When and How Do You Use Systems Thinking? — Part 2

Illustration by Hans Park

By Annie Neimand, Ph.D. and Ann Christiano. Part 1 is available here.

In his insightful article, Human-Centered, Systems-Minded Design, Thomas Both, Director of the Designing for Social Systems Program at the Hasso Plattner Institute of Design at Stanford, tells the story of Jill Vialet (Both, 2018).

Vialet is the founder and CEO of PlayWorks, an organization that helps schools help their kids do their best by improving how they play during class breaks. School leaders often told Vialet they faced a shortage of substitute teachers. When a teacher called in sick, they did not have a reliable pool of substitute teachers to cover the classes. This created a challenge for the schools, as they would have to move teachers around or rush to find a replacement. As a result, teacher absences had adverse effects on the classroom and on the students’ learning outcomes.

Vialet wanted to take this challenge on. Most people working to solve this problem would assume the school simply needed to hire more substitute teachers. But as Both reports, when Vialet studied the substitute teacher system, she found that the problem wasn’t that there weren’t enough substitute teachers. In fact, there was a large pool of people already in the system, but few taught on a regular basis.

Vialet spent time with substitute teachers to identify the root cause of the problem and how to change the system. She found that, “Substitutes felt they weren’t respected or valued by schools, and felt they didn’t have a community of support in schools or among their fellow subs,” writes Both. Schools had to change how they interacted with and supported substitute teachers. This is quite a different intervention than simply recruiting more people. As a result of this work, Vialet founded Substantial Classrooms to help schools improve how they train and support substitute teachers.

Vialet was able to solve this complex problem because she employed both human-centered design and systems thinking. As Both writes,

“For both human- and systems-level challenges, we need to identify the problems worth addressing if we are to create meaningful change. Understanding the right problem, we can better create effective solutions. A very simple characterization of a design approach is that we move from working to understand a challenge, to working on creating solutions in response to the challenge.”

Systems thinking is a methodology for understanding complex problems. It helps us understand the motivations and beliefs of people involved, as well as structural dynamics at play. Knowing the dynamics of the systems we are trying to innovate will help us know where to act.

Innovators should use systems thinking to identify the right problem to solve. In a report, the Humanitarian Innovation Fund shared a five-stage model for innovation (Obrecht and Warner, 2016). This model is similar in nature to those found across the humanitarian innovation sector and follows this five-stage process:

1. Recognition of a specific problem
2. Ideation on solutions to address the problem
3. Development of innovation by creating actionable plans and guidelines
4. Implementation of the innovation to test and compare solutions
5. Diffusion of successful innovation to scale for wider use.

Systems thinking is useful in both phase one and two. However, it’s important to note that not all problems require systems thinking. In a field guide to systemic design produced by CoLab, that describes the work of a team working within the government of Alberta, Canada to support communities in systemic design and strategy, the authors argue that systems thinking and design should be used to solve complex problems like climate change, rather than what they define as “complicated problems” or “simple problems.” Specifically, they suggest using this approach when there is a low level of agreement about the problem, a low level of certainty as to what to do, a high level of unpredictability, or a high level of diverse stakeholders involved.

Vialet’s work to better understand why schools were struggling to build a reliable pool of substitute teachers used systems thinking to identify the root cause of the problem: substitutes felt undervalued.

Here’s the process Jill followed:

Step One: Collect Data
To identify the problem and potential interventions, the first step is to conduct research that will allow you to build a model of the system (Mugadza, 2015, Both, 2018). To collect this data, innovators should conduct ethnographies of the system. An ethnography is an approach to research that requires the researcher to spend time with a community to get a deeper sense of its norms, values and structure.

You can use various methods in doing ethnographic research, including:

• Interviews with key stakeholders to uncover their perspectives and gain a deeper understanding of their behaviors.
• Review of documents that can help you understand hierarchy and processes within the organization. You might also interview external experts and people working in analogous systems to provide a different perspective.
• Observations to understand people in their contexts.
• Focus groups and brainstorming sessions with stakeholders.

It is important to be inclusive of the people who are affected by the system in identifying the problem and considering solutions. As the Humanitarian Innovation Fund writes, “Technical and field staff often know the problems facing humanitarian operations but lack access to reliable mechanisms for raising attention to these or for initiating problem-solving processes. Learning from affected people: A few case studies found problem recognition began with feedback from affected people, as gathered through evaluations.” By bringing those affected into the process, you will likely build rapport with them, and, as a result, they will be more open to experimenting with your intervention at a later phase.

You’re gathering this data to identify the element or parts of a system, how they are interconnected, and feedback loops (see part one for an explanation of these concepts), as well as individual (mindset), cultural (norms) and institutional (structures) factors that make up the existing system.

When a problem has been identified, the Humanitarian Innovation Fund suggests answering the following questions:

• Who sees this as a problem (those who will be receptive to a solution)?
• Who is invested in addressing the problem (and could act as a partner, advisory group members, etc.)?
• What are the underlying causes of the problem?
• What are the end-user’s needs and incentives?
• What were the previously attempted solutions? Why did they fail?
• Which existing practices are relevant to the problem?

Step Two: Map the System
Once you have collected your data, it’s time to put the pieces together to get a bigger picture of the system. The objective of modeling is to identify what the system is and what it must do to produce its current outcomes (Shalhoub and Qasimi, 2005). Lay out the elements of the system, their interconnections, and the feedback loops. They also identify the nature of those interconnections by naming the individual (mindset), cultural (norms) and structural (hierarchies, policies, procedures) factors upholding the existing output of the system.

