The Benefits of Behavioral Nudges: Using Choice Architecture to Improve Decisions and Shape Outcomes in Retirement Savings Programs

Journal of Financial Planning: March 2024

Executive Summary

An increasing number of private and public organizations have borrowed insights from behavioral economists to design nudges as part of their choice architecture in retirement plans to enhance financial outcomes on behalf of employees.
Lawmakers recently endorsed further nudging in retirement plans with the passage of the SECURE 2.0 Act. The new bill includes expanded automatic enrollment and auto-escalation features, higher catch-up contributions for late-career employees, an option to establish workplace emergency savings programs for qualifying employees, and the ability to offer small immediate incentives to plan participants.
Despite the widescale success of and broad support for these behavioral nudges, some detractors warn of the potentially intrusive nature of nudges, arguing that intentional paternalistic interventions can invade individual liberty and usurp autonomy. Other objectors claim that nudges are ineffective in addressing individuals’ preferences because those interests are too varied and are indecipherable to the retirement plan designer.
This paper explores (1) situations in which some level of paternalism is unavoidable; (2) how carefully crafted nudges can be used to inform decisions and shape outcomes without inhibiting individual agency; (3) how choice architecture (the deliberate arrangement of defaults, and the organization of options within the decision environment) can be used to bridge the gap between information and action; and (4) how new nudges in workplace retirement plans could significantly improve employees’ short- and long-term financial positions.

Shon J. Eckert, CF^P®, ChFC, BFA, is a CERTIFIED FINANCIAL PLANNER™ professional at Deseret Mutual Benefit Administrators, benefits administrator for employees of the Church of Jesus Christ of Latter-Day Saints. His research focuses on the use of behavioral finance tools to improve financial decision making and enhance economic outcomes among retirement plan participants.

Megan McCoy, Ph.D., LMFT, AFC, CFT-I, is an assistant professor in the personal financial planning program at Kansas State University. Her research focuses on financial therapy and has been published in top-tier family science, marriage and family therapy, and financial journals. Dr. McCoy volunteers for the Financial Therapy Association Board of Directors and serves as the co-editor of the Financial Planning Review.

JOIN THE DISCUSSION: Discuss this article with fellow FPA Members through FPA's Knowledge Circles.

FEEDBACK: If you have any questions or comments on this article, please contact the editor HERE.

In act 1, scene 2 of The Merchant of Venice, Portia laments to Nerissa, “If to do were as easy as to know what were good to do, chapels had been churches, and poor men’s cottages princes’ palaces” (Shakespeare 1995, 7). Shakespeare’s script epitomizes the struggle humans face when attempting to turn noble ideas into noble action. Research points to several behavioral biases that seem to suppress the average individual’s ability to act on information, even when the information is understood and the benefits to the actor are clear (Jain, Walia, Singh, and Jain 2021; Kahneman and Tversky 1982; Rock and Schwartz 2007). Barriers such as bounded rationality, present bias, status quo bias, an inclination toward mental accounting, and a pervasive difficulty to overcome inertia all combine to hamper behavior change even when significant consequences are at stake (Kahneman and Tversky 1979).

While this paper will focus on the realm of retirement plans, a health analogy illustrates the universal difficulty in bridging the gap between information and action, with one seminal study showing that even when coronary bypass recipients’ lives were on the line, only one in nine patients changed their behavior (Rock and Schwartz 2007). There are times when even clear consequences, such as harm to health or inadequate income in retirement, are not enough to overcome inertia and effect a change in human behavior. In these cases, nudging healthcare patients or, as evidenced in this paper, savings plan participants by using small and subtle prompts can encourage behavior change and help people move in a desirable direction. Thaler and Sunstein (2021) define this nudging process as any deliberate intervention in a program design with the objective of making the optimal behavior easier. These interventions can be as simple as the replacement of the historical default choice in a retirement plan to allow for automatic enrollment, the inclusion of an incentive to save, a limitation on withdrawals, or the deliberate placement of fruit in front of sweets in a school cafeteria (Thaler and Sunstein 2008).

Thaler and Sunstein (2021) describe the adoption and organization of such features in a program’s design as choice architecture and those responsible for offering and arranging any options that could influence a participant’s choice (for better or for worse) as choice architects.

Any feature in a program’s design or defaults that might make the desirable behavior easier by removing barriers (also known as friction) is one goal of good choice architecture. Despite the obvious difference between physical and financial threats to well-being, savings plan participants often fall victim to the same behavioral roadblocks as heart patients when it comes to acting on evidence-based prescriptions for success. The discovery of such persistent, detrimental patterns in decision-making led Kahneman and Tversky (1979) to experiment with intentional interventions to overcome behavioral biases. Their work laid the foundation for behavioral finance (Lewis 2017), the field that would spearhead the research that has effectively revolutionized the way workers save for retirement through subtle and simple behavioral-based interventions. This paper examines the efficacy of several of these interventions or so-called nudges and proposes the possibility of new nudges inside public and private retirement plans to reinforce long-term retirement savings and help workers prepare for short-term economic shocks as well.

