More than half of the votes in last week’s United Federation of Teachers leadership election came from retired union members, according to a final vote tally that the union released today.

The complete count, released three days after the election, showed 19,808 votes for president from people who currently work in schools and 21,813 from retired members. The distribution reflects longstanding trends in union voting that were accelerated by dramatically lower turnout and a policy change that increased the weight of retirees’ votes.

Retirees make up a potent, and unusual, voting bloc in the UFT, one of the only labor unions in the country that allows retired members to continue to vote in union elections. They turn out in droves and almost always cast their ballots for the union’s leadership. This year’s election was no outlier, with 38 percent of retirees returning their ballots, compared to just 17 percent of active members.

Ninety-three percent of the retirees backed Michael Mulgrew in a quest for his second full term as union president. Active members, on the other hand, cast their ballots 21 percent of the time for Julie Cavanagh, the teacher put forth by the Movement of Rank-and-File Educators, a minority caucus that opposes Mulgrew’s positions on charter schools, teacher evaluations, and other hot-button policy issues. (A third party, New Action, ran a slate of candidates but did not oppose Mulgrew.)

Norm Scott, a founding MORE member, cautioned last year about the growing significance of retirees’ votes. “The danger for the union is that it continues to shut out the voice of people in the classroom,” he said.

Active members would still have had a majority if the union had not revised its constitution last year to increase the weight of retired members. Since 1989, the number of retiree votes had been capped at 18,000, in a provision that union leaders said was aimed to limit retirees’ influence. Last year, with the ranks of retired union members having tripled, the union’s leadership amended the constitution to allow 23,500 retiree votes to count.

Retirees would also not have cast the majority of ballots had active teachers voted in the same numbers as they did three years ago. In 2010, retirees cast only 46 percent of votes — and that ratio was further reduced by the cap that was in effect. This year, turnout among retired members was down by 12 percent – but turnout among active members was down 30 percent.

Just 17 percent of active members — educators who are currently working in schools and with city students — actually voted, a turnout that dismayed some who want to see changes in the union and the school system.

“The union should study these results and find out why so many active classroom teachers aren’t participating in the process,” said Jonathan Schleifer, executive director of New York City’s chapter of Educators 4 Excellence, a teacher advocacy group that opposed the union’s decision to increase the weight of retiree votes. “We know from our experience and growth in New York City that teachers are craving ways to get involved.”

The final vote tallies also highlighted another longstanding division within the union. High school teachers, historically the most radical within the union, cast their votes much more often for Cavanagh. She got 41 percent of high school teachers’ votes, compared to 23 percent of the vote from middle school teachers; 17 percent from elementary school teachers; and 14 percent from “functional” members, such as guidance counselors and speech therapists.

The preliminary vote count showed Mulgrew winning with 84 percent of the vote, down from 91 percent in 2010. The final numbers, which include the ballots of voters who did not select a single slate of candidates, show that he actually received about 86 percent of votes. The American Arbitration Association handled the election.