Nickels Before Dimes: Concluding Remarks

That is the fourteenth in a sequence of weblog posts overlaying a report by Diego Escobari and Gary Hoover on the 2019 presidential election in Bolivia. Their conclusions don’t stand as much as scrutiny, as we be aware in our report nickels earlier than dimes. Right here we develop on varied claims and conclusions that Escobari and Hoover make of their article. Hyperlinks to posts: half One, Second half, Half three, half 4, half 5, half six, half seven, half eight, half 9, half ten, half eleven, half twelve and half 13.

To sum up:

This weblog submit sequence explored assumptions and prejudices within the 2020 report by Escobari and Gary A. Hoover, Evo Morales and Electoral Fraud in Bolivia: A Pure Experiment and Discontinuity Proof. We checked out bias within the census and located that assist for any given candidate needn’t be constant throughout the census. A candidate’s share of votes can development upwards and even swing upwards quickly in the direction of the top of the rely.

We’ve seen that the 2019 elections in Bolivia, in addition to the 2016 referendum voting, confirmed related tendencies as arguably “left” voters tended to report their votes later than others. Definitely, rural counties – the place voters had been extra pro-referendum – got here later. Districts in main cities — the place Mesa was notably standard even amongst opposition candidates — tended to report extra absolutely early within the census. Voting patterns on the time of the TSE announcement, mixed with information of which polling stations went unannounced, strongly prompt Morales would win within the first spherical of the 2019 presidential election.

This clearly proves the OAS’ declare that the late outcomes and a primary spherical win had been “inexplicable”. It’s in no way potential to clarify an election of any measurement right down to the final vote. Nevertheless, primarily based on knowledge supplied by the TSE on the time of the announcement, it seems that a Morales win within the first spherical was not solely potential, however extremely possible. At this level, the OAS has not been adequately held accountable for its half in destroying the credibility of the Bolivian electoral system. Subsequently, we reiterate our name for the OAS to be open and clear about their 2019 election-related actions.

We’ve seen that naive “distinction mannequin” estimates are deceptive. Based mostly on the suspicion that late polling stations had been tainted by fraud, these fashions measure the affect of disruptive elements resembling rurality.

We’ve seen that including even small quantities of details about the election, resembling county-level outcomes from 2016, goes a protracted method to explaining the 2019 election outcomes, however care have to be taken to not over-structure our mannequin with assumptions which are too strict. Benign variations between 2016 and 2019 might violate the parallel development assumption of the difference-in-difference mannequin and lead to a biased estimate.

We have seen that when Escobari and Hoover get thinking about these benign variations, little or no of the election goes unexplained. Nevertheless, Escobari and Hoover then reinterpret their outcomes by flagging the non-parallel traits themselves as indicators of fraud. We emphasize that this reinterpretation is totally inconsistent with their earlier mannequin assumptions. In every case, the reinterpretation resets far-reaching assumptions and discards crucial benign causes for these variations in traits.

We strongly query the “triple” distinction outcomes of Escobari and Hoover. The frequent baseline of the 2016 outcomes reduces the mannequin to a distinction in distinction by contrasting the margins of the massive social gathering (MAS-CC) with the margins of the smaller social gathering (MTS-21F), though these completely different margins are by no means anticipated will be parallel. We marvel how Escobari and Hoover may discover such small “Triple” variations no matter interpretation.

Even accepting Escobari and Hoover’s outcomes, their reinterpretation requires that the measured development variations once more point out 2.5 share factors of fraud, unjustifiably robust mannequin assumptions.

We reiterate our name for Escobari and Hoover to talk out brazenly and transparently on a delicate political problem through which they’ve interfered.