In the realm of personal finance, multiple studies reveal that despite the discomfort of unwelcome debt that often results from inadequate financial reserves, many families remain unprepared for long-term retirement or even short-term mild economic shocks because they persistently under-save (Beshears et al. 2020; Federal Reserve 2019). According to the Federal Reserve, one in four adults was found to have nothing saved for retirement, and nearly 40 percent of respondents said they could not meet a $400 expense without selling or pawning something to do so (Federal Reserve 2019). Moreover, a short-term dearth for financial emergencies tends to deplete an employee’s long-term assets for retirement. One recent study found that within the past year alone, more than 60 percent of plan participants who had nothing saved for emergencies dipped into their retirement accounts to cover an unexpected expense (Ratcliffe et al. 2022). Among those who had at least one month of personal expenses on hand, however, only 9 percent tapped their retirement account for help (Ratcliffe et al. 2022).

There are certainly some cases in which inflation-afflicted households are so cash-constrained from either underemployment, disability, or any number of income-restricting circumstances that they are unable to set aside any savings whatsoever. In these examples, other solutions may be required to assist limited-income households in building a cash cushion to help absorb further financial shocks. These solutions are beyond the scope of this paper. Our research concentrates instead on using choice architecture within retirement programs to encourage workers who can set aside money from each paycheck to fund retirement and emergencies but do not, simply because deep-rooted behavioral biases compounded by ever-present inertia keep them from doing so.

This paper will first summarize several successful examples of choice architecture, where the deliberate integration of behavioral nudges into retirement plan design dramatically improved participation and increased savings rates. A thorough discussion will follow regarding several of the debates surrounding the practical application and the morality of nudges. An examination of the inevitability of some degree of paternalism and the inherent necessity of nudges in certain situations will be presented, followed by an introduction of new nudges that could assist employees in establishing emergency savings funds, in addition to building more robust retirement accounts. The paper concludes with implications for research and practical implementations for financial planning practitioners and retirement program designers.

Literature Review

The Success of Behavioral Nudges in Retirement Savings Plans

The following retirement savings study illustrates how an almost imperceptible nudge can be used to overcome the present bias (behavioral preference for immediate results) and status quo bias (the human tendency to resist change) by turning inertia on its head (Kahneman and Tversky 1979). For many years, newly hired workers with access to a workplace retirement plan were handed paperwork within a few days of employment and asked if they wished to set aside a portion of their paychecks toward retirement. This approach produced detrimental and delayed results. Many employees simply never got around to signing up for their retirement plan. Some who reconsidered after months or years on the job were dissuaded because of another behavioral bias: loss aversion (Kahneman and Tversky 1979). Once workers adapt to any given level of take-home salary, they are more reluctant to surrender a percentage of their paycheck. Regardless of the potential reward in retirement later, the impact to present pay is perceived as a loss of income. Some may consent to sacrifice salary as their careers move on, but behavioral economics reveals most humans are loss averse (Gigerenzer 2018), which, when compounded by inertia, may help explain why many eligible workers never enrolled.

To improve retirement plan participation, the Department of the Treasury and the Internal Revenue Service, inspired by groundbreaking studies from Brigitte Madrian and others, formally sanctioned the use of automatic enrollment in voluntary retirement plans in 1998 (Gale, Iwry, and Orszag 2005). Instead of asking employees if they would like to opt in to their employer-sponsored retirement plan, under the automatic enrollment scheme, employees are auto-enrolled upon hire and then given the right to opt out if they do not wish to remain enrolled. This simple change in the default or “starting point” (Sunstein and Thaler 2003) sent 401(k) participation rates soaring in plans that adopted the new rule. Initial enrollments rocketed from 49 to 86 percent (Madrian and Shea 2001). Seventeen years later, a research paper by Vanguard (Clark and Young 2018) reported that over time participation rates were still nearly doubling, from 47 percent to 93 percent whenever automatic enrollment was used (Clark and Young 2018). The success of small, but significant, nudges such as auto-enrollment and auto-escalation (discussed later) in retirement savings plans can be seen in various studies (Choi, Laibson, Madrian, and Metrick 2003; Clark, Utkus, and Young 2015; Cribb and Emmerson 2019; Madrian and Shea 2001). These studies demonstrate how the deliberate and beneficial arrangement of defaults through choice architecture can impact decision-making (Münscher, Vetter, and Scheuerle 2016).

Ethical and Practical Arguments Against Behavioral Nudges

Despite the wide-scale success of these low-cost, high-impact interventions, there are ethical and practical aspects of behavioral nudging that some economists view as overly paternalistic. Addressing these concerns (and hopefully resolving them) allows us to proceed to the critical questions of how to enhance existing savings plan nudges to produce more robust retirement balances and how the introduction of new nudges could improve the average worker’s short-term financial position as well.

Some of the most persistent arguments against the ethics and the effectiveness of choice architecture are that paternalistic nudging interferes with individual liberty and cannot improve the actual behavior that drives deficiencies in decision-making. One prominent paper by White (2017) argues that nudges fail to account for the complexity of an individual’s interests, violate the decision maker’s prerogative to elect outcomes, and may reinforce the very detrimental behavior they are designed to enhance. In one example, White (2017) compares paternalistic choice architecture to bowling with bumper guards; the beginner’s score is sure to improve, but they may not become a better bowler.

Another publication argues that by expanding the definition of liberty to include autonomy (which the authors define as how much control a chooser retains over their own evaluations and options), nudges are overreaching (Hausman and Welch 2010). The authors worry that a plan designer who uses choice architecture over “rational persuasion” to leverage cognitive deficiencies to advantage the chooser is at risk of affecting the chooser’s autonomy.

Both White’s (2017) and Hausman and Welch’s (2010) work present points that are important to consider in the plan design process and are appreciated in the context of assessing the morality and merits of nudging. However, these arguments do not provide sufficient evidence to overcome the logic or offset the proven benefits of behavioral nudges.

Addressing Anti-Paternalism and Other Objections to Behavioral Nudging

White’s (2017) assertions that any form of behavioral-based intervention stifles self-education, and that policymakers are ill-equipped to promote people’s actual preferences through paternalistic nudges, require substantiation. Research has not corroborated the concern that retirement program behavioral nudges are not already aligned with plan participants’ preferences. An abundance of evidence suggests that several (now) mainstream nudges, such as automatic enrollment, auto-increase, and restrictions on pre-retirement withdrawals already appeal to most people’s preferences. The following findings should allay apprehensions by anti-paternalists that people’s preferences or true interests in the field of retirement and emergency savings cannot be deciphered. In Beshears et al. (2020), the authors describe numerous studies in which participants value restricted accounts to strengthen savings strategies. Traditional economic theory predicts that upon receiving an unconditional windfall, most people would choose a holding account with no restrictions to allow unfettered spending of their unexpected gain. Research revealed the opposite. Aware of self-defeating tendencies, most subjects preferred a more restricted option. In fact, the steeper the stipulations on withdrawal, the more that was voluntarily allocated to the (restricted) account.

In retirement savings plans that incorporate auto-enrollment, a vast majority of employees express relief over having been enrolled without having to take any action. In one example, even of the handful of workers who opted out later, 90 percent said they were happy their employer at least offered the option (Harris Interactive 2007).

In the context of perceived paternalistic interventions overshadowing free will, White’s (2017) analogy may miss the mark when behavioral nudges are compared to Harry Frankfurt’s (1969) demon who possessed the power to predict his subject’s intentions and to intervene if the agent’s decision differed from what the demon wanted. In other words, he offers the appearance of free will by preliminarily presenting any choice, but then quickly corrects whatever decision deviates from the demon’s desired outcome. Within the type of paternalistic choice architecture advocated for by behavioral economists in retirement plan design, however, all original options are always on the table both before and after the chooser makes their selection. Furthermore, no decision is corrected after the fact and no coercion is used to dissuade the decider once an option has been selected. In the example of automatic enrollment in a retirement savings plan, the employee can simply unenroll.

The assertion of White (2017) and others that these deliberately designed interventions are overreaching and misguided (and that neutral nudges are possible and preferable) can be resolved as follows. First, in most cases, someone must make a design choice that will in varying degrees inform how options are viewed and choices are made by the individuals to whom the options and choices are presented (e.g., product placement in grocery stores). In other words, this paper does not disagree that deliberate choice architecture is potentially paternalistic, simply that some level of paternalism is unavoidable—and therefore should be thoughtful. Secondly, paternalism does not have to imply coercion. In the case of automatic retirement plan enrollment, none of the original options are removed, they are merely rearranged. The designer has not taken coercive action and no constraints on the individual’s original options have been imposed. Autonomy is intact and the individual retains the responsibility to choose for themselves which option is best.

Other objectors still contend that if the employee is not aware they have been enrolled, this counts as coercion, suggesting the worker has somehow been duped into saving. Thaler and Sunstein (2003, 2008) resolve this by explaining that precisely because people’s preferences are influenced by default rules and starting points, policies and plans should be overtly arranged to help the individual make better choices. “We urge that such (default) rules be chosen with the explicit goal of improving the welfare of the people affected by them.” Thaler and Sunstein (2003, 2008) go on to state that the mere rearranging of food in a cafeteria or the inversion of a default rule in a retirement plan does not constitute coercion. No one choice is binding. The chooser always has the option to reach past the fruit in the cafeteria to get to the sweets or opt out of the retirement plan. There is no deception in the default and no coercion in the plan.

Finally, White’s proposals (2017) fall short of the promised “neutral” nudges that the paper asserts would still enhance financial decision-making while honoring the individual’s intentions. Only sparsely explained solutions are provided, such as the elimination of automatic enrollment in 401(k) programs and the “logical and aesthetic” rearranging of food are provided. These proposals seem to suggest a reversion to past practices already proven ineffective in facilitating behavioral improvements in finance and health.

In the context of retirement savings plan design, choice architects are not attempting to steer choosers in an undesirable, undisclosed, or deceptive direction. Nudges have been introduced overtly and with the express purpose of facilitating widely established intentions by workers who want to save but acknowledge they may lack the tools or the skill to do so entirely on their own. While future studies are needed to reveal whether individuals can learn as much from errors averted as they can from mistakes made, the bowling with bumper-guards analogy may not be relevant as far as any restrictions to agency are concerned—just like a bowler can always lower the guardrails, so can a retirement plan participant choose to opt out of their 401(k).

Thaler and Sunstein (2003) anticipated the potential for anti-paternalistic arguments such as these cited above and countered by simply stating that some form of paternalism is almost always unavoidable in any design context. When it comes to options offered within any given program, private or public, someone must make a design choice that will impact the behavior of someone else. Because what people choose largely depends on “starting points and default rules” (Thaler and Sunstein 2003, 178), some form of influence on the part of the plan designer (intentional or not) is inevitable. Nudges, in these cases, are unavoidable; therefore, they should be intentional and deliberately designed to benefit the chooser.

Automatically enrolling workers in a retirement plan illustrates this point perfectly. The worker is deliberately nudged in a specific direction—to save for retirement. This can only mean that programs that do not automatically enroll employees must nudge workers in the opposite direction to not save for retirement. It is difficult to imagine even the most adamant anti-paternalist considering this a viable design choice. Lawmakers seem to agree. The SECURE 2.0 Act mandates that all employer-sponsored retirement savings programs established after 2024 incorporate auto-enrollment and auto-escalation into their plan design.

The Power of the Default and the Logic of Automatic-Enrollment

Thaler and Sunstein (2003) use their own university’s parking program to illustrate the power that auto-enrollment can wield in any enterprise deemed universally desirable by beneficiaries and administrators. When the IRS added a provision that allowed employees to pay for employer-provided parking with pre-tax dollars, it was presumed that all eligible employees would want to use the new benefit. At organizations where parking costs are not trivial, the savings can be significant. Thaler and Sunstein (2003) recount that their employer (as well as several other universities) adopted a policy to automatically enroll all employees in the pre-tax program, presuming that every employee would appreciate the tax savings. Those who preferred to pay for their parking on an after-tax basis were given the option to opt out. Naturally, no one did.

This is a prime example of (unavoidable) paternalism, recognized and used to advantage the chooser. The university employee’s choice set is not limited. Their agency is uninhibited, and their autonomy is never altered. Both options (to participate or not) remains available. The only difference was in the default. The manual opt-in option assumes most employees are not interested in the tax savings and increases the transaction costs (friction) of enrolling for those who are. The auto-enroll or opt-out option assumes most employees are interested in tax savings and reduces the friction of enrolling. Either option is paternalistic in the sense that employees are bound to be nudged one way or the other by the starting point (default rule). The university, just like its employees, had to choose between one default option and the other. Assuming most (if not every) employee would want the tax savings, automatically enrolling employees was the logical (and responsible) choice, as it made the more beneficial (and preferred) option easier to activate. No forms needed to be filled out or deadlines remembered.

Enhancing Existing Retirement Savings Nudges

To improve the overall efficacy of auto-enrollment, Thaler and Benartzi (2004) proposed a retirement plan nudge called Save More Tomorrow, which would allow employees to increase their contributions automatically on an annual or quarterly basis after they were enrolled. Employees could commit in the present to making annual contribution increases in the future—for example when they received a raise. Of those who enlisted, 80 percent of the participants remained in the program and upgraded their savings rates from 3.5 percent to 13.6 percent in less than four years.

In 2017, Thaler built on this proposal by calling for a combination of auto-enrollment and auto-escalation in 401(k) plans, pointing out that auto-enrollment default rates were set too low by plan designers (Salisbury 2017). Plan designers feared participants might opt out of saving plans altogether, but by setting the default too low, Thaler suggested that many people who could (and potentially would) have saved more, did not (Salisbury 2017). Studies found that when companies like Google and Credit Suisse increased their default contribution rates, very few of their employees opted out (Grind 2015). In fact, in one of the samples, only two out of 8,500 employees exited the plan due to increased savings rates (Grind 2015). These studies further support the theory that employee’s preferences are known: workers want to save more, are willing to save more—and they welcome the help to do so. Raising default rates and bundling automatic increases with automatic enrollment in 401(k) plans would seem to be an effective and responsible way to assist employees with access to a retirement plan who might otherwise under-save. While care must be taken to not enroll employees at unreasonably high contribution rates, greater 401(k) default enrollment and auto-escalation rates should be considered to accommodate greater savings for the majority of plan participants. Fortunately, policymakers and legislators have recently signaled their support for studies like those cited from Brigitte Madrian et al., Richard Thaler, and others with the passing of SECURE 2.0. The bill, which represents a win for behavioral economics, expands auto-enrollment and auto-escalation in existing retirement plans and will require both features be incorporated into new retirement plans with more than 10 plan participants established after 2024.

Customizing Default Rates

Another significant behavioral barrier to saving enough for retirement is that many workers simply don’t know how much they should be saving; they have no idea what retirement contribution rate is right for them. In fact, one of the most common financial planning questions (in the experience of the authors) from plan participants and practitioner clients alike tends to be: “Am I on track for retirement?” This general lack of clarity can lead to inaction, especially for those with limited financial literacy. An insufficient default savings rate could leave plan participants ill-prepared for retirement without the worker even being aware of the deficiency until it is too late. Moreover, it has been shown that inertia will prevent most workers from ever deviating from the default contribution rate their retirement plan established (Thaler and Sunstein 2021, 198–217). Therefore, it is critical that the default be adequate (Thaler and Sunstein 2021, 198–217). SECURE 2.0 (2022) takes a step in the right direction by authorizing retirement plan designers to increase default contribution rates, however, this doesn’t necessarily clarify for the average plan participant exactly what savings rate is right for them. This is the issue with any standardized default scheme. Planning for retirement is not a one-size-fits-all proposition. In addition to higher default contribution rates, this paper proposes employees be presented with a personalized contribution rate, tailored to their own financial situation. This customized default contribution could be based on the employee’s age, current retirement savings balances, any other household retirement assets and income sources, a pre-established income replacement ratio for retirement, and/or a stated retirement age goal. To preserve the sovereignty of the savings plan participant, this information could be provided voluntarily or opted out of by allowing the enrollee to alternatively choose an age-based or overall standardized contribution rate. As noted, default savings rates and starting points provide powerful incentives to steer employees in their decision-making process. This type of arrangement could encourage older employees who may have under-saved to enroll at a higher rate, increasing the probability of a successful retirement. Younger employees could be enrolled at lower retirement savings rates (leaving more to be deflected to other goals such as debt reduction or a dedicated rainy-day fund). It could also provide clarity and confidence for employees of any age that they are saving at an adequate rate, tailored to their situation.

An example follows of how a customized retirement default rate might work in the case of a 30-year-old employee who has not previously saved for retirement. (These numbers are expressed for illustrative purposes only and not to be interpreted as explicit or empirical recommendations.) A plan designer could operate under the following assumptions:

The plan presumes that the employee wishes to replace 85 percent of income in retirement.
Social Security will replace 30 percent of pre-retirement earnings.
The employee will retire at age 67 and expect to live 25 years in retirement.
No other significant source of retirement income is available, apart from Social Security.
Pre-retirement investment returns will average 6 percent annually.
Post-retirement investment returns will average 5 percent annually.
Inflation will average 3 percent throughout.
Wages will increase by 2 percent each year.

Using these assumptions, a 30-year-old employee would be automatically enrolled in the plan at a contribution rate of 10 percent. The default for a 25-year-old employee would be 8 percent, for a 35-year-old 15 percent, and so on. These are only examples, and care should be taken to ensure the customized default rate is not set so high that the employee is discouraged from saving altogether (SECURE 2.0 allows automatic increases up to 15 percent). A field study of 10,000 employees found that workers were tolerant of steeper default rates and were willing to save more for retirement when higher recommended rates were displayed on enrollment screens, but only up to a point. The study results suggested this optimal starting point, before enrollment levels begin to taper, appears to approach 11 percent (Beshears, Benartzi, Mason, and Milkman 2017). Should an enrollee’s case call for a higher overall default than is anticipated would be acceptable, lower starting points with auto-increase could be used to ease the saver to the recommended rate over time, so as not to dissuade enrollment altogether. Employer matching funds, where available, could also be included in the customized calculation, helping to offset higher than customary default rates, making steeper starting points more accessible to savings plan enrollees.

This overall process should appear as effortless as possible to the employee. As they are automatically enrolled, the employee could receive notice of their personalized retirement savings rate and an explanation of why they were enrolled at that rate, along with the option to enter pertinent data that might impact the recommended rate, opt out, or reduce their contribution, if necessary. It is suggested that, for those who elect to opt out entirely or reduce their initial contribution rate, an option should appear nudging them to engage in an auto-increase feature that would put them on track to reach the recommended retirement goal gradually over time.

The above assumptions are not necessarily prescribed as the actual inputs the choice architect in a retirement plan must use but are merely provided to illustrate how a customized contribution scheme might work. An examination of optimal assumptions should be conducted to determine what works best for most employees based on various ages and financial situations.

The default might be designed so that while all enrollees are prompted to enter pertinent data to determine a customized contribution rate, for those who opt out of this process, new hires are still automatically enrolled at the plan’s standardized rate. One approach might be for the retirement savings plan provider to settle on a standard default rate toward the top of the 10 percent range allowed by the SECURE 2.0 Act (2022) for all employees and add an electronic survey as part of the onboarding process in which the individual is asked relevant retirement readiness questions. The survey could be as short as one question, asking the enrollee how much they currently have saved for retirement. Once the information is entered, the employee could be prompted to do one of two things: (1) consent to the suggested contribution rate and receive notice of enrollment with the option to opt out or contribute at a lower rate; or (2) remain at the plan’s standard default rate but allow the auto-increase to adjust upward over time to the recommended level.

The obvious implication of this scheme is to design a system that incorporates these additional inputs without unduly encumbering the enrollment process, which must remain as automated and effortless as possible for the employee. If the inputs required by the employee increase the transaction costs (friction) of finalizing enrollment so much that some employees are dissuaded and opt out, the entire system could backfire. It is anticipated, however, that much of this could be resolved with the use of artificial intelligence and the creation of an enrollment system that is simple, appealing, and virtually frictionless to the employee. Should the programming of this proposal prove too onerous (for the plan designer or the participant) perhaps a simpler, semi-customized approach could be taken. For example, plan designers could use a standard age (67) and income replacement rate (85 percent) for retirement, along with the employee’s age upon hire, to create a customized default rate that would nudge the employee through auto-escalation toward saving the right amount for retirement.

While logistical and legal hurdles for implementation of customized contribution rates may be manifold, with more than eight out of 10 adults reporting they are still not saving enough for retirement (Federal Reserve 2019) and overwhelming evidence demonstrating the protracted power of pension plan defaults, the benefits of such a system seem clear. Individualized investment allocations, through age-based target date funds, are already offered in most retirement plans. Contribution rates should be customized as well.

The Psychology of Saving

In addition to enhancing retirement savings nudges, plan designers should act on one of SECURE 2.0’s landmark features and turn their attention to helping households build an adequate cash cushion for emergencies. As noted in this paper’s introduction, studies from the Federal Reserve (2019) confirm that many households struggle to save for even modest unplanned expenses in the short term. The same behavioral biases that prevent people from investing for retirement also deter families from saving for emergencies. Consider the case of a worker who is paid on a biweekly basis. They will receive over 1,000 paychecks during an average career. Setting aside a portion of each paycheck would require over 1,000 individual acts of saving. This is like asking workers to enroll 1,000 times in their 401(k). This explains in part the high level of irregular saving when it comes to nearer-term goals, like accumulating savings for a rainy day. Workers’ wages tend to follow the path of least resistance, meaning once salary makes its way to a household’s checking (spending) account, inertia tends to keep it there—only to be consumed, without any reserves set aside for rainy days.

The widespread success in using behavioral economics to get workers to save more for retirement begs the question: why not develop similar default rules that automatically enroll employees in a short-term savings program and fund it through payroll deduction? Employees could opt out if they wish, but those who remain enrolled would effortlessly accumulate funds for emergencies. This idea was explored by Beshears et al. (2020) who wrote that operating under another behavioral bias, mental accounting, employees may overspend today while simultaneously enrolling in a 401(k) plan that could possibly reflect their preference to save in the future. This lack of liquidity through overspending and under-saving for emergencies leads to “401(k) leakage.” For every dollar that makes its way into 401(k)s, between 30 and 40 cents leak out before retirement (Argento, Bryant, and Sabelhaus 2015). It is projected this leakage could wipe out nearly one-quarter of retirement assets in the United States over the next three decades (Tergesen 2019).

Setting up a separate savings account dedicated to emergencies and funded through payroll deduction could help workers systematically prepare for unexpected expenses without relying on their retirement funds to address short-term economic shocks. Workplace emergency savings arrangements are not without precedent. Even prior to recent legislative support (detailed below), employers such as Kroger, Truist, Levi Strauss, United Parcel Service, and Home Depot were already offering rainy day savings accounts to employees, demonstrating an appetite among employers and plan participants for an emergency savings nudge (Tergesen 2019; Min 2019). Creating this kind of automatic payroll deduction to an employer-sponsored savings account could leverage the same behavioral economics doctrine of reverse inertia and mental accounting that helps employees save for retirement. If employees were automatically enrolled, only one simple decision (to remain opted-in or to opt out) would be required. Thus, over an average career, instead of manually trying to save for emergencies more than 1,000 times, the employee would only have to decide to save once.

Legislators listened, and an option for retirement plan providers to incorporate a small, designated account for emergencies, funded by after-tax (Roth) dollars recently passed as part of the SECURE 2.0 Act (2022). The new provision can be offered to non-highly compensated employees beginning as early as 2024. Workers earning less than $150,000 will be allowed to defer up to 3 percent of their salary in the form of Roth contributions to this designated emergency account. Unfortunately, the maximum balance designated for emergencies is limited to $2,500 (indexed), but a nudge is a nudge, and this new intervention represents recognition by policymakers that workers want help preparing financially for retirement and for rainy days. Moreover, plans that already match employee retirement contributions will be required to match designated emergency deferrals and not charge any distribution fees on the first four withdrawals each year, providing further incentive for workers to take advantage of this new nudge. We recommend that retirement plan designers do all they can to consider such a feature in their savings programs to make it easier for plan participants to save for rainy days and not just for retirement.

Implications

The use of research-based behavioral interventions within the choice architecture of public and private retirement plans may help workers overcome behavioral barriers that can prevent them from saving enough for retirement and rainy days. Nudging has proven to be extremely effective (e.g., the models of organizations creating automated savings accounts [Tergesen 2019; Min 2019]) without restricting an individual’s agency or usurping the employee’s autonomy. Retirement plans that incorporate interventions like automatic enrollment and auto-escalation in their choice architecture reduce under-enrollment and boost contribution rates significantly.

A customized contribution arrangement might encourage older employees who may have under-saved for retirement to enroll at a higher rate. It could also give employees of any age confidence that they are saving at an adequate rate for retirement, tailored to their situation, and allow those who may be over-saving for retirement and underprepared for emergencies to divert more earnings toward rainy day reserves or other financial goals.

Practitioner and Policy Implications

Individual financial planning practitioners can incorporate behavioral principles from this paper. Customized contribution rates for both retirement and rainy days can be recommended as part of an individual’s financial analysis. Clients can be significantly motivated by a clear-cut number. Providing personalized contribution rate recommendations, tailored to their financial situation and goals, can give clients the clarity and confidence needed to work toward a specific saving rate calculated to prepare them for emergencies and put them on track for retirement. Some clients may not be able to begin saving at their recommended rate at once. In these cases, the financial planner could encourage the use of auto-escalation in the client’s retirement savings plan for clients seeking to approach a retirement savings goal gradually over time. Finally, for clients seeking to save for emergencies and other short-term goals, practitioners can leverage mental accounting biases by encouraging the use of separate saving accounts. These separate accounts can be funded automatically through payroll deductions that bypass the client’s checking account(s) altogether, leveraging inertia to the client’s advantage.

SECURE 2.0 expands existing behavioral interventions in retirement plans and opens the door for new nudges, such as matching employer funds based on student loan payments, increased catch-up contributions for certain age groups, de minimis incentives such as gift cards for plan participation, and emergency savings accounts for qualifying employees.

While the SECURE 2.0 Act (2022) represents a big win for behavioral economics, more work must be done by Congress in the future. The (indexed) $2,500 maximum allowed in the emergency savings portion of an employee’s 401(k) is not nearly enough to accommodate the accumulation of a 3–6-month emergency fund that most financial planning professionals recommend. Future policy should consider increasing the $2,500 ceiling considerably. Policymakers must also consider whether mandating that these emergency accounts be established inside of sponsored plans (as opposed to allowing for separate outside accounts) will provide enough flexibility to appear accessible to the employee. Further, retirement plan designers will need to ensure distribution processes and paperwork are simplified before many employees will feel they have unfettered access to their emergency fund.

Research Implications

Research will be required to measure the success of enhanced and new nudges sanctioned by SECURE 2.0, such as increased retirement contribution rates through mandatory auto-escalation and auto-enrolled workplace emergency savings programs, as they are incorporated into qualified savings plans. Overall participation should be measured by looking at opt-out rates from new programs initially and as time goes on.

It is expected that mandatory provisions of the SECURE 2.0 Act (2022), such as expanded automatic enrollment and auto-escalation that allow plans to increase default and maximum contribution rates, will provide some of the first case studies for research to be conducted. When sufficient data are available, average participation, percentage of contribution, and opt-out rates should be compared to pre-SECURE 2.0 retirement plans to see how these nudges impact participant behavior in terms of willingness to remain enrolled and save at higher rates.

Studies are needed to evaluate how many programs adopt optional interventions such as the addition of emergency savings opportunities and offering de minimis incentives for plan participation.

Regarding age- and/or asset-based customized contribution rates, further research is needed to test tolerance levels for higher default rates. At a minimum, in many cases, recommending an optimal, personalized rate could anchor the employee to a higher number, motivating them to make that higher rate their goal. Auto-escalation could be offered to gradually get them to the recommended rate over time. The novelty of this nudge calls for further exploration into the practical and political considerations, and an examination of potential logistical and legal hurdles. However, such an arrangement could greatly benefit employees seeking to save the right amount for retirement. Surveys should be conducted to measure retirement plan policymaker and plan participant support for such an arrangement.

For those plans that adopt the emergency savings feature of SECURE 2.0, qualified employees will have access to an emergency savings scheme through their employer. Such an account could be auto-enrolled in and funded automatically through payroll deduction. Research is required to evaluate anticipated improvements in 401(k) leakage rates and the short-term financial position of workers who participate in these payroll-deducted emergency savings accounts versus those who do not enroll. These data should then be compared to employee populations who do not have access to a workplace emergency savings program.

Conclusion

Where at least two opposing options are available in any private or public program, some form of paternalism is unavoidable. Retirement program defaults are bound to nudge employees in one direction or the other and should, therefore, be thoughtfully constructed. Properly placed and based on behavioral observations and participant surveys, these nudges, or liberty-preserving paternalistic interventions, can help bridge the gap between belief (“I should save for retirement and rainy days”) and behavior (retirement savings shortfalls and underfunded emergency funds persist) by helping workers turn information into action without interfering in an employee’s autonomy or inhibiting freedom of choice.

Behavioral nudges, such as automatic enrollment and auto-escalation in defined contribution plans, help employees save for retirement by harnessing the power of inertia in favor of the saver. Customized contribution default rates based on an employee’s age, assets, desired retirement age, and income goals are recommended to help address retirement savings shortfalls, increasing the probability of a successful retirement for many employees. An employer-sponsored, automated emergency savings arrangement could create a cash cushion for many workers, helping them prepare for and address short-term economic shocks without raiding retirement accounts prematurely.

Citation:

Eckert, Shon J., and Megan McCoy. 2024. “The Benefits of Behavioral Nudges: Using Choice Architecture to Improve Decisions and Shape Outcomes in Retirement Savings Programs.” Journal of Financial Planning 37 (3): 72–86.

References

Argento, Robert, Victoria Bryant, and John Sabelhaus. 2015. “Early Withdrawals from Retirement Accounts During the Great Recession.” Contemporary Economic Policy 33 (1): 1–16.

Beshears, John, James J. Choi, Christopher Harris, David Laibson, Brigitte C. Madrian, and Jung Sakong. 2020. “Which Early Withdrawal Penalty Attracts the Most Deposits to a Commitment Savings Account?” Journal of Public Economics 183: 104144.

Beshears, John, Shlomo Benartzi, Richard T. Mason, and Katherine L. Milkman. 2017. “How Do Consumers Respond When Default Options Push the Envelope?” SSRN: https://ssrn.com/abstract=3050562 or http://dx.doi.org/10.2139/ssrn.3050562.

Choi, James J., David Laibson, Brigitte C. Madrian, and Andrew Metrick. 2003. “Optimal Defaults.” The American Economic Review, Papers, and Proceedings of the One Hundred Fifteenth Annual Meeting of the American Economic Association 93 (2): 180–85.

Clark, Jeffrey W., and Jean A. Young. 2018. “Automatic Enrollment: The Power of the Default.” Vanguard Research.

Clark, Jeffrey W., Stephen P. Utkus, and Jean A. Young. 2015. “Automatic Enrollment: The Power of the Default.” Vanguard Research.

Cribb, Jonathan, and Carl Emmerson. 2019. “Requiring Auto-Enrollment: Lessons from UK Retirement Plans.” Issue in Brief 19 (6): 1–19.

Federal Reserve. 2019. “Report on the Economic Well-Being of U.S. Households.” www.federalreserve.gov/publications/files/2018-report-economic-well-being-us-households-201905.pdf.

Frankfurt, Harry G. 1969, December 4. “Alternate Possibilities and Moral Responsibility.” The Journal of Philosophy 66 (23): 829–839.

Gale, William G., J. Mark Iwry, and Peter R. Orszag. 2005. “The Automatic 401(k): A Simple Way to Strengthen Retirement Savings.” Retirement Security Project.

Gigerenzer, Gerd. 2018. “The Bias Bias in Behavioral Economics.” Review of Behavioral Economics 5 (3–4): 303–336.

Grind, Kristin. 2015. “Companies to Workers: Start Saving More—Or We’ll Do It for You.” The Wall Street Journal. www.wsj.com/articles/companies-to-workers-start-saving-moreor-well-do-it-for-you-1444901580.

Harris Interactive. 2007. “Retirement Made Simpler.” www.aarp.org/content/dam/aarp/ppi/2018/03/retirement-made-simple-topline-report-110507.pdf.

Hausman, Daniel M., and Brynn Welch. 2010. “Debate: To Nudge or Not to Nudge.” The Journal of Political Philosophy 18 (1): 123–136.

Jain, Jinesh, Nidhi Walia, Simarjeet Singh, and Esha Jain. 2021. “Mapping the Field of Behavioral Biases: A Literature Review Using Bibliometric Analysis.” Management Review Quarterly: 1–33.

Kahneman, Daniel, and Amos Tversky. 1979. “Prospect Theory: An Analysis of Decision Under Risk.” Econometrica 47 (2): 263–292.

Kahneman, Daniel, and Amos Tversky. 1982. “The Psychology of Preferences.” Scientific American 246 (1): 160–173.

Lewis, Michael. 2017. The Undoing Project. Penguin Books.

Madrian, Brigitte C., and Dennis F. Shea. 2001. “The Power of Suggestion: Inertia in 401(k) Participation and Savings Behavior.” The Quarterly Journal of Economics 116 (4): 1149–1187.

Min, Sarah. 2019. “More Big Companies Help Workers Auto-Save to Build Emergency Funds.” MoneyWatch. www.cbsnews.com/news/big-companies-are-helping-workers-sock-away-emergency-savings/.

Münscher, Robert, Max Vetter, and Thomas Scheuerle. 2016. “A Review and Taxonomy of Choice Architecture Techniques.” Journal of Behavioral Decision Making 29 (5): 511–524.

Ratcliffe, Caroline, Brianna Middlewood, Melissa Knoll, Misha Davies, and Grant Guillory. 2022. “Emergency Savings and Financial Security.” Consumer Financial Protection Bureau (CFPB) No. 2022-01.

Rock, David, and Jeffrey Schwartz. 2007. “The Neuroscience of Leadership.” Reclaiming Children & Youth 16 (3): 10–17.

Salisbury, Ian. 2017. “Meet Richard Thaler, the Man Who Just Won the Nobel Prize for Helping You Save for Retirement.” https://money.com/thaler-nobel-economist-retirement-savings-nudge/.

SECURE 2.0 Act, H.R.2954. 2022. www.congress.gov/bill/117th-congress/house-bill/2954/text.

Shakespeare, William. 1995. The Merchant of Venice. Dover Publications.

Sunstein, Cass R., and Richard H. Thaler. 2003, May. “Libertarian Paternalism Is Not an Oxymoron.” The American Economic Review 93 (2): 175–179.

Tergesen, Anne. 2019. “Employer Help Workers Build Household-Emergency Funds.” The Wall Street Journal. www.wsj.com/articles/employers-help-workers-build-household-emergency-funds-11560418206.

Thaler, Richard H., and Cass R. Sunstein. 2003, May. “Libertarian Paternalism.” American Economic Review 93 (2): 175–179.

Thaler, Richard H., and Cass R. Sunstein. 2008. Nudge: Improving Decisions About Health, Wealth, and Happiness. New Haven, CT: Yale University Press: 14–38.

Thaler, Richard H., and Cass R. Sunstein. 2021. Nudge: The Final Edition. Yale University Press.

Thaler, Richard H., and Shlomo Benartzi. 2004. “Save More Tomorrow™: Using Behavioral Economics to Increase Employee Saving.” Journal of Political Economy 112 (S1): S164–S187.

White, Mark D. 2017. “Nudging Debt: On the Ethics of Behavioral Paternalism in Personal Finance.” Journal of Financial Counseling and Planning 28 (2): 225–234.

Topic

Psychology of Financial Planning