The Study About Too Much Choice Doesn't Hold Up. The One About Your 401(k) Does.

The retirement industry had a choice about how to respond to one of the most famous findings in consumer psychology, and it did not take the fix most people assume it did. In 2000, a jam-tasting table in a California supermarket became the empirical backbone of an entire genre of "paradox of choice" writing: more options, worse decisions. A decade later, a rigorous meta-analysis of fifty follow-up studies found the underlying effect was, on average, close to zero. What survived that reckoning, and what plan sponsors, Vanguard, and eventually Congress actually built in response to a narrower, real version of the finding, is a quieter story: the number of funds in your 401(k) menu measurably predicts whether you have a 401(k) at all, and the fix that scaled had nothing to do with the menu.

The experiment everyone remembers

Psychologists Sheena Iyengar and Mark Lepper set up two tasting tables at an upscale California supermarket: one displaying 24 flavors of jam, the other just 6. More shoppers stopped at the 24-flavor table, but when it came to actually buying, the results flipped: roughly 40% of shoppers at the 6-jam table bought a jar, versus about 3% at the 24-jam table. More options pulled people in, then somehow talked them out of buying anything at all. The finding became the empirical backbone of Barry Schwartz's bestselling The Paradox of Choice and a permanent fixture of writing about menus, dating apps, and streaming platforms.

The reckoning nobody retells

In 2010, psychologists Benjamin Scheibehenne, Rainer Greifeneder, and Peter Todd pooled 63 experimental conditions across 50 studies, more than 5,000 total participants, that had tried to replicate or extend the original result. Averaged across all of it, the choice-overload effect was statistically indistinguishable from zero. Some studies found a strong version of the jam-table effect; a comparable number found none, or found more options actually helped. The variance was enormous, enough to conclude the phenomenon isn't a stable law of psychology so much as something that shows up under specific, still poorly understood conditions and vanishes under others. That's an uncomfortable result for anyone who's built a talk or a marketing pitch on the jam study. It makes the one domain where a large, high-stakes, real-money version of the effect does hold up look considerably more interesting by contrast.

The study that isn't a supermarket jam table

Around the same time, Iyengar turned the same question toward retirement savings. Working with Gur Huberman and Wei Jiang, she analyzed enrollment data from nearly 800,000 employees across hundreds of 401(k) plans with varying numbers of fund options, published in 2004. Unlike the jam table, this measured a real, high-stakes financial decision, whether to enroll in a retirement plan at all, across plans that varied naturally in menu size. Each additional fund added to a plan's lineup was associated with a 0.15 to 0.2 percentage point decline in the probability an eligible employee participated. Plans offering a small, manageable handful of funds saw meaningfully higher participation than plans offering ten or more.

The likely reason the effect survives here but not broadly: choosing jam is low-stakes and reversible. Choosing a 401(k) fund lineup is technical, involves concepts most employees have no training in, starting with the Roth-versus-traditional choice before they even get to picking a fund, and the cost of freezing is real money compounding for decades. Overwhelm concentrates where the decision is already hard and the cost of doing nothing is real.

The fix nobody built by shrinking the menu

The retirement industry read this research and largely did not respond by cutting fund menus down to a handful. It built a mechanism that makes the size of the menu almost irrelevant: automatic enrollment, paired with a default investment nobody has to actively choose. Vanguard's 2026 "How America Saves" report, the 25th consecutive edition of the industry's most closely watched dataset, makes the mechanism clear: overall participation among eligible employees hit a record 86% in 2025, up from 65% when Vanguard began tracking the data twenty-five years earlier. The number that isolates the effect is starker: employees in plans with automatic enrollment participated at 94% in 2025. Employees in plans that still required an active opt-in participated at just 64%: a thirty-point gap, explained almost entirely by whether the employee had to make an active choice or simply had to actively decline one. Ninety-six percent of Vanguard plans now default un-engaged participants into a professionally managed target-date fund rather than an intimidating full menu; nearly 70% of all participants now hold that kind of default allocation.

Richard Thaler and Shlomo Benartzi's related "Save More Tomorrow" research, first tested in 1998 and published in 2004, applied the same insight to contribution rates: employees who pre-committed future raises, rather than cutting current pay, saw their savings rate climb from 3.5% to 13.6% of pay over 40 months, almost entirely by removing the moment of active, painful choice each increase would otherwise have required. It became one of the most direct academic-to-policy pipelines in behavioral economics and helped earn Thaler the 2017 Nobel Prize.

What had been optional plan-design best practice became federal mandate at the end of 2022. The SECURE 2.0 Act requires most new 401(k) and 403(b) plans, established on or after December 29, 2022, to automatically enroll eligible employees starting with the 2025 plan year, at a default deferral rate of 3%–10% that must climb at least 1 percentage point annually to at least 10%, unless the employee opts out.

The honest complication

Defaults are also anchors. The same psychology that makes a default effective at getting people enrolled makes most people simply stay at whatever rate the default sets, indefinitely. For years the near-universal default was a modest 3%, a figure that traces less to research on what employees need and more to it being a round, unthreatening number. Industry data now show the most common default has shifted to 6%, specifically because sponsors recognized 3% wasn't a floor employees built up from; for most people, it was becoming the whole plan. SECURE 2.0's mandatory 1-point annual auto-escalation is the direct fix for that second-order problem: a default that moves on its own, so the same inertia that used to trap people at too low a rate now carries them upward instead.

The jam table taught a tidy but overstated lesson about human psychology. The 401(k) data taught the retirement industry, and eventually Congress, a narrower, better-supported one: you don't have to solve choice overload by removing choices. If your own plan still leaves you facing the full menu, check whether you're in the default fund by accident or on purpose. The number of options was never really the problem. Whether the good outcome happens automatically is. Congress's fix only reaches new plans; existing 401(k)s, IRAs, and small legacy accounts were left outside the mandate entirely, so the choice-overload fix Washington actually legislated still doesn't reach most of the money already in the system. For accounts nobody auto-enrolls you in today, like an IRA you have to open yourself, the same logic argues for building your own default: an automatic transfer, on day one.

---

This article is educational and is not financial, retirement, or investment advice.

Quick answers

Does having too many choices really hurt decision-making? Not reliably, as a general rule. A 2010 meta-analysis of 50 studies found the average "choice overload" effect across contexts was close to zero, with results varying enormously by situation.

Does the number of funds in a 401(k) actually affect enrollment? Yes. A study of nearly 800,000 employees found each additional fund option was associated with a 0.15–0.2 percentage point decrease in participation, with meaningfully higher participation in plans offering a handful of funds versus ten or more.

What actually fixed low 401(k) participation? Automatic enrollment, not smaller fund menus. Vanguard's 2026 data show 94% participation in auto-enrollment plans, versus 64% in plans requiring an active opt-in.

Is automatic enrollment required by law now? For new plans, largely yes. The SECURE 2.0 Act requires most 401(k) and 403(b) plans established on or after December 29, 2022 to automatically enroll eligible employees starting with the 2025 plan year, with exemptions for small and very new businesses.