Incumbents in the U.S. House of Representatives are overwhelmingly re-elected. The explanation is simple. Most are gerrymandered into safe districts. But these graphics show that U.S. Senate incumbents are re-elected with about the same frequency. Senate seats are not gerrymandered. What is going on?
A related data point involves the widely cited Congressional "approval ratings." These have apparently gone from 73% in 1958 to 19% now. Why do commentators even cite these in light of the incumbent re-election frequencies?
There is a fascinating discussion of voter ignorance (rational, irrational, or any other kind) at Cato Unbound. Most people are basically sensible, but the opportunity costs of further investigation rise quickly while the marginal benefits are small, usually limited to the joys of team membership. The political parties understand the game and realize that obfuscation is a plausible strategy. This makes the costs of sorting out arguments much higher for voters. The political parties, of course, find common cause with interest group lobbyists who spend wisely and generously. We have seen that politicians cannot figure out a way to spend less of other people's money. The "sequester" was pathetic.
High opportunity costs of participation is good news and bad news. The fact that many people have bigger fish to fry is good news. The fact that successful obfuscation causes many to decide it's hopeless to bother is the bad news. We end up with a bad spiral when enough people are turned off for interest groups to have their way. That augments obfuscation by the entrenched, etc., etc., etc.