Differences among candidates are often more form than substance. It seems that all favor a strong national defense, adequate welfare for the truly needy, a balanced budget, and low taxes. Positions soon become as predictable as calls to support motherhood, apple pie, and the flag. Why do party platforms and political speeches so often seem like photocopies of one another?

The median voter model suggests that the median voter must be captured to achieve a majority vote, and it helps explain why political parties and candidates so often seem interchangeable.

The median voter model partially explains similarities among political parties and candidates. In Figure 1, we assume that the preferences and voting patterns of individuals can be ranked very simply and are normally distributed along a continuum from the extreme left (revolutionary communism, perhaps) to the extreme right (fascism?). Point M is in the precise center of this spectrum and identifies the median voter—exactly half of all voters lie to the left, with the other half being to the right of this position.

 

Figure 1  The Median Voter Model

 

 

Consider two candidates running for an office. The median voter can provide the margin of victory if the winner must receive 50% plus 1 votes. Thus, the candidate who adopts positions on the issues closest to point M captures the median vote and wins the election. Regardless of their initial positions (say, L and R, respectively), both candidates find that shifting their positions slightly toward point M boosts their standings in public opinion polls. Figure 1 also uses a slightly right-wing candidate to illustrate how movements toward the extremes lose votes, while shifts to the center gain votes. The same would be true of a left-wing candidate. As the candidates each try to maximize their expected votes, each creep toward point M, and their positions ultimately may be almost indistinguishable.

 

You might think a new third candidate could win this race by entering at, say, point L, leaving the original candidates to split the votes from halfway between points L and M all the way to the extreme right. But this third candidate would also find that moving toward point M increases support by voters. This suggests that, ultimately, virtually all serious candidates for election gravitate toward point M. Similar forces are at work in elections ranging from that for president of a local PTA to the U.S. presidency, pressuring not only political candidates, but political parties as well, to gradually adopt middle-of-the-road policies.

 

The median voter model grossly oversimplifies the dynamic world of politics; otherwise, there would be no differences between candidates or party platforms. Among other reasons, differences exist because:

1.       The opinions of voters on various issues never fall along a simple array from left to right, and opinions on specific issues may be only loosely correlated. For example, there may be no predictable connections between people’s opinions on welfare, the environment, international trade, and law-and-order issues.

2.       Voter distributions change, so it is impossible for politicians to know exactly where the median voter is on any controversial issue at any point in time.

3.       Some politicians do have strong personal beliefs about some issues.

 

Low voter turnout and rational political ignorance also weaken the predictive power of the median voter model. If most eligible voters apathetically stay away from the polls, then well-informed activists who belong to narrow special-interest groups may cast deciding votes in an election. In spite of these qualifications, the median voter model helps explain middle-of-the-road clustering by politicians and parties, which ignores the preferences of vast numbers of less moderate voters. It also hints at reasons for the dominance of the two-party system.