Professor Frances Lee of the University of Maryland says that debates about district size go back to the dawn of the Republic, when anti-Federalists argued that even 30,000 people was too large a number for a single elected official to fairly represent. Her research with fellow political scientist Bruce Oppenheimer suggests that modern voters may feel the same way: In states with larger populations, voters tend to approve of their senators less and subject them to tougher re-election races.
Research by professor Brian Frederick of Bridgewater State College has shown that as the population size of Congressional districts rises, perceptions of the local House of Representative member worsen. Relative to people in populous districts, people in low-population districts are more likely to describe their House member as highly in touch with and very helpful to voters. Lightly populated districts are also characterized by higher approval levels of House members.
It’s not just voter perceptions of House members that change: Representatives from more-populous districts are more likely to take political stances that deviate from what their typical constituent believes. Frederick describes all these effects of larger district population as mutually reinforcing. “In big districts it is harder for House Members to have as much personal contact with voters and to keep track of what they are thinking," he says. "This makes them more likely to drift away from their constituents, which further fuels the sense back home that they are out of touch.”
As constituency size seems to influence quality of representative government, trying to equalize population across districts — a goal endorsed by the U.S. Supreme Court and governments in other nations — has a logical basis. Frederick would also like the House of Representatives to expand, arguing that “it’s more efficient to freeze it at 435 members, but it would be more democratic to grow it by 250 seats.”
Keith Humphreys is a professor of psychiatry and behavioral sciences at Stanford University.