Every few years, in offices that rarely attract cameras, officials sit down with census tables and maps and decide where one constituency ends and the next begins. The work is technical, slow and almost entirely invisible to voters. It is also among the most consequential acts in democratic politics.

The quiet power of a line

A boundary determines who votes alongside whom. Draw it one way and a community with shared interests forms a decisive bloc; draw it another and the same community is split across three seats in which it is always outnumbered. Neither map is neutral, because there is no neutral map — only choices about which communities to keep together and which to divide.

The practice of manipulating these choices for partisan advantage has a famous name, gerrymandering, after an oddly shaped district approved by a Massachusetts governor in 1812. Its techniques are simple to describe: concentrate your opponents' voters into a few seats they win overwhelmingly, and spread the rest thinly across seats they narrowly lose.

Who should hold the pen?

Democracies have tried different answers. Some leave the task to legislators, which invites the obvious conflict of interest. Others hand it to independent commissions bound by rules about population equality, compactness and respect for existing administrative borders. Increasingly, mathematicians and computer scientists are being asked to measure fairness — to generate thousands of possible maps and ask whether the one actually chosen is an outlier.

No method removes judgement entirely. But the question of who draws the lines deserves far more public attention than it receives. By the time an election campaign begins, a great deal of its outcome has already been decided by people most voters will never hear of.