In 2014, it was revealed that Facebook had manipulated news feeds to study emotions. Facebook did not go through any sort of ethical review at the time and has been criticized for this and failing to get informed consent. See for instance the Guardian's article.


Assuming a utilitarian framework focused on harm, would the results incurred by preventing research from happening outstrip the results caused by not requiring informed consent? On what grounds could utilitarians justify consent if the benefits of the research outweigh the harms inflicted by the lack of consent?

  • 2
    There's quite a few interesting things about the question, but they're too different to be easily answerable here. Off the top of my head: (1) Who are the "many people" making this claim? (2) shouldn't we ask whether it was wrong for Facebook to do so w/o informed consent before we ask what review boards would say about it? (3) how do we define harm such in a way that quantifies "preventing research from happening"? Many of these are opinion-based land mines that need to be resolved before there's something we can really answer for you.
    – virmaior
    Commented Apr 27, 2016 at 13:24
  • Or maybe to word that another, there are interesting questions about utility and research; there are interesting questions about utilitarianism and review boards; there are interesting questions about utilitarianism and informed consent. It's hard to compose a good SE question that simultaneously raises questions about all of the above. Can you instead fill in the above within your question?
    – virmaior
    Commented Apr 27, 2016 at 13:25
  • @virmaior : I edited the question to focus on ethical review boards requiring informed consent.
    – Christian
    Commented Apr 27, 2016 at 13:55
  • 2
    I've tried to edit your question to something answerable. I don't know if you didn't read my comment above, but your edit (adding 10 words) did not really address the issues there. You needed less moving pieces so I've made the focus of this question informed consent vs. research results. Review boards add an extra layer if it can be shown that there's a way to quantify gains from research against loss of consent but first that should be resolved.
    – virmaior
    Commented Apr 27, 2016 at 14:07
  • The italicized bit above was my thought in considering this. I do not think that gains from research can ever use the same "currency" as peoples' right to self-governance. It is simply "comparing" two things that cannot be balanced against each other. What possible level of benefit could inspire you to volunteer a family member to an experiment without informing them? Further, we do not expect an entertainment and communication medium like Facebook to conduct any sort of research or experiments. Sorry, they are not in that business. This has nothing to do with Utilitarianism as best I can see.
    – user16869
    Commented Apr 28, 2016 at 0:50

3 Answers 3


There's several issues that make this question complex (some of which make it more interesting and other's less so).

To simplify, I'm going to basically ignore the Facebook example. As "no comprende" notes in the comments, they aren't really a research entity per se, and as such, most of their users don't expect them to be doing research of any kind let alone groundbreaking studies. (In fact, I doubt that they intended to make "scientific" discoveries at all).

Further to simplify, the specific issue of informed consent is only going to come into play in a limited way. For this to even get off the ground, informed consent must be something that is either what we are trying to maximize or minimize or something that we can equate to it (this is no small feat).

Limiting ourselves in the above ways, the problem becomes a series of variations on: can a Utilitarian framework ever make it such that the benefits of some given research (benefits calculated in whatever units of measure we are using in our utilitarianism) outweigh the costs of the forms of harm involved in achieving that research (e.g. the loss we associate with the lack of informed consent)?

To answer this, we first need to understand the nature of Utility. For Mill, Utility is measured in pleasure. Thus, the question is about what maximizes pleasure. But in Utilitarianism Book 2 Paragraph 5, Mill makes an important, interesting, and controversial qualification. He explains that there's a distinction between higher and lower pleasures. And that humans should pursue the higher pleasures. Thus, his calculus is that we should maximize pleasures of this sort (I just read something about this in Robert Adams, Finite and Infinite Goods (OUP: 1999) near page 200). ( to some extent this raises a second issue "no comprende" points to in a comment, if we're quantifying how can we have upper/lower distinctions.)

Even Mill's account is more complicated than what he says in Utilitarianism as he brings up an idea in On Liberty that we cannot in the process of maximizing pleasure cause harm. If we've got both utility and a harm principle we cannot say it's better do something that causes harm no matter how much utility it brings about.

Before we give in despair, we should realize however that most contemporary "Utilitarians" do not follow Mill in making the goal the maximization of pleasure. First off, there's the problem with defining pleasure. Second, there's other concerns that they want to incorporate that pursue many to either want to (a) minimize suffering or (b) maximize justice or (c) maximize freedom or (d) something more complicated. These contemporary utilitarians are often called "consequentialists."

For some of these consequentialists, there is a single thing we are trying to maximize. In that case, it's partially a question of quantifying how much the research gains produce this vs. how much is lost in the loss of freedom. (Obviously, if freedom maximization is the goal, then research loses unless it could make up for that loss).

But we're not out of the woods yet. A further complication is that utilitarians differ in terms of how we go about this maximization. There's three major groups that I'm aware of "act utilitarians", "rule utilitarians" and "preference utilitarians". The difference being act utilitarians thinks each situation should be evaluated to max/min the desired good. Rule believe that we can establish rules that will lead in general to maximization. Preference U. believe that we should cause people to have preferences that lead to maximal outcomes.

It seems much harder to prove either preference or rule utilitarians could allow this regardless of what they want to maximize.

To summarize and return to "informed consent," there's no consensus that immediately follows from "Utilitarianism" and its relation to informed consent and to research. If anything, there's several problems mired together:

  1. what are we trying to maximize/minimize?
  2. can we calculate the benefits of research in accordance with (1)?
  3. can we calculate the harm of lack of informed consent in terms of (1)?
  4. Does our utilitarianism permit us to make "harm"-causing actions (vis-a-vis (1)) or are these always illicit?
  5. does the species of utilitarianism we are using allow us to make particular bets or require us to consider principles of optimization? (i.e. are we act utilitarians? are we able to calculate on potential gains against real harms)
  6. Another ambiguity in Utilitarian accounts is whether the maximization is of (a) actual consequences, (b) reasonably expected consequences, or (c) intentions. This is going to matter a lot for how we can qualify the benefits in (2) against our criteria (1).

Maybe, it's easier just to describe when a Utilitarian could easily say doing the research is okay or not.

Example #1 - It's fine utilitarianism. If you're trying to maximize pleasure and don't value consent at all, then you are in the clear as long as the research is likely to increase pleasure for the most in the long run.

Example #2 - It's never going to be fine utilitarianism. If you're trying to maximize human freedom and have a harm principle about violations, then even if the research would lead to gains in human freedom, then you cannot pursue it because your species of utilitarianism does not permit this form of harm.

For most other cases, we're back to the questions I've outlined above.

  • Given that your answer doesn't talk about the principle of informed consent, I don't see how it answers the question.
    – Christian
    Commented Apr 28, 2016 at 12:43
  • I've amended the answer to make it clearer. I'll admit I didn't use the phrase "informed consent" but I'm not seeing how the answer was not on the more obvious rather than less obvious side of the bar in terms of how it does address it.
    – virmaior
    Commented Apr 28, 2016 at 13:01

On what grounds could utilitarians justify consent if the benefits of the research outweigh the harms inflicted by the lack of consent?

This is a question of whether the Ends justify the Means. The value of taking a Utilitarian position is that you focus on the Ends, which you calculate as having a great value. However, in cases where the Means inflicts harm, how do you measure the short term impact versus the long term benefits? It depends on the priority and arrangement of social and political norms, as well as an individual's personal ethical or practical dispositions.

Utilitarianism can require consent, either explicitly or tacitly, as long as that is a factor in quantifying the "utiles" you are comparing. Utilitarianism has many variations, but one could reasonably consider a society that upheld explicit consent as one of its highest tenets. In this world, a key factor of maximization would be to ensure consent is given before any action.

i.e. Based on a strictly utilitarian understanding; If we were to say that no doctor may treat a patient without explicit consent. Then it is a utilitarian imperative that the doctor cease treatment, since consent is tantamount to any treatment.

However, an alternate society could put less weight in individual consent and prefer advancement or obedience. In these societies, consent is secondary, if even considered.

i.e. Based on strict utilitarian understanding; Obedience in the military and respect of the chain of command exceeds any consent from lower ranking officers.

Without the specific methodology of the research done by Facebook, I cannot provide a precise answer regarding their ethical behavior. The key factors would be:

1. What was the purpose of the research? What was the value gained from the research? What was the certainty of the outcome? (Evaluate the Ends)

2. How was the research done? What, if any, impact did it have on the subjects involved? (Evaluate the Means)

For a more well discussed example of this conundrum, look into some critical analyses of the Francis Field Test (1954 - 1955), where the Polio Vaccine was first tested on the mass public in the United States.

Here is a link that covers the topic of utilitarianism well. (LINK)

Contributing Philosophers

  • John Stewart Mills (b. 1806 - d. 1873)

  • Jeremy Bentham (b. 1748 - d. 1832)

  • Henry Sidgwick (b. 1838 - d. 1900)

Anecdotal References

  • Jonas Salk (b. 1914 - d. 1995)

If you include autonomy and the feeling of safety as goods, then the distrust engendered by the risk of being manipulated is a harm, which you can simply include as one of the forms of harm. If one does not expect informed consent for planned mass manipulations then the peace of mind available to any individual in any interaction with any institution that handles and directs information decreases, and we are all harmed. The odds of us all being helped are lower.

Then again, you can basically create a Kantian version of Utilitarianism built around autonomy and empathy as principle goods, where accumulation has a very high rate of diminishing returns. The first few people whose autonomy is adequately infringed by a policy count a whole lot and affect us all almost as much as if it happened to everyone. And we are then bound by a numerical form of the categorical imperative.

Any input whatsoever can be included into the hypothetical computations required by different forms of Utilitarianism. This makes the label almost useless without more context. And since the computations themselves never approach tractability and measure something that basically cannot be defined, there cannot really be a limit on what consideration Utilitarian measures can be assumed to include.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .