A national poll released by Quinnipiac University Wednesday morning found 42 percent believe public employees are paid too much and 35 percent say they’re paid the right amount, while 15 percent say they’re paid too little.

The poll also found to reduce state budget deficits collective bargaining for public employees should be limited. Forty five percent of voters support limiting collective bargaining rights, while 42 percent oppose limits on collective bargaining, a hot button issue in the state of Wisconsin.

Connecticut Gov. Dannel P. Malloy’s staff is meeting with leadership of the State Employees Bargaining Agent Coalition Wednesday for the first time to talk about the $2 billion in unidentified labor concessions over the next two years that the governor will use to balance the budget deficit. Malloy has called state employee wages and benefits “unsustainable.”

Voters surveyed by Quinnipiac agree. They said 63 to 31 percent that government workers should pay more for benefits and retirement programs.

“There is a partisan tinge to American voter attitudes toward government workers: By wide margins, Republicans say these workers are overpaid; want them to pay more for their benefits and want to limit collective bargaining,” said Peter A. Brown, assistant director of the Quinnipiac University Polling Institute.

“Democrats say 2-1 that government workers are paid about right or too little, rather than too much and split over whether they should pay more for benefits, but they solidly defend collective bargaining. Independent voters, as usual, are in the middle.”

Support for limiting collecting bargaining to reduce state deficits is 59 – 25 percent among Republicans, while independent voters split 45 – 43 percent. Defending collective bargaining are Democrats 56 – 33 percent and voters in union households 62 – 29 percent. Men are split 47 – 45 percent while women would limit bargaining 43 – 39 percent. White voters split 44 – 43 percent. Black voters back limits 49 – 37 percent.

The poll conducted Feb. 21 through Feb. 28 interviewed 1,887 registered voters and has a 2.3 percent margin of error. Live interviewers called land lines and cell phones.