Fintech and responsibility: Buy-now-pay-later arrangements

Abstract

Buy-now-pay-later (BNPL) arrangements have rapidly emerged as a short-term debt option, and like other innovative and disruptive Fintech, challenge existing regulation. BNPL arrangements avoided prescribed ‘responsible lending’ legislative obligations, which applied to similar short-term credit products. Instead, BNPL relies on ‘responsible spending’ in providing a potentially cheaper option than alternatives such as credit cards. We describe the interplay of regulation and responsibility with BNPL. A survey investigates whether a key demographic (young adults) have an appetite or skill for responsible use. We analyse the preference for BNPL relative to credit cards and the role of financial literacy and traits including propensity to plan and save. The findings suggest that financial literacy reduces perceived BNPL benefits and that lower financial literacy is associated with more benefits and less risks.

JEL Classification: G53, D14, G51, G41, G18

Keywords

Debt financial literacy Fintech responsible lending responsible selling responsible spending

1. Introduction

How do new Fintech credit products, such as buy-now-pay-later (BNPL) arrangements, alter the relationships between consumers, business, financial organisations and regulators? A central question that has emerged in considering these relationships is the responsibility of each party. In particular, what responsibility attaches to the credit provider and should this be at their discretion or as prescribed by legislation and enforced by the regulator? What responsibility level can be assumed of consumers when they choose these products, and are they materially different from other sources of consumer credit? The regulatory approach chosen reflects the judgement of policy makers and regulators as to how these responsibilities can best be balanced.

Agarwal et al. (2009) identify a range of mutually compatible regulatory approaches available to policy makers and regulators when considering financial decision-making, distinguished by the level of paternalism of the policy and the regulator. Getting the approach right is challenging. Recently chronicled financial industry behaviour identified financial misconduct by many entities in an environment where there was a ‘marked imbalance of power and knowledge between those providing the product or service and those acquiring it’ (Commonwealth of Australia, 2019b: 2). Given this, the design of regulation is clearly important. Policy makers also consider the broader economic externalities from a given regulatory approach. For example, the justification for recent proposed changes to consumer credit legislation emphasises the need for ‘the efficient flow of credit by reducing the time that it takes consumers and businesses to access credit so that consumers can continue to spend and business can invest and create jobs’ (Commonwealth of Australia, 2020b: 3).

Fintech lenders ‘position themselves as more convenient, faster, and cheaper’ for consumers due to, in part, their ‘online presence, automation, and favorable regulation’ (Balyuk, 2019: 1). In the case of BNPL arrangements, providers could be characterised as having exploited existing legislation or loopholes to minimise regulatory oversight evidenced by initial debate as to whether the BNPL arrangements were deemed credit under the National Credit Act (for a discussion, see Commonwealth of Australia, 2019a, Chapter 5). BNPL arrangements vary across providers but can be seen as the mirror image of traditional lay-by arrangements. Traditional lay-by arrangements enable consumers to purchase, periodically pay down and then ultimately receive a product, whereas BNPL arrangements allow the immediate receipt of the product or service in exchange for an agreed set of periodic payments over a short-term period.¹

Crucially, in the Australian context, a number of BNPL arrangements appear specifically structured so as not to be regulated by the prevailing National Consumer Credit Protection (NCCP) Act. For some arrangements, this is because² they do not charge interest for the credit; whereas for others, it is because their fees are not large enough and their term is not long enough.³ Where applicable, the NCCP Act obligates credit providers to assess credit contract suitability and prohibits credit provision unless the credit contract is not unsuitable.⁴

Just as products such as BNPL emerge with structures that result in them not being subject to regulatory oversight, so can legislative changes emerge reflecting a change in philosophy of how responsibilities are managed. For example, recent proposals⁵ reveal a

move away from a prescriptive framework for lenders and borrowers and will support risk-based lending that is attuned to the needs and circumstances of the borrower and credit product . . . [which reflects] . . . system-level obligations rather than focusing on individual loans engaged in by licensees. (p. 5)

The regulatory mix that ultimately prevails reflects an assumed level of responsible behaviour by both credit providers and credit consumers. Greater emphasis on requiring responsible lending⁶ is justified if responsible use (borrowing and spending) is not reasonable to expect, or if the cost borne by individuals provided with unsuitable credit exceeds the benefits from meeting overall credit provider financial stability. A further dimension is the cost of acquiring the information necessary to make an informed judgement. If the cost to acquire information is too high, responsible use of the product is unlikely.⁷

The arrival of new Fintech products⁸ coincided with new legislation and regulatory powers for financial products generally. The Design and Distribution Obligations (DDO) and Product Intervention Powers (PIP) Act 2019, and supporting regulations (Australian Securities and Investments Commission (ASIC), 2020b), provide new powers for the consumer credit regulator, the ASIC. Some industry participants have stated that there is no need for BNPL to be subject to the NCCP Act or the new DDO and PIP Acts, given the simplicity of the BNPL product (Afterpay, 2020d). For example, the largest Australian BNPL provider, Afterpay, describes the imposition of credit regulations on them as ‘inappropriate’ and argues that ‘Afterpay is a modern-day payment service that services retailers and, in turn, encourages and rewards responsible spending’ (Afterpay, 2020b, emphasis added). Where there is discretion to do so, BNPL providers such as Afterpay argue that they are better left as the arbiter of unsuitability and have a better chance of being a lender that is responsible, than if required to be so by legislation and the regulator. In contrast, some BNPL providers voluntarily choose to conduct credit checks and some hold a credit licence. For example, zipMoney holds a licence for its regulated products (ASIC, 2018a) whereas AfterPay originally had a credit licence but cancelled it in 2019 arguing it had not been used (AfterPay Touch Group, 2019c).

Reference to responsible spending places obligations on the consumer. What remains unclear is the distribution of ability for responsible spending among consumers, particularly, given the age profile of BNPL users. For example, 18–34 year-olds comprise 61% of BNPL users, with 18–24 year-olds making up 23% overall (ASIC, 2020a). In terms of ability to engage in responsible spending, available empirical evidence identifies younger adults as the least financially literate. For example, using Australian household data from wave 16 of the Household, Income and Labour Dynamics in Australia (HILDA) survey,⁹ those aged 15–19 and 20–24 years have the lowest financial literacy among all 5-year age bands except those in their 80s and 90s.¹⁰ For those in the 18–24 year-old group, BNPL will be part of their choice set as they first encounter consumer credit; whereas for older consumers, it adds to a previously encountered choice set. The justification for regulatory intervention for credit contracts was, in part, due to the ‘considerable information asymmetries’ between consumers and credit providers (Commonwealth of Australia, 2009: S2.6) echoed in the Royal Commission into Misconduct in the Banking, Superannuation and Financial Services Industry (Commonwealth of Australia, 2019a: 490). The Royal Commission also noted the varied financial literacy in the population, which interacts with these information asymmetries.

The remainder of this study is structured as follows: Section 2 presents a review of BNPL arrangements in Australia and considers how business, credit providers and regulators interact with the changes wrought through Fintech innovation. Section 3 outlines the empirical approach of the article, which utilises data from a survey of an Australian sample of undergraduate students who are a key target demographic of BNPL arrangements. The section analyses key scales administered in the survey including: financial literacy, orientation to saving and propensity to plan. New scales are developed to explore attitudes and preferences for BNPL products in comparison with credit cards (CCs). Section 4 presents the empirical analysis with attention to gender and age in addition to the scales noted. The final section summarises the main findings of this study and discusses policy implications for regulators.

2. BNPL overview

BNPL is an arrangement that allows consumers to buy and receive goods and services, immediately, but pay for the purchase over an agreed schedule. The innovation is less in how the product operates and more in its delivery mechanism, that is, online and at the point of sale, and in its structure that enables it to avoid regulation that applies to similar products. The consumer applies at the point of sale (which can be in-store or also online) via an App and is approved (or not) instantly for the purchase. The consumer completes the transaction at the merchant and purchases the good or service. The BNPL provider pays the merchant the sale price minus a commission fee.¹¹ The BNPL provider then collects the full price from the consumer over a series of short (fortnightly) instalments.

BNPL arrangements share the structured payments of lay-by transactions and other credit agreements. In a lay-by, you: (1) pay for the products in at least two instalments and (2) do not receive the products until the full price has been paid (Australian Competition and Consumer Commission (ACCC), 2019). In contrast, BNPL transactions satisfy an instant gratification or utility need among consumers with the goods or services received now, and instalment payments following. Lay-by transactions are likened to saving towards a purchase (or ‘save now buy later’ (SNBL)), making regular deposits until the consumer has enough deposited at the lay-by provider to receive the goods or services. The consumption is delayed but not the commitment (Loewenstein, 1987).

BNPL is credit in the ‘ordinary everyday understanding of what credit is’ (Commonwealth of Australia, 2019a: 67) but not a credit product under the NCCP Act. This is important as while BNPL providers are assumed to generally engage responsibly with consumers, not being classified as a credit product means that they are not subject to specific ‘responsible lending laws’.¹² The NCCP Act defines credit as: (1) payment of a debt owed by one person (the debtor) to another (the credit provider) is deferred; or (2) one person (the debtor) incurs a deferred debt to another (the credit provider; NCCP, Schedule 1, Part 1, Section 3). However, for the NCCP Act to apply, interest must also be charged, which most BNPL arrangements do not. Late fees or account fees applicable to BNPL arrangement, where payable, also do not prescribe it as a credit product. Initial confusion over how BNPL products were classified was demonstrated in the case of Afterpay being required to conduct a compliance audit to determine whether it breached the Anti-Money Laundering (AML) and Counter-Terrorism Financing (CTF) Act 2006 (Australian Transaction Reports and Analysis Centre (AUSTRAC), 2019). The audit concluded that Afterpay’s BNPL ‘activity constitutes the provision of a loan and repayment of a loan’ while Afterpay’s press release indicated that Afterpay ‘does not require customers to enter into a traditional loan’ (Afterpay Touch Group, 2019a: 2, cf. p. 6).¹³

BNPL products are similar to CCs in the way they operate with three major differences: (1) there is (generally) no credit assessment in the traditional sense. Rather, they apply ‘algorithms, which enable them to investigate consumers’ financial circumstances and test their eligibility for funds’ (Commonwealth of Australia, 2019a: 66) at point of sale; (2) BNPL spending limits and repayment schedules are generally more restrictive than CCs; and (3) BNPL are presented as interest-free and only accrue fees after a missed or late payment, whereas CCs generally provide an interest free period, thereafter interest accrues on the outstanding balance.

