REPLY Limits to Growth Plan B (SD7018)

[SDMAIL Jack Harich]

Posted by  Jack Harich <jack at thwink.org>

Tom Fiddaman wrote:
> Jack and others have noted that managing the globe involves two 
> subproblems:
> 1. figuring out the problem (e.g., are there limits, what are they, 
> when do they become binding)
This is what the System Improvement Process (SIP) calls the problem 
definition step.

> 2. convincing others to act (hard when others may be uninformed or 
> predatory)
And this is what SIP calls the change resistance problem.

>
> I'd argue that there's a third problem, which we've tended to gloss 
> over: 3. figuring out how to act. For the climate problem, for 
> example, this is nontrivial. Is it better to allocate property rights 
> in the atmosphere and start trading, or to implement a carbon tax, or 
> start a social movement, or ... ? More practically, as I see it, 
> places like California that are committed (at least on paper) to 
> significant GHG emissions reductions are trying to solve a big messy 
> global problem with regulatory approaches that were designed to handle 
> fairly localized problems (like criteria air pollutants). Policies 
> under development leave much of the problem unaddressed, and could 
> prove counterproductive. It's a shame to waste strong support on 
> measures that are doomed to fail (as zero-emissions vehicle mandates 
> and electric power deregulation did). Then again, perhaps the reason 
> current measures don't go far enough is that #2 has not really been 
> accomplished, so strong support is an illusion. If so, then it would 
> be useful to recognize that the emperor has no clothes, i.e. that 
> current policies are a start, not an answer.
"Figuring out how to act" is what SIP calls the proper coupling problem.

Notice how I've just introduced three terms: problem definition, change 
resistance, and proper coupling. When problems become this big and 
messy, problem solvers need a standard, clear set of terms. This is one 
thing a formal process like SIP provides. I've gone through great pains 
to make each of these terms as unique, self-explanatory, and productive 
as possible.

So let's talk in terms of solving the proper coupling problem, which is 
what your post deals with. Presently, the human system is improperly 
coupled to the larger system it lives within: the environment. If it was 
properly coupled, then feedback loops would exist that would keep 
environmental impact below acceptable limits. This includes all three 
main types of limits: the depletion of non-renewable resources rate, the 
use of renewable resources rate, and the rate of pollution.

>
> I've been working with refiners and agencies on alternative fuel 
> issues. California has policies underway to control things like 
> vehicle fuel efficiency and fuel carbon intensity (none of which 
> tackles the underlying growth drivers, as Jay pointed out). The 
> fuels/energy/air quality crowd perceives that technical options have 
> somewhat limited potential, and that to achieve aggressive goals in 
> the long run will require changes in driving habits and land use. I've 
> been wondering whether the transport/transit crowd in the state shared 
> the same vision, because transport agencies have been conspicuously 
> absent from the energy/emissions conversation so far. On Tuesday, I 
> got the answer at a presentation from a regional transport agency: the 
> region is planning to accommodate a massive increase in driving, and 
> counting on technology to somehow accomplish emissions reductions. 
> Clearly the left and right hands have some coordination to do. Even 
> worse, at least one member of the audience expressed horror at the 
> very idea that goods previously regarded as free (freeways) might 
> become costly.
I think we can see from the above two paragraphs of yours, the posts of 
others on this list, efforts like Plan B 3.0, and the latest efforts 
like the G8 summit (see link below), that there is a gigantic hodgepodge 
of solution suggestions underway. Their complexity and differences in 
detail boggles the mind. Their tendency to usually not work, or not work 
well, or to work so awkwardly as to beg the question of how can we do 
better, is ubiquitous. If one pushes back from the table and takes a 
bird's eye look at all these efforts, they look like the work of 10,000 
sailors lost at sea. So, how can we help all those sailors find their 
way? How can we slice through this logjam?

(See http://www.mofa.go.jp/policy/environment/warm/cop/policy0803.pdf as 
a solution example. This was prepared by the ministry of foreign 
affairs, Japan, for the summit. Thanks to Bill Rathborne for this link.)

This is where the System Improvement Process (SIP) comes in handy. While 
it has a long way to go to maturity, it's already refined enough to help 
us quite a bit. We simply execute the process on the proper coupling 
subproblem. The steps involved are:

In the System Understanding / Analysis step:
A. Find the feedback loops that are currently dominant.
B. Find the root cause of why they are dominant.
C. Find the low leverage points (LLPs) and symptomatic solutions.
D. Find the feedback loops that should be dominant.
E. Find the high leverage points (HLPs) to make them go dominant.

And then in the Solution Convergence step:
- Use the above knowledge and experimentation to converge on a way to 
push on the HLPs.

 From the perspective of SIP, efforts like Plan B, Al Gore's Global 
Marshall Plan, Japan's document on what the G8 should consider doing, 
what you are seeing California attempting to do, etc, are all focused on 
the Solution Convergence step. Because they have mostly skipped the 
analysis step, they are forced to converge on solutions by the most 
unreliable method known: intuition. The result is the mess you described 
in your two paragraphs above.

