Haiti’s Provisional Electoral Council is changing how it reaches voters. Instead of relying solely on citizens visiting registration centers, it is sending electoral-register agents into workplaces and public spaces.

The October 26 announcement could make registration easier for employees and people conducting everyday business. But the CEP’s own public data points to a broader test: can this change reach people the existing network has missed, resolve obstacles to completed registration, and produce a voter list that citizens can actually use?

At 8:44 a.m. Haiti time on October 27, the public data feeding the voter dashboard showed 881,544 registered voters, 936,024 applications, 2,977 active operators and 1,306 active centers out of 1,442 listed centers. These are signs of continuing administration. They do not, on their own, establish equal access or readiness for a national election. [1–2]

The central finding is straightforward: mobile registration addresses part of the access problem. The press release does not demonstrate how the CEP will overcome identification barriers, insecurity, uneven geographic coverage or the gap between applications and completed registration.

WHAT THE OCTOBER 26 NOTICE ACTUALLY ANNOUNCES

Press note CEP/DC/080-2526 says mobile teams of Agents du registre électoral, or ADRE, will facilitate registration at work or while citizens are shopping. It says officials at public institutions and private businesses have already authorized teams to enter their premises. The operation targets employees and service users holding their national identification card, or CIN. [1]

The title emphasizes public spaces, while the body also describes workplaces and institutional premises. Those are potentially useful channels, but they are not identical to a published nationwide outreach schedule.

The note also reminds voters that they may freely choose their voting center when registering. It reaffirms inclusion, impartiality, transparency and electoral integrity.

What it does not provide is equally important for evaluating implementation: the number of teams, their geographic distribution, opening hours, budget, deployment duration, numerical targets or a reporting framework. Those omissions do not prove that no internal plan exists. They mean the public cannot assess the plan from this notice alone.

THE VOTER DATA: ACTIVITY, BUT AN UNEXPLAINED PROCESSING GAP

A direct retrieval of the dashboard’s public aggregate data produced this snapshot: [2]

Indicator | October 27 retrieval
Registered voters | 881,544
Applications | 936,024
Distinct applicants, as labeled by the data | 936,024
Active operators | 2,977
Active registration centers | 1,306
Listed registration centers | 1,442

The difference between applications and registered voters is 54,480.

That difference should not automatically be described as rejected applications, a processing backlog or excluded citizens. The public fields do not explain its status composition. It could reflect stages, eligibility checks or other counting rules; determining which explanation applies requires documentation.

The CEP should reconcile the figures by publishing counts of applications awaiting review, requiring correction, rejected, withdrawn or completed, with clear definitions and reporting dates.

The equality between the application and distinct-applicant counters also needs explanation. It does not establish that no one ever submitted twice; it establishes only what these two public counters reported at retrieval.

The ten departmental registered-voter counts sum exactly to 881,544. That is a useful arithmetic check. It is not an independent audit of identity accuracy, eligibility or completeness.

THE DAILY SERIES SHOWS A SHARP SLOWDOWN—WITHOUT EXPLAINING IT

The public timeline assigns these application counts to the days preceding and including the mobile-operation announcement: [2]

Date | Displayed application count
October 23 | 2,851
October 24 | 2,231
October 25 | 661
October 26 | 482

This is a meaningful operational signal. Far fewer applications appear against the final two dates than against the preceding two.

It is not yet a causal explanation. The series does not establish the effects of security incidents, weather, weekend schedules, outages, delayed reporting or changing outreach. Nor does it show that the mobile strategy failed on its announcement day. Announcement, field deployment, data entry and final validation can occur at different times.

There is another reconciliation issue: summing all retrieved timeline rows for the voter-application type gives 969,228, compared with the summary’s 936,024 applications. The 33,204 difference means these series should not be treated as interchangeable cumulative totals without an explanation of their counting rules and refresh cycles.

This discrepancy is a request for clarification, not evidence that the CEP invented voters. Daily events, unique applicants, historical corrections and snapshot totals can legitimately differ—but the published documentation should say how.

October 27 was still in progress at retrieval and is excluded from the daily comparison.

WHY MOBILE TEAMS MAKE SENSE—AND WHY LOCATION CHOICE MATTERS

Mobile registration can reduce the time and transport costs of a separate trip to a center. It may help someone who cannot easily leave work during opening hours. Bringing an official service closer to citizens is a reasonable administrative response.

But convenience is not the same as inclusion.

If deployment mainly follows institutions and businesses that are easiest to reach, the operation could serve already accessible populations more effectively than people in remote areas, insecure neighborhoods or displacement settings.

That is a design risk, not a finding that the CEP has deliberately favored any employer or political constituency.

The appropriate test is whether teams add successful registrations among people the fixed network was not reaching. A busy registration desk at a large institution may look impressive while contributing little toward geographic or social gaps.

The CEP should publish how locations are selected, how underserved communities can request visits, and how outreach extends to people outside formal workplaces.

