What can the health care crisis, Switzerland, and the US military teach us about network design?

Technology is redrawing our maps of our organizations, our communities, and even our nations along virtual terrain. In the age of large-scale networks, do boundaries really matter anymore? It’s an important question to ask if you are someone who makes executive decisions for your organization. And if you have any part in designing human networks, complexity science tells us that you should consider carefully how you design the boundaries of your networks to encourage or suppress connectivity.

No, we aren’t talking about building any walls. We’re referring to the boundaries between networks, where interactions create unintended consequences. Today, many of the world’s vital systems are global ones. Modern transportation can put you halfway around the world in a day. Our food system is global, tightly linked, and specialized. Our communication systems allow us to collaborate across a global network, making innovation accelerate more rapidly than we’ve ever experienced. The scale of the teams we can build has increased. Even the networked applications we use to produce innovation (e.g., collaboration tools, online communities, big data, new currencies, new organizational designs) are experiencing an explosion of new forms.

As hyperconnectivity becomes more of a reality everywhere, do we really understand its implications?

Dispelling the romance of emergence and connectivity

In Issue 1 of design4emergence, we learned from network scientists, AI researchers, and designers that when networks mix it up at the border of one system and another, you get unexpected novelty — or emergence. After talking with Yaneer Bar-Yam, president of the New England Complex Systems Institute (NECSI), we’re cured of the notion that emergence is always a good thing. “It can be good or bad,” Bar-Yam told me. “Emergence is just something that happens with networks.” So how do you design for the “right” kind of emergence? The answer is “it depends,” because emergence amounts to the unintended and indirect consequences of connectivity.

Let’s go deeper into how connectivity works in networks, because I want to diffuse the Pollyanna mentality we often have about it.

“There are no single causes anymore to big global problems. Connectedness creates complexity.”

— Bud Caddell

There is a general assumption, even if it’s unspoken, that everyone should get along. That diverse perspectives lead to innovation. That we should have melting pot societies and decentralized organizations and global systems where everyone cooperates.

What would that even look like? Probably like the AirBnB you just left in Tokyo for the Starbucks in Anycity, USA.

In reality, the result of a highly connected network is usually a high degree of interdependency. In complex systems, cause and effect are often related in ways that are not obvious at first glance. The effects can be indirect: Push here, and there’s an effect way over there.

In the best cases, a densely connected network results in rapid innovation, such as development of roads urging the spread of civilization itself. The flip side of that coin are the following scenarios:

• Communities whose ideological maps of the world are so drastically different from one another that they cannot (and do not want to) get along
• Epidemics that spread more rapidly now that our global transportation system is so densely networked
• A global food system vulnerable to price shocks
• A banking system that is so highly interdependent that it is susceptible to cascading failure
• An innovation center that must be protected from its organization in order to bring about change
• When the systems we design need to be effective at multiple scales

The scale-complexity trade-off

Bar-Yam and I address the first five scenarios in our interview in this issue’s podcast. It’s the last scenario however, the scale-complexity mismatch, that trips us up most in our business ecosystems. You can think of scale in terms of the number of parts that have to work together in a tightly coordinated way. Efficiency and high quality at multiple scales is a lot to ask of a system unless you understand the trade-off between scale and complexity.

The US military figured out that trade-off a long time ago, and began to understand it more formally through a complex systems lens after Vietnam, where traditional strategies failed.

“Improving any organization’s performance involves assessing and dealing appropriately with two aspects of the system’s capabilities and tasks: scale and complexity.”

— Yaneer Bar-Yam

Bar-Yam advised the Navy on applying complex systems tools and concepts to specific military concerns.

The overall efficiency or agility of a military organization is its ability to build teams to handle conflict at different scales. The key designing principle is that the complexity of the force applied must match the complexity of the environment (e.g., the terrain, the scale and complexity of opposing forces, etc). In other words, don’t bring a tank to a sniper fight.