The majority of BNPL users are millennials who use BNPL as a source of short-term funding representing a potential shift from traditional credit sources such as CCs. Figure 1 supports the hypothesis of an apparent shift away from traditional short-term credit sources towards BNPL. The graph displays Australian data of outstanding CC limits and balances accruing interest and, for comparison, the share price of Fintech company, Afterpay. Both CC series are in clear decline and thus being in stark contrast to the upwards-trending share price of Afterpay.¹⁴

Figure 1.

Outstanding credit card limits and balances accruing interest and Afterpay share price.

2.1. Responsibility

Responsibility is central to the regulation of credit products. In this section, we consider the case of CCs and BNPL arrangements, given their overlap in features for consumers but different regulatory treatment. Figure 2 presents a summary of participants’ interaction and responsibilities.

Figure 2.

Summary of participant relationship and responsibilities.

In the case of CCs, the NCCP Act reflects a more paternalistic policy approach, and arguably more pragmatic view of the regulator’s responsibility, given the history of poor consumer outcomes due to credit providers’ conduct. Responsibility is explicitly placed on the credit provider who must acquire a credit licence (from ASIC) and must disclose information about the true cost of the credit, for example, by disclosure of a comparison rate (S8.273; Commonwealth of Australia, 2009). In turn, the regulator has responsibility through licensing to ensure this occurs.

There is a clear interaction between regulator and credit provider. In this approach, as reflected in the governing Act’s title, the consumer is to be protected, cast as having to rely on the ‘intermediary’s skill and expertise’ to assess their suitability for any credit product offered (Commonwealth of Australia, 2009: S3.9). A borrower’s responsibility, or accountability, is described in terms of their need to ‘conduct themselves honestly and transparently’ (Commonwealth of Australia, 2009: S3.17) in the provision of information when they engage with a credit provider. It is only after this interaction with the credit provider that the consumer engages, separately, with the retailer who has no role in the credit provision.

As noted in the introduction, regulatory settings evolve and reflect different means of striking a balance in levels of responsibility required of participants.¹⁵ Recent proposals reflect a shift away from a prescriptive product approach towards ‘support [for] risk-based lending’ (Commonwealth of Australia, 2020b: 5).¹⁶

In the case of BNPL, the regulator has a less direct role and does not directly interact with the provider. However, through the DDO and PIP Act 2019, the regulator now can exercise responsibility and intervene using ‘product intervention powers’ and ‘design and distribution obligations’.¹⁷ The Act’s explanatory memorandum reflects the view of the Financial Services Inquiry (Commonwealth of Australia, 2014), who recommended these powers be provided, that effectiveness of a policy approach based on disclosure will be reduced if there is ‘consumer disengagement, complexity of documents and products, behavioural biases, misaligned interests and low financial literacy’ (Commonwealth of Australia, 2018: 4). The new powers allow intervention in anticipation of consumer detriment, not after the fact. The effectiveness of these new powers is yet to be determined. While the largest BNPL provider has rejected the need for inclusion of BNPL arrangements (Afterpay Touch Group, 2019b), other providers voluntarily undertake actions including income and credit checks (Zip, 2019).¹⁸ BNPL providers have also drafted an industry code of practice (Australian Finance Industry Association, 2020), which commits providers to specific standards as a means of reducing the need for regulatory oversight.¹⁹

Figure 2 also identifies another difference in the case of BNPL, which is the role asserted for consumers and retailers. BNPL providers argue that their product is a tool for consumer empowerment and assists in responsible spending (Afterpay, 2020b). The consumer is not a passive participant in the BNPL providers’ view. This assumes an ability, and disposition, of consumers to use these features, which we examine empirically in Section 4. BNPL providers argue that BNPL terms and processes are simple and the risks low (size and duration of loans) requiring only basic skills of consumers. The retailer is also integral to the process as they can ‘Increase average order value and frequency by offering Afterpay at checkout’ (Afterpay, 2020c). The retailer is a key agent in the provision of BNPL by facilitating access. However, as noted in the introduction, this is also the case with other credit products such as department store CCs and car yard finance.

2.2. Macro assessment and second-best outcomes

From a macro perspective, responsible lending and selling are supply-side (credit provider) responsibilities; whereas responsible borrowing (spending) focuses on the demand-side (consumer). Responsible regulation requires policy makers and regulators to weigh the costs and benefits of regulation and whether to lay more emphasis on supply- or demand-side responsibilities. The relationships are dynamic and the regulator, and other participants, anticipate and react to each other. Whereas regulation could either reduce or enhance competition and thus stifle or promote innovation, it is more likely that strong forms of regulation help incumbents and thus decreases competition and innovation; whereas light forms of regulation are more likely to be associated with low barriers to entry and thus more competition and innovation.²⁰

In the context of BNPL, and Fintech more generally, it seems that legislators and regulators in Australia have initially chosen a light-touch approach allowing BNPL providers to operate and be exempt from certain types of regulations to which banks and CC providers are not exempt from for comparable short-term credit. This is good news for the Fintech industry, for example, evidenced by the share price of BNPL providers such as Afterpay and Zip Co Ltd, but also for competition and ultimately consumer outcomes. If consumers at least partially switch from high-cost debt to low-cost BNPL debt, this may further enhance competition and lead to lower financing costs for consumers.

Consideration of BNPL arrangements is set within the context of an existing portfolio of short-term credit products. It is rarely the case that a new financial product can be argued as unambiguously providing benefits to all. Policy makers must assess the distributions of the costs and benefits of new products. This article does not consider BNPL relative to potentially problematic short-term sources of funds such as payday loans (ASIC, 2015) and short-term consumer leases (Commonwealth of Australia, 2016), other than CCs (ASIC, 2018a). In this context, BNPL can potentially offer second-best improvements. For example, BNPL is expected to be a better option than a payday loan, and BNPL is expected to improve financial inclusion. But there will be some who experience detrimental outcomes due to how it is used in itself or how BNPL is used with other short-term debt sources. For example, while BNPL can increase financial inclusion by providing a short-term source of credit to those otherwise unable to access credit (e.g. via a CC), it may remain nonetheless unsuitable to a consumer and begin a cycle of problematic debt use, even if initially small.

2.3. Mixing regulatory approaches – responsible lending and responsible borrowing

To better understand potential regulatory or financial problems that occur due to the introduction of new Fintech products such as BNPL, we use a simple thought experiment. Assume there are two states. State 1 only allows the use of CCs whereas State 2 only allows the use of BNPL. The regulatory approach of State 1 focuses on the credit provider and is based on prescribed responsible lending; whereas the regulation of State 2 focuses on the consumer and is based on facilitating responsible borrowing and spending. The characteristics of products in both states are comparable to reality: CC limits are multiples of the maximum amounts of BNPL; credit card interest rates are at or near 20% p.a., BNPL interest rates are at 0% over 6–8 weeks; and the maturity of credit card debt is up to 5 years, whereas for BNPL it is 6–8 weeks. Consequently, the consumers of State 1 can access larger debts at higher costs and over longer time frames than the consumers of State 2.

Now assume that there is a policy change and both states allow both products, CCs and BNPL. The problem that arises with the policy change is that neither the regulatory approach nor the responsibility is clear-cut anymore. While BNPL is not problematic on its own and potentially less so than CCs, the availability of both products in each state blurs regulatory approaches and increases the risks of (ir)responsible lending and (ir)responsible borrowing. It also increases the risks of sub-optimal regulation and the risks of sub-optimal debt-financed consumption decisions. While both products carry certain levels of risks, the combination of the two can be riskier than the sum of its components.

Adding CCs to a ‘clean-sheet’ (no excess debt) State 2 is less problematic than adding BNPL to State 1 in which some consumers have accumulated excess levels of debt. In this case, BNPL may be used irresponsibly, given the legacy of excess debt possibly due to irresponsible lending of CC providers in the past. Consumers who have been deemed un-creditworthy for CC use could also use BNPL. This issue may have been reflected in the policy and regulators’ concerns to allow BNPL to remain less directly regulated than CCs. BNPL has ‘arrived’ at a time where 3 in 10 households are over-indebted,²¹ contributed in part due to irresponsible lending (Moradi-Motlagh and Jubb, 2020). BNPL is a greater risk if introduced in an economy in which households and consumers have excess debt. In other words, in the absence of a history of irresponsible lending and over-consumption in the past, BNPL would be unequivocally positive for consumers and competition. Since BNPL enhances competition and lowers the cost of debt for consumers, BNPL’s overall, aggregate, effect is expected to be mostly positive explaining the Australian regulators decision to allow BNPL providers to operate outside the NCCP Act.

Finally, since BNPL debts are comparatively small and maturities short, aggregate risk is limited. A simple calculation based on non-conservative estimates further supports this view. Assume that 5 million Australian households used BNPL and had outstanding payments (debts) of $500 each, that is, $2.5bn in total. This is a small number compared with the $140bn of outstanding CC limits and the $28bn of CCs accruing interest as of March 2020 (as shown in Figure 2). These numbers suggest that the proportion of outstanding BNPL ‘credit’ is relatively small when compared to CCs, and regulation would primarily focus on the CC debts rather than BNPL arrangements.²²

This optimistic view of the role of BNPL, and its characterisation as less problematic, can be contrasted with challenging evidence to the contrary in the behaviour of a substantial minority of users. This includes related credit behaviours such as: more frequent BNPL users being more likely to use other loans to meet repayments; BNPL users being 50% more likely to incur interest on CCs than non-BNPL users; and BNPL users using more of their available CC limit than non-users (ASIC, 2020a).

Assumptions about consumers’ ability to use or understand financial products, such as CCs and BNPL, underpin policy and regulatory responses. As has been identified in this section, an assumed lack of understanding has most commonly been the explicit justification for approaches that aim to protect consumers. By design and through promotion, BNPL products more explicitly rely on consumer financial literacy to justify their exclusion from regulatory oversight. The following two sections investigate the basis of this empirically.

3. Method

To investigate the level of understanding that can be used as a basis to assume consumer responsibility, we begin by examining consumer attitudes towards BNPL and CCs and the behavioural beliefs that underpin these attitudes. We do so because attitudes are expected to play an influential role in the intention to use, and actual use, of BNPL and CC products. Our focus on intention to use BNPL and CCs is grounded in theory and pragmatism. Because we focus on a sample of emerging consumers, the data availability of actual usage is limited. That is, young adults are just beginning to make financial decisions where short-term financing is needed or available. The role of behavioural beliefs and attitudes in the intention to use these products is captured in the model proposed in Figure 3, which draws from the theory of planned behaviour (TPB; Ajzen, 1985, 1991).²³ TPB proposes that behaviour intention, here the intention to use BNPL and/or CCs, is a function of attitudes to the behaviour, subjective norms about the behaviour and perceived behavioural control of the behaviour. In turn, the intention to use the products and perceived behavioural control over using the products predict use of BNPL and CC products. TPB assumes that attitudes, subjective norms and perceived behavioural control are based on both the strength and evaluation of beliefs. TPB also assumes that the role of attitudes and norms are fully mediated by intention, whereas perceived behavioural control is partially mediated by intention.

Figure 3.

Model of BNPL, credit card use.

While TPB provides a theoretical basis for the current article’s empirical approach, we do not propose to test TPB and note several significant differences in the role of key variables. First, we propose that financial literacy has explanatory power in addition to beliefs, intentions and behaviour. Second, our data allow a cross-sectional analysis and, therefore, measures intentions and behaviour at the same time. Third, we do not include perceived behavioural controls in our empirical estimation, which is expected to have an influential role. For example, we expect that unexpected financial shocks have a significant explanatory role in predicting actual BNPL and CC use. We note that the samples for the analysis were collected prior to the COVID-19 pandemic. Finally, we also do not explore ‘feedback loops from behaviour to cognitions’ (Ajzen, 2015). That is, the extent to which favourable or unfavourable experiences from using the products reshapes beliefs and intentions.

3.1. Sample

Data were collected from surveys of undergraduate students at The University of Western Australia. As part of a programme evaluating the effectiveness of financial education interventions on financial literacy, students who enrol in an elective personal finance unit (Managing Your Personal Finances), and a control group of peers who could have otherwise enrolled, are invited to complete surveys, which examine financial literacy and financial behaviours. A module was included in the 2018 and 2019 end of semester surveys, which is distributed to the sample in each year as well as those who completed a survey in 2017. Gerrans and Heaney (2019) have previously shown that the profile of students who enrol in the unit is not substantively different from the general sample of students in the control group. Not all students who complete the initial survey respond to the follow-up invitation, which may result in a potential self-selection bias. The estimations check for this, but the results do not suggest the baseline results are impacted by self-selection bias. A further discussion of this is presented in the robustness section.

3.2. Key variables

As an important component of attitudes towards BNPL and CCs, underlying beliefs (positive or negative) about the products are important in predicting intention to use, and actual use, of the products. From a regulatory perspective, it is arguably beliefs (or awareness) of the risks of a financial product, rather than the benefits, which are more important. While efficient financial product markets rely on consumer awareness of both benefits and risks, in isolation, a failure to appreciate the benefits of a financial product represents a missed opportunity. However, a lack of awareness of the risks may directly result in a wrong product choice or its incorrect use and lead to a negative overall outcome.

Our interest is primarily with the role of three key variables in explaining benefit and risk beliefs: financial literacy, savings orientation and propensity to plan. Financial literacy was measured using the 13-item scale of Fernandes et al. (2014). We construct an index of financial literacy following the approach of von Gaudecker (2015) and assign don’t know responses with the probability that a guessed answer would be correct from the remaining options and then use an iterated principal factor method to estimate an index using the Bartlett score, given factor loadings.

Propensity to save was measured using the personal savings orientation scale (PSO; Dholakia et al., 2016). PSO measures the ‘habitual and lifestyle-oriented aspects of saving money in addition to planning and developing saving goals’ (Dholakia et al., 2016: 153). The PSO scale includes nine items, which cover day-to-day actions (e.g. ‘I do not spend money thoughtlessly’ and ‘Putting money into personal savings is a habit for me’) and savings lifestyle (e.g. ‘The goal of saving money is always at the back of my mind’) scored on a seven-point agreement, disagreement scale. PSO can be viewed as propensity for SNBL rather than BNPL. Dholakia et al. (2016) report a positive relationship between PSO and savings and negative relationship between PSO and CC debt.

Product providers suggest that BNPL allows better planning (budgeting) for large expenditures in contrast to some consumer experience, which suggests ‘this [BNPL] can increase the likelihood of making a purchase quickly without careful consideration’ (ASIC, 2018b: 30). Importantly, this risk also applies to CCs and even debit cards. An individual’s propensity to plan would, therefore, appear to be an important characteristic if a key feature of the product is to support better planning. We measured propensity to plan long-term (LTP) using the six-item scale of Lynch et al. (2010). The scale includes items such as ‘I set financial goals for the next 1–2 years for what I want to achieve with my money’ and ‘I actively consider the steps I need to take to stick to my budget in the next 1–2 years’. Items were scored on a seven-point agreement or disagreement scale.

3.3. Additional variables

Additional data were collected to measure students’ studies and characteristics. Of the former, we include indicators for whether the respondent had previously enrolled in the personal finance unit (Unit), whether they were a commerce student (Commerce major) and whether they were an international student (International). Controls for the latter include gender, age and indicators for having part-time work (Work) and living at home (Home). A final item sought perceptions of norms, specifically subjective norms, as ASIC (2018b) reported that 28% of users identified ‘word of mouth’ as an influence in BNPL use. Respondents rated they agreed that each product was ‘preferred by my friends’ (do not agree at all = 0 to completely agree = 5).

3.4. Expectations

Because BNPL is a new product, we expect financial literacy to have a more influential role in explaining consumers’ understanding of its benefits and risk relative to a more established product such as a CC, even among younger consumers. Those with an orientation towards saving are not expected to score the benefits of short-term debt as highly, given they are expected to have or see a reduced need. Therefore, we expect PSO to be negatively associated with the benefit beliefs of both products. We expect LTP to be positively associated with benefit belief scores. This aligns with the marketing of BNPL products highlighting that they provide consumers with a means of budgeting and planning for larger expenditures (Afterpay, 2020a). In terms of perceived risks, the role of PSO depends on what we expect to be the basis of PSO. If it is driven by a belief in the virtue of saving per se, there may be no relationship between PSO and product risks. If, however, PSO is driven by an aversion to debt and, therefore, a need for savings to fund larger expenditures, we might expect a positive relationship between PSO and product risks. It is difficult to identify an argument for a negative PSO and risk score relationship and, therefore, the net expectation is of a positive PSO and risk score relationship. In terms of LTP, those with a lower LTP are expected to be less aware of the risks, given they are less inclined to consider long-term obligations and, therefore, have lower risk scores. For those with higher LTP, we can argue for both a positive and negative relationship expectation. It may be that those who are more inclined to plan for the future are, therefore, more aware of the need and risks in ability to repay. However, having a propensity to LTP may provide confidence of being able to repay and hence a negative LTP and risks relationship. Therefore, the LTP and risks relationship depends on whether awareness or confidence results from propensity to plan.

4. Results

4.1. Estimating BNPL and CC attitudes

To identify beliefs that underpin attitudes towards BNPL and CCs, we developed statements to cover three broad areas, which reflect the views put forward by product providers, consumer advocates and regulators. The three areas covered are the features of the product ( ‘help plan for large purchases’, ‘help with budgeting’, ‘cheap’), experience of using the product ( ‘easy to use’, ‘easy to understand terms’) and consequences from using the product ( ‘leads to overspending’, ‘risky for my financial health’). The latter two items are based on reports that ‘more than half (55%) of buy now pay later users believed that they were spending more’ (ASIC, 2018b: 11). The overall ‘risk’ concern was a key finding from stakeholder consultation in the regulators review (ASIC, 2018b: 33). The same review identified, however, that almost two-thirds of consumers considered BNPL lower risk than other credit products (ASIC, 2018b: 29).

The preamble to the question stated: ‘Buy-Now-Pay-Later (Credit Cards) arrangements can be described as . . .’ followed by the seven descriptions as above. Respondents rated whether they agreed with each description for both products on a ‘Do Not Agree at All - 0’ to ‘Completely Agree – 5’ scale. Note that the survey did not qualify these statements or provide an indication of whether they were necessarily valid or correct. The survey also did not seek justification of response. We, therefore, primarily rely on the cited ASIC consumer surveys as support for the validity of the statements.

A summary of mean ratings is presented in Table 1. An exploratory factor analysis was undertaken using principal component extraction, separately for BNPL and CC item responses. For purposes of the exploratory factor analysis, we use a larger sample than in the main regressions as we include some respondents who completed the product module in their first survey but did not complete any subsequent surveys. The latter regression only includes respondents who completed the surveys at the end of the semester. The Kaiser–Meyer–Olkin measure of sampling adequacy was 0.73 for BNPL and 0.66 for CCs suggesting, respectively, ‘middling’ and ‘mediocre’ sampling adequacy (Kaiser, 1974). Bartlett’s test of sphericity χ² (21) = 1791.419 (BNPL) and χ² (21) = 1266.718 (CC) indicate that the correlation structure of the items for both BNPL and CCs are adequate for analysis. A cut-off criterion of 0.40 for factor loadings and Eigenvalues greater than one was implemented.

Table 1.

BNPL and credit card beliefs.

	Obs	Mean	Standard Deviation	Mean	Standard Deviation	Paired t-test
		BNPL		Credit cards		BNPL versuscredit card
Easy to use	865	3.46	1.39	3.92	1.27	−8.47***
Cheap	865	2.06	1.53	1.60	1.40	7.49***
Lead to overspending	865	4.10	1.18	3.95	1.19	3.19***
Help plan for large purchases	865	2.46	1.56	2.52	1.56	−1.09
Risky for my financial health	865	3.62	1.40	3.45	1.37	3.08***
Easy to understand terms	865	2.36	1.41	2.20	1.38	2.85***
Help with budgeting	865	1.69	1.41	1.76	1.29	−1.37

BNPL: buy-now-pay-later.

This table reports summary statistics of ratings to the question ‘Buy-Now-Pay-Later (Credit Cards) arrangements can be described as . . .’ rated on a scale of ‘Do not agree at all – 0’ to ‘Completely Agree – 5’. The final column presents a paired t-test of mean scores by product. The sample is all those who completed a survey to enable the factor analysis. For purposes of the exploratory factor analysis, we use a larger sample than in the main regressions as we include some respondents who completed the product module in their first survey but did not complete a follow-up survey. Significance is indicated by ***p < 0.01.

For both BNPL and CCs, two factors were supported accounting for 69.0% and 63.3% of common variance. The first is labelled ‘Advantages and Benefits’ and includes four items: cheap, help plan for large purchases, easy to understand terms and help with budgeting. The second factor is labelled ‘Risks’ and includes two items: lead to overspending and risky for my financial health.²⁴ For BNPL, all items had loadings exceeding 0.68 after rotation for the first factor and 0.69 for the second factor. For CCs, the minimum loading for factor one was 0.52 and 0.62 for factor two. One item ‘easy to use’ was deleted due to cross-loadings on both factors, more so for BNPL, with the difference between loadings close to 0.15 (Worthington and Whittaker, 2006). Deleting the item improved the Cronbach’s Alpha of the items in the Advantages and Benefits scale to 0.80 (0.74) for BNPL (CCs), and 0.74 (0.69) for the Risks scale for BNPL (CCs). The regression scores for each set of items were estimated for BNPL and CCs and used as manifest variables in the estimations using the 2019 sample discussed in Section 4.2.

With measures of beliefs that underpin attitudes to the financial products available, a series of hierarchical regressions were estimated to explain beliefs, intention and behaviour relating to BNPL products and CCs. A summary of the characteristics of the sample in these regressions is presented in Table 2.

Table 2.

Summary statistics.

	Mean (SD)
Used Afterpay (Yes 1, No 0)	15.8%
Used credit card	20.5%
Used both	6.5%
Likelihood BNPL use (1– Extremely Unlikely to 5 – Extremely Likely)	2.2 (1.3)
Likelihood credit card use (1– Extremely Unlikely to 5 – Extremely Likely)	3.1 (1.4)
Unit	47.6%
Financial literacy (score /13)	8.6
Male	45.9%
Age	21.1 years (4.9)
Home	67.8%
International	7.8%
Commerce major	30.0%
Work	67.8%
Personal savings orientation (PSO) (average 10 items scored 1 (low) to 7 (high))	5.3 (0.9)
Propensity long-term planning (LTP) (average 6 items scored 1 (low) to 6 (high))	3.6 (1.2)
First enrolled 2017	14.7%
First enrolled 2018	21.1%
First enrolled 2019	64.2%
Total	525

SD: standard deviation; BNPL: buy-now-pay-later.

This table provides summary statistics of the sample. The three ‘Used’ variables indicate whether respondent has or currently uses each product or both. The ‘Likelihood BNPL/CC use’ variables were responses to: ‘How likely is it that you will use the following in the future? Buy-now, pay-later’ options (e.g. zipPay, Afterpay); credit card. Unit is the proportion of students that had enrolled in the personal finance elective unit. The Financial literacy score is the number of correct answers to the 13-items in the scale by Fernandes et al. (2014). Home is an indicator of whether the respondent lives with parents; and International is an indicator whether the respondent is an international student. Commerce major is whether the student is enrolled with a first or second major in Commerce. Work is an indicator of any part or full-time work. PSO is the score from the nine-item scale by Dholakia et al. (2016). LTP is the score from the study by Lynch et al. (2010). ‘First enrolled’ reflects what year the student first enrolled at the university.

4.2. Belief estimations

We expect respondents’ beliefs about the benefits and risks of BNPL and CC to be correlated and, therefore, estimate a multivariate multiple regression of respondents’ benefit and risk factor scores for each product and the set of variables discussed above. That is, we estimate a single regression model with the four outcomes. Table 3 presents these estimates. Male students have significantly higher benefit scores for CCs but otherwise there are no significant differences by gender. There are no significant differences by age, whether the student lives at home, or is an international student. Commerce majors have significantly higher benefit scores and lower risk scores for BNPL products. This may reflect that commerce students have a more positive orientation to new financial products compared to students in other disciplines. However, there are no significant differences for CC belief scores by student major. Those with part-time work have significantly lower scores for both the benefits and risks of CCs. This may reflect that the income from part-time work reduces the need and perceived benefit of short-term credit and conversely provides confidence that debts can be repaid, hence lower scores. Those who had enrolled in the personal finance unit are not significantly different in the benefit or risk scores except that they have a lower score on the BNPL risk factor.

Table 3.

Buy-now-pay-later and credit card belief regressions.

	BNPL	BNPL	CC	CC
	Benefits	Risks	Benefits	Risks
Unit	−0.0499 (0.1093)	−0.2687** (0.1090)	0.0972 (0.1088)	0.0524 (0.1111)
Financial literacy	−0.1372*** (0.0511)	0.1082** (0.0509)	−0.1338*** (0.0508)	−0.0042 (0.0519)
Male	−0.0151 (0.0914)	−0.0386 (0.0911)	0.2251** (0.0910)	−0.0460 (0.0929)
Age	0.0138 (0.0097)	−0.0147 (0.0097)	0.0061 (0.0097)	−0.0145 (0.0099)
Home	0.0479 (0.1037)	−0.1103 (0.1034)	0.0342 (0.1032)	0.0343 (0.1054)
International	−0.1298 (0.1719)	0.0113 (0.1714)	0.2229 (0.1711)	0.0166 (0.1747)
Commerce major	0.3021*** (0.0944)	−0.1883** (0.0941)	0.1137 (0.0940)	0.0030 (0.0960)
Work	−0.1142 (0.0977)	0.0282 (0.0974)	−0.2426** (0.0972)	−0.3003*** (0.0993)
PSO	−0.1002* (0.0564)	0.1020* (0.0563)	−0.0515 (0.0562)	0.0063 (0.0574)
LTP	0.1060** (0.0420)	−0.0021 (0.0419)	0.1499*** (0.0418)	0.0531 (0.0427)
PSO×LTP	0.0037 (0.0377)	0.1046*** (0.0376)	−0.0567 (0.0375)	0.0952** (0.0383)
2017 (2019 base)	0.0148 (0.1323)	−0.0228 (0.1319)	−0.0100 (0.1316)	−0.0370 (0.1345)
2018	0.0169 (0.1164)	−0.1280 (0.1161)	0.0883 (0.1158)	−0.0303 (0.1183)
Constant	−0.2194 (0.2637)	0.5698** (0.2629)	−0.1530 (0.2624)	0.5008* (0.2681)
Obs	525
R ²	0.0589	0.0499	0.0804	0.0417

BNPL: buy-now-pay-later; CC: credit card; PSO: personal savings orientation scale; LTP: long-term planning.

This table presents estimates from a multivariate ordinary least squares (OLS) regression of product beliefs (benefits and risks) for each product types were jointly estimated. Each outcome is the regression score for each factor as discussed in Section 4.1. That is, we estimate a single regression model with the four outcomes. The outcome in each case is the predicted factor score, mean zero and standard deviation of one. Robust standard errors are shown in parentheses. Significance is indicated by ***p < 0.01, **p < 0.05, *p < 0.1.

4.2.1. Financial literacy

More financially literate students score the benefits of both BNPL and CC lower and there is no significant difference in the size of the effect between BNPL and CC. In contrast, financial literacy is positively associated with the risk factor scores for BNPL, but not significantly related to risk scores for CCs. Only the more financially literate score the risk of overspending and damage to financial health higher for BNPL. This is potentially problematic from a policy perspective as it suggests that responsible use, that is, awareness of benefits and risks, is conditional on adequate financial literacy. A non-significant financial literacy coefficient for CC risks may reflect that CCs are perceived more as a means of payment and not a debt source, or that the terms and risks of a CC are more widely understood.

4.2.2. Savings and planning orientations

BNPL benefits scores are significantly higher for those with a higher propensity to plan (LTP) and lower for those with higher savings orientation (PSO). The former aligns with expectations and with the marketing of BNPL products that it provides consumers with a means of budgeting and planning for larger expenditures (Afterpay, 2020a). The latter aligns with an interpretation of PSO as an orientation to SNBL rather than BNPL. Those with a savings orientation rate the proposed benefits of BNPL lower. CC benefit scores are also significantly higher for those with a higher propensity to plan²⁵ with no significant relationship with PSO. The former is consistent with CCs providing the same function for planners that BNPL does, and hence scores are higher for those with a planning orientation.

