Sustainability and Resilience

A philosophical diatribe to start your week.  I have no idea if this makes sense--but I'm sure you'll let me know in the comments.

I've always had a problem with the idea of sustainability.  I'm not opposed to sustainability of environmental and economic systems (who would be?), I just haven't had anyone give me a definition of sustainability that I can get my mind around.  Every definition of sustainability that I've seen requires that we define the path of consumption in terms of the unknowable future.  But if the future is unknowable, how can we dictate the present?

For that reason, there has been a fairly recent move away from sustainability as a working concept for economic and environmental systems toward resilience.  But is resilience any better than sustainability for decision making?

Resilience, as I understand it, is the ability of complex systems to absorb shocks and keep going.  Or put a little more precisely, the ability of complex systems to "resist disorder."  Unfortunately, that's about as precise a definition as I can give. 

Growing from engineering systems--how do we define production systems to be flexible enough to absorb changing economic and business conditions?--the ideas behind resilience are starting to be applied to economic and environmental systems.  The simplest question resilience asks is:  Is the economic/environmental system capable of absorbing unanticipated shocks?  But there are two fundamental problems that make this question impossible to answer:

1) If the shock is unanticipated, it is impossible to plan for.  It is logically impossible to design a system that is capable of handling all unanticipated shocks.  If a shock is unanticipated then it cannot be planned for, unless we always plan to avoid the worst possible outcomes. Sounds an awful lot like the precautionary principle.  A loose application of the precautionary principle would have us design a system capable of handling the worst imaginable consequence, not matter how unlikely.  But resilience takes it a step further--plan your system to avoid the worst unimaginable consequence. 

2) Resilience is self-fulfilling.  An example might help.  The digital revolution have been held up an example of resilience worthy of investigation.  In 1970, no one could have possibly anticipated the impact of computers and the Internet on everyday life.  Therefore, the system is resilient.  An unanticipated shock that the system absorb and kept chugging along.  But, can we name a shock that the system can't absorb?  Can't we always find a definition of the outcome that claims the system was resilient?  Or do we need a target?  The system is only resilient if it maintains a path defined as ___________.  But isn't that just the definition of sustainability? 

I'm not exactly sure where I'm going with this, but I'm pretty sure it helps to convince me that environmental economists are on the right track.  Consider all knowable costs and benefits--economic and environmental, today and in the future--and make the best decisions we can.  In the presence of an unknowable future, I see no alternative but to make decisions based on the knowable, and then watch from beyond as future generations adapt as our unknowable becomes their reality.  Isn't that how it has always worked?

Comments

[Don]

Dear Tim:
Why don't we focus on a few examples in which the ecology suggest a lack of both sustainability and resilience. Dead zones of the oceans, many caused by nitrogen pollution from farming on land lead to nearby seas without oxygen. I guess you could say that corn farming on the banks of the Mississippi is sustainable as long as the fertilizer trucks keep arriving. What are your thoughts about the sustainability and resilience of the biota in the gulf of Mexico?
Don

[odograph]

And haven't ocean reserves shown more resilience than managed fisheries?

[Ironman]

To draw from engineering, if you wish to design a sustainable system, or at least articulate one, you need to define your known limits (described by statistical variability) where failure might occur, then focus your system design so that it operates well within those limits.

What you're describing as resilience has its counterpart in engineering as the concept of margin of safety or factor of safety. Here, let's say you have a material that you wish to perform a given function (say a steel beam that you want to use to distribute the load of a heavy objects that you transport overhead).

You have a well-defined description of how the material behaves under given loading conditions (see Figure 2), which defines various points at which things happen. Since you don't know what loading conditions the beam will see in the future, you design it to operate well within the limits described by one of these known points of failure. In this case, the beam would be designed to handle a load that represents 20% of its maximum strength that it can handle within its known range of elasticity (the range at which the beam will, once loaded, return to its original shape without permanent deformation).

Why 20% of that maximum potential loading stress? Because that level has been established over time as providing a reasonable margin of safety - it allows for that unknowable future of how that beam might be loaded. The more you know about how the beam might be loaded (how well controlled that loading might be applied, for instance), you can increase that percentage limit within you may operate.

In some cases, you might be willing to go outside the material's limit of elasticity to operate at the margins to gain whatever benefit you might achieve by doing so. Of course, you do so at increased risk, but you'll have to accept that you will never be able to restore the beam to its original shape if you do load it above that limit.

Does this help? Would an example from public policy economics be useful?

