Systems change! Just saying the words aloud makes me feel like one of the cognoscenti, one of the elite who has transcended the ways of old-school philanthropy. Those two words capture our aspirations of lasting impact at scale: systems are big, and if you manage to change them, they’ll keep spinning out impact forever. Why would you want to do anything else?
There’s a problem, though. “Systems analysis” is an elegant and useful way to think about problems and get ideas for solutions, but “systems change” is accelerating toward buzzword purgatory. It’s so sexy that everyone wants to use it for everything. And even if it’s becoming one those phrases that means whatever the user wants it to mean, funders are calling for it, so the doers have to scramble to look like they’re doing it.
It doesn’t help that there are likely no two words in the English language with a broader definition than “system” and “change.” Once there is too much divergence on the meaning of a phrase, it ceases to be very useful as a way to drive concerted action. Variants can add richness if they coalesce back into a useful definition, but more commonly they create confusion: once the multiple-definitions horse is out of the barn, it rarely comes back. See “impact investing.”
But when you rummage through the growing literature on systems change thinking, there are in fact a few recurring themes. One is the need to tackle the root causes of any problem you take on. Another is that a broad coalition must be assembled ASAP. Finally, the most salient theme is the notion that the systems involved are transformed as a result of the work (although in many of the examples I read about, it’s not articulated clearly just what system is being changed).
Taken individually or as a whole, these themes point to some of the ways in which systems change is a less-than-ideal paradigm for the work we need to get done:
1. It’s too hard to know to what degree systems change is or isn’t happening. It may be the case that “not everything that matters can be counted,” but most of the stuff that matters can, and it’s hard to get better at something if you’re unable to measure it. But these words of a so-called expert on systems change measurement are typical of what I’ve seen in in the literature: “Measuring systems change is about detecting patterns in the connections between the parts. It is about qualitative changes in the structure of the system, about its adaptiveness and resilience, about synergies emerging from collective efforts—and more…”
Like I said, it’s too hard to know to what is or isn’t happening.
2. “Root cause” thinking can—paradoxically—bog down progress. “Root cause” analysis is a common feature of most systems change discussions, and it’s a wonderful tool to generate ideas and avoid unintended consequences. However, broad efforts to tackle all of a problem’s root causes can turn anything into a complicated, hard-to-replicate project. It can also make things look so overwhelming as to result in a kind of paralysis. And however successful a systems change effort might be, that complication makes it hard to replicate, and you’re often stuck with a one-off project.
There are good reasons why a lot of creative (and scalable) solutions bypass “root causes” entirely. Sometimes root causes are intractable. For example, a prominent root cause of many problems in the places where we work are dysfunctional governments, and no one is going to be able to fix them any time soon. To be useful, we have to find solutions that will work in spite of that root cause. Many good ideas do an end-run around some root cause that isn’t going away, and any number of successful solutions pay no attention to root causes at all. A good solution is like a fish that can thrive in the sea in which it swims, however murky and polluted that sea might be.
3. Systems change orchestration is fraught, at best. Another theme in the systems change literature is that somebody has to assemble a grand coalition to make it happen. But who is that someone? It’s a difficult role: you have to be acknowledged by others and prove yourself in the role. Too often that leads to a vacuum, and, as a result, funders often take over the role. It’s irresistible. We can’t help ourselves. Especially when we buy into the systems change paradigm, we feel compelled to step into the breach with our money, and since that money is the lifeblood of the doers, most have to dance to our tune. We become the de facto conductors. And we’re not very good at it. See “strategic philanthropy.”
4. Mostly we’re using a system, not changing it. Most of the most high-impact ideas I’ve seen use existing systems as platforms or channels for distribution of a product or service. They’re not changing governments or markets, they’re just using them to drive impact. It’s more like they’re leveraging systems or—in the most benign sense—exploiting them. You might be bolting something onto a system that makes it deliver something important but when you step back and look, it’s the same old system you started with. Most times we don’t really change the systems in any detectable way. Just because you managed to get a government to issue a sensible policy or even launch an effective program doesn’t mean you changed it. You used it—admirably, effectively—but that’s all.
So is there a better way? Sure! Call it “Scalable Solutions.” The phrase makes cameo appearances in discussions of systems change, but it actually represents a fundamentally different approach, and it’s worth taking that difference seriously. Scaling is more than when you add two more clinics to the one you have now; that’s just “growing.” If a “scalable solution” is one that has the potential to make a big dent in a big problem, then “scale” is the distant dream of that solution achieving its full potential, and the solution is “scaling” when the curve of impact over time steepens dramatically, even exponentially, in a sustained way.
What’s great is that the phrase “scalable solutions” has thus far been mostly confined to the business world. We still have a chance to define it tightly enough to be useful. So here we go. A scalable solution consists of five things:
- A Big Idea that drives and organizes it.
