This View of Life Anything and everything from an evolutionary perspective.
FIND tvol:
Why Best Practices Don’t Spread
Image credit: hobvias sudoneighm
David Sloan Wilson
David Sloan Wilson
is the SUNY Distinguished Professor of Biology and Anthropology at Binghamton University and Arne Næss Chair in Global Justice and the Environment at the University of Oslo

In 2009, the Binghamton City School District asked me to help design a “school within a school” for at-risk high school students called the Regents Academy. Students who were failing most of their courses during the 8th and 9th grades were almost certain to drop out unless something was done. Could I help?

I was approached because I had started to use my city of Binghamton as a field site for community-based research and action in 2006 (go here and here for more). My first project was a collaboration with the school system that involved measuring developmental assets of students in grades 6-12. Developmental assets are properties of individuals (internal assets) and their social environments (external assets) that contribute to healthy development. A survey to measure assets called the Developmental Assets Profile (DAP) was widely used and dovetailed nicely with my longstanding interest in prosociality, defined as any attitude, behavior, or institution oriented toward the welfare of others or society as a whole. DAP survey questions could be used to correlate the prosociality of the individual student (an internal asset) with the prosociality of their social environments (an array of external assets including family, neighborhood, school, religion, and extracurricular activities). This work received a lot of attention and established me as a worthy person to help design the Regents Academy.

I leapt at the opportunity because I was eager to put my scientific knowledge to practical use. At the same time that I began working in Binghamton, I was also helping to create the Evolution Institute, the first think tank to formulate public policy from an evolutionary perspective. EI projects enabled me to work with luminaries such as Nobel laureate Elinor Ostrom, Acceptance and Commitment Therapy (ACT) founder Steve Hayes, and prevention scientists Tony Biglan and Dennis Embry. I learned that while the quality of applied research is highly variable, the best is very good indeed. The Regents Academy was my first opportunity to try my hand at applied research and achieve the gold standard of assessment, the randomized control trial. I even had a graduate student named Richard A. Kauffman who had been a public school teacher and was available to take on the project as part of his PhD research.

Sign up for our newsletters

I wish to receive updates from:

To make a long story short, the Regents Academy was a spectacular success during its first year. Our guiding strategy was that all groups whose members must work together to achieve common goals require certain core design principles, no matter what the specific goal. Elinor Ostrom received the Nobel Prize in economics in 2009 for identifying eight design principles required for groups to successfully manage common-pool resources such as irrigation systems, pastures, forests, and fisheries. We used the same principles (plus two auxiliary principles relevant to the context of education) to design the Regents Academy. Not only did the Regents Academy students greatly outperform the comparison group in a randomized control design, but they performed on par with the average high school student in the city of Binghamton on the state-mandated exams. The nonacademic wellbeing of the students also increased as measured by the DAP.

But that’s not the story of this article. The story of this article is that the Regents Academy failed to survive – despite its outstanding success addressing one of the toughest problems in education.
My experience with the Regents Academy is disturbingly common. From my conversations with other applied researchers, I know that the world is full of programs that accomplish their objectives but don’t survive or replicate. Conversely, programs that don’t work and even cause harm can remain entrenched, no matter how strong the evidence against them: An example is D.A.R.E., which remains a fixture as a drug prevention program despite a mountain of evidence that it doesn’t work.

This seems paradoxical from an evolutionary perspective. Isn’t cultural evolution supposed to result in the adoption of useful practices? Yes, but the genetic mechanisms of cultural evolution evolved in the context of hunter-gatherer societies. We are well adapted to copy the practices that are most useful in terms of our immediate experience. We often rely upon surrogate information such as copying the most common behavior or the behavior of the most prestigious individuals. We often copy behaviors that benefit ourselves rather than the common good when these are not aligned with each other. Finally, learned traits are easily transmitted among individuals in societies where children grow up in the company of older children and adults, without requiring formal education.

A mismatch between ancestral and modern environments can lead to dysfunctions in cultural transmission comparable to immune system dysfunctions that are beginning to become well recognized. If so, then it isn’t good enough to design a program that works and to take its continuance for granted. Its continuance must also be designed with mismatches in mind.

Here are some of the problems that we failed to address, or were not in a position to address, which contributed to the demise of the Regents Academy.

Cultural transmission. During our first year, we worked closely with the principal of the Regents Academy to implement the design principles that made the program work. She was reassigned during the second year, forcing us to orient a second principal. Worse, we realized that many features of the program had been implemented during the first year without the teachers understanding their rationale. They were simply following the instructions of the first principal. Realizing that staff turnover would be a fact of life, we began to orient the second principal and the teaching and administrative staff at the same time. Since everyone was extremely busy, this needed to be done without excessively increasing workloads. After some common reading, we printed the design principles on posters that were displayed in every room. Each staff person was assigned a design principle to monitor and to lead a brief discussion in rotation during the staff meetings, which Rick and/or I attended whenever possible. We made plans to create a website organized around the design principles for the program so that they would become engrained among staff and students alike in the process of creating content for the website. These efforts were quite successful, and we were pleased by the quality of the discussion during the staff meetings. We felt that the staff had internalized the design principles and would be capable of transmitting them on their own rather than following instructions. Our program was now designed with transmission in mind, which we had neglected to do during the first year. This was fortunate, because we had two new principals to orient during the third year. This kind of instability is common in public school systems and undermines the continuity of all programs.

