Unbundling power transmission and distribution: whither the incremental power distribution network reform in China?

Abstract

China adopts an incremental approach to unbundling power transmission and distribution, allowing non-state actors to fund and operate power distribution networks that are additional to existing grid-affiliated networks. However, this approach is faltering, which could compromise China’s ‘dual carbon’ agenda, considering how integral a competitive power distribution market is for accommodating the rising distributed renewables. This article analyses the market design flaws and market power abuses contributing to this approach’s faltering state. As the market design is being corrected through energy policymaking, this article discusses whether and how the market power abuses could be addressed through antitrust enforcement.

Introduction

As the world’s largest energy-consuming and greenhouse gas-emitting country, China has recently adopted the ‘dual carbon’ goals to peak carbon emissions before 2030 and achieve carbon neutrality by 2060. This entails the formidable task of decarbonizing its electricity industry.¹ Meanwhile, China faces increasing difficulties in balancing power supply and demand, due to the diminishing coal-fired generation capacity and the rising extreme weather conditions, which spike up peak loads while limiting renewables generation. In response, policymakers have decided to expedite the protracted electricity marketization process, believing that better market design and greater competition are the keys to balancing the mandates of power decarbonization and grid stability most economically.²

Marketization necessitates fostering and preserving competition. This has been a tricky task vis-à-vis the Chinese power sector. First, there is the classic argument that more competition is not always desirable or feasible in the power sector, considering its natural monopoly tendency and close association with national strategic interests. This argument is reflected in the legal stance that no competitive concerns could normally exist vis-à-vis transmission and distribution grids,³ and it has been championed by certain reform policymakers.⁴ Secondly, China’s gradualist approach to electricity marketization,⁵ although with the merits of experimentalism, has created room for incumbent market players to amass market power and quasi-regulatory power to the detriment of further marketization.

Chinese policymakers are undoubtedly doubling down on power sector marketization to decarbonize the power sector and reduce power shortages in the long term. The ensuing question is how to carry forward the marketization reform, particularly the unbundling of power transmission and distribution and the promotion of competition in power distribution, a mission that the Chinese policymakers have outlined a long time ago but have been struggling to accomplish. On this question, recent discussions have been focusing on how policymakers and legislators should fix the long-overdue structural issues by adopting newer and more concrete rules. These discussions are certainly welcome, but they tend to overlook how existing legal institutions, particularly antitrust enforcement, could make a key contribution to marketization as well. Neglecting antitrust could increase the administrative costs and risk of regulatory capture relating to the marketization reform; it could also create unnecessary tension between power sector regulation and competition law, so an investigation, as attempted by this article, is desirable. Besides, in some other parts of the world, unbundling in the power sector usually focuses on separating networks (transmission and distribution) from generation and supply rather than the separation of transmission and distribution,⁶ so an analysis of how China goes about the latter would be of comparative value to regulators and businesses in other jurisdictions, especially now that energy transition has placed an ever more important role on distribution networks.⁷

This article examines China’s marketization reform vis-à-vis power distribution networks and how it could be carried forward by antitrust enforcement. Specifically, it investigates—through literature review, policy analysis, and legal analysis—why competition in power distribution is desirable in the ‘dual carbon’ context, what the competitive restraints are, and how competition law could be instrumentalized to tackle these restraints. The section ‘China’s electricity marketization and the liberalization of power distribution’ sets the stage by describing the conception and progress of the incremental power distribution network (IPDN) reform, highlighting the reform’s faltering state and its new-found importance under the ‘dual carbon’ goals. The section ‘The faltering incremental power distribution network reform’ analyses the market design flaws and market power abuses contributing to this approach’s faltering state. As the market design flaws are being dealt with by policymakers, the section ‘Addressing market power abuses through antitrust’ discusses whether and how competition law could be invoked to address the market power abuse problem. The final section concludes.

China’s electricity marketization and the liberalization of power distribution

China has been marketizing the electricity industry since 1985.⁸ Before 1985, the electricity industry was organized as a production function and managed directly by the government; then, structural changes inevitably took place as China began transitioning to a more market-oriented economy in 1978. The initial changes included the introduction of non-utility generators to boost generation capacity (1985–1996) and the establishment of the State Power Corporation (SPC) to separate enterprise from government (1997–2001).⁹ On that basis, two rounds of marketization took place.

The first round of marketization and the shelved unbundling mission

The first round of marketization began soon after the SPC was established. It sought to dismantle the SPC’s vertically integrated monopoly and inject competition into power generation and transportation. In February 2002, the State Council issued the Power System Reform Plan (‘Document No 5’) to guide the first round of power system restructuring. This Document outlined three unbundling missions: (i) separating power generation from grid operation, (ii) separating auxiliary businesses from grid operation, and (iii) separating power distribution from transmission.¹⁰

Measures to accomplish the first two missions materialized swiftly. On 29 December 2002, the State Council restructured the SPC into two grid operators—the State Grid Corporation of China (SGCC) and China Southern Power Grid (CSG), five generation group companies (‘Big Five’), and four auxiliary service corporations.¹¹ The CSG monopolizes grid operation in five southern provinces: Guangdong, Guangxi, Yunnan, Guizhou, and Hainan, while the SGCC monopolizes grid operation in the remaining parts of China.¹² In retrospect, the spinning-off of power generation was fairly successful: the Big-Five structure has remained over the years, with smaller and renewable power generators gradually catching up.¹³ Less successful was the separation of auxiliary businesses, as the four spun-off auxiliary corporations re-concentrated into two conglomerates (Power Construction Corporation of China and China Energy Engineering Corporation) in 2011, and the 2002-established SGCC had retained and acquired non-core business assets of more than $614 billion.¹⁴ The bloating of the SGCC has seriously obscured the true grid operation costs to the detriment of further unbundling plans, forcing the government to impose more stringent divestiture requirements in recent years.¹⁵

Most regrettably, reform measures for separating transmission and distribution never materialized. The main reason was the obscurity of grid costs. The endgame of separating transmission and distribution is to foster competition on the power supply side and prevent transmission grid companies from leveraging their natural monopolistic positions into the power distribution and sales markets. In a liberalization context, where vertically integrated state-owned utilities used to monopolize the entire power-supply chain, this endgame entails strictly limiting the business scope of grid companies to power transmission on one hand, while diversifying the channels for users to transact (directly or indirectly) with power generators on the other. On both accounts, it is necessary to delineate the grid companies’ power transmission costs so that, post-liberalization, the scope-limited grid companies can still operate profitably under government-controlled pricing schemes, while undistorted price signals guide transactions and spur competition in electricity wholesaling and retailing. Unfortunately, the incomplete divestiture of auxiliary business and the out-of-control expansion of the SGCC post-2002 have made the delineation of grid operation costs highly difficult.

The resistance from the SGCC also played a part in shelving the third unbundling mission. Under the size-driven (rather than efficiency-driven) performance review criteria set by the State-owned Assets Supervision and Administration Commission (SASAC) of the State Council, the SGCC had the incentive to disobey the auxiliary-grid separation mandate and expand its scale and scope.¹⁶ Meanwhile, thanks to the unclear power price structure and the fact that power price levels were heavily regulated, the SGCC had the possibility to do so as well. By alleging that government price control—instead of grid operation efficiency—had limited its core business profitability, the SGCC was able to use the revenue from the core business to expand its size instead of improving its core business profitability.¹⁷ As the SGCC became increasingly bloated, the State Electricity Regulatory Commission (SERC), which was the sector regulator from 2003 to 2013—fell short in wielding countervailing powers, ultimately bringing the transmission–distribution unbundling mission to a halt.

The second round of marketization and the IPDN reform

In March 2015, the Chinese Communist Party Central Committee and the State Council issued the Opinions on Further Reforming the Electric Power System (‘Document No 9’), which unleashed a new round of marketization reforms. Under the banner of ‘letting go of both ends and holding the middle’, this Document sought to liberalize power generation and sales while keeping power transportation under regulatory control.¹⁸ To that end, the National Development and Reform Commission (NDRC) and the National Energy Administration (NEA)—the sector regulator since 2013—adopted a series of policies addressing various structural issues relating to power generation and sales.¹⁹

Curbing the vertical expansion of transmission grid companies is integral to liberalizing power sales. So, the NDRC and the NEA launched the IPDN reform to open the investment and operation of power distribution to non-state actors.²⁰ Before the reform, power distribution networks were built and owned exclusively by the provincial subsidiaries of the SGCC and the CSG in their respective business territories, and both the SGCC and the CSG are themselves centrally state-owned enterprises (SOEs). However, after the reform, local SOEs and private companies may also build and claim ownership of power distribution networks. The word ‘incremental’ denotes that the reform does not seek to unbundle distribution networks that are already integrated with ‘grid companies’²¹; rather, it focuses on diversifying the ownership of distribution networks that are not yet constructed or those that are constructed but do not belong to grid companies. This reform has two purposes. The first one is to alleviate the shortage of distribution networks relative to the country’s rising level of electrification and the increasing generation of decentralized renewables.²² The second one is to facilitate the calculation of grid fees and bring competitive pressure to the transmission–distribution integrated grid companies. Admittedly, power distribution is naturally monopoly-prone, and grid companies are not required to spin off their distribution networks, but IPDNs can serve as yardsticks against which provincial grid companies’ costs of operating distribution networks become more measurable.

The IPDN reform is essentially a toned-down gradualist approach to the unfinished transmission–distribution unbundling mission outlined in Document No 5.²³ Unfortunately, even with this toned-down approach, there has been little success after 7 years of implementation. The section ‘The faltering IPDN reform’ expands on this. Before that, let us take a look at how the IPDN reform makes a key contribution to accomplishing the ‘dual carbon’ goals and therefore must not be abandoned.

A crucial role for distribution networks in the new electric power system

In pursuit of the ‘dual carbon’ goals, central policymakers have proposed to build a new electric power system that is low-carbon, secure, efficient, flexible, and smart.²⁴ The backdrop to this proposal is that the share of variable renewable power sources—particularly distributed solar photovoltaic—is rapidly rising thanks to the ‘dual carbon’ goals and technological cost reductions²⁵; but because these renewable sources are intermittent, wildly fluctuant, and seasonally uneven, grid balancing is becoming increasingly difficult in the traditionally centralized power system, so a more decentralized and renewables-based system is needed.

Building this new power system entails the integrative development of system flexibility resources relating to generation, demand response, storage, and (transmission and distribution) network operation, whereby power supply and demand can be balanced more locally and dynamically.²⁶ But in practice, the development of system flexibility resources is lagging behind, consequently dragging down the development of renewable sources.²⁷ Various factors contribute to the underdevelopment of system flexibility resources. For example, on the generation side, the most readily available flexibility resources in China remain coal-fired power and hydropower, but the former is to be reduced for decarbonization, while the latter is geographically and socio-economically limited to scale up; in terms of storage-related flexibility resources, pumped hydro is similarly limited as hydropower and electrochemical technologies are still being commercialized.²⁸ Meanwhile, demand-side flexibility resources are lacking because the power sales market itself is underdeveloped, with cross-subsidization excluding a large portion of power users outside of the market.²⁹

As to network-related flexibility resources, power distribution networks deserve the most attention, not only because there is an economy of scope between power distribution and sales (ie, demand-side flexibility), but also because the rise of distributed renewable sources will inevitably make grid systems more decentralized. The traditional grid system is largely centralized and one-sided, where transmission grids transport bulk generation from power plants directly to large industrial users or through distribution networks to residential and smaller industrial and commercial users. But with the rising uptake of distributed renewables (eg, rooftop solar) and power storage facilities (eg, electric vehicles) on the user side, electricity consumers are becoming prosumers, who are less dependent on transmission grids for power supply and may instead have additional power to feed back into the grid. This puts distribution networks in a unique position to operate localized ‘generation-grids-load-storage’ ecosystems and assist these prosumers in managing variable demand and engaging in peer-to-peer generation trading, thereby helping the entire grid system achieve balance.

The NDRC and the NEA have recognized the importance of distribution networks in promoting distributed renewable energy and have vowed to making distribution networks smarter and more active.³⁰ The SGCC and the CSG have followed the policy direction and are investing heavily in reinforcing and smartening their distribution networks. This is desirable but insufficient for preparing distribution networks for the rising distributed renewables. Improving operational efficiency and activeness (eg, provision of ancillary services) is also necessary, and it should be the priority as it can help defer the need for huge infrastructure investments.³¹ The consequential question is how the duopolistic grid companies can be incentivized to improve their distribution network efficiency and activeness. This is where the IPDN reform could help, as it subjects the duopolistic grid companies to comparison, and, in turn, competitive pressure.

The faltering IPDN reform

As the section ‘China’s electricity marketization and the liberalization of power distribution’ shows, by injecting competition into the power distribution business, the IPDN reform serves not only as a gradualist approach to the unfinished transmission–distribution unbundling mission but also as a crucial way to prepare China’s power system for the rising distributed renewable sources. In practice, however, this reform is faltering.

An overview

The NDRC and the NEA started rolling out IPDN pilot projects in November 2016. Until December 2023, they have approved five batches, totalling 459 pilot projects. After approving the last batch in August 2020, they delegated the approving responsibility to provincial authorities to normalize the establishment of IPDN projects. Since then, there has been no systematic information disclosure, and anecdotal evidence suggests that there have been few projects approved thereafter.

An approved pilot project needs to complete four steps before it can begin business operations: finishing planning, confirming ownership, establishing a power supply scope, and obtaining a power supply license.³² By December 2023, 329 of the 459 pilot projects had finished planning, 359 had confirmed ownership, 256 had established a power supply scope, and 227 had obtained licenses. These 227 licensed projects consist of 77, 44, 57, 34, and 15 projects from the first, second, third, fourth, and fifth batches, respectively. Time wise, it took 74 of the 227 projects 2 to 3 years to obtain a license, while 43 of them took more than 3 years.³³ The whole process is not necessarily tedious, but without rules laying down a clear timetable, it puts many approved-but-not-yet-licensed pilot projects in limbo, with investors unable to recoup costs and users unable to benefit from more competitive services.³⁴ Besides, after a project has begun operation, it may still struggle to make a profit because complementing rules are lacking and provincial grid companies engage in competition-foreclosing practices. This is explained in the next two subsections.

Inadequate market design

The NDRC and the NEA have designed the market for IPDN projects through several policy and legal documents (see Table 1). However, until recently, this design was inadequate vis-à-vis two issues that are vital for these projects to gain footholds: profit margin and capacity tariffs.

The first issue concerns pricing. Although it is legitimate to subject the natural monopoly-prone IPDNs to price control, the applicable rules did so in a way that was too crude to set a reasonable profit margin. According to the 2017 Guiding Opinions (see Table 1), the price users pay for the power distribution service provided by an IPDN at a certain voltage degree, plus the price they pay for transmission services provided by the provincial grid (of the SGCC or the CSG) at the higher voltage degree, must not exceed the price users pay when they get power at the same distribution voltage degree directly from the provincial grid. Provincial grids have six voltage degrees in total: 500 kV (700 kV), 220 kV (330 kV), 110 kV (66 kV), 35 kV, 10 kV (20 kV), and below 1 kV; the lower the voltage degree, the higher the price.³⁵ So, for example, when a user gets power at 10 kV from an IPDN, they pay the IPDN operator no more than what the provincial grid charges the same category of users for delivering power at 10 kV; then, under the 2017 Guiding Opinions, the IPDN operator must settle payments with the provincial grid by transferring to the latter an amount equal to the transmission service price charged to users at the next higher voltage degree, which is 35 kV. In other words, when an IPDN distributes power to a user at a certain voltage degree, the rules equate—rather simplistically—the distribution service price to the spread between the provincial grids’ rates for transporting power to the same category of users at the same voltage degree and the next higher degree. The problem here is that cross-subsidization has made the spread too narrow to reflect the true cost of power distribution: Residential and agricultural users are the recipients of cross-subsidies and are located mostly at lower voltage degrees. This means that commercial and industrial users at the same voltage degrees bear the brunt of the subsidization burden, which, in turn, translates to higher-than-usual prices for these lower-voltage commercial and industrial users. Consequently, the spread is narrower than it should be and so is IPDNs’ profit margin.³⁶ Admittedly, the 2017 Guiding Opinions provides an exception to the rule,³⁷ but this exception has never been invoked due to its lack of specification.

This price-setting method effectively enables provincial grids to treat IPDNs as ‘large-scale industrial users’,³⁸ even though IPDNs are officially recognized as part of the public grid system despite their private ownership status.³⁹ This leads to the second issue of capacity tariffs. Namely, questions arise as to (i) whether—and if so, how much—IPDNs should pay fixed capacity tariffs to provincial grids, as actual large-scale industrial users normally do and (ii) how IPDNs should share with provincial grids the capacity tariffs that they have collected from large-scale industrial users. National rules do not answer these questions. From the perspective of IPDNs, it seems unreasonable to pay a monthly fixed capacity fee when they have already been charged an industrial-user-level running fee,⁴⁰ but the 2017 Guiding Opinions do not explicitly make an exception for IPDNs when enabling grids to treat them as industrial users; instead, the issue is left to each provincial government to address. Until recently, there had been no national rules providing an unequivocal answer to the second question either, meaning IPDNs had to negotiate with provincial grids for their tariff-sharing arrangement on a case-by-case basis.⁴¹

Fortunately, central policymakers have started addressing these issues. The 2023 Notice (see Table 1) merges the two marketized user categories—‘large-scale industrial users’ and ‘general commercial and industrial users’—into a single ‘industrial and commercial users’ category. It also introduces a new two-part tariff scheme for charging these users transmission and distribution fees. A two-part tariff consists of a fixed charge based on the maximum demand or transformer capacity and a running charge based on the amount of electricity transacted. In the new scheme, the fixed charge is negatively correlated with a user’s voltage degree. This differs from the old scheme, where the fixed charge for a user category remained the same across different voltage degrees. Consequently, IPDNs gain a new revenue stream. To illustrate, suppose an IPDN at 110 kV distributes power to a two-part tariff user at 35 kV. In the old scheme, the operator’s revenue mainly lies in the running charge, namely the spread between the electricity prices corresponding to these two voltage degrees (measured in ¥/kWh), multiplied by the amount of electricity distributed (measured in kWh), plus a percentage of the monthly fixed charges that the IPDN may or may not have agreed upon with the provincial grid. But in the new scheme, the operator’s revenue would consist not only of the running charge but also of the spread between the fixed charges relating to these two voltage degrees (measured in ¥/kWm or ¥/kVAm), multiplied by the maximum demand or transformer capacity (measured in kW or kVA).

As the market design improves, there remains the problem of market power abuse by provincial grid companies. Now that the rules have given IPDNs a wider profit margin, it is even more likely that grid companies will perceive IPDNs as competitive threats and engage in exclusionary practices.

Three exclusionary practices by provincial grid companies

Provincial grid companies are transmission–distribution integrated regional monopolies owned directly by the SGCC or the CSG. IPDNs are competitors to these grid companies not only in providing power distribution services but also because these incremental networks make power distribution efficiency more measurable, thereby constraining the grid companies’ possibility to extract monopoly rents. Provincial grid companies are therefore incentivized to exclude IPDNs through three types of practices.

The first practice is refusals to connect. While putting competitive pressure on provincial grid companies, IPDNs are also downstream undertakings that rely heavily, if not entirely, on connection with provincial transmission grids for input. This allows a provincial grid company to foreclose an IPDN by refusing to connect the latter’s substations to the grid under ostensibly valid excuses. For example, a grid company may claim that an IPDN project has too low a capacity utilization rate to justify any additional grid connections.⁴² Consequently, an IPDN project could wait for years to get a grid connection, even after obtaining a power supply license. Additionally, a grid company may foreclose an IPDN during its license-obtaining process by refusing to sell assets that fall within the IPDN’s service scope.⁴³

Admittedly, these refusals may be based on legitimate concerns relating to grid security and efficiency, but they are prone to abuse as there is no independent third-party regulator to check their necessity and proportionality. The NEA has approached this problem by adopting rules prohibiting provincial grid companies from refusing to connect with IPDNs without justification.⁴⁴ However, the NEA does not provide details on how to apply this prohibition in individual cases. To make it worse, provincial governments hesitate to confront the SGCC and the CSG because they rely heavily on the grid companies for power supply and local infrastructure development. They prefer to mediate between the provincial grids and IPDNs towards reaching a compromise, which often comes at the IPDNs’ expense. This is what happened with the Zhengzhou Airport IPDN project in Henan province: due to the provincial grid’s refusal to provide a connection, the project could not operationalize a newly constructed substation and supply a highly coveted customer; this eventually forced the project’s shareholders—primarily a local SOE—to transfer the substation and the customer to the provincial grid company.⁴⁵

To be sure, IPDNs may not need to connect with provincial grids if they could find enough power sources elsewhere. For example, an IPDN in Guizhou province successfully reduced dependency on the provincial grid by partnering up with an independent regional grid to share hydropower sources.⁴⁶ This is a promising solution with the rising distributed renewables, but only to a limited extent. First, there are policy restrictions. While being encouraged to connect with nearby distributed renewables, IPDNs are also prohibited from (i) using coal-fired power units to establish dedicated user lines or microgrids; (ii) using off-grid power plants to establish incremental distribution network; (iii) transforming in any way on-grid power plants into off-grid ones.⁴⁷ These prohibitions are meant to limit coal-fired generation and maintain grid stability, but the vague wording makes them subject to expansive interpretations. For example, a question may arise as to whether the second prohibition also applies to self-owned renewable power plants. Secondly, a provincial grid company may jeopardize an IPDN’s plan to accommodate distributed renewables by abusing its grid-management roles and refusing to cooperate with the IPDN’s linking with a renewable generation plant. Grid companies are motivated to do so because more self-sufficient power users mean less revenue in the short term and less relevance of the transmission grids in the long term.⁴⁸ Overall, connection with provincial grids remains vital for most IPDNs for the time being.

The second exclusionary practice is acquisitions. First, the rules state that the ownership of an IPDN project should be established through a public bidding process in which all interested parties may contend to build and own the project, with grid companies and local SOEs being normally discouraged from owning controlling shares of IPDN projects.⁴⁹ This, however, fails to deter provincial grid companies from trying to acquire a controlling interest in an IPDN project when it is still at the ownership-confirming stage. This was attempted on a first-batch pilot project in Chongqing and another one in Henan, forcing the NDRC to step in and condemn the predatory behaviour.⁵⁰ Secondly, provincial grids may try to take over an already operational IPDN, as exemplified by the aforementioned Zhengzhou Airport project: after acquiring a substation and several large clients from the project, in March 2023, the provincial grid company (SGCC Henan) got the provincial government’s endorsement and reached an agreement with the shareholders to absorb the entire IPDN project.⁵¹ SGCC Henan could push forward the acquisition so aggressively because it used power infrastructure investment as leverage over the provincial and municipal governments.⁵² The local governments were captivated by the vast investment that SGCC promised to reciprocate, overlooking the long-term welfare loss associated with the exit of a competitive IPDN project.

The third practice is prohibiting IPDNs with captive power plants from feeding power back into the grid. When providing connection to these IPDNs, provincial subsidiaries of the SGCC would only allow power transportation to these IPDNs; they would install monitoring devices at the connection points and cut off the power supply if they detect reverse power transportation.⁵³ Consequently, IPDNs are reduced to being downstream distributors to the provincial grids when they could also be upstream power suppliers. This pattern of behaviour reflects the SGCC’s inertia to centralize and monopolize power supply when the marketization reform needs it to shift towards a ‘platform’ between generators and users. Admittedly, the SGCC may be prohibiting reverse power flow for valid system reliability reasons,⁵⁴ but, like the excuses for refusing connection, these reasons are susceptible to abuse. This will likely become a bigger issue as IPDNs accommodate more distributed renewable sources.

In sum, as central policymakers improve the market design vis-à-vis IPDNs, there remains the problem of competition foreclosure by provincial grid companies. This problem is somewhat structural: by creating a dual-track (ie, independent versus integrated) regime for distribution networks without effective safeguarding mechanisms, the IPDN reform subjects transmission–distribution integrated grid companies to competition on the one hand and enables them to abuse their monopolistic and grid-management positions on the other. In that light, a more fundamental change in approach to the transmission–distribution unbundling mission is needed. However, making this change requires significant time for building consensus and planning details, and the fact that key sets of rules have expired (see Table 1) for years without replacements suggests that more time may be needed still. It is therefore worth considering using the existing competition law to protect IPDNs from being unduly excluded, thereby ‘buying time’ for central policymakers to figure out how to advance the unbundling mission.

Addressing market power abuses through antitrust

This section discusses how the Chinese Anti-Monopoly Law (AML)⁵⁵ could be enlisted to address the problem of provincial grid companies foreclosing IPDNs. It first examines the applicability of the AML vis-à-vis provincial grid companies. Then, relying on relevant case precedents and antitrust theories of harm, it offers an analytical framework for assessing provincial grids’ exclusionary practices.

Applicability of the AML

Stakeholders in IPDNs have exposed the questionable refusal-to-connect practices of provincial grid companies to the NDRC, which has promised to heighten grid-access regulation but has not taken any concrete measures.⁵⁶ This inaction is in part due to a lack of legal grounds. First, as the product of an experimentalism-spirited reform, IPDNs are recognized only in ‘departmental rules’ (listed in Table 1) and not in any of the higher-ranking ‘administrative regulations’ or ‘laws’ (including the primary, 2018-amended Electric Power Law).⁵⁷ Secondly, although Article 36 of the 2021 Administrative Measures for Supervising Fair and Open Power Grids prescribes fines for grid companies that unjustifiably refuse grid access, it references Article 31 of the Regulation on Electric Power Supervision (an administrative regulation adopted in 2005) for the calculation of fines. The maximum amount is only ¥1 ($0.14) million, which seems under-deterrent to the SGCC and the CSG. The upshot is that sector regulation, in its current form, appears too feeble to lend legalistic support to IPDNs in defending their competitive interests.

Antitrust enforcement appears to be a more promising alternative. First, it is theoretically sound and legally grounded to apply the AML to provincial grid companies on suspicions of market power abuses. A literal reading of Article 8 of the AML suggests that, although provincial subsidiaries of the SGCC and the CSG are entitled to retain their monopolistic positions, they are obligated not to abuse those positions. This reflects a behaviourist approach to competition preservation vis-à-vis natural monopolies, according to which a monopoly-prone market structure does not preclude the need for antitrust outright because there may still be competition worth protecting around the monopoly, so antitrust authorities should be vigilant about incumbents’ behaviour, particularly behaviour leveraging market power from the monopoly to adjacent competitive markets. Secondly, it is practically feasible, albeit unprecedented, to apply the AML to provincial grid companies. The AML is enforced through civil litigations and administrative actions. No provincial grid company has been the defendant in AML civil litigations, but provincial subsidiaries of other central SOEs in the energy sector have.⁵⁸ As to public enforcement of the AML, multiple provincial subsidiaries of the SGCC have been investigated by provincial antitrust authorities, although these investigations were eventually terminated through commitment procedures.⁵⁹ The possibility of subjecting provincial grid companies to both private and public antitrust enforcement becomes even more real against the backdrop that over the years, provincial governments have been gaining authority and influence in managing the power sector within their respective jurisdictions.⁶⁰ Ultimately, it comes down to whether provincial governments are sufficiently incentivized to defend IPDN developments.

To be sure, market power abuses of provincial grid companies could be tackled through sector-specific regulatory rules (eg, access regulation and non-discrimination obligations⁶¹) instead of generally applicable antitrust rules. Ideally, sector regulation and antitrust enforcement would complement each other towards the goal of competition protection, with their roles adjusted dynamically as the liberalization reform deepens.⁶² In the case of the IPDNs, however, because power sector regulation in its current form is too feeble to tackle the abuses in question, the AML should be relied on more heavily, at least until more competent regulatory rules are devised. To that end, antitrust and sector regulatory agencies should utilize their respective strengths and establish coordination mechanisms for case referral, investigation, and remedy implementation in administrative actions launched under the AML.⁶³

Aside from antitrust enforcement and sector regulation, state ownership policy is an ostensibly useful but ultimately problematic approach to tackling market power abuses of provincial grid companies. Namely, it may seem like a straightforward solution to ask the SASAC to impose a code of competitive conduct on the SGCC and the CSG, since they are central SOEs under the SASAC’s supervision, but it is doubtful whether the SASAC would be incentivized to do that when increased competition inevitably limits (at least in the short term) the profitability of the SOEs and inherently contradicts the SASAC’s primary mandate of establishing state champions ‘to maximize a range of benefits extending from state revenues to technological prowess’.⁶⁴ In fact, it is a major concern of the AML that central and local administrations might distort market competition by affording unfair advantages to SOEs through state ownership policy, so the AML subjects individual administrative decision-making and general administrative ruling-making to abuse-of-administrative-power regulation (Chapter 5 of the AML) and fair competition review (Article 5 of the AML), respectively.

Analytical frameworks for assessing the exclusionary practices

Exclusionary practices in the form of acquisitions are relatively easy to tackle. First, there are no rules exempting the SGCC and the CSG from merger review under the AML; the State Administration for Market Regulation (SAMR), which is the national merger control authority, had already fined a provincial subsidiary of the CSG in 2021 for concentrating with other undertakings without prior notification to the authority. The SAMR did not require the parties to undo the merger, but only because it found no anticompetitive effects.⁶⁵ Therefore, when provincial grid companies acquire IPDNs, they are inevitably subjecting themselves to antitrust scrutiny and potential AML liabilities. Secondly, Article 33 of the AML unequivocally empowers the SAMR to take a comprehensive account of the social welfare effects resulting from an acquisition. So, much as a grid company may bring forward well-crafted justifications on how acquiring an IPDN would enhance provincial grid operation efficiency and stability, a more decarbonization- and marketization-minded merger control authority can always look further into how the IDPN in question could play a crucial role in accommodating decentralized renewables and how it exerts competitive pressure on the provincial grid company. The applicability of the AML and the possibility of being substantively scrutinized should suffice to deter provincial grid companies from engaging in predatory acquisitions.

Refusals to provide grid access and prohibitions of reverse power flow (ie, refusals to transport power sourced from IPDNs) can both be dealt with under Article 22(3) AML, which recognizes refusals to deal as a standalone type of abuse. By January 2024, the Supreme People’s Court, the ultimate judicial authority on interpreting the AML, had issued seven judgments that substantively assessed the legality of refusal-to-deal conduct under the AML.⁶⁶ A four-question analytical framework can be drawn from these assessments: (1) whether a deal was feasible in the case circumstances; (2) whether the dominant firm implicitly or outright refused to deal; (3) whether effects of competition restriction were caused; and (4) whether the refusal had any justifications.

Question (1) boils down to whether there were objective obstacles to prevent a deal from happening. Examples of such obstacles include government-imposed pandemic-control measures⁶⁷ and the lack of government-specified prices for a regulated product.⁶⁸ On question (2), the Supreme Court considers not only the facts proving the dominant firm’s refusal of trading conditions proposed by the counterparty but also the inferable motives behind the refusal. To be clear, anticompetitive motive is neither a sufficient nor necessary condition for establishing liability under competition law, but it is usually strong circumstantial evidence of (actual or potential) anticompetitive effects.⁶⁹ On question (3), presumptive effects of competition restriction would suffice, and to find such effects, the Supreme Court relies heavily on ascertaining whether the refused party was from a market adjacent to the dominated one and whether the dominant firm’s service/product was indispensable for the refused party’s operation. Notably, the dominant firm’s vertical integration into the refused party’s market is not a necessary condition for finding anticompetitive effects, as the dominant firm could be excluding the refused party simply to deter other uncooperative trading parties. On question (4), the Supreme Court examines the validity of the specific excuses advanced by the dominant firm based on proven facts, customary practices, and relevant regulatory requirements.

Nuancing the refusals to deal by grid companies

The above-mentioned four questions will need more nuanced examinations when applied to refusals to deal by grid companies. For example, when investigating the motive of a refusal under question (2), it may be useful to follow the single monopoly profit theorem and ask how it would be profitable for a transmission grid company to monopolize a downstream distribution market.⁷⁰ Here, we need to consider not only the market conditions affecting the profitability (ie, whether, in light of the circumstances, the loss of revenue in the transmission service market resulting from the exclusion of trading parties would be more than compensated by the monopoly rents gained in the distribution service market) but also a non-market factor: the SGCC and the CSG are inherently driven to expand because they are central SOEs subject to the SASAC’s size-based performance review.⁷¹

When assessing potential justifications under question (4), it is advisable to look closely at the regulatory obligations of grid companies to provide fair and non-discriminatory interconnection.⁷² These obligations understandably entail the limiting and conditioning of IPDNs’ grid access, and how that is done would be crucial for finding antitrust liabilities. The following two points deserve special attention.

First, provincial grid companies may claim that refusing an IPDN’s connection request is necessary to prevent system inefficiency, but short-term efficiency reductions in the current system could just be the price for the long-term welfare benefits promised by a liberalized and competitive power.⁷³ So, unless the ex ante regulatory rules laying down grid access obligations are comprehensive and specific enough to constrain grid companies’ arbitrary deviation from providing fair and non-discriminatory grid access, the possibility of grid companies justifying their refusal (or discriminatory) actions towards IPDNs in ex post antitrust cases should be strictly limited. For example, the grid companies should be required to substantiate the objective and technically and economically justified reasons for their refusal or discriminatory actions, as well as the proportionality of such actions⁷⁴; otherwise, they must show in light of the circumstances that there was no abusive refusal or discrimination in the first place.⁷⁵ That being said, liberalization could be a mitigating factor in determining the extent of antitrust liability after establishing the illegality of the conduct. This is because liberalization creates a situation of change in which a monopolistic grid company may not be fully aware of its obligations under competition law.⁷⁶ So, antitrust agencies and courts should focus less on prescribing hefty fines or damages and more on devising behavioural remedies that commit grid companies to building a fairer and more transparent third-party access regime. Admittedly, the AML does not give antitrust agencies and courts any ‘hard’ power to impose behavioural or structural remedies,⁷⁷ but they have the ‘soft’ power to nudge the accused undertakings into making behavioural or even structural commitments because it is standard enforcement practice to consider the undertakings’ pledges to correct wrongdoing and remedy competitive harms as a mitigating factor in determining the amount of fines or damages. In the case of grid companies being found violating the AML, it may be too politically sensitive and legally questionable for antitrust agencies and courts to adopt structural remedies (ie, horizontal break-ups and vertical unbundling),⁷⁸ but behavioural remedies remain feasible.

Secondly, seeing that decarbonization is a key policy drive for protecting competition in power distribution, a provincial grid company may argue that its pursuit of decarbonization goals (as mandated by the SASAC) inevitably entails competition restriction in power distribution in certain circumstances. This is essentially a clash of decarbonization-minded ownership policy⁷⁹ with decarbonization-minded market regulation. In this case, the SAMR would be forced to weigh decarbonization and competition protection against each other, and it should start with examining whether the allegedly inevitable restriction of competition is based on or acknowledged by sector regulatory rules of the NDRC and the NEA. If not, the argument should be invalidated; if yes, a more in-depth assessment should be conducted to determine whether the restriction of competition is proportionate with the gain in decarbonization.⁸⁰

Conclusions

The IPDN reform allows private entities to fund and operate power distribution networks that are incremental to existing, grid-integrated distribution networks. It serves as a gradualist approach to unbundling power transmission and distribution, but it has been struggling due to poor market design and market power abuse. Fortunately, China’s decarbonization policy agenda is breathing new life into the IPDN reform, as competitive power distribution networks are crucial for accommodating the rising distributed renewables.

As the market design flaws are being dealt with by policymakers, more attention should be paid to the market power abuses hindering the development of IPDNs. Three types of abusive practices are identifiable: refusal to connect, acquisitions, and prohibition of reverse power flow. To be sure, these abuses are somewhat enabled by the dual-track regime that the IPDN reform has created, so eradicating these abuses requires a more forceful approach to the unbundling of power transmission and distribution. But before that happens, antitrust enforcement is the best hope for protecting competition in power distribution and decarbonizing the power sector.

I thank Lea Diestelmeier and the participants of the ASCOLA 2024 Conference for their valuable feedback. This research is supported by the Fundamental Research Funds for the Central Universities [Grant ZK1105] and the Fujian Provincial Federation of Social Sciences [Grant FJ2024C035]. I have no conflicts of interests to disclose.

Footnotes