The objective is to identify causes and effects within the system that are enabling the identified problem. Work with diverse stakeholders to include their perspective of the problem as a gut check on your interpretation and to identify potential interventions they and others will likely adopt.

Now You Try:
Below is one of the many types of mapping exercises you can do to build a more comprehensive view of a system. With a team of diverse stakeholders and innovators, work through these steps.

1. Identify the system you are mapping (i.e. hiring process, organization’s partnerships, operations)
2. List all of the elements that make up the system individually on post-it notes, scratch paper, or on a whiteboard.
3. Draw lines between the elements that have interconnections. Include arrows that tell you the direction of those relationships.
4. Identify feedback loops within the system (balancing, reinforcing)
5. Describe the nature of those connections (individual: mindset, values and worldviews; cultural: norms; and structural factors: policy and procedures).
6. Identify places where the system is producing the problem.

Before you begin the next step, it is important to clarify what exactly you want the system to do that it is not doing now. Once you have mapped your system and have identified the root cause of the problem, your group can identify potential interventions or solutions to address the causes (rather than the symptoms). To begin, create a definition of the problem. This statement should identify the root cause of the problem within the system and the factors contributing to it. This will help your team focus on the problem you are trying to solve.

Step Three: Identify Potential Interventions

Once you have a definition of the problem, you can pinpoint where in the system you will need to implement solutions. This work requires you to identify your points of leverage. Doing this will help you identify the actions that need to be taken in order to change the dynamics and outcomes of the system.

Illustration by Hans Park

Systems thinking scholar Donella Meadows (1997) identified 12 leverage points or places to intervene in the system. She writes, “these are places within a complex system (a corporation, an economy, a living body, a city, an ecosystem) where a small shift in one thing can produce big changes in everything.” These 12 points can be helpful in brainstorming possible points of intervention. Meadows shares these levers in order of least effective (parameters) to most effective levers (power to transcend paradigms).:

1. The parameters of a system (i.e. minimum wage, number of dollars allocated to a program, taxes, etc.)
2. The size of buffers: buffers control how elements (that you can see, feel, or measure) operate within the system. These can be hard to change because they are often physical entities.
3. Structure: structures have an impact on how a system operates, such as creating limitations and bottle necks.
4. Information challenges: when the information is received too slowly or too quickly it can create delays in the system
5. The strength of negative feedback loops: negative feedback loops often help self-correct under different conditions.
6. The strength of positive feedback loops: positive or reinforcing feedback loops often speed up processes.
7. Information flow: who does and does not have access to information, and what kinds of information is being shared
8. Rules of the system: laws, incentives, sanctions, etc.
9. Power to add and change the system: adding new feedback loops, how and what information is shared and with whom, new rules
10. The goal of the system: changing the goal of the system inevitably changes everything. To achieve this goal, there will have to be changes in parameters, feedback loops, how and what information is shared and with whom, and how the system is organized (elements and their interconnections)
11. Mindsets and paradigms: changes to assumptions that form the foundation of the system
12. Power to transcend paradigms: changes to values and priorities that will fundamentally change assumptions with which a system is built upon.

Once you have a definition of the problem and have identified opportunities for intervention, you can begin to brainstorm possible interventions by asking “How Might We” questions, a technique from human-centered design, that allows you to narrow-in on the challenge you are trying to solve.

The trick is to strike the right balance between really big challenges and more specific challenges. For example, in the video below from IDEO, the speaker provides an example of a project he is working on to create new bathrooms in India. He says “how might we” questions like “How might we redesign public toilets in India” is too wide, and “How might we create a door knob for people in India that is clean, safe and invites people in” is too narrow. Instead, he says a question like “How might we create a sense of safety in public toilets” strikes the right balance to help a group focus their brainstorm to discover potential interventions.

Again, it is important to include stakeholders in this process to ensure that interventions are relevant and realistic for those working within the system. According to the Humanitarian Innovation Fund, interventions should consider the needs and preferences of stakeholders, identify interventions that are within their capacity, and include incentives for engaging in the intervention. The interventions that are produced should be seen as desirable, effective and doable by the people who will play a role in their implementation (Obrecht and Warner, 2016). Including them in the process helps you do that.

Step Four: Experimentation

Lastly, once you have identified your interventions, you can prototype and run small-scale experiments to test their effectiveness. When you are able to prove the effectiveness of an intervention, you can begin to scale it up through the organization by sharing the success stories of your prototypes and developing a communications strategy around the intervention. On how to build a strategy to build support for your intervention, read “The Back-of-the-Envelope Guide to Communications Strategy.”

Systems thinking as the norm

Systems thinking is a great methodology for identifying the root causes of complex problems. When we use this type of thinking we move beyond bandaids and quick fixes. We quickly come to see that every system is perfectly designed to deliver its results. Innovation then requires us to identify deeply rooted systems problems to build organizations and solutions that can have an impact.

Works Cited

Both, T. (2018). Human-Centered, Systems-Minded Design.

Meadows, D. H. (1999). Leverage points: Places to intervene in a system.

Obrecht, A., & Warner, A. T. (2016). More than just luck: Innovation in humanitarian action. HIF/ALNAP Study.