Abstract
Two rationales have emerged for why individuals keep their promises: (a) an emotional commitment to keep actions and words consistent, a commitment rationale and (b) avoidance of guilt due to not meeting the expectations of the promisee, an expectations rationale. We propose a new dichotomy with clearer distinctions between rationales: (1) an internal consistency rationale, which is the desire to keep actions and words consistent regardless of others’ awareness of the promise and (2) a communication rationale, which captures all aspects of promise keeping that are associated with the promisee having learned of the promise, including but not limited to promisee expectations. Using an experiment that manipulates whether promises are delivered, we find no support for the internal consistency rationale; only delivered promises are relevant. In a second experiment designed to better understand what aspect of promise delivery influences promisor behavior, we manipulate whether the promise is delivered before or after the promisee is able to take a trusting action. We find late-arriving promises are relevant though not as relevant as promises delivered before the promisee chooses whether to take the trusting action. We conclude that implicit contracting does not fully explain promise keeping, because had it done so, late-arriving promises would also be irrelevant.
Similar content being viewed by others
Notes
Kawagoe and Narita (2014) do not inform trustors that their first-order beliefs will be communicated to trustees.
Ellingsen et al. (2010) eliminate the potential for false consensus bias by directly informing participants of others’ beliefs. They find no evidence that the expectations of others matter. However, the design used by Ellingsen et al. (2010) does not include promises. Further, Khalmetski et al. (2015), using within-subject data, call into question the conclusions of Ellingsen et al. (2010), finding that individuals do condition their responses on the expectations of others, but that variation between individuals masks important behavior if analyzed only on an aggregate level. In neither study are trustors informed their beliefs will be communicated to trustees.
We prefer the term “communication” to “social obligation” as social obligation appears to make an a priori assumption regarding the promisor’s emotions.
If the promisor is given a binary choice on whether or not to return the trust, the communication rationale would be evidenced by finding promisors who would not break their promise to return the trust if their promise were delivered, but would not otherwise act in a trustworthy manner. It is not unreasonable to conjecture those who keep their promises would also be those who would act trustworthy even without a promise. In our design, a promisor can act less kindly if the message is not delivered than if it is delivered, but still not take maximum advantage of the other individual.
It is conceivable that delivered non-promises could decrease trustworthiness relative to undelivered non-promises. If a blank piece of paper or an irrelevant message were interpreted by the recipient as “don’t count on me,” it could ease trustees’ guilt for not acting in a trustworthy manner.
In our experiment I = 5 dollars and Y = 19 dollars. Although the growth factor Y/I is somewhat higher in our experiment than in Berg et al. (1995) study, 3.8 versus 3, the meta-analysis of Johnson and Mislin (2011) suggests that once the growth factor becomes this large trustees tend to keep for themselves any profit due to the increased factor.
In those instances where a promise of “half” was made, we coded as 9.5. Coding as either 9 or 10 would not change the nature of the results.
The correlation between the promised return and the decision regarding how much would be returned if the message were not delivered is 0.13, not significantly different from zero.
To perform the test, the data were cast in a 2 × 2 format as presented in Table 5.
Table 5 indicates that trustee #17 was omitted from the second-definition tests. This trustee’s promise was “will be fair, I promise.” Because there are two reasonable ways to define fairness (return at least 5 or return approximately half, i.e., 9 or 10), we were not comfortable in classifying this trustee’s promise contingent on non-delivery because the Ret|Inter amount was 7. In the Table 5 analysis we required observations under both delivery conditions and so had to delete that observation.
It is odd that the medians for the two conditions are both equal to 9, but the Wilcoxon signed ranks test yields a significant difference. This occurs because the test, strictly speaking, tests differences in signed ranks rather the median and therefore can detect other distributional differences. It is possible that the ranks of one group (except for those observations at the median) are substantially higher or lower than the ranks of the other group (again, except for those at the median). Also, note in Table 2 that there are no instances, considering promises only, in which Ret|Late is greater than Ret|OT, but several in the other direction, with some of them large in magnitude.
As mentioned, we did not collect expectations data.
References
Berg, J., Dickhaut, J., & McCabe, K. (1995). Trust, reciprocity and social history. Games and Economic Behavior, 10(1), 122–142.
Bicchieri, C. (2006). The grammar of society: The nature and dynamics of social norms. New York: Cambridge University Press.
Charness, G., & Dufwenberg, M. (2006). Promises and partnership. Econometrica, 74(6), 1579–1601.
Charness, G., & Dufwenberg, M. (2010). Broken promises: An experiment. Economic Letters, 107(2), 281–283.
Douthit, J., Kearney, L., & Stevens, D. (2012). Can agent cheap talk mitigate agency problems in the presence of a noisy performance measure? An experimental test in a single- and multi-period setting. Journal of Management Accounting Research, 24, 135–158.
Dufwenberg, M., & Gneezy, U. (2000). Measuring beliefs in an experimental lost wallet game. Games and Economic Behavior, 30(2), 163–182.
Ederer, F. & Stremitzer, A. (2016). Promises and expectations. Yale University working paper.
Ellingsen, T., & Johannesson, M. (2004a). Promises, threats and fairness. The Economic Journal, 114(April), 397–420.
Ellingsen, T., & Johannesson, M. (2004b). Is there a hold-up problem? The Scandinavian Journal of Economics, 106(3), 475–494.
Ellingsen, T., Johannesson, M., Tjøtta, S., & Torsvik, G. (2010). Testing guilt aversion. Games and Economic Behavior, 68(1), 95–107.
Hays, W. (1963). Statistics. New York: Holt, Rinehart and Winston.
Ismayilov, H., & Potters, J. (2016). Why do promises affect trustworthiness, or do they? Experimental Economics, 19(2), 382–393.
Johnson, N., & Mislin, A. (2011). Trust games: A meta-analysis. Journal of Economic Psychology, 32(5), 865–889.
Kawagoe, T., & Narita, Y. (2014). Guilt aversion revisited: An experimental test of a new model. Journal of Economic Behavior & Organization, 102, 1–9.
Khalmetski, K., Ockenfels, A., & Werner, P. (2015). Surprising gifts: Theory and laboratory evidence. Journal of Economic Theory, 159, 163–208.
Klein, D., & O’Flaherty, B. (1993). A game-theoretic rendering of promises and threats. Journal of Economic Behavior and Organizations, 21(3), 295–314.
Lundquist, T., Ellingsen, T., Gribbe, E., & Johannesson, M. (2009). The aversion to lying. Journal of Economic Behavior and Organizations, 70(1–2), 81–92.
Siegel, S., & Castellan, N. J. (1988). Nonparametric statistics for the behavioral sciences (2nd ed.). New York: McGraw-Hill.
Vanberg, C. (2008). Why do people keep their promises? An experimental test of two explanations. Econometrica, 76, 1467–1480.
Zizzo, D. (2010). Experiment demand effects in economic experiments. Experimental Economics, 13, 75–98.
Author information
Authors and Affiliations
Corresponding author
Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
About this article
Cite this article
Schwartz, S., Spires, E. & Young, R. Why do people keep their promises? A further investigation. Exp Econ 22, 530–551 (2019). https://doi.org/10.1007/s10683-018-9567-2
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10683-018-9567-2