When the pressure on voters comes from the government itself
Mexico's highest electoral court threw out a governorship because the government itself was the one doing the corrupting — and no audit in the world could have caught it before it happened.
It is the evening of October 22, 2015, and five magistrates in Mexico City have just done something that takes a moment to absorb: they have annulled a governorship.
Not a city council seat. Not a local race. A governorship — the kind of result that is supposed to be settled, certified, done.
The state of Colima had held its election. A candidate had been declared the winner. A certificate of majority had been issued. And now the Superior Chamber of Mexico's Electoral Tribunal of the Judicial Branch of the Federation — the TEPJF — was revoking that certificate and ordering the whole race run again.
The reason was not a hacked voting machine. It was not a broken chain of custody or a miscounted stack of paper. The problem was the government itself.
The governor's office as a campaign apparatus
What the TEPJF found, in resolving case SUP-JRC-678/2015, was a pattern of official conduct that turned public resources into campaign weapons. State officials had steered social programmes — benefits, services, the kind of things ordinary people depend on — toward voters in ways designed to benefit the coalition candidate. Opposition campaign workers had faced intimidation connected to arrests. The principles of equity and neutrality that Mexico's electoral law demands of officeholders had been violated in ways the tribunal found sufficient to taint the result.
The tribunal revoked the winner's certificate and referred the conduct of the officials involved for investigation. It ordered the National Electoral Institute to organize an extraordinary election.
Read that again slowly: an election was annulled not because anyone stuffed a ballot box, but because the people already in power used that power to tilt the contest before a single vote was cast.
This is the scenario that almost no post-election audit is designed to catch.
Corruption from above is still corruption
Most of what gets called "election integrity" focuses on what happens at or after the ballot box: whether machines counted correctly, whether ballots were stored securely, whether the tally matched the scanner tape. That is real and important work. But it rests on an assumption that the inputs to the counting process — the votes themselves — actually reflect free choices.
The Colima case breaks that assumption.
When government officials control the distribution of benefits, control who gets arrested, and control the machinery of the state, they can corrupt a vote without touching a single ballot. The coercion happens upstream: in the threat implied by an arrest, in the favour implied by a programme steered toward your neighbourhood at the right moment. By the time any ballot is cast, the field is already tilted.
This is what electoral scholars mean by "misuse of state resources." It is not an edge case in fragile democracies. It is a documented, recurring pathology that international election monitors flag repeatedly — in EU member states, in established democracies, everywhere the people running the government are also running in the election.
The monitors keep finding it
The OSCE/ODIHR mission that observed Albania's 2021 parliamentary elections documented "allegations of vote-buying by political parties" that were "widespread during the campaign," alongside concerns about "misuse of state resources and pressure on voters." In Kyrgyzstan's 2017 presidential election, the same monitoring body pointed to "cases of pressure on voters, vote-buying and the misuse of public resources" and urged the authorities to "guarantee the right to a free and secret choice." In North Macedonia's 2025 local elections, observers found "instances of misuse of public resources, allegations of pressure on public-sector employees and voters, and allegations of vote-buying."
Notice what all of these findings share: the word allegations, because this form of corruption is structurally hard to prove. You cannot hold a benefit envelope up to a light and see a fingerprint. You cannot run a forensic audit on a voter's fear.
The secret ballot — first institutionalized in Australia in the 1850s and introduced to Britain by the Ballot Act of 1872 — was designed as a structural defence against exactly this kind of pressure. If no one can prove how you voted, the bribe cannot be verified, the threat cannot be enforced. Secrecy does not just protect privacy; it dismantles the enforcement mechanism that makes coercion profitable.
But secrecy alone is not enough when the coercer is the government. A government that controls public employment, public contracts, and public benefits does not need to see your ballot. It just needs you to believe that compliance is being tracked and that non-compliance has a cost.
In Colima in 2015, according to the TEPJF, that belief was actively cultivated.
The last-resort safeguard and its limits
The Mexican tribunal's annulment was an extraordinary act of institutional courage. An independent court looked at what officials had done and said: this contest cannot stand. That matters enormously.
But notice what had to happen first.
A coalition had to win the election under those conditions. An appeal had to be filed. Magistrates had to review the evidence, deliberate, and reach a majority. The certificate had to be revoked months after it was issued. The entire machinery of justice had to work correctly while the people being investigated were still in office.
That process worked in Colima. It does not always work. And even when it does, it takes time — time during which the questionable winner holds office, makes appointments, and shapes the conditions of the re-run.
An annulment is a correction. It is not a prevention. The difference matters.
Courts that can void elections are essential. But they are the fire department. What you actually want is a building code that makes the fire less likely — and architecture that makes the fire immediately visible when it starts.