THE CIN REQUIREMENT LEAVES ANOTHER BOTTLENECK IN PLACE

The notice specifically describes registering people who possess their national identification card. It does not announce that electoral agents will issue replacement CINs or resolve every identity-record problem. [1]

That distinction matters for anyone whose card is missing, inaccessible or awaiting delivery. Bringing the registration service closer does not, by itself, supply the document needed to use it.

A separate OAS–Japan initiative announced in August supports identification services, including procurement of 350,000 blank biometric cards and continued deployment of ONI mobile units. That announcement establishes a planned intervention, not proof that all those cards have already reached citizens. [3]

ONI identification services and CEP electoral registration therefore need coordinated referral arrangements. Citizens should know what happens if they arrive without the required card, encounter inconsistent records or need to correct an error.

Without that connection, a mobile team can identify unmet demand without being able to complete the transaction.

AN ACTIVE CENTER IS NOT THE SAME AS AN ACCESSIBLE CENTER

The public summary classifies 1,306 of 1,442 centers as active: approximately 90.6%. The other 136 are outside that category. [2]

It would be inaccurate to say all 136 were closed by gangs. The data does not identify their individual status reasons.

It would be equally premature to treat the active-center percentage as a measure of population coverage. The figure counts locations, not citizens who can safely reach them.

The departmental data exposes uneven utilization of configured capacity. Artibonite records 58,450 registered voters against a configured capacity of 785,750—approximately 7.44%. Sud-Est records 95,862 against 418,250—approximately 22.92%. [2]

These are ratios to configured administrative capacity, not percentages of each department’s eligible population or forecasts of turnout.

They nevertheless provide a reason to ask where additional services should go. Low utilization warrants examination of operating days, staffing, connectivity, transport, public awareness and security before any single cause is assigned.

IOM’s September reporting described nearly 1.5 million internally displaced people and a crisis spreading beyond the capital. That context makes displaced voters’ access a central policy concern, although the CEP dashboard does not separately quantify their registration outcomes. [4]

CHOOSING A CENTER MUST REMAIN THE VOTER’S DECISION

The notice’s assurance of free voting-center choice deserves particular attention in workplaces.

An employer’s permission for agents to enter premises is not a substitute for an employee’s voluntary participation. Managers, political intermediaries or other people present should not steer applicants toward a preferred center or collect political information unrelated to the transaction.

These are safeguards the operation should visibly provide—not allegations that such interference has occurred.

Useful protections would include a private interaction with the electoral agent, a confirmation the voter can check, a clear correction procedure and a complaint channel accessible without employer involvement.

For displaced citizens, practical guidance should explain how to select a reachable center and what correction or transfer procedures apply. The press release’s promise needs procedures citizens can use.

A MOVING CANDIDATE DASHBOARD IS NOT A MEASURE OF MOBILE VOTER SUCCESS

The candidate dashboard is a different dataset. Its retrieved page bears an October 27, 8:30:04 a.m. update stamp and displays: [5]

Office | Candidate registrations
President | 18
Senator | 359
Deputy | 1,239
Mayor | 1,588
CASEC | 2,178
ASEC | 1,322
Total | 6,704

Compared with the earlier October 26 screenshot total of 6,657, the displayed total is 47 higher. One category decreases: deputy registrations move from 1,240 to 1,239. That cautions against treating every change as a simple count of new applicants.

The page labels the figures preliminary and reserves final approval for verification and validation. Women represent 29.2% overall, with 0% in presidential registrations, 13.4% in senatorial registrations and 6.9% in deputy registrations. [5]

These figures describe the candidacy pipeline. Mobile voter registration cannot be credited with their movement, and a registration total does not establish a final ballot.

THE PARTY-MEMBERSHIP REJECTIONS ARE A SEPARATE WARNING

Earlier screenshots supplied for Haiti Politic’s analysis showed political-membership rejection counters of 6,194,185 for invalid NINU numbers, 16,904,051 for non-matching information and 5,742,633 for duplicates. Their arithmetic sum is 28,840,869. These are historical screenshot figures, not refreshed October 27 totals. [6]

They concern organizational membership, not the mobile voter-registration operation.

The counters do not establish 28.84 million different people, a comparable number of fraudulent voters, or the size of the general electorate. Categories may overlap, submissions may be repeated and reporting periods may differ.

Their relevance is procedural: expanding registration should come with clear validation rules, correction rights, protected identity checks and explanations of what rejection counters actually measure.

More collection points do not eliminate the need for trustworthy processing.

MOBILE REGISTRATION DOES NOT SUPPLY A COMPLETE ELECTION CALENDAR

The CEP’s October 5 notice extended online candidacy registration and document submission to October 31, postponed the campaign period and said voter registration would continue. Its October 20 publication described consultations on updating the electoral calendar. [7–8]

The October 26 mobile-registration note does not announce a new polling date, a final campaign schedule or a voter-registration closing date.

The October 31 deadline shown for candidacies should therefore not be casually described as the deadline for this mobile voter operation.

These are distinct administrative processes. Conflating them creates confusion about who must do what, and by when.

THE RESULTS THAT WOULD JUSTIFY CONFIDENCE

The CEP should evaluate this operation through completed services and equitable reach.

