My print column examines efforts by political scientists to estimate the relative power
of voters to influence presidential elections, depending on what state
they live in.
John Banzhaf, a law professor at George Washington University Law
School, began studying the question of voter power in response to a
Supreme Court ruling, a half century ago, in Baker v. Carr.
A divided high court ruled that federal courts may intervene to address
districting that isn’t aligned with population. The ruling led to the
question: Just how should districts’ voting power be allocated such that
each individual voter has the same power?
His answer, building on earlier work by British mathematician Lionel
Penrose, is that a fair allocation would be proportional to the square
root of districts’ populations, rather than directly proportional to the
populations. Banzhaf didn’t advocate weighting votes as such, but
instead intended to point out that weighting voter power isn’t all that
Part of the reason is that truly testing which voters have the best
chance of swinging an election isn’t realistic. Tied races are very
rare, too rare to say how likely they are in any given election, let
alone a national one that occurs only every four years, giving little
opportunity for testing.