///

Using PDIA to tackle educational inequities in Brazil

5 mins read

Guest blog by Cesar Nunes, IPP 2021

It was a joy to know that the new edition of the Implementing Public Policy (IPP) course would be offered in 2021. I knew about the Problem Driven Iterative Adaptation (PDIA) approach and was already following the Building State Capabilities website and YouTube channel. I had already read and watched all of their online materials: books, papers, podcasts, and videos. At the end of 2020, I identified PDIA as a path to overcome lots of difficulties experienced in the Brazilian states I was collaborating within the public education sector.

The COVID-19 pandemic in Brazil was quite severe. Schools were closed for a long period, solutions based on technology were increasing inequities, and we were and still are amidst a national education reform. When the 2021 IPP Online started, we were already working for 5 months on the challenge of introducing hybrid learning with reduced inequities. We decided to combine PDIA with other approaches to reach scale and consider the specificities of the educational sector. The course came at a moment of the implementation process early enough to allow redirecting activities and efforts, and brought a lot of insights and contributions. Having a real and enduring process happening provided concrete meaning to tasks and reflections proposed along the course.

The course starts with discussing complexity and assuring that for complex problems processes must be designed and implemented accordingly. Education, consequently educational policy, is one of the most complex endeavors one can think of. Raising a child is usually used as an example of a ‘complex problem’ in comparison to sending a rocket to the moon that is ‘only complicated’. Many of the problems at the educational policy level can be treated by laws, logistics, training for using protocols, and monitoring using defined metrics according to a defined curriculum. However, the most sensitive and important parts of contributing to the development of about thirty students in a classroom, and a few hundred students in a school, cannot be reduced to protocols. It depends on the judgment, decisions, adaptability, and knowledge of teachers and principals. They must act in a discretionary way, especially if we want to reduce inequalities. We have wicked hard problems at the street level bureaucracy. Since PDIA is designed to develop state capabilities, our aim was to take it to this last level in the usual hierarchy, overpassing or coordinating with the many layers of plan-and-control already consolidated in educational systems.

Education itself should be seen as an ‘infinite game’—a moving target. The history of educational reforms demonstrates that what could be defined as the purpose of education nowadays won’t be suitable anymore in a few years. Education perpetuates a culture. At the same time, it creates conditions to modify it. There are as many views of education as there are world views. Modern and complex societies that acknowledge diversity as positive and necessary must undertake education as preparation and contribution to such a rich world.  In this sense not only students learn, but teachers must also become lifelong learners, and schools and educational systems must become learning organizations. Again, this challenging demand fits the very aim of building state capabilities for solving complex problems.

Technology in education has a long tradition of computer-supported collaborative learning and the use of visual tools for making thinking and learning visible. In the process of constructing and deconstructing the problem, we used Knowledge Forum (KF), a collaborative environment developed over the last thirty years as an international collective effort led by Carl Bereiter and Marlene Scardamalia from the University of Toronto and the Institute of Knowledge Innovation and Technology. The collaborative tool was used by a team of representatives of all departments from the central office of a state secretary of education after getting authorization from the secretary of education, the highest in the hierarchy. Following the PDIA approach, this team was the first ‘snowflake’ and constructed the problem: The necessary heavy use of technology for remote and hybrid learning introduced during the pandemic is increasing inequality and turning education more limited and inefficient than it was already.

The same collective and visual digital environment was used to deconstruct the problem by creating the fishbone with many hands writing, editing, and moving ‘bones’ at the same time. Due to the restrictions of social distancing, all the work was done at distance and all zoom sessions were recorded. The process generated discussions, reflections, commitment, and collective knowledge building. Everything got naturally registered/documented and was used for sharing knowledge, legitimacy, and expand authorization. Revisiting and editing the fishbone the team in a very engaging process added the triple-A to identify entry points and, in another view (a new blank space in the collective digital environment), filled the change space.

Contributions from the IPP Online course on leadership, teaming, relationships, and the Portuguese version of the PDIA toolkit were crucial to bringing quality and increasing the acceptance of the process. The existing culture is based on the use of a yearlong three-tier PDCA (plan-do-check, act for central office, districts, and schools) with a unidimensional indicator – student score on tests – to redirect actions. As expected, resistance to the changes was present all the time.

While building capabilities it is important to consider that education itself is a science devoted to promoting learning and building competencies that have had extraordinary advances in the last decades. Knowledge generated by Harvard researchers from the educational field was considered while subsidizing the design of solutions (external solutions in the change space): Fernando Reimers´ Five dimensions for changes in education; Chris Dede’s Design-based Implementation Research that incorporates iterative processes; Ron Ritchhart, David Perkins and Daniel Wilson’s Learning Innovation Lab, Making Learning Visible, Thinking Routines, and Creating a Culture of Thinking.

Inspired by the work from Michael Fullan and Andy Hargreaves on the role of districts, Dave Snowden’s Cynefin Framework for complexity and emergent behaviors, and NESTA’s approach for collective intelligence on solving complex problems, some departures from a pure PDIA was proposed to get to scale and tackle wicked hard problems at the frontline. Decentralization by empowering districts was done by letting them choose if they want to work on problems at their level, building teams and using the PDIA iterative tool, or if they want to run the full process, learning PDIA, constructing and deconstructing problems, exploring the change space, and fostering emergence through iterations.

In parallel to the district and central office snowflakes, a community of practice approach was introduced for ‘problem-solvers’ at the frontline (teachers from different disciplines and grade levels, principals, pedagogical coordinators, and educational technology experts). PDIA emphasizes learning in each snowflake during the iterations. Communities of practice based on knowledge-building principles add the possibility of fast dissemination and collective knowledge-building without the necessity to reduce complex wicked hard problems to complicated problems with protocols.  

As a closing remark, I would say PDIA is a very powerful approach that should be introduced carefully recognizing it is a paradigm shift that releases the public sector from a myriad of opportunistic consultants. It is not necessary and may be counterproductive for many kinds of problems or policies. Purely logistic problems can be tackled with traditional plan-and-control methods. Low ambiguity and low conflict policies do not need PDIA. Transaction-intensive services with known and accepted technologies and protocols can be tackled with other methods than PDIA. For such examples well defined metrics and evaluation processes are possible.

Most real problems do not fit into the categories above. When forcing them to fit we reduce complexity at the expense of someone, usually minorities, disempowered, underrepresented. PDIA as an inclusive approach to building capabilities is essential if we want to reduce inequities. For that, besides PDIA, metrics, evaluation, and culture must change.


This is a blog series written by the alumni of the Implementing Public Policy Executive Education Program at the Harvard Kennedy School. Participants successfully completed this 6-month online learning course in 2021. These are their learning journey stories.

Leave a Reply

Your email address will not be published.

Latest from Blog

%d bloggers like this: