Rethinking how risk aversion and impatience are linked with cognitive ability: experimental findings from agricultural students and farmers

ABSTRACT

Dohmen et al. (2010) find in their paper (‘Are Risk Aversion and Impatience Related to Cognitive Ability?'), which has been published in the American Economic Review, that risk aversion and impatience are negatively related to cognitive ability. This topic is important because controlling for cognitive ability might be necessary if someone is interested in the link of risk preferences or time preferences to real-world outcomes. We re-examine their key results by conducting an experimental study using two subject pools (agricultural students and farmers) and three levels of monetary incentives. Similar to Dohmen et al. (2010), our study finds the above-described negative correlations. However, the strength of the association is relatively small in the realm of risk aversion and negligible with impatience.

KEYWORDS:

• Risk aversion
• impatience
• cognitive abilities
• scientific replication

JEL:

• C91
• D91

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 It is not the goal to study intrinsic motivation. But briefly addressing this concept helps to illustrate that some activities are carried out because of other motives than money.

2 For example, the Cognitive reflection test is frequently used by cognitive psychologists (e.g. Pennycook and Rand 2020). They count the number of correct answers to the questions and treat the sum as metric.

3 In this paper, the terms agricultural students, students of agricultural economics, and students of the agricultural sciences are used interchangeably because the degree programmes had slightly different labels but very similar content.

4 In Germany, only 36% of the people working in agriculture are female. The share of women is particularly low among corporate farm managers (less than 10%; Deutscher Bauernverband Online 2018). In many empirical/experimental studies with farmers, the number of women is therefore low. For example, in the study by Mußhoff et al. (2018), all 114 subjects were male. While roughly half of the students in Germany studying agricultural economics are men and the other half women (e.g. HAZ Online 2016), considerably more women joined the study.

5 The multiple price lists described above to elicit risk aversion and impatience rest on the axioms of rational choice. For example, people are assumed to be in line with transitivity (i.e. choices are internally consistent) and monotonicity (i.e. better outcomes are preferred with a higher than a lower probability). However, multiple price lists are prone to inconsistent choices (Andersen et al. 2006; Engel and Kirchkamp 2019). In the Holt and Laury (2002) procedure, subjects can inconsistently switch back and forth between A and B, or choose the dominated option A in the 10th lottery pair. Similarly, inconsistent behaviors in Laury, McInnes, and Swarthout (2012) comprise multiple switching between option A and option B. Moreover, opting for option B instead of option A in the first row (same probability to win the prize, but nine weeks more waiting time associated with option B) could be interpreted as inconsistent unless subjects are indifferent between waiting and not waiting. Many studies ignore inconsistent subjects as long the fraction is relatively small (Holt and Laury 2002). In contrast, a large fraction of inconsistent subjects can distort the results, and as a consequence are mostly dropped out of the sample (but in a later step often subject of analysis; Jacobson and Petrie 2009; Hirschauer et al. 2013). We observe a low rate of inconsistent choices and followed the practice to keep these subjects in the sample: Only 10% (more precisely, 8% of the students and 12% of the farmers) of the subjects behave inconsistently in the Holt and Laury procedure. Even less inconsistent choices were made in the procedure of Laury, McInnes, and Swarthout (2012) to elicit impatience (5.33% of the subjects behave inconsistently). No differences in the population could be found.

6 This casts some doubts on the robustness of the stake size effect. However, it was not the primary goal of this study and merely a side effect. Thus, it should be addressed in future work with new data.

7 Noussair et al. (2013) find a positive association between being a member of a religious group and risk-averse behaviors. Carter et al. (2012) argue that the future is less discounted by religious people. In his literature review, Lane (2017) finds that the association of life satisfaction (subjective well-being) risk preferences is unclear. There is a plenty of evidence for both hypotheses positive mood leading to risk-averse behavior (mood maintenance hypothesis; Isen and Patrick 1983) and positive mood leading to risk-seeking behavior (affect infusion model; Forgas 1995). Lane (2017) states that life satisfaction reduces impatience over monetary rewards. There is also evidence that the individual's political attitude matter to explain risk preferences and time preferences. According to Kam (2012), there is a link between risk seeking and political participation, which can be explained via the mechanisms of excitement seeking and novelty seeking. Friehe and Tannenberg (2020) show that political regime shapes time preferences.

Additional information

Funding

The financial support of the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation – 388911356) is gratefully acknowledged.