Scaling Education Initiatives: Why and How

Cross-posted at Education Week.

Buried in last week’s blog—an open letter to Mark Zuckerberg and Priscilla Chan—I remarked on the unscalability of certain education initiatives praised by Checker Finn as “charter-like.”  One of them was New Leaders.  After it was published, I received the following statement from Jackie Gran, Chief Policy & Evaluation Officer of New Leaders:

“We are disappointed that Mr. Tucker describes efforts to get a well-prepared, highly skilled principal in every school as “unsustainable and unscalable” full stop.

Mr. Tucker cites New Leaders’ per-principal cost — $85K—as evidence. He doesn’t consider the cost savings achieved by hiring principals with the skills to manage the demands of leading a school without quickly burning out. 

Only 50 percent of principals nationally remain in their roles for more than three years (and retention rates in urban districts are considerably lower). By contrast, 74 percent of New Leaders trained are retained beyond that point. The average cost to districts of preparing and hiring each new principal is $75,000. For districts juggling scarce resources, investing in high-quality training that prepares principals to achieve a lasting difference makes sound financial sense.

And critically, we cannot carry out the strategies Mr. Tucker lists as requirements for school improvement at-scale—strategies that New Leaders strongly supports—without school leaders capable of instituting a high quality curriculum, bolstering instruction, recruiting and retaining talented teachers, and fostering a culture of achievement and shared accountability for student success. 

Strong school leadership matters—decades of research support this fact—so quality leadership training is essential. A recent report by the RAND Corporation found that New Leaders has the strongest evidence of effectiveness of any principal preparation program in the country, including both traditional and alternative models. A key to that success is our yearlong residency model—similar to a medical residency—which provides aspiring principals with real-world learning that equips them to be successful from day one. That’s why we promote proven practices such as job-embedded learning across all principal preparation programs. For example, with the University Council for Educational Administration, we recently released a toolkit ( for states on how to use program evaluation to improve the quality of principal preparation.

New Leaders does recognize that school systems have limited resources and we continue to identify cost-savings so we can deliver principal preparation at lower costs without sacrificing quality. For example, we expanded the scope of our programming to develop teacher leaders because providing future principals with leadership development opportunities earlier in their careers results in greater success as they grow while immediately infusing schools and systems with teams of skilled leaders capable of enacting ambitious improvement plans.

Strong demand for our programs—we work in over 20 districts and 100 charter schools—provides evidence that schools nationwide crave effective, research-based approaches to developing teacher, school, and system leaders. Only through such investments can we effectively carry out the strategies Mr. Tucker so rightly promotes.” 

NCEE, the organization I head, is deeply committed to the development of capable school leaders for all the reasons given by New Leaders in their statement.  But our earlier work in the delivery of technical assistance and training to schools had a big influence on the strategies we chose.  That work—on comprehensive school reform—began in 1989.  It was entirely paid for by private foundations.  The program grew rapidly.  But the foundations that had supported its very successful launch did not want to tie up their money indefinitely, much less put ever more of their resources into it as it expanded.  Other foundations were not interested in picking up where the initial supporters had ended, in part because they would get no credit for the success of programs others had begun.  So we decided, with some trepidation, to ask each school and district to pay the full cost of the training.  We aimed to offer a first-class program based on the best research available, a program that produced outstanding results and one that districts and states could afford.  If they could afford it and would buy it, then there would be no limit to the scale it could achieve.  The design is a train-the-trainer design, so that the district or state develops the capacity to deliver its own face-to-face training and support to the principals.

Many states and districts were willing to pay for it, enough to keep doubling the size of the program annually for years on end.  We found that the states and districts took the program much more seriously when they had to pay for it.  They were more prepared to change core policies and institutional structures to support the program than they had been before and to allocate more executive time to assure its success. They said when the program was supported by private foundations, they viewed it as soft money that could disappear overnight, so it made no sense to change their policies and practices to accommodate it.  The school boards wouldn’t refuse a free program, but were committed to it only as long as the money kept coming in.  But when the superintendent had to sell it to the Board and the Board had to fund this initiative instead of a something else they could use the money for, they owned it in a way they did not own it when it was free, and, furthermore, they held their superintendents responsible for its success in a way that they did not when it was free.

We created the National Institute for School Leadership more than 15 years ago at the request of a group of foundations led by Carnegie Corporation of New York.  We decided initially to concentrate on creating an executive development program for serving school principals, because the majority of school staff who sign up for principal training in the nation’s graduate schools of education have no intention of becoming school principals.  The program we ended up with combines the best research on leadership from the world’s top military and business schools with the best research on education in a system that involves both web-based delivery and face-to-face instruction over a period of what was once 18 months and is now down to 12 months.

Research performed by the RAND Corporation has found that the only program meeting the U.S. Government’s Tier II standards with evidence of improved student performance for the initial training of principals is New Leaders and the only program meeting the Tier II standards with evidence of improved student performance for the continuing training of principals is NISL’s Executive Development Program.  So the two programs meet the same demanding performance standards.

But there is a big difference between the two. The NISL Executive Development Program costs an average of $8,925 per principal trained, which appears to be on the order of one-tenth of the cost of New Leaders standard model. At first, this made no difference to the districts and charter management companies that used it, because the U.S. Congress subsidized the program with a massive earmark.  When the Congress stopped granting such earmarks, private foundations stepped in.  But the cost was so high that the support from the private foundations was not enough and so New Leaders had to ask the districts and charter managers to bear some of the cost.  But, even with the subsidy, the cost remained so high that only a small number of districts and charter companies could be served with this model.

We have a high regard for the quality of the New Leaders program and for its staff.  They believe that their Aspiring Principals program is a worthwhile investment for the results districts get from implementing it.  But they know that its expense imposes real limits on the rate at which they can grow it.  They have been working with their supporters to lower the cost and to encourage local funders and districts to put “more skin in the game.”  And they have also been working in an imaginative way to find ways to take advantage of their experience to create a range of other products and services that are less expensive and more likely to fit in with districts’ established ways of working.  We think they are on the right track.

The point I was making in my previous blog was addressed not to New Leaders but to Checker Finn’s use of the original New Leaders model as an example of what foundations ought to be supporting.  And it was simple:  New Leaders’ program for Aspiring Principals, like the Teach for America model, which Finn also cited, was not designed to scale.  Foundations that want to use their own money to leverage large changes in American education can’t think about scale after they have invested in the interventions of their choice.  They have to think about scale as they are choosing interventions, and, if they do that, they will not select the kinds of interventions that are likely to affect American education only at the margins because their cost is prohibitive for most districts and states. They need to find interventions that will result in products and services that districts and states will be ready and willing to pay for themselves once the interventions have been shown to work.  That is where the money is for large-scale implementation.  Indeed, that is the only place you will find it.