Review of Retail Payments Regulation – Conclusions Paper
October 2021 2. Introduction

3.1 Issues for the Review

Debit cards are now the most frequently used payment method in Australia. Around 90 per cent of debit cards issued in Australia are dual-network debit cards (DNDCs), which allow domestic payments to be processed via either eftpos or one of the international debit schemes (Debit Mastercard or Visa Debit). Prior to widespread use of contactless (‘tap-and-go’) technology, consumers would insert their DNDC card into the merchant's payment terminal and then select the network to process the transaction. Many cardholders selected the typically lower-cost eftpos network by pushing the CHQ or SAV button, and merchants could steer the customer to choose the cheaper network (through surcharges, for example). But with the advent of contactless technology, transactions using DNDCs defaulted to the international networks, resulting in an increase in merchant costs. In recent years, therefore, the Bank has been encouraging financial institutions to provide merchants with ‘least-cost routing’ (LCR) or ‘merchant-choice routing’ functionality on contactless payments. LCR gives merchants the ability to route contactless DNDC transactions via whichever of the two networks on the card costs them less to accept. This can help merchants reduce their payment costs and increase competitive pressure between the debit schemes, incentivising the schemes to lower the fees that are ultimately incurred by merchants. The Board has strongly supported the issuance of DNDCs and the provision of LCR because of this contribution to efficiency and competition in the payments system.

As LCR functionality has been rolled out, schemes have responded to the increase in competitive pressure by lowering their fees. The weighted-average interchange rates for Visa and Mastercard debit since late 2019 have been comfortably below the 8 cents benchmark, and international scheme fees on some routable transactions have also fallen by 40 per cent since mid-2019. This has translated into a reduction in the average cost of accepting debit card transactions through the international schemes for merchants over the past couple of years (see Graph 1 above). At the same time, however, there have been increases in interchange fees and scheme fees for some non-routable debit transactions, such as those made using mobile wallets, which are making up a growing share of total debit transactions.

Given the benefits to date from LCR, a key focus of this Review was whether policy action was warranted to promote the availability and wider take-up of this functionality. Following pressure from the Bank, most acquirers had implemented some form of LCR functionality by mid 2019. However, there remain some key differences in the LCR capabilities offered by different acquirers, with most not yet offering a version that maximises merchant savings by enabling ‘dynamic’ routing for each individual transaction. Furthermore, take-up among merchants remains relatively low, which may reflect a lack of awareness or understanding of the potential benefits, along with a lack of promotion by large acquirers. One major bank has automatically switched on LCR for eligible small merchants where it determined that they would benefit from the functionality. Another two major banks have implemented single-rate plans for smaller merchants with LCR implemented in the background. However, for many merchants the onus remains on them to understand the benefits of LCR and request it from their acquirer.

In addition, there are a number of emerging challenges to the viability of LCR. First, technological changes have driven a significant shift away from the use of physical (plastic) cards at the point-of-sale to the use of new ‘form factors’, such as mobile wallets, which may increase the pool of transactions that cannot be routed. For mobile wallets, LCR is currently not possible, because each network is separately provisioned and the wallet presents the credentials of only one network during payment; this network is typically the international debit network, which is set as the default, but it can be overridden by the cardholder. Nevertheless, DNDCs can still facilitate competition between schemes in the mobile context, as merchants may be able to attempt to incentivise the customer to choose a particular network in their mobile wallet during the checkout process. However, not all mobile wallets and issuers currently support the provisioning of both networks of a DNDC; in some cases, only the international scheme is provisioned.

A second challenge to LCR is that several smaller and mid-sized issuers have begun moving away from DNDCs towards single-network debit cards (SNDCs) which allow payments to be processed through only one (international) debit network. The switch to SNDCs reflects two factors. First, the international schemes have been keen to facilitate the issuance SNDCs for some time and at least one scheme is offering higher interchange rates on transactions on SNDCs. In making the case for issuance of SNDCs, the international schemes have noted that some issuers still have single-network, eftpos-only ‘proprietary’ cards on issue (which may also attract higher interchange rates than equivalent transactions on DNDCs); LCR is not feasible on either of these single-network cards. Second, issuers and international schemes pointed during the consultation process to the additional cost of issuing debit cards with two networks instead of one. Given the largely overlapping functionality provided by the three debit schemes, some smaller issuers felt that supporting a second debit network yielded little benefit to their customers but generated significant costs.

SNDCs reduce both customer and merchant choice, and so lessen competition between schemes at the point-of-sale. A particular concern is that a shift towards SNDC issuance could have the effect of making LCR less attractive, especially for large merchants that benefit from lower ‘strategic’ interchange rates if they send significant volumes through a particular scheme. When larger ‘strategic’ merchants adopt LCR and most of their DNDC transactions are routed via eftpos rather than an international scheme, they lose access to strategic interchange rates from that international scheme on any debit card transactions that continue to be processed through that international network; the latter transactions would include transactions on DNDCs where the customer actively selects the international network or where routing is not possible because they are online or due to some problem with the chip or the issuer, as well as transactions on SNDCs. An increase in the prevalence of (international scheme) SNDCs would increase the pool of non-routable transactions that must be processed through the international schemes, while decreasing the pool of routable DNDC transactions. This would raise the cost of losing strategic interchange rates – lowering the net savings from LCR – to the point where LCR might not be commercially attractive for large merchants that could benefit from strategic rates.

Stakeholders have highlighted that the ongoing shift towards mobile payments, noted above, is increasing the pool of non-routable transactions that are automatically processed through the international schemes. Accordingly, the financial case for large merchants to use LCR has already become marginal, with the Bank aware of two large retailers that were early adopters of LCR having recently decided to stop using it. Smaller merchants, which do not have access to strategic rates, might continue to benefit from LCR (albeit to a lesser extent) even with a shift occurring to SNDCs. However, if ePAL^[6] cannot compete for the volume of large merchants, its ability to compete for smaller merchants would also be weakened. In the extreme, as the lowest-cost network, its potential exit from the market would result in a significant lessening of competitive pressure in the debit market and would likely result in an increase in both interchange rates and scheme fees, impacting all merchants.

Online payments functionality is currently being enabled for eftpos, which raises the possibility of LCR in the online (or ‘device-not-present’) environment. Indeed, the Bank is aware of several payments service providers already offering LCR online. The Board supports the provision of LCR online, given the clear benefits that LCR has had in the card-present or ‘device-present’ environment, in terms of stronger competition and lower payment costs.^[7] However, the major banks, which are both the largest acquirers and the largest issuers in Australia, may not have strong incentives to provide LCR online. Further, the online payment process is distinct from the device-present environment, which raises additional policy questions. A key issue is whether customers should be notified when merchants choose to route online transactions and whether customers should be given a choice to override merchants' routing decisions. Some stakeholders have expressed concerns about the comparability of the debit schemes' online payment offerings, particularly in regard to security, and stressed the importance of customer choice and notification. Indeed, one of the international schemes has already imposed rules relating to customer notification and choice. However, some other stakeholders are concerned about the frictions that these rules would introduce into the checkout process, in part due to customers' poor understanding of payments, which could significantly deter the development and/or merchant take-up of LCR online.

Another challenge to the viability of LCR is the potential for the international schemes to link strategic interchange rates on credit card transactions to the value or volume of merchants' debit card transactions (or their decision to adopt LCR). Such ‘tying conduct’ penalises merchants that route debit card transactions to eftpos through higher interchange rates on their credit transactions, which could offset merchants' savings from LCR. In effect, the international schemes could leverage their market power in the credit card market to dis-incentivise the take-up of LCR. In early 2018, the Bank sought and received assurances from the international schemes that they would not respond to LCR in ways that would limit the competitive pressure in the debit card market. Despite these assurances, several merchants have alleged that both Visa and Mastercard have engaged in potentially anti-competitive tying conduct, which the Board is particularly concerned about. The ACCC has investigated Visa's conduct, due to its concerns that Visa may have limited competition by engaging in tying conduct, resulting in the ACCC accepting a court-enforceable undertaking from Visa in March 2021.^[8]

3.2 Options presented in consultation

3.2.1 Least-cost routing

In relation to the provision of LCR functionality, the Board requested stakeholder views on three policy options:

Option 1: Maintain current arrangements

The Bank would continue to monitor market developments in the provision of LCR across all relevant payment channels without any formal intervention.

Option 2: Explicit guidance on the provision of LCR by acquirers and payment facilitators

The Bank would state an explicit expectation that all acquirers and payment facilitators would offer and promote LCR functionality to merchants in the device-present environment; acquirers and payment facilitators would be expected to report to the Bank every 6 months on their LCR offerings and on merchant take-up. There would be no similar expectation regarding LCR in the online environment at this stage. However, the Board would set out a list of principles that it expects the industry to follow, to prevent the erection of barriers to the development and adoption of LCR online. If expectations for the provision of LCR are not met, the Board would consider formal regulation.

Option 3: Explicit regulation on the provision of LCR by acquirers and payment facilitators

The Bank would require – through a change to the Bank's standards – that relevant payments service providers offer or support LCR for both device-present and online DNDC payments. The Bank would also set explicit rules for LCR in the online environment to ensure that the interests of merchants and consumers are appropriately balanced.

For all 3 options, the Board also considered whether the Bank's information-gathering powers under section 26 of the PSRA should be used to require schemes to notify the Bank of all scheme rules and any changes to those rules (this would overlap with a similar proposal regarding scheme fee-related rules, discussed in the section on ‘Scheme fees’ below).

3.2.2 Dual-network debit card issuance

Given recent industry developments and the issues discussed above, the Board requested stakeholder views on three broad options in relation to the issuance of DNDCs, which represent an escalating degree of regulatory response. The Board in particular called for feedback on Options 2 and 3, because its preliminary view was that the relative merits of these two options was finely balanced:

Option 1: Maintain current arrangements

Issuers would continue to make the choice between issuing DNDCs and SNDCs based on their own commercial considerations. The Bank would continue to monitor market developments without any formal regulatory intervention.

Option 2: Explicit expectation of DNDC issuance for the major banks

The Bank would set an explicit expectation that the major banks would continue to issue DNDCs, with two card schemes to be provisioned in all form factors, including mobile wallets, offered by the issuer (where the functionality is supported by the scheme). There would be no presumption as to which two debit networks were included by issuers; various combinations of domestic and international schemes might be feasible. The Bank would also set a cap on any cents-based interchange fees that was lower for SNDC transactions than for DNDC transactions, which would lessen the incentive for SNDC issuance.

Option 3: Regulation mandating DNDC issuance for the major banks and medium-sized issuers

The Bank would require – through a change to Standard No. 2 of 2016 – that all issuers above a certain size threshold must issue only DNDCs, with two card schemes to be provisioned in all form factors, including mobile wallets, offered by the issuer (where the functionality is supported by the scheme). In designing the mandate, the Bank could draw on similar rules relating to DNDCs in other jurisdictions, such as the United States.^[9] Under Option 3, there may not be a case for a lower cap on cents-based interchange fees for SNDC transactions, depending on where the issuance size threshold were set.

3.2.3 Potential tying conduct by the international schemes

The Board requested views on two options to address the potential for international schemes to link strategic interchange rates on credit card transactions to merchants' value or volume of debit card transactions (‘tying conduct’).

Option 1: Leave the ACCC to investigate and take enforcement action against any anti-competitive tying conduct

Consistent with current practice, any alleged anti-competitive tying conduct would be investigated by the ACCC under the Competition and Consumer Act 2010 (CCA).

Option 2: Explicitly address tying conduct

The Bank would seek voluntary undertakings from the designated card schemes that they will not engage in tying conduct; if this was not feasible, it would introduce a new standard to explicitly prohibit such conduct by designated card schemes.

3.3 Stakeholder views

Many stakeholders expressed support for the continued issuance of DNDCs to some degree. The Government, in particular, through a letter from the Treasurer, strongly encouraged the Board to consider mandating DNDC issuance for major and medium-sized financial institutions. Most merchant groups and some other stakeholders went further, arguing that all issuers should be required to support DNDCs in all form factors to maximise the benefits of LCR. However, other stakeholders highlighted that there are significant costs associated with supporting DNDCs, particularly for small and medium-sized issuers, and were in favour of only requiring the major banks to support DNDCs. Issuers noted that there are limited cost synergies from operating two debit networks despite the similarities in their product offering. Issuers flagged that there are cost duplications relating to investment spending, product upgrades and mandate compliance. Small issuers argued that the time and opportunity costs associated with issuing DNDCs hinder their ability to innovate and compete with the major banks; most issuers said that they would like the freedom to make a commercial decision about the issuance of DNDCs. Some stakeholders recognised this cost burden on smaller issuers, but maintained they should still be required to issue DNDCs as part of their ‘social responsibility’ given the system-wide benefits DNDCs enable. The international card schemes and some major banks argued that mandating DNDCs for all issuers would reduce competitive pressure in the market, eliminating ePAL's incentive to innovate and attract issuers.

Some stakeholders argued that without DNDCs, customers would lose valuable functionality only available through eftpos, such as real-time Medicare rebates and cash-out at the point-of-sale. However, some issuers felt that these additional services provided little value to their customers. Issuers that have begun the process of switching to SNDCs argued that SNDCs reduced the complexity of their product developments and would allow them to launch new product offerings (such as the mobile ‘pays’) for their customers more quickly. Some stakeholders also noted that the shift towards digital ‘form factors’ such as mobile wallets made support for DNDCs more costly than was the case in a world with only physical cards.

Several stakeholders noted that the financial case for large merchants to adopt LCR would be undermined if SNDCs became more prevalent, particularly given the growth of non-routable mobile-wallet and online transactions. Merchant groups and some other stakeholders argued that the Bank should take action to ensure LCR is possible on mobile-wallet transactions given their growing importance. More generally, these stakeholders argued that the adoption of LCR remains too low, and that the Bank should mandate the provisioning of LCR on an ‘opt-out’ basis and require that all merchants be provided with ‘dynamic’ routing functionality which realises the full possible savings from LCR. On the other hand, some stakeholders agreed that the Bank's suasion has worked, and that regulating LCR is not needed given recent industry progress. Some stakeholders also noted that discussions over LCR risked placing too much emphasis on the cost of accepting payments, and ignoring other functionality valuable to the merchant, such as security.

In the online environment, payments industry stakeholders generally argued that the Bank should not mandate the provisioning of LCR. Some of the major banks contended that developing LCR functionality online would be costly, time-consuming and would likely require a coordinated industry approach. A number of stakeholders also argued that there are material differences between the security and product offerings of the debit card schemes in the online environment. Given the higher incidence of fraud in e-commerce transactions, some of these stakeholders argued that LCR online would increase fraud rates, adversely affecting all stakeholders in the online payments ecosystem. These stakeholders felt that consumer choice should be paramount, and that customers should be given a transparent choice between the debit card schemes or be clearly notified if their transaction is being routed.

However, other stakeholders argued that the Bank should mandate LCR online. They argued that the major banks – which are both the largest acquirers and the largest issuers in Australia – are not incentivised to provide LCR online, due to the resulting reduction in their interchange revenue and their deep relationships with their international scheme partners. These stakeholders also argued that schemes provide customers with comparable product offerings, and that there is (or soon will be) little difference between their security functionalities. They also claimed that customers typically do not have a preferred debit scheme, and have a limited understanding of their debit options, so notifying them would only create confusion. They believed that since merchants bear the cost of payments, merchant choice should take precedence over consumer choice. Nearly all stakeholders were opposed to any requirement to provide customers with an option to override a merchant's routing decision given the complexity and friction it would add to the checkout process.

