Bolivia dismissed its October elections as fraudulent. Our research found no reason to suspect fraud.

Bolivians will hold a new election in May — without ousted president Evo Morales

From The Washington Post, 27.02.2020, by John Curiel and Jack R. Williams

As Bolivia gears up for a do-over election on May 3, the country remains in unrest following the Nov. 10 military-backed coup against incumbent President Evo Morales.

A quick recap: Morales claimed victory in October’s election, but the opposition protested about what it called electoral fraud. A Nov. 10 report from the Organization of American States (OAS) noted election irregularities, which “leads the technical audit team to question the integrity of the results of the election on October 20.” Police then joined the protests and Morales sought asylum in Mexico.

The military-installed government charged Morales with sedition and terrorism. A European Union monitoring report noted that some 40 former electoral officials have been arrested and face criminal charges of sedition and subversion, and 35 people have died in the post-electoral conflict. The highest-polling presidential candidate, a member of Morales’s Movimiento al Socialismo (MAS-IPSP) party, has received a summons from prosecutors for undisclosed crimes, a move some analysts suspect was aimed to keep him off the ballot.

The media has largely reported the allegations of fraud as fact. And many commentators have justified the coup as a response to electoral fraud by MAS-IPSP. However, as specialists in election integrity, we find that the statistical evidence does not support the claim of fraud in Bolivia’s October election.

The OAS claimed that election fraud had happened

The primary support for claims of fraud was the OAS report. The organization’s auditors claimed to have found evidence of fraud following a halt in the preliminary count — the nonbinding election-night results meant to track progress before the official count.

The Bolivian constitution requires that a candidate either earn an outright electoral majority or 40 percent of the votes, with at least a 10-percentage-point lead. Otherwise, a runoff election will take place. The preliminary count halted with 84 percent of the vote counted, when Morales had a 7.87 percentage-point lead. Though the halt was consistent with election officials’ earlier promise to count at least 80 percent of the preliminary vote on election night and continue through the official count, the OAS quickly expressed concern over the stop. When the preliminary count resumed, Morales’s margin was above the 10-percentage-point threshold.

The OAS claimed that halting the preliminary count resulted in a “highly unlikely” trend in the margin in favor of MAS-IPSP when the count resumed. The OAS reported “deep concern and surprise at the drastic and hard-to-explain change in the trend of the preliminary results.” Adopting a novel approach to fraud analysis, the OAS claimed that high deviations in data reported before and after the cutoff would indicate potential evidence of fraud.

But the statistical analysis behind this claim is problematic

The OAS report is in part based on forensic evidence that OAS analysts say points to irregularities, which includes allegations of forged signatures and alteration of tally sheets, a deficient chain of custody, and a halt in the preliminary vote count. Crucially, the OAS claimed in reference to the halt in the preliminary vote count that “an irregularity on that scale is a determining factor in the outcome​” in favor of Morales, which acted as the primary quantitative evidence to their allegations of “clear manipulation of the TREP system … which affected the results of both that system and the final count.”

We do not evaluate whether these irregularities point to deliberate interference — or reflect the problems of an underfunded system with poorly trained election officials. Instead, we comment on the statistical evidence.

Since Morales had surpassed the 40-percent threshold, the key question was whether his vote tally was 10 percentage points higher than that of his closest competitor. If not, then Morales would be forced into a runoff election against his closest competitor — former president Carlos Mesa.

Our results were straightforward. There does not seem to be a statistically significant difference in the margin before and after the halt of the preliminary vote. Instead, it is highly likely that Morales surpassed the 10-percentage-point margin in the first round.

How did we get there? The OAS approach relies on dual assumptions: that the unofficial count accurately reflects the vote continuously measured, and that reported voter preferences do not vary by the time of day. If these assumptions are true, then a change in the trend to favor one party over time could potentially indicate fraud had occurred.

