The American educational landscape is cluttered with the ruins of grand
theories, erected far and wide before anyone thought to find out whether
they could stand on their own.

New math, open classrooms, self-esteem training, whole language, discovery
learning, multiple intelligences, heterogeneous grouping, and new
new math (again) - everybody will have a different list of follies.

Denver Public Schools has, in contrast, rather a modest theory. It's
that making part of teachers' pay dependent on better student performance
will lead to better student performance.

This year and next, it is testing its theory with a pilot program in
12 elementary schools with about 10 percent of the district's teachers.
Teachers in schools chosen for the pilot study would get $500 the
first year just for participating. Each teacher, together with the
school principal, would set two objectives related to student achievement.
The teacher would get $500 the first year for meeting each of the
objectives, and $750 each the second year.

If the experiment is successful - and if the union approves - the current
contract, which bases salary increases on years of service and educational
credentials, will eventually be replaced with one requiring all teachers
to earn their yearly increases by meeting their individually agreed-on
objectives.

Whether union members would vote for such a contract is uncertain at
best. They did approve the pilot study, but that was a no-lose proposition,
because teachers participating in the study get extra money, and nobody
gets less. Even so, the district had trouble finding schools where
at least 85 percent of teachers wanted to take part in the pilot study,
as the plan required. No middle schools qualified, and only the bare
minimum of 12 elementaries that the plan called for. Becky Wissink,
vice president of the teachers' union and a member of the four-person
design team that will oversee the study, said only that more than
half of the teachers at more the half of the schools wanted to participate.

That introduces another uncertainty. What if the pilot plan succeeds,
but only because it is operating in schools where almost all the teachers
favor it? 

The largest uncertainty, though, is whether the pilot program, ambitious
though it is, will yield conclusive results. There are too many variables.

Denver is a 69,000-student district. Almost 14,000 of them have limited
English proficiency, most of them Spanish-speaking, and more than
half the students are Hispanic. After two decades under a court desegregation
order, it returned to a neighborhood school policy three years ago.
But it has alternative schools, magnet schools and charter schools.
Also, Colorado allows both intra- and interdistrict public school
choice and statewide, more than 20 percent of students go somewhere
other than their neighborhood schools.

The pilot study calls for 12 elementary schools, divided into three
groups of four, each one trying a different variant of pay-for-performance.

The first group of four schools will rely on standardized testing.
The contract called for basing teachers' objectives on the Iowa Test
of Basic Skills, but the design team decided teachers could use other
tests, as appropriate, including La Prueba, for Spanish-speaking children,
Terra Nova, or in the grades when they're given, the Colorado state
assessments.

The four schools in this group are wildly different. Colfax is 81 percent
Hispanic, has 86 percent of its students in the free or reduced-price
lunch program, and a 23 percent transiency rate, that is, the percentage
of children who start the year there but leave before the end of the
year. On the state's fourth-grade reading test, 25 percent score proficient
or advanced. Smith is 75 percent black, with 72 percent of students
in the lunch program, a 19 percent transiency rate, and 12 percent
passing the fourth-grade reading test. Oakland is mostly black and
Hispanic, with 54 percent in the lunch program, 10.5 percent transiency
rate and 39 percent passing the reading test (above the average for
Denver, which is 31 percent). The fourth school, Traylor Fundamental
Academy, is a districtwide choice school with a back-to-basics curriculum.
It is mostly white and Hispanic, though neither is a majority, it
has 23 percent of students in the lunch program, 7.8 percent transiency
rate, and 77 percent of its students pass the fourth-grade reading
test.

Obviously the first task will be to figure out what objectives are
appropriate for such disparate situations. To be fair to teachers,
their objectives should be roughly equal in difficulty, but what does
``roughly equal'' mean between, say, Smith and Traylor?

Making the problem even worse, the schools have the option of choosing
school-wide, rather than individual, objectives. If one or two schools
succeed in the pilot program, is it because of pay-for-performance,
or because of some of the many other ways they are different?

The second group of four schools will set objectives based on tests
developed or chosen by teachers. The design team overseeing the pilot
will ensure that the tests are aligned with the school curriculum
and with the state standards, but the problems of comparability will
remain.

As it happens, all four schools in this group are predominantly Hispanic,
ranging from 55 to 90 percent.

The third group of four schools will test whether better teacher skills
improve achievement. The teachers will meet their objectives by demonstrating
they can use their training effectively in the classroom. The success
of the group, however, will be measured by student performance, in
some way yet to be decided.

One of schools in this group, Cory, is 81 percent white, with low rates
of poverty and transiency, and a 68 percent passing rate on the fourth-grade
reading test. Mitchell, like Smith, is mostly black and Hispanic with
a high proportion of at-risk children. And one school, Southmoor,
just reopened this fall, so it has no track record.

It's difficult to believe that any definitive conclusion will emerge
from such a complex experiment. But the district administration and
the teachers deserve credit for their efforts.

At least they may avoid building any more follies.