# The wisdom of the crowd (I): Can Democracy Trump Elites?

Reading time: 8 min.

----
Page title: Can democracy govern better than elites?

Meta Description: An enjoyable read investigating whether the wisdom of the crowd can create intelligent governments.

Keywords: decision-making; democracy; elite; diversity: ability; collective intelligence; wisdom of the crowd; cognitive science; complex systems

----

**Democracy vs. elites... an issue that has been part of the political debate since ancient times. From Plato's Republic to the structuring of the United Nations, to the workers' struggles at the beginning of the 20th century and May 1968. The question of how democracy can create good governments is maybe one of the problems with the greatest social and political consequences. Perhaps one of the most widely used arguments against democratic reason is that of the ignorance of the voters, but can a group of "average citizens" make intelligent decisions collectively?**

Image by i.imgut.com via https://i.imgur.com/9Gd255e.png

## An old problem rooted in the classics…

In book VI of Plato's Republic, the Greek philosopher addresses the question of who should govern the State by using what would happen with different decision systems as an analogy for choosing the captain of a ship. Namely, if the strongest of the sailors demands to be the master and imposes their law, we would be facing a Tyranny. However, not because of such condition will he know better the methods of navigation. If it is the richest who drives, we would have a Plutocracy. But neither would successful sailing be guaranteed because knowledge is a more useful tool than money when it comes to finding our way in an unfamiliar place. And if it were a popularly elected majority, that is, a Democracy, it would not be that ship that necessarily arrived at the best port either, because the masses may be full of ignorants. It is a government of wise people that would give the best results according to the Greek philosopher, a system known as Sophocracy. They are the ones who know the way and the techniques of navigation best. Today, this decision-making system is called Technocracy, the government of the technicians, the most skilled.

There has always been criticism around democracy, consistently deemed inefficient and slow. At the same time, it is not unusual to find criticism of some anti-democratic alternatives that, from time to time, have materialised in governments throughout history: dictatorships, plutocracies and technocracies. Yet, is democracy actually more inefficient than the technocratic alternative? We ask ourselves this question in the times of a super-specialized society, where skilful people are usually in charge.

## A Faulty Generalisation

A common critique of democracy is rooted in a simple psychological explanation called faulty generalisation. This fallacy consists in using one or more examples as a demonstration of a more general law. To illustrate faulty generalisation, we´ll use an everyday example. In a bar debate about whether democracy or elites are better at making collective decisions, who hasn't ended up listening to the examples of the doctor who operates better than the masses or the pilot who drives better than the crew?

In general, the well-defined problems of our daily lives come to mind more easily, we are more likely to think of them as concrete examples, and they are used more frequently to discuss whether or not democracy is a good system of decision-making. Therefore, everyday examples which usually favor technocracies present themselves more easily, and are used to generalise to other issues that might not be as well-explained by said system. For this very reason, the question of whether democracy is better than an elite decision-making system almost always [revolves around the kind of examples that favour technocratic decision-making](https://https://tomstafford.substack.com/p/democratic-reason).

**In general, the well-defined problems of our daily lives come to mind more easily, we are more likely to think of them as concrete examples, and they are used more frequently to discuss whether or not democracy is a good system of decision-making. Since well-defined problems are usually well known and well studied, they tend to favour a technocratic approach, which is based on technical knowledge. However, these concrete examples are often used to generalise to other issues that might not be as well-explained by said system. For this very reason, the question of whether democracy is better than an elite decision-making system very often [revolves around the kind of examples that favour technocratic decision-making](https://https://tomstafford.substack.com/p/democratic-reason).**

But what does science have to say on this matter? What do we know today about whether the decisions of a diverse group of people are better or worse than those of the elite?

## The Hong-Page theorem

While well-defined problems are best solved by specialists, there are problems for which we do not have a well-defined framework or whose complexity is so high that solutions are achieved through heuristic approaches. Think for example of problems with ethical dimensions where the interests of citizens are different, even opposed. For instance, the choice of a government, the rules that regulate the market or the measures to be taken in a totally unknown situation. It has recently been argued that in these cases a diverse mass is wiser than a group of experts. This is the conclusion reached in [2004 by Lu Hong and Scott Page](https://www.pnas.org/content/101/46/16385), where they used a theoretical model to demonstrate that a diverse group of individuals make better decisions than a non-diverse group of experts. For this to happen, however, it is necessary to have a sufficiently large group to have enough highly skilled individuals and enough individuals who think differently.

The conclusion of the article is that, when solving problems without a defined framework, individuals are valuable to the extent that they think differently, as this variety improves the collective decision. Therefore, an individual's ability to contribute to a solution is contextual, and it depends on the type of approach taken in a given problem. Thus, having contributions from individuals who think differently, even if they are not experts, can be crucial to improve collective decisions.

## Diversity is not the same as randomness

In a [paper published in 2014](http://www.ams.org/notices/201409/rnoti-p1024.pdf), Abigail Thompson opposes the Hong-Page theoretical proposal, considering that in their model it is the random process of group formation, and not diversity itself, which explains its success. Somehow, what Thompson asserted was that the model did not prove what it mathematically claimed to prove, but only illustrated the well-known fact that randomly formed groups are sometimes effective.

However, the disagreement continued, as [Daniel J. Singer (2018)](https://www.journals.uchicago.edu/doi/abs/10.1086/701074?af=R&mobileUi=0) defended that, in fact, the success of the groups selected by Hong & Page was due to diversity and not randomness. Now, it is necessary to understand that diversity is not the same as randomness: in these models, high diversity means that there is little overlap in the way individuals solve problems, while randomness simply means that individuals in a group are chosen at random. Indeed, it is possible, though unlikely, that a randomly selected group will have less diversity than an expert group. However, what Singer showed is that diverse groups generally perform better than random groups, and therefore he concluded that the groups Thompson used in her counter-argument did not represent the entirety of possible diverse groups. When one then compares the performance of diverse --and not simply random-- groups with expert groups, what Hong & Page initially claimed holds. That is, diversity, in general, trumps ability.

## But there's more...

The lessons of what agents do in simulated worlds are limited because of the complexities involved in real-world scenarios, with real people and countless practical and ethical constraints. Yet it may be useful to pay attention to the story that these models have to tell. While we should not consider the conclusions of these models to be entirely applicable to the real world, undervaluing them could lead to the devaluation of diversity as a source of wisdom. In an increasingly technological world, human cognitive diversity should not only be seen as a common heritage of humanity, but also as an excellent tool for finding solutions to problems for which we do not have a well-defined framework.

It should be noted, however, that the concept of democracy is ambiguous and not always linked to that of diversity. If we are talking about a system where all citizens contribute to the solution of a problem, then the ideas discussed here are valid and we can conclude that in general the diversity facilitated by democracy helps to make wise decisions. If, on the other hand, we are talking about a system where government representatives are appointed, then the group of elected rulers will not necessarily be diverse. In the latter case, there are also other problems, such as the so-called Arrow's impossibility theorem, which states that it is not possible to design a voting system with more than two alternatives that simultaneously fulfils a number of rational and democratic criteria. This problem, however, will be discussed in our **next blog post**. For now, we are left with the idea that, contrary to what everyday examples might suggest, when the framework of the problem is not well known, diversity can trump ability.

## Acknowledgment:

Thank you to Adoración Guzmán García (ENS-PSL, Université de Paris, EHESS) for her ideas and contributions.

## References:

Hong, L., & Page, S. E. (2004). Groups of diverse problem solvers can outperform groups of high-ability problem solvers. Proceedings of the National Academy of Sciences, 101(46), 16385-16389.

Singer, D. J. (2019). Diversity, not randomness, trumps ability. Philosophy of Science, 86(1), 178-191.

Thompson, A. (2014). Does diversity trump ability?. Notices of the AMS, 61(9), 1-24.

Stafford, T. "Democratic Reason". *Reasonable People*. Sep 1, 2020. https://tomstafford.substack.com/p/democratic-reason