In general, higher-scale forces correspond to lower complexity and more coordinated actions (ex: a fleet of ships), while smaller, more complex arenas (shorelines with civilians and complex terrain, guerrilla warfare) call for independent, highly trained individuals — aka special forces.

Questions to ask when determining scale of effort:

• How many people have to work together to perform it?
• At what distance can it be distinguished as a separate action?

Matching scale and complexity gives us the power to evaluate different organizational structures for different scenarios — whether we need a specialized systems analyst assigned to each client account, for example, or if we can get away with a streamlined call center to handle all support calls.

Trying to perform all the tasks using the same organizational structure is a bad idea.

Higher complexity generally means that there are more environmental demands to respond to, and the system’s parts need a greater degree of independence to respond to them. So the military has separate forces to match the conflict at the appropriate levels of scale and complexity.

Expecting efficiency and complexity at every scale is too much to ask of a system. The health care system, for example, is expected to achieve financial efficiency at the large scale and complexity at the fine scale of doctors treating patients. The growing administrative layer — billing systems, managed care, co-ops — has made the system complex at the large scale, where it should be simple. Turbulence in the financial flows from employer to insurer to provider creates conflicting priorities and threatens quality of care. An outrageous number of people are killed each year from medical error, the third leading cause of death in our country.

‘Organizational turbulence’ is when ‘a coherent flow is broken up into many smaller flows.’ What works at once scale doesn’t always translate to another.

Health care has historically dealt with its own growing complexity by attempting to streamline processes all the way down to the finest scale of care. This typically makes problems worse. That’s because physicians are the “special forces” of the health care system, uniquely trained to administer quality care in the form of complex and individualized treatments.

Functionally segregating a network by scale of behavior reduces organizational turbulence.

By applying efficiency for large-scale tasks and complexity for complex tasks, Making Things Work suggests that we could draw a boundary that separates the large-scale repetitive demands on the system from the complex, specialized ones and reduce system turbulence.

Inoculation, nutrition, wellness services, disease screenings, prenatal care, and minor issues such as allergies—these are services that could be handled at the population scale, Bar-Yam suggests, through employers or public health initiatives. We can relieve the burden on physicians by separating these parts and allowing doctors to focus on individualized care — in effect, forming two distinctly separate health care systems.

“Apply efficiency for large scale tasks, and complexity for complex tasks, these are the key lessons.”

— Yaneer Bar-Yam

The biggest problems in complex systems arise from the entanglement of scale and complexity and the misapplication of central control. What complexity scientists have learned from studying the health care crisis, military operations, and territorial conflict has implications for us as business practitioners and network designers to use boundaries wisely.

Most organizations need a variety of network structures to contend with the scale and complexity of the tasks they must perform. The structure must match the task and the terrain. As network designers, keeping these lessons in mind allows us to design interactions at the network scale. We can think past the false dichotomy of decentralized distributed systems and control hierarchies and create living networks that adapt and breed resilience. Boundaries help us match the structure to the task and to design teams with various degrees of independence and distributed coordination.

“Some missions call for completely coordinated actions, whereas others call for multiple partly independent actions,” Bar-Yam said. Giving your teams more autonomy doesn’t necessarily lose you any centralized control, just as the brain still regulates an entire network of distributed organs made of self-organizing cells. If you put the boundaries in the right place, we really can all get along.

References

Bar-Yam, Yaneer (2004). Making Things Work: Solving Complex Problems in a Complex World. NECSI, Knowledge Press.

Rutherford, Harmon, Bar-Yam, et al. (2014). Good Fences: The Importance of Setting Boundaries for Peaceful Coexistence. PLOS One. Web.

Gorski, David (2016). “Are medical errors really the third most common cause of death in the US?” Science-Based Medicine. Web.

Allan, Marshall and Pierce, Olga (2016). “Medical Errors Are No. 3 Cause Of US Deaths, Researchers Say.” Heard on NPR Morning Edition, Web.