You are here: Foswiki>Dmi Web>WinterSchool2016>WinterSchool2016ProjectPages>WinterSchool2024FBaudiencesMexico (04 Mar 2024, RichardRogers)Edit Attach

The sad fate of small Facebook audiences: Mexico

Team Members

Meg Kitamura

Natalia Alvarez Milan

Contents

1. Introduction

Parts of this report are redacted for privacy reasons.

Mexico is among the long list of countries that will undergo an electoral period in 2024, where around 98 million Mexican voters will elect a new president, all 128 members of the Senate, and 500 Deputies (Villegas, 2023). The country, currently under the mandate of President Andrés Manuel López Obrador, has been facing a period of social division when it comes to political opinion. Before 2018, electoral disputes always involved the three traditional parties: Partido Acción Nacional (PAN), Partido Revolucionario Industrial (PRI), and Partido de la Revolución Democrática (PRD). The candidacy of López Obrador under his own party, Morena, represented for many the long overdue promise of a left-wing president who would advocate for social well-being and firmly stand against corruption and impunity, which were issues that plagued previous parties. However, others claimed that his political background made him unreliable, and that his leftist discourse was purely a facade and would not amount to anything. The victory of López Obrador was not only driven by his proposals, but also by the established rivalry between the traditional parties and Morena (Castro Cornejo, 2023), leading people to find in his “us vs. them” polarizing discourse a firm stance against the traditional political establishment in Mexico.

The 2024 elections will not only be monumental for its scale, but this election will also see an unprecedented alliance of the three traditional parties going against Morena. The upcoming elections will show whether the people will choose to give Morena another chance even with President López Obrador not having met all of his promises (Flores, 2022), or if the discontempt towards Morena will be strong enough for people to return the power to the traditional parties, known for their malpractices, by voting for the PAN-PRI-PRD alliance.

Regardless of who is in power, Mexico’s social media has been plagued with corruption and malpractice during electoral periods. The Latin American Center for Journalistic Investigation (CLIP) found that a Mexican political communication agency, Neurona Consulting, was hired to “exalt the actions of the administration of Andrés Manuel López Obrador in Mexico” as well as other Morena candidates. At the same time, the agency spread information “to vilify opposition groups of those administrations, through web pages and social networks registered by its founder César Hernández and his partners” (“México, un semillero,” 2023). Similarly, during the 2012 elections, the Mexican government made use of bots and “troll farms” to fill social media platforms with spam and fake news, amplifying certain perspectives, buying likes and retweets, and overall guiding public perception in support of the then-candidate and former President Enrique Peña Nieto, who belonged to the PRI party (Martínez, 2018). As the 2024 elections approach, there continues to be an increase in political Facebook advertisements and content that are popping up in users’ timelines. With these revelations of malpractice even within a party that staunchly stood against corruption, it has become increasingly important to scrutinize digital campaign tactics leading up to the elections.

In an effort to identify untransparent campaigns, Meta’s Ad Library has proven useful as it provides some transparency to who is paying for certain political advertisements and how much they are spending for them. While investigations tend to focus on big spenders (such as the candidates who are running for election as well as political parties and renowned news sites), small spenders, which we have defined as Facebook pages spending 100 MXN or less, are often overlooked. When looking at the Meta Ad Library’s database that shows all political ad spenders in Mexico in the last 90 days (from October 8, 2023 to January 6, 2024), we observed a total of 14,841 ad spenders who, in total, spent 202,986,593 MXN. While small spenders only represent 0.31% of the total sum (637,600 MXN), they make up 43% of the total number of spenders. These proportions become more insightful when observing that, even though small spenders are not investing large sums, they often receive as many impressions as the ads ran by big spenders (Figure 1).

The actual scale of the reach of big and small ad spenders is difficult to substantiate because Meta’s ad library makes it difficult for investigators to acquire the cumulative number of impressions per ad spender. However, as the observation of a few big and small spenders revealed that there was no significant correlation between ad spend and impression rates, we believe it is misguided to presume that ads ran by small spenders have less influence than those ran by big spenders, especially when the cumulative impact and reach may rival or even exceed that of the latter.

