The moral behavior of ethics professors: A replication-extension in German-speaking countries

ABSTRACT

What is the relation between ethical reflection and moral behavior? Does professional reflection on ethical issues positively impact moral behaviors? To address these questions, Schwitzgebel and Rust empirically investigated if philosophy professors engaged with ethics on a professional basis behave any morally better or, at least, more consistently with their expressed values than do non-ethicist professors. Findings from their original US-based sample indicated that neither is the case, suggesting that there is no positive influence of ethical reflection on moral action. In the study at hand, we attempted to cross-validate this pattern of results in the German-speaking countries and surveyed 417 professors using a replication-extension research design. Our results indicate a successful replication of the original effect that ethicists do not behave any morally better compared to other academics across the vast majority of normative issues. Yet, unlike the original study, we found mixed results on normative attitudes generally. On some issues, ethicists and philosophers even expressed more lenient attitudes. However, one issue on which ethicists not only held stronger normative attitudes but also reported better corresponding moral behaviors was vegetarianism.

KEYWORDS:

• Experimental philosophy
• replication-extension
• moral attitudes
• moral behavior

Acknowledgment

We are especially thankful for the continued assistance of Eric Schwitzgebel, whose guidance has contributed greatly to this project in many ways. We are also grateful for helpful comments, suggestions, and generous support from Florian Cova, Ferenc Kemény, Edouard Machery, Lukas Meyer, Norbert Paulo, Thomas Pölzler, Joshua Rust, Peter Singer, Markus Seethaler, and Pascale Willemsen.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplemental material

Supplemental data for this article can be accessed here.

Notes

1. Our translation.

2. We owe this citation and its place to Schwitzgebel and Rust (2014).

3. These five states were California, Florida, North Carolina, Minnesota, and Washington.

4. With our replication study we also aim to address the problem of the replication crisis in social science fields (Schmidt, 2009) like moral psychology and adjacent areas of enquiry such as experimental philosophy, even though it has recently been suggested that experimental philosophy replicates at a much higher rate than similar disciplines (Cova et al., 2018). Moreover, in order to contribute to this ideal of a cumulative science and adopt the “best-practices” (Cova et al., 2018, p. 10) suggested, we pre-registered this study in the Open Science Framework (Schönegger & Wagner, 2018) following the “replication recipe” (Brandt et al., 2014).

5. In German: “Ethik,” “Angewandte Ethik,” “Normative Theorie,” “Ethik der X,” and „X-Ethik“.

6. We used Polldaddy as survey platform.

7. Following Schwitzgebel and Rust, we decided to limit the “eating meat” item to eating the meat of mammals, as there might be some who draw an ethical distinction between mammals and fish. Focusing on mammals was a way to preempt confusion. This means that any further classification of “vegetarianism” could also include, at least in part, a certain number of pescatarians. We do recognize that not including poultry could have distorting impacts on the interpretability of the result. However, we did decide to stick with the original version for replication reasons. All further mention of “meat” will be taken to mean “meat of mammals.”

8. The first part of the survey utilized the “morally good – morally bad” dichotomy, whereas the extension made use of the “strongly disagree – strongly agree” variant of the same scale.

9. Like Schwitzgebel and Rust (2014), we gave the option of free text commentary at the end of each page to give participants the opportunity to clarify their answers or raise objections to the wording or methodology. We gave these comments a cautious reading. Most of them raised objections to specific wordings, pointed towards the importance of context in certain item descriptions, demanded more background information on certain questions, expressed disagreement with our choice of some items or possible answers, and commented more generally on whether or not they liked the survey or what they felt it missed. None of these comments suggested a fundamental flaw in the study design, similar to the responses to the original study (Schwitzgebel & Rust, 2014, p. 298–299).

10. According to the discipline’s standards, we put the alpha-level of significance at 0.05.

11. Like the original, we collapsed the values 1 through 4 into “morally bad” and values 6 through 9 into “morally good” for proportional analysis. When analyzing the good side of the scale, the bad side and the neutral values were collapsed and vice versa.

12. We aggregated yes and no questions on moral behavior, counting “yes” as 1, and “no” or “do not recall” as 0. On behavioral measures where we only had a continuous variable, we split the distribution at the median and assigned “1” to the upper and “0” to the lower 50% of the distribution. As there was thus exactly one point to get for moral behavior on every normative issue, participants could get between 0 and 7–8 points (depending on whether vegetarianism was excluded or not).

13. Non-ethicist philosophers fell behind on the issue of “answering students’ e-mails” as well as on one aggregate measure of moral behavior, while no such differences were observed in terms of normative attitudes. Though this effect does not allow for an inference with regard to the effects of professional ethical reflection, it speaks for a local inert discovery view as an explanation for non-ethicist philosophers’ moral insights, as non-ethicist philosophers did not live up to the equally strong attitudes they expressed in this instance compared to ethicists and non-philosophers.

14. It is very important to note, however, that in our talk of “relative stringency or leniency,” we are not talking about relativity among normative issues, but relativity among groups with regards to particular issues. The issue of theft, for instance, was unequivocally perceived as the most morally reprehensible by all groups. In fact, the selected normative issues are not straightforwardly comparable in their absolute numbers, as the responses to them might differ with the intensity of the example chosen – one could ask about a theft of 50 euros instead of 1000 euros, for example. Hence, in our talk about somewhat and properly moral, we are talking about relative differences among ethicists, non-ethicist philosophers, and non-philosophers, arguing that the issues which ethicists favor compared to non-philosophers might reflect properly moral issues.

15. Participants who commented on charity were mostly concerned with background assumptions of duty and the classification used for the question, resulting in a demand for a better focus on additional circumstances such as family situation and other obligations.

16. Notice how this provides a straightforward explanation of ethicists’ more stringent attitudes and better behavior in terms of vegetarianism: It is highly recommended in terms of effectiveness, and in our contemporary society it’s not too difficult to implement. It also deserves mention, however, that the difference between the original US sample and our European sample was much larger than the difference that was observed within cultures, that is, between groups. This indicates that culture is a more influential factor on eating meat than ethical reflection, which, however, does not thereby undermine ethical reflection as a significant factor. The difference between samples might reflect cultural differences or a general trend towards the reduction of meat consumption in the course of the last decade, at least among academic populations, as only 4.3% of Germans aged 18 to 79 (Gert et al., 2016) subscribe to a vegetarian diet. Interestingly, the original study also found 27% of ethicists (Schwitzgebel & Rust, 2014, p. 307) reporting strict vegetarianism, which is comparable to our number of 30%. On this picture, while the proportion of full-vegetarian ethicists remained roughly the same, we were able to observe a significant trend toward eating less meat across groups.

17. Following Haidt’s understanding, one would expect a distribution favoring harm and fairness considerations, for example, in charity and vegetarianism, over others, such as calling one’s mother. By contrast, one would expect this prioritization to change in other cultures. Indeed, one generally finds a varying pattern in other contexts (Graham, Meindl, Beall, Johnson, & Zhang, 2016) and, more specifically, with regard to different religions (Levine et al., 2018). To give an example, in Confucian cultures civility takes a central role in morality (Buchtel et al., 2015), which would suggest a different weighing, one which would favor calling one’s mother substantially more. These cross-cultural analyses, however, can only illuminate the background of our research here, as they cannot explain the group differences studied.

18. A similar worry has been raised by the authors of the original study (Schwitzgebel & Rust, 2014, p. 299–300).

19. In addition, non-philosophers also judged more stringently than non-ethicist philosophers with regards to “paying academic fees.”

Additional information

Notes on contributors

Philipp Schönegger

Philipp Schönegger is a graduate student in moral, political and legal philosophy at the University of St. Andrews and the University of Stirling.

Johannes Wagner

Johannes Wagner is a graduate student in psychology at the University of Graz.