[Tim Haab]

I'll address more comments later, but Ironman, with the beam example, you have a very specific goal--maintain the shape of the beam (or structural integrity)--what is the economic/environmental system equivalent?

[nonlineardifferential]

From the sustainability point of view the economic/environmental system equivalent looks like a society with compact communities that are highly social with relatively equal income distributions that extracts minimal resources from the environment and dumps as little waste into the environment as possible.In other words the integrity of the society depends on healthy human relationships with other humans and the environment. Relationships that can sustain happy lives and not cause irreperable damage to the system over the very long run. This is not a vision that is defined well enough to quantify and compare outcomes. To an economist it's totally useless, it certainly cannot be evaluated against the efficiency metric, and to libertarian types it also means that someone is imposing and engineering a society where their vision dominates (assuming not everyone holds the sustainability vision in mind).
I am bothered by the vagueness of the sustainability camp, but it is clear that human behaviour is doing irreparable damage to the world's ecosystems that may result in outcomes that are detrimental to our well-being. I don't believe that the efficiency metric is sufficient to capture the potential losses, and perhaps we need to take a closer look at the ecosystem models to see if there is a way to quantify some of the uncertainty and bring it under the umbrella of risk.

[Larry Sheldon]

Lotta words for a simple concept.

A sustainable system is one that does not consume any resources from the set of resources I wish to go unconsumed.

See how simple that is?

[gormk]

"Every definition of sustainability that I've seen requires that we define the path of consumption in terms of the unknowable future. But if the future is unknowable, how can we dictate the present?"

Are you serious? What's the point of having discussions about climate change mitigation policies if it is a requirement that (states of) the future is (fully) knowable to formulate policies for the presence?

I recently went to a very inspiring seminar on sustainability based on the following paper:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1004637

I doubt we're moving away from using the term sustainability. The problem is the opposite. Scientists have not shown enough discipline in using the term in its originally intended macro-welfare preserving sense. Instead, the term has proliferated in applications (sustainable production, sustainable cities/communities, sustainable buildings, etc., bla-bla-bla).

Part of the problem is that economists have not taken strong enough ownership over the concept and its formal, official definition. To me, sustainability should be the ultimate welfare-evaluation criteria, a replacement of Pareto-optimality.

[Ironman]

Thinking about it, a good starting point for economics/environmental applications would be to consider the supply of X, plus or minus observed variation, over a rolling time period for an appropriate and discrete geographic volume. As an example, that could be number of blue crabs counted at regular intervals and tracked over rolling ten year periods for a given inlet on Chesapeake Bay.

The variability that you would observe over time would, when combined with what you know of the inputs (birth rates, food supply, pollutants, disease microbes, etc.) and outputs (natural life span, harvesting, etc.), give you a sense of what the effective elastic limit (long-term stable population) would be for that defined volume. That, in turn, is something that can be used to meter policy (catch limits, crab farming, allowable environmental discharges, etc.) for the affected geographic volume.

[Tim Haab]

Gormk,

I just want you to be very clear as to what you are proposing to "sustain," and then sit back and watch as it is sytematically picked apart with examples of every possible contingency, unlikelihood and unforeseeable event.

Personally I think Larry Sheldon nailed the definition: "A sustainable system is one that does not consume any resources from the set of resources I wish to go unconsumed."

Tim

[Ironman]

"Personally I think Larry Sheldon nailed the definition: "A sustainable system is one that does not consume any resources from the set of resources I wish to go unconsumed."

Ah, environmental inventory safety stock....

[joshua corning]

Economies are not designed they are emergent.

One tendency of emergent systems is that they are more sustainable and resilient the less you plan for them.

My example would be DNA based life.

Have a nice day.

[David Zetland]

Read this and this.

[Alan Dove]

A system's "sustainability" or "resilience" isn't quantifiable or measurable without some additional parameters for time and scope. Nothing is perfectly sustainable. The overwhelming majority of species that have existed are now extinct, and for the future, I wouldn't place geologically long bets on anything bigger than a bacterium. Even that might be a stretch in a few billion years, when that star sitting eight light-minutes away finally vaporizes the oceans.

Larry Sheldon's (rather circular) definition above is about as clear as this issue will get.

[Josh]

"A sustainable system is one that does not consume any resources from the set of resources I wish to go unconsumed."

Sounds a bit postmodern to me.

[john]

Joshua

Sure social systems are 'emergent'in some sense, but they are also self aware, which means design can, and does have a function. Have you read any history at all?