The OAS cites no previous research demonstrating that these assumptions hold. There are reasons to believe that voter preferences and reporting can vary over time: with people who work voting later in the day, for instance. Areas where impoverished voters are clustered may have longer lines and less ability to count and report vote totals quickly. These factors may well apply in Bolivia, where there are severe gaps in infrastructure and income between urban and rural areas.

Was there a discontinuity between the votes counted before and after the unofficial count? For sure, discontinuities might be evidence of tampering. In Russia, for instance, one allegation is that local election officials stuff ballot boxes to meet preset targets.

If the OAS finding was correct, we would expect to see Morales’s vote margin spike shortly after the preliminary vote count halted — and the resulting election margin over his closest competitor would be too large to be explained by his performance before preliminary count stopped. We might expect to see other anomalies, such as sudden shifts in votes for Morales from precincts that were previously less inclined to vote for him.

The x-axis shows the margin for Morales before the cutoff within the preliminary count of 1,477 precincts that reported data before and after the cutoff, and the y-axis is their final margin, as reported within the official count. The high correlation between preliminary count and final vote results suggests no significant irregularities in the election count, or Morales’s final vote margin. (Jack Williams. Data from Tribunal Supremo Electoral, 2019.)

We didn’t find any evidence of any of these anomalies, as this figure shows. We find a 0.946 correlation between Morales’s margin between results before and after the cutoff in precincts counted before and after the cutoff. There is little observable difference between precincts in the results before and after the count halt, suggesting that there weren’t any significant irregularities. We and other scholars within the field reached out to the OAS for comment; the OAS did not respond.

We also ran 1,000 simulations to see if the difference between Morales’s vote and the tally for the second-place candidate could be predicted, using only the votes verified before the preliminary count halted. In our simulations, we found that Morales could expect at least a 10.49 point lead over his closest competitor, above the necessary 10-percentage-point threshold necessary to win outright. Again, this suggests that any increase in Morales’s margin after the stop can be explained entirely by the votes already counted.

There isn’t statistical support for the claims of vote fraud

There is not any statistical evidence of fraud that we can find — the trends in the preliminary count, the lack of any big jump in support for Morales after the halt, and the size of Morales’s margin all appear legitimate. All in all, the OAS’s statistical analysis and conclusions would appear deeply flawed.

Previous research published here in the Monkey Cage finds that economic and racial differences make it difficult to verify voter registration in the United States, resulting in higher use of provisional ballots among Democrats — and greater support for Democratic candidates among votes counted after Election Day. Under the OAS criteria for fraud, it’s possible that U.S. elections in which votes that are counted later tend to lean Democratic might also be classified as fraudulent. Of course, electoral fraud is a serious problem, but relying on unverified tests as proof of fraud is a serious threat to any democracy.

John Curiel is a research scientist with MIT’s Election Data and Science Lab. He earned his PhD in political science from the University of North Carolina at Chapel Hill.

Jack R. Williams is a researcher with MIT’s Election Data and Science Lab.

Bolivia: il CEPR di Washington afferma che non ci sono prove di irregolarità o brogli nelle elezioni

Da L’Antidiplomatico, 11.11.2019

La narrazione dei golpisti in Bolivia è chiara: Evo Morales è stato rieletto grazie a brogli. Quindi deve rinunciare, e non partecipare alle nuove elezioni. 

Ma questi brogli, avallati anche dall’Organizzazione degli Stati Americani – che non smentisce la sua nomea di Ministero delle Colonie degli USA – che parla di presunte ‘irregolarità’ in un comunicato stranamente diffuso prima del rapporto ufficiale, sono realmente accaduti? Vi sono prove?

Una risposta ha provato a darla il Center for Economic and Policy Research (CEPR) di Washington. La risposta è no. Non ci sono stati brogli, Evo Morales ha nettamente vinto al primo turno.

«Semplicemente non esiste una base statistica o probatoria per contestare i risultati del conteggio dei voti che mostrano che Evo Morales ha vinto al primo turno», ha affermato l’analista di politica senior del CEPR e coautore del documento redatto dal think thank statunitense, Guillaume Long. «Alla fine, il conteggio ufficiale, che è legalmente vincolante e completamente trasparente, con i fogli di conteggio disponibili online, ha eguagliato i risultati del conteggio rapido». 

L’analisi statistica dei risultati elettorali e dei fogli di conteggio del 20 ottobre in Bolivia – scrive il CEPR – non mostrano alcuna prova che irregolarità o brogli abbiano influenzato il risultato ufficiale che ha dato al presidente Evo Morales una vittoria al primo turno. Contrariamente a una narrazione post-elettiva che è stata supportata, senza prove, dalla Missione di Osservazione Elettorale dell’OSA, l’analisi statistica mostra che era prevedibile che Morales avrebbe ottenuto la vittoria al primo turno, sulla base dei risultati del primo 83,85 percento dei voti scrutinati con il conteggio rapido.

Il documento «Che cosa è successo nel conteggio dei voti in Bolivia nel 2019? Il ruolo della Missione di Osservazione Elettorale dell’OSA», presenta una ripartizione dettagliata di ciò che è accaduto con il conteggio dei voti della Bolivia (sia il conteggio rapido non ufficiale, sia il conteggio ufficiale più lento), cercando di dissipare la confusione sul processo. Il rapporto include i risultati di 500 simulazioni che mostrano che la vittoria al primo turno di Morales fosse non solo possibile, ma probabile, in base ai risultati dell’83,85 percento iniziale dei voti scrutinati con il conteggio rapido. 

Insolito è stato il comunicato dell’OSA che ha dato nuovo vigore alle violenze golpiste e spinto il presidente Evo Morales a rinunciare nel mezzo di crescenti minacce contro vari esponenti del MAS. 

Mark Weisbrot, condirettore del CEPR, ha osservato come fosse altamente discutibile il fatto che l’OSA emettesse una dichiarazione stampa in cui si mettevano in discussione i risultati delle elezioni senza fornire prove a riguardo. Ha osservato che anche la relazione preliminare dell’OSA sulle elezioni non ha fornito prove del fatto che ci fosse qualcosa di sbagliato nel conteggio dei voti.

«La dichiarazione stampa dell’OSA del 21 ottobre e la sua relazione preliminare sulle elezioni boliviane sollevano inquietanti domande sull’impegno dell’organizzazione a un’osservazione imparziale, professionale ed elettorale», ha affermato Weisbrot. «L’OSA dovrebbe indagare per scoprire come tali dichiarazioni, che potrebbero aver contribuito al conflitto politico in Bolivia, siano state fatte senza alcuna prova».

Il documento verifica che le tendenze di voto storiche a favore di Morales nelle aree di votazione più lontane e impervie spiegano il motivo per cui il divario tra Morales e Mesa si è ampliato con il conteggio dei voti, terminando con un risultato ufficiale che ha portato Morales davanti a Mesa di 10,57 punti.

l documento mostra anche che le rispettive tendenze di voto per Morales e Mesa erano coerenti, contrariamente alle prime dichiarazioni post-elettive dell’OSA: «Né il conteggio rapido né il conteggio ufficiale mostrano improvvisi cambiamenti nelle tendenze nei risultati finali, e la stessa tendenza ben nota, spiegabile dalla geografia, è evidente in entrambi i casi».

«Incoraggiamo chiunque sia interessato a quello che è successo alle elezioni in Bolivia a fare il proprio esame dei fogli di calcolo e la propria analisi statistica», ha affermato Long. «Speriamo che la missione elettorale dell’OSA lo farà. Ma dobbiamo anche ricordare che una missione elettorale dell’OSA ha annullato i risultati elettorali ad Haiti nel 2011 senza basi statistiche o di altro tipo per farlo».