Let’s Keep Democracy

Thank you, your comment will be visible after it has been approved.


October 5, 2018

Let’s Keep Democracy

But let’s look for better alternatives

By Jason Brennan

  • Conservative
  • Democracy
  • Europe
  • Liberal
  • Politics
  • United States

The following essay were presented as part of the day-long conference “Democracy in Trouble?” at the University of Pennsylvania’s Andrea Mitchell Center for the Study of Democracy. As the post-Cold War democratic order is straining under the dual threat of authoritarian and exclusionary movements on the national level and transnational oligarchic networks, the goal of the conference was to take account of the different facets and causes behind these developments. Originally published on the Mitchell Center’s website, these pieces are also natural fits for Public Seminar’s Vertical “Liberal Democracy in Question.”

The BMW 3-series is wonderful, often the best in its class, but it nevertheless has significant flaws. BMW’s engineers acknowledge its shortcomings and continually attempt to improve it. Consumers love the car, but also recognize in a given year that competitors may be better, and often buy the competitors instead. Iron Maiden is a wonderful metal band, but some of their albums are awful. The band avoids playing bad songs from those bad albums live.

I’m a fan of democracy the way I’m a fan of the BMW 3-series and of Iron Maiden. Democratic forms of government have generally been the best performing systems we’ve had so far in human history. [1] But democracies also suffer from pervasive and systematic deficiencies, flaws built into democracy itself. Democracies perform well compared to dictatorships, single-party communist states, feudal systems, or oligarchies. But most such governments are terrible, so saying democracy is superior is setting a low bar. Most political systems throughout history existed more for the purposes of extraction and exploitation rather than for human development.[2]

Democratic systems of government do a better job. Nevertheless, they also greatly underperform compared to their potential. For instance, the economic argument for radically increased immigration is overwhelming, but democratic communities around the world engage in extreme restrictions to movement.[3]

Now, we must be careful when assessing just why democratic systems perform as well as they do. Democratic countries also tend to have liberal economic and civil institutions, themselves the most important ingredients for sustained prosperity. Just why democracy and liberalism tend to come and go together is fiercely debated. Political theorists tend to treat politics as primary, as if all institutions and all institutional results flow from and are explained by political design. But economists generally know better. Insofar as we can treat “democraticness” as an independent variable, its independent contribution to generalized prosperity is weaker than, say, secure private property and open markets. [4] Maybe economic and civil liberalism does most of the work, but democracy is just a tagalong.

Danny Oppenheimer and Mike Edwards says democracy works despite itself, despite “irrational…voters and flawed…politicians.” [5] It is indeed surprising that democracies do as well as they do, in light of the severe pathologies voters and politicians exhibit. Let’s review.

Ample empirical work finds that the mean, median, and modal voter is ignorant of most basic and simple political information. [6] Voters are not merely ignorant, but in many cases misinformed. For instance, during the Brexit vote, UK citizens vastly overestimated the number of EU immigrants in the UK, overestimated Chinese foreign investment, dramatically underestimated EU investment in the UK, and vastly overestimated how much the UK sends to the EU in terms of various welfare payments. [7] Further, voters are not merely ignorant or misinformed about basic political facts, but also misinformed about the social scientific knowledge needed to understand those facts and determine which policy platforms have the best chance of achieving their ends.[8]

Empirical work in political psychology general finds that citizens do not reason about politics in a scientific or truth-tracking way, not even in an “ecologically rational” way. Rather, voters engage in motivated reasoning; they tend to believe what they want to believe rather than what the evidence supports. For instance, they suffer from both confirmation bias (they tend only to see out, to pay attention to, and to accept information that reinforces their current beliefs) and disconfirmation bias (they tend to reject or ignore information that undermines their current beliefs). They suffer from affect contagion and prior attitude effects; their current emotional states change how they process new information. They also suffer from framing effects; the opinions they offer in response to questions depends on the wording of those questions, and they will switch answers if offered a logically identical question that replaces one word with a synonym.[9]

Democracies seem to work well not because the voters are good at selecting parties who will achieve their ends. Rather, political parties have an incentive to keep voters happy but also have the wiggle room to avoid implementing the policies voters actually want. [10] For instance, the US has far stronger protection of economic and civil liberties than we would expect if the median voter got what he wanted. [11] Perhaps democracy works because it doesn’t work.

The democratic realist paradigm in empirical political science claims that so far I might be giving voters too much credit. For instance, there is strong evidence the majority of voters lack organized or stable sets of political beliefs. What few opinions they have cannot be amalgamated into a coherent position. They may label themselves “liberal” or “conservative” but generally lack actual ideologies. [12] They nevertheless are strongly attached to political parties — indeed, they are tribalistic in the worst sense of the word — but their attachments to a particular party result from arbitrary historical contingencies, not from the tendency of that party to promote their interests and not from any genuine commitment to their party’s platform.[13] For the typical Democrat, the thinking is more “People like me vote Democrat, and Democrats are pro-choice, so I’m pro-choice” than it is “I’m pro-choice so I’ll vote Democrat.”

Hélène Landemore’s previous work attempts to defend democratic decision-making using a priori mathematical theorems. These theorems would, if applicable to real-life democracy, explain how certain collective decisions produce good outcomes even though most individual voters are not reliable.[14] However, her most recent work appears to be empirical rather than theoretical. She thinks certain case studies can help us determine under what conditions collective and deliberative forms of decision-making succeed.

I have some worries about her recent paper on Iceland’s experiment with crowdsourcing reforms to its constitution.[15] Iceland is a lovely country, though it arguably underperforms compared to other Nordic/Scandinavian countries. Iceland nevertheless benefits from a number of factors which make democratic deliberation and consensus easier to achieve: it is small, ethnically homogenous, highly educated, has very high rates of human development, has high rates of interpersonal trust among its citizens, and has a strong history of civic involvement, broadly understood. Iceland should be able to make democracy succeed if anyone can, though if they can do it, it does not imply France or the US can copy their example.

In her paper, Landemore describes a situation in which a panel of “seven government experts” was asked to suggest revisions to the Icelandic constitution’s provision on religious freedom. In turn, twenty-five lay citizens were also asked to propose a revision, which they partially crowd-sourced. In the end, the crowd-sourced proposal was, Landemore argues and I agree, superior to the “expert” version. (I put “experts” in scare quotes because I’m unsure I’d consider the government agents experts.)

This was a relatively easy task. They are not being asked to write fine policy details on a hard question, but rather a meta-legal principle for a constitution. They have hundreds of years of history and thousands of legal case studies from hundreds of different constitutions upon which to draw. They could, if they so desired, examine at great length how different religious freedom provisions in different constitutions have tended to work — a topic on which there’s a gigantic theoretical and empirical literature They could, if they wanted, largely copy the language from already existing, well-functioning constitutions written by experts.

Even then, and even with all the other things Iceland has going for it, the text the laypeople came up with was not particularly impressive or innovative, and it strikes me as rather obviously worse than the provisions in other liberal countries. A 75% atheist country could not even manage to de-establish the Lutheran church none of them attend, making the new constitution to that extent illiberal and somewhat ridiculous. Landemore discusses at some length why de-establishment was not a live option. However, I take her explanation itself to be a defense of the draftees but also a simultaneous critique of Icelandic voters.

Democracy is the best system we have, especially when it is appropriately limited by liberal rights and appropriately matched with liberal economic and civil institutions. But it is also seriously flawed. We should be open to improving democracy as best we can. We should also be open to experimenting with possible alternatives to democracy on a small scale, and if they succeed, trying them on a larger scale. Perhaps most of all, when we recognize how pathological democracies are, we should be less enthusiastic about trying to solve problems through political, rather than non-political, means.

Jason Brennan is Robert J. and Elizabeth Flanagan Family Professor of Strategy, Economics, Ethics, and Public Policy at the McDonough School of Business at Georgetown University, and Research Professor at the University of Arizona’s Freedom Center and Department of Political Economy and Moral Science.


Acemoglu, Daron, and James Robison. 2013. Why Nations Fail. New York: Penguin.

Achen, Christopher and Larry Bartels. 2016. Democracy for Realists . Princeton: Princeton University Press.

Althaus, Scott. 2003. Collective Preferences in Democratic Politics. New York: Cambridge University Press.

Arceneaux, Kevin, and Robert M. Stein. 2006. “Who Is held Responsible When Disaster Strikes? The Attribution of Responsibility for a Natural Disaster in an Urban Election,” Journal of Urban Affairs 28: 43-53.

Arian, Asher, and Schamir, Michal. 1983. “The Primarily Political Functions of the Left-Right Continuum,” Comparative Politics 15: 139-158.

Barnes, Samuel H. 1971. “Left, Right, and the Italian Voter,” Comparative Political Studies 4: 157-175.

Bartels, Larry. 1996. “Uninformed Votes: Information Effects in Presidential Elections,” American Political Science Review 40: 194-230.

Bartels, Larry. 2003. “Democracy with Attitudes,” in Electoral Democracy, ed. George Rabinowitz and Michael B. MacKeun. New York: Oxford University Press.

Brennan, Jason. 2013. “How Smart Is Democracy? You Can’t Answer that Question a Priori,” Critical Review 26: 33-58.

Brennan, Jason. 2016. Against Democracy. Princeton: Princeton University Press.

Brennan, Jason. 2018. “Democracy and Freedom,” in The Oxford Handbook of Freedom, ed. David Schmidtz. New York: Oxford University Press.

Campbell, Angus, Converse, Philip E., Miller, Warren E., and Stokes, Donald E. 1960. The American Voter. New York: John Wiley.

Caplan, Bryan. The Myth of the Rational Voter. Princeton: Princeton University Press.

Caplan, Bryan, Crampton, Eric, Grove, Wayne A., and Somin, Ilya. 2013. “Systematically Biased Beliefs about Political Influence: Evidence from the Perceptions of Political Influence on Policy Outcomes Survey,” PS: Political Science and Politics 46: 760-7.

Chong, Dennis. 2013. “Degrees of Rationality in Politics,” in The Oxford Handbook of Political Psychology, ed. David O. Sears and Jack S. Levy, pp. 96-129. New York: Oxford University Press.

Cohen, Geoffrey. 2003. “Party over Policy: The Dominating Impact of Group Influence on Political Beliefs,”Journal of Personality and Social Psychology 85: 808-22

Converse, Philip. 1964. “The Nature of Belief Systems in Mass Publics.” In Ideology and Discontent, ed. D. E. Apter. London: Free Press of Glencoe, 1964.

Converse, Philip. 1990. “Popular Representation and the Distribution of Information” in Information and Democratic Processes, ed. John A. Ferejohn and James H. Kuklinski. Urbana: University of Illinois Press.

Converse, Philip, and Pierce, Richard. 1986. Political Representation in France. Cambridge MA: Harvard University Press.

Delli-Carpini, Michael X and Keeter, Scott. 1996. What Americans Know about Politics and Why It Matters. New Haven: Yale University Press.

Erison, Cengiz, Lodge, Milton, and Taber, Charles S. 2014. “Affective Contagion in Effortful Political Thinking,” Political Psychology 35 (2014): 187-206.

Friedman, Jeffrey. 2006. “Democratic Competence in Normative and Positive Theory: Neglected Implications of ‘The Nature of Belief Systems in Mass Publics’,” Critical Review 18: i-xliii.

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

Inglehart, Ronald, and Klingemann, Hans. 1976. “Party Identification, Ideological Preference, and the Left-Right Dimension among Western Mass Publics,” in Party Identification and Beyond, ed. Ian Budge, Ivor Crewe, and Dennis Fairlie. London: Wiley.

Iyengar, Shanto, Sood, Guarav, and Lelkes, Yphtach. 2012. “Affect, Not Ideology: A Social Identity Perspective on Polarization,” Public Opinion Quarterly 76: 405-431.

Kahneman, Daniel, Slovic, Paul, and Tversky, Amos, eds. 1982. Judgment under Uncertainty: Heuristics and Biases. New York: Cambridge University Press.

Kelly, James Terence. 2012. Framing Democracy. Princeton: Princeton University Press.

Kinder, Donald and Kalmoe, Nathan. 2017. Neither Liberal nor Conservative: Ideological Innocence in the American Public . Chicago: University of Chicago Press.

Landemore, Hélène. 2012. Democratic Reason. Princeton: Princeton University Press.

Landemore, Hélène. 2017. Inclusive Constitution-Making and Religious Rights: Lessons from the Icelandic Experiment,” Journal of Politics 79(3): 762-779.

Lewis-Beck, Michael, Jacoby, William, Norpoth, Helmut, and Weisberg, Herbert. 2008. The American Voter Revisited. Ann Arbor: University of Michigan Press.

Lodge, Milton, and Taber, Charles. 2013. The Rationalizing Voter. New York: Cambridge University Press.

McCann, James A. 1997. “Electoral Choices and Core Value Change: The 1992 Presidential Campaign,” American Journal of Political Science 41: 564-583.

Oppenheimer, Danny, and Edwards, Mike. 2012. Democracy Despite Itself. Cambridge, MA: MIT Press.

Rasinki, Kenneth A. 1989. “The Effect on Question Wording on Public Support for Government Spending,” Public Opinion Quarterly 53: 388-394.

Somin, Ilya. 2013. Democracy and Political Ignorance. Stanford: Stanford University Press.

Somin, Ilya. 2016. Democracy and Political Ignorance, 2nd Edition. Stanford: Stanford University Press.

Taber, Charles, and Lodge, Milton R. 2006. “Motivated Skepticism in the Evlauation of Political Beliefs,” American Journal of Political Science 50: 755-69.

Taber, Charles, and Young, Everett. 2013. “Political Information Processing,” in Huddy, Sears, and Levy 2013, pp. 525-558.

Tversky, Andrew, and Kahneman, Daniel. 1973. “Availability: A Heuristic for Judging Frequency and Probability,” Cognitive Psychology 5: 207–233.

van der Vossen, Bas, and Jason Brennan. 2018. In Defense of Openness. New York: Oxford University Press.

Westen, Drew. 2008. The Political Brain. New York: Perseus Books.

Westen, Drew, Pavel S. Blagov, Keith Harenski, Clint Kilts, and Stephan Hamann. 2006. “The neural basis of motivated reasoning: An fMRI study of emotional constraints on political judgment during the U.S. Presidential election of 2004.” The Journal of Cognitive Neuroscience 18: 1947-1958.

Zachmeister, Elizabeth. 2006. “What’s Left and Who’s Right? A Q-method Study of Individual and Contextual Influences on the Meaning of Ideological Labels,” Political Behavior 28: 151-173.

Zaller, John. 1992. The Nature and Origins of Mass Opinion. New York: Cambridge University Press.


[1] See Brennan 2018.

[2] Acemoglu and Robinson 2013.

[3] See Bas van der Vossen and I argue in In Defense of Openness (New York: Oxford University Press, 2018) for a review of the evidence. Democratic governments have implemented highly inefficient immigration and trade policies which on very conservative estimate cost the world at least $40 trillion a year, and probably more on the order of $100 trillion/year, and which are responsible for the continued misery of the developing world.

[4] Acemoglu and Robinson 2013.

[5]  See Oppenheimer and Edwards, Democracy Despite Itself (Cambridge, MA: MIT Press).

[6] Campbell et al 1960; Delli-Carpini & Keeter 1996; Converse 1964; Friedman 2006; Caplan 2007; Somin 2013; Brennan 2016.

[8] Bartels 1996; Althaus 2003; Caplan 2007; Gilens 2012; Caplan et al 2013.

[9] Tversky and Kahneman 1973, Kahneman, Slovic and Tversky 1982; Rasinki 1989; Bartels 2003; Arceneaux and Stein 2006; Lodge and Taber 2006; Westen et al 2006; Westen, 2008; Kelly 2012; Haidt, 2012; Chong 2013; Lodge and Taber 2013; Taber and Young 2013; Erison, Lodge, and Tabor 2014

[10] Gilens 2012; Achen and Bartels 2016; Brennan 2016.

[11] Althaus 2003; Gilens 2012; Caplan 2007.

[12] Converse 1964; Barnes 1971; Inglehart and Klingemann 1976; Arian and Shamir 1983; Converse and Pierce 1986; Zaller 1992; McCann 1997; Goren 2005; Zechmeister 2006; Lewis-Beck et al 2008; Achen and Bartels 2016; Kinder and Kalmoe 2017; Iyenger et al 2012.

[13] Achen and Bartels 2016.

[14] Landemore 2012. For a critique, see Brennan 2013 and Somin 2016.

[15] “Inclusive Constitution-Making and Religious Rights: Lessons from the Icelandic Experiment,” Journal of Politics 79 (3): 762-779, 2017.


Leave a reaction with this article