2. Research Questions

As there has already been suspicion of media networks engaging in covert political campaigns, we plan to utilize small ad spenders as a case study that can be scaled to uncover a larger network of untransparent political agenda setting through marketing campaigns leading up to the 2024 elections in Mexico. Through this research process, we aim to answer the following research question: Can we confirm the presence of networks of suspicious “news” outlets that use Facebook ads to operate untransparent political campaigns?

Considering previous investigations, we expect to find suspicious Facebook pages that can be linked to covert media networks and campaigns. However, the process of answering our research question will help us define a reliable and replicable methodology that hopefully allows for further research to be conducted in regards to this topic.

3. Methodology and initial datasets

The Meta Ad Library is a publicly accessible database where an individual can look for ads that run across Meta’s platforms (even without a Facebook account, although an account is required to access their API) (“Meta Ad Library API,” n.d.). For ads about social issues, elections or politics, the database allows one to search ads including inactive ads dating back to seven years prior to the search date. As our investigation was interested in the small ad spenders on Facebook, we accessed Meta’s Ad Library Report rather than filtering for keywords directly through their Ad Library. After filtering for our country of interest, Mexico, we scrolled to the very bottom of the Ad Library Report to download the full report with a date range of “Last 90 days” and an end date of “January 6” (Figure 2).

After downloading the .csv file onto our computers, we uploaded the file to Google Sheets for easy collaboration with our team. The columns in the initial dataset that was directly retrieved from Meta’s Ad Library report included Page ID, Page Name, Disclaimer, Amount Spent (in local currency), and the Number of ads in library. In total, the .csv file was able to retrieve 14,841 unique Page IDs along with the above mentioned columns.

Step 1: Accessing Data

Once we downloaded the initial dataset (refer to 2. Initial Data Sets section in this report for further instructions) from the Meta Ad Library Report, we added a column next to the Page ID column to create a direct link to the Facebook page for each page ID by adding “facebook.com/” in front of the Page ID sequence. This new link enabled easy access to the top page of each Facebook page from our dataset.

Step 2: Filtering Data

As our investigation was interested in small ad spenders on the platform, we sorted the “Amount Spent (MXN)” column in ascending order and filtered out all ad spenders using more than 100 MXN. This list of page IDs spending less than or equal to 100 MXN enabled us to filter down the list from the initial 14,841 unique page IDs to 6,446.

We then manually looked at the first 100 page IDs on the list and highlighted suspicious sounding and looking pages. As one of our investigators is from Mexico, this investigation leveraged their cultural and language specific knowledge to highlight suspicious sounding pages by looking at the names of each page (refer to Table 1 for the criteria we used to determine whether a page had a suspicious sounding name). As this step revealed that many of the suspicious pages were or were posing as media/news companies, we expanded our list from the top 100 to the top 500 page IDs and started to look specifically for news outlets. Most news outlets self-identified as media/news companies, but we determined whether the page could be classified as a media/news company page by looking at whether they included a link to their external news website of the same name, although we kept them in our list if the Facebook pages seemed to share a substantial amount of news directly on their timeline.

We also looked at their “About” section which included information that could tell us more about a page. The information that we were specifically interested in were websites and social links (under “contact and basic info”), address and phone number (under “contact and basic info”), page history (under “page transparency”), and ads from this page (under “page transparency”). We flagged the page if there were any irregularities or discrepancies in their Timeline or About section (refer to Table 1 for specific factors we looked for for each section). After filtering for suspicious media/news companies, we ended up with a list of 67 unique page IDs.

Step 3: Conduct Targeted Investigations

After we created a list of suspicious page IDs in step 2, we conducted targeted investigations for each page by clicking into every button or link we could find on each Facebook page as well as the website that the page linked out to. Out of the 67 pages on our list, we found 6 Facebook pages that seemed to be part of a larger network of news outlets. As there were already allegations of political candidates creating networks of news outlets to illegally engage in their political campaigns in Mexico, we used these 6 Facebook pages as a starting point for our investigation.

