Complexity: Is the Presidency Doomed?

The President of the United States is both the executive in charge of our Government, and the symbol of our country’s power.  The president is in charge of a complex network of departments and agency designed to implement policies for the public good.  Increasingly, the President is facing complex crises. Responding to these crises means dealing with an internal labyrinth of career public servants and political appointees, as well as an external maze of opposing politicians and their expert opinions, media, industry and international concerns.  Each of these “agents“, as they are referred to in complexity parlance, is influenced by the behavior of other agents. That is what makes the system complex.

Before the President can take decisive action, he must receive advice from hundreds of reasoned but diverging opinions, some focused on the short-term, others on the long-term, each seeing the world with a different eye, and each weighing risks and uncertainty differently.  Decisions then must percolate through the system, be interpreted and followed through.

The oil spill crisis is a good example.  The public and the media demand that the President take decisive action to stop the oil from gushing from the bottom of the ocean.   The President must turn to his internal agents – the Coast Guard, the Department of the Interior, “Heck of a job” Brownies, and now HHS.  These agents have limited resources, and must comply with slow moving rules and regulations.  The President must then take into account the immediate and future impacts on the other stakeholders in the system.  Either the President tries to take action – with little or no impact on the complex situation or he chooses not to act at all (anticipatory regrets).

Ironically, the power of the Presidency is founded on the perception that the President can take decisive action and make an impact.  He/she is the most powerful person in the most powerful country on Earth… This is the shape of things to come:  If the President cannot take decisive action and react effectively to solve our most complex issues, is the concept of the Presidency doomed?

The Oil Spill and the Edge of Chaos…

Yesterday, I talked about the BP oil spill and the relationships between the parties involved from a complexity point of view.   Another interesting notion that applies to the spill is the concept of the  “edge of chaos”.  

We will all agree that the oil production and exploration field is a complex system.  This system evolves over time, as technology advances, as drilling goes deeper, as the price of oil fluctuates (based on supply and demand), and as the regulatory environment responds.   The behavior of this system can be characterized as adaptive (see Wikipedia for detailed information – although I do not agree with many of the assertions made there).  In adaptive systems, participants are referred to as “agents”  interacting with one another.  Transocean and BP are agents, and the stubborn oil field spewing under the sea is another.

Complex adaptive systems (CAS) evolve along a path, or multidimensional surface.  Where the system goes may not be predictable, because each agent is pursuing a behavior that has not, or cannot, be totally modeled because our lack of knowledge, incapacity to model or randomness.  For instance, the behavior of the oil well as it was capped could not be modeled and therefore was not anticipated.  Why is that?    Each agent seeks to “influence” system behavior through their own behavior and reaction.  BP, it is alleged, sought to maximize profit and accelerate the project by short-cutting safety.   This created the potential for instability by increasing the probability of accidents, and brought the system as a whole closer to the  “edge of chaos”.

Agents seek to push the system from areas where it is safe and predictable (but costly) to this edge of chaos where the system best performs.   Over the edge, the system is thrown into a chaotic and unpredictable state.   The signs of this chaotic boundary at the oil well were there: failed tests, forewarning incidents, which altogether should have warned BP and their partners that they we were stepping into dangerous territory.  Now we are left with a very unpredictable situation, not knowing if this top-kill procedure will work… until hopefully, soon, we can jump over this edge of chaos, and get the system back on safer grounds.

See the live oil well cam here

Oil Spill and Complexity…

Gail the actuary wrote a detail post (The Gulf Deepwater Oil Spill – Was Complexity a Factor?) where she explores the implications of complexity on the oil spill situation in the Gulf of Mexico.   She said:

When there are many employees and contractors with partial responsibility, it is all too easy for things to slip through the cracks.

Actually, one could argue that it may be more difficult for things to fall through the cracks when there is a third-party relationship involved.   The legal contract between the two companies should (and likely is) clearly specifying the scope of work and the respective responsibilities of the parties involved.   These terms of reference are the basis for the relationship, and they outline clearly what needs to be done, by whom, and who is responsible.  The contract between parties makes the system more reliable and reduces complexity by making outcomes more predictable, and reducing the level of integration in the system.

The difficulty comes when one party exerts pressure on the other, in violation of the agreement.  On one hand, (Wall Street Journal “BP Tries to Shift Blame to Transocean”, BP asserts that Transocean clearly had operational control over the rig. Yet, a rig worker, Mike Williams told “60 Minutes” that “The communication seemed to break down as to who was ultimately in charge,” and that BP ultimately imposed its point of view (to go faster) over the objections of Transocean.

