Performance Pays

As CIO for Denver Public Schools, Ed Freeman serves 72,000 students in 140 schools―a big change from his previous life as a consultant for venture capital firms. Since joining DPS three years ago, he has succeeded with unusual projects such as a teachers’ pay for performance system and a statewide forum for CIOs. School CIO spoke with Freeman about how private sector experience and enthusiasm for technology inspire his work.

Q. What is different about working in public schools rather than the private sector?
A. At first, the pace really impacted me; I was used to moving much more quickly. In the private sector, profit motive is the bottom line. The agendas are very clear: if you’re not generating profits for your company, you have to rethink what you’re doing. In the government sector, the matrix of possible choices you can make and priorities you can create is much broader, so it’s slower.

Q. Tell me more about the teachers’ pay for performance system.
A. It’s a system we built in-house, called ProComp, and it has been implemented over a five-year period. Over a third of the teachers have enrolled―it’s optional for existing teachers, but not for new teachers.

ProComp’s payout system has four components. The first is knowledge and skills, which includes payouts for professional development, graduate degrees and certifications, and tuition reimbursements. The second component is professional evaluation, which has payouts every year. Third, there are market incentives: payouts for hard-to-staff positions and hard-to-serve schools. The fourth component is student growth, which is based on annual student growth expectations, a standardized achievement test called the CSAP, and a distinguished schools system. If you have helped to create a school that has distinguished itself academically, then you get a payout.

Q. Why did Denver Public Schools decide to develop ProComp?
A. The teachers’ union representatives and the administrative leadership all agreed that it would be a good path to explore. They started off by doing quite a bit of research, and then created studies and pilot programs to gather more datA. They analyzed all of that data and came up with their final design. Not only did the administration and schools vote for the idea, but also the voting public approved the idea to the tune of a $25 million Mill Levy, which is a tax that citizens of a community take on in order to support key school programs.

Q. How will you measure the success of the system?
A. We are in the process of putting together a number of metrics, ranging from enrollment counts to student achievement. But this is the first year of significant implementation. Last year, only a few components were fully implemented, so enough of it wasn’t sufficiently in place to get a good feel for its effects. You’ll have to ask me again in a year.

Q. Can you describe how the technology of ProComp works?
A. The system is too large and complex to attempt to perform all tasks manually and still have time to teach kids. ProComp processes absolutely have to be automated. We have modules (subsystems within ProComp) that, for example, help principals assess student growth objectives and do performance evaluations with their teachers. We’re also building a module that will help teachers manage their professional development units. They will be able to find out about new courses, have courses tracked and have that information sent to the central data warehouse so that they can be paid properly for professional development. We also have a backend module that takes all the information from automated and manual sources, does the necessary salary calculations and produces accurate paychecks.

Q. Do you have other automated systems?
A. Yes. One example is the automated teacher staffing system. It has an online application on the front end, and goes all the way through the budgeting process for hiring teachers on the back end. It is part of a larger initiative; we want to do the same thing with all district employees. But we’re building this system in phases, which is what any good consultant would do.

Q. How do you know that automated systems are worth the effort and expense?
A. You tackle the biggest problems first, and then you look at outcomes by getting feedback from your affected customers, and by looking at statistics on how performance has changed. When the chief operating officer arrived here five years ago, all of the applications for job openings were in boxes on the floor of the HR office, and if you wanted to find a particular candidate, you had to come to the office and sift through the boxes. Now, people can connect to the Web site from anywhere in the world, do a job search, find the job that they want, apply online, and the system will take them through a set of steps, through the interview process and potentially through a hiring process.

Q. But there are still people involved in the hiring process, right?
A. Of course, the system doesn’t make the final hiring decision. Automating a workflow process still means that there is a human at every step, but automation helps to maximize the productivity of human beings. We are trying to get rid of all the grunt work, to let people do the work they need to do and make the best choices.

Q. You brought together a forum of CIOs in your state. Can you tell me more about it?
A. There are 13 CIOs and technology directors in the forum, and we represent about 500,000 students in the state. We initially organized to evaluate a new student information system. But after our third meeting, we realized that we had a lot more to talk about than the student information system! So we broadened our perspective and talked about various ways we could work more closely together.

Q. How has the forum been helpful?
A. One example is pooling our RFPs for hardware: you get better prices with higher volume. Rather than Denver Public Schools alone ordering 3,000 laptops, everyone in the CIO forum can put their needs together and maybe we can order 15,000 laptops, and allow the volume to help us get a much better price.

Lindsay Oishi is a graduate student in Learning Sciences and Technology Design at Stanford University.