- A Mission that focuses the idea on a specific outcome.
- A Theory of Impact that articulates the connection between the idea and the mission (the basic mechanism to achieve that outcome).
- A Model that lays out a systematic and replicable way to apply that theory.
- A Strategy that identifies who is going to replicate the model at really big scale and who is going to pay for all that replication. Call them the doer-at-scale and the payer-at-scale.
Replication of that model is what’s going to grow your impact, and so the model itself has to be inherently scalable. It needs to be “enough” in four ways:
- Effective enough: Nobody should try to scale up a model without strong evidence of impact, and effect sizes matter too: a 3 percent increase in income, or a 4 percent in literacy might not be worth bothering with. “Statistically significant” does not equal “meaningful.”
- Big enough to matter. It addresses a big problem, and the overlap between where it is needed and where it would work needs to be big enough to make a significant dent in an important problem.
- Simple enough that the doer-at-scale can do it. It needs to be systematic and replicable, yes, but that alone isn’t enough: Whether that doer is government, businesses, or NGOs, you need to find persuasive examples of that doer doing a decent job of something comparable.
- Cheap enough that the payer-at-scale will pay. Whether customers, governments, or Big Aid, every payer—from a mom buying a stove to a finance minister who will decide the fate of your health worker idea—has a price point. You have to figure out what that price point is and hit it.
Here are several examples of promising scalable solutions, all of which came from distinct organizations and are now scaling:
- Professionalized community health workers: governments are both doer and payer at scale.
- Pay-as-you-go home solar energy: businesses are the doer and customers are the payer at scale.
- Ultra-poverty graduation programs: governments will ultimately be both doer and payer at scale (with NGOs & Big Aid providing a step along the way).
- Teach At The Right Level: a simple classroom practice that improves the learning odds for every kid.
Oh, and this is critical: there’s no such thing as a scalable solution without one or more scale-obsessed organizations to drive it. Solutions don’t go anywhere on their own and a model is just a model: it’s execution that creates impact. The social sector does not function like a market for impact—you can’t dump an idea in the water supply and expect it to flourish. For the most part, if you have an idea, you own it, and if you don’t drive it toward scale, nobody will.
While a systems change approach makes it hard to know to what degree change has happened, and where you are in the process of change, successful scalable solutions follow a common pathway on their journey to scale. That means you can demonstrate progress with reasonable precision. That progress is captured in terms of stages:
- R&D: You’re a lab. You started with an idea and you iterate as rapidly as possible to turn it into a systematic, replicable model. It’s done when you can make a persuasive case for impact and that the model is scalable (though it’s still too early to be fully rigorous).
- Replication: You’re increasingly a factory, churning out replication of your model while you continue to iterate on it. You’re growing, not scaling, while you work out all the operations, the systems, and the monitoring and evaluation that will make your model maximally replicable. You can move on when you’ve made a rigorous case for both scalability and impact, using RCTs and/or other rigorous evaluations. This stage is mostly about what people refer to as “direct delivery.”
- Scaling: You’re mostly involved in the generation of other factories, of replication by others. You may still be a doer yourself, but you’re increasingly focused on recruiting other doers, and helping them replicate the model at high quality.
The Big Shift
The sort of dramatic and sustained uptick in impact that constitutes scaling—what we call the Big Shift—usually happens via some combination of six factors. Those Big Shift factors comprise a kind of checklist:
- Doer at Scale: When you’ve progressively handed off the heavy lifting to the putative doer-at-scale, be it government, businesses, or other NGOs. This should be the central obsession of any long-term strategy.
- Payer at Scale: When there has been a transition from philanthropy (which includes concessionary debt and equity) to your payer at scale, be it government, Big Aid, or customers. Because doers mostly do what they’re funded to do, this has to happen in conjunction with the transition to doer-at-scale.
- Model: When you figure out how to make the model simpler, cheaper, more broadly applicable and/or punchier in terms of impact.
- Tech: When reach is extended, transaction costs lowered, and efficiencies achieved as organizations find ways to use emerging tech, especially mobile and digital.
- Policy: Things start to look systems change-ish when, as organizations grow, they come to understand the ways in which changes in policy can open doors and lower barriers. In short, when they learn how to make top-down changes that will accelerate all that bottom-up momentum. At the same time, this is when the organization has gotten big enough for policy makers to notice them and care what they think. If the organization is smart, they’ve gotten to know and have involved those policy-makers very early in the journey.
- Collective Action: Finally, here’s the part that should make systems change aficionados happy. When organizations start scaling their solution, they often form networks and/or collaborations with other doers pursuing similar ends. They may join an existing movement, or even try to start one. Sometimes organizations with complementary, but distinct, solutions join the circus.