The expectation of a positive PSO and risk score relationship, reflecting an aversion to debt, is found for BNPL but not for CCs. For BNPL, at average LTP scores (LTP and PSO scores are mean centred), the PSO coefficient is significant (weakly) and positive. The positive interaction effect suggests that those who are both more inclined to plan than average and those with a higher savings orientation than average, score the risks higher. For CCs, a similar pattern is observed. At average levels of LTP, PSO is not significant; but at above average levels of both PSO and LTP, risks are scored higher.

4.2.3. Summary of predictors of attitudes towards BNPL and CC

In summary, both knowledge (financial literacy) and dispositions (planning and savings orientations) are significantly related to consumer beliefs about BNPL and CCs. Financial literacy significantly tempers beliefs about the benefits of both products. Risk beliefs appear independent of financial literacy for CCs but positively related to BNPL. Those who are stronger savers do not rate benefits as high and rate risks higher for BNPL. Interestingly, the combination of being above average level savings orientation and above average planning propensity is significantly positively related to risks for both products. In isolation, planning propensity is not significant in risk scores.

4.3. Intentions to use BNPL and CCs

Next, we examine the intention to use BNPL and CCs in the future. Given the response scale provides an ordered ranking (1 – Extremely Unlikely to 5 – Extremely Likely), ordered probit models were estimated. We expect the intention to use the products to be correlated and hence we estimate a bivariate ordered probit regression. The bivariate model is a natural extension of the univariate ordered probit and assumes a bivariate normal distribution of errors. Each estimation includes the same set of respondent characteristics as in the previous section. As per the structural model in Figure 3, we also include respondents’ predicted factor scores for benefits and risks beliefs as key predictors of intention to use each product.

Overall, students reported a higher likelihood to use a CC compared with BNPL (3.2 vs 2.2). Table 4 presents estimated marginal effects for each response for intended use of BNPL and CC in the future. The estimated correlation between residuals is positive (0.36) and significant, supporting a joint estimation strategy. The unconditional predicted probability for each response is shown at the top of each column. When interpreting the marginal effects estimates for a product, note that the sum of the marginal effects across the five responses for any variable is zero reflecting that the outcomes are mutually exclusive.

Table 4.

Estimated marginal effects on likelihood of future buy-now-pay-later and credit card use.

	BNPL					Credit cards
	Extremely unlikely	Unlikely	Neither	Likely	Extremely likely	Extremely unlikely	Unlikely	Neither	Likely	Extremely likely
	0.3936	0.2526	0.1362	0.1438	0.0737	0.2094	0.1530	0.1830	0.2749	0.1796
Unit	−0.0141(0.0364)	0.0008(0.0020)	0.0029(0.0076)	0.0056(0.0144)	0.0048(0.0125)	−0.0122(0.0281)	−0.0037(0.0085)	−0.0009(0.0022)	0.0057(0.0130)	0.0112(0.0257)
Financial literacy	0.0252(0.0181)	−0.0014(0.0011)	−0.0052(0.0038)	−0.0100(0.0072)	−0.0086(0.0063)	−0.0089(0.0133)	−0.0027(0.0040)	−0.0007(0.0010)	0.0041(0.0062)	0.0082(0.0121)
Male	−0.0228(0.0292)	0.0012(0.0016)	0.0047(0.0061)	0.0090(0.0116)	0.0078(0.0101)	0.0574** (0.0227)	0.0167** (0.0068)	0.0040** (0.0020)	−0.0261** (0.0106)	−0.0520** (0.0204)
Age	−0.0038(0.0035)	0.0002(0.0002)	0.0008(0.0007)	0.0015(0.0014)	0.0013(0.0012)	−0.0078** (0.0035)	−0.0023** (0.0011)	−0.0006* (0.0003)	0.0036** (0.0016)	0.0072** (0.0032)
Home	−0.0397(0.0369)	0.0028(0.0033)	0.0084(0.0080)	0.0154(0.0143)	0.0132(0.0117)	−0.0404(0.0277)	−0.0113(0.0074)	−0.0022(0.0014)	0.0187(0.0130)	0.0352(0.0230)
International	−0.0730(0.0487)	0.0003(0.0030)	0.0143(0.0091)	0.0301(0.0212)	0.0284(0.0212)	−0.0297(0.0429)	−0.0097(0.0154)	−0.0032(0.0062)	0.0132(0.0184)	0.0294(0.0460)
Commerce major	−0.0259(0.0299)	0.0012(0.0013)	0.0054(0.0062)	0.0104(0.0121)	0.0090(0.0105)	−0.0626*** (0.0220)	−0.0207** (0.0082)	−0.0069* (0.0035)	0.0283*** (0.0100)	0.0619*** (0.0235)
Work	0.0211(0.0337)	−0.0010(0.0015)	−0.0044(0.0069)	−0.0084(0.0136)	−0.0073(0.0118)	−0.0127(0.0245)	−0.0037(0.0070)	−0.0009(0.0016)	0.0059(0.0113)	0.0115(0.0218)
2017 (2019 base)	0.0346(0.0452)	−0.0021(0.0035)	−0.0073(0.0097)	−0.0136(0.0176)	−0.0117(0.0146)	0.0506(0.0375)	0.0144(0.0098)	0.0028* (0.0017)	−0.0241(0.0184)	−0.0438(0.0297)
2018	0.0433(0.0365)	−0.0029(0.0031)	−0.0091(0.0080)	−0.0169(0.0142)	−0.0143(0.0117)	0.0431(0.0296)	0.0126(0.0083)	0.0028(0.0018)	−0.0204(0.0144)	−0.0381(0.0248)
BNPL benefits	−0.2616*** (0.0146)	0.0142** (0.0063)	0.0543*** (0.0071)	0.1035*** (0.0105)	0.0895*** (0.0104)	0.0097(0.0137)	0.0029(0.0041)	0.0007(0.0011)	−0.0044(0.0063)	−0.0088(0.0126)
BNPL risks	0.0328* (0.0168)	−0.0018(0.0012)	−0.0068* (0.0036)	−0.0130* (0.0068)	−0.0112* (0.0058)	−0.0061(0.0142)	−0.0018(0.0042)	−0.0005(0.0011)	0.0028(0.0065)	0.0056(0.0129)
CC benefits	0.0319* (0.0186)	−0.0017(0.0013)	−0.0066* (0.0040)	−0.0126* (0.0074)	−0.0109* (0.0063)	−0.1502*** (0.0136)	−0.0448*** (0.0063)	−0.0114*** (0.0039)	0.0691*** (0.0084)	0.1374*** (0.0129)
CC risks	−0.0018(0.0173)	0.0001(0.0009)	0.0004(0.0036)	0.0007(0.0069)	0.0006(0.0059)	0.0115(0.0135)	0.0034(0.0040)	0.0009(0.0011)	−0.0053(0.0063)	−0.0105(0.0123)
BNPL preferred	−0.0160(0.0108)	0.0009(0.0007)	0.0033(0.0023)	0.0063(0.0043)	0.0055(0.0037)	−0.0008(0.0080)	−0.0002(0.0024)	−0.0001(0.0006)	0.0004(0.0037)	0.0008(0.0073)
CC preferred	0.0038(0.0116)	−0.0002(0.0006)	−0.0008(0.0024)	−0.0015(0.0046)	−0.0013(0.0040)	−0.0312*** (0.0085)	−0.0093*** (0.0027)	−0.0024** (0.0011)	0.0144*** (0.0040)	0.0286*** (0.0080)
Obs, Log_Likelihood, rho		525, −1310.3, 0.3621***

BNPL: buy-now-pay-later; CC: credit card.

This table presents estimates from two jointly estimated (bivariate) ordered probit regressions of the likelihood of using buy-now-pay-later (BNPL) and credit cards (CC) in the future. The joint estimation incorporates the correlation of residuals of the two ordered probits, and an estimate of this correlation (rho) is reported in the last row. Each column presents the average marginal effects for the probability of each of the five ordered responses (1 – Extremely Unlikely to 5 – Extremely Likely). Columns 2–6 present the marginal effects for BNPL responses and columns 7–11 the marginal effects for CC responses. When interpreting the marginal effect estimates for a product, the sum of the marginal effects across the five responses for any variable is zero reflecting that the outcomes are mutually exclusive. The unconditional predicted probability for each rating is presented at the top of each column. Robust standard errors are shown in parentheses. Significance is indicated by ***p < 0.01, **p < 0.05, *p < 0.1.

For both BNPL and CCs, higher benefit beliefs positively predict a higher likelihood of future use, though there is a difference in the size and pattern of effect. In terms of size, a unit increase in BNPL benefits reduces the probability of an Extremely Unlikely response for BNPL by 26% points. A comparable increase in beliefs about CC Benefits reduces the probability of an Extremely Unlikely response for CC by a statistically significant 15% points.²⁶ In terms of the pattern, for BNPL, the negative marginal effect for the Extremely Unlikely response is not offset solely by an increase in Likely responses. The Unlikely and Neither responses also have positive marginal effects. Contrast this with CCs where increase in CC benefits has a negative marginal effect not only on each Unlikely response and Neither. This could be interpreted as there being less certainty in likelihood of future use, notwithstanding perceived BNPL benefits, and greater certainty in intended future use of a CC, given increased benefit beliefs. In contrast, risk beliefs do not appear important to intended future use. The negative marginal effect of BNPL risks is (weakly) significant, for the Likely responses. CC risks are not significant for intended future CC use.

In the case of BNPL, beliefs appear to fully mediate the role of financial literacy and all other controls. That is, there is no direct role for financial literacy to future use once beliefs are included, which is also the case for CCs. There is a significant gender difference, but only for CCs. Males have a lower (higher) probability of answering they are likely (unlikely) to use CC in the future. This is consistent with US evidence (Li, 2018), using a comparable young and educated sample of consumers, which found that females had more intensive use of CCs. An investigation of HILDA data using the 2016 wave and restricting to those aged 20–25 years indicates that 21% had ‘credit cards, charge cards or store accounts’. A linear probability regression using the HILDA data (not tabulated) of ownership of CCs accounts as a function of gender, age and financial literacy indicates no significant gender difference.

