Complex Systems

In Public Health, Purpose-Built Solutions are the Worst Kinds of Solutions


Bee populations have collapsed for mysterious reasons around the world, but few places have been as hard hit as the Maoxian region of China. There, farmers must now lean small ladders against pair trees, climb up into their branches, reach upwards with slender rods, and pollinate every flower by hand. 

Pollinating flowers by hand is an example of a purpose-built solution. The farmers have a problem and they are devoting specific time and energy to solving it. Googling the words “purpose-built solutions” reveals that the words are almost universally regarded as positive. Companies brag that their products deliver “precise specifications” that are “right for you,” just like the farmers’ rods are right for every flower. Googling the words “side effect” brings up nothing good — mostly the unintended consequences of drugs. And yet, I suspect the farmers were happier when bees pollinated flowers as a side effect of gathering food.

There is a general principle at work here, and it offers a critical lesson for public health. Trying to achieve something directly cannot compete with thousands or millions of people achieving it as a side effect of something else they are already doing. Consider the kind of accurate, real-time traffic data that is available thanks to millions of people simultaneously using Google Maps, and compare that to a traffic engineer counting cars once or twice a year. For farmers and traffic counters alike, the purpose-built solution creates two kinds of cost: it will never accomplish the goal as effectively, and resources must be devoted to trying, whereas the side effect would be free.

Diabetes Canada aims to reduce the number of diabetes cases in Canada by 500,000 by 2020. How should it approach that goal? Traditionally, they should brainstorm actions, such as an outreach campaign or coaching programs, then select the most cost-effective, and hire people to implement them. It makes sense intuitively: accomplish goals with solutions designed to accomplish goals. The approach is also destined for failure.

A study analysed the results of 122 interventions in which professionals tried to coach people to eat better or exercise more, and it finds they only have a small effect size. While billions have been invested in public health campaigns over the decades in the United States and elsewhere, rates of physical activity have dropped by a third and obesity has tripled.

In one intervention, the Finnish Diabetes Prevention Study, nutritionists and clinicians met regularly with 522 participants, organized voluntary walking events, and offered free personalized training over three years. The project successfully reduced the risk of contracting diabetes by 36% over a seven year period, as compared to a control group. While these results are encouraging for those 522 individuals, they are less encouraging for society overall. The project was expensive enough to require support from one government ministry and three foundations. For the project to continue helping people this way, it would need to continue receiving funding year after year, and that for only five hundred people.

In contrast, another tool was found to have a similar impact on preventing diabetes, but this one costs nothing, and has this impact year after year for a thousand times more people. That tool is called “the neighbourhood.”

In Toronto, residents who live in places where they can walk to a variety of destinations have a 33% lower risk of developing diabetes, as compared to people living in places where driving is the only option. And these neighbourhoods have this impact for half a million people. None of them were designed to prevent diabetes, and neither were their bakeries, grocery stores, offices, bars, and cafes. Yet the combination of decent sidewalks and businesses makes walking a good, attractive, daily option, preventing diabetes as a mere side effect more productively than any government program could ever hope to. It is worth noting that many of these neighbourhoods were built at a time when diabetes was a rare disease.

Direct public health interventions may nonetheless be worth the money for the people they help, but they will not create progress for society overall if the way cities are designed do not support health as a side effect.


The Trouble with Implementing Side Effects

But if these side effects are more powerful than purpose-built solutions, we have a problem: governments make decisions based on solutions, not side effects. Governments frequently, for example, put hospitals on cheap land outside of cities. In such places, hospitals can perform their primary function — treating patients — sufficiently well. Ironically, however, these are often the worst locations for supporting public health goals, such as encouraging people to walk, bike, or take transit to the hospital. (Using transit tends to substantially increase physical activity, because it involves walking).

In a recent case in Halifax, Nova Scotia, the city’s Chief Planner publicly criticized the province’s decision to put a hospital in a business park, where “community building” and providing adequate transit would be “nearly impossible.” Nothing will surround the hospital but enormous parking lots, dangerously fast roads, and big box stores. While doctors in the hospital will work on curing diseases, its location will actively undermine efforts to prevent them.

A frustrating asymmetry discourages governments from prioritizing side effects. The location of one hospital will, by itself, have a tiny impact on the total amount people in the province walk, and so it seems irrelevant next to the project’s own priorities, such as having enough land to erect a large building. And yet, if the government follows the same logic for all buildings, the impact on healthy behaviour will be enormous: the tens of thousands of people who use and work in all hospitals, schools, and government offices will walk and bike less every single day.

In fact, if buildings fail to support physical activity, it is unlikely that any amount of amount of public health funding can ever make up for the deficit. Farmers cannot fully replace the work of thousands of bees. Ad campaigns and aerobics classes cannot replace the amount that people would walk or bike every single day if hundreds of buildings were conveniently located near homes.


Solving the Asymmetries and Mismatches

For governments to make policy based on side effects, they must overcome two problems. First, those with the authority over projects tend not to have the mandate to care about the side effects. Second, the system-wide impacts of all projects seem inconsequential for individual projects.

The two problems point to a simple solution: analyse the system impacts to define mandates. That’s it. If building locations is critical for supporting physical activity, then anyone making those decisions must necessarily have the mandate to support physical activity — no matter their job title or department. No progress can be made in any other way.

One hurdle for this solution, however, is that decision makers can only prioritize so many things. A hospital could support an infinite number of good ideas, such as providing space for community gardens or special parking spaces for autonomous cars. A mandate cannot prioritize infinity. We need a principle to separate those responsibilities that are irreplaceable, and those that are just good ideas.

To define that mandate, the word “irreplaceable” is instructive. From the perspective of the orchard, bees have the role of pollinating, because nothing can fully replace them. We can therefore say that bees have this “system role.” Other good ideas, in contrast, can be accomplished elsewhere just as well. Plenty of properties can be a good home for a community garden, so the person in charge of the hospital project does not have the “system role” to create one. He or she does, however, have the system role to support walking.

The idea of a “system role” can clarify the ethical responsibilities of professionals in all kinds of fields. In The RainForest, Victor Hwang and Greg Horowitt point out that it is far more productive when business people generate trust with every business transaction than to rely on contracts and courts. Courts are purpose-built solutions that cannot fully compensate for millions of shady transactions, and so each business person has the system role of supporting trust, if they want to enjoy prosperity. Architects and developers have the system role of designing buildings that make streets inviting and attractive, because if they don’t, nothing can. News organizations have the system role of holding government publicly accountable. Plants have the system role of producing oxygen. These roles may not figure explicitly in these job descriptions, but from the perspective of the system, we need them to play that role. Nothing can fully replace them.

To create a society in which every sidewalk, street, building, park, and workplace consistently support health and wellbeing, we cannot make progress unless every decision maker has the mandate to fulfil their system role. A bureaucrat making the decision for where a building is located can no longer that they do not have the mandate to support walking and biking. From the perspective of the system, and to improve the health of millions, they necessarily have that role.

Realigning mandates to match system roles will require a wholesale transformation of government responsibilities. We should get started.


- Tristan Cleveland

Photo credit: Full Spectrum Biology.