3.4 The Board's assessment and conclusions

The Board continues to support the issuance of DNDCs and the provision of LCR because they significantly enhance competition and efficiency in the payments system. Since the issues affecting the availability of DNDCs and LCR are highly inter-related, the Board has developed a package of reforms to address its concerns, drawing on the various specific options presented in the Consultation Paper. In doing so, the Board has carefully taken into account the views of stakeholders, the likely costs and benefits of regulatory intervention, and the scope of its mandate and powers, all in the context of a payment ecosystem that is changing rapidly.

3.4.1 Least-cost routing

Mobile wallets

One of the key challenges to LCR, and the competition and efficiency benefits that it brings, is the rapid growth of mobile-wallet transactions, which at present cannot be routed by merchants. Without LCR for mobile-wallet transactions, the pool of non-routable device-present transactions is likely to continue growing, further undermining the viability of LCR for large merchants and limiting the savings from LCR for smaller merchants. Accordingly, some stakeholders argued strongly for the Bank to require industry participants to enable LCR in the mobile-wallet context.

The Board acknowledges that extending LCR to mobile-wallet transactions would yield benefits in terms of lower payment costs for merchants and ultimately lower prices for consumers. Merchants would see a direct benefit by routing to the lowest-cost network, and there would be indirect effects on payment costs from increased competitive tension between the debit schemes (putting downward pressure on the interchange rates and scheme fees that apply to mobile-wallet transactions).

The Bank understands that there are 2 possible implementation models:

1. Under a ‘single token’ model, the mobile wallet would continue to present a single scheme token to payment terminals, but the owner of the scheme token (or token service provider (TSP)) would de-tokenise the payment credentials if needed and route the transaction to the merchant's preferred network. The card schemes, along with their TSPs and other industry participants, would need to coordinate to develop such a solution.
2. Under a ‘2 tokens’ model, mobile wallets would present tokens from both schemes to the terminal during checkout. Mobile-wallet providers would need to change their solutions so that their devices present 2 tokens simultaneously, which they have indicated may be feasible, but would be a significant and complex change. This model would likely also require all cards on mobile wallets to be reloaded (or ‘reprovisioned’). And many payment terminals would need significant modification (or replacement) to be able to make routing choices based on the 2 scheme tokens (rather than the single physical card number).

Both options would require a significant change to the technical implementation of mobile payments for the whole industry, which would involve significant costs and be time consuming. There is also very limited international precedent for such functionality; the Bank is only aware of the first and second models being used, to a limited extent, in the United States and France respectively. The first model may be somewhat easier from a technical perspective, but by requiring schemes to detokenize payment credentials for competing schemes, it raises the likelihood of ongoing disputes about the commercial terms of, and access to, this service; it may also partly unwind the benefits of tokenisation in reducing fraud. The second model would be not only more complex but would require the cooperation of some payment service providers – most notably the mobile-wallet providers – in relation to which the application of the Bank's powers under the PSRA is unclear.

Accordingly, while the benefits of enabling LCR in the mobile-wallet context could be substantial, the Board's view is that these would likely be outweighed by the significant implementation costs, as well as other legal and practical challenges. The Board is also mindful that mobile payment methods could change significantly in coming years (through, for example, the use of quick response (QR) codes).

However, in the Board's assessment there are other policy responses that would strengthen competition in mobile payments. First, the Bank has already begun to engage with mobile-wallet providers that do not currently support the provision of both networks on DNDCs and will be encouraging them to do so; legislative reforms following Treasury's Review of the Australian Payments System may allow this outcome to be achieved through regulation if necessary. Second, as discussed further below, issuers that are expected to issue DNDCs will also be expected to provision both networks in mobile wallets (where supported by the relevant schemes and mobile-wallet provider). In combination, this will increase the proportion of mobile payments for which consumers have a choice of debit network, thereby increasing competitive tension between the schemes.

LCR for device-present (or in-person) transactions

Another key policy issue is that while the broad availability of LCR for device-present transactions has put downward pressure on payment costs, merchant take-up of LCR has been quite limited. This has raised the question whether policy action is warranted to promote the wider take-up of this functionality, to generate further cost savings for merchants and greater competitive tension between debit schemes. Some stakeholders argued strongly that the Bank should require acquirers and payment facilitators to enable LCR for merchants as their default option (with merchants able to ‘opt-out’ if they wish). These stakeholders argue that the major banks do not have an incentive to provide and promote LCR to their merchant customers, because it results in less interchange revenue for the card issuing side of their business. Also, the complexity of payments issues in general and merchant pricing in particular create significant barriers to merchants understanding LCR and realising its benefits.