Social norms were assessed by respondent ratings of whether they agreed that their friends preferred each product ( ‘Do not agree at all – 0’ to ‘Completely Agree – 5’). The mean rating of 2.2 for each was low, reflecting the low use rate of each product in the sample. For BNPL, there was no significant relationship (BNPL preferred); but for credit cards (CC preferred), a small positive effect was estimated.

A limitation of the estimated marginal effects for likelihood of future use reported in Table 4 is that they are unconditional on the intention to use the other product. To investigate this, those responding with a Likely or Extremely Likely rating were collapsed to one Likely group with remaining ratings grouped as Not Likely. Given the two dichotomous outcomes for each product, and consistent with the evidence of a correlation between behaviours, a bivariate probit model was estimated. Marginal effects were then estimated for each of the four possible permutations, which allows investigation of the role of a variable for one product conditional on use of the other. The four permutations are: Both BNPL and CC Likely, BNPL Likely only, CC Likely only and both Unlikely. Table 5 presents the average marginal effects estimates with the predicted probability for each combination of intended use shown at the top of each column. When interpreting the marginal effect estimates, the sum of the marginal effects across the four outcomes is zero reflecting that the outcomes are mutually exclusive.

Table 5.

Marginal effects on future buy-now-pay-later and credit card use combinations.

	BothBNPL and CC	BNPLonly	CConly	Neither
	0.1366	0.0813	0.3222	0.4598
Unit	0.0223(0.0263)	0.0018(0.0178)	0.0165(0.0416)	−0.0406(0.0433)
Financial literacy	−0.0072(0.0120)	−0.0199** (0.0084)	0.0492** (0.0195)	−0.0222(0.0203)
Male	−0.0231(0.0213)	0.0166(0.0148)	−0.0691** (0.0331)	0.0755** (0.0351)
Age	0.0038(0.0024)	−0.0005(0.0016)	0.0050(0.0039)	−0.0083** (0.0041)
Home	0.0274(0.0238)	−0.0058(0.0175)	0.0451(0.0387)	−0.0667(0.0408)
International	0.0304(0.0436)	0.0215(0.0318)	−0.0293(0.0664)	−0.0227(0.0694)
Commerce major	0.0303(0.0240)	−0.0024(0.0140)	0.0343(0.0341)	−0.0622* (0.0375)
Work	−0.0205(0.0249)	−0.0150(0.0167)	0.0219(0.0370)	0.0136(0.0388)
2017 (2019 base)	−0.0240(0.0306)	−0.0003(0.0197)	−0.0219(0.0477)	0.0462(0.0546)
2018	−0.0493** (0.0252)	–0.0099(0.0177)	−0.0277(0.0430)	0.0870* (0.0471)
BNPL benefits	0.1093*** (0.0133)	0.0749*** (0.0100)	−0.1052*** (0.0183)	−0.0789*** (0.0198)
BNPL risks	−0.0017(0.0124)	–0.0075(0.0074)	0.0195(0.0168)	−0.0103(0.0198)
Credit card benefits	0.0318*** (0.0122)	–0.0534*** (0.0093)	0.1821*** (0.0189)	−0.1605*** (0.0185)
Credit card risks	−0.0019(0.0121)	0.0012(0.0078)	−0.0051(0.0181)	0.0058(0.0196)
BNPL preferred	0.0097(0.0071)	0.0056(0.0048)	−0.0063(0.0113)	−0.0090(0.0119)
CC preferred	0.0056(0.0081)	–0.0110** (0.0055)	0.0365*** (0.0122)	−0.0312** (0.0127)
Obs	525
Log_Likelihood	−462.6
Rho	0.3956***

BNPL: Buy-Now-Pay-Later; CC: credit card.

This table presents estimates from a bivariate probit regression of the likelihood of using buy-now-pay-later (BNPL) and, separately, credit cards (CC) in the future. The likelihood of future use of each product is recoded to a binary yes (likely or extremely likely) and no (neither, unlikely, extremely unlikely). Combined, the two probits produce four mutually exclusive groups: those who are likely to use both (column two); those likely to only use BNPL (column three); those likely to only use CC (column four); and those likely to use neither (column five). The unconditional predicted probability for each group is presented at the top of each column together with average marginal effects on the probability of each of the four groups. When interpreting the marginal effect estimates, the sum of the marginal effects across the four outcomes is zero reflecting that the outcomes are mutually exclusive. The bivariate estimation reflects an expectation that expected BNPL and CC usage are correlated, and an estimate of this correlation is reported in the last row (rho). Robust standard errors are shown in parentheses. Significance is indicated by ***p < 0.01, **p < 0.05, *p < 0.1.

The results confirm the significant positive role of BNPL and CC benefit beliefs. The estimates also suggest that higher BNPL benefit beliefs are associated with a higher probability of intention to use both BNPL and CC together, and BNPL only, with a reduced likelihood of a CC only, and neither product. A similar result is observed for CC benefits. In addition, the role of controls is more refined. The gender difference reported in Table 4 for CCs is again observed. Males are significantly more likely to intend to use neither, and less likely to intend to use CCs only.

4.4. Prior BNPL and CC use

As noted in the ‘Method’ section, our analysis focuses on behaviour intention, given that intention along with perceived behavioural control are deemed sufficient predictors of future behaviour (Ajzen, 2011). Prior use of BNPL or CCs may also be influential in explaining future behaviour though prior behaviour is usually better ‘considered a proxy for habit strength’ (Ajzen, 2011: 1120). For example, the more an individual uses BNPL or a CC the more the behaviour may become habitual, under their control and a likely future behaviour. However, more interesting is the role of experience on intentions (Ajzen, 2011), particularly in the case of an emerging product such as BNPL and from a regulatory perspective as the drivers of future use may be different. We, therefore, re-estimate the intention regression (Table 5) separately for those with and without experience with either product to examine whether the role of financial literacy and perceived product benefits and risks, given experience.

Table 6 presents the estimate marginal effects. The top row of the table presents the unconditional probability for each of the combinations and, as expected, prior experience is a good predictor of future use. For example, 58% who have not used either product previously do not expect to use either product in the future, whereas the proportion is 18% for those with experience. The estimated rho is almost twice as large (0.4998) for the estimation using those with no prior experience than for those with prior experience (0.2754). This suggests future use of both is more strongly correlated for those without experience relative to those with experience. In terms of the key explanatory variables, the positive marginal effect of financial literacy on use of CC only reported in Table 5 is isolated to those who have no experience as reported in Table 6. Similarly, the negative financial literacy marginal effect for the BNPL only group is only significant for those with no experience. Across both groups, perceived benefits are better predictors of future use rather than perceived risks of the products. The significant marginal effects estimated for the perceived benefits of each product are robust across experience groups. For example, higher perceived benefits of BNPL arrangements are associated with a higher probability of future BNPL use and a lower probability for CC use (and vice versa for CC benefits).

Table 6.

Marginal effects on buy-now-pay-later and credit card use combinations by product experience.

	Have no prior experience with either				Have prior experience with either
	BothBNPL and CC	BNPLonly	CConly	Neither	BothBNPL and CC	BNPLonly	CConly	Neither
	0.0762	0.0364	0.3072	0.5801	0.2867	0.1849	0.3507	0.1777
Unit	0.0004(0.0257)	−0.0116(0.0143)	0.0659(0.0505)	−0.0547(0.0542)	−0.0839(0.0734)	0.0399(0.0471)	−0.0342(0.0701)	0.0781(0.0570)
Financial literacy	−0.0081(0.0120)	−0.0142** (0.0069)	0.0575** (0.0252)	−0.0352(0.0279)	0.0456(0.0300)	−0.0157(0.0191)	0.0100(0.0289)	−0.0399* (0.0225)
Male	−0.0086(0.0199)	0.0082(0.0121)	−0.0639(0.0424)	0.0642(0.0453)	−0.0626(0.0579)	0.0505(0.0334)	−0.0557(0.0494)	0.0678(0.0440)
Age	−0.0047(0.0060)	−0.0002(0.0030)	−0.0099(0.0100)	0.0148(0.0124)	−0.0013(0.0049)	−0.0081** (0.0034)	0.0121** (0.0051)	−0.0028(0.0038)
Home	0.0580*** (0.0203)	0.0220(0.0137)	0.0497(0.0502)	−0.1297** (0.0519)	0.0355(0.0609)	−0.0419(0.0383)	0.0513(0.0593)	−0.0448(0.0462)
International	0.0591(0.0604)	0.0859(0.0535)	−0.1447** (0.0735)	−0.0003(0.1044)	0.0194(0.1015)	−0.0772(0.0472)	0.1195(0.0984)	−0.0617(0.0663)
Commerce major	0.0259(0.0238)	0.0096(0.0134)	0.0065(0.0441)	−0.0420(0.0481)	−0.0308(0.0596)	−0.0711* (0.0381)	0.1127** (0.0553)	−0.0108(0.0453)
Work	−0.0262(0.0295)	−0.0146(0.0163)	0.0171(0.0459)	0.0238(0.0527)	0.0099(0.0582)	−0.0213(0.0378)	0.0281(0.0566)	−0.0167(0.0455)
2017 (2019 base)	0.0019(0.0291)	−0.0015(0.0174)	0.0121(0.0661)	−0.0125(0.0674)	−0.1523(0.0928)	0.0325(0.0463)	−0.0110(0.0673)	0.1307(0.0961)
2018	−0.0138(0.0234)	−0.0078(0.0135)	0.0096(0.0542)	0.0119(0.0578)	−0.2059*** (0.0635)	0.0149(0.0436)	−0.0009(0.0661)	0.1919** (0.0778)
BNPL benefits	0.0765*** (0.0155)	0.0454*** (0.0109)	−0.0663** (0.0271)	−0.0556* (0.0296)	0.0951*** (0.0336)	0.0896*** (0.0184)	−0.1579*** (0.0264)	−0.0269(0.0260)
BNPL risks	0.0059(0.0119)	−0.0016(0.0065)	0.0222(0.0229)	−0.0265(0.0262)	−0.0297(0.0291)	−0.0016(0.0166)	0.0107(0.0258)	0.0206(0.0225)
CC benefits	0.0185(0.0113)	−0.0258*** (0.0083)	0.1826*** (0.0237)	−0.1752*** (0.0237)	0.0646* (0.0331)	−0.1341*** (0.0225)	0.1775*** (0.0341)	−0.1080*** (0.0286)
CC risks	−0.0001(0.0121)	0.0018(0.0072)	−0.0099(0.0252)	0.0083(0.0261)	0.0343(0.0289)	0.0049(0.0170)	−0.0169(0.0262)	−0.0223(0.0224)
BNPL preferred	0.0051(0.0066)	0.0019(0.0039)	0.0016(0.0140)	−0.0086(0.0149)	0.0310(0.0208)	0.0216* (0.0128)	−0.0404** (0.0195)	−0.0122(0.0156)
CC preferred	0.0091(0.0079)	−0.0036(0.0047)	0.0407*** (0.0152)	−0.0462*** (0.0153)	−0.0174(0.0226)	−0.0176(0.0152)	0.0307(0.0227)	0.0044(0.0170)
Obs	368				155
Log_Likelihood	−280.9				−146.6
rho	0.4998***				0.2754***

BNPL: buy-now-pay-later; CC: credit card.

This table re-estimates Table 5 separately for those with and without experience with BNPL or a CC. Estimates from a bivariate probit regression of the likelihood of using buy-now-pay-later (BNPL) and, separately, credit cards (CC) in the future. The likelihood of future use of each product is recoded to a binary yes (likely or extremely likely) and no (neither, unlikely, extremely unlikely). Combined, the two probits produce four mutually exclusive groups: those who are likely to use both (columns two and six); those likely to only use BNPL (column three and seven); those likely to only use CC (column four and eight); and those likely to use neither (column five and nine). The unconditional predicted probability for each group is presented at the top of each column together with average marginal effects on the probability of each of the four groups. When interpreting the marginal effects estimates (within either the no prior or prior experience estimation), the sum of the marginal effects across the four outcomes is zero. The joint estimation reflects an expectation that expected BNPL and CC usage are correlated, and an estimate of this correlation is reported in the last row (rho). Robust standard errors are shown in parentheses. Significance is indicated by ***p < 0.01, **p < 0.05, *p < 0.1.

4.5. Robustness

The survey responses were collected from students who responded to a follow-up survey. Whereas the initial survey was completed by those in the unit and peers in the broader university, there is a potential selection bias given the choice of responding to an additional survey invitation. To explore the robustness of results, a probit regression was estimated of those students who participated in the follow-up survey and those who did not. For the original survey, students who completed the survey were able to elect to enter a draw to win a gift card. Students were given the option that, in the event they were randomly selected, that they could either receive the gift card or donate the equivalent gift card value to a nominated charity. This choice was expected to be related to the willingness to complete a follow-up survey but not related to BNPL and CC beliefs, intentions or use and, therefore, a useful exclusion restriction. It was expected that those who selected the donation were less motivated by the gift card prize and, therefore, less likely to respond to a follow-up survey. The first-stage regression is presented in the Appendix Table A2. The prize variable is significant and negative. An Inverse Mills ratio was estimated and included in each of the models which were re-estimated with bootstrapped standard errors. Each table in the supplementary file appendix has a corresponding estimated table to the main article. The Inverse Mills ratios are not significant in the estimations, with one exception where it is weakly significant for one of the four beliefs. We, therefore, rely on the estimations reported in the main text.

We acknowledge that the causal direction flowing from financial literacy to use and intention is not supportable, given the data, as is the case for other variables, given the cross-sectional data. As noted in the introduction, the estimated model does not incorporate feedback loops, for example, that financial literacy increases with use.

5. Summary and conclusion

BNPL is an example of a Fintech innovation that has combined a simple financial product (payment instalments) with novel delivery (online and in store) to create a significant disruption to short-term credit products. The dominant BNPL business model offers a product that does not charge interest, which enables the arrangement to avoid requirements applying to other similar products such as credit checks. In doing so, questions have been raised as to whether BNPL arrangements lead to over-indebtedness, given that the suitability of the product for the consumer is not directly assessed at the individual consumer level. BNPL providers argue that their product is a proactive budgeting tool for consumers, which drives better outcomes than alternatives, such as CCs. Furthermore, BNPL providers argue that they enable credit for those who otherwise have limited credit access and are, therefore, more vulnerable to more problematic debt such as payday loans.

We reviewed the overlapping responsibilities that are implicit and explicit in more traditional short-term credit products (CCs) and the new Fintech enabled BNPL. This review identified that for ‘traditional’ credit products that currently fall under the NCCP Act, an explicit responsibility attaches to the credit provider who must apply for a credit licence. The regulator has a direct responsibility in this case over the conduct of the credit provider, which can extend to responsible lending laws. We also noted that just as new financial products such as BNPL emerge over time, so does the legislative and regulatory mix used, which reflects an ongoing tension in ensuring responsible behaviour by all BNPL participants.

Consumers’ responsibility requires at least honest engagement with the credit provider. With BNPL arrangements, more responsibility has been shifted to consumers requiring responsible spending and borrowing. BNPL providers avoid the need to assess the suitability of the product for a specific consumer but argue that this is effectively achieved through algorithms applied in real-time approvals. On one hand, the regulator’s direct responsibility is reduced, given that BNPL products fall outside the scope of the NCCP, but has increased, given powers to intervene when a financial product such as BNPL ‘has resulted, will result or is likely to result in significant consumer detriment’ (ASIC, 2020b).

Empirical results based on data collected from a young demographic of BNPL consumers focussed on beliefs about BNPL and CCs as a basis of attitudes towards the products and their intention to use, and actual use of the products. We found financial literacy was associated with lower perceived benefits of both products and higher perceived risks of BNPL arrangements. This is problematic, given that the primary users of BNPL arrangements are younger adults (ASIC, 2018b) who are the age group with the lowest financial literacy (Wilkins and Lass, 2018). We also note that this finding is from a sample of undergraduate students, which have higher average financial literacy than their peers not pursuing higher education (Wilkins and Lass, 2018).

The empirical analysis also identified that personal savings orientation, or a preference for SNBL, reduced perceived benefits of BNPL. Propensity to plan was also significant in explaining preference for BNPL, consistent with what BNPL providers advertise. Perceived risks for BNPL was higher for those with a higher propensity to plan and a higher personal savings orientation. The results for financial literacy and savings and planning orientations suggest means by which the regulator can exercise its responsibility indirectly. ASIC’s national financial capability strategy (Financial Capability, 2018) includes an objective to assist consumers in ‘making informed money decisions’.

There are limitations in our analysis which affect the generalisability of results. Our sample of young, higher educated, consumers is expected to be an attractive group to financial service providers (Milner and Rosenstreich, 2013). This is also true when it comes to BNPL arrangements, which makes the sample an important segment to consider. However, undergraduate students remain ‘an admittedly special population’ (Brugiavini et al., 2015) in that they are more educated, less attached to the workplace, and more attached to parents with the financial protections this can provide. They are also expected to be more comfortable with the mobile and online technology favoured by BNPL providers (see study by Caddy et al., 2020), which may skew perceptions of benefits and risks.

Further questions remain that can be addressed in future research. First, while BNPL has opened a new access to credit for some, it remains unclear for how many this becomes problematic in the absence of suitability assessments. Second, it is unclear how consumers prioritise BNPL over other competing short-term obligations, given that 23% use a CC to make a repayment (ASIC, 2018b). Finally, we find evidence that there is some substitution of BNPL for CCs, consistent with suggestions that Millennials have a different attitude to debt use. More investigation into the role of debt aversion in the preference for BNPL across age groups could provide insight into whether BNPL will remain a sustainable innovation.

Supplemental Material

sj-docx-1-aum-10.1177_03128962211032448 – Supplemental material for Fintech and responsibility: Buy-now-pay-later arrangements

Supplemental material, sj-docx-1-aum-10.1177_03128962211032448 for Fintech and responsibility: Buy-now-pay-later arrangements by Paul Gerrans, Dirk G Baur and Shane Lavagna-Slater in Australian Journal of Management

Footnotes

Final transcript accepted 17 June 2021 by Millicent Chang (AE Finance).

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Paul Gerrans

Dirk G Baur

Supplemental material

Supplemental material for this article is available online.

Notes

References

Afterpay (2020a) Afterpay fact sheet. Available at: https://www.afterpay.com/attachment/44/show

Afterpay (2020b) Afterpay is committed to responsible spending. Available at: https://www.afterpay.com/en-AU/responsible-spending

Afterpay (2020c) Partner with the most loved way to pay. Available at: https://www.afterpay.com/en-AU/business

Afterpay (2020d) Submission (83) to the Senate Select Committee inquiry on Financial Technology (FinTech) and Regulatory Technology (RegTech). Available at: https://www.aph.gov.au/DocumentStore.ashx?id=79fc0dda-c2ad-4930-9192-5a03a726df16&subId=675279

Afterpay (2020e) Submission from Afterpay to the Reserve Bank of Australia (RBA) Review of Retail Payments Regulation: Issues Paper (RBA Review). Available at: https://www.afterpaytouch.com/images/06022020-Afterpay-Submission-to-RBA-Issues-Paper-on-Retail-Payments-Regulation.pdf

Afterpay Touch Group (2019a) AUSTRAC Update – Final audit report. Available at: https://www.afterpaytouch.com/images/25112019—AUTRAC-Update—Final-Audit-Report.pdf

Afterpay Touch Group (2019b) Joint consumer submission on the proposed Design and Distribution Obligations and Product Intervention Powers. Available at: https://www.afterpaytouch.com/images/21012020-Afterpay-submission-to-Senate-FinTech-and-RegTech-inquiry.pdf

Afterpay Touch Group (2019c) Afterpay Touch Group Limited ASX announcement. Available at: https://www.afterpaytouch.com/images/1936384.pdf

Agarwal

Driscoll

Gabaix

, et al. (2009) The age of reason: Financial decisions over the life-cycle with implications for regulation. Brookings papers on economic activity. Available at: https://www.brookings.edu/wp-content/uploads/2016/07/2009b_bpea_agarwal.pdf

10.