Wait…did you just say systems change?
As you start to make some serious scaling progress, it does start to look something like what people talk about in the systems change literature. But the critical difference is that it is emergent, driven by scale-obsessed doer organizations. They drive it in a more organic way that answers to their ambitions and needs as they go all-out for maximum impact in the lives of those they serve.
It’s always good to have an example: The progress of Professionalized Community Health Workers is a beautiful story of a scalable solution. Over the last four decades, countless numbers of community health worker (CHW) programs were launched as a way to get primary care to people who lack access to doctors and clinics. In the end, most of these CHW programs failed to deliver because they relied on volunteers who were insufficiently supported and chronically under-supplied. But over the last decade or so, a number of organizations—increasingly well-known actors like Last Mile Health, Muso, Partners in Health, Integrate, Living Goods, and Amani—have developed various permutations of the common professionalized community health worker (PCHW) model. All of their health workers are:
- Salaried and certified: They are treated like professionals.
- Skilled like professionals: They are well-trained, and work proactively to provide doorstep care, treating a broader range of conditions (Digital tech is playing a big role).
- Supported in the field: Systems are set up to ensure that they are systematically managed and are visited and mentored on site in their communities (digital tech here, too).
- Supplied: They have the medicines and materials they need, reliably (and digital tech here, as well!).
The professionalized community health worker is an eminently scalable solution (and yes, it’s nice that it can be captured by things that all begin with “S”). It can be described simply:
- Idea: Professionalized Community Health Workers.
- Mission: Save Lives.
- Theory of Impact: “If we put PCHWs in the field, they will deliver quality doorstep care, and lives will be saved.
- Model: “4 S’s”
- Strategy: Doer = Government; Payer = government
And it checks all the scalable solution boxes:
- All the variants take the form of well-described systematic models.
- All these organizations have recognized that this will only get to scale with government as doer and payer (sometimes backed up by Big Aid).
- There is a lot of emerging impact evidence, some of it spectacular.
- It’s big enough: There’s no reason to believe it wouldn’t work with most governments that are trying to provide primary care.
- It’s simple enough: Most governments currently run stuff at least this complicated.
- It’s cheap enough: Costs are in line with what most of the target countries are paying for primary care overall. Many countries are realizing that the impact and reach of PCHWs justifies the reallocation of resources and BIG Aid is increasingly interested in helping governments pay.
Armed with a replicable model, virtually all the protagonist organizations have followed a recognizable version of the “journey to scale” playbook: They’ve iterated on their models to boost scalability, embraced digital tech in just about every conceivable way, and become effectively involved in national and regional policy-making.
Perhaps the coolest thing is that they self-organized into a meta-organization called the Community Health Impact Coalition (CHIC). CHIC is emergent, brought into being by the doers themselves. It is wisely organized around a set of explicit principles for an effective PCHW model, and if you want to join up, you need to sign on to them. It’s got a staff now and has become a vehicle to pursue a shared research, policy, and funding agenda. It’s also become a way to share best practices, and has attracted supporting organizations like Medic Mobile (digital health tools) and Village Reach (optimized supply chains). A wonderful side effect of CHIC was that the members realized that funding is not a zero-sum game. They started introducing the others to their funders, which had the effect of attracting more money overall.
So that is the story of professionalized community health workers, a solution that is really taking off in Africa and beyond. Scalable solutions and systems change approaches can eventually land you in similar places, but the example of obsessed organizations scaling their models in multiple countries demonstrates how much speed and efficiency scalable solutions can achieve. And with most scalable solutions, it is the doer-at-scale that matters the most. And we’re not going to fundamentally change those doers any time soon, be they low-performance governments, inefficient markets, or organizations caught up in the dysfunctional BINGO/Big Aid nexus. We have to use them even if we can’t change them.
Am I saying that I think we shouldn’t try to change systems? Hell no! it’s noble and necessary. It’s just that we’re doing ourselves no favors by conflating the two. Sometimes, with wrenching and prolonged effort, we do manage to fix a big dysfunctional system, and it’s essential that we keep trying. However long it takes, we do need to transform systems like healthcare, energy, education, governance, and industry so that they actually work for the benefit of all. Ultimately, all scalable solutions need a top-down enabling environment to achieve their full potential.
In the meantime, though, we need to change lives at big scale. We have a chance to make big things happen with scalable solutions, and conflating that with systems change isn’t helping. It might even be that a focus on scalable solutions will make real system change happen that much faster, and spectacular things happen when bottom-up momentum meets a top-down accelerant. But in the meantime, we have to use that which we’re not yet able to change. You must swim in the sea that you’re in, not the one you wish you were in.
This article was originally published by Stanford Social Innovation Review on April 8th, 2021 with the headline - We’re Beating Systems Change to Death