Complexity science helps us make better decisions—from detangling trade webs to preparing for climate crises.
By Susan Beauchamp
Sep. 28, 2017
In 2010, major wheat-growing regions of eastern China suffered a drought, causing the failure of vital crops for two years. But it wasn’t the people in China who suffered most from this disaster. Instead, it was nations downstream in the trade network that depended on the food they sourced from China.
One of these countries was Egypt. With the wheat scarcity caused by the drought, the cost of bread rose far above the means of the average worker. Demonstrations about food prices soon fomented into riots against the government and its leaders as a whole. People took to the streets chanting “Bread, freedom and social justice!” Eventually, the Egyptian economy collapsed and President Hosni Mubarak was ousted.
These activities were some of the earliest, catalyzing events of the Arab Spring, a series of protests and uprisings across North Africa and the Middle East.
Surely, no one could have predicted that crop failure in China would spark a revolution on the other side of the globe. Or could they? Shade Shutters and his colleagues believe that complexity science can help us track and contain potential crises before they spiral out of control.
“Our argument was that the transfer of those socio-political vulnerabilities is through trade networks, and if we can analyze those trade networks, we might be able to anticipate future hot spots of instability,” says Shutters, a research scientist at Arizona State University’s Global Security Initiative.
Detangling the data
Researchers have always relied on data to validate their work and support their conclusions. But the volume of information available to researchers today is exploding. Because of advances in technology, more data was created from 2013–2015 than in all of human history up to that point. By the year 2020, about 1.7 megabytes of new information will be created every second for every human being on the planet. While these advances make far more information available, the challenge now is to make it more valuable.
Enter the relatively new and emerging study of complexity science, which enables researchers to organize and analyze seemingly random facts into useful decision-making tools. Complexity science capitalizes on recent advances in technology and revolutionizes our capacity to gather, store and make sense of massive amounts of information, or “big data.”
There is an evolutionary arms race in all ecosystems. And in human-generated systems, that sweet spot is always moving — just like it is in nature."
Increasingly powerful computers, along with highly sophisticated algorithms, have allowed researchers to transcend traditional linear thinking, which examines the cause and effect of a limited number of variables. Today’s models and simulations can explain phenomena and predict outcomes from interactions between the interconnected and constantly changing elements that make up a complex system.
“If you want to understand the human body and how it functions, you could take classes on just the kidneys or just the bones, and that's good to know all the facts. But at some time, you have to step back and put it all together to really understand and be a good medical practitioner. Complexity science has that holistic approach to any system,” says Shutters.
New problems need new science
Complexity science can help us address problems we’ve never faced before. One example is what to do when entire countries disappear because of climate change.
The Republic of Kiribati spans 33 coral atolls in the Pacific Ocean, halfway between Hawaii and Australia. The average land elevation in Kiribati is just six feet above sea level. If sea levels continue to rise at their current rate, the entire country will be submerged within 60 years.
In anticipation of this, local leaders have purchased land on the island of Fiji, and Kiribati citizens have already begun to migrate to their new homeland. But there is already conflict. The Fijians want the new arrivals to observe the governance and laws of Fiji, while the Kiribatis intend to re-establish their area of the island as a separate and sovereign nation. Additional questions about access to jobs, education and food for the increased population are among the issues that will inevitably arise as more Kiribatis are forced to migrate.
“Throughout history we have had populations that have been temporarily displaced because of famine or war, and eventually those people assimilate or return to their home country. But what do you do when you know an entire country is going to disappear and is not coming back?” asks Shutters.
Computer models can predict these kinds of problems and offer potential solutions before they become a crisis. By asking the right questions and understanding the facts, the violence, poverty and civil unrest that may result from an influx of refugees can be alleviated, if not averted altogether. Already, Anote Tong, the president of Kiribati, in conjunction with other nearby nations that will be most heavily impacted by the disappearance of these islands, has launched a “migration with dignity” policy to forestall these issues.
Shutters is quick to point out that complexity science is not about finding the one “right” answer, but ensuring that stakeholders have all the facts they need to make informed decisions.
“Sitting with people on opposite sides of a debate and spelling how things should be done — I don't do that. My goal isn't to find an optimal decision, but to highlight all sides of the issue clearly and objectively,” he says. “My goal is to make more information available so the debates that do happen in the public policy arena are well-informed. I like to know that the people making the decisions have all of the facts.”
An evolving discipline
Complexity science is so new that even its name is still evolving. The terms “data science,” “complexity science” and the now less-popular term “informatics” are sometimes used interchangeably. And no one knows just exactly how to define it.
“Complexity science isn’t really one discipline. It's a response to reductionism in science gone to its extreme. It takes it back the other way to take a holistic view of systems,” Shutters explains. “I’m particularly interested in applying complexity science to human social systems, but it could be a biological ecosystem, or it could be a global system of trade and diplomacy. Anything that you can define as a system that has intelligent components that evolve and adapt, that makes it a complex system.”
Surprisingly, Shutters’ formal education is not in statistics or computer science. He holds a PhD in biology from ASU. But Shutters views his background in the natural sciences as a fortuitous fit to his work in informatics.
“I feel privileged to have done doctoral work in ecology because I see there is an evolutionary arms race in all ecosystems. And in human-generated systems, that sweet spot is always moving — just like it is in nature. There is always a trade-off and that trade-off is always shifting,” he says.
Although some universities have begun to offer degrees in complexity science, ASU has set a course to incorporate it across all disciplines. A floor in the renovated Hayden Library will become a state-of-the-art data science center, providing students and faculty with opportunities to conduct research that would have been too difficult or too expensive without the benefits of advanced computers and the methodologies of data science.
“Data science is one of those weird academic fields where right now no one really knows if it’s a methodology or if it's going to be its own discipline,” says Shutters. “At its heart, it's the same thing scientists have asked for hundreds of years. ‘How does the world work?’ ‘What are the secrets of nature and how do we uncover those?’ But it is a slightly different way that they approach it. It’s very complex, but it’s kind of a cool problem to have.”
Shutters' research has received funding from Skoll Global Threats Fund, the National Science Foundation and the U.S. Department of Defense.