NOTE-1273: Leaving a Legacy: Intergenerational Allocations of Benefits and Burdens

Contribution

An experimental paper from the legacy programme of Wade-Benzoni and colleagues. It asks whether passing on burdens (debt, waste, disease) differs from passing on benefits (energy, money, food) when the recipient lives later and has no voice. It argues that intergenerational decisions are distinctive because they join a time delay to a self–other trade-off, and that this combination is what lets a decision bear on one’s legacy. Three studies measure attitudes and three measure allocations. For the record it is also the only available account of Wade-Benzoni 2006 (LIT-1716), which it cites three times.

Key insight

Bad outcomes weigh more than good ones, but which way that pushes depends on the context. In negotiation, burdens make people more selfish. Here, where the recipient is in the future and powerless, burdens made people less selfish. The authors explain the reversal by legacy: “people care more about avoiding leaving a burdensome negative legacy than about leaving a beneficial positive legacy” (p. 13). Legacy, in turn, is a matter of the self: “the motivation underlying legacies is largely based on the desire to extend the self into the future” (p. 26).

Assumptions

• Legacy needs others. “One cannot create a legacy by having a fleeting impact, or by affecting merely one’s own future self” (pp. 11–12, citing Wade-Benzoni 2006b). A legacy is an enduring effect on other people who come later.
• Legacy is symbolic immortality. Leaving a legacy “provides a symbolic form of immortality” (p. 7) and expresses “a desire to extend the self beyond mortal life” (p. 12, citing Becker, Kotre and McAdams).
• Bad is stronger than good (Baumeister et al. 2001, quoted at length on p. 10).
• Organizational successors stand in for generations. A replacement employee or another subsidiary in five years counts as a future generation (pp. 8, 20, 24).

Key results

• Experiment 1 (pp. 13–15; 76 undergraduates, waiting for a chance at basketball tickets). A global-warming passage framed as diminished food and water for future generations (benefits) or as spreading disease (burdens). A nine-item “modified” generativity scale (α = .86) was higher after burdens, M = 4.90 (SD 1.01), than after benefits, M = 4.10 (SD .96), F(1, 74) = 12.44, p = .001. Its items are about oneself living on: “I want to be remembered after I die”, “I hope that in some way, part of me will live on after I die”, “I feel that I have done something that will survive after I die”.
• Experiment 2 (pp. 15–17; 101 MBA students, class exercise). Consuming beneficial resources versus minimizing burdensome substances.
  • Legacy concern, two single items analysed separately because they correlated only r = .191: higher after burdens (5.89 vs 5.31), F(1, 99) = 9.42, p = .003. Avoiding a negative legacy mattered more than leaving a positive one (5.79 vs 5.44), F(1, 99) = 4.72, p = .03; no interaction.
  • Responsibility to future generations, one item: 6.0 vs 5.5, F(1, 99) = 7.22, p = .008.
  • Affinity for future generations, one item: 5.62 vs 5.0, F(1, 99) = 9.62, p = .003.
  • Ethical concerns, four items (α = .88): 5.5 vs 5.1, F(1, 99) = 4.64, p = .03.
• Experiment 3 (pp. 17–19; 64 MBA students). Recalling a time one allocated a burden, rather than a benefit, between oneself and a future person who could not reciprocate raised moral emotions (disgust, regret, guilt, shame, embarrassment; α = .88): 2.18 vs 1.54 on a 1–5 scale, t(62) = 2.63, p = .01.
• Experiment 4 (pp. 19–23; 32 university staff, paid $5plusupto$10). A departing employee splits a $100 sales fund between profit now (5% to oneself) and marketing that pays a successor 50% more, and decides how many of 20 boxes to deliver now or leave to the successor at 50% more cost. Everyone made both decisions. On a common metric, self-interest was higher for benefits (M = 13.10, SD 5.51) than for burdens (M = 8.03, SD 7.61), F(1, 28) = 10.58, reported p < .001. Having made the sales oneself made no difference.
• Experiment 5 (pp. 23–26; 131 MBA students). An energy executive keeps “Delta” energy or passes it to another subsidiary that will use it 50% better, either now (dictator context) or in five years (intergenerational context); or neutralizes Delta’s waste or passes it on at 50% more cost.
  • Present self-interest (Table 1, higher = more for oneself now): benefits 47.00 to a present other and 64.80 to a future other; burdens 44.03 and 37.91. Valence main effect F(1, 127) = 6.06, p = .02; interaction F(1, 127) = 2.89, reported p = .05. Only benefits with a future recipient differed from the other three cells.
  • “There were ethical considerations that came into play”, one item (Table 2): benefits 4.58 present and 3.80 future; burdens 4.26 and 5.37. Interaction F(1, 128) = 7.50, p = .007. Ethics predicted lower self-interest, B = −4.52, and with it entered the valence × timing interaction dropped to B = −16.91, p = .17; the test of the reduction was marginal, z = 1.84, p = .065.
• Experiment 6 (pp. 26–27; 92 undergraduates). The same task, all with a future recipient, but the present party was either oneself or another member of the other subsidiary (Table 3): benefits 52.43 with self and 38.48 without; burdens 32.18 and 37.59. Valence F(1, 88) = 4.50, p = .04; interaction F(1, 88) = 3.78, p = .05. Removing the self lowered present favouritism for benefits, t(88) = 2.91, p < .01, and did nothing for burdens.
• Practical recommendations (pp. 29–30). Have managers identify with their successors, as in common-ingroup work (Gaertner et al. 1989) and minimal groups; highlight ethics and the burdensome side of decisions.

Claims

id	claim	strength	support
C1	Framing an intergenerational decision as leaving burdens rather than using up benefits raises concern with one’s legacy, felt responsibility and affinity toward future generations, and ethical concern	moderate: two studies, one with single items, student and MBA samples, self-report right after a short passage	Experiments 1–2
C2	People are more concerned to avoid a negative legacy than to leave a positive one	weak to moderate: two single items correlated .19, one sample	Experiment 2
C3	People are more generous to future others with burdens than with benefits	moderate: three studies; the effect lies in keeping benefits when the recipient is in the future	Experiments 4–6
C4	The valence effect needs both a future recipient and the self as a party	weak: two interactions at reported p = .05, one of which (F = 2.89) is p ≈ .09 two-tailed	Experiments 5–6
C5	Ethical considerations underlie the valence effect	weak: one item measured after the decision, marginal reduction test	Experiment 5
C6	Legacy concern, the wish to extend the self beyond death, explains the allocation effects	weak: legacy is measured only in the attitude studies and never with an allocation	General Discussion

Concepts

• legacy: an enduring effect of one’s behaviour on other people who come later; a route to symbolic immortality.
• intergenerational dilemma: a choice in which what serves future others requires a sacrifice now, marked by power asymmetry, no direct reciprocity, a self–other trade-off, a time delay, role transition and growing consequences (p. 19).
• intergenerational generosity (beneficence): allocating more benefit, or less burden, to future others than self-interest would.
• resource valence: whether what is allocated is a benefit or a burden.
• generativity: as used here, a nine-item scale of wanting to be remembered and to make a lasting mark, presented as “concern for and commitment to the well-being of future generations” (p. 14).

Connections

• Wade-Benzoni 2006 (LIT-1716), unread. Cited for legacy as symbolic immortality (p. 7), for the claim that a legacy cannot be left to one’s own future self (p. 12), and among the sources of the list of intergenerational features (p. 19). See “Read in its place”.
• Wade-Benzoni et al. 2012 (LIT-1760), unread. The next step: mortality reminders instead of valence.
• McAdams and de St. Aubin 1992 (LIT-1777), unread. The source of the modified generativity scale. The items kept are those on living on and being remembered.
• Lifton (LIT-1714) and terror management. Becker’s Denial of Death is cited for the wish to extend the self beyond death (p. 12).
• Common in-group identity (NOTE-1286). The authors recommend making managers identify with their successors and cite Gaertner et al. 1989, the same programme.
• Hernandez, Chen and Wade-Benzoni 2006, not held. Cited for the claim that thinking about one’s legacy turns future generations “from a social category of total strangers that they will never meet to one of life partners” (p. 15).

Bearing on the record

• The legacy question (QUESTION-301). These are inferences; the paper does not discuss egoism.

  • The authors take the thesis’s reading as their premise. Legacy is the wish “to extend the self beyond mortal life” (p. 12) and “to extend the self into the future” (p. 26). But they do not test it. Nothing measures how far the self extends, and legacy concern is never measured in a study that measures generosity.
  • What extends is the self, not the self’s circle. The generativity items are about oneself being remembered and living on. Future people figure as the medium of that survival, not as part of the self. That is the temporal extension in QUESTION-301, symbolic immortality, rather than a social extension to include those who come after.
  • Experiment 6 cuts against an extended-self reading of the generosity. Putting the self into the trade-off raised self-favouring with benefits and left burdens unchanged. What the self’s presence did was make people keep more, not give more. The generosity with burdens is as well explained by a moral brake on imposing harm on the powerless, which the authors themselves invoke (p. 12), as by legacy.

• The owner’s thesis (CLAIM-491). The paper’s framing is the thesis in time: acting for future generations as acting for a self that outlives the body. Its data do not separate that reading from plain self-interest restrained by ethics.

• The triviality objection (CLAIM-492). “Affinity for future generations” rose with the burden framing (5.62 vs 5.0), alongside responsibility and ethical concern. One item, moving with every measure of caring, cannot show that affinity comes apart from caring.

• Batson’s objection (CLAIM-494). Not addressed.

• A dated note on QUESTION-301 is proposed. No note on CLAIM-491 itself: the paper names the thesis’s reading but gives it no test. Proposed text:

  Wade-Benzoni, Sondak and Galinsky (NOTE-1273), read in place of Wade-Benzoni 2006, define legacy as the wish to extend the self beyond death and show that leaving burdens, rather than using up benefits, raises concern with being remembered and generosity to future others. But legacy concern and generosity are never measured in the same study, and the generosity effect is people keeping benefits for themselves when the other person will get them later. Including the self in the trade-off made people keep more, not give more. The experiments show self-interest discounting future others and a moral brake on passing harm forward, more than a self that includes them.

• No THEORY filed.

• No ML instruction.

Limitations

• Two halves never joined. Experiments 1–3 measure legacy, affinity and emotion; Experiments 4–6 measure allocation. No study has both, so the claim that legacy concern drives allocation is inferred across studies.
• Small samples and class exercises. Experiment 4 had 32 participants in four cells; Experiments 2, 3 and 5 were MBA class exercises debriefed in class.
• Hypothetical stakes, except Experiment 4’s few dollars. Future generations are replacements and subsidiaries, five years on at most.
• Single items for legacy, responsibility, affinity and ethics. The two legacy items correlated only .19.
• The generativity scale is a selection. The nine items kept are about one’s own survival and mark, not care for the next generation, so it measures legacy striving more than generativity in McAdams’s sense.
• Reported p-values that do not match their statistics.
  • Experiment 4: F(1, 28) = 10.58 is printed as p < .001; it is p ≈ .003.
  • Experiment 5: the interaction F(1, 127) = 2.89 is printed as p = .05; two-tailed it is p ≈ .09. The paper does not say it tested one-tailed. This is the result behind Hypothesis 8.
• Typos.
  • “we a conducted” (p. 16); “a greater a sense” (p. 20); “SE = 12,31” for 12.31 (p. 25).
  • Eisenberg is cited as 2000 in the text (p. 17) and 2001 in the references, with volume 51 (the 2000 volume).
  • Taylor 1991 is given as Psychological Bulletin 11 (the volume is 110).
  • “Organizational Behavioral and Human Decision Processes” (twice, for Organizational Behavior).

Open questions

• Does legacy concern, measured in the same participants, predict or mediate the allocation? A study that measured both would test the paper’s central inference.
• Is the burden effect legacy or the moral asymmetry between imposing harm and withholding benefit? Telling participants that no one will ever know who left the burden would remove the legacy route and leave the moral one.
• What did Wade-Benzoni 2006 argue about whether intergenerational beneficence is altruism or egoism? The 2010 paper does not say; Wade-Benzoni and Tost 2009, “The egoism and altruism of intergenerational behavior”, which the record does not hold, is the likely place.