Democracy is Autonomy

Why Democracy?

In a year of election ballots filled with questionable candidates, social media fights over which convention scored points for the best celebrity cameo, and policy concerns falling on deaf ears, you might be asking yourself why we, the citizens of apparently the world’s greatest democracy, put ourselves through such an ordeal and whether it is still worth it.

Ask a democracy enthusiast and they will tell you that democracy might be worth it because democratic institutions have desirable consequences. For instance, some claim that democracy preserves stability in the state, that it generates economic growth, that it is particularly good at respecting the rights of citizens, and that sometimes, it might even help prevent famine. These are all worthy goals for institutions to pursue. However, they fall short of explaining why we value democracy independently. Think of a state that is governed by a benevolent algorithm overlord. A benevolent algorithm is one that legislates in a morally perfect way, always calculating the most efficient manner of arranging social conditions to achieve ends. Say that the algorithm was, in fact, much better at achieving these ends than many of our democratic states. It would not have to organize lengthy and costly elections or referendums, write and argue through competing bills, or worry about pleasing competing constituents. All of the desirable ends listed above would be more easily met in some non-democratic form of government. So, why democracy?

We might instead suggest that democracy is not instrumentally valuable for achieving some end, but rather, it is intrinsically valuable. What, then, makes democracy intrinsically valuable? There are two usual suspects: equality and freedom. Relating to each other as equals in a democracy seems pretty important and indeed, in contemporary discussions it has become a popular way of understanding the value of democracy. If we want to avoid relations of inequality among citizens, it seems that democratic structures best institutionalize against this. Furthermore, democracy might be particularly compatible with the freedom of citizens. The state is much more likely to not interfere with the freedom of the citizens if the state manifests the will of the people (I am unlikely to desire my own unfreedom). But does this pass the benevolent AI overlord test? It seems like government by algorithm would be compatible with both these features. If we were all equally subject to a benevolent AI, then there would be no political inequalities (at least with regard to political power) among citizens; all the citizens would be equally subjected to the algorithm. Furthermore, the freedom of citizens might be equally protected under an AI overlord insofar as the AI might plausibly arrange for the minimum amount of interference while also achieving political goals.

Yet, we still seem to have an intuition that something would be missing if we were to hand over decision-making power to the benevolent algorithm overlord. What is this something? I think it has to do with freedom, but not just freedom in the negative sense (not being unfree) but in the positive sense (being determined by your own laws). This is what is often called autonomy.

Immanuel Kant is the premier philosopher of autonomy. He famously argues, in his Groundwork for the Metaphysics of Morals, that every other philosopher up to that point had failed to define the principle of morality for the same reason: they had failed to account for the fact that the will is autonomous. What does Kant mean by this? Kant takes autonomy to be a necessary concept to make sense of moral obligation. This is because, when we examine the kind of obligation that follows from moral concepts like the good will and duty, we arrive at a concept of law that binds unconditionally. Two things are implied by this: first, that the bindingness of moral obligation is completely independent of antecedent desires (it is not dependent on anything like inclinations or further ends) and second, that bindingness of the law is based on the lawful form rather than the substance of the obligation (it is motivated by lawfulness alone). Now, in order for this bindingness to be possible in the real world, the will must be autonomous. Otherwise, how could the will be bound through nothing but the lawful form of its maxims if it did not give itself laws. Legitimate moral obligation requires that I self-legislate. Thus, without the autonomy of the will, moral laws are, as Kant puts it, mere fictions.

While being the premier philosopher of autonomy, Kant is usually not applauded as a great philosopher of democracy. Despite this, in his main works of political philosophy, he argues that there is a parallel between his theory of autonomy in the human will and the state. As he puts it, just as there is internal and external freedom, there is internal and external legislation. Internal freedom and legislation are about the possibility of legislating in the individual will while external freedom and legislation are about the possibility of legislating through a public will. However, these significant differences (and the fact that Kant doesn’t use the word “autonomy” in his political works), I think, does not change from basic place of autonomy in Kant’s theory of the will (Private or public). The main question in the political sphere, as in the moral sphere, is how political obligation is possible. Kant suggests that in the state, in addition to the principles of freedom and equality there is a third principle which he sometimes calls independence and sometimes dependence. This principle articulates the subject’s simultaneous independence from other private wills and a complete dependence on the legislation of a publicly constituted will. This is autonomy in the public sphere, to be dependent on no other will except one which you co-legislate. This is not merely because it has particularly good outcomes for us, or is particularly compatible with equality and freedom, but rather because this independence through dependence on a public will makes laws, political obligation, and our political freedom under laws, possible.

So why democracy? It seems like we cannot appeal to some purpose outside of democracy, but rather that democratic rule is conceptually necessary to make sense of political obligation. What this means for us is that democratic institutions are necessary independent of their difficulty or even their sometimes bad outcomes. Democratic rule is not an instrument which we use to achieve political goals and so cannot be evaluated on whether it adequately delivers on these points. It is, itself, our freedom. Democracy is autonomy.

Mike Gregory

Mike Gregory is currently a Postdoctoral Research Fellow at University of Edinburgh Law School and Edinburgh Futures Institute under the project "Democracy, Rights and the Rule of Law in a Data-Driven Society". Mike's work is on Kant's Legal and Political Philosophy, the Rule of Law and Automation, as well as republican political theory. His work has appeared in Ratio, The British Journal of the History of Philosophy, Kant-Studien, and Kantian Review. He is currently working on a book on Kant and Democracy as well as a volume on AI and Public Law.