not work as the PRI lost the election by six percentage points. In the 2000 presidential election, political parties also began to broadcast ads with dubious but favorable public opinion data for their causes. I remember watching the second presidential debate on the night of May 26, 2000. Less than an hour after the debate ended, a paid spot about Vicente Fox being the winner of the debate appeared on the main national TV network. Given the technological restrictions of those years and the short period of time between the end of the debate and the diffusion of the spot, it was impossible to conduct, 2

in less than an hour, a serious poll about who the winner of the debate was. Naturally, the purpose of the spot was not to convey accurate information but to shift public opinion in favor of Vicente Fox. From then on, polls have been widely used in Mexico as an instrument to circulate misleading claims about popular support for governments and candidates. Push polls (polls aimed at swaying voters using loaded or manipulative questions) frequently receive wide diffusion in established media outlets. Why Do Politicians Promote Fake Polls? Notwithstanding the lack of evidence regarding the impact of published polls on public opinion or voting behavior, Mexican politicians keep pushing for the publication of favorable results to their causes. In some cases, politicians may believe that there is a bandwagon effect that will benefit them. In some other cases, politicians want to appeal to strategic voters, that is, voters who do not want to waste their vote supporting a candidate or party with slim chances of winning. By revealing that a party has a low probability of winning, polls whether fake or not may influence the decision of strategic voters. The rationality behind a bandwagon effect or strategic voting rests however in informing the voters about the actual state of electoral preferences. Given the low levels of information voters have, it is not clear that they will find out what the polls are actually saying and even less clear how much weight will be given to that information. Furthermore, poll results sometimes conflict, so polls effect on voters may cancel out. Indeed, one can argue that the main purpose of any fake poll is to misinform the public about the actual state of electoral preferences. Fake News and Fake Polls in the 2018 Election The use of fake polls has become the rule rather than the exception for electoral processes in Mexico. Just days before the 2017 Estado de México gubernatorial election, a fake survey, falsely claiming to be from Consulta Mitofsky, a Mexican polling firm, spread throughout social media, and different political actors used these numbers to support the MORENA 3

candidate, despite Mitofsky denying its authenticity. Furthermore, the day of the election, fake exit polls circulated on social media. However, the diffusion of fake polls is not restricted to social media as sometimes the whole process starts in the mainstream media. For instance, a series of questionable polls regarding the 2018 presidential election were published for a long time in different media. The polls were attributed to the Office of the President or to CISEN, Mexico s equivalent of the National Security Agency. The reports of these polls had blatant errors, such as a huge margin of error for a national poll of 5,000 interviews, but no one in the media questioned their authenticity. These fake polls circulated widely for many months, and it was not until recently that the Press Secretary to the Mexican President issued a statement rejecting the claims that these were governmental polls. The purpose of these polls was to present some parties and politicians as appearing more popular than they actually were. As long as there are more unpopular politicians than popular ones, the demand for fake polls will continue. If the mainstream media keeps echoing these fake polls, social media will just pick it up and will disseminate them broadly. Unlike other countries, in Mexico fake news and fake polls are a problem because the mainstream media does not stop false information from reaching the public. Hence, it is likely that more fake news and polls will start to appear in the current presidential campaign. How Receptive are Mexicans to Fake News? If the supply side of fake news is alive and kicking, what about the demand side? We know citizens filter out information that clash with their ideology, partisanship, interests, prejudices, or preconceptions. While social media allows a degree of voter segmentation unforeseen a decade ago, voters are not just passive recipients of information. Further, access to the internet or social media is not universal in Mexico. Only around half of the eligible voting population has internet access in Mexico. To succeed, fake news requires a dual strategy, a strategy that works in both the mainstream media and in social media. 4

Given this context, a relevant question to ask is whether Mexicans are prone to believe in fake news and if social media consumers are more likely to believe in these types of stories. To answer this question, I included a set of questions in a national face-to-face survey of 1,003 Mexican adults aged 18 and over (November 2017). The battery of questions aimed to measure if respondents were inclined to believe in a series of fake statements and conspiracy theories. In the survey, respondents were asked if several false statements were true or false. Roughly half of the Mexican adults (53%) think that the PRI and the PAN have a secret plan to stop Andrés Manuel López Obrador from becoming president. A polarizing figure, Lopez Obrador is currently the favorite candidate to win the 2018 presidential election. Polling data and past electoral campaigns suggest Mexicans are likely to believe questionable stories surrounding his candidacy. For instance, the results of the survey show that one in five (22%) respondents think that advisors to Hugo Chávez and Nicolás Maduro work with López Obrador. Even though the left-wing candidate has stated that he does not have any relationship with the Venezuelan government, an important number of citizens think the opposite. Hence, fake stories about López Obrador s links with the Venezuelan government will fall on fertile ground in the upcoming presidential campaign. 5

Figure 1. Share of people that believe in a series of fake statements Please tell me if you think the following statements are true or false. TRUE FALSE 0 20 40 60 80 100 The PRI and the PAN have a secret plan to stop Andrés Manuel López Obrador from becoming president 53 33 The government knew the earthquakes of last September were coming but they did not want to tell the citizens 29 62 The leaders of our country are members of secret societies like the Masons and the Illuminati 27 50 The advisors of Hugo Chavez and Nicolás Maduro work with Andrés Manuel López Obrador 22 42 In September 2017, Mexico was shaken by two major earthquakes. Coincidentally, the second quake occurred on the 32 nd anniversary of one of the most devastating tremors of the country. Even though scientists cannot predict earthquakes, the survey shows that 29 percent of Mexican adults think that the national government knew in advance that an earthquake was going to hit the country, but they decided not to tell people. Finally, even though there is no evidence to support this idea, roughly one in four (27%) Mexicans concur with the idea that Mexico s leaders are members of secret societies such as the Masons and the Illuminati. Poll respondents were also asked if they believe in a series of conspiracy theories that have circulated not only in Mexico but also in other countries. The purpose of these questions 6

was to quantify if respondents were prone to believe narratives that are hard to prove. A significant number of people think these stories are true. Nearly half of the respondents (43%) expressed that vaccines do not prevent disease but that pharmaceutical companies sell them just to make money. A similar share of people (41%) believe that governments hide information on extraterrestrial life. Additionally, 23 percent of the respondents consider that climate change is a myth created by international organizations and some countries. Furthermore, one in five (22%) think that the United States created the AIDS virus. Overall, 70 percent of Mexican adults believe in at least one of these narratives. Needless to say, the proclivity to believe in conspiracy theories is not exclusive to Mexicans. For instance, research suggests that over 55 percent of Americans believe in some conspiracy theory. People, in general, seem to be inclined to accept these kinds of stories. The data thus reveal that citizens are receptive to fake stories and hard-to-prove theories. In a world where it has become relatively easy to report and spread false news, someone who wants to shift public opinion could succeed given that people tend to believe in questionable information. Given that social media has become the favorite place to disseminate fake news, it is important to know if social media consumers are more likely to believe in false narratives. In Mexico, no evidence was found to support this idea. Social media users are as likely to believe in fake news and/or conspiracy theories as non-social media users. The most likely reason is that access to the internet or social media in Mexico is skewed towards the younger, more educated, and more affluent social groups, and they tend to be more critical of the information they receive. 7

Figure 2. Share of people that believe in a series of conspiracy theories Please tell me if you think the following statements are true or false. TRUE FALSE 0 20 40 60 80 100 Vaccines do not prevent disease, but pharmaceutical companies sell them to make money 43 51 The governments hide information about UFOs 41 46 Climate change is a myth created by international organizations and the governments of some countries 23 68 The U.S. government created the AIDS virus 22 60 What the data show is that there is receptivity to fake or controversial stories. However, there is a high degree of variance depending on the topic, although it is safe to say that stories related to the political realm will be more credible given the discredit of political institutions. This is particularly relevant because fake news can be used not only to promote positive/negative information about a candidate but also to discourage turnout, especially among younger voters who have more access to social media. The current presidential election in Mexico takes place in a context where parties and candidates have incentives to spread fake news, and voters show a proclivity to believe in them. While the main battleground will be in the digital realm, the role of the mainstream media cannot be understated: if they do not filter out the fake news and fake polls, all Mexicans will be exposed to false stories increasing the chances for poor decision-making. 8

About the Author Jorge Buendía is the Director of Buendía & Laredo and a Global Fellow at the Wilson Center s Mexico Institute. Mr. Buendía conducts research on electoral marketing, electoral studies, electoral behavior, and public opinion. His research findings on these topics have been published in various academic journals in the United States, Latin America, and Mexico. Mr. Buendía received his PhD and master s degree in Political Science from the University of Chicago. In addition, he has a master's degree in Comparative Government from the London School of Economics and Political Science. He is a member of the National System of Researchers and a professor-researcher of the Autonomous Technological Institute of Mexico (ITAM) and the Center for Economic Research and Teaching (CIDE). 9