When Representation Becomes the Grid’s Weakest Link

South Africa’s electricity crisis offers a hard test for a principle that is often discussed in moral language but judged in practical terms: what happens when demographic transformation begins to outrank technical competence in institutions that must work? The Eskom collapse, and the wider argument over race-conscious policy in America, suggest that a society can pursue redress with good intentions and still pay a severe price when it treats expertise as a political obstacle rather than a public asset.

The argument leaves the seminar room

Heather Mac Donald’s address at Hillsdale College in February 2024 was framed as a warning about race, excellence and the future of the United States. Its central claim was not that history should be forgotten or that discrimination never existed. It was that a new presumption has taken hold inside public institutions: whenever two racial groups produce different outcomes, the institution must be at fault.

Mac Donald called that presumption “the most consequential falsehood governing public policy today”. Her argument was aimed at universities, professional regulators, corporations, cultural bodies and police departments that have come to treat disparity as evidence of discrimination before asking what else might explain it. Academic preparation, family circumstances, geography, individual choice and differences in performance are pushed to the side. The gap itself becomes the charge sheet.

That argument is easy to dismiss as another instalment in an American culture war. It becomes harder to dismiss when the subject is not a university admissions office or a corporate diversity report but a power grid. A grid does not care whether its operators satisfy a representation target. It cares whether they understand turbines, boilers, transmission systems, maintenance schedules and failure modes. A hospital, an aircraft cockpit, a bridge and a nuclear facility impose the same blunt requirement. They must function.

This is where the Eskom story matters. It does not prove that every employment-equity policy produces institutional failure, nor does it show that one racial group possesses a monopoly on competence. It does, however, provide a case study in what can happen when a technically demanding organisation loses experienced personnel, weakens its professional pipeline and rewards administrators for satisfying political measures while the underlying system deteriorates.

The issue is not whether representation has value. It is whether representation can be pursued without reducing the standards on which collective safety depends. That distinction is often blurred because the language of equity is morally powerful. But moral language does not keep the lights on. Engineers do.

Eskom was not destroyed by one policy alone

South Africa’s electricity crisis has a complicated history. Eskom’s deterioration cannot be explained by a single statute, a single racial policy or the departure of one category of employee. Political interference, corruption, procurement failures, poor management, ageing plants, construction delays, fuel and coal problems, inadequate maintenance and years of delayed decision-making all played a role. Any account that turns the entire collapse into a simple morality tale is too neat to be trusted.

Yet the transformation framework formed part of the institutional environment in which those failures occurred. After the end of apartheid, South Africa faced an unavoidable political and moral demand: state institutions and major employers could not continue to operate as though the country’s racial history had no bearing on access to work, promotion or ownership. Employment Equity and Broad-Based Black Economic Empowerment policies were designed to alter an economy in which skilled positions, management authority and capital ownership had been distributed through a system that excluded the black majority.

The political case for intervention was therefore not imaginary. The country’s past had left a large and visible imbalance. A government that refused to confront it would have preserved a different kind of institutional risk: permanent exclusion, declining legitimacy and a state seen as belonging to one group rather than the whole population.

The problem arose when transformation became detached from the question of institutional purpose. In a technically dependent organisation, a policy can be defensible in principle and damaging in operation if it causes expertise to leave faster than it can be replaced. The relevant question is not simply how many people from each group occupy a department. It is whether the department retains enough tested knowledge to train successors, diagnose failures and make unpopular decisions before a crisis becomes visible to the public.

Reports and commentary about Eskom have described the loss of experienced white technical staff after 1998, the shrinking of engineering capacity and the growing importance of transformation scorecards in management decisions. Exact numbers vary by source and by definition. The broader pattern is less disputed: the utility struggled to retain and reproduce the knowledge required to manage a complex generation fleet.

That is not a racial law of nature. It is an organisational one. When expertise is treated as replaceable before a replacement exists, the institution becomes fragile.

The cost of losing knowledge is hidden until the system breaks

An experienced engineer carries more than a qualification. He or she carries memory: which boiler behaves unpredictably under a certain load, which component was altered during a previous repair, which warning signal can be ignored and which one cannot, which supplier delivered a part that failed in the field, and which maintenance shortcut will create a larger problem six months later.

Some of that knowledge is written down. Much of it is not. It is passed through supervision, argument, repetition and the quiet accumulation of mistakes that an organisation learns not to repeat. A new graduate may be bright and well trained, but cannot instantly possess decades of operational memory. Nor should anyone expect them to. The task of management is to create a pipeline in which experience is transferred before it disappears.

This is why the departure of a thousand people can have an effect larger than the number suggests. A departing specialist may have trained ten others, approved a maintenance plan, managed a crisis or served as the final check on a decision made by a less experienced team. Once gone, that person’s contribution is not replaced by adding a name to a staffing table.

Eskom’s technical difficulties appeared in familiar forms. Maintenance was postponed. Plant availability declined. Tube leaks and breakdowns increased. Repairs took longer. The utility had less room to absorb a failure because fewer systems were operating with a reliable margin. Load shedding then ceased to be an occasional inconvenience and became a routine feature of national life.

The public experienced the result as darkness, lost income and disrupted plans. Businesses ran generators or closed. Hospitals and small firms faced higher costs. Food, transport and communications were affected. The economic damage was measured not only in lost electricity sales but in investment postponed, production interrupted and confidence weakened. Estimates of the total cost reached into the trillions of rand, though calculations differ according to the period and assumptions used.

The important point is that institutional decline rarely announces itself at the beginning. It looks like a vacancy, a delayed inspection, an inexperienced manager, a deferred repair or a scorecard that gives more weight to process than performance. By the time citizens see the consequence, the underlying capacity may have been eroding for years.

Transformation was meant to repair legitimacy, not abolish competence

The strongest case for race-conscious policy begins with the reality that formal equality can preserve the effects of historic exclusion. If one group has been denied decent schools, professional networks, property ownership and access to skilled work for generations, applying identical rules at the end of the process may simply reproduce the starting advantage of another group.

That is why South Africa’s transformation agenda cannot be judged only by asking whether a quota or preference looks unfair in isolation. The country inherited an economic structure built through law. Correcting it required more than declaring that the future would be colour-blind. A black South African arriving at a professional labour market did not arrive with the same accumulated opportunities as a white South African whose family had benefited from the previous order. The demand for representation had a historical foundation.

But a legitimate objective does not validate every method used to reach it. The state’s duty is not merely to change the appearance of an institution. It is to make the institution more capable, more legitimate and more useful to the public it serves. If transformation produces a broader leadership class but a weaker power system, the policy has met one measure while failing the purpose of government.

This is the point at which the language of redress can become confused with the mechanics of selection. Redress may mean investing in schools, funding apprenticeships, widening recruitment, removing informal barriers, improving transport and housing, mentoring candidates, and ensuring that promotion is not controlled by old networks. Those measures can expand the pool of qualified people without lowering the standard applied to the work.

A different approach is to set demographic targets first and then search for ways to make the results fit. When that happens, a qualification becomes a political symbol rather than evidence of readiness. Managers may feel pressure to appoint the candidate who improves a transformation scorecard, even where another candidate has more relevant experience. Existing staff learn that competence is not enough; they must also navigate an ideological system whose demands may change.

That uncertainty has a cost. Skilled employees leave. Younger employees stop believing that performance will determine advancement. Training becomes defensive. Managers protect themselves by meeting visible targets rather than solving less visible problems. The institution can still look busy while losing the ability to do its core job.

The American debate began with a gap in outcomes

Mac Donald’s Hillsdale speech placed this same question inside American education and professional life. Her evidence centred on persistent differences in academic performance. She cited National Assessment of Educational Progress results showing that in 2019, 66 per cent of African American twelfth-graders had not reached even partial mastery of basic twelfth-grade mathematics, while only 7 per cent were judged proficient. Reading results were also deeply unequal, with about half of African American twelfth-graders unable to demonstrate partial mastery of basic skills.

Those figures are not a verdict on an individual child. They are a warning about preparation. A young person cannot be expected to compete in advanced mathematics, medicine or engineering without having acquired the foundations those fields require. Nor can a country close an achievement gap by declaring that the measurement of achievement is itself discriminatory.

Mac Donald argued that similar disparities appear at professional gateways. She pointed to the medical licensing examination known as Step One and to the pressure to convert it from a meaningful differentiator into a pass-fail test. She also cited California bar-examination results from February 2020: 5 per cent of black law-school graduates passed on their first attempt, compared with 52 per cent of white graduates and 42 per cent of Asian graduates.

These figures raise several questions at once. One is whether the exams measure knowledge that lawyers and doctors actually need. Another is whether professional schools are admitting students whose preparation is insufficient and then asking licensing bodies to conceal the problem. A third is whether the solution to unequal results should be better preparation or the removal of distinctions that reveal unequal preparation.

The answer should not be predetermined by racial politics. If an examination is defective, reform it for everyone. If a standard is essential, keep it and help more students reach it. What should not happen is the quiet substitution of group outcomes for individual competence, followed by the claim that no standard has been lowered because the language has changed.

A hospital patient does not receive safer care because a licensing score has been made less informative. A client does not receive better legal representation because a pass rate has been equalised. The public receives better service when the profession knows what its members can do and refuses to confuse a political objective with proof of ability.

Disparate impact turns difference into accusation

The legal doctrine of disparate impact gave the modern debate a powerful mechanism. Under its broadest interpretation, a rule can be treated as discriminatory even when it is neutral in wording and applied without animus, if it produces different outcomes among racial groups. The purpose was to address hidden barriers that could survive after explicit discrimination had been prohibited.

That purpose was understandable. Employers could not be allowed to write “no black applicants” and then escape scrutiny by claiming that the decision was neutral. Rules that served no genuine business purpose but reproduced exclusion needed to be challenged. The difficulty begins when any disparity is treated as proof that the rule itself is illegitimate.

A workforce will not automatically mirror the population in every profession. A cancer laboratory, a law partnership, an airline cockpit and a police department draw from different educational pathways, geographic pools, interests, physical requirements and standards. Their racial composition may differ for many reasons. If proportional representation becomes the expected endpoint in every institution, the institution is placed under permanent suspicion whenever reality fails to match the target.

Mac Donald’s formulation is deliberately stark: under the disparate-impact lens, a laboratory with fewer than 13 per cent black oncologists or a law firm with too few black partners can be presented as evidence of racism. The same logic can reach a cockpit, even though aviation demands a tightly defined set of skills and safety procedures.

The problem is not that institutions should be immune from scrutiny. It is that the burden of proof can move from the person alleging a particular act of discrimination to the institution trying to explain why its results are not proportional. Once that happens, administrators have a rational incentive to alter the standard, change the measurement or give preference to group identity simply to avoid being accused of failure.

A society needs a way to identify discrimination without declaring every unequal outcome discriminatory. It also needs a way to distinguish an unfair barrier from an unavoidable distribution of talent, preparation and choice. Without that distinction, the doctrine becomes less a tool for finding hidden bias than a demand that institutions engineer equal results.

Equal treatment of individuals and equal representation of groups are not the same thing. Treating them as interchangeable produces policies that may look compassionate at a distance but create arbitrary decisions up close.

Admissions reveal the pressure placed on objective standards

The controversy over Asian applicants at elite American universities illustrates the tension between group representation and individual assessment. The source material describes a familiar pattern: Asian-American students admitted to highly selective colleges often have higher test scores than white, black or Hispanic students admitted to the same institutions. The explanation offered by diversity advocates is that an institution seeks a student body with varied backgrounds and experiences, not a simple ranking by test scores.

There is a legitimate point here. Universities are not examination boards. They select students for a community as well as for a classroom. A student’s family circumstances, obstacles overcome, intellectual interests, artistic ability, public service and contribution to campus life may matter alongside scores. A university that admitted solely by one test would be efficient but narrow, and the test itself would not capture every quality worth cultivating.

The objection begins when subjective criteria become a way to obscure group quotas. If “leadership”, “fit”, “geographic diversity” or “personal qualities” are applied more harshly to one population than another, the institution has not created a holistic system. It has created a discretionary system in which the preferred result comes first and the reasons are supplied afterwards.

That history is not confined to current disputes. In the early twentieth century, geographic diversity was used at Harvard and other universities to limit Jewish enrolment without openly declaring an anti-Jewish quota. Because many Jewish applicants came from New York and the East Coast, an institution could claim that it wanted students from across the country while reducing the number of Jewish students it admitted. The language was respectable; the effect was not.

The historical comparison should not be used carelessly. Today’s admissions policies arise from different circumstances and are defended in different terms. But the comparison raises a question that no institution should evade: when a flexible criterion consistently disadvantages a particular group, what evidence shows that the criterion is measuring educational value rather than managing demographics?

The Supreme Court’s treatment of diversity added to the ambiguity. Justice Lewis Powell’s 1978 opinion rejected racial quotas while permitting race to be considered as part of educational diversity. That distinction allowed universities to claim that they were not assigning fixed numbers while still pursuing an approximate racial balance. The later legal battles showed how quickly a constitutional compromise can become a bureaucratic language game.

The fairest system is not necessarily one that ignores background. It is one that states the purpose of each criterion, applies it consistently and permits outsiders to see whether the standard is real. Secrecy is where principled selection turns into preference by another name.

Musk’s challenge is best understood as an engineering question

Elon Musk’s 2024 interview with Don Lemon became part of the wider American dispute over diversity, equity and inclusion. Musk’s critics saw a billionaire using his platform to attack efforts intended to correct historical and contemporary exclusion. Supporters saw an engineer and industrialist asking whether institutions were sacrificing performance to satisfy political expectations.

The safest way to understand his argument is not to treat it as a complete theory of race. It is an engineering question: when a system has a critical function, should selection be based on the characteristics most relevant to that function, or should it be adjusted to achieve a preferred demographic composition?

That question becomes particularly sharp in technical hiring. If a rocket fails, a battery overheats or a power station shuts down, the public does not ask whether the team achieved a target in representation. It asks who understood the risk, who checked the design and why the warning was missed. Diversity may bring useful perspectives, and a workforce that excludes large parts of society may miss talent. But those benefits must be demonstrated in the job itself rather than assumed as a substitute for qualification.

The same tension appears in elite admissions. If Asian applicants face higher barriers because an institution wants fewer of them, the policy resembles a quota regardless of whether administrators call it diversity. If, on the other hand, an institution considers context to identify students who achieved highly despite weaker opportunities, the policy can be defended—provided the method is transparent and applied without racial double standards.

Musk’s South African background gives the argument a particular charge. He grew up in a country where racial classification was not an abstract subject but a governing reality. His concern that competence could be subordinated to political targets reflects a fear that a state trying to correct injustice may create a new form of institutional dependence, in which access and promotion turn on official categories rather than demonstrated ability.

That concern should not be treated as self-validating merely because Musk built companies. Nor should it be dismissed because he is controversial. The argument stands or falls on evidence. Where DEI programmes widen recruitment and remove arbitrary barriers, they may improve an organisation. Where they require managers to ignore relevant qualifications or conceal unequal preparation, they create risk. The label tells us nothing. The selection rule does.

South Africa’s lesson is not that redress must end

The Eskom case is often presented as proof that transformation destroys competence. That conclusion goes too far. South Africa has a real obligation to broaden opportunity, and the country cannot build a stable democracy around a permanently excluded majority. An institution staffed entirely through old networks may be technically capable for a time but politically unsustainable. Legitimacy is not an optional extra in public infrastructure.

The lesson is narrower and more useful: redress must be designed around the work an institution must perform. A power utility cannot treat an engineer as interchangeable with an administrator whose primary achievement is compliance with a transformation framework. It cannot replace operational knowledge with a promise that the knowledge will be developed later. It cannot let a scorecard reward a manager for demographic movement while the plant’s availability, repair times and safety record deteriorate.

A serious transformation programme would begin before recruitment. It would invest in mathematics and science education, support technical colleges, fund apprenticeships, rebuild professional associations, create paid training routes and require senior experts to transfer knowledge to new staff. It would measure whether recruits become competent, not merely whether they were hired. It would protect the independence of technical decisions from political pressure while ensuring that the profession is open to every qualified South African.

It would also distinguish between temporary preference and permanent institutional design. A country emerging from apartheid may justify special measures for a period. But a temporary measure should have a clear purpose, a review date and evidence that it is producing a larger pool of qualified candidates. If the same preference continues indefinitely without improving the pipeline, it is no longer a bridge. It is the structure.

South Africa’s partial recovery efforts underline the point. Bringing back veteran engineers or relying on experienced specialists can restore stability, but it also reveals the weakness of a system that failed to retain and reproduce its own expertise. Rehiring the people a policy once encouraged to leave may be necessary in an emergency. It is not a long-term workforce strategy.

The country needs both justice and competence. The false choice is to say it must choose one. The real choice is whether its policies are intelligent enough to pursue both.

Public safety shows what happens when institutions retreat

Mac Donald’s most forceful warning concerned policing and public safety. She argued that disparate-impact analysis has encouraged “deprosecution and depolicing”: prosecutors, police departments and retailers become reluctant to enforce rules because enforcement may fall unevenly across racial groups. The result, she contended, is that institutions retreat from their core responsibility rather than face accusations of discrimination.

The argument is difficult because crime enforcement has not always been fair. Minority communities have experienced over-policing, abusive tactics, wrongful convictions and unequal treatment. A society that demands more enforcement without demanding lawful enforcement will produce resentment and, in some cases, further injustice. Trust matters. So do constitutional limits.

But the correction can go wrong if the existence of unequal enforcement is treated as evidence that enforcement itself is illegitimate. Walgreens, CVS and Target have faced the practical consequences of retail theft and store closures. Security decisions can carry racial implications because suspects in a particular area may not be evenly distributed across racial groups. That fact does not by itself prove that the store is targeting a race. Nor does it justify abandoning staff and customers to theft because stopping it might look unequal in aggregate.

The cost of public disorder is also unequal. Families with money can move, install security or shop elsewhere. Poorer residents are left with shuttered pharmacies, fewer shops and streets where victims expect no response. A policy intended to protect minority communities can end by removing the basic institutions those communities need.

Mac Donald cited a sharp rise in homicides in 2020 and argued that the burden fell heavily on black communities. She also referred to data showing black youths in the post-George Floyd period facing shooting rates far above those of white youths, including a reported rate one hundred times higher in some comparisons and drive-by shooting fatalities at twenty-four times the rate among white youths aged 10 to 24.

Such statistics require careful definition and verification; dramatic ratios can depend on small denominators, geography and the years selected. But the underlying reality is clear enough: violent crime imposes a devastating burden, and it is often borne by people who have the least ability to escape it.

The public deserves both impartial policing and effective policing. A department should not be excused for racial bias. It should also not be discouraged from stopping violence because the people arrested are disproportionately from one group. The relevant question is whether officers act lawfully and whether the intervention prevents harm.

Culture cannot be separated from standards forever

Mac Donald extended her argument beyond institutions that regulate life and work to the institutions that define culture. She criticised the tendency to treat classical music, Western art and theatre as evidence of oppression rather than as bodies of work that can be judged for achievement and meaning. She referred to the Metropolitan Museum of Art’s 2023 exhibition “The Fictions of Emancipation”, which interpreted some nineteenth-century white artists’ representations of slavery through a framework that, in her account, risked treating their efforts to depict slavery’s cruelty as a claim about the natural condition of black people.

Art deserves interpretation, including hostile interpretation. Museums should not be forced to praise every work they display, and no tradition should be protected from moral criticism. The concern is different: whether the interpretive framework has become so predetermined that the work’s artistic, historical or intellectual qualities are secondary to the identity of its maker.

The same problem appears in science. Mac Donald cited statements from organisations such as the American Medical Association and the journal Nature that described medicine or science through the language of white supremacy and colonial power. Such language may be intended to expose exclusions in professional history. But it can also produce a strange result: a discipline’s universal methods are treated as suspect because they developed within societies that committed injustice.

The scientific method is not morally pure because scientists are morally pure. It is valuable because claims can be tested, challenged and corrected. Cause and effect are not “white” properties. A mathematical proof does not become less valid when its author belongs to an unfashionable group. If institutions reject those distinctions as symbols of oppression, they weaken the tools by which error can be discovered.

A culture that cannot praise excellence without first checking the creator’s demographic category will become fearful of its own inheritance. A culture that cannot criticise its inheritance will become complacent. The proper response to history is neither worship nor demolition. It is judgment.

That judgment should be open to new voices. But new voices do not become valuable because they replace standards. They become valuable when they produce work that expands understanding, challenges a settled view or achieves something that can withstand scrutiny. Inclusion is a condition for a larger conversation. It is not the conclusion of the conversation.

The geopolitical competition does not pause for American arguments

The most practical point in Mac Donald’s address was geopolitical. The United States can debate the moral vocabulary of race for as long as it likes, but China is not waiting for the debate to resolve. A country that wants technological leadership must produce scientists, engineers, mathematicians and technicians capable of meeting difficult standards.

The source material places China first globally in K-12 mathematics, science and reading while the United States ranks around 25th. International comparisons are sensitive to the assessment used, the year selected and the populations included, so rankings should not be treated as an all-purpose measure of national intelligence. They do, however, provide a warning about educational preparation. A country cannot maintain scientific leadership while allowing large numbers of students to leave school without basic mastery.

The competition is not only between school systems. It is between institutional habits. China can direct resources toward strategic industries without first asking whether every outcome is demographically proportional. The United States, by contrast, has created a large administrative apparatus devoted to measuring representation, equity and inclusion across almost every major institution. Those goals may be morally defensible, but they consume attention and can become substitutes for the work they were meant to improve.

The danger is not that diversity officers exist. It is that a nation begins to judge an engineering department more by its demographic profile than by the reliability of its designs, a medical school more by the composition of its class than by the competence of its graduates, or a university more by its statements than by the knowledge its students acquire.

The same warning applies to South Africa. A country with unreliable electricity cannot afford to make energy expertise subordinate to political signalling. Its competitors will use the failure. Investors will move. Skilled workers will leave. Every blackout will become a tax on national ambition.

Excellence is not a decorative word in a contest between states. It is capacity: the ability to build, maintain, discover, repair and defend. A country that weakens the institutions that produce capacity may win an argument about representation and lose the future that representation was meant to improve.

The standard should be universal, and the route to it should be broad

There is a way to answer both the moral objection to exclusion and the practical objection to lowered standards. Make the route to excellence wider while keeping excellence real.

That means identifying barriers that have no connection to performance. It means ending informal patronage, expanding access to rigorous schools, providing financial support where it is needed, recruiting beyond familiar networks and ensuring that examinations measure relevant knowledge rather than inherited advantage. It means making the expectations clear before applicants enter the process, rather than changing them after results expose a gap.

It also means admitting that equal opportunity will not produce identical outcomes. People differ in interest, preparation, discipline, health, family responsibilities and aptitude. Some differences will reflect injustice; others will not. A serious institution must investigate the difference rather than assuming one explanation in advance.

For universities, transparency should be non-negotiable. Applicants and the public should know what admissions criteria mean, how they are weighted and whether the same qualities are rewarded across groups. If a college values adversity, it should measure adversity. If it values intellectual achievement, it should explain how achievement is assessed. If it uses race as a proxy for disadvantage, it should show why race is the appropriate proxy and how it avoids penalising people who do not fit the preferred narrative.

For professional regulators, the rule is simpler. Do not license people who cannot perform essential tasks. If pass rates differ, improve preparation and examine the school pipeline. Do not hide the weakness by making the test less informative. A doctor who has not mastered anatomy is not made safer by a more equitable score distribution.

For employers, representation targets should never replace job requirements. Recruitment can be broad, but selection must remain tied to the work. Promotion should reward results, judgment, teamwork and the capacity to carry responsibility. If the workforce is unrepresentative because the pipeline is too narrow, fix the pipeline. Do not pretend the final appointment can repair every earlier inequality.

This is not a rejection of fairness. It is a demand that fairness be measured at the level of the individual rather than inferred from the final racial arithmetic. The aim should be a society in which more people are genuinely qualified to enter demanding professions, not one in which the definition of qualification is adjusted whenever the composition looks wrong.

Politics must decide what institutions are for

Mac Donald’s call for Republicans and other defenders of merit to oppose disparate-impact standards and race-based allocation of resources is a political programme, not a neutral conclusion. It should be debated as such. A government can choose to prioritise proportional representation, individual assessment, historical redress, public safety or some combination of these aims. Courts and administrators should not be allowed to settle the matter through euphemism.

If voters want demographic targets, legislators should state them plainly and take responsibility for their consequences. If voters want professional standards to be universal, officials should defend them without apologising for the possibility that outcomes will differ. The present system often combines the political ambition of quotas with the legal deniability of diversity. That arrangement avoids honest argument while leaving institutions to absorb the risk.

The same applies to energy policy. South Africa’s transformation rules were not merely technical procedures; they expressed a national judgment about who should hold economic power after apartheid. That judgment had to be made. But power policy also requires an honest account of trade-offs. If a staffing rule reduces plant reliability, the public should be told. If a target requires a longer training period, the budget should reflect it. If experienced engineers must be retained while new candidates are prepared, the policy should make room for both rather than treating retention as ideological surrender.

No country can repair history by pretending that competence is irrelevant. No country can preserve competence indefinitely while telling a large part of its population that the institutions it depends on are closed to them. The difficult work is to build institutions that are open at the entrance and exacting at the point of responsibility.

That work is less satisfying than a slogan. It produces no instant equality of outcomes. It demands investment in children years before they apply to university and patient training years before they manage a power station. It requires leaders to accept that some targets will take time and that public trust is lost faster than technical expertise can be rebuilt.

But the alternative is a politics that confuses the appearance of correction with correction itself. An institution can change its workforce and still fail its mission. A university can issue statements about inclusion and still graduate students without adequate skills. A utility can satisfy a transformation formula and still leave a nation in the dark.

The test is not whether a policy sounds virtuous. The test is what happens when the lights go out, the patient is on the table, the aircraft is in the air or the bridge is under load. At that moment, a society discovers which standards it truly valued.