CHICAGO -- What started here as a traditional labor fight over pay, benefits and working conditions has exploded into a dramatic illustration of the national debate over how public school districts should rate teachers.
At stake are profound policy questions about how teachers should be granted tenure, promoted or fired, as well as the place standardized tests will have in the lives of elementary and high school students.
One of the main sticking points in the negotiations here between the teachers union and Mayor Rahm Emanuel is a new teacher evaluation system that gives significant and increasing weight to student performance on standardized tests. Personnel decisions would be based on those evaluations.
Over the last few years, a majority of states have adopted similar systems, spurred by the desire to qualify for the Obama administration's Race to the Top education grants. The Education Commission of the States says that 30 states require that evaluations include evidence of student achievement on tests, and at least 13, and the District of Columbia, use achievement measured by test scores for half or more of a teacher's rating.
Proponents say these measures are needed to improve teaching in a country where 33 percent of fourth graders are not reading at grade level and about one-quarter of public high school students do not graduate on time, if at all. They say the new rating systems will help districts identify the best and worst teachers.
These efforts are stirring skepticism and anger among teachers, some of whom express a sense that those behind the new evaluations know little about what it is like to be in a classroom. Others fear that heavy reliance on scores will turn schools into test-taking factories.
That sentiment certainly permeated the picket lines and rallies in Chicago this week.
"Children are so much more than data points on a grid," said Elizabeth Coughlan, a third-grade teacher of gifted bilingual students, who was marching in a rally where teachers clogged downtown streets on Monday. Another teacher held a sign that read, "Let's teach kids to think outside the box not fill in circles."
Advocates of the new evaluation systems say test scores should not be the only measure of a teacher's quality. Even those who believe that such systems can work in theory say that it is important to get teacher buy-in.
"It's tough work because it's hard to get it to be fair," said Kathy Christie, a vice president at the Education Commission. "We've only recently started getting student data that could be traced back to the classroom. It's all very intertwined and complex, and it could fail very easily if people don't get it right. Teachers have very valid concerns."
Still, she said, efforts to reach a consensus could cause the rating systems to collapse in practice. "It's like trying to put a man on the moon by committee," Ms. Christie said. "At some point, decisions have to be made."
In Chicago, the teachers union has bristled at what it sees as a unilateral effort to install a system that will start by basing 25 percent of a teacher's rating on student achievement, going to 40 percent in five years.
The Illinois legislature passed a law in 2010 that requires all districts to develop teacher evaluations based in part on student performance, with Chicago being the first district to begin its system this year. The law, which passed unanimously in the Senate and received only one opposing vote in the House, requires that various test results be used for at least 25 percent of a teacher's rating in the first two years, growing to 30 percent. Classroom observations also figure prominently in the evaluations. A separate law passed in 2011 allows teacher evaluations to be used in tenure and layoff decisions.
Chicago's teachers say they would accept a rating where 25 percent was based on student achievement on tests, but balk at the increase to 40 percent, higher than the state standard.
Across the country, critics have seized on the Chicago fight to blast the use of a teacher's ability to raise scores as an unreliable measure. The National Center for Fair and Open Testing put out a statement from its public education director, Robert Schaeffer, saying, " 'Enough is enough' to so-called reforms based on standardized exam misuse."
Several studies have shown that teachers who receive high value-added scores -- the term for the effect that teachers have on student test performance -- in one year can score poorly a year later. "There are big swings from year to year," said Jesse Rothstein, associate professor of public policy and economics at the University of California, Berkeley. But other studies have shown that students taught by teachers who achieve high value-added scores go on to have lower teenage pregnancy rates, are more likely to go to college and earn higher incomes as adults.
Some studies, including one that looked at a pilot of teacher evaluations here, have shown correlations between teachers whose students' test scores improve and those who receive high marks in classroom observations and on student surveys.
Sara Ray Stoelinga, senior director at the Urban Institute at the University of Chicago who conducted the study, said that using student test scores protects teachers from arbitrary decisions by principals. "The theory," she said, "is that if you have multiple pieces of information, it gives the most fair and accurate measure."
But with research at an early stage, other districts and states have stepped carefully. In Colorado, where a sweeping education law passed three years ago stipulating that half of a teacher's evaluation should be tied to student performance, the state is slowly introducing the programs, with training. "We want to make sure to do it right rather than do it fast," said Michael Johnston, a Democratic state senator who sponsored the bill.
And in New Haven, Conn., the district and the union spent more than six months discussing a new evaluation system, and union members felt their feedback was valued. "We knew we were being treated as equal partners," said David Cicarella, president of the New Haven Federation of Teachers. "It can't just be one test that the kids take one week in March."
Even those who say there is value in using test scores to measure a teacher's performance say there are plenty of other factors. "There are other things that teachers do that aren't captured by test scores," said Douglas Staiger, an economist at Dartmouth College who has studied the effects of teachers on student achievement. The debate over whether test scores accurately reflect a teacher's ability, he said, should ultimately be about "how much importance we want to place on academic achievement defined by the test."
This article originally appeared in The New York Times.