On quality assurance and accreditation: The price of trust

This article first appeared on University World News

Quality assurance is an unglamorous phrase for a glamorous promise. A degree is not a bundle of pages; it is a sentence that says: “You can rely on me.” Employers, graduate schools, licensing bodies – none of them can interview an entire cohort. They need a shortcut.

Accreditation is supposed to be that shortcut: a way to convert uncertainty into confidence without reading every assignment or sitting in every class. When it fails, the market writes its own precautionary rules. It discounts everyone.

This is what we are living through: a crisis of credibility masquerading as a paperwork problem. The symptom is familiar.

A few poor-quality providers flood the market with credentials whose educational content is thin.

Employers, unable to distinguish at speed, begin to distrust degrees in general. Reputations congeal around a handful of brands.

Graduates from institutions without international name recognition pay the tax – more tests, lower starting offers, requests for additional certificates to prove what the degree should have settled. Economists call it a market for lemons. Universities call it unfair.

The safety of paperwork

Our current quality regime often responds with theatrical certainty. It counts inputs that are easy to count and declares victory on a checklist: contact hours, library holdings, committee minutes, a syllabus with all the requisite words.

The file looks excellent while the class loses the ability to write, calculate or solve a problem that has not been pre-labelled.

Auditors leave with a binder; students leave with a credential that requires explanation. The result is a paradox. The more forms we produce, the less the world believes us.

We should be candid about the politics behind this theatre. Many quality agencies are neither independent nor predictably funded. Their boards are appointed with a care for loyalty; their budgets arrive with a discretion that teaches deference. Under pressure, they prefer harmony to consequence.

Renewals become automatic; sanctions are reserved for scandals; poor providers learn to pass the exam. Universities adapt by optimising for auditability rather than for learning.

What can be documented outruns what should be done. In this environment, even good institutions begin to look defensive; they point to the paperwork because it is the only thing that cannot be punished.

Confidence flows from outcomes

If credibility is a commons, then one polluter can ruin it for all. The solution is not more paperwork. It is better market design.

We must rebuild trust as a public infrastructure, the way we treat aviation safety or pharmacy regulation: independent, boringly regular, and intolerant of shortcuts.

Start with independence that can be verified, not declared. Quality councils need funding formulas insulated from ministerial whim, multi-year mandates for board members staggered across political cycles, and due process rules for decisions that can survive court scrutiny.

This is not to place universities beyond the reach of the state. It is to protect the state’s own interest in being told the truth in time to act, rather than after an inquiry.

Then change what we ask for and what we publish. Input checklists will always have a place, but confidence flows from outcomes. Evidence that graduates can do what the degree implies should be visible beyond marketing.

Capstone work should be sampled and assessed by external examiners whose names and reports are made public. A programme should publish a short annual dossier: graduation rates, supervisor-to-doctoral ratios, time to repair essential equipment, supervision loads, moderation reports, employer feedback, and revision actions taken. None of these numbers goes viral. That is precisely why they are credible. They resist performance art.

Risk-based accreditation

Risk should shape effort. A mature department with stable outcomes should not face the same procedural burden as a new programme or a provider with erratic results.

Move from uniform compliance to risk-based accreditation: light touch monitoring when history warrants it – deep dives; surprise audits; and provisional approvals where it does not.

Enforcement must be swift where fraud is clear: stop admissions orders, teach out provisions for students, public notices of noncompliance. Consumer protection belongs in higher education as much as it belongs in banking.

Because credibility crosses borders whether we like it or not, we need a continental layer. A pan-African seal – earned, not assigned – can operate as a reputational backbone: a common baseline that signals to employers and regulators beyond our borders that a programme has been judged by standards not written to flatter local weaknesses.

Mutual recognition then becomes practical rather than aspirational. External examiners and accreditors should be drawn from cross-border rosters, creating a professional class whose reputations depend on the quality of their judgments, not on their proximity to the ministry.

Demonstrations of competence

Assessment needs to grow up. As large language models cheerfully write essays and exams, doubling down on invigilated recall will only teach students to hide better. Design for authenticity rather than for surveillance.

Require demonstrations of competence that are difficult to outsource: oral defences recorded for audit; open-resource exams that test judgment rather than memory; team projects with individual accountability mapped and defended; portfolios of work linked to real clients where possible.

Use AI as a servant – clustering common errors, generating draft feedback for human refinement, flagging anomalies – while keeping final judgment human and documented. Academic integrity systems should publish their policies, their error rates, and the recourse available to students. A regime that cannot be questioned is not a regime; it is a rumour.

Digital credentials can help if they carry evidence rather than hype. A secure, portable degree supplement should include verifiable links to capstone projects, datasets, code repositories, and examiner reports.

This turns a line on a CV into a dossier that travels. The aim is not to glamorise the exceptional student but to make the average claim legible to the average employer without an extra certificate purchased after graduation.

Privacy legislation will need to keep up; so will our instincts. The default should be consented transparency, not public relations.

What corrodes credibility?

We must also confront the cost of getting this right. Quality assurance cannot be a hobby assigned to the most conscientious person in the department.

It requires staff, training, systems, and the authority to say no. At the same time, bloat is a threat.

The point is a small apparatus with large consequences, not a large apparatus that confuses activity with assurance. A crude rule suggests itself: the fewer meetings, the more transcripts we trust.

Replace ceremony with cadence: brief, regular reviews aligned to cohort cycles; sunset clauses that retire modules politely; revision logs that show how criticism becomes syllabus.

The private sector is not the villain by definition, and the public sector is not virtuous by default. Both contain excellence and mediocrity. Both should face the same standards, the same sanctions, and the same opportunities to improve.

Degree mills should be closed, not tolerated as relief valves for political promises. Glorified training centres should not be accredited because they have attractive buildings; universities should not be shielded from consequence because they have attractive histories.

There is a role here for students beyond being surveyed. A trusted system protects dissent. Student ombuds offices need independence and teeth; whistleblowers need anonymity and protection from retaliation; grievance processes need timelines and the possibility of external review.

Nothing corrodes credibility faster than an institution that cannot hear its own complaints without reaching for a lawyer or a police officer.

We own students’ reliability

The AI era makes all of this both harder and more urgent. If knowledge can be summarised well enough to fool a tired marker, then the value of a degree shifts to what cannot be faked at scale: reasoning under constraint, ethical judgment, collaborative problem-solving, the ability to set up and interpret an experiment, the capacity to explain trade-offs to a lay audience.

Quality assurance must, awkwardly, learn to look at these things without turning them into more paperwork. It must sample, question and publish. It must be comfortable with recordings and with ambiguity. It must treat faculty development as a core intervention, not as a conference.

A seal that can be bought is a sticker. A report that cannot be read is a confession. A regulator that frightens only the honest is a comedian.

If we want the global marketplace to stop discounting our degrees, we must stop discounting our own standards. Build independence that outlasts speeches. Publish what matters and ignore what flatters. Sanction quickly when trust is abused and support steadily when quality is earned the slow way.

If we do, something interesting happens to the price of trust. Employers learn that our seal is cheaper to believe than to duplicate.

Graduates discover that their qualifications no longer require an additional certificate to be understood.

Universities regain the right to be proud without being defensive. And Africa’s contribution to the global conversation arrives marked, not by suspicion, but by a familiar sign: reliability – the most persuasive argument in higher education, and the one we owe our students most.