Higher merchant take-up of LCR would lead to further cost savings for merchants and benefit payment system efficiency. However, the Board is not convinced that formal regulatory intervention, such as requiring all acquirers and payment facilitators to offer LCR as their default option, is warranted, for the following reasons:

• The payments industry has made considerable progress in the provision of LCR in the device-present environment without any explicit regulatory requirements, albeit following considerable suasion by the Bank over a number of years. All major acquirers now offer LCR in some form, although the functionality offered by some acquirers is rather limited.
• Three of the 4 major banks already provide LCR, or soon will do so – on an opt-out basis, or in the background – for smaller merchants. Most recently, CBA has announced that it is moving all small business customers with annual turnover equal to or less than $250,000 to a new simple rate plan charging 1.1 per cent for all in-store card transactions, and 1.5 per cent for online transactions, regardless of the interchange rate or the type of card. This will be a significant reduction relative to previous plans for these merchants, with CBA implementing LCR ‘in the background’.
• Numerous stakeholders have noted that competition in the acquiring market is strong in many respects, with a range of new global and technology-focused providers entering the market in recent years, providing merchants with viable alternatives to the major banks. These new providers, along with existing smaller players, often offer superior technology capabilities to merchants (including more sophisticated LCR functionality) and are putting pressure on the major banks to improve their own payment services and pricing.
• Merchant take-up is not the sole measure of success of LCR. Even with low take-up, there is clear evidence that the availability of LCR has intensified competition between the debit schemes and led to lower wholesale payment costs for all merchants, with both interchange and scheme fees on routable transactions falling over the past few years. This has contributed to further falls in the average fee paid by merchants to accept card payments over recent years, with the evidence that is available suggesting that the cost of accepting debit payments in Australia is amongst the lowest in the world (Box C).
• Requiring acquirers and payment facilitators to enable LCR for merchants as their default option would not necessarily guarantee that the savings would be passed on to merchants.^[10] In contrast, strengthening competition in the acquiring market, including by improving merchants' understanding and awareness of payments issues, and their willingness to switch providers, would both increase the competitive pressure on acquirers and payment facilitators to improve and promote their LCR offerings as well as ensure that more of the savings are passed on to merchants. The Bank is separately proposing reforms in this area, as outlined in the section on ‘Competition in card acquiring’ below; in particular, the Bank is optimistic that extending the Consumer Data Right to smaller merchants' card transactions would be especially beneficial for boosting competition in that part of the acquiring market.

The Board's preferred policy to promote LCR is to state an explicit expectation that all acquirers and payment facilitators will both offer LCR functionality for device-present transactions and promote the functionality to their merchant customers. Acquirers and payment facilitators will be expected to report to the Bank on their LCR offerings, and on merchant take-up of LCR, every six months. In light of the progress made by the industry to date, combined with the competitive dynamics in the acquiring market and other initiatives in this Review, the Board sees this as a simpler and more balanced approach to achieving the desired improvement in LCR functionality and awareness.

Box C: The cost of debit transactions in Australia

Average merchant fees for all card payments have fallen over the past decade, from more than 0.8 per cent of the value of transactions to just over 0.6 per cent (Graph C1; right-hand panel). The Bank's various card payment reforms have contributed to this decline, with merchant fees for most payment systems falling over the period (see Graph 1 above). The gradual switch from credit to debit cards has also helped lower average card payment costs, because debit cards tend to be less expensive for merchants to accept than credit cards.

Average merchant fees for debit card payments have generally fluctuated between 0.4 and 0.5 per cent of the value of transactions over the past decade, despite a sizeable fall in the share of transactions processed through the domestic eftpos network, which tends to be cheaper for merchants to accept on average than the Visa and Mastercard debit networks.

There has been a fall in the average cost of debit payments since 2017. A number of reforms implemented by the Bank are likely to have contributed to this decline, benefitting smaller merchants in particular. Most notably, new standards implemented in July 2017 reduced the weighted-average interchange fee benchmark for debit card transactions, from 12 cents to 8 cents, and introduced caps on individual interchange fees. As noted earlier, larger merchants typically benefit from preferred low (or ‘strategic’) interchange fees on all their card transactions. Smaller merchants, on the other hand, usually bear the full cost of high interchange fees on some cards and transaction types. Capping interchange fees has therefore put downward pressure on the costs of accepting such payments for smaller merchants. The reduction in the debit interchange cap from 15 cents to 10 cents in this Review, discussed in the ‘Interchange fees’ section below, will similarly benefit smaller merchants.

More recently, the increased availability of LCR functionality starting in early 2019 has contributed to a modest decline in the average cost of debit card payments. This has lowered debit costs both directly by allowing merchants to route transactions through the lowest-cost network, and indirectly by increasing the competitive pressure on debit schemes to lower their fees. Indeed, over recent years there have been sizeable declines in the interchange and scheme fees charged on routable transactions. For example, the interchange fees charged by the international schemes on standard card-present transactions – which are in-person transactions at non-strategic merchants, using physical cards that are not premium cards – have fallen from 12½ cents (Mastercard) and 8 cents (Visa) in mid 2017 down to 4 cents currently (see Table 1 in the ‘Interchange fees’ section below). However, this has been partly offset by increases in interchange and scheme fees for some non-routable debit transactions, such as those made using mobile wallets, which are growing quickly and making up an increasing share of total debit transactions.

While similarly detailed data for other countries are not available, the available evidence suggests that the cost of debit card transactions in Australia is amongst the lowest in the world. As shown in (Graph 3) in the ‘Interchange fees’ chapter below, interchange fees on a $50 transaction in Australia are lower than in the United States, Canada, Europe and New Zealand. Further, average merchant fees for debit card transactions are much lower in Australia (around 0.4 per cent) than in the United States (around 0.7 per cent); indeed, merchants of all sizes in Australia, except those that are very small, tend to pay average merchant fees for debit transactions that are lower than the economy-wide average for the United States (Graph C2).

While this Review is implementing a number of measures to safeguard competition in the debit card market, the Board will be continuing to monitor the market and there would be a number of additional regulatory responses, including changes to the weighted-average interchange benchmark, that the Bank could consider in the event that there was some reduction in competition and upward pressure on merchant fees.

LCR for device-not-present (or online) transactions

The Board supports the provision of LCR online, given the clear benefits that LCR has had in the ‘device-present’ environment, in terms of stronger competition and lower payment costs. However, the Board is concerned that the provision of online LCR may be hindered by industry participants taking slow, divergent, or restrictive, approaches to its implementation. In particular:

• some stakeholders have argued that the major banks – which are both the largest acquirers and the largest issuers in Australia – are not incentivised to provide LCR online, due to the resulting reduction in their interchange revenue and their deep relationships with their international scheme partners;
• two of the major payment gateways in Australia are owned by the international schemes, and so may have less incentivise to provide online LCR, because it would increase competition between the international schemes and eftpos;
• one of the international schemes has implemented a rule requiring acquirers and merchants to notify customers of LCR in the online context and to provide them with an override option; merchant stakeholders argue that this rule would add considerable friction to the checkout process, and could significantly deter the development and take-up of LCR online.

Stakeholders have put forward two main arguments against online LCR, and for rules like the one noted above. First, they maintain that consumer choice should take precedence over merchant choice when paying online. Second, they consider that there are material differences between the security capabilities and other product offerings of eftpos compared with the international schemes.

However, the Board is persuaded that the majority of cardholders do not have a strong preference between debit card schemes. In addition, following extensive liaison, the Bank is not persuaded that differences in the security capabilities of the schemes – once ePAL has finished building its online capabilities – are likely to have a material impact on security or fraud in the online payments ecosystem. Accordingly, given that merchants incur the cost of processing a transaction and bear much of the fraud risk, the Board considers that they should be able to route transactions via their preferred network, without significant friction being added to the checkout process. Regulators have reached a similar view in the United States, where issuers and schemes are not permitted to prevent merchants from routing debit transactions, including in the online environment.^[11]

The Board therefore sees merit in taking action to facilitate the broad-based availability of online LCR for merchants. Given the Bank's traditional presumption in favour of self-regulation, and that eftpos' online functionality is still being rolled out, the Board is of the view that it is too early to intervene with formal regulation. Instead, the Board will set two explicit expectations regarding online LCR, and will consider more formal regulation if they are not met:

1. First, all acquirers, payment facilitators and gateways will be expected to offer and promote LCR functionality to merchants in the online environment by the end of 2022. In setting this deadline, the Board notes that ePAL is due to finish building out its online capabilities by the end of 2021, with issuers and acquirers due to comply with ePAL's mandate to enable their systems to process online eftpos transactions (for all risk levels) by mid 2022. In line with the expectation for the device-present environment, acquirers, payment facilitators and gateways will be expected to report to the Bank on their LCR capabilities and offerings, and on merchant take-up of LCR, every six months.
2. Second, industry participants will be expected to abide by the principles set out in Box D. These principles are intended to help the industry coalesce around an implementation model that, in the Board's view, appropriately balances the interests of merchants, consumers and the schemes. Under the principles, merchants would not be required to provide customers with a choice of debit network, but customers would always be informed if routing could occur. Box E provides some stylised examples of how the principles might work in practice.

A number of stakeholders noted that the complexity of the online payments ecosystem means that the development of LCR for online transactions would require considerable changes to infrastructure and systems across a number of participants, and so is best managed through a coordinated industry approach. Accordingly, the Board encourages industry to work together as appropriate to develop online LCR functionality, and welcomes the AusPayNet working group that has begun this process.

Finally, the rule implemented by one international scheme relating to online LCR has highlighted once again that scheme rules can have significant policy implications. Accordingly, the Board has decided that the Bank should be notified of all scheme rules and any changes to those rules. This will be implemented as an exercise of the Bank's information gathering powers under section 26 of the PSRA. The Bank expects this to impose minimal compliance burden on the schemes, as they would be required to simply provide the same access to rules and notification of changes that is already provided to scheme participants.

Box E: Online LCR in practice

This box steps through two stylised online transactions to clarify some aspects of how the principles outlined above might work in practice. These examples are illustrative only, and in practice merchants would be able to present debit payment options to their customers in many different ways while still adhering to the principles.

Consider a customer shopping online at two different merchant stores. Once the customer finishes adding items to their virtual carts, they proceed to the checkout pages and eventually reach the list of payment options. The customer prefers to use their (dual-network) debit card for online purchases.

At the first online store, the customer is presented with a ‘debit card’ option, which they select. The customer is then presented with the debit schemes accepted by the merchant, and is required to select their preferred debit network using a checkbox (as shown in Figure E1). Once they select their preferred network, Network 2 in this example, they enter their card details and finalise the payment. The payment is then routed through Network 2, regardless of the merchant's own preference (provided Network 2 is indeed enabled on the customer's card). This is because the customer's explicit choice of network cannot be subsequently overridden by the merchant, or any other party in the transaction process, and routed to another network (Principle 1).

At the second online store, the customer is presented with a ‘debit or credit card’ option, which they select, before entering their account number and details (as shown in Figure E2). However, they are not asked to explicitly choose a debit network. In this case, the merchant would be free to route the transaction to their preferred network (one of the networks on the card), but the customer should be notified that routing may occur (Principle 2). There are a number of ways in which customers could be notified of potential routing without disrupting the online shopping experience. Individual merchants could make their own decision about the most appropriate method of notification for their online store, including the location and the wording of the notice. The image below shows one stylised example (immediately below the scheme logos) of how a merchant could notify customers directly on a guest checkout page without interrupting the payment process. The customer would proceed to fill in their debit card details and finalise the payment (unless they were concerned about not having a choice of debit network, in which case they could instead choose another payment option). The merchant's payment service provider would then route the transaction through the merchant's preferred network.

Similar processes would apply if a customer was setting up a recurring payment or saving a DNDC on file with a merchant for future transactions – some merchants might give the customer an explicit choice of debit network for processing their future payments (in which case, the customer's selection should not be subsequently overridden), while other merchants might choose not to do so (in which case, they should provide some notification to the customer if routing could occur).

If a customer already had a DNDC saved on file with a merchant or had an existing recurring transaction that used a DNDC (and had not explicitly selected a debit network), the merchant – upon adopting LCR – would similarly need to notify the customer if they planned to route any future payments; this could be done, for example, via an email or a notification when the customer next chose to use their saved payment method.^[12]

Importantly, when customers were not given an explicit choice of network, the principles require that merchants must not mislead customers about which payment networks their transaction may be processed through (Principle 3). Using the second example above, in which the merchant displayed a selection of scheme logos, a DNDC transaction should only be processed through one of the networks displayed in that list.

3.4.2 Dual-network debit card issuance

Another key challenge to LCR, and the competition and efficiency benefits that it brings, is the shift by some smaller and mid-sized issuers away from issuing DNDCs. A widespread shift towards SNDCs would threaten the viability of LCR, which could impose significant efficiency costs on the system as a whole due to the loss of competitive tension between the debit schemes. Accordingly, the Board's view is that policy action to limit the shift to SNDCs is necessary.

However, in determining the appropriate policy response, the system-wide benefits of DNDC issuance must be weighed against the associated costs for individual issuers and others to support two debit networks. Many issuers have told the Bank that they incur significant additional costs from issuing debit cards with two networks instead of one, as there are limited cost synergies in connecting to two debit networks. In particular, technical differences between the networks were said to result in material duplication of issuers' compliance and development costs. Differences in scheme rules and back-office processes also reportedly mean that supporting two networks increases the ongoing day-to-day costs of operating a debit card portfolio. Given the largely overlapping functionality provided by the different debit schemes, some smaller issuers felt that supporting a second debit network yields little benefit to their customers but generates significant costs. The additional cost burden reportedly also makes it harder for small and medium-sized issuers to compete with the major banks, which can spread the costs of supporting two networks over a larger customer base. While estimates vary regarding issuers' costs of supporting two networks, the compliance and development costs alone are estimated to be more than a million dollars per year for mid-sized issuers. They are likely to be lower for smaller issuers, which rely more on aggregators such as Cuscal, ASL and Indue, but they are still significant amounts in the context of the overall costs of running a debit card portfolio (particularly on a per-transaction basis).

Overall, given that the issuance of DNDCs by the major banks is the main influence on the number of DNDCs in the market, the Board's view is that the economy-wide benefits from requiring DNDC issuance by the major banks would easily outweigh the costs to those banks (especially since those banks would likely issue DNDCs in any event). However, for very small issuers it is highly unlikely that the public benefits of DNDC issuance would outweigh the material fixed costs they would incur, because their contribution to the prevalence of DNDCs is not significant. Accordingly, the optimal threshold, at which the public benefits of DNDC issuance match the costs, is likely to lie somewhere between these two outcomes.

After further consideration of the various costs and benefits, informed by stakeholder feedback through the consultation process, the Board's judgement is that the requirement to issue DNDCs should extend beyond the major banks, to cover any issuers with a market share of more than 1 per cent of the value of debit transactions (corresponding to around $4 billion in transactions in 2020); this threshold would currently capture 4 issuers in addition to the major banks, but would increase the proportion of the debit market covered by around 10 percentage points (to almost 90 per cent).^[13] At lower thresholds, the Board has much less confidence that the cost imposed on issuers to support DNDC issuance would be outweighed by the public benefits. The Board's approach will ensure that DNDCs continue to account for a large majority of all debit cards in the market. Many smaller debit card issuers are likely choose to continue issuing DNDCs reflecting their assessment of the benefits to their customers, though they would not be required to do so. Given the vast majority of cards will be DNDCs, the possibility of LCR should continue to exert significant competitive pressure on interchange fees and scheme fees, even though the commercial case for adopting LCR will be somewhat weaker for large merchants that could benefit from strategic rates than if all debit card issuers were required to issue DNDCs.

There would be no presumption as to which two debit networks will be included on DNDCs by issuers. Various combinations of existing domestic and international schemes might be feasible, or indeed new networks that can process card transactions, provided that the two networks enabled were unaffiliated.

The Board continues to favour setting an explicit expectation for DNDC issuance, rather than imposing a formal regulatory requirement through a standard. This ‘expectations’ approach, rather than formal regulation, is consistent with the Board's traditional presumption in favour of self-regulation to address policy concerns. However, if this expectation was not met, the Board would consider imposing formal regulation in response. More broadly, if the cost of payments were to rise because LCR was no longer a viable option for many merchants and interchange fees or scheme fees were rising, the Board could consider if any other policy actions might be in the public interest.

