What is the Value of Democracy

Sketch of liberal pluralism

In the previous paragraph I suggested that most empirical political scientists are, in an implicit and haphazard way, using an operating model of democracy that most resembles what has come to be known as liberal pluralism.⁶ ‘Liberal’ in the historical use of the term as upholding the values of freedom of choice, reason, and tolerance, and being situated against the overreaching tyranny of the Church or state (see Held, 2006, p.59 for a historical overview of the development of what we call liberalism). And ‘pluralist’ because it assumes and accepts heterogeneity in the beliefs and preferences of citizens. Or we could call this a concern with being respectful of diverging conceptions of the good.

This leads directly to the institutions of democracy we are most comfortable and familiar with, namely things like voting and representation. And, unlike, other models, it is typically not about achieving any sort of consensus or common good; it is about giving legitimacy to outcomes that are reached in a fair manner (i.e. the winner is based on whoever had the plurality of votes). And although there are various flavours of liberal pluralism, at its core I would describe it thusly: because human beings tend to differ in their beliefs and preferences in how to order society, about morality, and many other things, the best we can do is somehow aggregate those preferences or interests so that the preferences or interests of the greatest number of people get fulfilled.

What is the value of democracy?

There has to be a reason why we want democracy in the first place. If we know why democracy is good, then we can better answer what democracy is supposed to achieve for those within the polity. Although usually unstated, much empirical political science on democracy seems to put the value of democracy on the idea that citizens are supposed to get what they want to the greatest extent (i.e. preference maximization). In which case, why even have a representative democracy? More direct forms of democracy will achieve that far better. And the American system of democracy seems to be founded exclusively against that mindset. Writing in Federalist No.10, for example, Madison worries about the “superior force of an interested and over-bearing majority” (1982, p.43) who, if left unchecked, would lead the country into “spectacles of turbulence and contention”, not respect the rights of property, and eventually ensure the overthrowal of the regime (p.46). Instead, Madison proposes a Republic, or a representative form of democracy where elected officials “whose wisdom may best discern the true interest of their country” will be more likely to achieve the public good (p.47). The institutions that Madison and his colleagues imposed as a reflection of this view further hinder the pure fulfillment of preferences. That is, not only the act of representation itself, but other practices such as judicial review, indirect election of senators (initially), and the electoral college. Who is right between this competing view of Madison and contemporary political scientists is less important to the purposes of this essay; indeed, one of the problems is that I am making an assumption about what most political scientists view the value of the democracy to be. But I have to make that assumption because very few actually explicitly spell that out. So, I will correct that by stating as clearly as possible what I see as the value of democracy.

I believe the value of democracy is autonomy.⁷ That is, democracy is the most desirable form of government because it values the autonomy of citizens within the polity. Of course, autonomy rests upon another principle–the belief that every human being is infinitely precious and valuable, and that every human being is (and ought to be respected) as an individual entity with a right to order their own life and pursue the ends they see fit. I like the term ‘agency’ here too, but agency doesn’t capture the quality of self-government that autonomy implies. Of course, the principle of autonomy has a natural antipathy to collective government. As soon as I extend the size of a polity to include more than a single individual, there is a possibility of some curtailment of autonomy, of some surrendering of one’s ability to govern oneself or to order one’s affairs as one sees fit. That is the natural tension that exists in any democracy. Total freedom cannot exist in any ordered system of government. Given, however, the irrepressible fact that human beings have to live together in some form of collective entity, we need a way to govern ourselves that maximizes our chance to achieve autonomy. And that is what democracy does which no other form of government can do. That is because only democracy truly values the governing capacity of every citizen by allowing them a degree (however small) of participation in the political system. Therefore, I would say that democracy is both substantive and procedural. The substantive value it promotes is autonomy, but that autonomy is expressed through the procedure of participation in the political system. This can only happen in a democracy. The most enlightened despot could work to maximize citizen autonomy, but would never succeed in providing true autonomy to their citizens because they do not allow citizens to actually practice that autonomy by governing themselves.

Furthermore, I think most political scientists either are operating with this value under the hood or would be comfortable with it as the fundamental value of democracy.⁸ For example, returning to the “Statement of Concern”, we might say that overly-restrictive voter ID laws designed to make voting more of a nuisance violate this principle of autonomy. That is, they fail to recognize that each person should have the same right to participate in self-government; instead, they try to restrict the government of the polity to only a certain segment of the polity.

And autonomy works well with the unstated assumption that many political scientists have about preference aggregation being the heart of democracy. After all, that is how political scientists often consider voting–the chance to aggregate preferences about policy, with the majority (or plurality) representing the greatest proportion of preferences. I would adjust this focus slightly. While I do consider preference satisfaction important for LP+ democracy, it is not the end of democracy, but a means to achieve the goal of autonomy. Certainly, if autonomy is the goal, then it makes perfect sense to listen to what citizens say they prefer in terms of policy. Of course, this necessitates that we are comfortable with those stated policy preferences often being misguided or simply bad public policy. If I took this argument to its full extent, we would, in the name of prioritizing autonomy, have to only concern ourselves with satisfying what people say they want, not on what may actually be in their best interests. The logic being that this is the only way to truly respect their autonomy as individuals to decide on what life is best for them to lead.

While that argument resonates strongly with me, I think it can be made more palatable by adding in the concept of positive liberty (as opposed to negative liberty). Although negative and positive liberty would be familiar to many philosophers, not least of all John Locke and Jean Jacques Rousseau, the best exploration of these concepts comes from Isaiah Berlin (1969). Berlin calls negative liberty “freedom from” (p.146). That is, liberty is freedom from interference from others (especially the government). The Declaration of Independence contains clear examples of negative liberty: “We hold these truths to be self-evident, that all men…are endowed by their Creator with certain unalienable Rights, that among these are Life, Liberty and the pursuit of Happiness.” Jefferson (and the other Founders) clearly believed that human beings have natural rights that they deserve simply by virtue of being human; these rights mostly being in being left alone to pursue one’s life’s goals as one sees fit. This is a conception of liberty that is most familiar with Americans, and tends to resonate the most given our value system.

Isaiah calls positive liberty the “freedom from unrealizable desire” (p.136), and later quotes Rousseau when he says that “he is truly free who desires what he can perform, and does what he desires” (p.139). Freedom then is about having the capacity to fulfill the measure of one’s being. This is a corrective to only relying on negative liberty. Consider Berline’s logic (p.139-140):

“If I find that I am unable to do little or nothing of what I wish, I need only contract or extinguish my wishes, and I am made free. If the tyrant (or ‘hidden persuader’) manages to condition his subjects (or customers) into losing their original wishes and embrace (‘internalize’) the form of life he has invented for them, he will, on this [negative] definition, have succeeded in liberating them.”

What Isaiah is expressing here is the possibility of human beings changing their desires in response to adversity or persuasion. If I’m North Korean, and the autocratic state has convinced me, say, that what I really want is to spend the precious years of my life grubbing for silkworm pupae for the Peerlessly Great Man, then I will be free if the government stays out of my way and allows me to do that. But is that really freedom? Isaiah would say no. Freedom would be to have the education and resources at one’s disposal to learn that such grubbing for such a dictator is degrading. Freedom would be gaining the knowledge to pursue a life plan that better satisfies one’s immense human potential.⁹

I think there is room for both positive and negative liberty within a democracy, especially as it relates to autonomy. Take the case of highly addictive drugs such as heroin. If we value autonomy and negative liberty, presumably we should allow a person who expresses a preference for imbibing heroin to do so unfettered. After all, who are we to say what is a good use of their life? We have to respect their wishes to be autonomous, which is freedom from interference from the state in telling them how they are and aren’t allowed to live their life. But, someone who values positive liberty and autonomy would reply that drug addiction is the antithesis of freedom. It retards one’s ability to fulfill the measure of their being. Instead, one is a slave to a synthetic substance.

Freedom, in this case, is freedom from such a dependency which opens up the possibility for one to live a more fulfilling life. But one could obviously go scary places rather quickly if we overemphasizes positive freedom alone, without the accompanying check of negative freedom. For example, why not let the government ban all entertainment that does not obviously uplift the human spirit? Surely that would increase our capacity to fulfill the measure of our being if we didn’t all waste time sitting on our couches watching The Great British Baking Show? I think it would, but I also wouldn’t want to live in a world where I can’t watch contestants melt under the pressure of the steely gaze of Paul Hollywood. Of course, my example is deliberately extreme, but there are real examples of positive liberty all around us; examples where the line of how much is too much is often difficult to judge. For example, seat belt laws, laws regarding the regulation of alcohol, tobacco, and marijuana, laws regarding the permissibility of pornography, and so forth.

As it relates to preference aggregation then, autonomy can be fulfilled by both positive and negative liberty, and preferably by some sort of balance between them.¹⁰ Or more specifically, we can fulfill the demands of autonomy both by seeking to follow the preferences people state, and allowing them the freedom to pursue such preferences without interference from the state, while at the same time recognizing that autonomy requires restricting human behaviour when that behaviour actually harms autonomy in the long run. This is a precarious tightrope but there are often sensible solutions. One example would be universal childhood education. Yes, it restricts the autonomy of individuals to force them to be educated, but it also enhances their autonomy in the long-term because it makes them more capable of understanding and navigating their world. It gives them the tools to truly order their life as they see fit.

Thus, while I consider preference aggregation or satisfaction to be an important component of democracy, it is not the be all, end all. Preference aggregation is merely one path to achieve the goal of autonomy; it is not in and of itself the value of democracy. Furthermore, there may be multiple ways for preference aggregation to work. The default view of majority rule¹¹ (while crucially important) is not the only way we might think about satisfying preferences. We might, for example, weight them so that some recognized groups get an allocation of influence in the aggregation process. Many examples of this have been tried or are being currently tried. Quebec is a good example. Although French-speakers constitute a minority of Canada, they have certain guarantees built into the political system to ensure their minority preferences are met. Historically, Lebanon operated on similar lines. Christians and Muslims alike had to have a certain allocation of representation in government, not based on population but on creed.¹²

These two examples highlight another natural tension to democracy that comes from valuing autonomy above all else. Namely, the problem of citizens using their autonomy to curtail the autonomy of others. If we hold only to the principle of majority rule, there is little reason to expect that whoever’s preferences represent the majority view will refrain from using their greater numbers to restrict or destroy the autonomy of those who hold minority views. Such real world examples as slavery or the Jim Crow laws in the American South show that this is not merely an academic concern. The possibility of majoritarian terror obviously defeats the goal of seeking as much autonomy as possible for as many people as possible.

What this necessitates is not abandonment of majority rule, but something that is already familiar to American citizens. The imposition of a constitution that safeguards against the naturally tyrannical impulses of the majority. Or, we might think of the institutional structure of the democracy–designing institutions that perform a similar function. Judicial review in the U.S. is one such institution. It puts a check on legislation that presumably reflects mostly the views of the majority. So, although LP+ seeks to deliberately treat all citizens’ views as worthy of consideration by virtue of them being important to them, we have to accept that there are some values that have to be accepted no matter what the individual belief. That is, there have to be foundational values that are binding to all, no matter what a person’s individual beliefs. ¹³ To be specific, all citizens have to accept (no matter their personal views on the matter) that they cannot pursue policy preferences that destroy (or seriously harm) the ability of others to pursue their policy preferences.

The role of the citizen

In the previous paragraph, I posited that citizens cannot pursue preferences that rob others of their opportunity to self-govern. Citizens cannot, for example, choose to enslave the minority portion of the population. Doing so robs all those individuals of their autonomy. What else is required of citizens? Two more things. Firstly, citizens should not use their right to participate in a democracy in order to destroy democracy. Secondly, citizens should follow the law.¹⁴ Apart from these basic criteria, there is not much else that is absolutely required of citizens.

Importantly, there is no expectation citizens be highly informed about politics. People are busy; they have limited time, interest, and opportunity to participate directly in politics.¹⁵ That is the whole point of a representative system. The legislator can act as an agent for your interests. And, if they do so poorly, can always be voted out by the citizens they represent. For citizens, there is no expectation that they are well-informed about politics, they simply need to fulfill the baseline expectations I mentioned previously.¹⁶ The system should ensure that every citizen can participate if they want (this should be part of the initial constitution), but citizens are free to ignore politics as much as they want. Of course, their preferences will probably be more likely to be met to a greater degree if they are more informed about politics, but that is beside the point of democracy. Democracy is about giving them the opportunity to exercise autonomy, not ensuring that their preferences are met. If they are met, so much the better though. Likewise, more responsiveness on the part of representatives is better (in terms of maximizing autonomy if not necessarily in terms of achieving beneficial outcomes), but what matters more than responsiveness is the theoretic opportunity for citizens to participate (no matter if their representatives are responsive or not).

Now, traditionally, political scientists have often believed that citizens need to have some baseline level of competency. At the very least, they need to have some ideological constraint (see Converse, 1964), and know what their preferences are in order for elected officials to represent them. Without stable preferences, so the thinking goes, there can be no real accountability or even representation. I disagree. All that matters is that citizens have the opportunity to participate in politics if they choose to. Participation can be as minimal as having the opportunity to cast a vote. Gerrymandered voting probably doesn’t meet this criteria, but most other voting would. Obviously more participation, and more informed participation is better, but it is not a necessary condition for citizens. Allowing citizens the right to choose not to participate in politics, or participating in an ignorant fashion, is part of respecting their autonomy.

The problem of endogeneity

Critics of the liberal pluralist model often point to the fact that preference formation is not exogenous to the political environment (e.g. Barber, 1984). Citizens’ preferences are certainly influenced by elite rhetoric and discourse (see Zaller, 1992), and often end up reflecting their preferences. That is how we end up with policy outcomes that favour the wealthy, despite their numerical inferiority (Bartels, 2008; Gilens, 2012; Gilens & Page, 2014).

Given that preference formation is undoubtedly manipulated by elites in the system, why should we seek to fulfill those preferences? It is a fair question and critique. All else equal, certainly I would prefer that citizens have a roughly equal chance to influence one another in terms of preference formation, and thus have policy that reflects all strata of society more equitably. But that is a high standard; too high a standard for really any democracy that we have seen heretofore. As long as democracy exists in countries with economies mostly based on the principle of the market (i.e. nearly every democratic country), then there will exist a class of people who (by dint of possessing greater resources) have more opportunities to be better informed, more connected, and more able to influence elected officials. That is the reality of the environment in which every democracy in the world today exists. And, however normatively displeasing, that is not a death knell for democracy. If it were, no democracy would exist, or would have existed up to this point. Each one has a significant amount of inequality of political influence baked in. We simply have to accept the reality that preference formation will be influenced by elites. At a minimal level, satisfying autonomy only means that we treat citizens respectfully–that is, that we treat them as if they hold rational preference that were developed independently of elite manipulation. That we treat citizens (rightly or wrongly) as if they had taken the time to truly consider whether their preferences benefited themselves, or were based on logic or even the common good, and parsed out what they believed to be elite manipulation. Sound like wishful thinking? Perhaps, but rule by the people means that we have to live with people participating ignorantly or people participating who merely parrot the beliefs of their social “betters.” For my part, I am more sanguine about the role of elites given what we know about collective identities. Elites generally share the same social identities as the rank and file of the citizenry, and thus seek to benefit their own ingroups (which includes ordinary citizens). Elites are still black and white, male and female, religious and secular, gay and straight, American and foreign. By seeking to maximize the benefits to these social groups, they benefit ordinary Americans who share these identities.

Of course, there are things that can be implemented in the system to try and minimize the role elites play in influencing citizen preference formation, or at least to fracture it. One example is a free press where citizens can (if they choose to) expose themselves to elite and non-elite arguments from all sides of an issue. Another example is developing the capacity of the citizenry to resist elite framing by formal education. Another might be reliable fact-checking of candidates. A different approach might be to limit campaign contributions and airspace to the wealthy. None of these will truly overcome the problem of endogeneity, that is insurmountable, but they can theoretically curtail the most deleterious aspects. Of course, if we borrow from deliberative models of democracy we might say that endogeneity is the opportunity, not the problem. It is a chance for citizens and elites to deliberate about policy, and the idea that preferences need to be formed outside of that deliberation unnecessary at best, and naive at worst. Whichever way we choose to look at it, however, endogeneity need not mean that we give up on satisfying preferences. But, as I’ve also written, preference satisfaction is only part of the equation. We only care about satisfying preferences that do not harm autonomy, preferences that do should be left unfulfilled.

Where does this leave us?

So far my main point has been this: the reason why we should want to live in a democracy is because a democracy is the only system that values human autonomy. It does not grant total autonomy because no system can when we are dealing with a collective of hundreds of millions of people. Democracy is better, however, when it both uses a process that promotes autonomy (for example, extending each citizen the right to have a say in the outcome of a political decision) and achieves outcomes that promote autonomy. We can therefore judge how democratic a country is by considering the amount of autonomy extended to citizens to make decisions, as well as by the outcomes (the laws) that result from that process. This is how the system gains legitimacy. It should be respected as legitimate by its citizens when the procedure of making decisions gives some (possibly small) measure of autonomy to each citizen to exercise political sovereignty, and when the outcomes reached through that process also value the autonomy of citizens. It is the procedural part, however, that I think should do the heavy lifting in terms of legitimizing the system. Evaluating outcomes is more fraught, more subjective. I share the same concerns as Estlund (2008, p.99) on relying too heavily on democracy’s epistemic ability. Since “the good” as Estlund would put it (in this case, whether a decision promotes autonomy) is often contested, this can lead to instability in the system–people only believing that the system is legitimate if outcomes they think are good are reached. In other words, a situation where laws are only obeyed (or rules followed, or election results accepted) when they are favoured by that citizen. That seems overly-precarious from the standpoint of someone who values stability. We should accept democratic decisions as legitimate because they were arrived at through a democratic process. Better decisions, of course, only help in the process of legitimation, and thus (agreeing with Estlund) I would say democracy does have some epistemic value too, despite mainly gaining legitimacy procedurally. To put it in terms of autonomy again, democracy is legitimate because the process of participation allows citizens to exercise autonomy. That is key. Supplemental to that, but still important, is the legitimacy gained when outcomes enhance the autonomy of the citizenry.

And, given the tendency of people to mostly pursue what’s good for their own groups (even at the expense of other groups), a democratic system should be undergirded by a constitution that espouses values that cannot be superseded,¹⁷ whatever the will of the majority. This constitution could also outline certain institutions that further buttress this protection for certain values that are important but which could be put under threat by a hostile majority. This is one of the primary reasons I call it liberal pluralism plus. The plus is the recognition that satisfying preferences is not enough; we should not always be interested in following the wishes of the greatest number of people. Democracy also requires that certain values be respected regardless of what the people want. Do I think such values could survive a permanently hostile majority? Probably not. At the heart of the democracy, however, is the recognition that human beings deserve respect and dignity and the ability to self-govern (as far as that is possible). If we can inculcate those values in citizens, a constitution would seem less vulnerable. In this respect, I agree with aspects of Anderson’s (2009, p.214) argument:

“Democracy is more than a set of governing practices. It is a culture or way of life of a community defined by equality of membership, reciprocal cooperation, and mutual respect and sympathy and located in civil society…democracy as a mode of government cannot be fully achieved apart from democratic culture…democratic government is a manifestation of democratic culture; its point is to serve the democratic community”.

Without citizens embracing that democratic culture, the tenets of democracy are going to be very difficult to maintain. And this is an area where Americans need a lot of work. Using both a natural and set of lab experiments, Graham and Svolik (2020) found that Americans universally support democracy in the abstract, but routinely put partisan considerations first when evaluating candidates who exhibit anti-democratic behaviour. As they note, “Only a small fraction of Americans prioritize democratic principles in their electoral choices” (p.393). How small? Only 4% of partisan voters in their experiments were willing to punish candidates who exhibited anti-democratic behaviour, and only 13% of partisans were willing to defect to the other party’s candidate who supported democracy. Clearly, we need something (perhaps civics classes?) to better inculcate support for democracy in this country.

I wrote about this problem in the months leading up to the last presidential election, trying to convince conservatives to defect from former President Trump. I argued that voting third party was a good way to punish his undemocratic behaviour (i.e. he threatened before the election that he wouldn’t commit to supporting a peaceful transition of power), while still valuing one’s ideological position (more or less). Ultimately, there was less defection than I hoped. Even milder conservatives largely held their noses and voted for Trump. In this they replicated Graham and Svolik’s finding–centrist voters in their study could only tolerate about a 15% movement away from their preferences before they chose a candidate who valued democracy more.

Are Political Scientists right to be worried?

Let’s return to the three issues the signers of the Statement of Concern raised: election deligitimization, overly-restrictive voter ID laws, and gerrymandering. How do these behaviours fare in the LP+ conception of democracy? I would say that respecting the outcome of elections has to be a democratic principle that is accepted regardless of how the people at large (i.e. the majority) feel about the loss of their candidate. This is a case of principle triumphing over preference maximization, which, as discussed, is par for the course in LP+. Obviously, if the results of elections are not respected, the idea of self-government by popular participation is in serious jeopardy. So, yes, I would agree with the signers that this trend is a serious threat to democracy. We have reached an unprecedented point with this disbelief in the legitimacy of elections in this country. Recent polls indicate that the majority of Republicans believe that the current President was only elected by fraud, and only a third of Republicans indicate that they are willing to trust the next election results, regardless of who wins, despite no real evidence of any widespread fraud.¹⁸. If citizens believe they are being ruled by a dictator who seized power illegitimately, it is their right to try to overthrow the government. That is the logical conclusion of such thinking, which would obviously destroy democracy as we know it in this country.¹⁹

Overly-restrictive voter ID laws are also problematic under LP+ because they violate the principle of autonomy. Each citizen has to have an opportunity to participate politically, so that they can have some say in how they are governed. These laws, when they are needless and designed only to penalize certain constituencies, thus go against the very value of democracy. I would say gerrymandering is similarly bad on similar principles–it makes self-government harder for some people rather than others.

So, yes, the signers are justified in their censoriousness. What I hope this essay has done, however, has to be to lay out a vision of democracy that more fully articulates why such violations are actually problematic. They are problematic because they violate the value of democracy. In my reading of democracy, democracy is about human beings who have to live in a collective being able to come up with a system of government that ensures them the greatest degree of autonomy. Autonomy can be represented both in freedom from government interference, but also freedom to achieve a greater fulfillment of life as a result of living in such a system. Thus, elected leaders face the tricky problem of trying to fulfill both positive and negative freedoms, which are often at loggerheads. Navigating that balance is the heart of democracy, and it requires preference satisfaction at times, and interest satisfaction at others, which may not always converge. It also requires that citizens buy in to certain principles that have to be fundamental to the system and not popularly mandated.

From this theoretical base, many different institutions or “on the ground” set-ups of democracy are possible. That is one of the great attractions to LP+. It sets up a certain framework of governance (the fundamental values that I mentioned), has a guiding principle (autonomy), and within those constraints, then allows the real hurlyburly of democracy to take place (i.e. the preference aggregation). It is a more sustainable and satisfying conception of democracy than relying on preference satisfaction alone, not to mention more flexible. It does not say which values have to be built into the system externally, only that these values should exist. You could, using this set-up, construct various visions of democracy that I would be quite happy living in. I think that is the value of such admittedly abstract theorizing. It allows us to evaluate the features of democracy that exist in our own systems of government according to an abstract system. For example, institutions such as the Electoral College obviously can be evaluated quite easily given the underlying stress on promoting autonomy that LP+ emphasizes. In this way, I hope this conception of democracy could be useful to both theorists and policy-makers who grapple with democracy “in the raw”.

Update (November 2023)

Every Friday I run some discussion sections for Political Science 100. In my last one, I decided to have the students undertake the exercise themselves. That is, I asked them to define the fundamental value they wanted government to express, then choose a form of government and set of institutions based on that chosen value.

88 students participated, and we ended up with 10 total values chosen between these 88. The most common value chosen was equality, but note that that only represents about one fifth of the class. The chosen values were more evenly distributed than I would have thought. This was reflected in the forms of government chosen to express those values, and more particularly the institutions. We ended up with a wide array of institutions designed to express each particular value. This really helped the students contextualize the institutional set–up of the United States, a government that is clearly designed to express the value of negative liberty (freedom from interference from external authority). A government that was designed primarily to express a different value, like welfare of its citizens, would end up with very different institutions. Building the system from the bottom up, with the value defined explicitly at the beginning, is a much better way to understand what institutions are designed to do.

Works Cited

Althaus, Scott L. 2006. “False starts, dead ends, and new opportunities in public opinion research.” Critical Review 18 (1-3):75-104.

Anderson, Elizabeth. 2009. “Democracy: Instrumental vs Non-instrumental Value” in Contemporary Debates in Political Philosophy, eds. Christiano & Christman: 213-228.

Appiah, A. 2005. The ethics of identity. Princeton University Press.

Barber, Benjamin. 1984. “Citizenship and participation: Politics as epistemology.” In Strong Democracy: Participatory Politics for a New Age. Berkeley, CA: University of California Press.

Bartels, Larry M. 2008. Unequal Democracy. New York: Russell Sage.

Berlin, Isaiah. 1969. “Two Concepts of Liberty.” In Isaiah Berlin, Four Essays on LIberty. London. Oxford University Press.

Converse, Philip, E. 1964. The nature of belief systems in mass publics, Critical Review, 18:1-3, 1-74.

Dahl, Robert. 1989. Democracy and Its Critics

Downs, Anthony. 1957. An economic theory of democracy. New York: Harper Collins.

Estlund, David. 2008. “The Irrelevance of Jury Theorem” in Democratic Authority, chapter 12.

Gilens, Martin. 2012. Affluence and Influence. Princeton, NJ: Princeton University Press.

Gilens, Martin, and Benjamin Page. 2014. “Testing Theories of American Politics: Elites, Interest Groups and Average Citizens.” Perspectives on Politics 12: 564–81.

Graham, M. H., & Svolik, M. W. 2020. Democracy in America? Partisanship, Polarization, and the Robustness of Support for Democracy in the United States. American Political Science Review, 114(2), 392–409.

Held, David. 2006. Models of Democracy, 3rd ed. Cambridge: Polity Press.

Kolodny, Niko. “Democracy for Idealists” unpublished manuscript.

Madison, James. 1982 [1787-1788]. “Federalist No. 10.” In The Federalist Papers, edited by G. Wills. New York: Bantam Books.

Mill, J. S. 1998. On Liberty and Other Essays. Greece: Oxford University Press.

Mill, J. S. 1861. Considerations on Representative Government. United Kingdom: Parker, Son, and Bourn.

Przeworski, Adam. 1993. “The Minimalist Conception of Democracy: A Defense” in Democracy’s Value.

Riker, William. Liberalism Against Populism (1982)

Schattschneider, E. E. 1983. The Semisovereign People: A Realists View of Democracy in America

Schumpeter, Joseph. [1942] 1976. Capitalism, Socialism, and Democracy.

Sunstein, Cass R. 1991. “Preferences and politics.” Philosophy & Public Affairs 20 (1):3-34.

Zaller, J. 1992. The Nature and Origins of Mass Opinion (Cambridge Studies in Public Opinion and Political Psychology). Cambridge: Cambridge University Press.