Political scientists know a lot about how voters behave, but much less about the arguably more important question of how politicians think voters behave. As the eminent political scientist V.O. Key Jr. wrote half a century ago, “If politicians see voters as most certainly responsive to nonsense, they will give them nonsense. If they see voters as susceptible to delusion they will delude them. If they see an electorate receptive to the cold, hard realities, they will give it the cold, hard realities.”

David Broockman has a new paper exploring the views of nearly one thousand incumbent state legislators regarding electoral accountability. Broockman’s findings suggest that these elected officials see a lot of scope for nonsense and delusion in their relationships with their constituents.

For example, only 15 percent of state legislators agreed that voters “usually know who in government to blame” when they “don’t like a particular public policy.” If voters don’t know who to blame, why not shut down the government?

Are you surprised that elected officials are ideologically “polarized”? Only 42 percent of state legislators said that “Moderate candidates and politicians win significantly more votes.” Many expressed doubt that most voters “decide who to vote for based on the issues” (38 percent in general elections, 46 percent in primaries).

Do politicians see the electorate as “a rational god of vengeance and of reward,” as Key colorfully put it? Not so much. Almost 60 percent agreed that “voters usually base their choices on only very recent events,” and 35 percent agreed that voters “sometimes decide whether to vote for incumbents based on things completely unrelated to politics, like whether their favorite football team recently won a game.”

In each of these respects, the politicians were considerably less impressed by the constraints of electoral accountability — and, I would argue, considerably more realistic in their assessments — than their constituents were. For example, 47 percent of the public claimed that they “usually know who in government to blame” when they don’t like a particular public policy. Overwhelming majorities claimed that they decide who to vote for based on the issues (94 percent in general elections and 90 percent in primaries), and only 4 percent admitted to voting on the basis of “things completely unrelated to politics.”

One of the few instances in which legislators and constituents agreed was in finding fault with the news media. Only 51 percent of state legislators and 44 percent of constituents agreed that the media “pay close attention to whether elected officials are serving the public interest.”

The differences Broockman found between legislators and constituents may be partly attributable to differences in question wording. Citizens were asked about their own behavior, and probably thought about national and statewide elections, whereas state legislators were generally asked about “most voters” in their own local races. Nevertheless, it seems clear that these incumbent politicians have a much more jaded view of the electorate than their constituents do.

That is a recipe for unaccountable government and a frustrated citizenry.