Buy now, pay later consumer credit behavior: impacts on financing decisions

The buy now pay later (BNPL) model represents a rapidly evolving trend within the financial sector, capturing the interest of scholars, policymakers and industry experts alike (Alcazar and Bradford, 2021a; Alcazar and Bradford, 2021b; Fisher et al., 2021; Di Maggio et al., 2022; Powell et al., 2023; Filotto et al., 2024). This credit mechanism is becoming increasingly popular among consumers, notably among the younger demographic (Gobbi, 2022; Relja et al., 2024). The surge in attention toward BNPL coincides with the growth of online shopping, a trend accelerated by the COVID-19 pandemic-induced movement restrictions (Relja et al., 2024). According to Cervellati et al. (2022), the consequences on the agents’ economic behavior caused by COVID-19 have changed people’s payment preferences, especially in Italy. Filotto et al. (2024) highlight how BNPL services have become popular in Italy and the preference for that payment service has shifted Italians’ payment behavior from credit cards to BNPL. In the present study, we want to verify if the BNPL service could shape people’s credit behavior as well as their payment credit behavior, as shows by Filotto et al. (2024). Furthermore, we want to observe and analyze the consumer perceptions and preferences regarding BNPL with the aim of to uncover its potential impacts on the consumer credit market and regulatory landscape. Specifically, the research investigates how BNPL usage or the intent to use influences individuals’ financial decisions and their engagement with additional consumer credit services.

Our research is built on a survey conducted with 1,457 Italian consumers, using the Computer-Assisted Web Interviewing (CAWI) method. Italy’s distinctive consumer behavior, financial cultural norms and the dynamics of its credit market offer an insightful backdrop for this analysis (Guiso and Jappelli, 1991; Guiso et al., 1996; Guiso and Paiella, 2004; Casolaro et al., 2006). Despite traditionally lagging e-commerce penetration, the onset of COVID-19 lockdowns significantly propelled the adoption of digital commerce and payment methods in Italy, heightening interest in alternative financing options like BNPL (Ardizzi et al., 2021; Papich, 2022; Guttman-Kenney et al., 2023). Given the existing literature’s depiction of Italian consumers as generally cautious regarding borrowing (Guiso et al., 1996), focusing on an Italian cohort is particularly relevant for determining whether BNPL schemes can alter financing behaviors even among traditionally conservative credit users.

This study enriches the BNPL literature by offering an empirical analysis from the perspective of consumer demand, especially within the Italian setting. It sheds light on the complex dynamics between BNPL, consumer characteristics and financing decisions, further outlining how consumers might shift toward additional credit services. The insights gleaned from this research are invaluable for market participants, regulators and policymakers, providing them with critical data to craft precise strategies and initiatives. These efforts are aimed at enhancing credit risk management within the financial ecosystem and supporting responsible credit usage among consumers.

Data collection was achieved through a web-based questionnaire distributed in Italy from February 1st to June 30th, 2023, using the CAWI method. Leveraging social media platforms like Facebook, Twitter and LinkedIn, we amassed 1,500 responses. After excluding responses with missing values, the count was narrowed to 1,457. According to criteria laid out by Dattalo (2008), our sample is representative of the Italian population.

We conducted a comprehensive survey that included a range of questions designed to evaluate participants’ financial literacy, confidence in the financial system, behavior regarding consumer credit and their shifting preferences from BNPL to traditional consumer credit options. Our objective was to gain insights into participants’ familiarity with and use of BNPL services to better understand their decision-making processes in financial matters. This approach led us to identify two primary variables for analysis: one related to participants’ awareness of BNPL services and the other to their actual utilization of these services. In addition, the survey collected vital socioeconomic and demographic information such as age, gender, marital status, educational attainment, employment status, income level and the frequency of e-commerce purchases.

To assess financial literacy, we adopted a widely recognized methodology, as outlined by Klapper and Lusardi (2020). Table 1 displays the specific financial literacy questions included in our survey (Table 1).

To evaluate trust in the financial system, we applied the methodology defined by Sapienza and Zingales (2012). We asked participants to rate their level of trust on a seven-point Likert scale, with a score of 1 representing “Not at all” and a score of 7 signifying “Very much.” To analyze the transition in preference from BNPL to traditional consumer credit financing, participants were invited to express their level of agreement with specific statements using the same seven-point Likert scale, where 1 stands for “Strongly Disagree” and 7 for “Strongly Agree.” For assessing consumer credit behavior, we used an 11-item questionnaire detailed in Table 2. Responses were gathered using a seven-point Likert scale to gauge various aspects of consumer credit usage, ranging from 1, indicating “Not at all,” to 7, indicating “Very much” (see Table 2).

To deepen our analysis of sensitivity toward Consumer Credit Behavior Questions, we carried out a factor analysis to pinpoint the main components within these queries. Using the maximum likelihood method and the Promax rotation criterion facilitated the handling of the Consumer Credit Behavior Questions. The key components unearthed through this factor analysis were then used as dependent variables in our linear regression models. Additionally, our survey covered questions on sociodemographic and economic characteristics, such as age, gender, marital status, educational background, occupational status, income and place of residence. It also explored participants’ attitudes toward e-commerce and their online purchasing frequency.

Table 3 offers a descriptive comparison of our sample. The age distribution indicates a concentration of participants in the middle age brackets, with those between 30 and 40 years constituting the largest group at 37.7%. Significant representations are also seen in the 41–50 and 51–60 age groups, at 20.3% and 30.1% respectively. The older demographics, particularly 61–70 and over 70 years, are less represented, making up 9.1% and 2.8% of the sample, respectively. Gender distribution is nearly balanced, with females making up 49.0% and males 50.9% of the sample, and a marginal percentage (0.1%) identifying as non-binary. The most common marital status is being married with children, accounting for 45.3% of respondents, followed by living alone (20.0%), married without children (19.4%) and living with parents (7.4%). Other living situations are less common. Educationally, the largest segment of respondents has completed high school or college (41.9%), with the next largest groups holding a master’s degree (26.0%) and a bachelor’s degree (13.4%). A minority have attained a PhD/MBA or have only elementary or middle school education. The majority are full-time employees (72.2%), with part-time employment and retirement also notable (13.3% and 6.5%, respectively). Few respondents are homemakers, disabled/unable to work or unemployed. Income levels are diverse, with the most common range being €20,000–€40,000 per year (39.7%). Other significant income groups include those earning €10,000–€20,000 (23.3%) and €40,000–€80,000 (17.8%), with a smaller portion of the sample earning either less than €10,000 or more than €80,000 annually. E-commerce purchase frequencies vary, with the most common being once every 2–3 months (25.0%), followed by once a month (24.7%) and once every 15 days (17.0%). Less frequent shopping patterns include at least once a week or only once in 6 months. Regarding BNPL usage, 12.0% of respondents currently use the service, while a larger proportion, 33.6%, are prospective users. Financial literacy levels are categorized into Poor, Medium and High, with a plurality of respondents displaying high financial literacy (41.9%). Medium and poor financial literacy levels are also significant, at 34.8% and 23.3%, respectively. (see Table 3)

To assess the feasibility of grouping Consumer Credit Behavior and Financial Literacy queries into main components (henceforth referred to as MCs), a Principal Components Analysis (PCA) with Promax rotation was performed. As per the findings of Jolliffe and Cadima (2016), PCA effectively simplifies the interpretation of complex data sets by reducing complexities associated with eigenvalues and eigenvectors. Concerning the PCA’s rotation technique, Russell (2002) suggests the Promax approach is more suitable when measurement scales are interrelated, as observed in our study. The outcome of the PCA is detailed in Table 4. The pattern matrix indicates a distinct four PC breakdown with no overlapping factors, allowing for a clear delineation of each PC. PC1 labeled “Financial Literacy,” PC2 as “Financial Diligence,” PC3 titled “Financing Preference,” and PC4 named “Financial Over-Indebtedness.” Following the guidelines of MacCallum et al. (1999) and Raubenheimer (2004), at least three questions are recommended to define the Consumer Credit Behavior construct. The validity of these measurement scales was further confirmed through an analysis of the Cronbach’s alpha (α) values for each construct, revealing that all constructs have α values exceeding the minimum threshold of 0.7, considered acceptable as per Taber (2018).

Aiming to ascertain whether characteristics connected to consumer credit behavior are more likely to occur among users who have used BNPL services (BNPL Users) or those aware of BNPL and wanting to use it in the future, Table 5 shows the results from logistic regression analysis. Compared to younger responders, older ones exhibit a greater inclination of using BNPL. Where a non-significant coefficient would indicate the necessity of more research to thoroughly investigate gender dynamics in financial behavior, BNPL usage seems to be unaffected by gender, marital status and employment position. Likewise, the income level participants indicated did not seem to match BNPL use. Lower educated participants show more inclination toward BNPL use, maybe suggesting a lack of knowledge of the problems connected with installment payment systems like BNPL, which could immediately affect user financial stability. Remarkably, BNPL users are those who visit less often online stores and businesses. For BNPL users, trust in the financial system has a more noticeable negative impact, maybe reflecting higher prudence or risk perception among these consumers. Particularly since participants with low financial literacy are more likely to use BNPL, the statistics on BNPL usage and financial literacy raises alarming questions. The results of the link between financial literacy and BNPL usage highlight the need of encouraging financial education programs to improve people’s financial skills. This result is consistent with the evidence provided by (Guttman-Kenney et al. (2023) and Filotto et al. (2024). Examining the link between consumer credit behavior, the results show that those more likely to use BNPL are those more inclined to finance purchases of goods and services (Financing Preference) and less likely to sufficiently inform themselves before signing financing contracts (Financing Diligence). This mix of data on financial literacy indicates how a potentially reckless attitude toward credit management can cause BNPL users to experience possible economic problems. The non-significance of the variable linked to worry for possible over-indebtedness situations among BNPL user participants emphasizes the importance of creating financial education programs targeted at raising consumer knowledge of the hazards involved with payment systems like BNPL. Examining the group of people aware of BNPL and intending to use it in the future (Prospects), patterns in many spheres coincide with those of present consumers. Older folks in particular usually feel more familiar with BNPL than younger ones. Like active users, familiarity with BNPL seems not to be influenced by gender, marital status or career condition. Though income is a major determinant of who among various income levels is more likely to become a future BNPL user, education level did not link with BNPL knowledge unlike past observations. Lower income people especially show increased interest in future BNPL usage. This lower financial capacity and the inclination toward future BNPL use could cause over-indebtedness conditions because of consumers’ income insufficiency in handling the debt from BNPL use. Against all the assumptions, prospects also often make less regular internet purchases. Prospects are more negatively impacted by trust in the financial system than by existing users, which would suggest more possible doubt of financial institutions among prospects. Like existing consumers, among prospects those with lowest degrees of financial literacy are most likely to use the BNPL service going forward. This strengthens the case already underlined on the need of supporting financial education programs to raise people’s financial competency. Prospects exhibit a low inclination to sufficiently inform themselves before making financial decisions and are also ready to engage in financing for the acquisition of goods and services, although to a less degree than current users. This tendency paired with unfavorable data on financial literacy points to possible economic challenges resulting from reckless credit management (Table 5).

The provided regression analysis of Table 6 presents a comprehensive and complex picture of consumer financial behavior influenced by the use of or interest in BNPL products. Prospects thus came front stage when one considered the percentage of the sample comprising both respondents already using BNPL and those aware of it, wishing to use it in the future. The study indicates that age has a positive effect and that older persons show stronger tendency toward consumer credit. Gender has a clearly negative impact; hence, given their experience with BNPL women, men may be less likely than women to seek for financing. Although the degree of education is clearly a major determinant of inclination toward the use of finance for purchasing goods and services, marital status and employment position have not shown any significant influence in individuals’s tendency to pick for financing. More precisely, a lesser degree of schooling is correlated with a change in preferences; this outcome is compatible with earlier studies applying logistic regression. Although declining faith in the financial system shows a clear negative coefficient, implying that this could assist to explain the change in preferences, income and online purchases are not main determinants of the financing choice. The negative impact of the choice shift brought about by BNPL adoption toward consumer credit and financial literacy indicates the requirement of extra financial awareness in avoiding the use of more conventional loan solutions. This result confirms already mentioned patterns and offers a concerning future picture of participant financial security. Moreover, linking the knowledge of financial literacy with that related to over-indebtedness reveals that respondents are impacted by too high confidence in their abilities to manage financial resources, therefore exposing them to possible financial calamities (Table 6).

The objective of this study was to investigate whether the usage or the intention to use BNPL services impacts individuals’ financial decision-making processes. It delves into aspects of financial literacy and consumer credit behavior, emphasizing their interaction with BNPL usage. This research adds to the scholarly discussion on the viability and future prospects of BNPL in the finance industry (Guttman-Kenney et al., 2023). BNPL usage could be change consumer credit behavior and, as consequence, their payment inclination. BNPL handling lead people to consider consumer finance instruments like consumer credit to do payment or purchase services or goods. This evidence could lead people, especially who have a very low level about financial literacy and skills, to take an incautious payment behavior. For this reason and as consequence of our findings, we point out how if people increase their preference on BNPL usage, in particular among less educated individuals calls for a shift in educational policies toward enhancing awareness of the risks tied to such financial instruments. Those results are consistent with the evidence provided by Gathergood (2012) and Lusardi and Tufano (2015). Additionally, the observed lower tendency among more knowledgeable consumers to use BNPL suggests that a robust financial education might act as a deterrent against engaging with potentially hazardous credit options. This highlights the critical role of financial literacy in preventing over-indebtedness, as highlighted by Hastings et al. (2013) and Kaiser et al. (2022). Thus, advocating for financial education initiatives is paramount, not only to bolster individual financial competencies but also to encourage more deliberate and cautious credit use.

Regulating BNPL services must take these factors into account by implementing policies that promote transparency and provide adequate consumer protection, especially for those most at risk. A strategy that combines enhanced financial education with stringent regulation may serve as a vital approach to reduce the dangers associated with BNPL, aiding in safeguarding consumer financial stability and contributing to overall economic health. A thorough comprehension of the factors driving BNPL usage is vital to steer such measures effectively, ensuring that advancements in payment technologies enrich consumer economic welfare without subjecting them to undue financial peril.