On Cognitive Democracy and New Technology


In many ways, the work that we do here at the Sunlight Foundation is built on the premise that democracy functions best when citizens can get good information about what their government is doing. Widespread transparency makes citizens better and more active participants and makes politicians more accountable.

We’re also bullish on the ability of technology to facilitate this. The revolution in web and mobile communications have made it easier than ever to keep tabs on what government is doing, and we’re trying to make it even easier to do so.

So it’s very encouraging to read a new essay by Henry Farrell and Cosma Shalizi on “Cognitive Democracy.” They’re bullish, too, on what they call “novel forms of collective cognition that are facilitated by new media.”

They think democracy is a superior problem-solving institution (as compared to markets or hierarchy) because it is most capable of aggregating diverse perspectives. It also has cognitive benefits (in that it improves how we think generally). The age of the Internet holds great promise for unlocking the full potential of democracy as a collective decision-making institution.

Farrell and Shalizi start with the premise that complex problems are hard to solve, and “Individual agents have limited cognitive abilities, and (usually) limited knowledge of the landscape.” Individual people are prone to make bad choices because they don’t know any better.

Research on problem-solving shows that the institutions tend to come to the best decisions are institutions that bring together diverse perspectives on a relatively equal footing. Democracies tend to do this better than markets or hierarchies.

Farrell and Shalizi let Friedrich Hayek make the case for markets as superior problem-solving institutions. Hayek’s argument is that, given the complexity of the world, no one person can ever know enough to make the right decision for other people, and hence government planning will always err. Local and tacit knowledge that cannot be meaningfully centralized. Only decentralized individual decision-making can adequately aggregate this dispersed intelligence.

Farrell and Shalizi see limits in the fact that market purchases are incredibly blunt signals. They don’t allow for exchanges of ideas, only of money. Markets also tend to lead to inequalities, and unequal resources impair collective decision-making.  “In Hayek’s markets, people communicate only through prices,” they write. “But there are many useful forms of knowledge that cannot readily be converted in this way.”

Next up: Hierarchy. Here Farrell and Shalizi turn to Richard Thaler and Cass Sunstein, authors of the influential book Nudge, to make the best case for a modified version of hierarchy, “libertarian paternalism.”  Their claim is that “choice architects” (i.e. enlightened bureaucrats who are experts in human psychology and behavior) can help individuals make the best decisions by properly setting the defaults.

But Farrell and Shalizi see problems with the top-down approach. The choice architects, they write, “will have difficulty in eliciting feedback, even if they want to.”

So onto democracy. “Since, as we’ve argued, power asymmetries inhibit problem-solving, democracy has a large advantage over both markets and technocratic hierarchy,” Farrell and Shalizi write. Democracy, at least in its ideal form, promotes equality of power. Democracy promotes debate. Democracy can bring together individuals with high diverse viewpoints. Debate and deliberation, they argue, (as long as it is competitive) forces people to improve and strengthen their arguments.

Yes, in theory. But in practice? The current partisan gridlock in American politics hardly seems to be a model of effective problem solving.

I suppose Farrell and Shalizi would respond, as they do in their essay, “we have no reason to think that actually-existing democratic structures are as good as they could be, or even close.”

And so here, to me, is where things get exciting. “We do not yet know the possibilities of Internet-mediated communication for gathering dispersed knowledge, for generating new knowledge, for complex problem-solving, or for collective decision-making,” The write. “But we really ought to find out.”

Amen. We really ought to.

Here’s my take:

From a historical perspective, Both Athenian and American democracy began at a scale at which “cognitive democracy” was possible. The polis was small enough that individual citizens could engage with each other and meaningfully exchange knowledge, and power was widely enough dispersed to facilitate optimal problem-solving. But diversity was lacking. This was largely democracy for rich white males.

Over the years, diversity has obviously increased, but so has scale, in a way that has impeded the ebb and flow of exchange and debate. Government has become increasingly distant and abstract for most individuals, and politics has (perhaps by necessity) become more performative and less interactive. Money has become far more important, and the growth of professional lobbying alongside it has undermined the equality of power. So the modern era has not exactly been a golden time for the problem-solving potential of democracy. No wonder that the allure of markets and hierarchies to solve complex social problems has grown, and that the “government is the problem” meme has caught on.

But, as Farrell and Shalizi explore, technology has the power to change that.

New technology facilitates collaborative decision-making in a way that has not been possible before in a modern-scale democracy, through wikis and other knowledge aggregation possibilities.

We at Sunlight have advocated for public markups and started our own website to do this.  I’ve also personally argued (in a Brookings paper) that all lobbying (both professional and constituent) should be done through an online clearinghouse in which everybody can productively engaged with everybody else in an open and transparent way.

If Farrell and Shalizi are correct in their assertion that institutions that maximize relative equality and diversity viewpoint are best able to solve problems (And my guess is that they are), the potential for an Internet-enabled democracy is tremendous. This begs the question of how we can get there. We think that more transparency and more public data would be a good start.