A useful weekly public report would show team-days by area, people served, completed new registrations, corrected or transferred records, unresolved cases and the time needed to resolve them. Personal identification numbers and individual political affiliations should remain protected.

It should also show costs per completed service, while distinguishing outreach to hard-to-reach communities from cheaper activity at convenient locations. A higher cost in an underserved area may be justified; raw productivity rankings alone could discourage inclusion.

Independent observation and a documented complaint process would help test whether center choice remains voluntary and personal data stays protected.

The CEP’s new approach deserves to be evaluated on those terms. It responds to a practical barrier, but it has not yet demonstrated that it reaches those most excluded.

Haiti’s electoral challenge is not simply to raise a registration counter. It is to ensure that a citizen can obtain identification, register correctly, choose a usable center, seek a remedy when something goes wrong, and eventually vote safely under a clear timetable.

The mobile operation will earn confidence when the CEP can demonstrate that it connects those steps.

METHODOLOGY AND LIMITATIONS

The October 26 press note was read directly from the supplied CEP image. Its contents are attributed to the CEP; no field verification of the announced deployments was obtained.

The voter dashboard’s public JavaScript identifies aggregate data requests for stats-summary, stats-timeline, stats-geography and geography. Those public, unauthenticated reporting requests were retrieved on October 27 at approximately 12:44:54 UTC, or 8:44:54 a.m. Haiti time. Analysis filters the reporting records to applicationType 1, matching the public frontend’s voter-registration selection. No individual voter records or private systems were accessed.

The retrieved departmental registered-voter counts sum to the summary total. The timeline does not reconcile to the summary application count; this limitation is discussed rather than suppressed. Retrievals are not a transactionally synchronized database export, and individual counters may refresh at different times.

The candidate page was retrieved separately and displayed an 8:30:04 a.m. update stamp. Its totals and the earlier screenshots must not be treated as final validated candidacies. Party-membership rejection figures are explicitly historical and were not independently refreshed.

No field visits, interviews, mobile-team attendance sheets, procurement files or independent operational audit were obtained. The article distinguishes documented statements, arithmetic findings, design risks and unanswered questions.

SOURCES

[1] CEP, press note CEP/DC/080-2526, October 26, 2026, “Le CEP lance une opération mobile d’inscription des électeurs au sein des espaces publics.” Supplied image: CEP-press-release.jpg.

[2] CEP, voter-registration dashboard and its public aggregate reporting data, retrieved October 27, 2026.
https://cephaiti.ht/statistiques-inscription-des-electeurs/
https://stats.cephaiti.ht/
Public frontend inspected:
https://stats.cephaiti.ht/assets/index-Bo8dlgrH.js
Public reporting requests used POST with an empty JSON search object:
https://stats.cephaiti.ht/api//stats-summary/search
https://stats.cephaiti.ht/api//stats-timeline/search
https://stats.cephaiti.ht/api//stats-geography/search
https://stats.cephaiti.ht/api//geography/search

[3] OAS, August 18, 2026, cooperation agreement with Japan supporting identification services.
https://www.oas.org/en/media_center/press_release.asp?sCodigo=E-090%2F26

[4] IOM, September 25, 2026, “New IOM Data: Nearly 1.5 Million Displaced in Haiti as Hurricane Risks Mount.”
https://www.iom.int/news/new-iom-data-nearly-15-million-displaced-haiti-hurricane-risks-mount
Public reproduction consulted:
https://ebs.publicnow.com/view/8910FC411A926ED1B942DFBFF607005D0632B3FE

[5] CEP, candidate-registration dashboard, retrieved October 27, 2026; displayed update 08:30:04 a.m.
https://tableau.cephaiti.ht/

[6] Previously supplied CEP dashboard screenshots, transcribed and analyzed in Haiti Politic’s October 26 article, “Millions of Rejected Entries, No Settled Election Calendar.” Historical figures retained for context.
Official module entry:
https://cephaiti.ht/statistiques-enregistrement-adherent/

[7] CEP, October 5, 2026, extension of candidate registration and adjustments to electoral deadlines.
https://cephaiti.ht/prolongation-de-la-periode-dinscription-en-ligne-des-candidats-et-ajustement-de-certaines-echeances-du-processus-electoral/

[8] CEP, October 20, 2026, consultations on updating the electoral calendar.
https://cephaiti.ht/actualisation-du-calendrier-electoral-le-cep-consulte-les-acteurs-politiques-et-la-societe-civile/

Additional institutional context:
CEP, Electoral Register Directorate, responsibilities concerning the general electoral list.
https://cephaiti.ht/direction-du-registre-electoral/
CEP, September voter-registration outreach notice discussing first-time voters.
https://cephaiti.ht/la-direction-de-communication-du-cep-intensifie-la-campagne-de-sensibilisation-et-de-motivation-autour-de-linscription-des-electeurs/

The displayed registration total should not automatically be equated with Haiti’s entire eligible electorate. Its relationship to existing electoral lists requires a documented definition.