How to: harness complex systems for social change

Most systems we care about are complex – and devilishly difficult to redesign. But, if we can't be the 'architects' of new systems, can we be the 'gardeners' that tend them? Social change strategist and former GuideStar boss Jacob Harold explains how, in an extract from his new book The Toolbox: Strategies for Crafting Social Impact.

Jacob HaroldJust about every aggregation relevant for social change is some kind of a system. Health care delivery, higher education and criminal justice are systems. The energy economy and the political environment are systems. Sometimes big change happens when we restructure the order of the system itself: labour unions restructured labour markets, social media restructured the public conversation about gender identity, and organic certification restructured entire supply chains. 

When considering social systems, we can start by looking for a simple explanation by mapping out the components and interactions of a system. If we’re lucky, we might identify a simple intervention. Maybe we just need to put more money into social media advertising to get people to come to our art fair. Maybe the best way to help people without resources is to give them money. 

Other times, the system may be understandable – it’s just really complicated. In a complicated system, it may in fact be possible to disentangle all its parts. But it will take patience and resources to address those parts one at a time. 

We can’t just rebuild a system from scratch to match our perfect vision. What we can do is tend to, alter, and care for the system

Most of the time, though, the systems we care about are complex. The whole is fundamentally different from the sum of its parts. We can’t just rebuild a system from scratch to match our perfect vision. What we can do is tend to, alter, and care for the system. Or, to put it another way: even if we cannot be architects of a complex system, we can be gardeners.


Building high-performing systems

Systems analysis is relevant from both the inside and the outside. Consider a police department with a history of discriminatory practices. Activists might be trying to change the department from the outside – just as a new police chief is trying to change it from within. Or people inside and outside a dysfunctional legislature may wish to transform its procedures. Both activists and pharma­cists may wish to transform the system for funding prescription drugs. 

In general, we want a system to show the emergent characteristics of complex­ity: learning, responsiveness, adaptation, resilience and efficiency. Below, I’ll suggest six ways to build – or reno­vate – high-performing systems. 

This might seem abstract, so I’ll try to ground this in a concrete example that is close to my heart. Over the past two decades, many people – myself included – have worked to make the system of philanthropic giving in the US function better for nonprofits and the world.

Already, that system has seen sig­nificant changes. Consider, for example, the proportion of American donors who proactively consider several different nonprofits before giving (instead of simply saying yes or no to individual requests). According to Camber Collective’s “Money for Good” research, that proportion tripled from 2010 to 2015. If extrapolated to the country as a whole – admittedly a big “if” – this finding would suggest a change in how 12m people give away US$15bn each year. Additional evidence suggests that increased transparency can bring huge increases in donations – as much as a 50% increase – to nonprofit organisations.

No matter how we choose to interpret the quantitative data, there’s been a clear shift in the public discourse about giving in the US. There are rising expectations for transparency and results reporting from donors, government officials and the media. And, as importantly, there’s been an unmistakable call for all stake­holders – notably nonprofits and their beneficiaries – to be treated with dignity throughout the process.

So, below, I offer six general lessons for interacting with complex systems. In each case, I show how this played out in the work to improve the system of philanthropy, some of which I was involved in while working with Hewlett Foundation, GuideStar and Candid.


Lessons from philanthropy 

1. Be intentional about connectivity 

The relatively weak connectivity of many systems dampens the ability to respond to the needs of the outside world. To increase connectivity, perhaps the simplest thing to do is to increase the ease of connection – or, in simple terms, the convenience for the end user.

In philanthropy, we sought to do this by (a) standardising interactions (eg, providing donors’ information in a predictable format) and (b) making information available near where actors are already operating (eg, adding information about nonprofits into donors’ online finance interfaces).

2. Simplify the rules 

If all interactions among components follow their own separate rules, it is hard to achieve any systems-level efficiency (ie, there are no economies of scale) or cumulative learning (ie, components are con­stantly reinventing the wheel). 

In philanthropy, we sought to simplify rules by (a) creating and adopting common standards for information transfer (eg, grant applications and reporting) and (b) removing barriers to the transfer of money.

3. Support feedback loops 

Feedback is a prerequisite for learning: try something, see what happens, and adjust. Weak feedback loops make for weak learning.

In philanthropy, we worked to build feedback loops between (a) foundations and grantees (eg, through surveys like the Center for Effective Philanthropy’s Grantee Perception Report) and (b) nonprofits and beneficiaries (eg through the Fund for Shared Insight). 

4. Encourage goal orientation

When com­ponents in a system have clear goal ori­entation, they are more likely to adjust behaviour and optimise goal achieve­ment. This also can – consciously and unconsciously – make it easier for components (people, organisations, or ants) to align and aggregate behavior.

In philanthropy, (a) advisory groups like Bridgespan worked to support nonprofit strategy processes, and (b) Candid sought to gently guide nonprofits towards goal orientation by prominently highlighting organisations’ goals, strategies and metrics on the GuideStar platform.

To get people to engage in a shared purpose or common pro­tocols, you may first have to address what people mean when they say, “we”

5. Embrace a common identity

In social change, the parts of a system are often people, and those people may or may not see themselves as part of a community. To get people to engage in a shared purpose or common pro­tocols, you may first have to address what people mean when they say, “we”. At its best, ‘community building’ can be far more than a feel-good activity. It can have critical strategic consequences if it clarifies a constitu­ency’s identity and shared goals.

In philanthropy, (a) we began with a recognition that the tax code offered a ‘boundary’ that defined the group for both legal reasons and shared identity, which provided a very useful initial frame, and (b) we also recog­nised the blurring boundaries across sectors, so we engaged with partners in business and government who did not share the same legal status. In an age of multiple bottom lines, we needed an expansive and explicit sense of shared identity: the common purpose of building a better world.

6. Enhance the flow of information through the system

For people to work together, they each have to know what the others are doing. Often, there is an opportunity to improve the flow of information among different nodes in a given system. 

In philanthropy, much of our focus was on flow of data about nonprofit organisations. We sought to (a) build on existing data (the financial information available from the Form 990) with new data (questions about programmes, operations, and governance) while also (b) supporting the flow of that data through data standards, new pricing models, and technology like APIs (application programming interfaces).


The community of social change is growing in scale, impact and complexity. With insights from complex systems science, we can think clearly about how we – members of a global community driven to shape a better world – might become greater than the sum of our parts. For that to happen, we need to move beyond a view of only our individual parts and see our collective efforts as a single, immensely complex organism.


Top photo by Christin Noelle on Unsplash

Thanks for reading Pioneers Post. As an entrepreneur or investor yourself, you'll know that producing quality work doesn't come free. We rely on our subscribers to sustain our journalism – so if you think it's worth having an independent, specialist media platform that covers social enterprise stories, please consider subscribing. You'll also be buying social: Pioneers Post is a social enterprise itself, reinvesting all our profits into helping you do good business, better.