South Africa has just crossed one of the most important political thresholds since the end of apartheid. In the 29 May general election, the African National Congress fell to roughly 40% of the national vote, losing the parliamentary majority it had held continuously since the country’s first fully democratic election in 1994. The ANC remains comfortably the largest party, but its decline from 57.5% in 2019 to around 40% today means it can no longer govern alone and must negotiate with parties it previously defeated or marginalised. The Democratic Alliance has won roughly 22%, former president Jacob Zuma’s new uMkhonto weSizwe party around 15%, and the Economic Freedom Fighters around 9%. South Africa has therefore not rejected the ANC entirely. It has done something more institutionally significant: voters have separated the party’s historical legitimacy from its contemporary performance and ended the assumption that liberating the country confers an indefinite right to govern it.
That distinction matters because the ANC occupies an extraordinary place in South African history. It was the political movement of Nelson Mandela, Oliver Tambo and generations of anti-apartheid activists, and its 1994 victory represented much more than ordinary partisan alternation. The ANC entered government as the organisation most closely associated with the dismantling of a racial state that had denied the majority of South Africans meaningful political rights. For millions of Black South Africans, voting ANC therefore carried historical, emotional and political significance extending far beyond whether a municipal road had been repaired recently. Liberation movements accumulate what might be called historical political capital: citizens support them partly because past sacrifice establishes legitimacy that ordinary governing parties never possess. The remarkable feature of this week’s result is that this capital has depreciated enough that performance increasingly outweighs memory.
Political capital, like economic capital, does not remain valuable automatically. The ANC could reasonably claim enormous credit for helping construct South Africa’s democratic constitutional order, expanding access to housing, electricity and public services after apartheid, and presiding over the emergence of a substantial Black middle class. Yet historical achievements create a diminishing political return when citizens evaluate present government. A voter born in 2002 has no personal memory of apartheid government and experiences the ANC primarily through today’s unemployment, electricity supply, policing, corruption and municipal services. The reference point therefore shifts across generations. Older voters can compare the ANC government with apartheid. Younger voters increasingly compare the South Africa they experience with the South Africa they believe should exist three decades after apartheid ended.
This is the generational depreciation of liberation legitimacy. Governments often rely on historical narratives much longer than voters rely on historical comparisons. Political parties remember the circumstances under which they acquired authority; citizens gradually begin asking what that authority produces now. The shift does not imply ingratitude towards the struggle against apartheid. It means democratic legitimacy must be renewed through performance rather than inherited permanently through history. That is precisely what makes this result institutionally healthy. If the ANC’s role in liberation had become an argument that defeating it electorally was somehow disloyal to South Africa, the movement would have transformed historical legitimacy into political entitlement.
Instead, voters have imposed a distinction between the nation and the party that once liberated the nation. South Africa can honour Nelson Mandela without voting ANC indefinitely. This sounds obvious in a mature multiparty system, but liberation movements across southern Africa have often fused themselves politically with the states they helped create. Criticism of the governing party can then be framed as criticism of the liberation struggle itself, while opposition parties are portrayed as threatening achievements won through decades of resistance. That narrative can remain powerful long after it stops providing information about administrative competence.
The ANC’s electoral decline therefore reflects one of democracy’s most important mechanisms: retrospective accountability. Citizens cannot monitor every procurement decision, municipal contract or policy failure personally. Elections allow them to aggregate experience into a broad judgement about whether government should continue under the same terms. South African voters appear increasingly dissatisfied with what they receive in exchange for thirty years of political dominance. Official unemployment remains above 30%, poverty remains widespread, inequality is among the highest in the world, violent crime remains severe and many communities continue facing failures involving electricity, water, transport and municipal administration.
The persistence of inequality is especially politically explosive because apartheid produced one of the most unequal distributions of land, wealth, education and economic opportunity in the modern world. Democratic government inherited that structure rather than creating it from nothing. Expecting three decades to erase centuries of dispossession would be historically unrealistic. Yet the persistence of inequality cannot provide the ANC with an indefinite defence either, because thirty years is long enough for citizens to evaluate whether the direction and speed of change are satisfactory. Historical constraint explains why transformation is difficult; it cannot explain every present failure.
South Africa therefore illustrates the distinction between structural inequality and governmental performance. The enormous racial wealth gap was inherited from apartheid. Corruption in municipal procurement was not. Spatial segregation created by apartheid planning remains difficult to reverse. Repeated failures to maintain infrastructure are contemporary administrative problems. Educational inequality has historical roots. A dysfunctional local department today still has current managers and political accountability. Serious analysis needs both levels simultaneously because blaming everything on apartheid removes responsibility from current institutions, while pretending apartheid’s economic structure disappeared in 1994 removes the historical conditions under which those institutions operate.
Electricity provides perhaps the clearest example. Eskom, the state-owned electricity utility, has struggled for years with ageing infrastructure, debt, corruption and insufficient generating capacity, producing repeated periods of scheduled power cuts known as load shedding. Electricity shortages impose costs far beyond household inconvenience. Factories stop production, small businesses purchase generators, telecommunications become less reliable and investors price additional risk into projects. One failing state enterprise therefore reduces productivity across the entire economy. The result is a classic network externality: because electricity is an input into almost every other economic activity, failure inside one institution propagates across millions of firms and households.
This is why South Africa’s problem cannot be described simply as insufficient redistribution. The state may possess strong redistributive intentions while lacking enough implementation capacity to convert fiscal resources into reliable services. South Africa operates one of the developing world’s largest social-grant systems, providing important income support to millions of people. Those transfers alleviate poverty and stabilise household consumption. They cannot substitute permanently for electricity, schools, policing, infrastructure and productive employment. Redistribution can move income through an economy; development requires institutions capable of increasing the economy’s productive capacity itself.
The employment problem is particularly severe because unemployment is not simply lost income. Long periods outside formal work reduce skill accumulation, weaken professional networks and alter expectations about economic mobility. Youth unemployment therefore produces hysteresis: temporary exclusion can generate permanent effects. A twenty-two-year-old unable to enter formal employment today may possess lower lifetime earnings even if the economy improves later because several years of experience were never accumulated. South Africa’s economic challenge is therefore partly a race against time.
This makes the ANC’s electoral decline understandable even among voters who recognise its historical achievements. Democracy allows people to distinguish gratitude from contract renewal. A government can deserve credit for one period and dismissal in another. Indeed, institutionalising that distinction may be one of the clearest signs that South Africa’s post-apartheid democracy has matured. The ANC’s declining vote share has occurred through elections rather than coups, constitutional suspension or mass political violence. A liberation movement has lost its majority because citizens stopped voting for it in sufficient numbers.
This is especially important given the democratic deterioration occurring elsewhere across Africa. Military coups have returned across parts of the Sahel, while leaders in several countries have attempted constitutional manipulation to extend tenure. South Africa instead demonstrates that an enormously powerful ruling party can lose dominance while electoral institutions continue functioning. Democracy becomes most meaningful when an incumbent confronts an outcome it genuinely did not want and accepts that the arithmetic of votes now constrains it.
The next test is harder because South Africa has almost no national experience with coalition government. For thirty years, the ANC’s parliamentary majorities allowed it to govern without negotiating continuously with opposition parties. The election has changed the structure of political incentives overnight. To elect a president and sustain a government, the ANC now needs support from parties whose economic and constitutional programmes differ substantially from its own. The country is moving from dominant-party government towards bargaining government.
Coalitions are often described negatively as instability, but they perform an important democratic function: they force political parties to translate broad campaign demands into negotiable priorities. A party can promise radical land reform, extensive nationalisation or fiscal conservatism while campaigning alone. Once several parties need one another’s votes, they must decide which commitments are essential and which are negotiable. Coalition politics therefore converts ideology into prices: how much is one party willing to concede in exchange for another party’s support on the issue it values most?
The economic implications depend enormously on whom the ANC chooses. The Democratic Alliance broadly favours market-oriented reforms, stronger fiscal discipline and measures aimed at improving administrative competence and private investment. The Economic Freedom Fighters, led by Julius Malema, advocates considerably more radical redistribution and greater state ownership, including nationalisation proposals. uMkhonto weSizwe, associated with Jacob Zuma, has also advocated radical constitutional and economic changes, including stronger state control over land and parts of the economy. These are not interchangeable coalition partners. Each would pull an ANC-led government towards a different policy equilibrium.
Markets therefore will interpret coalition negotiations as economic information. Investors do not need to agree morally with one party or another before changing asset prices. They estimate probabilities around fiscal policy, property rights, regulation, infrastructure and institutional stability. A coalition perceived as increasing the probability of unpredictable nationalisation can raise the risk premium demanded on South African assets. A coalition interpreted as providing policy continuity and stronger administrative reform could lower it. Political bargaining therefore enters interest rates and exchange rates through expectations long before new legislation is enacted.
This is why institutional credibility has a price. A government does not need to confiscate property before investors respond to the possibility. Conversely, announcing reform does not produce investment unless firms believe implementation is credible. Expectations convert politics into economic behaviour. Investors delay projects, households move savings and currencies adjust according to what actors expect politicians to do next.
The ANC now faces a particularly difficult bargaining problem because the coalition partner easiest to explain economically may be among the hardest to explain politically. An arrangement with the DA could reassure parts of the business community and potentially create a centrist policy coalition. Yet the DA has historically struggled to overcome perceptions that it disproportionately represents wealthier and white constituencies, even though its support is more diverse than that description alone suggests. For sections of the ANC’s traditional base, governing with its principal historical opposition could appear like abandoning the redistributive project associated with post-apartheid transformation.
An alliance with the EFF or MK could be easier to frame through the language of Black economic transformation but much harder to reconcile with existing ANC policy and with concerns around institutional and macroeconomic stability. This creates a political-economic trilemma: the ANC wants to retain its historical identity, reassure markets and maintain enough parliamentary support to govern. Depending on coalition design, maximising any two can make the third harder.
Jacob Zuma’s return complicates the equation further. Zuma governed South Africa from 2009 until he resigned in 2018 amid extensive corruption allegations and political pressure, although he has denied wrongdoing in various cases. His new MK party has now performed extraordinarily strongly in its first national election, particularly in KwaZulu-Natal, where it displaced much of the ANC’s traditional support. The result demonstrates that the ANC did not lose voters in one ideological direction. Some migrated towards the centrist DA, while others moved towards populist and more radical alternatives. The governing party’s coalition has fragmented simultaneously to its left, right and through personalised political loyalties.
That means there may be no single policy capable of “winning back the ANC voter” because the category increasingly describes several distinct constituencies. One voter wants faster redistribution. Another wants lower corruption and more reliable services. Another remains loyal to Zuma personally. Another prioritises business confidence and economic growth. Dominant parties often appear ideologically broad because winning majorities requires containing these groups within one organisation. Electoral decline reveals the internal contradictions that dominance previously concealed.
This is coalition decomposition. The ANC historically functioned as a broad church containing social democrats, trade unionists, communists, Black nationalists, business-oriented pragmatists and regional political networks. Maintaining those factions inside one party was easier while the party reliably controlled government and distributed political opportunity. As electoral support falls, defecting becomes more attractive because former insiders can create separate parties and potentially become coalition kingmakers. The weaker the dominant party becomes, the more weakness itself encourages fragmentation.
This feedback loop can accelerate political transition. A dominant party falls below 50%. Breakaway groups discover they can gain influence independently. More politicians leave. Voters encounter more credible alternatives. The dominant party loses additional support. Eventually the political system reorganises around permanent coalition competition rather than one party with satellite opponents.
That could improve South African democracy by making elections more consequential and increasing accountability. It could also create policy instability if coalitions become fragile and governments repeatedly collapse. Local government provides warnings: coalition arrangements in cities such as Johannesburg have sometimes been unstable, with repeated changes in mayors and alliances. National coalition politics will therefore require norms and institutional discipline South African parties have not previously needed at this scale.
The central requirement is separating coalition bargaining from patronage bargaining. Coalition partners should negotiate public policy, cabinet responsibility and legislative priorities transparently. If agreements become principally exchanges of ministries, procurement opportunities or state positions, coalition government can fragment accountability further. Citizens then struggle to know which party caused a policy failure because every party blames another while sharing government resources.
Transparency therefore becomes more important under coalition government than majority government. Written coalition agreements can specify objectives, red lines and dispute mechanisms before disagreements occur. Northern European democracies have developed this infrastructure over decades because coalitions are routine. South Africa will need to learn much of it quickly.
The election also raises a broader question about whether dominant-party systems are necessarily bad for democracy. The answer is more complicated than it appears. One-party dominance can increase policy continuity, allowing governments to implement long-term infrastructure and development programmes without constant coalition renegotiation. It can also reduce accountability because voters begin believing one party will govern regardless of performance, while bureaucracies and businesses learn that access to the state flows through the same political networks indefinitely.
Competition therefore creates economic benefits beyond ideological pluralism. Governments facing a realistic probability of losing have stronger incentives to improve performance before elections. Officials know corruption scandals can transfer power to opponents. Businesses know political access may need to survive changes in government rather than depending on one party relationship. Alternation can therefore help depersonalise the state.
South Africa’s election may create exactly this effect even if the ANC remains in government. Losing its majority tells officials and voters that the party is electorally replaceable. That expectation changes incentives. A mayor, minister or provincial leader who previously assumed ANC identification provided sufficient political protection now has stronger reasons to care about performance. Opposition parties know governing responsibility is becoming attainable, which should theoretically increase their incentives to develop credible policy rather than remain permanent protest movements.
Competitive politics therefore can improve state capacity through threat of replacement. Markets use competition to discipline firms because inefficient companies lose customers. Democracies use electoral competition imperfectly to discipline governments because ineffective parties lose votes. Neither mechanism works perfectly: consumers can remain trapped by monopolies, and voters can choose according to identity rather than performance. But competition provides information and consequences that monopolistic environments lack.
The ANC’s dominant status effectively weakened that mechanism for decades. Its historical legitimacy, organisational reach and opposition fragmentation made national defeat appear remote. That did not make elections meaningless, but it reduced the expected political cost of underperformance. Today’s result changes the expected cost.
The comparison with India is interesting. The BJP currently dominates national politics but faces meaningful state-level opposition and remains a much newer national hegemon. The ANC governed continuously for thirty years partly because the apartheid transition gave it an unparalleled historical mandate. Both cases demonstrate why large electoral victories should not be treated as permanent political equilibria. Voters update, generations change and performance modifies identity.
South Africa’s younger voters are particularly important because their political identities are being formed after apartheid. This does not mean race has disappeared from politics; economic inequality remains deeply racialised because historical wealth structures persist. It means the political mechanism connecting racial identity automatically to one party can weaken over time. A young Black South African can view the ANC as the party of liberation historically while still voting against it because of unemployment today.
This is democratic normalisation. Politics moves gradually from “Which movement represents liberation?” towards “Which party can govern better?” That transition is politically healthy if it preserves memory of historical injustice without allowing memory to exempt governments from accountability.
The danger is that unresolved material inequality can push dissatisfaction towards increasingly radical political alternatives rather than producing ordinary centrist alternation. South Africa remains one of the most unequal societies in the world, and land ownership, unemployment and racial wealth gaps provide fertile ground for parties promising much faster structural redistribution. The rise of the EFF and MK reflects partly the belief among many voters that the post-1994 economic settlement changed political rights much faster than economic power.
That criticism contains an important truth. Democracy transformed legal citizenship dramatically while asset ownership and economic opportunity changed much more slowly. The question is which reforms can accelerate inclusion without damaging the productive capacity required to finance inclusion. Redistribution and growth are complements up to a point and substitutes beyond another. Redistribution can increase human capital, social stability and demand when poorer households receive resources. Poorly designed expropriation or state ownership can reduce investment and productivity if property rights or management become unpredictable. The optimal policy therefore depends on the mechanism, not the ideological label.
Land reform illustrates this complexity. Apartheid and earlier colonial policies dispossessed Black South Africans of enormous amounts of land, making redistribution a legitimate question of historical justice as well as economic policy. Yet transferring land without financing, infrastructure, agricultural knowledge and secure tenure can produce symbolic redistribution without productive transformation. The developmental objective should be increasing both equity and productive use, which requires more institutional sophistication than simply choosing between existing ownership and uncompensated seizure.
The same applies to Black Economic Empowerment. Policies designed to reverse racial exclusion from ownership and management address a genuine historical distortion. Poor implementation can concentrate benefits among politically connected elites without transforming opportunity broadly. A policy can therefore succeed at changing the racial composition of ownership among the top 1% while doing much less for unemployment among the bottom half.
This is the elite-capture risk of redistribution. If programmes transfer rents through political connections rather than building capabilities, inequality can remain high while the identities of some elites change. Citizens then become frustrated with both existing markets and the state interventions supposed to reform them.
That frustration helps explain why corruption has been so politically damaging to the ANC. Corruption is not merely theft from government budgets. In a society where the ruling party justified authority partly through transformation, corruption can appear as a betrayal of liberation itself. Money intended for housing, electricity or transport being diverted through patronage carries a different political meaning when the governing movement claims moral inheritance from anti-apartheid struggle.
Historical legitimacy therefore magnifies the reputational cost of corruption once citizens stop viewing it as isolated misconduct. The stronger the moral claim made by a political organisation, the more damaging evidence of self-enrichment becomes.
The Zuma era contributed heavily to this deterioration through the phenomenon commonly described as state capture, where private interests associated particularly with the Gupta family were accused of influencing appointments and government contracts. Extensive commissions and investigations later documented severe institutional failures. Zuma and others have disputed various allegations, and specific legal responsibility belongs to courts, but the political effect has already been enormous: citizens learned that institutions created to transform South Africa could themselves be captured by networks seeking private advantage.
President Cyril Ramaphosa entered office partly promising institutional renewal and anti-corruption reform. His government has made changes in some areas, yet the ANC’s electoral collapse suggests voters do not believe reform has proceeded fast enough or that governance failures can be attributed only to the Zuma period. Long-serving parties eventually lose the ability to compartmentalise blame across administrations because citizens experience the organisation as continuous.
The ANC now faces the hardest type of institutional reform: reforming itself while depending on itself for survival. Networks benefiting from patronage exist inside the same organisation whose leadership needs to remove them. Anti-corruption therefore creates internal losers capable of obstructing reform. External electoral competition may provide the pressure internal reform lacked, because losing votes changes the cost-benefit calculation for party elites.
This is where democracy can become a mechanism of organisational renewal rather than merely replacement. A party does not need to lose power completely before elections improve its behaviour. The threat of losing can be enough. The ANC can respond to 40% by reforming candidate selection, improving local government, reducing patronage and focusing on implementation. If it does, today’s electoral punishment may strengthen the party in the long run.
Alternatively, leaders can interpret declining support primarily as a communication problem, assume voters simply misunderstand achievements and intensify historical rhetoric. Dominant parties often choose this explanation because it is psychologically easier than admitting institutional decline. The next several years will reveal which diagnosis the ANC adopts.
The election therefore should not be interpreted as the end of the ANC. Forty percent remains an enormous national constituency, substantially larger than any competitor. Liberation history continues to matter, organisational networks remain extensive and the party retains deep support across many communities. What ended this week is not the ANC’s political relevance. It is its ability to assume relevance automatically produces unilateral governing power.
That difference is enormous.
For thirty years, South African politics asked which faction would control the ANC and therefore much of the state. It may increasingly ask which parties can construct a majority across the entire electorate. Internal ANC bargaining gives way partly to interparty bargaining. Political competition moves outward from one dominant organisation towards parliament.
That can make policymaking slower.
It can make government less coherent.
It can also make more citizens believe their votes influence actual power.
This is the basic trade-off of pluralism.
South Africa’s constitutional system was designed to survive precisely this transition. The president is elected by the National Assembly rather than directly by voters, making coalition negotiation constitutionally manageable. Parties receive parliamentary representation broadly according to proportional support. No tanks are needed to resolve the result. No constitutional improvisation is required. Politicians need to negotiate.
That institutional mundanity is worth emphasising.
A party that liberated the country has just lost its majority after thirty years.
The constitutional order remains intact.
In many political systems, loss of dominant-party power creates existential conflict because the party and state have become indistinguishable. South Africa is now testing whether enough separation survived that the ANC can become one participant in coalition politics rather than treating reduced power as a threat to the national order.
If it can, that may become one of the strongest pieces of evidence for the durability of the democracy created in 1994.
The liberation movement does not prove its democratic legacy by governing forever.
It proves it by helping create institutions capable of telling it no.
South African voters have now done that without rejecting the constitutional system the ANC helped build. They have not erased the party’s history. They have placed history back where democratic history belongs: as information about why a party deserves respect, not an exemption from judgement about whether it still deserves power.
That is the political transition occurring this week.
Apartheid ended thirty years ago.
The era of automatic ANC majorities has now ended too.
The next South African political order will be negotiated rather than inherited.
That will be messier.
It may also be more democratic.
