Diversification heuristic
Also known as: Variety-seeking bias, 1/n heuristic
Asked to choose several items at once, people spread their picks more than they really want.
What it means
The diversification heuristic is the tendency to spread choices across the available options - a bit of everything - far more when the options are chosen together than one at a time. Offered several snacks in a single sitting, people pick variety; choosing week by week, they repeat a favourite, though tastes have not changed. The pull is partly that simultaneous choice frames the decision as a portfolio and partly that we hedge against not knowing our future appetite. A celebrated special case is the 1/n heuristic in retirement investing, where contributors tend to split money fairly evenly across the funds they actually use; lab and early field work suggested the menu's composition therefore drives the portfolio, though larger field studies find real savers use only a few funds and their equity exposure tracks the menu only loosely. Add a fund and the split shifts - the menu nudges the risk more than it sets it.
Why it happens
Two forces combine. First, presenting several choices together frames the decision as a portfolio, and a portfolio invites a spreading rule: variety feels safer and more sophisticated than putting everything in one basket. Second, when you choose now for consumption later, you cannot picture your future appetite. Facing that uncertainty, you hedge, buying one of everything so whatever mood arrives is covered. Sequential choice removes both pressures. Each decision is its own small bracket, appetite is known in the moment, and the natural move is to pick the single thing you most want right now. So the same person diversifies under simultaneous choice and repeats under sequential choice, not because tastes changed but because the format changed which rule felt appropriate.
What the evidence shows
The founding demonstration is Simonson's 1990 snack study: 64 percent of students who picked three snacks in one sitting chose three different ones, against 9 percent who chose fresh at each weekly class. Read and Loewenstein replicated the discrepancy and ruled out standard utility-maximizing explanations, showing it as a genuine framing effect rather than rational hedging. Benartzi and Thaler carried it into finance, documenting that retirement savers split contributions roughly evenly across plan funds, so a menu heavy in stock funds pushed savers into more equity. The core simultaneous-versus-sequential gap is robust and has been reproduced across snacks, lotteries and financial menus, though effect sizes vary and later work stresses that the effect is moderated by how intuitively people decide.
Where it breaks down
The strong reading, that menus mechanically dictate portfolios, does not survive scrutiny. Huberman and Jiang, studying half a million real 401(k) participants, found people spread evenly across the funds they use, but use only three or four regardless of whether four or fifty are offered, and equity exposure barely tracked the menu's stock share. So the 1/n tendency is real but conditional and bounded, not the automatic menu-capture the lab implied. There is a deeper caveat: equal weighting is not obviously a mistake. DeMiguel, Garlappi and Uppal showed that none of fourteen optimized models reliably beat the 1/N portfolio out of sample, because estimation error in the fancy models swamped their theoretical edge. Naive diversification can be a poor snack strategy and a defensible investment one.
Using it in practice
For anyone designing a menu, the lesson is that composition is a lever. Because people anchor on an even split and on the set they are shown, adding or removing options quietly reshapes the average allocation, so a plan's fund lineup or a form's default categories carry more weight than they appear to. If the goal is a considered choice rather than a reflexive spread, decouple the decisions: ask people to choose closer to consumption, or one at a time, and the variety inflation shrinks. If a sensible spread is the goal, simultaneous framing works with you. Either way, audit what your menu implies when split evenly, since for many users that is exactly the portfolio they will end up holding.
Examples
Picking three days of snacks at once, people choose three different bars; choosing fresh each day, they pick their favorite all three times.
Offered four funds in a workplace pension, savers put roughly a quarter in each. Add a fifth and the split becomes fifths — the menu, not the saver, sets the risk.
Picking a term's optional modules in one sitting, a student spreads across four departments; choosing term by term, they would have taken everything from the one subject they actually like.
Splitting a year-end donation across the six charities on a workplace giving portal, a donor gives each a sixth; add two more causes and the shares shrink to eighths, though their real priorities never moved.
Ordering a mixed case of wine in one online basket, a shopper picks twelve different bottles; buying one bottle before each dinner, they would keep repurchasing the single label they reliably enjoy.
First described in Simonson (1990); Read & Loewenstein (1995); Benartzi & Thaler (2001).
Key references
- DeMiguel, V., Garlappi, L., & Uppal, R. (2009). Optimal versus naive diversification: How inefficient is the 1/N portfolio strategy? Review of Financial Studies, 22(5), 1915-1953. doi.org/10.1093/rfs/hhm075
- Huberman, G., & Jiang, W. (2006). Offering versus choice in 401(k) plans: Equity exposure and number of funds. Journal of Finance, 61(2), 763-801. doi.org/10.1111/j.1540-6261.2006.00854.x
- Benartzi, S., & Thaler, R. H. (2001). Naive diversification strategies in defined contribution saving plans. American Economic Review, 91(1), 79-98. doi.org/10.1257/aer.91.1.79
- Read, D., & Loewenstein, G. (1995). Diversification bias: Explaining the discrepancy in variety seeking between combined and separated choices. Journal of Experimental Psychology: Applied, 1(1), 34-49. doi.org/10.1037/1076-898X.1.1.34
- Simonson, I. (1990). The effect of purchase quantity and timing on variety-seeking behavior. Journal of Marketing Research, 27(2), 150-162. doi.org/10.2307/3172842