This is likely when the situation got more complex. The system is composed of may parts. Complexity depends on two dimensions: how variable the output of each part is, and how integrated the system is (how much each part’s outcome influences the other parts.   When BP allegedly stepped over its bound by overriding and overtaking Transocean’s responsibilities, it introduced in the system additional complexity by twisting the system and coupling decision making that should have been decoupled.

So, Gail is right, complexity is a factor in this crisis.  The division of labor is a good thing. It reduces variability in each part of the system by clearly identifying expectations and outcomes, and the relationship between parties.   Complexity jumps when stress is applied to the system (pressures to go fast, failures, inconsistent test results = actuators) that resulted in violating the expectations embedded in the contracts.  This increased complexity likely contributed to breakdowns in decision making (confusion over who is in charge)  which probably contributed to the accident.

Looking forward to your comments.

Herding sheep: it’s complicated! No, it’s complex :)

Someone (my brother) asked me the difference between “complicated” and “complex”.  I touched on that in the post “What is Complexity?”  Let’s try another way, using the example provided by Michel Cotsaftis in “Complex Systems and Self-organization Modelling“:  Dogs herding sheep.

The dog has two strategies in order to herd the sheep to its destination:

1. The dog can target each sheep individually, chasing each one of them in the direction of the destination.   The dog targets a sheep, and engages into a “bilateral” relationship with that sheep trying to get it home.  This will not work (or may take forever) because the speed and direction of each sheep will affect the sheep around it in the herd.  That’s why good herding dogs will go to Strategy #2.
2. The dog makes moves that affect the herd as a whole, understanding that each moving sheep will affect the other sheep. In so doing, the dog builds a model and projects in its head how the herd will react before it makes a move. Then it may correct or make additional moves until the herd gets home.

As you may have guessed it, the first strategy views the problem as “complicated“, while the second views the problem as “complex”.  It is the interaction between the parts (how one sheep moving affects the others) that makes the problem complex.

Take a look at this video to see what I mean:

My Complex Lunch…

To someone interested in complexity, everything seems like a complex system.  Take my lunch for instance… I sampled the delicious Vifon Pho Ga  Chicken Flavor Vietnamese Pho Rice Noodle Soup pictured to the side.  Its preparation was anything but complex.  But the $2.50 lunch was itself made of 10 distinct parts:  A cellophane wrapper, a bowl with lid and separate label on top, dried rice noodles, a soup base pack, a dried vegetable pack, a flavor oil pack, a hot sauce pack and a plastic fork.

The list of ingredients was also mind-boggling: Rice Noodle: Rice, Salt, Sugar, Guar Gum,Vegetables Pack: Textured Wheat Gluten, Eryngium?, Leek. Oil Pack: Palm Oil, Chili, Artificial Chicken Flavor. Soup Base: Salt, Sugar, Monosodium Glutamate, Artificial Chicken Flavor, Onion, Disodium 5 – Inosinate & Disodium 5 – Guanylate, Ginger, Caramel. Chili Sauce: Chili, Garlic, Salt, Sugar, Vinegar, Corn Starch.

Whenever I look at food in front of me, I can’t help but think about how many ingredients are involved, and the complex production and supply chains behind each of them.  How did it get this way?  What forces have let us to such complexity?  I scoured the web but have not yet found a  relevant discussion of what drives food complexity.  Perhaps readers can suggest some sources?

In a complex world, Vifon and other prepared food manufacturers can no longer control their entire production and supply chain chains, worrying about sources of supply, consistency of taste and appearances, and food safety. It’s too complex. They trust and rely on other manufacturers to isolate these complexities from them by buying soup packets and oil packets produced to their specifications.   The network forces driving producers to evolve this way must be very strong to end up with such incredible solutions.  What is amazing to me is that Vifon and others manage to produce, distribute and sell this 10-part product for less than $3.00. Most of the cost must be in the packaging, by the way.

Modeling Complexity

“If a tree falls in a forest and no one is around to hear it, does it make a sound?“… This riddles raises some interesting questions about observation and the knowledge and understanding of reality.   Philosopher George Berkeley postulated that ” …the objects of sense exist only when they are perceived“

Likewise, we may ask: “If we do not reveal the complexity of a system, is it really complex? ” This conundrum suggests that we need to perceive and understand a system’s behavior before we can appreciate its complexity.  Understanding a system’s complexity has been a focus of the field of mathematics since the mid-thirties when British mathematician Alan Turing came up with the so-called Turing machine as a way to describe the behavior of algorithms in a computer.

Photo by Spike 55151 on Flickr

Our models of reality always fall short of describing accurately the behavior of complex systems.  We are hoping that we can capture and explain the most significant behavior of the system so that we can predict the system’s behavior with relevant accuracy under changing conditions.  Different people or organizations with differing vantage points may create different models of the complex system, with different rules, yielding differing results.  Planetary warming is a good example of a complex system where different modelers arrive at different results.

As our knowledge, mathematical sophistication and ability to develop detailed representations grow, our models of complex systems become more accurate and thus more closely describe the behavior of their real counterparts.   “Around 650 B.C.”, for instance, “the Babylonians tried to predict short-term weather changes based on the appearance of clouds and optical phenomena such as haloes.” (NASA)   “With the formation of regional and global meteorological observation networks in the nineteenth and twentieth centuries, more data were becoming available for observation-based weather forecasting.”  Today, weather observations are collected and used by large computer programs using physics and fluid dynamics to simulate temperature, moisture and wind, and accurately predict weather changes.  In 650 BC, the Babylonians used their understanding of the relationships between cloud appearance and future weather to predict the future.   To them, weather variety could not be understood and modeled accurately, and changes in weather could not be predicted effectively.  Today, we understand better the complexity of weather dynamics (changes in temperatures, wind and humidity)  because of our ability to represent this natural phenomenon more accurately.

Now it is time for lunch…

Measuring Complexity

In my previous post, I touched briefly on the definition and meaning of complexity.  There are two important characteristics of complex systems: variety and connectivity (Francis Heylighen).

They are the yin and the yang of system complexity.  Variety describes the degree to which each part of a system can change, on its own.  Variety provides the potential for differentiation and change, which creates complexity.  The other dimension of system complexity is connectivity.   Connectivity creates the potential for integration, where parts or subsystems influence one another, creating complexity.

Other measures of complexity are sometimes used, such as numerousness (the number of parts), but the number of parts is not necessarily a mark of complexity if these parts are just replicated with limited interactions with one another.  There are interesting discussions in linguistics for instance, about the complexity of language and its implications on society.

Measures of complexity describe the behavior of systems, rather than their static structure.  We can think of many examples where this is true:  The World Economy, for instance, depends on each country’s economy, which will change based on local policies and circumstances.  The economy of each country is linked to others through global trade.  If one country suffers from economic flu, it is likely that others will be affected too.  Systems with low variety and low connectivity are simple. Those with high variety and connectivity are complex.   This is essentialy a networked view of the World, describing complex systems in terms of the behavior of their nodes (parts, subsystems) and links (connectivity, integration).

Variety and connectivity. Why is this important?  This conceptual construct helps us approach complexity and prepare ourselves to deal with it.

What is Complexity?

A definition of Complexity will tell us that the concept applies to things intricate or complex. Most definitions apply to a specific domain such as biology or cybernetics.  These definitions generally fall short of what we think today is important about complexity:  A connected system of parts,  requiring specific knowledge, and likely hard to understand.  In “Complex Systems and Self-organization Modelling”,  Michel Cotsaftis contrasts the meaning of “Complex” from the Latin “cum plexus” (tied up with) with the word “Complicated” from the Latin “cum pliare” (piled up with).

The notion that something complex is made up of connected parts implies that the parts interact with the whole and are inseparable from their whole.  For instance, the recent news about a BP oil well that blew up in the Gulf of Mexico brings to mind the complexity of oil exploration and production, tying together the economy of oil exploration, exploration and production technology and the world of incident remediation.  Each of these areas is itself a complex system.

The word complexity also brings the notion of order.  Without order, a system would be chaotic, unruly, and impossible to define.   With a complexity, we assume, at least, that some measure of order can be uncovered and understood.   Complexity also depends on our point of view.  From far away in space, our world looks like a dot.  Yet as we come closer, we start to appreciate the complexity of its geography, its ecology and its inhabitants.

More to come…

Hello world!

Welcome to my new blog on complexity…  The purpose of this blog is to explore the notion of complexity, how it affects our lives and how it affects and will affect society today and tomorrow.  Complexity spans many scientific fields.  I am primarily interested in how complexity affects decision-making, objectivity, and our ability to progress as the human species.