That means something more than post-hoc adjudication. It means an electoral process whose neutrality is verifiable in real time, not merely assertable after the damage is done.
What verifiability actually means here
This is where the conversation about election integrity tends to split. One camp focuses on the count: was the machine accurate, was the paper trail intact, was the audit sound? Another camp focuses on the campaign: was there undue pressure, were resources misused, were voters free? These two conversations almost never happen in the same room.
They should. Because they are both, at root, arguments about the same thing: whether the announced result represents what voters would have chosen in a genuinely free contest, and whether anyone can independently verify that it does.
A system that publishes instant, precinct-level, machine-readable results the moment each precinct closes takes away one set of tools from a would-be manipulator: the ability to quietly adjust totals after the fact, or to create confusion about what the numbers actually were. That is real and worth building. See how common the reporting gap is across different systems.
But it does not, by itself, stop a government from steering benefits before the vote or arresting opposition workers during the campaign. For that, you need independent oversight bodies that are genuinely independent — not appointed by the officials they are supposed to restrain. You need published, auditable records of how public resources were distributed, by whom, and when, in the period before elections. You need results that are so granular and so immediately verifiable that any precinct-level anomaly produced by intimidation becomes visible to any citizen with a spreadsheet.
None of this is radical. It is just what "verifiable" actually requires when the threat is coming from inside the building.
The architecture problem
Consider what a fully transparent electoral process would have to do to resist the kind of pressure documented in Colima.
It would need results that are published at the precinct level, instantly, in a format any citizen can download and cross-reference. It would need an auditable chain of custody for every resource — social programme, government vehicle, official communication — that touches election-adjacent activities. It would need candidate-registration and campaign-finance data that is current and public, not filed months later in formats that require a law firm to parse.
And critically, it would need those things to be verified by people with no stake in the outcome, not simply asserted by officials whose interests run in one direction.
This is the throughline that runs through every case where an election ended up in court. Kenya's Supreme Court annulled its 2017 presidential election because results forms could not be independently verified against the digital tally — the chain of transmission was opaque. Malawi's courts voided their 2019 election because tally sheets had been altered with correction fluid and no tamper-evident record existed. Austria had to re-run its presidential election because postal ballot procedures were broken in ways that made the outcome unverifiable — even though no fraud was found.
In each case, the answer the court implicitly demanded was the same: produce a record that an independent observer can check, without taking anyone's word for it. Not "the minister confirmed it." Not "the vendor certified it." A record. Checkable. By anyone.
That standard applies just as much to the campaign environment as it does to the count.
The real lesson from Colima
The Colima governorship annulment is often cited as evidence that Mexico's electoral institutions work. And it is — the TEPJF functioned as it should. But that framing points the moral in the wrong direction.
The lesson is not that tribunals will catch this. The lesson is that the government was able to systematically distort a statewide election, and the only thing that stopped the distorted result from standing was a court acting months after the fact, under political pressure, with the burden of proof on the people who had already lost.
That is an uncomfortably thin margin.
The durable defence is not a better post-hoc remedy. It is a process so transparent, so granular, and so independently verifiable that pressure from above shows up in the data the moment it is applied — not as a judicial finding after an appeal, but as an anomaly that any observer with internet access can see and name.
Replace trust in officials with verifiability that anyone can check. That is not a slogan. In Colima, it was the difference between a corrected result and an uncontested one.
An election that can only be verified by the people running it is not verified at all.
See how these gaps show up across different electoral systems or read the two-minute overview of what end-to-end verifiability actually requires.
Sources
- TEPJF — Boletín: 'El TEPJF anuló la elección de gobernador en Colima' (SUP-JRC-678/2015)
- OSCE/ODIHR, Republic of Albania Parliamentary Elections 25 April 2021 — ODIHR Election Observation Mission Final Report
- OSCE/ODIHR — Kyrgyz Republic, Presidential Election, 15 October 2017: Final Report (8 Mar 2018)
- OSCE/ODIHR — North Macedonia, Local Elections 2025: ODIHR EOM Final Report (7 Apr 2026)
- UK primary legislation — Ballot Act 1872 (35 & 36 Vict. c. 33), legislation.gov.uk
- Supreme Court of Kenya, Presidential Election Petition No. 1 of 2017 (Odinga v IEBC), Judgment of 20 September 2017
- Supreme Court of Appeal of Malawi, Mutharika & Electoral Commission v Chilima & Chakwera, MSCA Constitutional Appeal No. 1 of 2020, Judgment of 8 May 2020
- Verfassungsgerichtshof, Decision W I 6/2016-125, 1 July 2016 (official English translation)