Extermination Extermination of Evil Sendan Kendatsub, public domain Evil Sendan Kendatsub

Do reflective people agree about ethics?

You might think that most people will share some big-picture beliefs about morality (a lacommon morality“). And you might think that this agreement is the result of reflective reasoning about ethics. For example, most people might think about ethics for awhile and accept a consequentialist principle like this: we should try to achieve the greatest good for the greatest number. Well, it turns out that people don’t agree about such ethical principles — not even people who often reflect on such matters. Before I get to the evidence for that claim, take a look at someone who thought that reflective people do agree about ethics.

1.  Will Reflective People To Agree About Ethics?

Here’s Henry Sidgwick:

“The Utilitarian principle […that there is a] connexion between right action and happiness […] has always been to a large extent recognised by all reflective persons.” (The Methods of Ethics, Book I, Chapter 6, Section 3)

Sidgwick is claiming that…

  1. there is a connection between happiness and right conduct (and)
  2. all reflective people recognize this connection.

What do you think? Do these claims sound right?

2.  The Evidence

Notice that 2 requires evidence. Alas, 2 is not well-supported by evidence: reflective people do not seem to agree that there is an important ethical connection between happiness and right conduct.

Common Morality

Consider that there is widespread disagreement about 1 among philosophers. To quantify this disagreement a bit, let us look at some data. Of about 1000 philosophers surveyed in 2009, 25.9% of leaned toward or accepted deontology, 18.2% leaned toward or accepted virtue ethics, and 23.6% leaned toward or accepted consequentialism (Bourget and Chalmers 2013). Consequentialism is the view most associated with 1 — the idea that there is a connection between happiness and right conduct — and yet fewer than a quarter of philosophers are partial to it. So, contrary to Sidgwick’s claim, the consequentialist’s connection between happiness and right conduct does not seem to be recognized by all reflective people. Indeed, it does even seem to be recognized be even most reflective people.


In situations like this, an intuitionist like Sidgwick might want to press on the notion of ’reflective’. After all, the finding (above) is only a problem for Sidgwick if — among other things — philosophers count as ‘reflective.’ If they do, then Sidgwick’s hypothesis is falsified. If they do not, then Sidgwick’s hypothesis might still be intact.

So if you want to defend Sidgwick’s hypothesis 2 from the evidence (above), then you need to argue that philosophers do not count as reflective — and do not thereby pose a counterexample to 2. One cannot, of course, merely stipulate that philosophers do not count as reflective. That would be ad hoc. In order to defend Sidgwick’s 2 from the aforementioned data, you will need to appeal to independent evidence. Fortunately there is independent evidence about the relative reflectiveness of philosophers and non-philosophers.

Alas, the evidence does not support Sidgwick’s hypothesis (2). Rather, the evidence suggests that philosophers are significantly more reflective than non-philosophers. In a sample of 4000 participants, those with training in philosophy performed up to three times better on tests of reflection — e.g., the Cognitive Reflection Test (Frederick 2005) — than those without such training (Livengood et al 2010). This result has been replicated and expanded. For example, those with (or a candidate for) a PhD in philosophy also performed significantly better than others — F(1, 558) = 15.41, p < 0.001, d = 0.32 (Byrd 2014). And these findings are not new. Over 20 years ago, Deanna Kuhn found that philosophers demonstrated “perfect” and domain-general reasoning competence (Kuhn 1991, 258-262).

So it seems that if any group of people should count as reflective, it is philosophers. And these reflective people do not — contrary to Sidgwick’s hypothesis 2 — unanimously recognize a connection between happiness and rightness.

3. So what now?

The idea that people share a “common morality” via “reflective equilibrium” might fly in the face of evidence. It certainly does for Sidgwick. After all, it seems like reflective people (e.g., philosophers) simply don’t agree about the alleged connection between happiness and right conduct. And if you try to respond to this evidence by denying that philosophers are reflective, then you run into another problem: that claim also flies in the face of evidence. So those objections won’t work.

A better strategy might be to reject my claims about the association between Sidgwick’s claims and consequentialism. That is, you might say that non-consequentialist approaches to ethics acknowledge the connection between happiness and right conduct just as much as consequentialist approaches — sort of like Andy Hallman does in the comments. If that claim is right, then Sidgwick might have been on to something. I leave it to you to decide if that kind of objection is promising.



Featured image: “Extermination of Evil Sendan Kendatsuba” via Wikipedia Commons (in the public domain).

Published by

Nick Byrd

Nick is a cognitive scientist at Florida State University studying reasoning, wellbeing, and willpower. Check out his blog at byrdnick.com/blog

8 thoughts on “Do reflective people agree about ethics?”

    1. One possible answer would be to say that what matters more is convergence on object-level moral judgements (Is abortion wrong? Is homosexuality wrong? etc). I think research shows Reason (broadly understood as IQ, CRT, and analytical thinking) supports correlations between that and values we would regard as consistent with broad liberal (not in the American sense) view. Michael Huemer argued (In one of Robin Hanson’s posts) that one prediction of his views would be precisely those correlations.

      1. Hi Artir,

        Interesting claim/prediction. Do you have citations in mind for evidence of the convergence hypothesis?

        Thanks for engaging! I wish you well!

        1. For analytical thinking, we have Pennycook, Fugelsang & Koehler’s “Everyday consequences of analytic thinking”:
          -Analytical thinkers are able to reject Haidtian-type disgust based moral scenarios (incest, dead chicken intercourse), assesing that there is no negative consequences done.
          -Negative correlation between CRT and binding moral foundations in Haidt’s framework.
          -Positive correlation between CRT and individualizing moral values
          -More likely to be mild altruists in economic games

          One such study is Pennycook, Cheyne, Barr, et al (2014) ‘The role of analytic thinking in moral judgements and values’

          Haidt, Talhelm, et al (2015) Liberals think more analytically than conservatives. (Holds also in China)

          For IQ

          *Woodke (2016) “Are smart people less racist?” argues that intelligence is correlated with less racism (but not pro-equality policies)

          *Then Carl (2015) “Cognitive ability and political belief in the United States” (See also his debate with Solon for more nuances) and Carl 2014 ‘Verbal intelligence is correlated with socially and economically liberal beliefs’

          *Onraet, van Hiel et al (2015) ‘The association of cognitive ability with Right-wing ideological attitudes and prejudice: a meta-analytic view’
          Reviewed in my blog here https://artir.wordpress.com/2015/10/16/are-conservatives-less-intelligent/

          *Mollerstrom & Seim (2014) show that cognitive ability is related to increased altruism, and lower preference for redistribution

          *Rindermann et al. (2012), Pan & Xu (2015), Kemmelmeier (2008) replicate Carl’s findings.

          *Deary, Batty & Gale (2008) ‘Bright Children become Enlightened adults’ show also asociation with social liberalism

          From Bekker & Wiepking (2007) ‘Generosity and philantrophy: A literature review’, higher IQ and higher abstract thought predict higher charitable giving. Same result in Bekker & Wiepking (2011) ‘Who Gives? A Literature Review of Predictors of Charitable Giving’

          Then there are other series of things like less crime, less corruption and such that are described in Garett Jones’ book Hive Mind, but those could be explained by enlightened self-interest, so would not count as moral proper.

          All of this seems to support that something that looks like liberalism is the ideology picked by those higher in ‘Reason’.

          This is compatible with saying that an emotivist account of morality, from a descriptive point of view is valuable. For most people, moral judgment will involve a degree of emotionality. Correlations between ‘Reason’ and moral views are not r=1, but Huemer’s model only requires a limited correlation to induce moral progress over time.

          Currently, there is no model in moral psychology to explain these correlations. Most other correlates of morality (pathogenic load, GDP, ‘fast life-stories’, ) are also correlated with IQ. Kanazawa tried (intelligent individuals prefer novelty in general, and so prefer non-adaptative moral values) but got heavily criticised by Dutton (2013).

          1. So after giving it some thought, I would say the following.

            First, the following disclaimer is key: “Correlations between ‘Reason’ and moral views are not r=1, but Huemer’s model only requires a limited correlation to induce moral progress over time.” Sidgwick (and his ilk) do not seem to have a view that is so well-supported by these data. Huemer seems to have a view that can receive more support from these data.

            Second, it would be interesting for philosophers and scientists to address the following question: what is the nature of the connection between reasoning competence(s) and liberalism? After all, it might be true that improved reasoning competence is associated with being (socially) liberal, but this does not tell us whether reasoning competence should or must lead to liberalism (according to some rational standard(s) and/or according to some causal story). And it doesn’t rule out latent variables that explain the connection. So perhaps more reflective people are more liberal, but maybe they need not be. The connection might be contingent on something hitherto not carefully understood. (E.g., I imagine that some would want to point to the dominance of liberalism in higher ed. and amongst the authors of the studies being cited). This is just to echo your point about there not being an accepted theory of this connection, of course.

            Aside: I think you win the prize for most well-cited comment on the blog in 2016…maybe ever.


  1. Hi Nick. Nice post.

    I have a question.

    You seem to interpret belief in deontology with a disbelief that there is a connection between happiness and right action. Is that true? Why can’t deontologists believe there is a connection between happiness and right action, but that there are other things that matter, too?

    I think it would definitely be wrong to say all reflective people are utilitarians, since the evidence doesn’t support that, but Sidgwick appears to be making a weaker claim that that.

    1. Hi Andy,

      Good point. Mark Lebar also raised this point via email.

      I didn’t mean to say anything which entailed that only consequentialists accept the connection (or that only consequentialists are reflective). But as I reread “…not even most reflective people…” I see that I implied this.

      My response: if it can be demonstrated that “all” reflective people (deontologists, virtue ethicists, “other”-ists) accept the connection, then I will rescind my point.

      But if the thought is just “Look, it is conceptually possible for a deontologist or a virtue ethicist to accept the connection…” or “Actually, a handful of published deontologists and virtue ethicists have said things which imply that they accept the connection…” then I will hold my ground. After all, Sidgwick’s claim entails that the connection is, in fact, accepted by “all reflective persons.” So neither conceptual/logical possibility nor textual/historical precedent will vindicate the claim.

      But maybe you want to say more now that I have laid more cards on the table.

      Thanks for engaging! And nice blog! I look forward to reading more of it.

Comments are closed.