Denver’s Professional Compensation System for Teachers

An Outcomes Evaluation of Denver’s Alternative Teacher Compensation System
2010 Report

Denver’s Professional Compensation System for Teachers (“ProComp”) is among the most prominent alternative teacher compensation reforms in the United States. Although teacher compensation that departs from the traditional “single_salary schedule” 1 is not new, it has recently gained popularity again in policy circles as an approach for increasing student achievement and improving teacher quality. Accordingly, two primary goals of ProComp are to: 1) increase student achievement; and 2) attract and retain high_quality teachers to the district.

This is the first of two reports detailing possible effects of Denver’s ProComp on student achievement, educator attitudes and behaviors, and teacher retention. This first report describes outcomes that may be associated with ProComp at the program level; a subsequent report (to be released in September, 2010) describes outcomes at a finer level of granularity to betterunderstand differential outcomes of the program’s various elements for educators of various backgrounds.


Key findings include:
District_wide student achievement trends
_ Growth in mathematics and reading achievement has increased substantially from 2002_
03 to 2002_09.
_ Teachers’ median conditional growth percentiles have seen a 3_8% increase over the last
eight years, based on a conditional achievement measure similar to that of the Colorado
Growth Model.
_ Trends are similar as measured in value_added teacher “effect” estimates.
Student achievement outcomes
_ Some evidence of ProComp composition/selection effects
o Teachers hired after the implementation of ProComp (subsequent to which
program participation has been mandatory) exhibit higher first_year achievement
than those hired prior to the program.
o This finding is:
• _ Positive for both mathematics and reading
• _ Consistent across school levels (though more pronounced at elementary
• level)
• _ Outcomes similar regardless of whether measured in terms of conditional
• growth percentiles or more common estimated value_added achievement
• effects
_ Less evidence of ProComp productivity effects
o Teachers who have voluntarily opted into the ProComp program slightly
outperform their non_participant colleagues, though differences are less
pronounced when adjusted for individual differences between teachers who
choose to participate and those who do not. Mixed effects models offer little
evidence of significant changes in effects for voluntary participants after
implementation compared to the effects before implementation

Teacher and principal attitudes
_ Teachers:
o Generally, teachers who participate in ProComp hold more favorable views of
ProComp than those who are not in ProComp
_ Teachers who voluntarily entered ProComp are most supportive of the
program
_ Teachers who were automatically enrolled in ProComp (due to joining
the district after January 1, 2006) reported attitudes that were more
similar to voluntary participants than to the attitudes of teachers who
were not in ProComp (This suggests positive selection into DPS for new
teachers.)
o A majority of ProComp participants indicated that they believed the ProComp
program could motivate teachers to improve instructional practices, with
positive respondents outnumbering negative responses by a 3_to_1 margin.
o Participants indicating that the ProComp program would ultimately improve
student achievement outnumbered those who disagreed by a 2_to_1 margin.
o Participants were evenly split regarding whether they believed ProComp had
improved teacher collaboration.
_ Principals:
o Generally, principals reported more favorable beliefs about ProComp than did
teachers
_ More also reported favorable beliefs about ProComp than about the
traditional salary schedule
o In a pattern opposite that of teachers, more principal respondents believed
ProComp could increase student achievement than believed ProComp could
improve instructional behaviors
Attitudes, Instructional Behaviors, and Student Achievement
_ Teachers who reported favorable attitudes towards ProComp were more likely to report
they had changed their instructional behavior and practices
_ The relationship between favorable attitudes toward ProComp and reported changes in
instructional behaviors persisted regardless of whether models included controls for
prior achievement history
_ Changing the way attitudes were conditioned on prior achievement resulted in different
effects on reported changes of instructional behavior
o Teachers’ percentages of high_growth students (students with student growth
percentiles above the 55th percentile) had little relation to teachers’ reported
changes in their instructional behaviors
o Greater relationships appeared between changes in teachers’ instructional
behaviors and percentages of students with high_growth students than between
instructional behaviors and the dichotomous distinction of an effective teacher
_ Suggests prior student achievement gains that occur around the 55th
percentile may be more influential on teachers’ attitudes
2010 ProComp Evaluation Report
P a g e 9
Retention in Hard_to_Serve Schools
_ Over the past decade, DPS has generally experienced an upward trend in teacher
retention
_ Schools with greater rates of ProComp participation have experienced higher rates of
retention in recent years
_ Retention trends at hard_to_serve schools lag behind those of schools that are not
designated “hard_to_serve”
_ Schools with greater rates of ProComp participation that are designated “hard_to_serve”
experienced a sharp increase in retention rates in the first full year ProComp was
implemented (2006_07)
o These schools also saw the greatest increase in retention (from 74% to 86%) over
the past decade
o Suggests there may be a positive effect on retention trends associated with
ProComp and the hard_to_serve bonus
This study is observational (rather than experimental) in nature.