Purpose. This note sets out, for each exhibit published on Bailiwick & Ballot, the sources on which it rests, the manner in which it was constructed from those sources, the reconciliations performed against independent records, and the qualifications subject to which it should be read. It is intended to permit a reader to trace any figure appearing on the dashboard to the record from which it was taken, and to assess for himself the weight that figure will bear.

Nature of the work. The dashboard is independent research and analysis prepared by the author in his personal capacity. It was not commissioned by, prepared for, or undertaken at the instance of the Commission on Elections, of any party-list organization, or of any other person or body. It is not an official government record or publication, and no part of it has been adopted, reviewed, verified or endorsed by the Commission on Elections or by any other government agency.

Reservation as to legal advice. Nothing in this dashboard, in this note, or in any exhibit or table published with them constitutes legal advice, and nothing in them is to be construed as legal advice or as the rendering of a legal opinion upon the affairs of any person or organization. Statements of law appearing in Part II are included to disclose the framework against which the registration and delisting exhibits were constructed, and for that purpose only. No lawyer-client relationship arises from the publication of, or from reliance upon, any part of this work. A reader whose interests may be affected by the matters discussed here should take his own advice upon his own facts.

Status: not peer-reviewed. This work has not undergone peer review. It has not been submitted to, accepted by, or published in any scholarly or professional journal, and it is not offered for such publication. It has not been refereed, editorially reviewed, or independently replicated by any third party. Where this note adopts the conventions of scholarly reporting — a register of sources, numbered references, a statement of limitations — it does so because those conventions make the basis of the work legible and its defects locatable, and not in order to suggest that the work carries the warrant that peer review confers. The reader is accordingly asked to treat every finding here as the author’s own, resting on the sources disclosed in Part III and subject without exception to the limitations stated in Part V.

Currency and revision. The dashboard states a position as at the date shown in the masthead. The great majority of the materials on which it rests are publicly available documents and datasets held by third parties, principally the Commission on Elections. Such materials are liable to be amended, superseded, re-tabulated, withdrawn from publication or decommissioned without notice, and several relied upon here have already suffered one or more of those fates during the currency of this project. A figure correct against the record as it stood on the date of collection may therefore cease to correspond to the record as it stands when read. Dates of collection are given for every source in Part III for that reason, and the limitation is developed at Part V(B).

Governing statute. The party-list system is governed by Republic Act No. 7941, the Party-List System Act.[1] Registration of a party-list group, and the cancellation of that registration, are matters committed by the Act to the Commission on Elections.

Grounds relied upon. Section 6 of the Act enumerates the grounds upon which the Commission may, motu proprio or upon verified complaint of any interested party, and after due notice and hearing, refuse or cancel a registration. Two of those grounds are material to the exhibits on the Registration tab:

Section 6(8) — “It fails to participate in the last two (2) preceding elections or fails to obtain at least two per centum (2%) of the votes cast under the party-list system in the two (2) preceding elections for the constituency in which it has registered.”

A third ground is material to the treatment of a single group on the Election tab:

Section 6(5) — “It violates or fails to comply with laws, rules or regulations relating to elections;”

Construction of section 6(8). Section 6(8) states two grounds, not one. In Philippine Guardians Brotherhood, Inc. v. Commission on Elections[2] the Supreme Court, En Banc, read the word “or” in the provision as disjunctive and held that “[t]he plain, clear and unmistakable language of the law provides for two (2) separate reasons for delisting”, so that “[a] delisting based on a mixture or fusion of these two different and separate grounds for delisting is therefore a strained application of the law.” The Court abandoned Minero, which had treated a group that did not participate as having, by default, obtained less than two per centum. This dashboard reads the provision accordingly: a group that participates but polls below the threshold does not thereby fail to participate; a group that sits out an election has no vote in it to be measured; and the two counts are never added together anywhere on this dashboard.

Consequence for the exhibits. The 21 card, the 38 card and the Comeback groups for 2028 cohort on the Registration tab are each constructed upon the reading at ¶ 8, applied to the ballot population disclosed at Part IV(F). Each card accordingly identifies groups whose electoral record, on the published returns, answers the description in one limb of section 6(8). It does not follow, and is nowhere asserted, that any group so identified stands delisted, is the subject of any proceeding, or will be acted upon. Cancellation is an act of the Commission; the cards record exposure on the face of the returns, and nothing more. The point is stated on the cards themselves and is repeated at Part V(G).

The published orders. The Commission’s exercise of the power is recorded on this dashboard through three published delisting resolutions, transcribed in full and segmented by the ground each cancellation was laid upon: Resolution No. 10273 of 19 March 2018,[3] Resolution No. 10696 of 17 February 2021,[4] and Resolution No. 11071 of 1 October 2024.[5] The grounds shown in the Delisting Resolutions exhibit are the orders’ own characterizations, not this project’s. One cancellation in the 2018 order rests on a footing of its own and is carried separately as a special ground.

The 2028 registration deadline. Section 6 of Resolution No. 11204[6] fixes 18 December 2026, within regular working hours, as the last day on which a petition for registration as a party-list organization intending to contest the 2028 elections may be filed. The countdown card on the Registration tab is computed against that date in Philippine standard time.

An Waray. The treatment of An Waray in the 2022 exhibits follows An Waray Party-List v. Commission on Elections,[7] in which the Court dismissed the petition for certiorari and affirmed the Second Division’s resolution of 2 June 2023 and the En Banc’s resolution of 14 August 2023 in SPP No. 19-008. The ground was section 6(5), and the violation dated from the 2013 term. An Waray won a seat in 2022 and was unseated after the election by cancellation of its registration; it is plotted on this dashboard as a group that won and was later unseated, and not as a group that missed. The same convention is applied to Duterte Youth in 2025.

Reservation. Parts II and IV disclose a reading of the Act and of the authorities cited. That reading is the author’s. It is not a determination, and it is not offered as one. The Commission on Elections determines registration and cancellation, and its determination controls. Where this note states that a group is within the description of a ground, it states a conclusion about the published returns; it states nothing about the outcome of any proceeding, whether pending or prospective.

This is an independent, single-author study conducted without institutional affiliation, external funding, research assistance or editorial oversight. It has not been peer-reviewed, refereed or independently replicated, and it is not offered for publication in any scholarly or professional journal. The apparatus of this note — the source register, the per-exhibit construction statements, the numbered references — is adopted because it makes the work auditable, and the reader is asked to audit it. It does not carry, and is not intended to imply, the warrant that peer review confers. Every finding is the author’s own and should be weighed as such.

A consequence follows for how the findings should be used. Where a figure here is to bear weight — in argument, in a filing, in advice to a client — the prudent course is to verify it against the source named in Part III rather than to rely upon this dashboard as authority. The register exists so that such verification is possible.

The study rests very largely on publicly available materials held by third parties, and principally by the Commission on Elections. Such materials are held at the publisher’s discretion. They may be amended, re-tabulated, superseded, withdrawn or decommissioned without notice, and no undertaking is given by any custodian that what was published will remain published. The consequence is that a figure verified against the record as it stood on the date of collection may not be reproducible against the record as it stands when read; and the study is to that extent contingent upon conditions outside the author’s control.

This is not a theoretical concern in the present case. Three instances arose during the currency of the project and are material to the weight the exhibits will bear. COMELEC’s 2016 and 2019 per-precinct results sites have been permanently decommissioned; the domains no longer resolve, and the underlying data was never archived, so the sub-national record for those two elections cannot now be obtained from the Commission at all. COMELEC’s 2022 per-board results interface refused access from 13 September 2026; this project’s walk of the preceding day is consequently the only copy in its possession and cannot be re-collected or independently re-verified from source. And the 2016 national canvass report survives only as scanned tally sheets which the project has been unable to read, so a question about that year’s ballot composition is closed by decision rather than resolved by evidence.

Where a source has become unavailable the exhibit that depended on it is not silently rebased. The basis is stated, the date of collection is given, and the limitation is disclosed at the point of use.

Because of the decommissioning described above, the sub-national layers for 2016 and 2019 are not drawn from official returns. They are drawn from municipal returns collected by the OpenHalalan project from two broadcaster election feeds — GMA for 2016 and ABS-CBN for 2019. These are secondary sources, assembled under election-night conditions by newsrooms rather than by the canvassing authority, and they are used here because they are the surviving record and not because they are equivalent to one.

The ground for using them at all is a control rather than an assurance. The same project’s 2022 layer returns a national party-list total of 36,656,659, identical to the vote to this project’s own independent walk of all 1,659 COMELEC municipal canvass boards — the same number reached by two unrelated paths. That agreement is evidence that the collection pipeline is sound. It is evidence about one year, and it is offered as no more than that. The 2016 and 2019 feeds fall short of the official national totals by 3.12% and 1.53% respectively; the province tables are scaled to the official figure by largest-remainder rounding so that each file sums exactly, which corrects the aggregate but cannot correct the distribution. Sulu and Lanao del Sur are under-reported in both feeds, and neither carries overseas voting.

Province shares from these two years are accordingly treated as reliable and individual small-province figures as approximate. No 2016 or 2019 sub-national figure in this study should be cited as a COMELEC figure, and none is so described.

Several measures the study would otherwise report across the whole series exist for part of it only, because they were never published for the remainder. Undervote and overvote are separable at national level for 2022 and 2025 and for no earlier year; sub-national blank-ballot figures exist for 2022 and 2025 alone; registered-voter tables are published by municipality from 2019 and by province only for 2016, with Metro Manila published for that year as a single region. These gaps are shown as gaps. They are not interpolated, modeled or carried forward from an adjacent year, and an exhibit that cannot cover a year says so on its face.

A related limitation attends the registration exhibits. This project has not obtained the Commission’s register of registered party-list organizations. The populations it works with are the groups that appeared on each ballot and the groups named in the three published delisting orders. Those are records of participation and of action taken; neither is a record of who is registered. Inferences about registration status are correspondingly indirect and are graded rather than asserted.

The advocacy labels applied at IV.D are this project’s own. They use the CPBRD category set, with Regional Political Parties and Regional Groups combined into a single Regional Development row, but the assignment of individual groups to those categories is the author’s. It has not been verified against the Commission’s registered sector classifications, which this project has not obtained. Every group on the charts carries a label — a best assignment, supported by researched evidence where evidence could be found — and there is no Unclassified bucket; the effect is that the charts contain no visible signal of where the evidence is thin.

The coding is graded internally by evidentiary strength: documentary evidence, a published third-party classification or the registered name, or a name-based assignment alone. A material proportion rests on the weakest of those tiers. The grades are not shown on the charts, and the reader should therefore not treat any single bar as a precise count of groups sharing a sectoral character; the charts are offered as a picture of the distribution’s shape, and they should be read at that resolution.

Two further coding constraints bear on the series. Group naming differs across years and across sources — only 27 groups appear under an identical name in all four elections — so every cross-year count depends on a name crosswalk, and 49 rename cases remain unexamined by decision. Two sector-coding disagreements are recorded and unresolved, neither of which reaches a published chart.

Four discrepancies are known, disclosed and unattributed. They are stated here together so that a reader need not assemble them from the exhibits.

(i) Two 2022 winners’ published totals differ from their canvass sum by more than one per cent — TUCP by 25,776 votes and Patrol by 4,138. The difference lies between COMELEC’s published canvass report and this project’s board walk, and has not been attributed to any canvassing unit. Those two groups carry the published figure on the vertical axis and their own canvassed share on the horizontal, so that neither axis mixes tables.

(ii) A residual difference of 47 votes in Duterte Youth’s 2025 total is recorded and unresolved. It did not arise from, and is not corrected by, the choice of figure the dashboard carries.

(iii) The 2025 precinct file is systematically low against the published winner totals — never high, median 0.128% low, worst 7.00%. Every non-winner figure derived from it is therefore a floor and every shortfall computed from it a maximum. This is a known bias, quantified but not corrected, because correcting it would substitute an estimate for a measurement.

(iv) The 2022 seat line is Akbayan’s published 236,226 rather than the canvass figure of 235,312. The choice is deliberate, is recorded once, and is applied consistently.

The registration and delisting exhibits identify groups whose electoral record, on the published returns, answers the description in one limb of section 6(8) as construed at ¶ 8. That is the whole of what they assert. They do not assert, and nothing in them should be read as asserting, that any group identified stands delisted, is the subject of any proceeding, has been classified by the Commission as at risk, or will be acted upon. Cancellation of a registration is an act of the Commission upon a case before it, and the Commission acts upon what is before it. The exhibits record exposure on the face of the returns; the distance between exposure and cancellation is not a matter these exhibits can measure.

The same reservation governs the cohort at IV.E.4 in the opposite direction. That a group does not appear on any of the three published delisting orders is not evidence that its registration survives. The orders record what the Commission has published; an order not obtained by this project, or a proceeding not yet concluded, would not be visible to it. The cohort is offered as a set of graded leads for inquiry, and the grades are the measure of how far each has been taken.

Finally, the study is descriptive. It reports what the returns show about where votes came from and what a seat cost. It does not establish why any group polled as it did, and no causal claim about electoral strategy, organizational capacity or voter behavior is made or supported by these exhibits.

The study covers the party-list elections of 2016, 2019, 2022 and 2025 and the registration position as it stands at the date in the masthead. The 2010 election is present in one of the underlying datasets but covers only 948 municipalities against some 1,634 in later years; it is excluded as unusable without heavier qualification than the exhibits could carry.

Two matters remain live and will alter the position if they move. Pilipinas Babangon Muli’s appeal against the cancellation of its registration is pending before the Supreme Court, and the 2025 seat-allocation basis shown here is the sheet that includes it; should the appeal be resolved the other basis may become the operative one. And the filing deadline at ¶ 11 falls on 18 December 2026, after which the registration exhibits describe a window that has closed. Neither is a defect in the study. Both are reasons to read it against its date.