SAMPRA Held R303.7m in Unpaid Royalties at the End of 2024, 1.72 Times What It Paid Out That Year

South Africa’s neighbouring-rights society ended 2024 holding R 303,662,657 in royalties it had allocated to rightsholders but had not yet paid out, which is 1.72 times the R 176,700,000 it processed and paid to members and sister societies over the whole of that year.

Both numbers sit in one document, SAMPRA’s Annual Review 2024. The unpaid balance is audited to the rand at company level, signed off by PricewaterhouseCoopers Inc. The payout figure is the society’s own, stated twice in the report’s prose and rounded to the nearest hundred thousand rand.

Company-level figures are used throughout, not group, because the group also consolidates SAMPRA Development Fund NPC, a separate non-profit that receives an internal allocation from SAMPRA rather than licensing music itself. This measures needletime and other neighbouring rights in sound recordings only, meaning the royalties owed to performers and record companies when a recording is played in public or broadcast. Performing rights in the underlying musical works sit with SAMRO and mechanical rights with CAPASSO, and neither is measured here.

Sponsored

What the 1.72 actually measures, and what it does not#

It is a stock against a flow, and the two are on different clocks. SAMPRA states that it processed the 2023 distribution in September 2024 and intended to process the 2024 distribution in September 2025. So the R 176,700,000 that moved during 2024 discharged the previous year’s allocation, and the closing balance contains a distribution that was scheduled rather than overdue at the reporting date. Anyone reading 1.72 as a backlog of unpaid money is reading it wrong.

The subtraction is the part that survives that objection. Take the audited distributable amount after tax for the year, R 213,565,286, out of the closing balance of R 303,662,657, and R 90,097,371 is left over: money in the unpaid balance that the year’s own distribution does not account for. One component of that residue is known and not sized. The statements define the balance as royalties available for distribution plus cumulative interest on funds invested, so some of it is interest rather than royalty, and the report does not say how much. Nor does it age the balance. This desk does not claim to know how old the residue is, who is owed it, or why it has not moved. It reports only that it is there.

Where the money comes from, and what the society’s “total income” includes#

Radio is not the biggest line. Of the R 223,119,341 SAMPRA collected in licence fees in 2024, 58.4 per cent came from Communication to the Public licence fees, the category covering needletime royalties for recordings played to the public other than by radio broadcast, including television simulcast, retail, hospitality venues and streaming platforms. Radio broadcast licence fees were 40.9 per cent. Foreign income, collected on SAMPRA’s behalf by sister societies abroad under reciprocal agreements, was 0.7 per cent, the smallest of the three lines by a wide margin.

There is a second definition worth pinning down before anyone quotes a rate off this report. The financial summary states total income for 2024 of R 282,500,000, without saying whether that is the group or the company figure. The audited licence-fee line, at company level, is R 223,119,341. The difference is 21.0 per cent of the stated total, and it is not licence income. That matters because a cost ratio or a distribution ratio struck against total income is not the same ratio struck against what licensees actually paid for music, and the two get quoted interchangeably.

That revenue mix does not resemble SAMRO’s either. This desk has already measured South Africa’s performing-rights society, where broadcasting alone is 55 per cent of licence revenue and digital is 15 per cent. SAMPRA’s recordings-side business leans further into non-broadcast public performance, the kind of licensing that depends on a venue’s or a platform’s own reporting rather than a radio log, which is a harder stream to audit and a slower one to collect.

Two collecting societies, the same measure, opposite directions#

Expressed against a single year’s revenue, SAMPRA’s unpaid balance was 136.1 per cent of the R 223,119,341 it collected in 2024, against 136.7 per cent of the R 218,464,977 it collected in 2023. It went down. SAMRO’s equivalent ratio went the other way, from 168.3 per cent to 189.1 per cent across its own two most recent audited years, a move this desk reported separately when SAMRO’s payable balance climbed while its licence revenue fell.

SAMPRA’s own year moved in step across all three of its headline lines rather than in one of them: licence revenue up 2.1 per cent, the distributable amount after tax up 2.4 per cent, the unpaid balance up 1.7 per cent. The unpaid balance grew more slowly than the money coming in, which is what a ratio falling from 136.7 per cent to 136.1 per cent means in cash terms.

The counter-reading#

The first objection is that a two-point comparison is not a trend, and it is a fair one before it is anything else. SAMPRA’s FY2023-to-FY2024 change and the SAMRO figures set beside it are each two audited years for one organisation. The two societies measure different rights, needletime in recordings against performing rights in the underlying songs, and their financial years do not even share an end date: SAMPRA reports to 31 December, SAMRO to 30 June. This is a comparison of direction, not of magnitude, and it is not evidence that one society is better run than the other.

The second objection is that one unusual year could be carrying the result. The robustness scenarios answer that directly rather than reassuringly. Dropping the larger year, FY2024, from the two-year total takes the mean of R 301,167,502.5 down to R 298,672,348, a change of 0.83 per cent, while the total falls 50.41 per cent because that is what dropping one of two rows always does. With only two years on record this test can rule out a single distorted year and it cannot establish a trend. What it shows is that FY2023 and FY2024 sit close enough together that neither is doing the work on its own.

The third objection goes at the headline ratio itself, because its denominator is rounded in the source. Taking the payout at the top of its rounding band gives 1.718 and at the bottom gives 1.719, so the rounding cannot move the finding. That is the whole sensitivity: the imprecision that worried the reviewer is smaller than the last digit reported.

The fourth is regulatory, and it cuts in SAMPRA’s favour. The distributable amount after tax ran at 95.7 per cent of licence revenue in 2024 and 95.5 per cent in 2023, well inside the floor: Regulation 6(2) of the Regulations on the Establishment of Collecting Societies in the Music Industry requires SAMPRA to distribute at least 80 per cent of the funds it collects in a year, and Regulation 8(1) requires it to distribute those funds at least every 12 months. Nothing measured here shows a society breaching either. It shows a payment cycle that runs a year behind the collection year, and a balance carried across that cycle that is larger than the cycle explains.

Two limits belong on the page rather than only in the method box. The R 176,700,000 is paid to members and to sister societies abroad, and the report does not split it, so nothing here supports a per-member figure and none is offered. And this is one organisation in one financial year: a South African performer’s total rights income runs through more than one society, and only one of them is measured here.

What this means if you have recordings out in South Africa#

Do not book an allocation as cash. The society’s own calendar puts the 2023 distribution in September 2024 and the 2024 distribution in September 2025, so needletime money earned in a calendar year is a payment that lands in the following one. A cash-flow model that books South African neighbouring-rights income in the year it is earned is a year early.

Size the pool before you price your share of it. SAMPRA collected R 223,119,341 in licence fees in 2024 across the entire South African recorded-music sector, and R 176,700,000 went out to members and sister societies during the year. Those are the totals a claim on this society’s money gets measured against.

Weigh where the money actually comes from. Communication to the Public, at 58.4 per cent, is a blend of retail, hospitality, television simulcast and streaming rather than a streaming number on its own, and radio broadcast is 40.9 per cent. A rights holder budgeting South African neighbouring-rights income around Spotify and Apple Music alone is budgeting a share of a share.

And treat registration as the lever, because the number that should worry you is R 90,097,371. That is the part of the unpaid balance the year’s own distribution does not account for. This desk cannot say from the published record how much of it is unmatched recordings, and SAMPRA does not break the balance down by cause. What is not in doubt is the mechanism that moves money out of a balance like that one: a recording matched to a registered rightsholder with correct performer credits and banking details on file. That is administrative work, and on these figures it is competing for the same rand as any rate negotiation available to a South African performer.