John Huber
Professor of Political Science

It’s a great privilege to be able to stand here and congratulate all of you on earning your PhD from Columbia. You may not actually realize how special you’ve become by obtaining you doctorates, and in the spirit of the careful scholarly research that has brought you to this room, I decided to quantify your specialness. A reasonable estimate is that there are 1 million people in the US who have PhDs from good research universities, which means that among people aged 25 and over, only 1 out of every 2,000 is as educated as you are. One in 2,000. That’s very special, and the faculty are very proud of your accomplishments.

Before I talk about democracy and majority rule, I thought I’d offer you some practical advice on how you should live in the months ahead, now that you are special. The advice is simple: “Live poor.” My first job as an assistant professor resulted in a 400% pay increase over my grad stipend. I imagined I was rich and I bought a new car, my first real bed, a couch, and breakable dishes, among other things. The 400% was not nearly enough to cover my splurge, so over the following months, life was no fun. A night out meant a trip to the public library, and my diet consisted of oatmeal and pasta. I urge you all to avoid my mistake, and to transition slowly out of your “living poor” habits.

Now, about democracy and majority rule. Democracy is a hot topic in the world and in political science, as many countries are creating new democracies, and many existing democracies, including the US, wrestle with how to fine tune the functioning of our democratic institutions. In the next few minutes, I want to describe a small corner of the political science research agenda on democratic processes.

For me, an interesting starting point is to consider abstract mathematical models of majority rule. These models, which date from the 1970s, undermine perhaps the most common normative justification for democracy, which is that democracy is desirable because it produces outcomes that represent the will of the people, or the will of the majority.

To convey the intuition from these models, consider a three-person democracy—an historian, an economist and a sociologist—that must decide by majority rule how to divide one dollar. Each person wants as much of the dollar as possible, and each is allowed to make proposals about how to divide up the dollar among the three people. The proposals are then subjected to majority voting.

Suppose the beginning distribution of the dollar is rather egalitarian: it gives the historian 34 cents and the other two 33 cents each. Is there a majority that prefers some other division of the dollar? Clearly there is. The economist, for example, could propose to give the sociologist 1 penny more, give the historian nothing, and keep 64 cents for himself. This makes two voters better off and would defeat the egalitarian proposal. But could this new division, which gives the sociologist 34 cents and the economist 66 cents, be defeated? Of course if could. The sociologist could propose to keep 99 cents, and to give the historian a single penny, which makes them both better off, and which would thus defeat the previous division by majority rule. But could this be defeated?

Of course it could, and the point of the models is that the process could continue forever because for any distribution of the dollar among the voters, there exists some other distribution that makes a different majority of voters better off. Think about it! Think of any distribution of the dollar to the three voters. You can now think of another one that makes two of them better off.

We care about this for two reasons. First, the models teach us that we cannot justify democracy based on a belief in some abstract notion of majority will. There simply is no such will. For any outcome that emerges from a democratic process, there always exists some other outcome that is preferred by some other majority.

Second, we care because the models underscore that the outcomes of any democratic process are influenced as much by the institutional arrangements for making choices as they are by the preferences and tastes of individual voters. In the example with the three PhDs, an institutional prerogative to make the last policy proposal would obviously confer great advantage. Or to take a real example that is close to many of our hearts, if the framers of the American constitution had prescribed direct election of the American president rather than election by an electoral college, we might reasonably expect that world history over the past 6 years might have unfolded quite differently.

Of course, we do not know if Al Gore would have won had there been direct election. With direct election, an extra vote in Florida would have been no more valuable than an extra vote in Kansas or California. Under these different rules, candidates would have been forced to wage different campaigns than they did under the Electoral College. We never observed these campaigns, so we will never know what would have happened if the rules had been different.

A central preoccupation in political science has been to address questions about how the nuts and bolts of democracy—the institutional arrangements that govern decision-making processes—affect political behavior and outcomes. Seemingly small details, such as the presence or absence of an electoral college, shape strategic incentives of political actors, thereby shaping who wins and loses the battle to shape public policy.

The study of electoral rules provides one example of how we study these institutional details. Some electoral rules encourage two-party systems, such as we have in the US, with the Democrats and the Republicans. Other electoral rules encourage multiparty systems, as in the proportional representation systems of many European countries. Which electoral rules are better?

Political scientists would argue that the question, posed so broadly, makes little sense. Instead, the right way to pose the question concerns tradeoffs. We believe, for example, that relative to multi-party systems, two-party systems generate higher rates of economic growth, greater accountability by politicians, and higher levels of political stability. At the same time, however, we believe two-party systems produce greater inequality, lower citizen satisfaction with their institutions, and weaker representation of minority viewpoints. So which is better? It depends on how you value, for example, economic growth versus inequality, or political stability versus minority misrepresentation.

The point of this whirlwind tour is not to argue about the advantages of particular forms of democratic institutions. Instead, it is to underline the fact that democracy, as a concept, is rather vague and amorphous. Democracies emerge from institutional arrangements that shape incentives, and that thereby influence different aspects of democratic performance. Thus, the choice of any institutional form over another involves tradeoffs. Our goal as political scientists is to obtain a relatively clear understanding of these tradeoffs.

To conclude, I would like to describe an unfortunate side-effect that sometimes emerges from the process of studying these tradeoffs. Political science has responded to the complexity of this task by doing exactly what you would expect. The best scholarship breaks down the questions about institutions into small, manageable pieces. We offer arguments about things like electoral rules or agenda control. We gather evidence in an attempt to support our arguments. There then occur on-going efforts to discredit existing arguments and replace them with better ones. Those arguments that appear least wrong survive for a while, and we try to build on them.

The unfortunate side-effect of this standard scholarly process is that it can condition us to be reticent about asserting ourselves in debates that unfold outside the university. If successful scholarship emerges from posing narrow and manageable questions, it’s not hard to convince ourselves that we have little useful to say about the big problems in the world, which are not narrow or manageable at all. And if the scholarly process rests on a foundation of intrinsic skepticism about the ultimate validity of any argument, it’s not hard to convince ourselves that our thoughts on the big problems are probably sufficiently wrong so as to be of little value. Too often, our response is to remain cloistered within the academic community, comforted by the fact that our scholarly reputations do not depend on any direct relevance of our work to the outside world.

This is too bad. When we, as scholars, back away from hard problems in the world because we don’t trust our answers, or because we see only complexity, we cede the debate to individuals with weaker tools for thinking clearly and analytically about the problems. That makes little sense, and it need not be so. In earning your PhDs from Columbia, you not only have distinguished yourselves as scholarly experts, you have earned a credential and a license that enables you to assert yourself in the myriad important issues that arise outside of our campuses. I wish you all the courage needed to suspend any of your own disbelief about the value of what you have to say, and to get into people’s faces, sending a loud signal about which tradeoffs are the right ones for our societies to make.