Almost every summary of the B-BBEE skills development element says the same thing. Spend 6% on training.
Three things are wrong with that sentence, and each of them changes the number. It is not 6% of payroll. The 6% is a shorthand for several separate targets. And if your sector has its own code, the figure may not be 6% at all.
Here is how to arrive at the actual number.
What is the leviable amount?
The base for the calculation, and a specific legal definition rather than a loose synonym for payroll.
The B-BBEE Commission has set out the position directly. The target for skills development is based on the leviable amount, and that term is defined in the Skills Development Levies Act as the total amount of remuneration paid or payable, or deemed to be paid or payable, by an employer to its employees during any month, determined in accordance with the Fourth Schedule to the Income Tax Act for the purposes of determining the employer's liability for employees' tax, whether or not the employer is liable to deduct or withhold that tax.
The final clause is the one that catches people. Whether or not the employer is liable to deduct employees' tax, the remuneration still forms part of the leviable amount. So the base is not limited to the staff you actually withhold PAYE for, which is a distinction that matters in businesses with a mix of employment arrangements.
Not sure what your own SETA has actually paid you? Send us a payroll summary and your SETA registration and we will tell you what you were entitled to and what was claimed.
Request a call backTwo consequences follow. First, the Income Tax Act sets out amounts that must be excluded when determining the leviable amount, so the leviable amount is generally lower than a company's gross payroll figure. Second, it is measured across the full measurement period, not annualised from a convenient month.
An employer calculating the target off total payroll cost, including items the Fourth Schedule excludes, sets themselves a higher target than the codes require and then spends against it. That is an expensive kind of caution.
How do you find the number without rebuilding payroll?
You already submit it every month. It is on the EMP201.
The levy on that return is 1% of the leviable amount for the month. So the arithmetic runs backwards cleanly: multiply each month's levy figure by 100 to recover that month's leviable amount, then add those figures across your measurement period.
Worked through, with three months for illustration. A levy of R1 000 in March implies a leviable amount of R100 000 for March. R1 150 in April implies R115 000. R1 200 in May implies R120 000. Added together, the leviable amount for those three months is R335 000.
Run that across twelve months of returns and you have the figure the whole element is calculated from, taken from documents SARS already holds and your verification agency will accept. No estimates, no reconstruction.
One caution worth stating plainly. Check the addition yourself. Published examples of this calculation contain arithmetic errors, and a target built on a mistyped sum is wrong in a way nobody notices until verification.
What does the 6% actually consist of?
Three separate measured items on the generic scorecard, each with its own weighting and its own target, which is why treating 6% as one pot produces a plan that misses points it paid for.
Skills development expenditure on learning programmes for black people, measured as a percentage of the leviable amount, carries a target of 3.5%. Bursaries for black students at higher education institutions carry a target of 2.5%. Skills development expenditure on black employees with disabilities carries a target of 0.3%.
Add those three and you get 6.3%, which is where the familiar 6% comes from and why it is approximate. Alongside them sit two further items that are not spend targets at all: learnerships, apprenticeships and internships measured as a percentage of total employees, with a target of 5%, and bonus points for the full-time employment of black people at the end of those programmes.
The planning implication is direct. An employer who spends 6.3% of the leviable amount entirely on general training for black employees has overshot one target and scored nothing on two others. The bursary line and the disability line are separate, and they are the two most commonly left empty.
Why is this a priority element?
Because the consequence of underperforming is disproportionate to the points involved.
Skills development is a priority element, which means an entity has to achieve a minimum proportion of the element, commonly expressed as at least 40% of it, or be discounted a full level on the overall scorecard regardless of how the rest of the card looks.
That asymmetry is the whole reason to calculate the base correctly. A company that is comfortable on ownership, management control and procurement can still drop from a level 2 to a level 4 by missing a sub-minimum on this element, and levels are what procurement decisions actually turn on.
Is your target even 6%?
Possibly not, and this is the question construction and civil engineering employers should ask first.
Sector codes set their own targets, and they differ from the generic scorecard. The construction sector code has applied a materially lower skills development expenditure target, in the region of 2% of the leviable amount for contractors, phasing upward at defined intervals to 2.5% and then 3%, with its own weighting points and its own sub-elements.
Those sub-elements are structured differently too, measuring the portion of skills development expenditure using the adjusted recognition for gender across African employees relative to the economically active population, across black executive, senior and middle management, across black junior management, and on bursaries and scholarships. A mentorship programme features as a separate requirement.
An employer in a sector with its own code who plans against the generic 6% is not being prudent. They are measuring themselves against a scorecard that will not be applied to them, and possibly spending two or three times what their own code requires while still missing the sub-elements that code actually measures.
So the first question is not what percentage. It is which code, and the answer depends on the sector you operate in rather than on which scorecard is easiest to find online.
Does company size change the target?
Yes, and the thresholds are worth confirming rather than assuming, because they move the target by half.
The generic scorecard applies to larger measured entities. Qualifying small enterprises are measured on their own scorecard with a lower skills development expenditure target, commonly cited at 3% of the leviable amount against the generic 6%. Exempted micro enterprises are treated differently again and are generally not measured on the element in the same way.
Turnover is what determines which category you fall into, and it is assessed for the measurement period rather than set once. A business that grows across a threshold during a year moves scorecard, and the first time most employers discover this is when a verification agency applies the larger entity's targets to a plan built for a smaller one.
Establish your category before calculating anything. The sequence is category, then code, then base, then sub-targets. Getting them in a different order means redoing the work.
Which training counts towards the spend?
Defined categories rather than anything with an invoice attached.
The codes work from a learning programme matrix that specifies which categories of learning count and how they are treated. Programmes leading to registered qualifications or part-qualifications are treated most favourably, which is the same principle that governs which training attracts SETA discretionary funding.
Training for people who are not your employees can also count in defined circumstances, which is how bursaries for students at higher education institutions enter the calculation at all. And spend has to be measured and verifiable, which in practice means the invoice, the attendance record, the programme identification and the outcome all have to exist and agree with each other.
What does not work is retrofitting. Deciding in month eleven that the year's assorted short courses will be categorised into the right lines is how companies discover that spend they genuinely incurred cannot be claimed in the form they need.
Does a levy holiday change the target?
No, and the point was tested.
When SARS granted a four-month levy payment holiday as pandemic relief, the question arose whether the skills development target fell proportionately. The B-BBEE Commission addressed it directly, stating that the payment holiday was a relief measure under disaster management regulations and did not alter the requirements for B-BBEE implementation. The target remained based on the leviable amount for the full measurement period.
The principle generalises usefully. The leviable amount is a measure of your payroll, not a measure of what you paid SARS. Relief on the payment does not reduce the base, and neither would an exemption or an administrative dispute.
The sequence that gets this right
Establish which scorecard applies to you, generic or a sector code, before anything else. Then pull twelve months of EMP201 returns and build the leviable amount from the levy figures, checking the addition.
Apply the sub-targets separately rather than as a single percentage, so the bursary and disability lines are planned rather than discovered. Then categorise the training you already intend to run against the learning programme matrix, and see which lines are empty.
The empty lines are the whole exercise. Most companies are already spending enough. What costs them points is that the spend landed in one category when the scorecard was measuring four.