Tag Archives: ancient Athens

Is Democracy the “Collective Capacity”?

Stanford classicist and democratic theorist, Josiah Ober, looks to ancient Greek etymology to redefine our idea of democracy. The ancient Greeks invented the word, after all, so the root might illuminate some of our problems getting a good, theoretical grasp on it.

Twentieth century political science and philosophy seemed to settle the meaning of democracy as “aggregating preferences”. That just means that we vote for the choice of law, candidate for office, or party platform, and the choice that receives the most votes wins. And political scientists have tinkered with the formula for counting votes, so that more voters can get what they want out of voting. This aggregative model stems from our understanding of the word “democracy” as meaning the “rule of the many”.

This account of democracy is the result of importing the assumptions and methodologies of economics into political science. The assumptions of 20th century economics are that human beings seek to maximize their own best interest and will accurately make the trade-offs that get them the best deal. The assumptions are as mistaken in politics as they are in economics. Preferences are not simple entities, but psychologically complex constructions that react counterintuitively to all sorts of little triggers. In any case, the aggregative model of democracy results in several logical puzzles (such as Arrow’s impossibility theorem, a story for another time).

This vote-counting notion of democracy is a very thin account of its theoretically richer potentials, though. Ober has been developing a new and comprehensive theory of democracy based on the ancient democracy of Athens. In “The Original Meaning of Democracy”, he explores the root suffixes of political terms in ancient Greek. In particular, he compares the suffix, -kratos, from which we get the word “democracy”, and –arche­, from which we get words like “monarchy” and “oligarchy”. Why are some forms of government “-cratic” and some “-archic”?

Ober suggests that the ancient Greeks used –arche to denote a “monopoly of office” and used –kratos­ to denote the form of power. So a monarchy describes a government in which one man controls the public offices of the city, while anarchy tells us that no one holds any public office. The suffix –arche describes the number of people who hold offices.

What we do not have is an ancient Greek term for “rule of the many”, which would be something like “polyarchy” (polyarchia). Robert Dahl actually coined the term in the mid-twentieth century as a description for the government of the United States. In his view, the “many” that ruled were the many, organized special interests that competed for advantage. (In those days, believe it or not, organized special interests weren’t just corporations, but civic organizations, labor unions, and other organizations of the popular classes.)

Meanwhile, the suffix –kratos means “power”. But “power” alone is ambiguous. Looking at some modern terms is not helpful. For example, we take the term “aristocracy” to be a class of landed wealth, when in the literal Greek the term would mean “power of the excellent”.

Instead, by analogy with other terms, Ober suggests we should understand –kratos­ as a “capacity to do”. For example, isokratia (or “isocracy”) meant to the ancients an “equal capacity to public action”. The Athenians were fond of the prefix iso-. They referred to their government as an isonomia (“law of equals”) and guaranteed the right of isegoria (“equal access to speak in public fora”). The founder of Athens’ democracy, Solon, even referred to the new order as isomoiria, or the “equality of shares”.

In any case, let’s return to democracy. Given that the demos were the people as one body, democracy would instead mean “the power of the people”. The people as a whole have the capacity to act in the world.

Thus, democracy does not mean “rule of the many”, but “the capacity of the people to act”. Ober’s suggested definition does relax the focus of contemporary political science on counting up votes. What this definition does not answer is, how do a people become a people? Granted that, once united, they have the capacity to act, how do they unite in the first place?

In some sense, a group of individuals becomes one people through shared goals and established practices that add up to their social institutions. We want, or should want, those institutions to be democratic. But if the only criteria is that the institutions give that group the ability to act in concert, then we still don’t know what those institutions should be like. Any set of institutions will give a group the power to act collectively, whether a republic or a dictatorship.

Our difficulties with democracy might have been avoided if, instead of picking up “democracy” from Aristotle, we took the term “isonomy”. The “law of equals” leaves little ambiguity as to how we should understand “isonomic” social institutions.


Josiah Ober, “The Original Meaning of Democracy

Picking Legislators at Random Would Improve Congress (Part II)

Previously on Democracy in Principle…

Earlier this week, we looked at why elected legislatures pass so much legislation that only serve the personal interests of the legislators. The Italian researchers PluchinoGarofaloRapisardaSpagano, and Caserta showed in a mathematical model that an elected legislature is likely to pass a great quantity of legislation, because its members are bound by party affiliations. But that legislation will be unlikely to serve the public interest.

Perhaps the solution is to select legislators by lottery, as was the ancient Athenians did with their officials. The Pluchino et al. model though shows us that a legislature whose members are selected by lot will only be a little more efficient.

Instead, the researchers find that mixing legislators selected by lot with elected legislators can greatly improve the likelihood that the legislature will produce legislation in the public interest. For the purposes of discussion, we will call this mixed system the lottery-added system.

How Does This Work in Practice?

To implement such a system, we must have a method of election and a method of random selection.

If our legislature has a fixed size (suppose 500 members), then we have a problem with traditional electoral systems. Our current electoral system has a fixed number of districts from which voters choose legislators, one for each district. But under the lottery-added system, the proportion of elected legislators varies with the size of the majority. Thus, we could not even redraw the electoral districts before the next election, because we would not know ahead of time how many districts we would need.

Instead, we should adopt a system that elects a legislature “at large”. That is, we can treat the whole country as an electoral district. Voters would not vote for individual legislators, but would vote for political parties. Each party then receives the number of seats in the legislature proportional to the number of votes received. For example, if the Democratic Party receives 51% of the vote, then they receive 51% of the seats in the assembly. Each party then fills seats with legislators from a party list. This is a system of proportional representation called the “party list” system. In fact, the party list system is the most widely used system of proportional representation.

Then we alter the party-list method by the addition of some number of independent legislators. To optimize the efficiency of the legislature, we add a number of legislators selected by lot according to how large the majority is. The larger the majority, the greater the proportion of legislators selected by lot. (Mathematically minded readers can find the formula in Pluchino et al.)

Thus, if the Democratic Party has 51% of the vote, then the national election commission selects 20 people at random to serve in the legislature. If the assembly is composed of 500 members, then 20 would be selected by lot, 249 (or 51% of 480) would be Democrats and 245 would be Republicans. If the Democrats have 80% of the vote, then 280 of the seats will be filled by lottery, 176 by Democrats, and 44 by Republicans.

Pluchino et al. recommends that legislators selected by lot be limited to one term. This prevents them from forming a new party and eliminating the efficiency of the legislature.

Where do we get the independent legislators? In experimental public decision-making by randomly selected citizens, those chosen to participate have the right to refuse. This reduces the randomness of the selection process. Disadvantaged segments of the population – the working poor, the disabled, women, and others who bear significant social burdens – may be less likely to choose to serve.

We would have to make legislative service easier and enticing. The election commission will have to assure randomly selected citizens that their job or their business will be there at the end of their legislative service. Lottery-added legislators will require guaranteed benefits, such as free housing and good pay, for their term of service.

We also may reduce the term of service of each independent legislator to half or a quarter of that of elected members. In the United States, that means that while an elected Congressperson would serve two years, a randomly selected representative would serve only one year, and replaced for the remainder of the congressional term.

Questions about the Lottery-Added System

We elect our legislature because we believe that this will make our legislatures representative of the public. (We’re wrong about that though, but that’s for another post.) Since the development of political parties, we call a legislature representative when the distribution of parties in the legislature reflects the distribution of parties in the voting population. In most countries, this means the adoption of some form of proportional representation. The traditional plurality elections that we have in the United States, where the candidate with the most votes wins, does not produce such representation.

The lottery-added system, however, dilutes this representativeness by including legislators who do not represent any party. In fact, that is their virtue. Our traditional idea of representation assumes that election, with the incentive for reelection, is enough to get legislators to represent the public. But elected legislators can hide their motivations from their voters, and pursue their personal interests behind the closed doors of the backroom.

Randomly selected legislators each have diverse and specific personal interests. Theoretically, they do not have personal interests in common the way party-bound elected legislators do. This is why they produce better legislation than elected legislators do. Randomly selected legislators are collectively incorruptible. Thus, their presence may make the legislature more representative. They force the elected legislators to pursue the public interest voters elected them to represent.

More recently, we also want legislatures to represent the public in various demographic ways. We want legislatures to look more like the population. If 50% of the US population is women, then 50% of the House of Representatives should also be women. If 13% of the US is African-American, then 13% of the House should be African-American. We can’t rely on the party bosses to put forward more women and minorities in elections. But the random selection of legislators would provide opportunities to increase the demographic diversity and representativeness of legislatures.

I will leave aside whether the Pluchino et al. model of a legislature is accurate. There are some questions about it. For example, randomly selected legislators may be partisans; they pass over the definition of “public interest”, leaving it vague. The model is plausible, and the resulting solution of a lottery-added system is interesting.

The idea of using lotteries to elect public officials (often called sortition) continues to be regarded with surprise or even horror by elite and public alike. However, philosophers identified the method with democracy until the eighteenth century, because lotteries provided every citizen the opportunity to govern. If we are to achieve democracy in the 21st century, we should resurrect the idea that every citizen can govern.


Pluchino,C. Garofalo,A. RapisardaS. SpaganoM. Caserta. “Accidental Politicians: How Randomly Selected Legislators Can Improve Parliament Efficiency.” Physics A 390 (2011): 3944-3954. Web. 7 March 2011.

Picking Legislators at Random Would Improve Congress (Part I)

Nobody would call Congress efficient. It can’t pass anything. The government shut down just last year because Congress wouldn’t exceed an artificial borrowing limit. The legislation Congress does pass is rarely beneficial to the public. We all know this: Gallup polls show that only 14% of us approve of Congress’ ability. Unfortunately, people identify this with who is in Congress. The real problem is how people get into Congress: private campaign contributions, party approval, manipulative advertising. Is this how democracy is supposed to work? There must be a better way.

Unlike modern democracies, classical and medieval republics used lotteries to select officers. The archetype is classical Athens, whose lottery-based democracy flourished until its military destruction at the hands of Macedon. In fact, until the eighteenth century, political philosophers followed Aristotle in considering lotteries a defining feature of democracy. By contrast, Aristotle and his successors considered elections oligarchical, encouraging only the wealthy few to rule.

Contemporary lottery advocates, including John Burnheim, Brian Martin, Josiah Ober, and the Kleroterians at Equality by Lot, claim that we should revive this concept of democracy. I am sympathetic to this viewpoint. But is there any evidence that such a body would give the people more of what we want?

Italian researchers from the University of Catania (Pluchino, Garofalo, Rapisarda, Spagano, and Caserta, 2011) constructed a computational model to see how randomly selected legislators would perform compared to elected, partisan legislators. read more »

5 Things from Ancient Athens that We Need (Besides Democracy)

Demosthenes, Athens’ Great Orator (it’s a marble statue so you know it’s time to talk about ancient Greece)

One might expect that the staff here at Democracy in Principle have an enthusiasm for the ancient democracy of Athens.  Athens was undoubtedly the most democratic civilized society in history.  Modern people are typically appalled at its exclusion of women and its slavery, and rightly so.  I just hope that the future does not judge us so harshly, though, for our failures – to bring about the democratic inclusion we claim to believe in, or to accept our own very recent history of slavery.

Typically we say that the ancients have nothing to teach us.  We suppose that their cultural and historical conditions were so different from our own that they and we have nothing in common.  In particular, the city of Athens was small by our standards, barely maybe a tenth the size of my own city of Richmond.  But the issue of size is only relevant to the central organization of the Athenian state, the Assembly of its citizens.  Its other state institutions and practices – its lotteried administration, its popular jury courts, its Council – are not limited by the need for the population to gather, and thus not limited by a larger population.  Today, let’s look at some of the lesser practices and principles that the Athenians observed, and how we might apply them. read more »

The Democracy of Ancient Athens

It is commonly claimed that the story of democracy begins in ancient Athens, but this is pretty clearly false.  Generalizing from technically simpler cultures still existing or having been recorded before their destruction by Europeans, democracy in various direct forms was probably pretty widespread in early human history.  The turn towards authoritarianism in human societies has been speculated to have been the result of sudden shifts in the material basis of that society.  The ways and means of getting what is needed to live were rapidly eroded, perhaps due to various changes in local ecosystems.  But I’m no anthropologist.  What is important about Athens is that it is one of the best recorded democracies available, and one of the most extensive.  The ancient polis is the paradigmatic democracy. read more »