CNN’s Fraudulent Analysis of Fraud in the Venezuelan Presidential Election

In the quarter century of the Bolivarian Revolution in Venezuela, initiated by the election of Hugo Chávez and continued by his successor Nicolás Maduro, 31 national elections have been held. Invariably, the US-backed opposition claimed fraud in all but the two contests that they won. Equally unvaryingly has been the corporate press’s supporting and indeed embellishing of the claims of fraud as described below.

Blackmailing the Venezuelan electorate

The moment that Venezuela’s election authority (CNE in its Spanish initials) reported that incumbent Maduro had prevailed in the July 28 presidential contest over his nearest challenger, Edmundo González, howls of fraud were to be heard throughout the world’s corporate press. Just a few hours after the CNE announcement, CNN reported that the vote “was marked by accusations of fraud and counting irregularities.”

The next day, CNN provided a backgrounder on “what you need to know” about the election. Leading off was the risible description “of a unified opposition movement that overcame their divisions.” Such a misrepresentation is indicative of a press that confounds the “Washington consensus” for the will of the Venezuelan people.

In fact, there were nine candidates on the ballot besides Maduro; all of them were opposition. They were not unified and for good reason. In its meddling into the internal affairs of Venezuela, Washington has backed the far right in the opposition spectrum.

The US-backed González ran on a platform of radically reorienting Venezuela’s foreign policy from support of the Palestinians to unconditional approval of the current campaign of genocide by Israel and the US. Domestically, he stood for an extreme neoliberal program of privatizing practically everything – schools, transportation, national oil company, housing, hospitals, food assistance. These were not stands that could possibly unify the opposition and certainly were not stands that would appeal to the vast majority of Venezuelans.

So why would the US choose a completely unknown and inexperienced candidate running on an unpopular platform when they could have given the nod to a moderate opposition politician with far more political backing and experience? This is the elephant-in-the-room that CNN and the rest of the corporate press ignore.

Left out of the media barrage against the government of Venezuela has been reporting on the context of the 930 unilateral coercive measures that Washington has levied on Venezuela with the explicit intention to asphyxiate the economy and cause the people to abandon the Bolivarian Revolution. The US banked on blackmailing Venezuela to vote out Maduro or continue to suffer what are euphemistically called “sanctions.”

Battle of the bean counters

Now that the election is over, followed by the predictable accusations of fraud, the corporate media has intensified its campaign to delegitimize the results. In particular, we analyze a CNN article that reports “after Venezuela’s contested presidential vote, experts say government results are a ‘statistical improbability.’”

CNN casts doubt on the results of the Venezuelan presidential election based in part on a post published by the Quantitative Methods in Social Science program at Columbia University. The post cites the official results published by the CNE showing the votes for each candidate followed by their percentage of the total rounded to a single decimal place.

Maduro            5,150,092        51.2%
González          4,445,978        44.2%
Others                 462,704           4.6%

Total              10,058,774       100.0%

Based on the CNE data, the post goes on to calculate the percentages vote for each candidate to seven decimal places:

Maduro            5,150,092        51.1999971%
González          4,445,978       44.1999989%
Others                  462,704         4.6000039%

Total               10,058,774      100.0000000%

Then the post notes that using the original percentages (51.2, 44.2, and 4.5) to compute the votes for candidates results in fractional vote counts that do not match the original vote totals published by the CNE:

Maduro            5,150,092.288             51.2000000%
González          4,445,978.108             44.2000000%
Others              462,704.604                  4.6000000%

Total                10,058,774                 100.0000000%

Of course, the counts don’t match but that is an artifact of the misuse of statistics by the post. For example, let’s apply the same method to the official results of the 2020 US presidential election, published by the US Federal Election Commission (FEC):

Biden    81,268,924      51.31%
Trump  74,216,154      46.86%
Total  158,383,403    100.00%

If you multiply the percentage given for Biden 51.31% times the total number of votes, (158,383,403), the result is 81,266,524, which is different from the total reported by the FEC above (81,268,924).  Perhaps CNN has uncovered and should be reporting on statistically significant evidence of fraud that was committed by the FEC in the 2020 presidential election?

Nevertheless, the post concludes, referring to the discrepancy that they created: “If it is not evidence of fraud by itself I do not know what is.” Then the posts wildly speculates: “Anyhow, the image of the Chavista bosses fabricating the results with a napkin and their phone calculators seems to be as plausible as amusing.” The post continues: “That seems fishy… Can it be, that instead of calculating the percentages from the number of votes, someone decided the percentages and then calculated the number of votes?”

So there you have it. The “Chavista bosses,” starting from their desired fabricated percentage results, computed candidate totals (using phone calculators on a napkin!), only to be exposed because their clumsily computed fabricated totals do not match their fabricated percentages!

But what CNN is really doing is statistical arm waving. The supposed discrepancy that CNN exposes is nothing more than their misuse of statistics.

The CNN article then goes on to provide additional dramatic and equally sketchy evidence for fraud, citing Andrew Gelman, a professor of statistics and political science at Columbia University who “ran a mathematical simulation with a probability model.” Gelman apparently concluded that “there is about a 1 in 100 million chance that this particular pattern will occur by chance.”

Are they correct? The short answer is that CNN is trying to baffle us.

There is a simple response that provides an easy (and highly probable) explanation of the results published by the Venezuelan CNE: the percentages were computed by dividing the reported vote totals for each candidate by the total number of votes cast, and then the result was rounded to one decimal place.

This is the normal and standard practice. Voting percentages in elections aren’t usually published and displayed to seven decimal places.

If the starting point is the observed vote totals, the percentages can be computed and then rounded to one decimal place as shown in the CNE report. This is nothing more than a simple case of rounding numbers for the purpose of displaying them in a readable form.

Professor Gelman’s “dramatic” simulation only shows that rounding can have a significant impact on decimal calculations; not any evidence of fraud.

In technical jargon, rounding X/N and then multiplying the result times N almost never equals N, unless N is a multiple of X.  If you randomly generate a million numbers and divide them each by their sum, it is extremely unlikely that any of the numbers is a factor of the sum. It says nothing about whether the proportion X/N is fabricated or not.

In lay terms, Gelman is misleading readers by misapplying statistical modeling.

Misuse of statistics

The starting point for the post cited by CNN is the assumption that there are “strong allegations of fraud in the recent Venezuelan elections.”  The goal of exposing an imagined scenario of “Chavista bosses fabricating the results” as well as other political comments present in the post (comparing, for example, the supposed fraud in Venezuela to the US election deniers and January 6th), indicates that this “expert analysis” providing “strong evidence of fraud” had political motivations.  There may be other aspects of the election pointing to fraud, but none of the above does so.

By falsely generating suspicions of fraud in the election, CNN heightens the risk of violence, extra-constitutional resolution of the crisis, and support for outside interference in the electoral process.  If you are looking for trustworthy, unbiased reporting and analysis of the Venezuelan election, stay away from CNN

Allan Miller has a PhD in economics with a research focus on quantitative modeling. He teaches statistics in the Data Science program at the University of California Extension program in Berkeley, California. Roger D. Harris is with the Venezuela Solidarity Network and the US Peace Council. Read other articles by Allan Miller and Roger D. Harris.