Focal Points (Equal Division Heuristic)

Focal Points

Historical precedent is not the only way a default option might be established. Sometimes, an option may become a default by virtue of its conspicuousness or psychological prominence. For example, Schelling (1960) has argued that equal division is often chosen merely by virtue of its mathematical simplicity – not because equality is compelling in a logical, moral, or legal sense. He notes that the results of long negotiations involving many complicated issues and parties with greatly differing bargaining power often converge on some crudely simple, but psychologically conspicuous benchmark like “splitting the difference” or “the 38th parallel,” because those focal points possess prominence, uniqueness, simplicity, or something else that makes them qualitatively differentiable from the continuum of alternatives that one could consider.

Schelling’s speculation is supported by experimental evidence. Harris and Joyce (1980) found that when people allocated profits among members of a group venture, they chose to distribute profits equally, even when members generated different amounts of revenue. However, when asked to allocate expenses, equal division was again favored, even though this left profits unequally distributed. The authors commented that respondents’ choices seemed to be motivated by the simple heuristic of dividing equally rather than a true preference for equality of outcome per se; that responses reflected an overlearned share-and-share-alike rule “without regard to what is being shared or what each participant’s ultimate fate is” (p. 177). Similarly, Rabin (1997) observed that people generally have a “one-pie-at-a-time” conception of allocation problems, favoring equal distribution of the focal pie, without considering the initial or final wealth levels of the parties involved.

Allison and Messick (1990) and Messick (1993) argue that “divide equally” is a social decision heuristic that is often applied with little deliberation. Results by Roch, Lane, Samuelson, Allison, and Dent (2000) support this view. In that study, respondents participated in a confidential, 8-person, resource-sharing task, and were told that they could stake the first claim to some amount of $60. Half of the respondents were put under cognitive load (by being required to remember the number 91704305) and half were not. The mean request of the cognitive load group was $8.18, which was not significantly different from their “fair” share of $7.50 ($60 ÷ 8 people). However, the control group, who was not impaired by the additional memory task, requested $17.00 – more than twice their fair share. The authors concluded that individuals first anchor on equal division, and adjust from that when they possess sufficient cognitive resources to do so. In the language used by Kahneman and Frederick (Chapter 2, this volume), being fair required only System 1, because equal division is an “obvious” solution, whereas being greedy required the help of System 2 to override the “obvious” solution proposed by System 1.

Much evidence suggests that the pull of equal allocation is strong for intra-personal allocation as well. Samuelson and Zeckhauser (1988, pp. 31–33) reported that about half of all Harvard employees divided their retirement funds equally between TIAA (a portfolio of bonds, commercial loans, mortgages, and real estate) and CREF (a broadly diversified common stock fund), and that most retained that equal division regardless of stock market performance. Benartzi and Thaler (2001) proposed, more generally, that investors often apply a “1/n heuristic” dividing their resources equally between whatever set of funds they happen to consider. Equal division may apply across time as well. Frederick and Loewenstein (2001) found that when intertemporal choices were framed as allocating goods among time periods, the modal response was equal division. The popularity of equal division contrasts with the typical findings of positive time preference (preference in favor of early periods) found when people are asked to choose between a smaller proximate reward and a larger distal reward and also contrasts with the preference for improvement when people are asked to choose between declining, flat, or improving sequences.