On the Legitimacy and Authority of Governments

In Democratic Authority[1], David Estlund argues that democracy’s value is captured by his theory of epistemic proceduralism - that democracy derives its authority and legitimacy from fair procedures that have a tendency to produce the correct answer to questions in a way that may be considered objective. I believe Estlund’s argument to show that epistemic proceduralism has epistemic value is flawed – his reliance on primary bads is insufficient to show this value. A government’s performance in dealing with primary bads is not a reliable indicator of how it will deal with lesser issues. I propose a theory of government legitimacy based on values – a law is only legitimate if it embodies the shared values of the public.

Estlund argues that we need an epistemic procedural theory because the existing theories of democracy, which he characterizes as either wholly procedural or wholly epistemic (correctness theories), are each flawed. He believes a theory that combines the two approaches can best account for democracy’s authority and legitimacy by better answering difficult questions, such as that of moral deference i.e. why a citizen should obey laws that they disagree with morally.

Procedural theories hold that democracy’s value comes from the use of fair procedures. Estlund believes they do not produce a sufficient reason for citizens to obey laws. A coin flip is a fair procedure, but we wouldn’t expect people to obey laws that were determined so randomly. People have some expectation that their laws have been produced more thoughtfully than that.

Correctness theories hold that democracy has value because it produces the correct answer to the questions it faces. It expects the minority to accept the majority’s decision as correct. In questions of justice, Estlund views this as unacceptable because each individual has equal authority. There is not an objectively correct answer to those kinds of questions. If people have strong moral beliefs, we do not expect them to change their minds because more people disagree with them.

Estlund believes epistemic proceduralism overcomes this problem of deference. It does this by providing people a moral reason to obey laws even if they find those laws morally objectionable. This can be achieved by discovering the right kind of procedures, ones that are generally acceptable to all reasonable points of view and have some epistemic value. We would expect people to abide by decisions made this way, much as we abide by the decisions made by a jury. The outcome of a jury deliberation is generally accepted as legitimate, and it is on this sort of moral authority that Estlund’s argument rests.

Estlund notes that laws wouldn’t be allowed that weakened the democracy (such as disenfranchising minorities) or that violated certain rights (such as rights of expression that are necessary for democracy). We may also wish to protect certain rights, such as freedom of religion or preventing cruel and unusual punishment, even if those rights may not have obvious democratic value.

The question becomes how much epistemic value is needed to justify democracy’s authority and legitimacy. Since a non-democratic epistemic arrangement would privilege someone’s point of view over someone else’s, it would be generally unacceptable to those who are ignored. Therefore Estlund claims that a democratic arrangement with any epistemic value at all is better than any non-democratic arrangement, and he only needs to show that epistemic proceduralism’s value is better than random.

This is still a problem because, without privileging anyone’s moral code or sense of justice, it is difficult to judge the correctness of some democratic outcomes objectively. Estlund proposes adopting a formal framework that can be used to objectively judge epistemic procedural arrangements. He uses what he refers to as primary bads, situations that all reasonable points of view would agree should be avoided - such as war, famine, economic collapse, political collapse, epidemic, and genocide. He believes that if an arrangement would do much better than random in preventing these things, it can be interpolated that it would as least do better than random in other areas as well.

He proposes an arrangement modeled on a concept of deliberation - one that stresses equal access, equal opportunity to speak, honesty, equal consideration, equal power, the ability to recognize reason, the representation of all parties involved, and a devil’s advocate. Estlund believes that deliberation conducted under such conditions will generally answer questions correctly in an objective fashion. He also emphasizes that this model should not be the actual goal of a real world democracy. The intent is to show that actual deliberation can have epistemic value, even if it deviates significantly from the model. Given that Estlund merely needs to show that epistemic proceduralism has any epistemic value at all, it is not a stretch to imagine a world in which a deliberative democracy produced some epistemic value.

Estlund’s theory seems right. If we accept, as he does, that any political arrangement needs to be generally acceptable to all reasonable points of view, it seems clear that, of the currently theorized systems of government, democracy would be the most generally acceptable. It also seems unquestionable that people would prefer their laws to be crafted with at least an attempt of capturing the truth, but not so beholden to the outcome that they are expected to disregard any moral qualms about those laws that they may have.

Unfortunately, Estlund’s argument is insufficient to show that epistemic proceduralism has the epistemic value he claims. His theory relies on showing that a government that does very well preventing primary bads would do better than random in other areas as well. In extrapolating from how a government does in preventing primary bads to how it would do on other matters, I believe he is relying on a false comparison. I plan to show that even if a government may do well preventing primary bads, its procedures may have little epistemic value in other areas.

Let’s re-examine the primary bads - war, famine, economic collapse, political collapse, epidemic, and genocide. Estlund admits this list is not exhaustive but merely representative of the sort of major issues a government may face. He also admits that there are some exceptions. For instance, some wars are justified. Economic or political collapse may be acceptable if the continuation of the current economic or political systems would be worse. But with the possible exception of war (depending on how justifiable one believes certain wars to be), these are things that don’t occur in modern Western democracies like the United States. Therefore I would argue that there are already governments, including the United States, that do much better than random at preventing primary bads.

According to Estlund’s theory, these governments should also do better than random in other areas as well. However, this does not seem to be the case in the United States. Democracy and Ignorance[2] by Ilya Somin shows empirically that the American public is largely ignorant of political matters, without enough knowledge to effectively cast a vote that is consistent with their intentions. If the public does not possess enough knowledge to understand what effect their vote will have, that vote is essentially random. The current system could be considered to have little to no epistemic value, even less than Estlund’s theory would require.

This reveals the flaw in Estlund’s argument. He assumes that a government that does well preventing primary bads would do well in other areas. Yet the current system does well at preventing primary bads, but performs randomly in other areas. Therefore Estlund’s assumption is wrong, and his argument fails to provide sufficient evidence of epistemic value.

One might object that the current system doesn’t prevent all primary bads as well as one that had epistemic value. As I have mentioned, there are still wars that should probably be avoided. And as the list of primary bads isn’t exhaustive, we might be able to come up with other bads that are nearly as serious that the current system also doesn’t do well preventing. These are fair points so I will not rest my argument solely on the empirical record.

I think I can show why Estlund’s assumption is invalid. He essentially compares primary bads with other issues a government handles. However, I believe this is a false comparison because primary bads are fundamentally different from other issues.

First, the goal of the government – avoiding the primary bad – would be well established. The debate would only be over the procedures used to attain that goal. People will propose various solutions, one of which may be better than the others, but many of the options may be better than doing nothing. A random choice, even if it is “wrong”, might still help solve the problem. There may be epistemic value in the process of creating the options – one would assume they were produced by human reason and so far likelier to produce a better outcome than an option that was truly random – but it is important to note that Estlund’s theory maintains that epistemic value is derived from choosing the best option from those provided. So any epistemic value generated by the option producing process itself cannot be used to justify his theory.

Second, the public can judge the performance of the government on primary bads fairly easily. There are simple, objective standards that can be used to measure these problems. It is easy to see if large numbers of people are being sent to war, starving, catching a disease, or being systematically murdered. Popularly elected government officials know that they would have to address these issues to be re-elected.

If we consider other issues a government faces, we can see how different they are from primary bads. First, there is large disagreement about what the goals of the government should be. Should a nation’s gun policy protect the public’s right to bear arms or control guns to protect the public? Should tax policy maximize tax revenue or burden the public as little as possible? Without a fixed goal, there is no objective standard to measure a government’s performance in these areas. Also, the public does not pay attention to these issues as carefully (as Somin has showed) so they would not be able to judge the government’s performance even if standards were available.

So when we compare primary bads to other political issues, we see some differences. In the case of primary bads, the government’s goal is apparent. Most solutions derived by thinking people would probably be better than doing nothing. Therefore choosing one at random may still be effective. This would account for modern democracies’ ability to prevent primary bads. On the other hand, we have issues where the goal is not apparent. Presumably some goals would be better and some would be worse. The goal may be determined randomly, and the policy may be determined randomly. This extra layer of randomness and the ability to make things worse would account for the lack of epistemic value in current democratic systems. So Estlund’s assumption that preventing primary bads would be an indicator of epistemic value is incorrect, and the parts of his argument that rely upon it fall apart.

However, I have to agree with Estlund that a government that reliably avoids primary bads has value. If a government has the power to prevent these things, I assume people would view it as at least somewhat legitimate. Surely it is better to submit to an authority that can prevent primary bads than to live in a system where those things cannot be prevented. However, as I have shown, that legitimacy might not come from the epistemic value of the government’s procedures. So I believe we must search for the source of government authority and legitimacy elsewhere.

I propose that the authority and legitimacy of a government are derived from the shared values of the public. I define a shared value as one that is generally acceptable to the public. For the purposes of tractability, I will be talking about these values as broadly defined concepts (such as security, liberty, equality, moral or religious values, and epistemic values), but clearly each of those concepts represents many separate values. A political outcome is legitimate if and only if it embodies these shared values. It is deference to their own values that provides citizens reason to obey the laws of these systems.

I am not attempting to privilege any set of values, so I will be looking at authority and legitimacy through an objective lens. This will require a different view of these concepts to account for a more general theory. I take it that if a government or political outcome has legitimacy, it also has authority. It is difficult to imagine a circumstance in which an outcome had legitimacy but not authority. So I will limit my discussion to legitimacy, with the assumption that legitimacy leads to authority. The common definition of legitimacy is the popular acceptance of authority. Therefore, in an objective sense, an outcome is legitimate if the people accept the authority of that outcome, whether or not they have any normative justification for doing so. While most normative theories would seek to explain why the public should view certain outcomes as legitimate or not, I will seek to explain why they do view those outcomes at legitimate or not. I believe the public judges the legitimacy of political outcomes according to their values, and therefore there is no way to say that they are right or wrong without unfairly privileging certain values over others.

My argument runs roughly as follows. First, that people have different values and prioritize them differently, but they may still share values. Second, political systems can also be considered to have values and a priority of values, and they can roughly be defined by the values embodied by their outcomes. Third, people judge political outcomes according to their own value system. Fourth, political outcomes are legitimate when the values they embody match the shared values of the public. And fifth, such a conception of legitimacy overcomes the problem of moral deference because the values the public defers to are already their own.

I take it for granted that different people hold different values and prioritize or weigh those values differently. For instance, people have a wide range of religious views, many of which affect the values they hold. However, some values may be generally acceptable. For example, nearly everyone values their security (for the purposes of this discussion, protection of their body or property from harm). This is what I mean by a shared value. Other shared values may be liberty (the protection of rights and freedom), equality (that liberty should be given equally to all citizens), or epistemic values (such as an objective or theological pursuit of knowledge). However, even if people share values, they may prioritize them differently. One person may value their security more than their liberty. Another person may value equality more than the objective pursuit of knowledge. Such priorities are not absolute. I presume that there is some point at which the loss of one value outweighs the gain of another value, even if someone values the latter more highly. For instance, people who value their security more highly than their liberty would likely not agree to be locked in a bunker indefinitely just to protect themselves from the risk of a terrorist attack. Their slight gain in security would be outweighed by their massive loss of liberty.

I believe that systems of government can be viewed the same way, and furthermore that they can be defined by the values they embody. For instance, democracy is a system of government where every qualified citizen has an equal vote on political matters. This concept of universal suffrage embodies the values of equality and liberty. All citizens are allowed to vote (equality), and they are allowed to vote as they wish (liberty). These values define democracy because, without these values, democracy is not possible. However this also implies that those values are more important than any others. A democracy can still have other shared values, but only insofar as they do not threaten liberty and equality, and as such they would not be defining values. A democracy with freedom of religion but a separation of church and state would still be a democracy. If, on the other hand, a government held religious values as more important than liberty and equality, those religious values would be its defining values and we would call that government a theocracy, even if it had some democratic elements. Therefore I think that a government and its outcomes can be defined by the values they embody and the priority of those values.

Now consider what happens when an individual judges the legitimacy of a single law. In the case that an individual’s values and the values embodied by the law align perfectly, there is obviously no question that the person will believe the law to be legitimate. Who would consider a law illegitimate if it agreed with their personal beliefs? But in real life, this will rarely be the case. Suppose we have three groups of people, A, B, and C. A’s values (in order of priority) are equality, liberty, and the objective pursuit of knowledge. B’s values are religion, liberty, and equality. C’s values are liberty, equality, and religion. Now consider a law that protects a woman’s right to abortion. This law embodies the values of liberty (the right to abortion) and equality (in this case women’s) but disagrees with the religious values of B and C. For simplicity, assume each group views the gain in liberty and equality equal to the loss in religious values. How will each group react? Case A is uninteresting. Since their values align with the law’s, they will believe the law legitimate. In case B, we would expect them to believe the law illegitimate. It disagrees with their most highly held values. In case C, however, even though the group believes the law to be immoral, they should believe it legitimate because they hold the values of liberty and equality higher than their religious values. Again, I am not making the case that this is how these groups should behave, but that it is how they do behave. There is clear evidence of all three groups in the United States, for instance. There are liberals who support abortion wholeheartedly, the religious right that clearly thinks any abortion laws are clearly illegitimate, and those who consider abortion immoral but support women’s equality and rights.

Imagine a society where the public’s shared values aligned perfectly with the values embodied by the government. This may be a democratic nation where the public values liberty and equality above all else. The public believes the government to be legitimate and by any normative theory of democracy they should. But the imagined society does not have to be democratic. It could be a nation with a completely homogeneous population that believed in a single religion. Its government bases its laws on the holy text of this religion. Due to their belief, the public completely accepts the authority of these laws. Is such a government legitimate? According to any of the normative theories of democracy we have discussed, the answer would be no. Yet this contradicts the objective fact that the public unanimously believes their government to be legitimate. It does not seem reasonable to me to conclude that a government is not legitimate if every single citizen believes that it is.

Therefore a political theory must explain how such different governments can both be legitimate. This can be accomplished if we judge the legitimacy of a government as a measure of how well the values embodied by a government’s laws match the shared values of the public. If these values match, as we have seen, individuals will believe the government legitimate, making it legitimate in some objective sense. Therefore different forms of government may be legitimate if the publics of various nations have different values. A theocracy is legitimate if the public holds their religious values above all others, and a democracy is legitimate if the public holds the values of liberty and equality above all others.

Now I turn my attention to the problem of deference. Suppose we have a democratic government, and a public with the shared values of liberty and equality. My previous example of abortion law is illustrative. We’ll ignore group A since they are in agreement with the law so no moral deference is needed. Let’s consider group C first this time. Here we have a group that believes abortion to be morally wrong. Yet they obey the law and believe it legitimate because they are deferring to their own values of liberty and equality. However, group B would not be expected to do so. The law violates their most highly held values. If I am correct that individuals use their personal values to judge the legitimacy of political outcomes, this leads to the obvious objection that group B cannot be expected to morally defer in this matter. Therefore my theory cannot adequately explain why someone should obey a law they believe is immoral.

However, I believe there is one value to which we can objectively expect people to hold above all others - our collective need for survival. We recognize that some government is necessary to prevent widespread catastrophe and the violation of the values we do share. This brings us back to where we started, that any government that does well on primary bads must have some legitimacy. What are primary bads if not situations that threaten our collective survival? So even if someone lives under the authority of a government that does not embody the values they hold most dear, they can defer to their value of survival. It should be understood that I am talking about the survival of social groups, not individuals. While a person may hold some other values as more important than their individual survival, I would expect that such a sacrifice came from the belief that it benefited the survival of their social group. I believe than any group that did not value their collective survival would eventually die out. Thus, unless a political outcome threatens their survival, groups that hold values that are not shared by the general public should recognize the need to work within the political system to avoid the kind of prosecution that would threaten their survival.

So, given the need for everyone to work together for survival, we must recognize that a government that embodies our shared values is the best we can do. However, while is not fair to privilege certain values over others, in a modern, heterogeneous society, it may be that liberty and equality are the only generally acceptable values. If this is the case, then democracy will be the only legitimate form of government. There is room for a normative theory of democracy, but it would have to rely on an appeal that liberty and equality are the values that we should prioritize over all others, perhaps because it is precisely these values that will best help us survive.

[1] Estlund, David. Democratic Authority. Princeton, New Jersey: Princeton University Press, 2008. Pp. 98-116, 159-183. Print.

[2] Somin, Ilya. Democracy and Political Ignorance: Why Smaller Government Is Smarter. Stanford, California: Stanford University Press, 2013. Print.