Thesis – Fintech Adoption Barriers Among Older Australians

Abstract

Australian banks are withdrawing physical service channels faster than many older customers are moving to digital alternatives. This thesis examines the factors that shape the intention of Australians aged 65 and over to adopt consumer fintech services, defined here as mobile banking applications, digital wallets and app-based payment channels. An extended UTAUT2 model incorporating trust and perceived risk was tested on survey data from 238 older Australians and analysed using partial least squares structural equation modelling. The model explained 58 per cent of the variance in adoption intention. Effort expectancy (β = 0.31), trust in the provider (β = 0.27) and perceived risk (β = -0.24) were the dominant predictors, while facilitating conditions, largely family based, also contributed (β = 0.18). Social influence was not significant. The findings indicate that usability, institutional trust and risk mitigation, rather than persuasion about benefits, should anchor bank strategy and public policy for the digital inclusion of older Australians.

Introduction

Australian retail banking has moved decisively from the branch to the smartphone. Australian Prudential Regulation Authority data show that the number of bank branches fell by more than one third between June 2017 and June 2024, with regional communities losing services at nearly twice the metropolitan rate (APRA 2024). Over the same period, cash declined to around 13 per cent of consumer payments in 2022, down from roughly 70 per cent in 2007, yet Australians aged 65 and over remain the most cash-reliant group in the economy (RBA 2023). A Senate inquiry into regional branch closures concluded that the burden of this transition falls disproportionately on older customers, the heaviest users of face-to-face banking (Senate Rural and Regional Affairs and Transport References Committee 2024).

The customers being asked to change are also the least digitally included. Although most Australian adults now use the internet, regular use and confidence decline sharply after age 65 (ABS 2023), and Australians aged 75 and over record the lowest digital inclusion scores of any age cohort in the Australian Digital Inclusion Index (Thomas et al. 2023). The risk environment compounds the problem: people aged 65 and over lost more than $120 million to scams in 2023, the largest reported losses of any age group (ACCC 2024). Fintech adoption by older Australians is therefore a financial inclusion problem with clear policy stakes.

This thesis investigates the factors that shape the intention of Australians aged 65 and over to adopt consumer fintech services offered by authorised deposit-taking institutions. It extends the second Unified Theory of Acceptance and Use of Technology (UTAUT2) with trust and perceived risk and tests the model on survey data. Three research questions guide the study:

1. RQ1: Which factors most strongly influence the intention of Australians aged 65 and over to adopt fintech banking services?
2. RQ2: To what extent do trust and perceived risk add explanatory power beyond the core UTAUT2 constructs?
3. RQ3: What role do facilitating conditions, including family and community support, play in shaping adoption intention?

Literature Review

From TAM to UTAUT2

Technology acceptance research descends from the Technology Acceptance Model, which established perceived usefulness and perceived ease of use as the core drivers of adoption (Davis 1989). Venkatesh et al. (2003) synthesised eight competing models into the Unified Theory of Acceptance and Use of Technology, identifying performance expectancy, effort expectancy, social influence and facilitating conditions as the principal predictors of behavioural intention. UTAUT2 adapted the framework for consumer settings by adding hedonic motivation, price value and habit (Venkatesh, Thong & Xu 2012). This study retains the four core constructs but omits the consumer extensions: habit presupposes existing use, which most respondents lack; price value discriminates poorly where banking apps are free; and hedonic motivation has limited theoretical purchase for utilitarian financial tasks.

Trust and perceived risk in digital finance

Digital banking asks customers to transact with an institution they cannot see, so trust substitutes for physical assurance. Gefen, Karahanna and Straub (2003) demonstrated that trust operates alongside ease of use and usefulness rather than through them, justifying its treatment as a separate construct. Perceived risk is conceptually distinct: Featherman and Pavlou (2003) decomposed it into financial, privacy, performance and psychological facets and showed that it suppresses e-service adoption even where trust is present. For older Australians the construct carries unusual salience because scam activity is heavily publicised and heavily concentrated in their age group (ACCC 2024). Both constructs therefore enter the model as separate paths.

The digital divide and older users

Divide research has shifted from access to skills and use. Friemel (2016) found that the divide among seniors is stratified within the cohort itself, with age, education, prior occupational computer exposure and social embeddedness predicting uptake. Australian evidence mirrors this second-level divide: the Australian Digital Inclusion Index scores older Australians lowest on the ability dimension even where access exists (Thomas et al. 2023), and ABS (2023) data show the steepest decline in confident use after age 75. Critically, Friemel (2016) identified hands-on help from family as a stronger predictor of uptake than generalised social pressure, a distinction the present model captures by separating social influence (normative expectations) from facilitating conditions (practical support).

Conceptual model and hypotheses

Figure 1 presents the conceptual model. Performance expectancy (H1), effort expectancy (H2), social influence (H3), facilitating conditions (H4) and trust in the provider (H5) are hypothesised to increase adoption intention, while perceived risk (H6) is hypothesised to reduce it. Age group, region and prior internet experience are modelled as controls.

Methodology

Research design and sample

The study employed a cross-sectional survey design. Participants were community-dwelling Australians aged 65 and over, recruited between March and May 2024 through Council on the Ageing state networks, University of the Third Age groups and seniors’ computer clubs in New South Wales, Victoria and Queensland. Questionnaires were offered online and in large-print paper form so that low-digital respondents were not excluded, a sampling bias common in technology acceptance research. Of 246 returned questionnaires, eight were removed for excessive missing data, yielding a final sample of 238. Respondents were 54 per cent female; 58 per cent were aged 65-74, 33 per cent were 75-84 and 9 per cent were 85 or over; and 39 per cent lived in regional or rural areas.

Measures

All constructs were measured with multi-item scales using five-point Likert response formats. Performance expectancy, effort expectancy, social influence, facilitating conditions and adoption intention were adapted from Venkatesh, Thong and Xu (2012); trust from Gefen, Karahanna and Straub (2003); and perceived risk from Featherman and Pavlou (2003). Items were reworded in plain English after pre-testing with twelve older adults, and the facilitating conditions items explicitly referenced help available from family members as well as from the bank.

Analysis and ethics

Data were analysed using partial least squares structural equation modelling (PLS-SEM), which suits prediction-oriented models and makes no multivariate normality assumption (Hair et al. 2022). Significance was assessed with 5,000 bootstrap subsamples. The study was approved by the university’s Human Research Ethics Committee and conducted in accordance with the National Statement on Ethical Conduct in Human Research (NHMRC 2023). Participation was voluntary, responses were anonymous and no incentives were offered.

Findings

Descriptive statistics and measurement quality

Table 1 reports construct means, standard deviations and reliability. All Cronbach’s alpha values exceeded the conventional 0.70 threshold and composite reliability ranged from 0.86 to 0.96, indicating internally consistent scales (Hair et al. 2022). Average variance extracted exceeded 0.50 for every construct and all heterotrait-monotrait ratios fell below 0.85, supporting convergent and discriminant validity.

Table 1: Construct descriptive statistics and reliability (five-point scale, n = 238)

Effort expectancy (M = 2.74) sat below the scale midpoint, indicating that respondents on balance found banking apps difficult rather than easy, while perceived risk (M = 3.67) was the highest-scoring construct. Adoption intention was lukewarm (M = 2.91, SD = 1.04), the large standard deviation signalling a divided cohort rather than uniform reluctance.

Structural model and hypothesis tests

As shown in Table 2, the structural model explained 58 per cent of the variance in adoption intention (R² = 0.58; Q² = 0.41), a substantial result for a behavioural model. Five of the six hypotheses were supported. Effort expectancy was the strongest predictor (β = 0.31, p < 0.001), followed by trust (β = 0.27, p < 0.001) and perceived risk (β = -0.24, p < 0.001). Facilitating conditions (β = 0.18, p = 0.003) and performance expectancy (β = 0.14, p = 0.028) contributed more modestly. Social influence was not significant (β = 0.06, p = 0.327), so H3 was rejected.

Table 2: Structural model results and hypothesis tests (5,000 bootstrap subsamples)

Discussion

The dominance of effort expectancy confirms that, for this cohort, the decisive question is not whether digital banking is worthwhile but whether it is manageable. This echoes the ease-of-use tradition running from Davis (1989) and the skills-based second divide described by Friemel (2016). Open-ended comments identified small text, frequent app redesigns and multi-step authentication as concrete deterrents, suggesting the barrier is located in product design as much as in user capability.

Trust and perceived risk operated as separate channels, consistent with Featherman and Pavlou (2003). Trust reflects confidence in the institution; risk reflects the perceived hostility of the wider environment. Given the scale of publicised scam losses among older Australians (ACCC 2024), elevated risk perceptions are not irrational and cannot be dismissed as technophobia. Interventions that build trust without addressing risk, or the reverse, are therefore likely to underperform.

Performance expectancy mattered far less than in mainstream consumer applications of UTAUT2 (Venkatesh, Thong & Xu 2012). A plausible interpretation is that perceived benefit is no longer the binding constraint: branch closures make the usefulness of digital banking self-evident and partly coerced (Senate Rural and Regional Affairs and Transport References Committee 2024), so variance in intention is governed by whether adoption feels achievable and safe rather than whether it feels useful.

The rejection of H3 alongside support for H4 is the study’s most distinctive finding. Older Australians appear largely indifferent to normative pressure to bank digitally, yet responsive to practical, relational support: the facilitating conditions items referencing help from family were the strongest-loading indicators of the construct. This family facilitating conditions pattern aligns with Friemel’s (2016) finding that hands-on encouragement outperforms social expectation, and with the ability dimension of the digital inclusion index (Thomas et al. 2023). Adoption among over-65s is better understood as a household project than as an individual decision.

Implications for Banks and Policy

Implications for banks

• Design for effort, not persuasion: offer simplified interface modes with larger type, stable layouts and plain-language error messages, and test releases with panels of customers aged 65 and over.
• Provide human-assisted onboarding: staffed digital coaching during branch transitions, phone-based walk-throughs and assisted service through Bank@Post outlets where branches have closed.
• Calibrate risk communication: pair specific, actionable scam guidance with clear reimbursement commitments, since generalised fear messaging raises perceived risk and suppresses intention.
• Build family-inclusive features: supported onboarding, view-only access and dual-authorisation options for trusted helpers, designed with explicit safeguards against elder financial abuse.

Implications for policy

• Sustain public funding for community digital capability programs delivered through libraries, University of the Third Age groups and seniors’ organisations, targeting confidence as well as skills.
• Adopt enforceable transition standards for branch closures, as recommended by the Senate inquiry (Senate Rural and Regional Affairs and Transport References Committee 2024), including minimum notice periods and assisted-service arrangements.
• Maintain regulator pressure on scam disruption (ACCC 2024), because ecosystem-level risk cannot be removed by any single bank.

Conclusion

This thesis asked why many older Australians hesitate at the threshold of digital banking while physical alternatives disappear around them. The extended UTAUT2 model provides a clear answer: intention is governed first by whether digital banking feels manageable, trustworthy and safe, and only secondarily by its perceived benefits, with family-based facilitating conditions supplying the scaffolding that normative pressure cannot. The study contributes an Australian empirical test of UTAUT2 in an ageing cohort and demonstrates the analytical value of separating normative influence from practical support.

Three limitations qualify the findings. The cross-sectional design cannot establish causality; intention is an imperfect proxy for behaviour; and recruitment through community organisations may over-represent socially connected seniors, meaning barriers in the wider population are plausibly higher rather than lower. Future research should track adoption longitudinally and examine adopter-helper dyads within families. As Australia’s banking system completes its digital transition, the pace of that transition should be set with, and not merely for, its oldest customers.

References

Australian Bureau of Statistics 2023, Household use of information technology, Australia, ABS, Canberra.

Australian Competition and Consumer Commission 2024, Targeting scams: report of the ACCC on scams activity 2023, ACCC, Canberra.

Australian Prudential Regulation Authority 2024, Authorised deposit-taking institutions points of presence statistics, APRA, Sydney.

Davis, FD 1989, ‘Perceived usefulness, perceived ease of use, and user acceptance of information technology’, MIS Quarterly, vol. 13, no. 3, pp. 319-340.

Featherman, MS & Pavlou, PA 2003, ‘Predicting e-services adoption: a perceived risk facets perspective’, International Journal of Human-Computer Studies, vol. 59, no. 4, pp. 451-474.

Friemel, TN 2016, ‘The digital divide has grown old: determinants of a digital divide among seniors’, New Media & Society, vol. 18, no. 1, pp. 313-331.

Gefen, D, Karahanna, E & Straub, DW 2003, ‘Trust and TAM in online shopping: an integrated model’, MIS Quarterly, vol. 27, no. 1, pp. 51-90.

Hair, JF, Hult, GTM, Ringle, CM & Sarstedt, M 2022, A primer on partial least squares structural equation modeling (PLS-SEM), 3rd edn, Sage, Thousand Oaks.

National Health and Medical Research Council 2023, National statement on ethical conduct in human research, NHMRC, Canberra.

Reserve Bank of Australia 2023, Consumer payment behaviour in Australia: results of the 2022 consumer payments survey, RBA, Sydney.

Senate Rural and Regional Affairs and Transport References Committee 2024, Bank closures in regional Australia, Commonwealth of Australia, Canberra.

Thomas, J, McCosker, A, Parkinson, S & Ganley, L 2023, Measuring Australia’s digital divide: the Australian Digital Inclusion Index 2023, RMIT University and Swinburne University of Technology, Melbourne.

Venkatesh, V, Morris, MG, Davis, GB & Davis, FD 2003, ‘User acceptance of information technology: toward a unified view’, MIS Quarterly, vol. 27, no. 3, pp. 425-478.

Venkatesh, V, Thong, JYL & Xu, X 2012, ‘Consumer acceptance and use of information technology: extending the unified theory of acceptance and use of technology’, MIS Quarterly, vol. 36, no. 1, pp. 157-178.