About S.C.A.M.

S.C.A.M. is an independent scorecard that tracks documented, publicly reported corruption-related cases involving members of South Africa's parliamentary parties. It gathers what has already been established or alleged in official and mainstream sources and presents it in one consistent, comparable place.

It is not an investigation and it does not decide guilt. It records where each matter stands in the legal or official process — and updates that record when a case moves, including when someone is cleared.

Each scored case is placed in one of four categories, each with a default weight:

• Convicted — weight 4
• Currently criminally charged — weight 3
• Implicated by commission / ethics body — weight 2
• Accused / under investigation — weight 1

Complaint lodged — no official action yet — never scored. Listed for transparency. Not scored until an investigating body opens an investigation or makes a finding.

Complaints by political rivals or members of the public are shown as Pending and are not scored until an official body (NPA, SIU, Public Protector, a commission, or Parliament's ethics committee) has opened an investigation or made a finding. This rule applies to every party.

Attribution. A matter is counted under the party the person represented at the time of the alleged conduct. If they did not represent any party at the time (for example, they were a state-owned-company executive or a private contractor), it is counted under the party they represent now. The rule applies to every party.

A party's score is the sum of its cases' weights. Resolved cases (for example, charges withdrawn or an acquittal) stay on record but no longer count. Per-seat score divides by the party's 2024 seats (per 10 seats) so parties of different size can be compared more fairly. You can adjust every weight on the leaderboard and watch the ranking change — the weighting is an editorial choice, not a fact.

Default ordering. Every page — leaderboard, party pages, compare, embed widget and share images — uses the same default: the per-seat score, highest first. Only parties with a score above 0 get a rank number; parties with no scored matters are listed separately, unranked and alphabetically. That reflects what has been publicly reported and verified — it is not a clean bill of health.

Named people per 10 seats. The number of distinct people with at least one unresolved, scored matter (people, not cases — one person with two cases counts once), divided by the party's 2024 seats, times 10. Named people need not be current MPs, so treat this as an approximate ratio, not a count of MPs.

Composite index (optional view). One optional way to combine size, severity and recency. Higher = more or more serious recorded matters. It does not measure a party's integrity. It is a 0–100 score with three parts, each scaled so the highest party scores 100:

• Size-adjusted score (50%) — the weighted score per 10 seats. Parties under 10 seats are counted as 10 seats, so a single case in a very small party can't outweigh everyone else.
• Case severity (30%) — convictions × 4 plus charges × 3, divided by the party's total weighted score: how much of what is documented is serious rather than merely alleged.
• Recency (20%) — each case is weighted by its most recent year: this year 1.0, last year 0.85, two years ago 0.70, older 0.50.

The Composite index updates live with the weight sliders. Per-seat (the default), raw, Severity % (convicted + charged as a share of all scored matters) and named-people views are always available. Parties with no scored matters score 0 — reflecting absence of reporting, not a clean bill of health.

In October 2026 every one of the 18 parliamentary parties was searched using the same criteria: current MPs, ministers, deputy ministers, premiers and national office-bearers (and former ones where the conduct happened while they represented the party); criminal charges, convictions, and findings or referrals by the NPA, SIU, Public Protector, commissions, and parliamentary or provincial ethics bodies, since 2019. Each result is checked by an editor against its sources before it is published. Parties with no published matters were searched with the same effort.

• Sources — publicly reported court proceedings, National Prosecuting Authority statements, Zondo Commission findings, parliamentary ethics findings, Public Protector matters and mainstream South African news reporting.
• Initial compilation — the first set of cases was compiled in September 2026 with AI assistance, from the model's knowledge of public reporting. Each case names its source, but not every case yet has a direct link; we are adding links progressively.
• Ongoing monitoring — Updated: public tips, right-of-reply responses and a monthly editor check of every case — all reviewed and approved by a human editor before anything changes.
• Monthly verification check — on the 2nd of every month every case, including resolved ones, is listed for an editor with an automatic check of its source links. The editor re-checks each case; the check itself never changes a case.
• Community tips — submissions through the flag-a-case form also go into the review queue.
• AI assistance — Editors may use an AI tool to help spot relevant developments in news articles. The AI only drafts suggestions with quotes from the article; an editor checks every item against the source before anything is published or changed.
• Human review — an editor checks each item against a primary or reputable source before anything is published or changed.
• Verification dates — each case shows a last-verified date; the leaderboard shows the most recent one.
• Limitations — the dataset is not exhaustive and reflects only what has been publicly reported. See the attribution rule above; where a person has since moved party, that is noted on the case.

Every case card has a "Named in this case? Give your side" button. The person named, their party's media office or their legal representative can send a response, which goes to a private review queue. An editor verifies the sender before publishing, and verified responses appear next to the case under "Their side", edited only for length or legal reasons. Editors also record public statements already made in the press, with a link where available.

S.C.A.M. is run by JLT IT, a brand of Just Like That Productions (Pty) Ltd (South Africa).

This project is independently run and reader-funded. It receives no funding from any political party, politician or PAC and has no government affiliation. Donations never buy influence over what is published or how it is scored.

If something is wrong or out of date, please tell us — see the corrections process. Every edit to case data is listed publicly in the edit and corrections log.

General and press enquiries: hello@scamsa.org. Corrections and information requests: corrections@scamsa.org. Right of reply: reply@scamsa.org or the "Give your side" button on any case. Data can be downloaded from the export page, which includes a citation format.

Everyone who has been charged, implicated or accused is presumed innocent until proven otherwise in a court of law. This scorecard tracks allegations and legal processes — not guilt. The data reflects public sources as of its stated date and is not exhaustive; further scrutiny may surface more cases for any party.