The current DPS and DCTA debate on ProComp offers fresh data from each side’s proposal and a chance to revisit the original plan.  Still in its infancy, it is too early to judge ProComp a success or failure; but it is worth another look.  One should not need to build a whole house to judge if the foundation is any good.

Changing teacher compensation is critical to education reform, and Pay for Performance plans can be a significant catalyst to improve student achievement. Their value is threefold: 1) to mandate focus on a limited number of objectives; 2) to measure teacher effectiveness to see what works; and 3) to reward specific behaviors. Administrators often struggle with the first, as it requires choosing among seemingly endless objectives (and their constituencies). Unions dislike the second, as the implicit teacher ranking undercuts the single-salary structure at the core of collective bargaining contracts. Rewards get most of the attention, but in my view are often least important; it is the reinforcing cycle between focus and measurement that provides much of the value.

Ideally, Pay for Performance entails a specific bonus (the “Pay”) received upon achieving defined educational outcomes (the “Performance”). These educational outcomes should address a central and acknowledged problem: in Denver, as in most urban centers, the primary need is to both raise overall student performance and close the achievement gap.  How do the current proposals and original plan measure up?

First, let’s look at ProComp’s pay.  Receiving pay as a bonus is important to both directly correlate to objectives and for financial solvency. It is unwise for a goal achieved in 2008 — particularly if it is not sustained — to become a salary increase for 15 years and a pension boost for another 25.

Under the DPS proposal, 90% of compensation goes into base salary, while the union’s plan puts 98%. Whoever prevails, there will be no more than 10% – and as little of 2% — of ProComp dollars paid as a specific bonus.  So the direct link that is one of the central tenants of Pay and Performance is largely absent.

Now to ProComp’s performance objectives: ProComp is divided into 10 elements under four groups: 1) Knowledge and Skills; 2) Professional Evaluation; 3) Market Incentives; and 4) Student Growth.

In the first three of these groups there is no required educational outcome for students.  “Knowledge and Skills” consists of tuition reimbursement and the completion of degrees, licensing, or professional development units.  No study has ever shown a correlation between increased credentials and improved student achievement.  “Professional Evaluation” is composed entirely of salary increases granted upon a rating of “satisfactory.”  I don’t know the percentage of teachers currently ranked “satisfactory” or better, but I’d wager it is both pretty high and far greater than the percentage of proficient students.  “Market Incentives” are bonuses for teaching specific subjects or in certain schools. There is no variation for the quality of teaching: the bonus is for showing up.  Based on DPS data (http://denverprocomp.org/paychart) this means roughly 80% of ProComp dollars are available without either meeting an educational outcome or with any improvement in student performance.  This is true under both proposals.

Of the 20% remaining for Student Growth, about 7% is a bonus for “serving in a distinguished school” – regardless if one’s performance helps the school achieve distinction or not (and assuming as few as 25 teachers per school, this nets out with the impact an individual teacher merits at about 0.3%).

The remaining 13% is linked to educational outcomes: 10% is a “sustainable increase” (assume salary) for exceeding expectations on the CSAP, and another up to 3% is based on meeting one or two “Annual Student Growth Objectives.” This is better, but again, the link from Pay to Performance is not direct: of the 13%, only 1.5% is a bonus.

This gets complicated, and I’ve rounded some numbers, but the part of ProComp that is a specific bonus for achieving a defined educational outcome comprises roughly 2% of available dollars.  Fully 87% is paid primarily for activities regardless of quality. The remaining 11% includes educational outcomes but the direct link to pay is tenuous as it is paid as salary and continues into future years and pensions, regardless if the Performance is sustained. In fact, currently over 90% of ProComp builds pension costs, which given the existing pension burden could well be considered financially imprudent.

It also turns out that the annual cost of administering ProComp is $2 million.  Initially this is not funded by taxpayers, but clearly this $2M cost was not well publicized in a District that just went through a gut-wrenching, 18-month process of closing schools that resulted in savings of just $2.5M per year.

Like so much else about education reform, the praise afforded ProComp seems deeply and prematurely self-congratulatory: lauded as “groundbreaking” and “unprecedented” well before anyone knows if the ground broken or precedents set actually work. Education too often applauds process and inputs, and never bothers with outcomes. Perhaps it is not surprising if ProComp suffers the same imbalance.

In the current debate, I think DPS recognizes the structural constraints, and their proposal makes some marginal improvements, but even if uniformly adopted, very little changes.  DCTA’s proposal lacks even minor improvements: it would almost entirely preserve the “cake” of single-salary structure, and views ProComp as the complementary “icing” (their words).

While the current ProComp debate is important financially there is little innovation at stake. It seems increasingly that ProComp will be an argument between adults that does very little for kids.  As an avid supporter of both changing teacher compensation and Pay for Performance plans, I want to avoid discarding the bathwater because of the baby.  But it is abundantly clear to me that the chance of ProComps’s success in addressing the educational needs of our city is badly limited by its design.

ProComp is often defended as the best plan that could be achieved given the circumstances.  This is, in all likelihood true, although the flaws are larger than acknowledged.  Overall, the DPS and DCTA debate is a sharp indictment of the limited possibilities under a system dominated by centralized bureaucracies and unions.  The final lesson of ProComp is likely to be that that real work of innovation and public education reform is still to come.