There was wide agreement that the economy still has far to go - three-fourths of voters said it was poor or not so good, according to preliminary results of exit polls. Only a fourth thought they were better off financially than four years ago when President Barack Obama was elected.
The survey of voters as they left polling places showed 6 in 10 ranked the economy the top issue. The majority who don't yet see economic improvement were roughly divided over whether things were getting even worse or just stuck in place.
Joseph Neat, a stay-at-home father in Hagerstown, Md., said he voted for Republican Mitt Romney because Obama has had enough time to deal with the economic troubles affecting families, especially gasoline prices that he called "insane."
"We don't have time for him to make changes. We need the changes now," Neat said of Obama. "And four years is plenty of time."
Voters pointed to years of high unemployment and rising prices as the biggest problem for people like them; those two worries far outstripped concerns about the housing market or taxes in the exit polls conducted for The Associated Press and television networks.
About half of voters say the previous president, George W. Bush, shoulders more of the blame for economic problems than Obama.
About 4 in 10 blamed Obama. They don't include William Mullins of Lansing, Mich.
"Obama had a lot to deal with when he came to office," Mullins said. "You can't change everything overnight."
Only a quarter of voters were feeling enthusiastic about Obama's administration; almost as many said they're angry about it.
But voters were more likely to say Obama stands for the middle class or the poor.
Half of voters said they think Romney's policies generally favor the rich and barely any thought he favors the poor. Only about 1 in 10 said Obama favors the wealthy. The biggest group - 4 in 10 - said Obama's policies help the middle class, with the poor coming in a close second.
The survey of 15,825 voters was conducted for the AP and the television networks by Edison Research. This includes preliminary results from interviews conducted as voters left a random sample of 350 precincts nationally Tuesday, as well as 4,389 who voted early or absentee and were interviewed by landline or cellular telephone from Oct. 29 through Nov. 4. Results for the full sample were subject to sampling error of plus or minus 2 percentage points; it is higher for subgroups.