Episode 275: Live from Future Proof 2023: Decoding Financial Decision-Making with Hal Hershfield

Episode Summary

Executive Summary: In a live Future Proof episode, UCLA professor Hal Hershfield discusses a review of consumer financial decision-making research and how it can inform advice. The field has focused heavily on saving and spending, while couples, insurance, budgeting, major life decisions, aging, and end-of-life choices remain under-studied. Hershfield urges advisors and academics to collaborate, test interventions, account for individual differences and context, and use AI as a complement to human guidance.

Main Topics: What financial decision-making research covers: Hershfield defines the field as the accumulation and use of resources over time, reflected in behavior and choices. A topic-modeling review grouped research into saving and spending, trade-offs between money and other things, budgeting, negative outcomes such as fraud and inequality, and biases. Couples and differences in financial knowledge: Research discussed on couples suggests that a partner’s perceived knowledge can influence joint decisions beyond that partner’s objective knowledge. Making roles and confidence explicit—and checking actual understanding—may improve how couples divide financial responsibilities. Payment frequency, subjective wealth, and client context: More frequent pay can make money feel more certain and increase perceived available wealth, potentially leading to more frequent spending. Hershfield recommends that advisors consider subjective perceptions and payment patterns alongside objective financial measures. Evidence-based practice and collaboration: Hershfield argues that firms can learn from academic randomized trials and A/B testing to determine which messages and approaches work, while researchers can learn from practitioners about major decisions and real-world client needs. Research gaps across the financial life cycle: The review identifies a need for more research on later-life decisions, budgeting among middle- and mass-affluent households, insurance, mortgages and education, decumulation, wills, and advance directives. Generative AI and the role of human advisors: AI could provide timely, situation-specific financial education and help people explore possible futures. Hershfield views it as potentially enhancing advisors, whose human contribution includes emotional understanding and attention to client well-being. Personalizing advice and client engagement: Meeting cadence and support should vary by client rather than follow a one-size-fits-all model. Hershfield suggests using observed behavior, preferences, and potentially transaction patterns to identify when clients may need more engagement or face consequential decisions.

Key Arguments: Financial decisions matter beyond money because resources shape many other life outcomes; understanding their causes, context, and consequences can help improve decisions. The existing research base is uneven: saving and spending have received substantial attention, while insurance, budgeting in affluent households, aging, and transformative life decisions need more study. In couples, perceived expertise can affect decisions independently of actual expertise. Explicitly discussing who knows what and checking competence may prevent decision-making roles from arising only through inertia. Income frequency can influence spending through subjective certainty and perceived wealth, so advisors should ask about how and when clients are paid, not rely solely on balance-sheet measures. Financial firms should test interventions rather than rely only on intuition. Randomized experiments can clarify whether an approach works and why, though practical constraints and client risk must be considered. Financial education appears most useful when delivered at the point of need; AI could help tailor just-in-time education to the decision a person is facing. AI is more likely to complement than replace advisors in the near term, particularly when paired with high-quality information and human support for emotional and well-being-related needs. Advice should account for individual differences: some clients may want frequent contact and deeper involvement, while others prefer to delegate. Observed behavior may be more informative than stated preferences alone.

Data Points: Episode number: 273 - The live recording at Future Proof is introduced as episode 273 of the Rational Reminder Podcast. Hal Hershfield's prior appearances: Two previous episodes mentioned: 256 and 141 - The hosts describe Hershfield as a third-time guest and refer to his earlier appearances. Future Proof attendees: 3,000 - The hosts describe approximately 3,000 financial professionals as attending the event. Financial education research: Almost 10 years ago - Hershfield refers to a paper by Fernandez, Lynch, and colleagues on the impact of financial education, which he characterizes as finding little to no impact after accounting for self-selection, except for just-in-time education. Social Security claiming interventions: 13 - Hershfield says his research team tested 13 different interventions related to encouraging later benefit claiming; the findings discussed were hypothetical because the Social Security Administration did not permit changes to consumer messaging. Future-self horizon mentioned: 20 or 30 years - The hosts discuss how AI might help people consider what they need to do now to reach a future-self goal.

Pivotal Quotes: "The one case where it seems to really matter is what we call just-in-time education." — Hal Hershfield: He explains that financial education may be most influential when people are actively facing the decision the information concerns. "I think the right advisor will know how to use the tools in conjunction with that emotional appreciation or the sort of focus on well-being" — Hal Hershfield: He describes a complementary role for advisors alongside AI and automated tools. "I would love them to be excited about sort of operationalizing concepts" — Hal Hershfield: He encourages advisors to translate behavioral-science concepts into practical, tested approaches.

Implications: Advisors can improve service by tailoring contact and education to clients’ circumstances, testing interventions, and addressing decisions beyond investment products. Researchers and firms should collaborate on underserved areas, while using AI to deliver timely support without neglecting human judgment and well-being.