The Behavior of Ethicists

for the Blackwell Companion to Experimental Philosophy

Eric Schwitzgebel

Department of Philosophy

University of California at Riverside

Riverside, CA 92521-0201

Joshua Rust

Department of Philosophy, Unit 8250

Stetson University

421 North Woodland Boulevard

DeLand, FL 32723

January 23, 2014

The Behavior of Ethicists

1. Introduction.

Arguably, one of the aims of studying ethics is moral self-improvement. In ancient philosophy, moral self-improvement is often treated as the foremost aim for the student of ethics – for example in Aristotle (4^th c. BCE/1962), Confucius (5^th c. BCE/2003), and Epictetus (2^nd c. CE/2008). Though 20^th and 21^st century philosophers appear, overall, to aim their ethical reflections more toward theoretical discovery than toward self-improvement, moral self-improvement plausibly remains among the goals of a significant portion of professional ethicists to the extent they use their philosophical training in ethics to help them reflect on, for example, to what extent they have a duty to donate to charity or whether it is morally permissible to eat meat, with the thought of acting upon their conclusions.

Two related questions thus invite empirical treatment: Is philosophical moral reflection of the sort practiced by professional ethicists in fact morally improving? And what sort of relationship is there between professional ethicists’ explicitly espoused moral principles and their practical moral behavior? Individual ethicists’ lives are sometimes examined with these questions in mind, especially the life of Martin Heidegger, notorious for his endorsement of Nazism (e.g., Sluga 1993; Young 1997; Faye 2005/2009); and general claims about the behavior of ethicists are sometimes made based on personal experience or broad plausibility considerations (e.g., Posner 1999; Knobe and Leiter 2007; Moeller 2009). However, until recently, systematic, quantitative research on these issues has been entirely lacking. To date, all published quantitative studies of the issue have been led by Eric Schwitzgebel and Joshua Rust, the two authors of this article, usually in collaboration with each other. Our general finding is this: On average, professional ethicists’ behavior is indistinguishable from the behavior of comparison groups of professors in other fields. Also, in one multi-variable study, we find ethicists neither more nor less likely than other professors to act in accord with their expressed moral attitudes.

2. Moral Behavior.

It is difficult to study the moral behavior of ethicists. Ethicists are a thinly distributed group that cannot normally be brought into the laboratory or observed in high rates in their daily lives. Self-report surveys can be conducted, but self-reports of moral behavior are likely to be distorted by the general tendency of survey respondents to present themselves in ways seen as socially desirable. Sometimes it is possible to directly observe the behavior of a substantial number of ethicists – such as in conference settings – but behavior in such settings might not be representative and might be distorted if the subject is aware of being observed. Also, naturalistic observational studies are likely to be confounded by other factors influencing the target behaviors. Furthermore, it is often contentious what behavior counts as moral or immoral. For these reasons among others, it is crucial to look for a diversity of convergent evidence before drawing firm conclusions.

In most of our studies, we have found no statistically detectable difference between the behavior of ethicists and non-ethicists. Below is a complete list of known attempts to find differences between ethicists’ and non-ethicists’ moral behavior. Some basic proportion data, confidence intervals, and test statistics are included parenthetically below to aid the reader in interpreting the null results. All the studies summarized below also included secondary measures and more complex measures (such as multiple regressions). We will not report those other measures here unless they generated materially different results.

Missing library books. Using online information about library holdings, we examined the rates at which relatively obscure philosophy books were missing from leading academic libraries in the U.S. and Britain (Schwitzgebel 2009). Ethics books were more likely to be missing than other philosophy books: 8.5% of ethics books that were off the shelf were missing or more than one year overdue, vs. 5.7% of non-ethics books, a risk ratio of about 1.5 to 1 (66/778 vs. 52/910, CI for difference +0.3% to +5.2% , Z = 2.2, p = .03).

Peer opinion about ethicists’ behavior in general: In a survey conducted at the 2007 Pacific Division meeting of the American Philosophical Association, we asked passersby if ethicists behave on average morally better, worse, or about the same as philosophers not specializing in ethics (Schwitzgebel and Rust 2009). Only a minority of respondents, 35%, expressed the view that ethicists behave on average better than do other philosophers, while 46% expressed the view that ethicists behave about the same and 19% expressed the view that ethicists behave worse (48/136 vs. 62/136 vs. 26/136, CI for “same” 37%-54%).

Peer ratings of the moral behavior of individual ethicists. At the same meeting, we asked a different group of respondents two questions about the moral behavior of specific, arbitrarily selected ethicists and specific, arbitrarily selected specialists in metaphysics and epistemology (the specialists in the respondents’ departments whose names came next after the respondents’ in alphabetical order, looping back from Z to A if necessary). 44% of respondents gave higher moral ratings to their ethicist colleagues than to their colleagues specializing in metaphysics and epistemology, while 26% rated them the same and 30% rated the ethicist worse – a statistically marginal trend to rate the ethicist better (55/125 vs. 33/125 vs. 37/125, one-proportion test of 55/92 vs. .5, exact p = .08).

Voter participation. On the assumption that voting in public elections is a civic duty, we looked at the voting participation rates of professors in five U.S. states (states that make individual voter participation data easily available to researchers; Schwitzgebel and Rust 2010, forthcoming). Among tenure-track professors, recorded voting participation in state databases was virtually identical for the three groups analyzed: 1.1 votes per year on average for professional ethicists, philosophers not specializing in ethics, and a comparison group of professors from departments other than philosophy (square-root transformed ANOVA, F = 0.3, p = .76). Specialization in political philosophy was also not predictive of voting (1.1 vs. 1.1, T = 0.6, p = .57).

Audience talking during formal conference presentations. On the assumption that discourteous acts are morally significant, at two APA meetings we coded the rates at which audiences in ethics sessions spoke audibly to each other during the formal presentation vs. the rates at which audiences in non-ethics sessions did so (Schwitzgebel, Rust, Huang, Moore, and Coates 2012). We detected no difference in talking rates between the groups: .010 vs. .009 instances per audience-hour in ethics vs. non-ethics sessions (15/1476 vs. 12/1324, CI for difference -.006 to +.008, Z = 0.3, p = .77).

Attempting to shut doors quietly when entering or exiting during formal conference presentations. At the same meetings, we coded the rates at which audiences entering or exiting during formal conference presentations allowed the door to shut noisily vs. made an effort to close the door quietly (Schwitzgebel et al. 2012). The simplest test showed a statistically marginal tendency toward fewer “slams” in ethics sessions: 18% of entrances or exits in ethics sessions vs. 24% in non-ethics sessions (52/286 vs. 77/315, Z = -1.9, p = .06). However, post-hoc analysis suggested the trials were not sufficiently independent, and we did not see a corresponding difference in the median percentage of slams per session: 18% in the ethics sessions vs. 15% in the non-ethics sessions (Mann-Whitney, W = 451, p = .95).

Leaving behind cups and trash in conference meeting rooms. In our view, it is generally polite for audience members at philosophy conferences to carry away their cups or trash, including handouts, when leaving a session. At four meetings of the APA, we examined the rates at which audiences in ethics and non-ethics sessions left behind cups and trash (Schwitzgebel et al. 2012). We found no difference for either measure: 17% of audience members in ethics sessions left behind cups, vs. 18% in non-ethics sessions (197/1173 vs. 284/1594, CI for difference -3.9% to +1.8%, Z = -0.7, p = .48); and 12% of both groups left behind trash (136/1173 vs. 188/1594, CI for difference -2.6% to +2.2%, Z = -0.2, p = .87). However, we did find that audiences in environmental ethics sessions left behind less trash: 3% vs. 12% (2/67 vs. 322/2700, Fisher’s exact, p = .02).

Paying conference registration fees. On the assumption that it is normally morally required to pay the relatively modest conference registration fees charged by the American Philosophical Association if one is participating in the conference, we examined the rates at which philosophers appearing on the Pacific Division program from 2006-2008 were recorded by the division as having paid their registration fees (Schwitzgebel 2013). Registration compliance was not statistically different between participants in ethics sessions and participants in non-ethics sessions: 74% vs. 76% (552/744 vs. 674/891, CI for difference -5.7% to +2.8%, Z = -0.7, p = .50).

Replying to student emails. We assume that it is generally better to reply to emails from undergraduates than to ignore them. With this in mind, we sent to three groups of U.S. professors email messages that were designed to look as though written by undergraduates (Rust and Schwitzgebel 2013). Ethicists replied to 62% of the emails we sent compared to a 59% reply rate for non-ethicist philosophers and a 58% reply rate for a comparison group of professors in fields other than philosophy – a trend that does not approach statistical significance despite over 3,000 trials (630/1021 vs. 641/1083 vs. 580/1005, χ² = 3.4, p = .18; CI for difference between ethicists and non-ethicist philosophers: -1.7% to +6.7%). (See Rust and Schwitzgebel 2013 for a discussion of the ethics of the use of deception in this study.)

Membership in one’s main academic disciplinary society. It is debatable whether there is a duty to support one’s main academic disciplinary society through payment of membership dues. In a sample of U.S. philosophers, we did not find higher membership rates in the American Philosophical Association among ethicists than among non-ethicists: 62% of sampled professors from both groups were listed as members on the APA’s website (218/354 vs. 224/362, CI for difference -7.4% to + 6.8%, Z = -0.1, p = .94; Schwitzgebel and Rust forthcoming).

Staying in touch with one’s mother. In 2009 we sent a multi-item survey on “professors’ moral attitudes and behavior” to a sample of about a thousand professors (Schwitzgebel and Rust forthcoming). We received replies from 198 ethicists, 208 philosophers not specializing in ethics, and 167 comparison professors from departments other than philosophy. Some of the questions asked for self-report of one’s own moral behavior. One such question was: “About how many days has it been since your last face-to-face or phone contact with your mother?” Assuming a norm of at least monthly contact with one’s mother, and excluding respondents with deceased mothers, we found the groups to report similarly high levels of contact with their mothers: 10% of ethicists reported its having been more than 30 days since last contact, vs. 11% of non-ethicists and 8% of non-philosophers (11/114 vs. 13/123 vs. 8/99, χ² = 0.4, p = .82; CI for difference between ethicists vs. non-ethicist philosophers -8.6% to +6.8%, Z = -0.2, p = .81). One secondary measure finds non-philosophers reporting more regular contact than either group of philosophers: 83% of non-philosophers reported contact twice a month or more, compared to 70% of ethicists and 74% of non-ethicist philosophers (χ² = 8.0, p = .02).

Meat eating. It might be morally good to refrain from eating meat. Another question in our survey was: “Think back on your last evening meal, not including snacks. Did you eat the meat of a mammal during that meal?” 37% of ethicist respondents answered “yes”, compared to 33% of non-ethicist philosophers and 45% of non-philosophers, a statistically marginal difference among the groups (69/185 vs. 65/195 vs. 75/165, χ² = 5.7, p = .06; ethicists vs. non-ethicist philosophers CI for difference -5.6% to +13.6%; ethicists vs. non-philosophers CI for difference -18.5% to +2.2%). One secondary measure finds ethicists more likely than the two other groups to report eating the meat of mammals at zero meals per week: 27% vs. 20% and 14% (χ² = 8.6, p = .01).

Organ donation. We also asked respondents whether their driver’s licenses indicated their willingness to be organ donors. 68% of ethicists answered “yes”, compared to 65% of non-ethicist philosophers and 69% of non-philosophers (125/184 vs. 126/193 vs. 111/161, χ² = 0.6, p = .75).

Blood donation. We also asked respondents how long it had been since they had donated blood. Excluding those reporting being ineligible to donate blood, 13% of ethicists reported having donated blood in 2008 or 2009 (the year of the survey), compared to 14% of non-ethicist philosophers and 10% of non-philosophers, not a statistically detectable difference (15/115 vs. 17/123 vs. 10/100, χ² = 0.8, p = .67).

Charitable donation: self-report. We also asked respondents what percentage of their income they had donated to charity in 2008. Assuming a norm of at least 3% charitable donation among professors, 66% of ethicists reported donating at least 3% of 2008 income, compared to 42% of non-ethicist philosophers and 63% of non-philosophers (120/181 vs. 80/190 vs. 98/156, χ² = 25.6, p < .001). Several other ways of analyzing these data (e.g., geometric mean reported percentage donation and percentage of respondents who reported donating nothing) yield similarly lower rates of self-reported charitable behavior among non-ethicist philosophers than among those in the two other groups.

Charitable donation: directly measured behavior. We also had one direct measure of charitable behavior. Half of our surveys were sent out with a charitable incentive to reply: a promise that we would donate $10 to the respondent’s choice among several well-known charities. In contrast with the self-report results, only the non-ethicist philosophers were detectably more likely to respond with the charity incentive than without it: ethicists 59% vs. 59%, non-ethicist philosophers 67% vs. 59%, non-philosophers 55% vs. 52% (104/177 vs. 94/160; 116/172 vs. 92/157; 84/154 vs. 83/160; non-ethicist philosophers one-tailed Z = 1.7, p = .048).

Survey response honesty. Survey respondents often overreport behavior or attitudes they see as socially desirable and underreport behavior or attitudes they see as socially undesirable. Our survey used several means to detect such a bias toward such “socially desirable responding”. The two groups of philosophers did not differ by this measure (the non-philosophers were not directly comparable), with 56% of both groups giving at least one suspicious survey response (110/198 vs. 117/208, CI for difference -10.4% to +9.0%, Z = -0.1, p = .89).

Summary. In total, this is 18 different measures of the moral behavior of ethicists. One of these measures finds ethicists behaving worse (the missing books measure), two find contrary results that depend on which comparison group is chosen (the two charity measures), and the remaining 15 primary measures reveal no statistically significant differences. Secondary measures suggest that those who attend environmental ethics sessions of the APA might litter less than participants in other sessions, that ethicists might be more likely than other professors to be strict vegetarians, and that ethicists might be less likely than other professors to stay in at least twice-monthly contact with their mothers. However, the p values for these effects are mediocre given the number of measures.

Meta-analysis. Most of the studies have the statistical power only to detect moderate to large effect sizes, and some studies show a trend favoring ethicists. To explore the possibility that these non-significant trends are manifestations of a small population difference undetected by the individual measures, we have combined all of the studies, excluding study of peer opinion about ethicists in general, into a meta-analysis. We have converted the results of each of the 17 target studies into a two-proportion comparison of the percentages of counternormative behavior. We recognize that some of the norms are controversial, especially avoiding meat and membership in one’s disciplinary society. However, in many cases, including specifically disciplinary membership and vegetarianism (Schwitzgebel and Rust forthcoming), we have evidence that the majority of professional ethicists in the U.S. endorse these norms. We converted the voting data into percentages by assuming a norm of at least one vote per year. We converted the individual peer ratings into percentages of ethicists vs. specialists in metaphysics and epistemology rated below the midpoint in comparison to other philosophers in the department. The remaining percentage comparisons are the ones reported above. Table 1 displays the relative risk of behavior coded as counternormative for each of the 17 studies:

TABLE 1: Relative risk of behavior coded as counternormative

measure	ethicists' % counternormative	non-ethicists % counternormative	relative risk	two-proportion Z score
Missing library books	8.4%	5.7%	1.48	-2.20
Peer ratings of individuals	19.7%	20.9%	0.94	+0.24
Less than one vote per year	37.6%	31.8%	1.18	-1.11
Audience members talking	1.0%	0.9%	1.12	-0.30
Conference door slamming	18.2%	24.4%	0.74	+1.88
Leaving behind cups	16.8%	17.8%	0.94	+0.70
Leaving behind trash	11.6%	11.8%	0.98	+0.16
Not paying registration fee	25.8%	24.4%	1.06	-0.67
Not replying to student emails	38.3%	40.8%	0.94	+1.18
APA non-membership	38.4%	38.1%	1.01	-0.08
Not recently calling mother	9.6%	10.6%	0.91	+0.23
Eating meat the previous night	37.3%	33.3%	1.12	-0.81
Not being an organ donor	32.1%	31.7%	0.92	+0.55
Not recently donating blood	87.0%	86.2%	1.01	-0.18
Self-reporting less than 3% of salary to charity	33.7%	57.9%	0.58	+4.82
Non-response to charity incentive	41.2%	32.6%	1.27	-1.69
At least one suspicious survey answer	55.6%	56.3%	0.99	+0.14
Equal weighted average	30.1%	31.0%	0.99 (geometric mean)	+0.17

A simple merge of all observations yields a total of 8,477 observations of ethicists’ behavior and 9,568 observations of the behavior of non-ethicist philosophers, among which we counted as counternormative 1,763 instances (20.8%) for ethicists vs. 2,022 (21.1%) for non-ethicists. If we can treat as representative this admittedly hodge-podge and unbalanced collection of observations, we can derive a nicely narrow 95% confidence interval for the difference in rates of counternormative behavior, an interval centered almost at zero: -1.5% to +0.8% (Z = 0.6, p = .57). Converting r to z', we can also create a confidence interval for the correlation between being an ethicist and exhibiting counternormative behavior: -.02 < population ρ < .01 (sample r = .004). However, since this simple merging fails to account for ethicists’ and non-ethicists’ different rates of participation in the studies, and it treats as homogeneous what is surely diverse, and it puts a large relative weight on the high-N conference and email studies, we thought it useful to offer an alternative analysis as well – one that weights all the studies equally. Again, we found no evidence of a difference in moral behavior between the two groups: 30.1% counternormative for ethicists vs. 31.0% for non-ethicists (Z = 0.7, p = .48, using the procedure of Rosenthal and Rosnow 1984/2008).

Despite these aggregate results, we leave it open as a possibility that ethicists really do behave morally better in some ways and perhaps morally worse in others, and that we have unfortunately chosen a set of measures and weightings that fail to reveal these differences. Our measures have been convenient and quantifiable. We haven’t examined whether ethicists, for example, are more likely to use a gentle tone with their children or whether they are more likely to resist the call of false ideology and authority (though see Schwitzgebel 2010 for some preliminary data on philosophers’ involvement in Nazism).

3. Relationships Between Moral Behavior and Moral Attitude.

Even if the study of ethics doesn’t make a difference to what we do, might it nevertheless at least make a difference to how we think about what we do. For example, ethicists’ attitudes and behavior might be more tightly correlated than non-ethicists’ (e.g., if they shape their attitudes to match their pre-existing behavior), or it might be be less tightly correlated (e.g., if ethical reasoning tends to explode excessively self-congratulatory norms without changing practical behavior).

We know of only one quantitative study exploring these issues: Schwitzgebel and Rust forthcoming (with some data further analyzed and discussed in Rust and Schwitzgebel 2013). This study asked U.S. ethicists, non-ethicist philosophers, and a sample of professors from other departments their opinions about the moral goodness or badness of ten types of behavior: stealing $1000, paying membership dues to one’s main disciplinary society, regularly voting in public elections, not keeping in at least monthly face-to-face or telephone contact with one’s mother, regularly eating the meat of mammals, being a blood donor, not being an organ donor, not consistently responding to student emails, donating 10% of one’s income to charity, and responding dishonestly to survey questions. For all the behaviors except theft, we collected self-reports of behavior later in the survey, and in some cases we had direct observational data on the same issues: The survey population was the same population for whom we had voting data, disciplinary membership data, and some of our email responsiveness data; and the survey itself generated observational data both about response honesty and about the effectiveness of our charity incentive on response rates. The behavioral data from this study we have discussed above in Section 2. This section relates those behavioral data to the data about moral attitudes. We emphasize that all data was de-identified for participants’ privacy, with names replaced by code numbers.

We found a general tendency for ethicists to embrace more stringent moral views overall than did the other two groups, but we found no systematic differences among the groups in the degree of consistency between their espoused norms and their self-reported or directly observed behavior.

The groups differed most in their opinions about vegetarianism and charitable donation. Asked about the morality of “regularly eating the meat of mammals, such as beef or pork”, 60% of ethicists rated it on the morally bad side of our response scale, compared to 45% of non-ethicist philosophers and only 19% of non-philosophers (χ² = 64.2, p < .001). Asked “About what percentage of income should the typical professor donate to charity? (Enter 0 if you think it’s not the case that the typical professor should donate to charity.)” 9% of ethicists entered “0” compared to 24% of non-ethicist philosophers and 25% of non-philosophers (χ² = 18.2, p < .001), and among the non-zeros, ethicists’ geometric mean answer was 5.9% of income, compared to 4.8% for the other two groups (ANOVA, F = 3.6, p = .03). As discussed above, despite these differences in expressed normative attitude, we did not find unequivocal evidence that ethicists ate less meat or gave more to charity.

We used correlational measures to examine the relationship between expressed normative attitude and either observed or self-reported moral behavior. Ethicists showed a higher correlation than the comparison groups in their expressed attitudes about the morality of voting, as measured on a 9-point scale – and their state-recorded voting rates: r = .36, compared to .14 for non-ethicist philosophers and only .01 for non-philosophers. (Using pairwise z_r conversions, the difference in correlation is non-significant between ethicists and non-ethicist philosophers (p = .11) but significant between ethicists and non-philosophers (p = .02).) However, ethicists showed a lower correlation than the other groups between their expressed attitude about how much the typical professor should donate to charity and how much they themselves reported having donated in 2008: r = .33, compared to .46 for the non-ethicist philosophers and .62 for the non-philosophers. (Using pairwise z_r conversions, the difference in correlation is marginally significant between ethicists and non-ethicist philosophers (p = .09) and significant between ethicists and non-philosophers (p = .003).) Merging all measures of attitude-behavior correlation in a meta-analysis, we found no overall differences among the groups, with ethicists showing a total estimated correlation of r = .20 compared to r = .24 for non-ethicist philosophers and r = .16 for non-philosophers (combining z_r’s as described in Rosenthal and Rosnow 1984/2008; final p’s ≥ .48).

4. Conclusion.

Evidence from a diversity of measures suggests that professional ethicists tend to behave on average very similarly to other professors, and one multi-measure study suggests that ethicists are neither more nor less likely than other professors to behave in accord with their expressed moral attitudes. However, all of the research so far is from a single research group; it remains to be confirmed by independent investigators. Also, it remains possible that different or more sensitive measures would reveal moral differences between ethicists and other groups of professors.

It does not follow that philosophical moral reflection is powerless to change moral behavior. But what influence philosophical reflection actually has on practical behavior, and in what directions, and under what conditions, are questions that – despite their centrality to a proper understanding of the value of philosophy as a discipline and the relationship in general between moral reasoning and moral behavior – remain wide open and virtually unexplored.

References:

Aristotle (4^th c. BCE/1962). Nicomachean ethics, trans. M. Ostwald. New York: Macmillan.

Confucius (5^th c. BCE.2003). Analects, trans. E. Slingerland. Indianapolis: Hackett.

Epictetus (2^nd c. CE/2008). Discourses and selected writings, trans. R. Dobbin. New York: Penguin.

Faye, Emmanuel (2005/2009). Heidegger, trans. M.B. Smith. Yale: New Haven.

Knobe, Joshua, and Brian Leiter (2007). The case for Nietzschean moral psychology. In B. Leiter & N. Sinhababu, eds., Nietzsche and Morality. Oxford: Oxford.

Moeller, Hans-Georg (2009). The moral fool. New York: Columbia.

Posner, Richard A. (1999). The problematics of moral and legal theory. Cambridge, MA: Harvard.

Rust, Joshua, and Eric Schwitzgebel (2013). Ethicists’ and nonethicists’ responsiveness to student emails: Relationships among expressed normative attitude, self-described behavior, and empirically observed behavior. Metaphilosophy 44, 350-371.

Schwitzgebel, Eric (2009). Do ethicists steal more books? Philosophical Psychology 22, 711-725.

Schwitzgebel, Eric (2010). Nazi philosophers. Blog post on The Splintered Mind, December 30, 2010, http://schwitzsplinters.blogspot.com.

Schwitzgebel, Eric (2013). Are ethicists any more likely to pay their registration fees at professional meetings? Economics & Philosophy, 29, 371-380.

Schwitzgebel, Eric, and Joshua Rust (2009). The moral behaviour of ethicists: Peer opinion. Mind 118, 1043-1059.

Schwitzgebel, Eric, and Joshua Rust (2010). Do ethicists and political philosophers vote more often than other professors? Review of Philosophy and Psychology 1, 189-199.

Schwitzgebel, Eric, and Joshua Rust (forthcoming). The moral behavior of ethics professors: Relationships among self-reported behavior, expressed normative attitude, and directly observed behavior. Philosophical Psychology.

Schwitzgebel, Eric, Joshua Rust, Linus Ta-Lun Huang, Alan T. Moore, and Justin Coates (2012). Ethicists’ courtesy at philosophy conference. Philosophical Psychology 25, 331-340.

Sluga, Hans (1993). Heidegger’s crisis. Cambridge, MA: Harvard.

Young, Julian (1997). Heidegger, philosophy, Nazism. Cambridge: Cambridge.