#298 1998 · Richard Thaler & Shlomo Benartzi (Save More Tomorrow) · Retirement finance / behavioral economics
Employees wouldn't save more of today's paycheck, so Thaler and Benartzi asked them to save more of a paycheck they didn't have yet.
the problem
people know they should change a financial behavior but the change itself feels like an immediate loss, so they keep delaying it
background
Retirement savings programs in the 1990s consistently struggled with low enrollment and low contribution rates even when employers offered matching funds, because asking an employee to increase their 401(k) contribution meant asking them to accept a smaller paycheck starting immediately — a visible, current loss that behavioral economics research showed people weighed far more heavily than the equivalent future gain from a bigger retirement balance. Standard financial education campaigns explaining the long-term math rarely overcame this immediate loss aversion.
Economists Richard Thaler and Shlomo Benartzi designed an alternative around a specific behavioral insight: loss aversion only bites when the loss is felt now. If a savings increase could be timed to future pay raises instead of current pay, an employee's take-home pay would never actually decrease — they'd simply keep a smaller slice of a raise they hadn't received yet, a choice people find far easier to commit to in advance.
what everyone would do
The standard approach was financial education, explaining the long-term math of compound retirement savings and asking employees to voluntarily increase their current 401(k) contribution, a change that meant accepting a smaller paycheck starting immediately.
what they saw
Thaler and Benartzi saw that the problem wasn't a lack of understanding of the long-term math, it was that any increase to a current contribution rate registered as an immediate, visible loss, and loss aversion research showed people weighed that present loss far more heavily than an equivalent future gain, no matter how clearly the retirement math was explained. Rather than trying to overcome loss aversion through better education, they timed the commitment to only take effect against a future raise the employee didn't have yet, so an employee's take-home pay never actually decreased, they simply kept a smaller slice of money they hadn't received yet.
the move
The Save More Tomorrow (SMarT) program, field-tested at a midwestern manufacturer in 1998, let employees pre-commit today to automatically increasing their 401(k) contribution rate by a set percentage at each future pay raise, with the option to opt out at any time — but requiring an active decision to stop rather than an active decision to keep contributing more.
why it works
Letting employees pre-commit today to automatically increasing their contribution rate at each future pay raise, rather than asking them to increase it against their current paycheck, meant the loss aversion that blocked immediate contribution increases simply never activated, since no current, felt loss ever occurred, the change only ever touched money the employee hadn't yet received. Requiring an active decision to opt out, rather than an active decision to keep contributing more, used default inertia in the same direction as the beneficial behavior, which is why 78% of participants stayed enrolled through four subsequent pay raises rather than reverting. This combination of never-felt loss and default persistence is what drove participating employees' savings rates from an average of 3.5% to 13.6% of pay over 40 months, a result financial education campaigns explaining the identical underlying math had consistently failed to achieve, and why the model scaled to more than 15 million American workers and eventually became a legally required default under the SECURE 2.0 Act.
the payoff
Participating employees' savings rates rose from an average of 3.5% to 13.6% of pay over 40 months, with 78% of participants still enrolled through four subsequent pay raises rather than opting out. The program has since scaled to more than 15 million American workers, and in December 2022 the US Congress made auto-escalation a required default feature of most new 401(k) plans under the SECURE 2.0 Act, effective 2025.
where it breaks
The mechanism depends on there actually being a predictable future resource, like a raise, to attach the commitment to — a behavior change with no equivalent future windfall to time against would have nothing to defer the felt loss onto, meaning this specific technique doesn't generalize to every loss-averse resistance, only ones where a future gain the person doesn't yet possess exists to absorb the change. It also depends on the opt-out mechanism genuinely being available and not so burdensome that it becomes coercive, since the program's ethical and practical legitimacy rests on the increase remaining a real, reversible choice rather than a trap disguised as a default. And because the mechanism specifically neutralizes loss aversion tied to a change in a resource level, it doesn't address other reasons someone might resist the same underlying behavior, distrust of the institution managing the funds, a genuine need for more current income regardless of framing, or skepticism about the retirement system itself, all of which would require separate interventions this technique alone can't solve.
what came after
Save More Tomorrow is a foundational case in behavioral economics and libertarian paternalism, cited alongside default-opt-out organ donation and Nudge-style choice architecture as proof that timing a beneficial commitment against future rather than present resources can neutralize loss aversion at essentially zero cost — the same auto-escalation logic now underlies retirement, savings and even carbon-offset programs designed to convert stated intentions into sustained behavior.
references
- [1]Why Planning To "Save More Tomorrow" Actually WorksKitces.com, 2023kitces.com
- [2]Save More Tomorrow™: Using Behavioral Economics to Increase Employee SavingUCLA Anderson School of Management, 2004anderson.ucla.edu