11th August 2020
By: Andi Mirviss, Programme Manager, North America, and Josh Sorin, Program Director, City Innovation, North America, Centre for Public Impact
This summary was originally published on the Centre for Public Impact website, where you can read the full ‘How To Fail (Forward): A Framework for Fostering Innovation in the Public Sector’ report.
Well before COVID-19 presented local governments with some of the greatest obstacles of the modern era, public institutions across the United States faced issues already rapidly increasing in their complexity and magnitude.
Many sought out innovation as a means to address these issues. While the word “innovation” is often associated with complex experiments that only a handful of people with specialised skills get to participate in, we believe innovation means anything that makes things better for people.
At its core, all innovation, from common process improvements to complex experiments, starts with failure. In its most fundamental form, this process entails identifying that something has not gone according to plan (i.e. that a failure occurred), and learning from that failure (or failing forward).
Innovation begins to take shape through the testing and iteration of new ideas to determine what works best — and learning from the things that do not work is as critical as learning from what does.
Despite its importance in the innovation process, learning from failure is very difficult. It is challenging to admit to oneself that a failure occurred and more challenging still to admit it to others. In government, where the stakes are higher and the resources tighter, talking about failures can be an anxiety-inducing, stressful experience. So stressful that many avoid it whenever possible. And yet — failing forward is critical to a government’s ability to innovate, and therefore to make things better for people. With this in mind, this project considers two basic research questions. Why is it so hard to learn from failure in government? What can leaders and their allies do about it?
Our findings are principally informed by on-the-ground research with the six local governments, 20 departments, and over 150 public servants who participated in Failure Foundries. These day-long workshops enabled government employees to identify internal challenges to learning from failure and develop action plans to address these challenges. We supplemented this research through interviews with 25 city leaders from all over the country and a literature review on relevant topics, including organisational learning, the psychology of failure and human learning systems.
We argue that workplace culture is the critical determinant of a government team’s ability to learn from failure. Try as we might, we found no silver bullets or simple solutions to promoting learning from failure in local government. Instead, our research revealed four primary opportunities to develop a workplace culture that promotes failing forward in government.
Our new report, How To Fail (Forward): A Framework for Fostering Innovation in the Public Sector, details how cultural and structural barriers limit failing forward within these four opportunities and proposes potential strategies, largely devised by failure foundry participants, to break down the barriers.
Mindsets and beliefs
Acknowledge that failures are already occurring in the status quo and view failure as a necessary step towards positive change.
Governments typically operate under the popular adage that failure is not an option. However, this mindset contributes to a deeply embedded fear of failure that discourages those in government from identifying (or acknowledging) problems and from engaging in the experimentation necessary to develop new solutions.
Both of these actions are critical to failing forward. First, governments must accept that failure is already occurring in the status quo, or else they are not likely to improve it. Second, because failure is inherent in the innovation process, public servants must view failures as learning opportunities.
Collectively, these two core beliefs can establish a baseline for a government that promotes failing forward: taking stock of what is not working fosters improvements in programmes and policies that are already in place. Accepting that failure is part of the innovation process empowers governments to identify failures when they occur, enabling them to safely learn critical information that informs subsequent iterations.
Foster internal teams and relationships rooted in psychological safety and empowerment across all levels.
Leadership is essential to building a fail forward culture. Leaders have the most critical leverage point to create the kinds of human relationships organisations need to fail forward. As the most visible members of the department, they set the tone and have power to create and enforce structures that maintain this culture.
Despite their unique positions, leaders face substantial barriers to realising their visions. Particularly strong barriers include reshaping a culture after staff had negative experiences with previous leadership, regular turnover that leads to change fatigue, and sceptical middle management.
Leaders can attempt to break down these barriers by developing closer relationships with staff built on trust and modelling the behaviour they seek to foster, such as being vulnerable about their own failures, providing support and coverage when failures occur, and promoting learning opportunities whenever possible.
Team cohesion is critical to failing forward. Learning from failure is an interpersonal process; it requires questions, honesty and collaboration to make improvements. Our research revealed that many government teams do not have the psychological safety (e.g. the confidence that one will not face professional or personal retribution if they discuss challenging topics or mistakes) necessary to engage in this process. Working to increase trust within government teams is essential to enabling failing forward.
Systems & Processes
Redesign internal systems and processes to promote identifying, learning from, and taking action about failures.
Sharing power with staff can motivate them and incentivise failing forward. Hierarchical decision-making power in governments can demotivate and disempower frontline staff from addressing the failures they notice. Frontline staff might know the programmes they implement do not work for residents, but without inclusion in decision-making processes they often feel it is ‘not their place’ to raise issues to management. Involving staff from all levels in developing and improving existing policies and programmes can help to ensure those with the greatest knowledge of residents’ experiences are empowered to make necessary improvements.
Rebuild performance management processes to promote learning and development. Existing performance management processes are typically designed from the outside-in. Departmental goals are designed to meet externally ascribed metrics (often coming from the state government), and staff is evaluated based on how well they meet individual benchmarks based on these metrics. Such a system diminishes the capacity for failing forward because many metrics do not capture the complexity of government work. This can cause public servants to focus on activities instead of outcomes and prioritise some issues at the expense of others. It also can limit the capacity for holistic learning. Additionally, externally imposed goalposts can diminish intrinsic employee motivation. When professionals are not able to apply their expertise to define what is most important, they can lose their desire to go the extra mile to affect change. Emerging research in ‘human learning systems’ provides promise for new performance management structures.
Break down silos within and across departments to improve communication and learning. Government silos pose physical and social barriers to failing forward. Many government teams that work on the same topic area or cover the same physical space do not have effective mechanisms to speak with one another if something appears to fail. These physical barriers can reinforce social ones; separate teams can develop animosity or mistrust of others, particularly if they do not understand how each other operates. Developing approaches to poke holes in, rather than completely tear down, these silos (e.g. quarterly meetings for teams that focus on similar challenges or cross-departmental process walks) are essential to promote learning.
Build learning from failure into policy design and organisation operations. Despite the ubiquity of failures, government teams do not typically include learning from failure into programme design or existing programmes and processes. This can make it challenging to identify failures when they occur and to find time to reflect on what is and is not working. To that end, government teams should intentionally weave learning from failures into systems, processes and programmes. This might mean holding group brainstorms to discuss what failure could look like at the beginning of programmes, carving out ‘sacred time’ for reflection, or hosting quarterly ‘pivot parties.’
Reshape the narrative and ecosystem to be supportive of local government innovation.
Build trust with residents to create opportunities for learning. Public mistrust and low civic engagement have long been challenges for local government. Such distrust can inhibit failing forward because public servants operate with a baseline fear of sparking public outcry. This limits the potential to test out ideas or communicate learnings from things that do not go according to plan. Innovation methodologies that require public servants to meet and test out ideas with the public in low-stakes environments, such as human-centred design, can be helpful paths forward.
Redefine relationships with local media. While local government and the media share an essential partnership in promoting government transparency and public awareness of local issues, pervasive negativity bias can limit government capacity to experiment and learn from failures. Public servants fear that when the media reports on things that do not go well, they are not likely to discuss lessons learned along the way. This diminishes interest in trying new things, and it perpetuates the desire to maintain the status quo. Finding opportunities to bolster learning-oriented narratives in local press might ameliorate these anxieties.
Amend federal and state funding and accountability mechanisms. Many local governments rely on federal and state grant funding for programming. Much of this funding comes with tight accountability mechanisms and little flexibility for experimentation and learning. This forms a major barrier for a local jurisdiction’s ability to fail forward: it simply may not have the budget to try something new. If it does have the budget, accountability mechanisms can cause public servants to try things that are guaranteed to meet certain metrics but may not solve the root problem. While there are no simple solutions to this challenge, federal and state agencies should consider adopting the principle of subsidiarity, in which those with the most proximate local knowledge have the greatest authority in decision-making and spending.
In responding to COVID-19, local governments are innovating at an unprecedented pace. They have proven they are capable of generating new ideas, fostering resident trust, and making a way out of seemingly no way, all at a speed that many — including themselves — thought improbable. Critically, they have shown resilience in the face of ambiguity and high-stress situations, finding new methods to work together across teams, departments, and even jurisdictions. Government leaders must develop cultures to ensure this capacity for innovation sticks beyond the current pandemic. We hope this research serves as a jumping-off point for the hard work local governments can do to develop the fail forward cultures they seek. Some of the strategies listed in the report might be helpful places to start, and we are eager to support future work — regardless of whether it fails the first time.