I’m in Bolivia with a @codepink delegation to observe Sunday’s elections. Spent most of the day with the OEP (Plurinational Electoral Body – highest election authority) where they explained the process. The DIREPRE system raised alarms.

Thread

DIREPRE = System for disseminating preliminary results; it’s basically a non-binding quick count designed to give the media and citizens an idea of how the results are going. Last year’s system was called TREP; this system has since been “updated”

In 2019, by 8 p.m. on election day, the TREP was supposed to have counted 80% of the actas (documents that have vote tallies and more info about specific voting stations). By 6 p.m. the TREP had completed only 60% of the tally.

To make up for the shortfall, the TREP abandoned their random sampling of actas and instead loaded actas from Santa Cruz, an opposition stronghold, to reach the 80% mark

This led to a much closer preliminary result between Evo Morales (MAS) and Carlos Mesa (Comunidad Cuidadana) as it the prelim results excluded rural areas. See the @CEPRDC report for more info. https://cepr.net/report/analysis-of-the-2019-bolivia-election/

Given how problematic the TREP was, the OEP decided to implement a new system called DIREPRE. While TREP was outsourced to a private company, DIREPRE is allegedly fully in the hands of the OEP

One of the main changes to DIREPRE is that they are not publishing the actas at a voting station level, instead they’re publishing them at a precinct level.

OEP claims that this makes the system faster and that it avoids duplicating work, as the official count will count all the actas.

OEP also has banned photographs of actas (typically used as a means to audit the vote). They reason that since it is not an official count, any actas that subsequently become invalidated will cause anxiety and confusion among the public

Yet OEP said that typically less than 1% of actas are ever invalidated (for not having sufficient signatures & fingerprints, using the wrong ballots, security violations, more votes than number of voters, etc.)

When asked, OEP officials could not say what percent of actas DIREPRE would tally on election night (presumably less than 80%). OEP officials also did not answer a question as to which companies they hired to audit the DIREPRE system.

DIREPRE and the official count are both brand new systems. They’ve been tested twice. 2019 was the first time that Bolivia used a non-binding quick count system to report on a presidential election.

DIREPRE’s unofficial results will be published roughly at 8p.m. local time on Sunday night (voting ends at 5 p.m.).

Today officials from the OEP gave three dates for when the official count might be ready: Oct 21, 23 or 25. Observers are encouraged to go home on the 20th.

The main concern I heard about DIREPRE today was that it will be used to establish a narrative in the media about a second round of voting (if no candidate wins at least 50% or more than 40% plus a 10 pt lead, it goes to a 2nd round).

The fear is that if the non-binding & incomplete DIREPRE results show a need for a 2nd round, the government will refuse to count the full ballots, using security issues as a justification.