Establishing new routines. On numerous occasions, we decided to employ a new practice that should have been straightforward to implement. Everyone wanted to do it, and it would be no more costly or time consuming than the old practice. Nevertheless, it often proved difficult to implement the new practice – as if we couldn’t collectively put one foot in front of the other. In retrospect, we realized that institutional routines are like individual habits that can be difficult to break. Change per se can be difficult, even when it is desired by all. We were unable to address this issue before the program as a whole was terminated, but in the future we will design programs with the positive reinforcement of change in mind.

Program autonomy: One of the core design principles (#7) is the autonomy of group members to govern their own affairs. The Regents Academy could not have worked during the first year if we were not given a free hand in its design, but this was an exception to the social organization of the Binghamton City School District as a whole, which like most public school systems is hierarchical. Superiors are accustomed to issuing directives to subordinates without their authority being challenged. As soon as the Regents Academy showed signs of succeeding during its first year, the Administration wanted to transfer other at-risk students into the program. This would have compromised a number of the design principles (such as the strong group identity that we had established), not to speak of the randomized control design required for assessment; yet we had to argue long and hard to limit the influx of new students. In subsequent years, meetings among the top administrators that were highly relevant to the Regents Academy took place without the principal, Rick, or me even being present. Eventually, a new school superintendent decided that it was wrong for at-risk students to be isolated from the rest of the student body. If the Regents Academy were to continue at all, it would need to cope with the fact that the students were being thrown back into the normal high school environment. At this point, Rick and I decided that the Regents Academy did not have enough autonomy to govern its own affairs, and we terminated our involvement with the program, which ceased to exist in any form during the following year.

What lessons can be learned from the demise of the Regents Academy? I have become pessimistic in some respects but remain optimistic in others. The Regents Academy failed primarily because of problems at the level of the whole school system. The systemic problems can be ranked in the following order:

1) A hierarchical social organization that violates the design principle of local autonomy.

2) Staff turnover, placing demands on the cultural transmission of successful practices.

3) The absence of positive reinforcement procedures for changing routines.

4) Decisions that are not based on solid theory and empirical evidence. These systemic problems will stand in the way of any successful school program, and they are so deeply entrenched in the American public school system that it is difficult to see how they can change.

Optimistically, the demise of the Regents Academy is not due to a failure of the design principle approach, but rather to violations of the design principles that we could not control due to a lack of autonomy. As long as the program could govern itself, it was successful. We were even able to extend the design principles approach to cope with some of the systemic problems, such as high staff turnover. The program, therefore, remains a model for implementation in schools where the systemic problems are not so great, such as charter schools, private schools, home schooling cooperatives, and public school systems that manage to solve some of their systemic problems. In addition, the design principles are relevant to all groups whose members must work together to achieve common goals, not just schools. The Evolution Institute is therefore creating a framework called PROSOCIAL that will enable any group to adopt the design principle approach while contributing to a scientific database for further advancing our knowledge about the efficacy of groups. We aim to create a social organization that maximally favors the implementation of the core design principles and the spread of practices that work.



Join the discussion


  1. Lillie Crowley says:

    Dave, I’d love to discuss with you a project we did some years ago in the community college system in Kentucky. We were trying to address what we call the “intermediate algebra problem”. We had NSF money, publishers wanting to publish our reform algebra book (we did publish it), alleged administrative support (especially so long as they had indirect funds coming in), supposed buy-in across the state, etc. And the actual courses had excellent outcomes: much improved success rates; increased immediate enrolment in follow-up courses; and improved success rates in same, although they were traditional format. Still, the project, long-term, failed miserably.

    Cheers, Lillie

  2. Joe Brewer says:

    Thanks for sharing what you’ve learned along this journey of discovery, David.  Your analysis points to what all practitioners of organizational change struggle with on a daily basis—namely that the process of change must occur at all levels, on different time scales, and in a manner that is open-ended such that obstacles arise along the way that require agile learning to navigate them toward success.

    All of this can be elucidated with an “applied memetics” approach that builds knowledge of the factors and pattern-formation mechanisms that shape real-world cultural evolution.  I look forward to exploring this more with you in the months ahead!


    Joe Brewer
    Change Strategist for Humanity

  3. Dennis D. Embry says:

    Dear David,

    Welcome to the road-rash of schools, which I have thousands of scars from the work in the US and Canada. The first point you made is the most difficult one to overcome, especially with encroachment of “big business” of the Broade Academy for superintendents, big publishing companies, and the decline of local control coupled with media scrutiny. It’s as if Ostrum’s principles were bought out by the mob.

    Issues a 2 and 3 are the easiest to manipulate, especially if you give significant control and implementation to the students. Adults may come and go; the kid’s don’t. Now you know the carefully constructed elements of students/children as change agents in all my work—egged on by symbotypes—including symbotypes of belonging so profoundly manipulated in PeaceBuilders and PAX GBG.

    #3 has to happen by group based contingencies, not just individual contingencies. Even better, if they have interlocking contingencies up and down hierarchies that convey status.  Most people don’t think of group contingencies, yet those scream in Ostrum’s work.

    #4 Can only happen with frequent data feedback on progress or success. That brings people under stimulus control. Right now schools play a football game but don’t have any scoreboard feedback until the end of the year or even several years. That cannot shape behavior. That’s why we build in all sorts of feedback loops on a daily basis to regulate behavior.

    2, 3 and 4 are easier to deal with than #1. But if you have 2, 3 & 4 in place, it helps dampen the toxic element of #1