In order to connect these news outlets to each other, we decided to look at the “paid for by” column for each ad as this was one of the only pieces of information that provided us with information of who may be behind these Facebook pages. Although recent ads have organizations listed as the ad spenders, older ads often list individuals as the funders of the ads. Using this critical piece of information, we used Google’s search engine to conduct background checks on these individuals. By tracing their careers and connections, we were able to connect the individuals to government officials and political parties that may be funding these advertisements on Facebook.

4. Findings

As we have outlined the way in which we filtered out the suspicious Facebook page IDs in the Methodology section above, this section will mainly focus on the 6 pages that we have identified to be engaging in highly suspicious behavior. Out of the 6 pages, we were able to uncover three media networks (Selvamar Información y Publicidad, Central de Medios Digitales, and Medios de Comunicación Digital Mx), two of which we were able to directly trace back to government officials and parties. While we will describe the successful case studies of Debate Social and Notidiario Tuxtla, we understand the importance of outlining the limitations of our investigation with the remaining four case studies (Última Hora, Enfocando Los Hechos, Grupo 50 mx, Hablemos del Campo).

Debate Social

The first Facebook page we investigated was Debate Social (https://www.facebook.com/100087247254413). This Facebook page self-identifies as a media/news company and although it shares articles that directly link to their external website (https://debatesocialmx.com/), the link is not listed under their “About” section. While scrolling through their Facebook page timeline as well as their external website does not immediately call for suspicion, we found that the only ad they had posted had a link that directed users to a different news outlet called La Opinión Quintana Roo, which had an estimated audience size of 500K to 1M users (Figure 3).

When we accessed La Opinión Quintana Roo’s website (https://laopinionqr.com/), we found that this news outlet also had a Facebook page (https://www.facebook.com/laopinionqroo). After a thorough investigation of the Facebook page, we found that [redacted] was paying for their ads between November, 2021 and February 2022 (Figure 4). It is important to note here that although the content of La Opinión Quintana Roo’s articles and posts did not seem to skew in favor of any political party upon a quick inspection, their ads were predominantly anti-Morena.

We then looked up “[redacted]'' through the Google search engine and found that he was contracted to work for local and federal political campaigns within certain districts of Quintana Roo between September, 2021 and July, 2022, which coincides with the time range of the anti-Morena ads that were ran on La Opinión Quintana Roo’s Facebook page (Figure 4). He has also been working for a local news outlet called Noticaribe Peninsular (https://noticaribepeninsular.com.mx/) as their Senior Associate Brand Manager since November, 2015 and we were able to confirm that a handful of Noticaribe Peninsular’s ads were also being paid for by [redacted] (Figure 4).

While [redacted]’s background in political campaigns alone shows that La Opinión Quintana Roo, may have a certain political agenda while giving the facade of neutrality, we also noticed while we were looking through Noticaribe Penninsular’s ad library that other ads were being paid for by Selvamar Información y Publicidad, a publishing company owned by [redacted], who is also the director of Noticaribe Peninsular (https://www.facebook.com/NoticaribeInfo). When we searched his name alongside “Selvamar Información y Publicidad,” we found an article stating that the former Secretary General of Congress, [redacted], who is part of the PAN party, had used government funds to pay over 8 million MXN to different advertising companies, owned by his acquaintances, one of which was [redacted], owner of Selvamar Información y Publicidad (“Publicidad, mina de oro”, 2023).

While it was difficult for us to find a direct link between [redacted] and the small news outlets running hyperpartisan ads (Debate Social and La Opinión Quintana Roo), working backwards from the ad library enabled us to find digital trails of connections between those who are listed under the “paid for by'' column and government officials who may likely be funding these ads. It is unknown what the money that was paid to Sergio Alonzo was used for, but the hyperpartisan ads that skew in favor of the party that the government official involved should at least be reason to be cautious of suspicious activities on their Facebook pages.

Central de Medios Digitales

Similar to Debate Social, the case study of Notidiario Tuxtla found a network of seven news outlets that were connected by an individual, [redacted]. He was paying for political ads, which were predominantly pro-Morena, for each Facebook page that was linked to these outlets (Figure 5). Through googling his name, we found an article that accused him of working with [redacted], the Director of Social Communication for the Government of Tuxtla Gutiérrez, who was allegedly paying over 2.5 million MXN a month to finance five news outlets (“Ayuntamiento de Tuxtla Gutiérrez gasta,” 2023). We confirmed that [redacted] was behind the ads on these five news outlets mentioned in the article (Notidiario Tuxtla, Noti Zoque, Tuxtla Habla, Miralo Vos Noticias, Tuxtlequeando) as well as two other news outlets which we confirmed by looking up his name directly on Meta’s Ad Library (Noticiero Libre Tuxtla and Ciudadanos Responsables). All of these pages listed the same phone number and had a link to the Central de Medios Digitales’ website (http://centraldemedios.tv).

While we were able to link [redacted] to these seven news outlets through the Ad Library, our next step involved linking [redacted] to them. When we looked up his name alongside “Central de Medios Digitales,” we found a Facebook post accusing the news anchors for each small news outlet for being registered as government workers for the Coordination of Social Communication of Tuxtla, which is the department in which [redacted] heads. We fact checked this Facebook post by accessing the official government database that shows the salaries of all government workers (https://consultapublicamx.plataformadetransparencia.org.mx/) and we were able to confirm that the news anchors are in fact registered as Administrative Assistants in the Coordination of Social Communication of Tuxtla. This branch of government falls under the supervision of the Mayor of Tuxtla, [redacted], who was accused of paying [redacted] to promote ads in favor of Morena, which is the mayor’s political party (“Ayuntamiento de Tuxtla Gutiérrez gasta,” 2023). As the ads found under each of the seven news outlets show Morena in a favorable light, we believe [redacted] and [redacted] are most likely paying [redacted] to run favorable ads and the news anchors are being paid from government funds to aid in this process.

Limitations

While we were able to engage in a journalistic inquiry of Facebook ads in the above two case studies where we uncovered media networks and successfully connected them to government officials and parties by looking at the “paid for by” column of each ad, there are certainly limitations to this methodology. The first limitation was the fact that using individual names may not generate useful leads and may, in some cases, lead to misleading information. In the case of Última Hora and Enfocando Los Hechos, through which we uncovered a network called Medios de Comunicación Digital through their Telegram link, we found that [redacted] was behind the marketing of a few of the Facebook pages but could not use his name to make meaningful connections. Although it could also be the case that he is contracted by different news outlets to manage their Facebook ads and marketing channels, the fact that all of the external websites have similar HTML formatting and run identical or similar-looking ads that only promote the Morena party should be cause for suspicion. Another case worth mentioning is Grupo 50 mx. We found two ads that openly supported Morena candidates in Jalapa, which were paid for by [redacted], who we believed to be a government official after a quick Linkedin search. However, looking at Grupo 50 mx’s official Instagram, the only Angel Vera García we found on its following list was a different Angel Vera García who did not have connections to the government. As individuals running social media sites tend to follow the page with their personal accounts, we found it important to double check for false positives in our investigation.

5. Discussion

Both of the networks of smaller news sites that we were able to confirm as having a connection to government officials and government funds reveal a similar modus operandi, where the money that is put into the creation of these “legitimate” news sites and the ads that each of them promote comes from and returns to government instances and political parties. By placing the Facebook ads as the central element from which our investigation stemmed, we defined a scheme (Figure 6) that explains how the ads and the news sites are funded, and what they result in in terms of the impact they have on users and their potential electoral decisions.

In the scheme, we observe how there is typically a government official deviating funds to pay a middle man, who can be either another government official or the owner of a news network. This middle man is responsible for creating or managing several small Facebook pages disguised as different neutral and reliable news outlets. They often use names such as ‘Social Debate’, ‘The Opinion’, or ‘Notimedios’ (a combination of the words noticias, which means “news”, and medios, which means “media”), all of which allude to objectivity, openness, public opinion and events happening in a city or region. To further reinforce this facade, some of these Facebook pages also create their own websites where they publish articles and “conduct journalistic work” that strategically speaks in favor of specific political parties and candidates.

After the pages are created, these “news” outlets use the money they receive to promote hyperpartisan Facebook ads in an attempt to reach users who will believe that they are seeing content from legitimate news sites and will therefore follow them. This is how these networks draw attention and views to their multiple “news” pages, all of which push biased political content that favors specific politicians, candidates, or even political parties instead of reliable journalistic work. By exposing users to this hyperpartisan content, they seek to increase the approval rates of government officials currently in service, or to influence the public opinion on specific candidates to eventually earn them more votes in the upcoming elections. This results in a “return of investment” for the political parties that use funds to make this process function, making this system a closed loop of money, political content and attention fueled by corruption.

Corruption, malpractice and fraud have unfortunately been a constant in both media and politics in Mexico. Mass media have historically been closely tied to government interests, with the largest television channel in the country (Televisa) having direct agreements with the government (Ramos, 2016). Sánchez Ruiz (2005) describes how television never played a role in the process of democratization in the cultural industries of information and entertainment in Mexico, but instead reinforced a simulation of democratization while simultaneously shaking hands with important government officials. The rise of social media appeared to many as a new hope for media democratization, when in reality the same corruption has also permeated these digital platforms. Social media and the Internet in general are becoming a more common source of information among Mexicans, and as Mexico enters the electoral period, this is decisive in terms of citizen participation rates. Temkin & Flores-Ivich (2014) provided empirical evidence that showed that when citizens used the Internet to inform themselves, the probabilities of them going out to vote significatively increased. Problems arise, however, when the information that they are exposed to online is not reliable or is deeply influenced by political parties through processes that lack sufficient transparency.

While hyperpartisan news have always been particularly problematic in Mexico, them being shared by outlets pretending to be legitimate news sources funded by the government represents yet another obstacle in the transparency and legality of the country’s electoral process. These outlets engage in a process of agenda setting, through which they “construct news” by strategically including, excluding and prioritizing certain issues over others. In the dynamic of agenda setting, Gómez-Vilchis (2020) explains how the media does not openly tell citizens what to think, but how they work does affect what citizens think. In this way, the work of the media allows public opinion to obtain a certain kind of knowledge based on the news that has been “constructed” from a particular perspective. If this knowledge comes from outlets like the “news” outlets analyzed in this research, the opinion of citizens will be clearly biased in favor of the parties who are paying for this content on social media without platforms like Facebook detecting it as biased informative content.

6. Conclusions

While our investigation was able to tie suspicious Facebook ads and pages to government officials, the significance of our research lies in the methodology we have established to identify media networks that may be engaging in untransparent political campaigns. However, this investigation was limited in scope in the sense that we were only able to look at 500 page IDs of small ad spenders out of more than 14,000 in our initial data set. As election day approaches, we expect even more intense political campaigns to take place digitally, especially on Facebook where many Mexicans obtain their news and information. While we believe it is important to continue uncovering suspicious news networks using our methodology to try to fundamentally dismantle these covert media campaigns, we also understand that tackling this issue will require more manpower and resources than what investigators, researchers and journalists may have. Thus, one recommendation we have is for Meta to build better safeguards against deceitful pages such as those we have identified in this investigation. Similar to what they did in Brazil before its 2022 general elections (“How Meta Is Preparing,” 2022), Meta should create country-specific safeguards that address unique issues that are relevant in each country.

Although fake news may still be an issue for Mexico, Meta’s policy that allegedly addresses misinformation will not work in dismantling suspicious Facebook pages and news networks that promote hyperpartisan news, as the news being shared may not necessarily be misinformation. Therefore, we recommend adding another layer of safeguards such as displaying a disclaimer for non-verified news sources under all the content being shared by them, especially on sponsored ads, so that not only can users see whether the news being shared is true or not, but also check if the news is being shared by a credible news source. We believe this will also prevent those operating the suspicious media networks from easily creating Facebook pages to disseminate their hyperpartisan ads and content. In the meantime, we recommend journalists and academics to further investigate small ad spenders alongside big spenders, as not only can the total reach of small spenders exceed that of big spenders, but also because it may be easier for corruption to go undetected in small audiences.

7. References

  • Castro Cornejo, R. (2023). The AMLO Voter: Affective Polarization and the Rise of the Left in Mexico. Journal of Politics in Latin America, 15(1), 96-112. https://doi.org/10.1177/1866802X221147067

  • ● Flores, A. E. (2022, June 30). A cuatro años del triunfo en las urnas: Aciertos, errores

y desafíos de AMLO. IMER Noticias. Retrieved from https://noticias.imer.mx/blog/a-cuatro-anos-del-triunfo-en-las-urnas-aciertos-errores-y -desafios-de-amlo/ News Mundo. Retrieved from https://www.bbc.com/mundo/noticias-america-latina-44302996

I Attachment Action Size Date Who Comment
Areagraph_03_Tavola disegno 1.jpgjpg Areagraph_03_Tavola disegno 1.jpg manage 302 K 21 Oct 2019 - 13:36 EmilieDeKeulenaar  
Atlantis_WikiTimeline_Tavola disegno 1.jpgjpg Atlantis_WikiTimeline_Tavola disegno 1.jpg manage 86 K 21 Oct 2019 - 13:28 EmilieDeKeulenaar  
Crusade_WikiTimeline-02.jpgjpg Crusade_WikiTimeline-02.jpg manage 70 K 21 Oct 2019 - 13:27 EmilieDeKeulenaar  
Screenshot 2019-07-22 at 15.22.51.pngpng Screenshot 2019-07-22 at 15.22.51.png manage 429 K 21 Oct 2019 - 13:20 EmilieDeKeulenaar  
Screenshot 2019-07-22 at 16.42.17.pngpng Screenshot 2019-07-22 at 16.42.17.png manage 527 K 21 Oct 2019 - 13:37 EmilieDeKeulenaar  
Screenshot 2019-07-23 at 12.25.46.pngpng Screenshot 2019-07-23 at 12.25.46.png manage 60 K 21 Oct 2019 - 13:24 EmilieDeKeulenaar  
Screenshot 2019-07-23 at 16.10.01.pngpng Screenshot 2019-07-23 at 16.10.01.png manage 327 K 21 Oct 2019 - 13:31 EmilieDeKeulenaar  
WW2_WikiTimeline-03.jpgjpg WW2_WikiTimeline-03.jpg manage 66 K 21 Oct 2019 - 13:28 EmilieDeKeulenaar  
cluster 2.pngpng cluster 2.png manage 1 MB 21 Oct 2019 - 13:44 EmilieDeKeulenaar  
image-wall-e3b55f6d8e296e95f13bd18fc943dd55.pngpng image-wall-e3b55f6d8e296e95f13bd18fc943dd55.png manage 934 K 21 Oct 2019 - 13:33 EmilieDeKeulenaar  
pasted image 0.pngpng pasted image 0.png manage 1 MB 21 Oct 2019 - 13:23 EmilieDeKeulenaar  
pasted image 2.pngpng pasted image 2.png manage 1 MB 21 Oct 2019 - 13:32 EmilieDeKeulenaar  
unnamed-2.pngpng unnamed-2.png manage 12 K 21 Oct 2019 - 13:34 EmilieDeKeulenaar  
unnamed-3.pngpng unnamed-3.png manage 11 K 21 Oct 2019 - 13:34 EmilieDeKeulenaar  
unnamed-4.pngpng unnamed-4.png manage 54 K 21 Oct 2019 - 13:37 EmilieDeKeulenaar  
Edit  | Attach | Print version | History: r2 < r1 | Backlinks | View wiki text | Edit wiki text | More topic actions
Topic revision: r2 - 04 Mar 2024, RichardRogers
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback