Contra Ira Katznelson, it is not democracy that is in crisis today but representative government. The solution, however, is not to tinker around the edges of representative government—it is to rethink democracy. Equipped with a clarified understanding of democracy in today’s world, we can imagine new institutions that would restore the democratic properties that Katznelson rightly celebrates: fairness and good policy.

This rethinking, as I see it, must question three central features of contemporary representative democracies: 1) the role of elections as the privileged selection mechanism for democratic representatives; 2) the idea that democracy is chiefly located at the level of the nation-state; and 3) the idea that democracy is only a matter of political institutions.

Elections are a historically contingent feature of democracy. The Greeks did not have them. At the rebirth of democracy, in the eighteenth century, the choice was between election, widely seen as an aristocratic principle, and sortition, the democratic principle inherited from ancient Athens whereby public officials were selected by lot from a pool of eligible candidates. The principle of election triumphed and with it a version of democracy, representative government, that is in fact a mixed regime, with an oligarchic tendency deeply entrenched in its DNA. Is it really surprising that 200 years later elections time and again produce a class of politicians with homogeneous socioeconomic profiles, very low turnover rates, and a propensity to support laws and policies disconnected from what the majority actually wants?

Campaign finance reform and voter education are not enough.

According to Katznelson, via Forster, we should “cheer” democracy for allowing “variety” and “criticism,” two features he suggests helped twentieth-century democracies outlast and overcome Nazism, socialism, and fascism. Variety helps to bring about better policies than those fostered by monolithic ways of thinking. Similarly, criticism is a source of learning available to democracies but not to regimes that silence opposition. Yet if contemporary representative democracies routinely empower homogeneous elites who are relatively well insulated from criticism, then our model of representative government fails to deliver on two of democracy’s cardinal virtues.

In view of these shortfalls, we need to consider new selection methods. Random selection has the merit of preserving the actual variety of the larger group within the smaller one, as well as ensuring genuine rotation of offices. It is already successfully practiced in the selection of citizens’ assemblies, deliberative polls, citizens’ juries, and other bodies appointed by lotteries. Self-selection is another option, successfully practiced in offline and online participatory democratic schemes, such as participatory budgeting and crowdsourcing experiments that have flourished around the world since at least the late 1980s. These selection mechanisms blur the line between direct and representative democracy by bringing to power ordinary citizens. They invite us to consider models of delegation, authorization, and representation beyond the current electoral one.

Another important cause of our problems is the obsessive focus on national sovereignty, which was on full display in the recent standoff between Greece and Germany that nearly tore apart the European Union. In the globalized age, when multinational corporations dwarf many nation-states in wealth and power, democratic citizens and their governments should come to terms with what Thomas Piketty calls “the myth of national sovereignty” and start demanding and implementing democracy at the inter- and supranational levels, where real power can be regained. The site of democratic claims should thus move upward from national governmental spheres to global organizations and institutions.

Finally, a related limitation is the idea that democracy is strictly a matter of political institutions. As Colin Crouch argues, we enjoy the hallmarks of democratic governance—regular free and fair elections, a free press, and universal suffrage—yet are moving toward “post-democratic” societies, devoid of meaningful democratic content. A more demanding notion of democracy would expand laterally to a greater swath of society. One site in particular seems urgently in need of democratic reinvention: the workplace.

As Katznelson concludes, democratic self-doubt should not turn into self-hatred. He is probably right, also, that extreme populist or technocratic alternatives on offer present the same totalitarian risks Nazism, socialism, and fascism did in the 1930s. What we should acknowledge, however, is that modest tweaking—with talk of campaign finance reform and voter education—will not do. This is not sufficient to overcome the deficits of representation. Our conception of democracy needs to be expanded and reshaped.