An issuer that is expected to issue DNDCs will also be expected to provision both card schemes on their DNDCs in all form factors, including mobile wallets, offered by the issuer (where the functionality is supported by the relevant schemes and mobile-wallet providers). As noted earlier in relation to mobile wallets, this will help ensure that both debit networks are enabled in as many payment contexts as practicable, thereby increasing competition between the debit schemes. The expectation to issue DNDCs will extend to issuance by all divisions and subsidiaries of the relevant banks, because issuance by these entities can be significant and could be used to circumvent the requirement. However, the expectation will not apply to cards issued by a bank under a white-label arrangement, provided the bank's client is not affiliated with the bank.^[14],[15]

For those issuers that are expected to issue DNDCs, the Board does not expect them to replace any existing SNDCs – either international scheme SNDCs or eftpos proprietary cards – on issue with DNDCs; SNDCs issued on or before 31 December 2021, and the accounts they relate to, will be grandfathered. In contrast, new SNDCs issued after 31 December 2021 would be expected to be replaced with DNDCs as soon as practicable.^[16] A ‘new’ SNDC is one that relates to a new account or an account that did not previously have an SNDC attached to it; it does not include the reissuance of an SNDC that existed prior to 2022. This is particularly relevant for affected issuers' stock of eftpos proprietary cards; issuers will be able to keep the relevant accounts open and reissue the existing cards, but will not be able to open new proprietary-card accounts. Grandfathering these cards will minimise disruption for issuers and cardholders, with little negative effect on competition for a number of reasons: eftpos proprietary cards account for a relatively small share of cards on issue; they appeal to a limited demographic; and they are declining in importance, with most issuers in the process of phasing them out and ePAL hoping to replace them with DNDCs with eftpos as the first (or ‘front of card’) network. Also, while the average interchange rate on proprietary cards is higher than the average for transactions on eftpos DNDCs, it is similar to the average interchange fees on transactions through the international debit schemes. From a level-playing-field perspective, the important point is that both eftpos and the international card schemes would be similarly able to support and incentivise new SNDC issuance by smaller issuers. As discussed below, all SNDCs, including eftpos proprietary cards, will also be subject to the same regulation regarding interchange fees.

The Board's view is that it is not appropriate for schemes to provide issuers with interchange-based incentives to issue SNDCs. Switching to SNDCs would reduce the cost burden faced by an issuer, so there is little justification for schemes to provide higher interchange revenue at the same time. Accordingly, the Board's view is that the interchange standards should be amended to limit the possibility of schemes using interchanges rates in ways that would reduce competition and efficiency in the debit market. The preliminary proposal in the Consultation Paper was for the cap on interchange fees that are set in cents-based terms to be set at a lower level for SNDCs than for DNDCs. However, stakeholder feedback suggested that smaller issuers of SNDCs would be particularly affected by such a cap. In light of this feedback, and the broader coverage of the expectation for DNDC issuance, the Board has decided to take a different approach to interchange on SNDCs. Specifically, the potential for high interchange rates on SNDCs will be limited by introducing a ‘sub-benchmark’ for SNDCs, such that the weighted-average interchange fee on transactions on SNDCs from a given scheme must be no more than 8 cents. The Board could consider reducing the level of this sub-benchmark in the future if policy concerns arose regarding a significant shift towards SNDC issuance.

3.4.3 Potential tying conduct by the international schemes

The Board is particularly concerned about alleged tying conduct by the international schemes, which would undermine the benefits of LCR and limit competitive pressure in the debit card market, imposing considerable costs on the payments system. As noted earlier, a recent investigation by the ACCC has resulted in the ACCC accepting a (time-bound) court-enforceable undertaking from Visa that it would not engage in tying conduct.

The Bank could leave it to the relevant provisions of the Competition and Consumer Act 2010 (CCA) and the ACCC's enforcement powers to address tying conduct by the international schemes. However, given the potential negative impact of tying conduct by international schemes on competition in the debit card market, coupled with the fact that not all schemes have offered an undertaking and Visa's undertaking is limited to a 3-year term, the Board will consider separate action by the Bank to prevent such conduct across all international schemes; most stakeholders also supported this option. The Bank will seek to obtain specific undertakings from the international card schemes that they will not engage in tying conduct; if the schemes are not willing to provide voluntary undertakings, the Bank will consult on the introduction of a new standard to explicitly prohibit such behaviour (separately to this Review). Such a standard would provide certainty to all parties about permissible conduct and would be expected to prevent tying conduct taking place, rather than relying on regulators to take enforcement action after such conduct has occurred. It would be a low-cost policy that supported competition in the debit card market by helping to ensure that schemes compete solely on the basis of their debit card offerings.

The provisions of any voluntary undertaking will be underpinned by the principles in Box F. Compliance will be monitored through an annual certification requirement (and potentially enforced through the Bank's ability to issue directions under the PSRA, as well as any actions from the ACCC where there was also a breach of the CCA). If the schemes are not willing to provide voluntary undertakings, they could expect that consultation on a new standard would proceed on the basis of these principles.

3.4.4 Conclusions