Ajzen

(1985) From intentions to actions: A theory of planned behavior. In: Kuhl

Beckman

(eds) Action-Control: From Cognition to Behavior. Heidelberg: Springer, pp. 11–39.

11.

Ajzen

(1991) The theory of planned behavior. Organizational Behavior and Human Decision Processes 50: 179–211.

12.

Ajzen

(2011) The theory of planned behaviour: Reactions and reflections. Psychology & Health 26: 1113–1127.

13.

Ajzen

(2015) The theory of planned behaviour is alive and well, and not ready to retire: A commentary on Sniehotta, Presseau, and Araújo-Soares. Health Psychology Review 9: 131–137.

14.

Ambuehl

Ockenfels

(2017) The ethics of incentivizing the uninformed: A vignette study. American Economic Review 107: 91–95.

15.

Australian Bureau of Statistics (2019) Household Income and Wealth, Australia, 2017-18 (Cat. No. 6523.0). Available at: https://www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/6523.02017-18?OpenDocument

16.

Australian Competition and Consumer Commission (ACCC) (2019) Lay-by agreements. Available at: https://www.accc.gov.au/consumers/contracts-agreements/lay-by-agreements

17.

Australian Finance Industry Association (2020) Code of practice for buy now pay later providers. Available at: https://www.afia.asn.au/s/AFIA-Web-Code-of-Practice-for-Buy-Now-Pay-Later-Providers_Final.pdf

18.

Australian Securities and Investments Commission (ASIC) (2015) Payday lenders and the new small amount lending provisions (Report 426). Available at: https://download.asic.gov.au/media/3038267/rep-426-published-17-march-2015.pdf

19.

Australian Securities and Investments Commission (ASIC) (2018a) Credit card lending in Australia (Report 580). Available at: https://download.asic.gov.au/media/4801724/rep580-published-4-7-2018.pdf

20.

Australian Securities and Investments Commission (ASIC) (2018b) Review of buy now pay later arrangements. Available at: https://download.asic.gov.au/media/4957540/rep600-published-07-dec-2018.pdf

21.

Australian Securities and Investments Commission (ASIC) (2020a) Buy now pay later: An industry update. Available at: https://asic.gov.au/regulatory-resources/find-a-document/reports/rep-672-buy-now-pay-later-an-industry-update/

22.

Australian Securities and Investments Commission (ASIC) (2020b) Product intervention power (Regulatory Guide 272). Available at: https://download.asic.gov.au/media/5633261/rg272-published-17-june-2020.pdf

23.

Australian Transaction Reports and Analysis Centre (AUSTRAC) (2019) Notice to appoint an external auditor to conduct compliance audit. Available at: https://www.austrac.gov.au/sites/default/files/2019-06/Afterpay_Pty_Ltd_-_section_162_notice.pdf

24.

Balyuk

(2019) Financial innovation and borrowers: Evidence from peer-to-peer lending. Available at: https://sites.google.com/site/tetyanabalyuk

25.

Brugiavini

Cavapozzi

Padula

, et al. (2015) Financial education, literacy and investment attitudes. SAFE Working Paper No. 86. Frankfurt: University of Venice and SAFECenter, University of Frankfurt.

26.

Caddy

Delaney

Fisher

, et al. (2020) Consumer payment behaviour in Australia. RBA Bulletin, March, pp. 9–19.

27.

Commonwealth Bank (2020) What is buy now, pay later? Available at: https://www.commbank.com.au/guidance/consumer-finance/buy-now-pay-later-understanding-payment-programs-201802.html

28.

Commonwealth of Australia (2009) National Consumer Credit Protection Bill 2009, Explanatory Memorandum. Available at: https://www.legislation.gov.au/Details/C2009B00148/Explanatory%20Memorandum/Text

29.

Commonwealth of Australia (2014) Financial system inquiry final report. Available at: https://treasury.gov.au/publication/c2014-fsi-final-report

30.

Commonwealth of Australia (2016) Review of the small amount credit contract laws: Final report. Available at: https://treasury.gov.au/consultation/review-of-small-amount-credit-contracts-final-report

31.

Commonwealth of Australia (2018) Treasury Laws Amendment (Design and Distribution Obligations and Product Intervention Powers) Bill: Explanatory Memorandum. Available at: https://www.aph.gov.au/Parliamentary_Business/Bills_Legislation/Bills_Search_Results/Result?bId=r6184

32.

Commonwealth of Australia (2019a) Credit and Hardship: Report of the Senate Inquiry into Credit and Financial Products Targeted at Australians at Risk of Financial Hardship. Canberra, ACT, Australia: Senate Printing Unit, Parliament House.

33.

Commonwealth of Australia (2019b) Royal Commission into Misconduct in the Banking, Superannuation and Financial Services Industry, Final Report. Available at: https://financialservices.royalcommission.gov.au/Pages/reports.aspx#final

34.

Commonwealth of Australia (2019c) Restoring trust in Australia’s financial system: The Government response to the Royal Commission into Misconduct in the Banking, Superannuation and Financial Services Industry. Available at https://treasury.gov.au/sites/default/files/2019-03/FSRC-Government-Response-1.pdf

35.

Commonwealth of Australia (2020a) National Consumer Credit Protection Amendment (Supporting Economic Recovery) Bill 2020. Available at https://www.aph.gov.au/Parliamentary_Business/Bills_Legislation/Bills_Search_Results/Result?bId=r6656

36.

Commonwealth of Australia (2020b) National Consumer Credit Protection Amendment (Supporting Economic Recovery) Bill 2020: Exposure Draft Explanatory Memorandum. Available at https://treasury.gov.au/consultation/c2020-124502

37.

Consumer Action Law Centre (2021) Submission on National Consumer Credit Protection Amendment (Supporting Economic Recovery) Bill 2020. Available at https://consumeraction.org.au/wp-content/uploads/2021/02/Consumer-advocates-submission_National-Consumer-Credit-Protection-Amendment-Supporting-Economic-Recovery-Bill-2020-002.pdf

38.

Dholakia

Tam

Yoon

, et al. (2016) The ant and the grasshopper: Understanding personal saving orientation of consumers. Journal of Consumer Research 43: 134–155.

39.

Fernandes

Lynch

Netemeyer

(2014) Financial literacy, financial education, and downstream financial behaviors. Management Science 60: 1861–1883.

40.

Financial Capability (2018) National financial capability strategy. Available at: https://financialcapability.gov.au/strategy/#home

41.

Gerrans

Heaney

(2019) The impact of undergraduate personal finance education on individual financial literacy, attitudes and intentions. Accounting & Finance 59: 177–217.

42.

Kaiser

(1974) An index of factorial simplicity. Psychometrika 39: 31–36.

43.

Klarna (2019) Terms and conditions. Available at: https://www.klarna.com/uk/terms-and-conditions

44.

Laybuy (2019) Consumer terms and conditions. Available at: https://www.laybuy.com/au/consumer-terms

45.

(2018) Gender-related Differences in Credit Use and Credit Scores. FEDS Notes. Washington: Board of Governors of the Federal Reserve System. Available at: https://www.federalreserve.gov/econres/notes/feds-notes/gender-related-differences-in-credit-use-and-credit-scores-20180622.htm

46.

Loewenstein

(1987) Anticipation and the valuation of delayed consumption. The Economic Journal 97: 666–684.

47.

Longs

(2016) New methods of interpretation using marginal effects for nonlinear models. EUSMEX 2016: Mexican Stata Users Group. Available at https://www.stata.com/meeting/mexico16/slides/Mexico16_Long.pdf

48.

Lynch

Jr Netemeyer

Spiller

, et al. (2010) A generalizable scale of propensity to plan: The long and the short of planning for time and for money. Journal of Consumer Research 37: 108–128.

49.

Milner

Rosenstreich

(2013) Insights into mature consumers of financial services. Journal of Consumer Marketing 30: 248–257.

50.

Moradi-Motlagh

Jubb

(2020) Examining irresponsible lending using non-radial inefficiency measures: Evidence from Australian banks. Economic Analysis and Policy 66: 96–108.

51.

Paterson

Howell

(2018) Everyday consumer credit overview of Australian law regulating consumer home loans, credit cards and car loans (Royal Commission in Misconduct in the Banking, Superannuation and Financial Services Industry Background Paper 4). Available at: https://financialservices.royalcommission.gov.au/publications/Documents/australian-consumer-credit-protection-law-overview-background-paper-4.pdf

52.

Reserve Bank of Australia (2019) Review of retail payments regulation: Issues paper. Available at: https://www.rba.gov.au/payments-and-infrastructure/review-of-retail-payments-regulation/pdf/review-of-retail-payments-regulation-issues-paper-nov-2019.pdf

53.

von Gaudecker

H-M

(2015) How does household portfolio diversification vary with financial literacy and financial advice? The Journal of Finance 70: 489–507.

54.

Wilkins

Lass

(2018) The Household, Income and Labour Dynamics in Australia Survey: Selected Findings From Waves 1 to 16. Melbourne, Victoria: Melbourne Institute of Applied Economic & Social Research, University of Melbourne.

55.

Worthington

Whittaker

(2006) Scale development research: A content analysis and recommendations for best practices. The Counseling Psychologist 34: 806–838.

56.

Zip (2019) Buy now pay later policy recommendations. Available at: http://zipmoneylimited.com.au/files/Zip_Policy_Recommendations_Final.pdf

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB