.png)
Submission on the Revised Electricity Pricing Policy, 2026
Government Notice No. 7852, Government Gazette No. 55257, 28 August 2026
By the
Izwe Lami Freedom Foundation (NPC)
28 October 2026
The Freedom Foundation & Izwe Lami is a policy think tank which researches and analyses government policy and positions in the national interest. It has no political affiliations or allegiances.
The energy sector is perhaps the most critical South African sector as it supports and affects every individual’s life and determines the economic prospects of the country for than any other single factor.
We have brought together several experts who provide valuable insight to energy market transformation in addition to making our internal submission from CEO Leon Louw.
Contents
Submission Details
To: The Director-General, Department of Electricity and Energy, for the attention of Mr Joseph Maraba. By email to EPP.Comments@dee.gov.za.
From: Izwe Lami Freedom Foundation (NPC),
\Reg. No. 2023/726599/08, P
BO No.930083077.
Unit 12, Valle Verde, 8 Riley Road, Bedfordview, Germiston 2007,Johannesburg, South Africa.
Telephone +27 67 401 8589.
Email info@freedomfoundation.co.za.
Contact person: Leon Louw, Chief Executive Officer.
Closing date. This submission is made within the period for comment as extended to 28 October 2026 by Government Notice No. 7948 (Government Gazette No. 55436 of 22 September 2026).
Willingness to engage. We would welcome the opportunity to make oral representations and to assist the Department in drafting alternative provisions.
Summary
The parrot problem. The Revised Electricity Pricing Policy (EPP 2026) correctly states that "a perfectly competitive market would produce efficient prices".
This is seemingly profound statement is repeated parrot-fashion endlessly in supposedly sophisticated discourse, yet it is obviously meaningless beyond being a purely hypothetical idea for teaching 101 entry-level economic theory.
There never has been, never will be and never can be “perfect competition” or a “perfectly competitive market”. All that is required for “efficient pricing” is a contestable market in which prices are determined by buyers and sellers, and influenced by existing or prospective competitors.
The EPP goes on to propose that, because South Africa's market is not "perfectly competitive", NERSA should "institute economic regulation aimed at mimicking competitive conditions" (section 2.2). “Mimicking” imaginary markets, however imperfect, is impossible. A parrot, for instance, can “mimic” speech, but it cannot hold a conversation, let alone a hypothetically “perfect” conversation.
A regulator can mimic prices, but it cannot discover them. There is no economic methodology by which it can do so. It can go through seemingly sophisticated processes which are pretence not rational calculus.
Prices are not data waiting to be discovered or calculated. They are the outcome of countless decisions by people who know vital context-specific facts that a regulator cannot know. That is the central flaw of EPP 2026, and every other concern in this submission flows from it.
No market is perfectly competitive. "Perfect competition" is a textbook abstraction that exists in no industry anywhere. To make it the test for when prices may be freed is to guarantee that they never will be. The relevant test is contestability: whether entry and exit are free. Once they are, price control is not merely unnecessary; it is harmful.
The market is being liberated. The Electricity Regulation Amendment Act, 2024, the removal of the licensing threshold for embedded generation, open network access and the emergence of an independent transmission system and market operator have made generation, trading and retailing contestable in law. Households and firms have shown, by installing rooftop and embedded generation on a large scale, that they can and will bypass any supplier whose prices are too high. That is contestability in action.
Price control in a liberated market does harm. Administered prices distort the signals that tell producers what to produce and consumers what to conserve; they discourage investment by substituting regulatory discretion for commercial certainty; and they penalise competitiveness by loading levies, standby charges and approval requirements on precisely the entrants and self-generators who discipline incumbents.
Our principal recommendations. We recommend that the Department:
confine price regulation to monopoly network services that are genuinely not contestable, and even there prefer published, non-discriminatory tariffs reviewed after the event to prices approved in advance;
free all prices for generation, wholesale trading, retailing and ancillary services, including Eskom's, from NERSA approval, with competition law as the safeguard;
fund social subsidies openly from the national Budget, targeted at poor households, rather than through hidden cross-subsidies and unavoidable levies on electricity users;
ring-fence legacy costs in a transparent, time-limited charge or move them to the fiscus, so that they do not become a permanent tax on competition;
replace compulsory standby charges on self-generators with voluntary back-up contracts priced by those who supply them;
abolish discretionary Negotiated Pricing Agreements in favour of the freedom of every user to contract on terms it negotiates; and
include a sunset clause requiring each remaining price control to lapse unless its continued necessity is demonstrated afresh within five years.
About the Izwe Lami Freedom Foundation
The Izwe Lami Freedom Foundation is a non-profit public benefit organisation that promotes economic freedom, the rule of law and limited government as the surest routes to prosperity for all South Africans, especially the poor. Its Chief Executive, Leon Louw, is the founder and former chief executive of the Free Market Foundation.
The Foundation has recently completed an Electricity Policy Review (October 2026), which examines Eskom's restructuring, transmission ownership, market liberalisation and contestability. This submission draws on that Review, among other sources.
The Central Flaw: Mimicry Is Not Competition
What EPP 2026 proposes
EPP 2026 speaks the language of markets: unbundling, cost reflectivity, transparency, non-discrimination and competition. Its substance is administrative price setting from one end of the value chain to the other. Under it:
generator prices must reflect cost of supply or a NERSA-approved power purchase agreement, and new supply is benchmarked against the "avoided system cost" of the incumbent and subject to ex ante approval (Policy Positions 9 and 10);
wholesale prices are "hedged back to the regulated pricing approach" until market platforms exist, and thereafter carry a regulated capacity charge, a legacy-recovery charge and a subsidy charge (section 5.1 and Policy Position 11);
transmission, distribution and retail prices are set to recover NERSA-determined revenue requirements calculated from NERSA-standardised regulatory accounts and NERSA-approved cost-of-supply studies (Policy Positions 1, 17, 24 to 30); and
every distributor's tariff structure is to be standardised by NERSA (Policy Position 28), and every new product, customer category and tariff change justified to it (Policy Positions 6, 26 and 29).
Even the policy's genuine improvements — ending hidden cross-subsidies, unbundling bills and stopping municipalities from recovering losses and bad debt from paying customers — are framed as further instructions to the regulator rather than as freedoms for suppliers and users.
Why a regulator cannot discover efficient prices
The knowledge problem. Friedrich Hayek showed that the knowledge on which economic decisions depend "never exists in concentrated or integrated form but solely as the dispersed bits of incomplete and frequently contradictory knowledge which all the separate individuals possess". The price system is the mechanism that communicates that knowledge. No regulator, however competent, can assemble it in a regulatory account.
Costs are not facts but choices. EPP 2026 repeatedly requires tariffs to reflect "the full and efficient cost" of an "efficient and prudent operator". But efficient cost is not observable. It can only be revealed by competition, because it depends on technologies not yet adopted, contracts not yet negotiated and risks not yet taken. A regulator auditing historic costs learns what was spent, not what should have been. Hayek called competition a "discovery procedure" precisely because it reveals what could not otherwise be known.
The fiction of the efficient operator. The "efficient and prudent operator" assumed throughout EPP 2026 is a construct of the regulator's imagination. Its costs are whatever the regulator decides they would have been. That is not cost reflectivity; it is cost assertion.
Mimicry in practice. The policy recognises that "these objectives are sometimes in conflict, and trade-offs need to be made during the tariff determination process" (section 2.1). In a market those trade-offs are made continuously by millions of buyers and sellers, each weighing price against quality, reliability and convenience. Under EPP 2026 they are made periodically by a single body, under political pressure, and imposed on everyone. The parrot repeats the words without understanding the conversation.
The wrong test
Perfect competition is nowhere. EPP 2026 justifies regulation on the ground that the industry "is currently not structured to deliver perfect competition" (section 2.2). Perfect competition — many identical firms, identical products, perfect information and no barriers to entry — is a teaching device. No market in the world satisfies it: not food, not clothing, not telecommunications. If it were the test for price freedom, every price in the economy would be regulated.
The right test. The question that matters for policy is not whether competition is "perfect" but whether markets are open: whether anyone may enter, supply, buy and exit on terms they negotiate. That is the question contestability theory asks and answers.
Contestability: Why a Liberated Market Needs No Price Control
The theory
Contestable markets. William Baumol, John Panzar and Robert Willig showed that a market with free entry and costless exit is disciplined by potential competition even where there are few, or only one, actual suppliers. The threat of entry is enough to keep prices close to cost, because any supplier who raises prices above that level invites entry that will take its customers. Baumol showed that contestable markets can deliver competitive outcomes even where they are served by a monopoly or a few firms.
Monopoly properly understood. As our Electricity Policy Review explains, a sole supplier is a harmful monopoly only if entry and competition are prohibited. The only filling station in a small town is not a "monopoly" in the sense that matters: its market is contestable, and it must price as if competitors were present, because they would be if it did not.
The paradox of the sole supplier. A sole supplier facing free entry may behave more competitively than one with existing competitors, because it has more to lose: it protects not a share of the market but the whole of it.
Consumer sovereignty. The eminent South African economist William Hutt coined the term "consumer sovereignty" to describe the reality that, in open markets, consumers rule. They "hire and fire" the largest firms every day by choosing one product or supplier over another. In a contestable market, "dominance" belongs to consumers, not suppliers.
Competition for the field. Harold Demsetz showed, in "Why Regulate Utilities?", that even where economies of scale make a single supplier efficient, competition among potential suppliers for the right to serve — or the ability of customers to switch to alternatives — can deliver competitive prices without price regulation.
The evidence in South Africa
Eskom's turnaround. Our Review observes that Eskom's recent turnaround is at least partly attributable to the threat of competition. Eskom now faces the prospect of entry into every segment of its business and is compelled to act competitively, even before significant actual entry has occurred. That is contestability working before the regulator has finished designing the market it is supposed to mimic.
The self-generation revolution. When generation licensing thresholds were removed and load-shedding made grid supply unreliable, South African households, farms, businesses and mines installed embedded generation at a scale no planner predicted. Every rooftop panel and every private plant is a vote cast against the incumbent's price and reliability. The ability of customers to bypass the grid, partly or wholly, is the most powerful check on network and supply prices that exists — far more powerful than any tariff determination.
Barriers to entry were regulatory, not natural. Our Review records that potential competitors who opened offices in South Africa in anticipation of liberalisation left when they concluded that investment was not viable. The reasons included regulated tariffs based on old, depreciated plant that were too low for new plant to compete with, the absence of a stable framework, and government-backed credit available only to Eskom. These were artefacts of policy, not features of electricity.
Transmission is less of a natural monopoly than supposed. It is commonly assumed that transmission is a "natural monopoly". Yet two or three power lines often run side by side and could have different owners; private and municipal lines, mine and industrial networks, cross-border interconnectors and merchant lines all exist. Network services may warrant a lighter form of oversight than competitive activities, but that is a reason for confining regulation to what is genuinely monopolistic, not for extending it to everything else.
The implication for pricing policy
Free entry is the regulator. Where entry is free, the market itself polices prices. Any supplier who charges too much loses customers to entrants, to self-generation, to energy efficiency or to substitutes such as gas, solar water heating and storage. Price control adds nothing to this discipline except cost, delay and error.
An incumbent's low prices are a benefit, not a problem. Our Review makes a point that EPP 2026 overlooks: should Eskom's efficiency and its accumulated, depreciated plant allow it to price low enough to make new entry uneconomic, that is a benefit to consumers and the country, not a market failure requiring correction. New entrants everywhere face the advantages properly enjoyed by established operators. Policy should not raise the incumbent's prices to make room for entrants, nor lower them below cost to suppress entry; it should let them be set by the market.
Competition law is the backstop. The Competition Act, 1998 already prohibits abuse of dominance, collusion and exclusionary conduct. Where genuine anti-competitive conduct occurs, it should be addressed after the event, on evidence, by the competition authorities, not anticipated by fixing every price in advance. We caution, however, that the Competition Commission tends to presume that a large market share is evidence of abuse; in a contestable market it is not.
How Price Control Harms a Liberated Market
It distorts markets
Shortages and surpluses. The working of the price mechanism is straightforward. If prices are fixed above market-clearing levels, there are surpluses; if below, there are shortages. South Africa has experienced both. For decades regulated electricity prices were held below the cost of replacing capacity, which suppressed investment and helped bring about load-shedding. When prices were later raised steeply to recover costs, demand fell: Eskom's sales have never regained their 2007 peak. Neither outcome was a market failure; both were failures of administered pricing.
Frozen signals in a changing system. EPP 2026 itself acknowledges that "the demand and supply dynamics in an integrated electricity system change constantly" (section 5.1.2). Its answer is to "periodically review" time-of-use definitions. A market answers continuously, hour by hour and place by place. A regulated time-of-use schedule revised every few years is a photograph of a river.
Prices that cannot move cannot clear. Without NERSA price control, Eskom and every other supplier would be free to promote balance between supply and demand through changing prices, including different prices for different users, periods, times of day and levels of demand. EPP 2026 permits "special products" only "to meet specific goals", subject to regulatory guidelines and public consultation (Policy Position 6). Innovation that must first be approved is innovation postponed.
It discourages investment
Regulatory risk is the largest risk. Investors in long-lived electricity assets need confidence that they will be able to earn a return. EPP 2026 offers an "opportunity" to earn a regulator-determined revenue that is expressly "not guaranteed", with reopeners only for industry-wide shocks (Policy Position 1). The investor bears the downside of the regulator's errors and is denied the upside of its own efficiency. That is the worst of both worlds.
Credit ratings depend on pricing freedom. EPP 2026 wants utilities to "achieve stand-alone investment-grade credit ratings" (Policy Position 1(h)). Rating agencies have repeatedly identified tariff regulation as a constraint on Eskom's credit quality; when Moody's affirmed Eskom's B2 rating on 29 May 2026, it again flagged tariff challenges, alongside municipal arrears, capital needs and the risks of unbundling. Investment-grade credit cannot be decreed by a regulator whose own decisions are among the principal risks being rated.
Ex ante approval deters entry. Requiring new supply options to be benchmarked against the incumbent's avoided cost and approved in advance (Policy Position 10) puts every entrant's business case at the mercy of an estimate of the incumbent's costs that the entrant cannot see and cannot challenge in time. Bilateral contracts between willing buyers and sellers need no approval of price.
Gold-plating and the wrong incentives. Rate-of-return and revenue-requirement regulation reward capital spending, because allowed returns are calculated on the regulatory asset base. Averch and Johnson showed that such regulation tends to induce over-investment in capital. It also weakens the incentive to cut costs, because savings are clawed back at the next determination.
It penalises competitiveness
Taxing the competitors. Several provisions of EPP 2026 load costs on precisely those who discipline incumbents: self-generators, wheeling customers and new entrants. They include compulsory standby charges on all customers, "including customers with their own generators" (Policy Position 15); the rule that no transmission or wheeling customer may "avoid contributing to approved subsidies or legacy costs" (Policy Positions 16(d) and 31(e)); a capacity charge on all wholesale customers (Policy Position 11(b)); and fixed network charges designed so that customers with rooftop solar "bear the same network costs as other customers" (section 7.7).
Exports credited below value. Net-billing customers must be on time-of-use tariffs, pay network and retail charges that are never offset against exports, and be credited only at the distributor's "avoided purchase cost", at rates set by NERSA (Policy Position 33). Every wheeling customer must contribute to NERSA-approved subsidies and municipal surcharges "irrespective of the source of energy supplied" (section 7.11), and licensees "may impose conditions on wheeling transactions" (Policy Position 34(e)). Each provision raises the cost of the alternatives that discipline incumbent distributors.
Insurance nobody bought. EPP 2026 likens the standby charge to an insurance premium. But insurance is a contract voluntarily entered into. A premium that must be paid whether or not cover is wanted is a tax. Self-generators who want back-up will buy it; those who prefer batteries, their own spare capacity or the risk of interruption should be free to choose.
Legacy costs as a barrier to entry. The costs of legacy procurement programmes, including sub-economic REIPPPP contracts, were incurred by the State. Recovering them through unavoidable levies on every unit that moves across the network makes new, cheaper supply carry the cost of old, expensive supply. That suppresses the very competition that would bring prices down.
Discretionary deals for the favoured. EPP 2026 retains Negotiated Pricing Agreements, defined as incentive pricing so that "priority and strategic industrial sectors can operate competitively, retain capacity, and attract new investment". Under section 10.4 and Policy Position 54 they are to be approved "only if it is in the national interest", must be supported by the Department of Trade, Industry and Competition, and any revenue shortfall "may be recovered through the applicable regulatory process" — that is, from other customers. Special prices for politically selected firms are paid for by everyone else, invite lobbying, and are inconsistent with the policy's own commitment to non-discrimination (Policy Position 4). The non-discriminatory alternative is simple: let every user negotiate its own supply contract.
It invites capture and politicisation
The economics of regulation. George Stigler showed that regulation is acquired by the regulated industry and designed and operated primarily for its benefit; Sam Peltzman extended the analysis to the political interests of regulators. Price control concentrates enormous value in a single decision, and therefore attracts lobbying by incumbents, organised users, municipalities and the State itself.
The South African record. Our Review documents how Eskom was repeatedly subjected to political interference contrary to what its management and Board might have preferred: it was barred from building new capacity in 2001, burdened with municipal debt and social obligations, and constrained by price determinations. Its recovery has come with greater commercial freedom, not with more detailed price control.
It is costly and slow
The administrative burden. EPP 2026 requires NERSA, typically within 12 months, to adopt an asset-valuation methodology, an investment prudency framework, standardised regulatory accounts, a national tariff framework, a cost-of-supply standard, a wheeling and congestion methodology, a national subsidy framework and a national distribution charging framework, and to review the tariff structures of every licensed distributor, including more than 150 municipal distributors. These are years of work for a regulator whose existing determinations are frequently challenged in court. Every month spent designing administered prices is a month in which market prices could already have been doing the job.
Delay is not neutral. The companion Market Transformation Position Paper envisages only a "soft launch" of the wholesale market in 2026, with participants beyond Eskom admitted gradually from 2027. Until the market is open, EPP 2026 hedges wholesale prices back to regulation. The longer the transition, the more entrenched the regulatory apparatus becomes and the more firmly the transitional arrangements harden into permanent ones.
Price Control Reaches Even Where Competition Already Works
Prepaid vendors. Policy Position 45 is the clearest illustration of the policy's instinct. It acknowledges that the sector of prepaid electricity vendors is "unregulated", then proposes that NERSA within 12 months "define allowable service fees and commission structures to prevent excessive margins", with monitoring, investigations and penalties. No evidence of excessive margins is offered — only "the potential" for them. Vending is a textbook contestable market: thousands of retailers, spaza shops, banks, cellphone networks and apps compete to sell tokens, entry costs little, and customers switch at the click of a button. Capping commissions in such a market will not lower prices; it will shrink the vending network, above all in townships and rural areas where the cost of serving customers is highest.
Resellers. Policy Position 44 requires NERSA to set "electricity resale pricing principles" for landlords, body corporates and other resellers. These are contractual relationships already governed by leases, sectional title rules and consumer law. Tenants and owners who are dissatisfied can negotiate, move, or install their own supply.
Price signals by permission. Policy Position 35 requires any price signal "over and above recovering the costs" to be "itemised and motivated specifically and be approved by NERSA". But a price is not a signal designed by an authority to steer behaviour; it is the outcome of supply and demand, and it conveys information precisely because no one designs it. Requiring permission for prices that depart from regulated cost is requiring permission to respond to scarcity.
Compulsory tariff designs. Policy Positions 36 to 38 make time-of-use tariffs compulsory for whole classes of customer within fixed periods, prescribe a three-tier national structure for residential tariffs, and require NERSA-driven phased increases in tariffs at lower voltages. Time-of-use pricing can be valuable, but whether it suits a particular customer is a matter for supplier and customer to agree. A supplier that offers tariffs customers prefer will win them.
Regulator as investor. Policy Positions 48 and 49 have NERSA decide "the amount of funds to be allocated" to demand-side flexibility and to utility-owned embedded generation, including microgrids and rooftop solar. A regulator deciding how much utilities may invest in competing with private installers is not regulating a monopoly; it is directing a competitive market.
What We Support
Several provisions point in the right direction and should be retained: no special tariffs for organs of state, which must budget and pay full cost (Policy Position 46); public lighting charged to municipalities as consumers rather than to electricity customers (Policy Position 40); exclusion of customer-funded assets from the regulated asset base, so that customers do not pay twice (Policy Position 39); the refusal to let excessive losses and bad debt be passed to paying customers (Policy Positions 42 and 43); the prohibition on adding a transparent municipal surcharge on top of hidden surpluses (Policy Position 47); itemised, transparent bills (Policy Position 3); and the requirement that subsidies be disclosed and capped (Policy Position 50). Each of these brings prices closer to cost and makes subsidies visible. The logical next step is to let competition, rather than the regulator, discover what cost is.
Applying Contestability Segment by Segment
Contestability theory asks of each activity not how many suppliers there are but whether entry and exit are free. Applied to the value chain addressed by EPP 2026:
Activity | Is it contestable? | Implication for price regulation |
Generation | Yes. Licensing thresholds have been removed; private, embedded and imported generation compete, and customers can self-generate. | No price regulation. Bilateral and market prices; competition law as backstop. |
Wholesale trading and retail supply | Yes, once access to networks is open: traders, resellers and self-supply all contest the incumbent's customers. | Free prices for all suppliers, including Eskom and municipalities. |
Ancillary services, storage and demand response | Yes. Generators, batteries, customers and aggregators can all supply them. | Competitive procurement by the system operator, not regulated prices. |
Prepaid vending and reselling | Highly contestable, with very low entry costs. | No price regulation; consumer law suffices. |
Embedded generation, microgrids and rooftop solar | Yes. Private installers compete with utilities. | No regulator allocation of funds; utilities compete on equal terms. |
Distribution networks | Partly. Self-generation, storage, microgrids, private and embedded networks and bypass contest network charges. | Published, non-discriminatory tariffs; monitoring and review after the event; no prohibition of bypass. |
Transmission | Partly. Parallel, private, merchant and independent lines are possible, and the Department itself plans independent transmission projects. | Lightest effective oversight: open access, published tariffs and cost disclosure. |
The pattern is clear. Most of what EPP 2026 proposes to regulate is contestable now. What remains only partly contestable becomes more so every year, as storage and distributed generation spread. A pricing policy for the next decade should be designed for that market, not for the vertically integrated monopoly of the last century.
Experience Elsewhere
Regulation as a holding operation. Stephen Littlechild, architect of British utility regulation, described price caps as a means of "holding the fort" until competition arrives, and argued that regulation should be withdrawn wherever competition becomes effective. EPP 2026 builds the fort but makes no plan to leave it.
Wholesale markets work without administered prices. Wholesale electricity markets in many jurisdictions, including Texas, the Australian National Electricity Market, New Zealand and much of Europe, set energy prices through competitive bidding rather than regulatory determination. Many also allow retail choice. They have their difficulties, but they show that electricity is not uniquely incapable of being priced by markets.
Failures were failures of price control and design. The California crisis of 2000 to 2001 is often cited against liberalisation. It arose in large part because retail prices were frozen by regulation while wholesale prices were exposed to the market, so that distributors could not pass on costs and consumers had no reason to reduce demand. The Texas winter blackout of 2021 is attributed mainly to the failure of unweatherised generating plant and gas supply in extreme cold. Neither failure resulted from too little price control, and neither is a failure of the principle that free prices clear markets.
Success depends on conditions, not controls. Our Review finds that countries succeed or fail mainly according to four conditions: cost-reflective prices, competent and independent oversight, sufficient investment in the grid and political stability. Reforms fail when prices are held too low, oversight is weak and governments interfere.
The Case Against Government Intervention in Pricing
The general arguments
Knowledge. Those who set prices administratively lack the knowledge that market prices embody and transmit. The more complex and fast-changing the system — and an electricity system with millions of prosumers, storage, variable renewables and cross-border trade is about as complex as systems get — the greater the gap between what the regulator knows and what it would need to know.
Incentives. Market participants who misprice lose their own money; regulators who misprice do not. Price setters with no stake in the outcome have weaker incentives to be right than those who bear the consequences of being wrong.
Calculation. Ludwig von Mises showed that without market prices for capital goods, rational calculation of costs and returns becomes impossible. Administered electricity prices feed into the investment decisions of the entire economy; if they are wrong, every downstream calculation is wrong too.
Capture and rent seeking. Wherever government determines prices, effort shifts from serving customers to influencing the price setter. The beneficiaries of each subsidy and special deal are concentrated and organised; those who pay for them are dispersed and often unaware.
Unintended consequences. Every intervention changes behaviour in ways its designers did not foresee. Price ceilings create shortages, cross-subsidies create bypass, and levies on wheeling create incentives to leave the grid altogether — which raises costs for those who remain.
Intervention begets intervention. Mises also observed that each intervention produces problems that seem to call for further intervention. EPP 2026 illustrates the pattern: cross-subsidies cause bypass, so self-generators must pay standby charges; standby charges deter self-generation, so special products must be designed; special products need approval, so a quick-turnaround approval framework is required (Policy Position 6(c)). The alternative is to remove the first intervention.
Certainty and the rule of law. Investors and consumers need stable, general rules, not discretionary determinations that change with each regulatory cycle. Freedom to contract within clear general law offers more certainty than any multi-year price path, however carefully published.
Freedom of contract and trade. Every citizen has the right to choose their trade, occupation or profession freely (section 22 of the Constitution), and the common law protects freedom of contract. A buyer and seller of electricity who agree a price harm no one. Requiring regulatory approval of that price restricts freedoms that should not be limited without compelling justification.
Answering the objections
"Networks are natural monopolies." Even if parts of the transmission and distribution networks are monopolies, that is no reason to regulate the prices of generation, trading, retailing and ancillary services, which are contestable. Where network regulation is retained, it should take the lightest effective form: open access, published non-discriminatory tariffs, price monitoring and review after the event, rather than ex ante approval of every charge. And the scope of the monopoly should be tested, not assumed: private, municipal, mine and merchant lines, embedded networks and off-grid supply all contest parts of the network business.
"The poor need protection." They do — but price control is a poor way to provide it. Below-cost prices for everyone help the rich more than the poor, because the rich consume more electricity. Hidden cross-subsidies are paid in part by poor households themselves, through the prices of goods and services produced with electricity. The better instrument is a transparent, targeted subsidy, such as Free Basic Electricity or a cash or token transfer funded from the Budget, so that its cost is visible and can be weighed against other spending. Our Review notes that the Homelight 20A tariff carries a subsidy of about 43 per cent of allowed costs, paid by other customers through an affordability subsidy charge. Such transfers should be on the Budget, not in the tariff.
"Incumbents will abuse market power during the transition." The remedy for market power is entry. The Electricity Regulation Amendment Act opens the market; competition law punishes abuse; and the ability of customers to self-generate caps what any supplier can charge. Price control during the transition delays entry, because it signals that prices will continue to be decided politically.
"Security of supply requires central planning." Security of supply is a service like any other. Users who value it can contract for firm supply, back-up or storage; those who do not can accept interruptible supply at lower prices. Central planning of capacity gave South Africa both the surplus capacity of the 1980s and the shortages of the 2000s and 2010s.
"Prices must be predictable." Markets produce predictability through contracts: long-term power purchase agreements, hedges and fixed-price retail offers, freely chosen. A five-year regulated price path is a forecast with the force of law, and therefore a commitment the State may be unable to keep.
"Legacy contracts must be paid for." They must. But they are obligations of the State that procured them. They should be ring-fenced, disclosed and recovered through a transparent, time-limited charge or funded from the fiscus, not embedded permanently in every network and wholesale tariff where they penalise competition.
Comments on Specific Policy Positions
The table below sets out our principal concerns with specific provisions and our recommended alternatives.
Provision | Concern | Recommendation |
Section 2.2 and Policy Position 1 (revenue requirement) | Makes administrative revenue determination, "mimicking competitive conditions", the basis of all regulated pricing; NERSA to define assets, prudency and returns. | Confine revenue regulation to non-contestable network services. Replace "mimicking competitive conditions" with a commitment to remove price regulation wherever entry is free. |
Policy Position 2 (cost reflectivity in five years) | Cost reflectivity determined by the regulator is not cost reflectivity; "efficient cost" cannot be observed. | Achieve cost reflectivity by freeing competitive prices immediately and removing cross-subsidies from network tariffs within the five-year period. |
Policy Position 4 (non-discrimination) | Undermined by the expansion of Negotiated Pricing Agreements and unavoidable subsidy levies. | Retain, and apply it to the State itself: no special prices for favoured sectors; all users free to contract. |
Policy Position 5 (network access) | Welcome open access, but "economic bypass" is not prohibited by economics: bypass happens because prices exceed the cost of alternatives. | Retain open access; delete the prohibition of economic bypass; let network owners compete to retain users. |
Policy Positions 6 and 7 (special products; price path) | New products require guidelines and consultation; a five-year regulated price path entrenches regulation. | Allow suppliers to offer any product to willing buyers without approval; publish network price forecasts only for regulated network services. |
Policy Position 8 (scope) | Brings all licensed and registered generators and importers within the EPP. | Exclude from price regulation all generators, traders and importers selling under bilateral contracts or into a competitive market. |
Policy Positions 9 and 10 (generator pricing) | Prices tied to cost of supply, NERSA-approved PPAs and the incumbent's avoided cost; ex ante approval of PPAs. | Delete ex ante price approval for generation sold bilaterally or into the market; limit approval to procurement by organs of state. |
Policy Position 11 (wholesale structure) | Regulated capacity, legacy and subsidy charges on all wholesale customers. | Let the market price energy and capacity; ring-fence legacy costs in a disclosed, time-limited charge; fund subsidies from the Budget. |
Policy Position 12 (Central Purchasing Agency) | The CPA may become a permanent single buyer under another name. | Limit the CPA to administering existing legacy PPAs, with no power to procure new capacity, and a sunset date. |
Policy Position 15 (standby charges) | Compulsory standby charges on self-generators penalise the most important source of contestability. | Replace with voluntary back-up contracts priced by suppliers; the system operator may procure reserves competitively. |
Policy Positions 16(d), 19 and 31(e) (transmission and wheeling) | No transmission or wheeling customer may avoid subsidies or legacy costs; fixed 50/50 split. | Remove social subsidies and legacy costs from network charges; let network charges recover network costs only, with locational signals based on cost causation. |
Policy Positions 23 to 30 (distribution) | Uniform national tariff structures, standardised by NERSA, for all distributors; fixed charges designed so that rooftop solar customers pay the same network costs. | Require transparency and ring-fencing, not uniformity; allow distributors to design tariffs to retain customers; let competition, including self-generation, discipline their prices. |
Policy Position 33 (net billing) | Exports credited only at "avoided purchase cost" on NERSA-set rates; compulsory time-of-use; charges never offset. | Let distributors and customers agree export prices; publish network charges; allow third parties to buy exported energy. |
Policy Positions 35 to 38 (price signals and tariff design) | Price signals above cost need NERSA approval; compulsory time-of-use and prescribed residential tiers. | Delete the approval requirement for price signals; allow suppliers to offer any tariff design customers accept. |
Policy Positions 44 and 45 (resellers and vendors) | Price control of reseller pricing and of vendors' fees and commissions in contestable markets, without evidence of abuse. | Delete both; rely on contract and consumer law and competition among vendors. |
Policy Positions 48 and 49 (demand-side and utility-owned embedded generation) | The regulator decides how much utilities may spend competing with private providers. | Remove regulator allocation of funds; utilities compete on equal terms, outside the regulated asset base. |
Policy Position 54 (Negotiated Pricing Agreements) | Discretionary special pricing "in the national interest", supported by the dtic, with shortfalls recovered from other customers, is discriminatory and invites lobbying. | Abolish NPAs; let all large users contract directly with suppliers of their choice on published network terms. |
Policy Positions 50 to 53 (subsidies) | Subsidies carried in tariffs and levies are hidden and poorly targeted. | Fund Free Basic Electricity and other social support from the Budget, targeted at poor households and disclosed annually. |
Policy Position 47 and section 7.11 (municipal surcharge) | Surcharges on electricity make electricity a tax base for municipalities. | Require municipal surcharges to be disclosed as taxes on bills, and phase them out in favour of general municipal revenue sources. |
A Liberated Pricing Framework
Principles. We propose that the final policy rest on the following principles in place of administrative price setting:
freedom to price: suppliers of generation, trading, retail and ancillary services, including Eskom, set their own prices, subject only to general law;
freedom to contract: any user may buy from any supplier on any terms they agree, including bilateral, wheeled and embedded supply, without approval of price;
open access: network owners must offer access to all users on published, non-discriminatory terms;
network charges for network costs only: network tariffs recover the cost of networks, not social subsidies, legacy costs or municipal revenue;
light-handed network oversight: published tariffs, monitoring and review after the event, with ex ante approval only where a network is shown to be an uncontested monopoly;
subsidies on the Budget: social support funded transparently by the fiscus and targeted at poor households;
legacy costs ring-fenced: disclosed, time-limited and declining, or moved to the fiscus;
competition law as the safeguard: abuse addressed on evidence, after the event, with proper regard to contestability;
Eskom liberated: Eskom freed from NERSA price control for its competitive activities, and from municipal debt, non-commercial obligations and political interference, so that it can compete on equal terms; and
sunset: every remaining price control to lapse after five years unless its continued necessity is demonstrated afresh in a public process.
The role of NERSA. Under this framework NERSA's role changes from price setter to referee: licensing where necessary, enforcing open access, monitoring network tariffs, publishing information and resolving disputes. That is a role it can perform well. Setting thousands of prices for an industry undergoing the most rapid technological change in its history is not.
Conclusion
EPP 2026 rightly recognises that competitive markets produce efficient prices and that South Africa's electricity industry is becoming competitive. Its error is to conclude that, until competition is "perfect", a regulator should imitate it. A parrot can mimic speech but cannot hold a conversation, and the more fluent its mimicry, the easier it is to mistake for the real thing. In a liberated, contestable market the real thing is available. The policy should let it speak: free prices where entry is free, regulate only what is genuinely monopolistic, fund social goals openly, and set a date by which the remaining controls must justify themselves or lapse.
We thank the Department for the opportunity to comment and would be pleased to assist further.
Sources
Averch, H. and Johnson, L.L., "Behavior of the Firm Under Regulatory Constraint", American Economic Review 52(5), 1962, pp. 1052–1069.
Baumol, W.J., "Contestable Markets: An Uprising in the Theory of Industry Structure", American Economic Review 72(1), 1982, pp. 1–15.
Baumol, W.J., Panzar, J.C. and Willig, R.D., Contestable Markets and the Theory of Industry Structure, Harcourt Brace Jovanovich, 1982.
Constitution of the Republic of South Africa, 1996, section 22.
Competition Act 89 of 1998.
Demsetz, H., "Why Regulate Utilities?", Journal of Law and Economics 11(1), 1968, pp. 55–65.
Department of Electricity and Energy, Revised Electricity Pricing Policy, 2026, Government Notice No. 7852, Government Gazette No. 55257, 28 August 2026.
Department of Electricity and Energy, Extension of the Period for the Submission of Public Comments on the Electricity Sector Market Transformation Position Paper, 2026 and the Revised Electricity Pricing Policy, 2026, Government Notice No. 7948, Government Gazette No. 55436, 22 September 2026.
Electricity Regulation Act 4 of 2006, and Electricity Regulation Amendment Act 38 of 2024.
Hayek, F.A., "The Use of Knowledge in Society", American Economic Review 35(4), 1945, pp. 519–530.
Hayek, F.A., "Competition as a Discovery Procedure" (1968), translated in Quarterly Journal of Austrian Economics 5(3), 2002, pp. 9–23.
Hutt, W.H., Economists and the Public: A Study of Competition and Opinion, Jonathan Cape, 1936.
Izwe Lami Freedom Foundation, Electricity Policy Review, October 2026.
Joskow, P.L. and Schmalensee, R., Markets for Power: An Analysis of Electric Utility Deregulation, MIT Press, 1983.
Littlechild, S.C., Regulation of British Telecommunications' Profitability, Department of Industry, London, 1983.
Mises, L. von, "Economic Calculation in the Socialist Commonwealth" (1920), in Hayek, F.A. (ed.), Collectivist Economic Planning, Routledge, 1935.
Mises, L. von, Middle-of-the-Road Policy Leads to Socialism, 1950 (reprinted in Planning for Freedom, Libertarian Press, 1952).
Peltzman, S., "Toward a More General Theory of Regulation", Journal of Law and Economics 19(2), 1976, pp. 211–240.
Stigler, G.J., "The Theory of Economic Regulation", Bell Journal of Economics and Management Science 2(1), 1971, pp. 3–21.
Comments