the sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any of their number is self-protection.
Or, in other words:
That the only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others. (Mill, On Liberty, Chapter 1)
Good quote elaborating on the Harm Principle: Mill writes:
His own good, either physical or moral, is not sufficient warrant. He cannot rightfully be compelled to do or forbear because it will be better for him to do so, because it will make him happier, because, in the opinion of others, to do so would be wise or even right. These are good reasons for remonstrating with him, or reasoning with him, or persuading him, or entreating him, but not for compelling him or visiting him with any evil in case he do otherwise. To justify that, the conduct from which it is desired to deter him must be calculated to produce evil to someone else. The only part of the conduct of anyone for which he is amenable to society is that which concerns others. In the part which merely concerns himself, his independence is, of right, absolute. Over himself, over his own body and mind, the individual is sovereign. (Mill, On Liberty, Chapter 1)
Implications of the Harm Principle: This principle is the primary source of Mill's theory of rights. That is to say, according to Mill, the reason that we have the rights that we do is because they are derived from the harm principle, which is justified by utilitarianism. It turns out that according to Mill we have a right to do anything at all that does not involve harm to others. This winds up allowing us to have a vast amount of negative rights, which tends to limit the amount of positive rights that can be granted.
What rights we should have: If you give people strong negative rights that usually includes the right not to be deprived of their money. Positive rights require the mandatory allocation of resources (usually money) to some people, and this money must come from somewhere (i.e., it must come from other citizens, usually in the form of taxes). But if those citizens have the negative right to their money, then you cannot take their money away to ensure other people their positive rights, so you cannot grant those positive rights in the first place. Of course, the government will be allowed to take money from citizen to provide the basic services of society such as civil protection, national defense, and other essential services - and the government will be allowed to compel people to do other jobs necessary in civil society (e.g., jury duty, military service). This is justified, not because there was any tacit consent by the citizens to give up their negative rights, but instead simply because such sacrifices are necessary for the very existence of civil society. Remember, according to Mill's theory of rights, people don't have any rights aside from those granted by society, so people never give up their right to the money taken in taxes, they never had the right to it in the first place.
Postponing two questions: There are two questions that I am going to explicitly put to the side temporarily. They are: 1) What does it mean to say that something involves a harm to others? (i.e., how much do I have to harm them for it to count?) [See part 7 below]; and, 2) Why shouldn't we allow government to keep people from harming themselves? [See part 6 below] I will deal with these issue below.
4) Exceptions to the Harm Principle: Mill does have some exceptions to the harm principle which help to make it a reasonable and plausible principle. It is clear that Mill would allow for all of these exceptions, and his theory should be understood as including them. I state them here for two reasons, so that you will have a greater understanding of Mill's theory, and so that no one can try to object to Mill by raising a counter-example that falls under one of these exceptions.
a) Children and the mentally underdeveloped: The harm principle is not intended to apply to children (the ages included under the term "child" are not clearly defined, nor are they of particular interest philosophically or in debate) nor does the principle apply to others who require substantial care (e.g., the elderly, the mentally ill, or anyone whom we have objective evidence to believe is incapable of rational judgment). Children are the primary case, and Mill maintains that we are justified in compelling children in paternalistic ways.
b) Necessary social duties: A second exception is that a government may compel its citizens to perform certain tasks that are essential to the preservation of itself as an ordered society. For example, the government must be allowed to draft people for the common defense of the nation, and must be allowed to compel people to testify in court - since both of these are essential to social well-being.
c) Intervention in emergencies where the agent is ignorant: The third exception is that an individual may forcibly intervene and prevent some action if it is evident that the agent does not know the consequences of her action and there is not time to inform her of those consequences and allow her to make a free choice. For example, if I am talking to a friend while waiting for an elevator and then I begin to walk when the doors open, even though there is no elevator in the shaft, then you should feel free to use physical force to prevent my walking into the shaft and falling to my death. I suppose that if there were time it would be best for you to simply ask me if I wanted to walk into the empty shaft and if I knew that this would surely cause my death, this would allow me to make my own decision and to choose to continue if I really wanted to.
d) Consenting adults: Lastly, if two (or more) people are engaging in an activity to which they both consent, then the activity should not be considered as involving harm to others even if the people are causing harm to each other. For example, if you and I agree to a boxing match, then the government should not forbid it (assuming that we are not harming anyone who is not consenting), even though we are doing harm to each other.
5) Freedom of Speech, Press, Expression, and Thought: Mill has an entire chapter in On Liberty on this topic, which is excellent. Time and energy prohibit me from providing a synopsis at this time. I will, however, provide the following brilliant quote from the second chapter of On Liberty.
If all mankind minus one were of one opinion, mankind would be no more justified in silencing that one person than he, if he had the power, would be in silencing mankind.
6) Against paternalism
a) The Basic Argument: It is now time to deal with the question of why a government should not be paternalistic. Mill argues that such legislation has the strong potential to be damaging to the society by stifling individuality and prohibiting people from having control over their own lives. Paternalistic laws will also tend to inhibit the evolution of society and social customs by mandating that people follow the currently established norms. It is certainly the case that paternalistic legislation inhibits individuality. Mill argued that individuality is of crucial importance to the health of a democratic society because without it there develops stagnation and complacency, which discourage people from becoming educated and politically involved. Paternalism also prevents the emergence of the truth concerning both scientific and social questions. A government might choose to compel all of its citizens to go to church on Sunday (or even to practice a particular religion) because they believe that it is better for those people if they do so, but such a policy is oppressive as well as deterring individuality. Such a law would certainly have been possible in the past, but society has evolved to the point where we would no longer accept such a law. This is a progression in society. Maybe a better example would be the historical government restriction on scientific inquiry that went against a state endorsed religion, this clearly had detrimental effects on social and scientific development. A contemporary example would be our laws against homosexual relationships and restrictions on homosexual marriages. Such laws are clearly paternalistic, and they inhibit the development of alternative ways of living which may turn out to be beneficial to the individuals and thus to the society. Suggested laws against smoking in private (where others aren't harmed) are another modern example of paternalistic laws.
b) Refinements for more difficult cases: These arguments may seem to have little to do with something like seat belt laws, and you might ask what is wrong with those kinds of laws. There are two reasons that these kinds of laws should not be passed according to Mill's theory. The first reason is simply that the government has no particular interest in what I do with my own life and my own possessions, certainly no interest that compares to the overwhelming interest that I have in my own life, health, and property. I also have far more information about my own situation and the peculiar circumstances that I am in at any given time. So, it would seem to be reasonable that the government would trust my judgment about what it is best for me to do in matters which involve only my own interests. (The government might reasonably try to convince me that what I am doing is a mistake by means of advertising and education, but this is very different from government compulsion.) So, I should be trusted to judge when it is reasonable to wear a seat belt because I am the one in a position to best make that judgment. The other response that Mill can make is that the government cannot be trusted to tell which paternalistic laws are innocuous (like seat belt laws may be) and which are harmful, so they should not be allowed to make any laws that are paternalistic. This argument is quite important. The point is that politicians are very bad at being objective about their motives and also not good at understanding or caring about the effect of legislation on the minority. Thus, they cannot be trusted to be able to determine which paternalistic laws are fair to minorities. Even if they could tell, they could not be trusted to refrain from passing oppressive laws. An analogy here is this: you can imagine that there is a surgical procedure that will have some minor beneficial effects, but that there are some people who have a very serious negative reaction to this procedure and it is not possible to tell who those people are (or even what percentage of the population will have that reaction). In such a case, it would be a bad idea to risk this procedure - and it would surely be wrong to impose such a risk on someone without their consent (which is analogous because the government does not ask individuals' consent when it passes laws). Clarifying the analogy: Passing a single paternalistic law is analogous to imposing this surgery on a single person - because each of these things might have some good effects but each also might have some very bad effects, in each case we simply cannot tell which will happen. Thus, in neither case should we take the risk. [See part 9,b below for further explanation]
7) What constitutes harm?
a) Basic answer: This is a very tricky question, and there have been very large books published which try to answer just this question. I'll do my best in a paragraph or so. Harm certainly includes most any form of physical harm (e.g., you punching me out, or you smoking near to me). It also would include most forms of financial harm (e.g., you taking my car, or you breaking my watch), but there are certain financial harms that will certainly not be included (e.g., me moving in next to you and lowering your property values because of my race, or me influencing people not to do business with you when I tell them that you overcharged me). The latter type of cases do harm you financially, but the real source of the financial harm is not me, it is other people's racism and your own poor businesses practices respectively. There are also mental harms which are included (e.g., you threatening me, or you excessively harassing me), but the standard is very strict in this area and the presumption is that a mental harm does not constitute a real instance of harm to others (e.g., I am offended by your joke, or I am disgusted by your style of dress). There are going to be a lot of hard cases concerning this (e.g., your offense at the nudity of me on the beach, or your financial harm when I move in next door and cut down all the trees on my property and use my back yard as a smelly compost heap). It is not clear whether these cases fall under Mill's theory as involving harm to others or not.
b) Harm and Political Speech: One area deserves particular note, the area of political speech. Some political demonstrations and speech can cause social unrest. The best example of this is when the Nazi's wanted to march in the predominately Jewish town of Skokie, Illinois where many holocaust survivors live. Such a demonstration would certainly cause extreme mental harm to others, and would likely cause a riot which would cause severe physical harm. Mill wants to defend freedom of expression and speech, and this type of speech would certainly be protected. It can be difficult to determine the difference between political speech which will cause a riot and rabble rousing which will incite a riot (which Mill does not think should be allowed). There is also the more modern legal classification of some speech as hate speech, and Mill would probably not support the protection of this kind of speech - but I am not at all clear on what the criteria for hate speech is. The point is that speech can harm others, but it is also strongly protected by Mill, so such cases are very difficult. [See On Liberty, chapter 2 for a detailed discussion.]
8) Mill's refinement of the assignable obligation: Some philosophers claim that Mill abandons the simple harm-to-others principle in the latter part of On Liberty in favor of the new principle of the assignable obligation. This new principle is stated in this way by Elizabeth Rapaport:
A person ought to be subject to social coercion only to prevent a violation of a "distinct and assignable obligation to any other person or persons." (From the editor's introduction to Mill's On Liberty, quotation marks surround Mill's words)
a) What is a distinct and assignable obligation? A distinct and assignable obligation is where there is someone who has either a right, or a legitimate claim or expectation, which the obligated person is bound to honor. These obligations can be from a promise or contract, a social position (e.g., spouse, parent, employee, citizen), or possibly some other source. It is important to note that not all obligations are distinct and assignable. For example, I may have a moral obligation to give to charity on occasion, but there are no specific charities that can claim that I have a distinct and assignable obligation to give them money. One has a distinct and assignable obligation only if someone else has a right that you must fulfill (even a negative right). For example, you have a negative right not to be hit by anyone, so I have a distinct and assignable obligation not to hit you.
b) How this affects Mill's Theory: This revision does not radically change Mill's theory, it merely changes the focus from the vague idea of a harm to others, to the supposedly more precise idea of a distinct and assignable obligation. (Personally, I prefer the harm-to-others formulation, even though it is in need of extensive clarification with respect to what constitutes a harm.)
9) Connection between rights and utility: An understanding of utilitarianism is important to understand this section. I could write for pages about this, but I'm going to try not to.
a) Basic issue: There seems a first look to be a fundamental incompatibility between Mill's political theory of rights expressed in On Liberty and Mill's moral theory expressed in Utilitarianism. In On Liberty, he claims that government should never interfere with an individual, except to prevent harm to others (i.e., never be paternalistic). But in Utilitarianism he claims that the right thing to do is whatever will maximize utility. He also admits (what would be absurd to deny), that there are some possible laws which are paternalistic, but which do actually maximize utility. These seem totally incompatible, but they are not. Take a few minute to make sure that you see the apparent incompatibility before you read on. I'm totally serious, stop reading!
b) Mill's consistent answer: Mill recognizes that it is practically impossible to calculate utilities of citizens, so government cannot make decisions by directly applying the principles of utilitarianism (just as individuals cannot do so). Thus, government must use some other principles to rule well. They need to use principles that can be followed easily and cannot be abused (since governments have a tendency to abuse their power), and ones that will come as close as possible to maximizing utility. The harm principle is fairly easily followed, and is very difficult to abuse, so long as exceptions are not permitted to the principle. Almost all of the laws which will contribute to maximum utility will be allowed by the harm principle. The laws which might contribute to utility, but are not allowed by the harm principle, could only be passed if a government allowed exceptions to the harm principle. But since we are unable to consistently judge which paternalistic laws are harmful and which are beneficial, more harm would be done by allowing for exceptions than would be done by not allowing for any exceptions at all (including those few paternalistic laws that would contribute to utility). [See the medical analogy at the end of part G,6,b]
c) Summary: Essentially, the theories are compatible because Mill claims that the best way for a government of imperfect people to maximize utility is to have a strict policy of respecting individual rights by not passing any law that violates the harm principle, even when it seems to be a good and safe idea to do so. Mill claims that the reason that we ought to have a particular set of rights protected by our society is not because they were natural or god-given, but rather because those are the rights which, when protected, will lead to maximum utility. The ultimate justification of rights for Mill is his moral theory, not some fictional state of nature or imagined natural rights.
10) Objections and replies: Mill's theory on rights is very interesting, largely because it is able to answer many questions which prove embarrassing for other theories (e.g., where do rights come from, how do we know that these are natural rights, what evidence is there that you have such a right).
a) Effective against Social Contract: Social contract theories seem to be unable to avoid a tyranny of the majority (Rousseau also fails in this because he cannot make clear what the general will is or how to discover it). For Mill there are no natural rights, a right is societal and is embodied in a law - for social contract theorists natural rights are essential, but they are peculiar sorts of entities that are not ever embodied or clearly seen. Claiming to have a natural right is sort of like claiming to have an invisible friend that cannot be detected in any way, not even you have ever seen her; it sounds kind of dubious to base a theory on something we have no evidence about.
b) Supported by the Bill of Rights: It is also worth mentioning that the Bill of Rights was put in place in large part because of the same considerations that motivate Mill's theory. The Bill of Rights is supposed to prevent government from infringing on the rights of minorities, and to provide absolute boundaries beyond which government is not allowed to legislate. These boundaries are designed to prevent the tyranny of the majority, and to allow for freedom of expression and open debate.
c) Arguing against Mill: Mill can be criticized because he does not allow for what seem to be very reasonable and helpful paternalistic laws. For example, thousands of lives have been saved by seat belt laws, and no one has seemingly been harmed. Any theory that would prevent this must be seriously flawed. Mill can also be criticized on the grounds that it is not possible to determine what actually constitutes harm. Consider some of the hard cases that I mentioned above to bring this out, and think up some more of your own. It may be that the harm principle sounds good, but that it cannot be practically applied because it is far too vague concerning what exactly counts as harm.