The exact cause of this mess (or muddle, as those across the pond would 
say) is the problem solvers involved don't know what the root causes of 
improper coupling are, what the LLPs are, and most importantly, what the 
HLPs are. This is an important claim. Pause to consider its validity. 
It's what my entire case rests upon.

First consider that if a doctor hasn't correctly diagnosed a patient's 
problem, how can she determine the right treatment?

Then consider that if one doesn't know the root cause of a problem, how 
can you resolve it? If the LLPs are unknown, how can you persuade people 
to stop pushing on them? If the HLPs are unknown, then how do you know 
where in the system to direct your solution intervention strategies?

This is where the world stands today. The answers to all three of the 
above questions are unknown. But yet individuals, NGOs, governments, and 
international agencies are acting as if they were known and correct! 
(Here's hoping this list can do something about that. :-)
The Dueling Loops paper is a sample analysis of the change resistance 
subproblem. I've also done a sample analysis of the proper coupling 
subproblem. Unfortunately, there's a timing problem here. I have several 
months to go on my second iteration. After that I could refer you to a 
100 page writeup and related model. So let me briefly describe the 
current analysis and solution conclusions.

Remember now, this is just an example of how to apply the process. But 
it may contain a few useful ideas that make it into a more mature 
analysis and a fully workable solution. By that I mean one that resolves 
the true root causes. (Root cause is another useful, standard term. It's 
crucial to a good analysis. That's why I introduced the term to this 
list a month ago. Further useful terms are LLPs and HLPs.) Here goes:


Step A. Find the feedback loops that are currently dominant. - These are 
the PAT growth loops identified so well in the World2 and World3 models. 
I like to think of this as one main loop: Consumption Growth. It could 
be modeled that way, but the more detailed World3 model has infinitely 
more credibility, because it is roughly calibrated. It's a very 
persuasive model.


Step B. Find the root cause of why they are dominant. - Here we take an 
analytical leap. Consider first that in the World3 model, there are 
about 20 places where policies need to be changed to create scenario 9, 
which is somewhat sustainable. The model is a simplification. Thus in 
the real world, there would be millions of places needing policy 
changes. Therefore it's logical to extend the World3-03 model, which 
I've done, to include a node we can call "sustainability preference." In 
percent, this is a measure of the world's systemic tendency to behave 
sustainably. As this percent goes from zero to 100%, the model moves 
from scenario 2 (today's unsustainable mode with optimistic assumptions) 
to scenario 9 (sustainable).

The "sustainability preference" node is a wonderfully simplifying 
abstraction, because it let's us ask a WHY question in a manner that is 
realistically answered and modeled: Why is the world's sustainability 
preference so low?

The answer is the world's Impact Reduction Capital (IRC) stocks are low. 
The World3-93 model has three of these: Land Yield Technology, Resource 
Conservation Technology, and Persistent Pollution Technology. But the 
other dozen or so policy places in the model have no such explicit IRC 
stocks. We could extend the model and add them. But there's a simpler 
way, one which is just as convincing: we add an Impact Reduction Capital 
subsystem to the model. This has several abstract stocks that represent 
all the IRC stocks. These can be modeled many ways.

A simple and educational way is two groups of stocks, for physical and 
social capital. Each group has three stocks, one each for R&D, 
Technology, and Capital itself. An influence flow occurs, starting in 
the R&D stock, whose discoveries allow accumulation of Technology, which 
when applied causes the Capital stock to build up. When the physical and 
social Capital stocks are sufficiently high, so is the world's 
sustainability preference. (BTW, it is the social IRC stocks that are 
the most important.)

This leads to the next WHY question: Why are the Impact Reduction 
Capital stocks so low?

Here's where we take the next leap. The answer is that the systemic 
tendency of social agents to accumulate the IRC stocks is too low. This 
is fairly obvious. The key agents are the dominant ones. As the Dueling 
Loops paper (the long version) explained, far and away the most dominant 
is the New Dominant Life Form. This is the modern corporation and its 
allies, notably the rich. (The Dueling Loops and New Dominant Life Form 
are additional standard terms.)

Looking at the real world, we can see that corporations have a low 
tendency to accumulate IRC stocks. So the model is on track. Now then, 
WHY is this tendency so low? That leads to the four root causes:

1. Corporations and humans have mutually exclusive goals. One wants to 
maximize the net present value of profits. The other wants to optimize 
the common good for all and their descendants.

2. Behaving sustainably is not nearly as profitable as behaving 
unsustainably.

3. The social system role definition for accumulating IRC is low in 
maturity. There is no standard way to go about doing it. This forces 
NGOs, governments, and even those corporations who do want to live 
sustainably to invent a myriad of different approaches, which is 
inefficient.

4. IRC laws are not general enough. New ones have to be passed for each 
new problem. This causes such an intolerably high transaction cost for 
new IRC laws that the result is few get passed, and those that do are 
immature to save on transaction costs.

Sorry this is so brief. Hope you get the main idea. Notice how different 
these root causes are from conventional wisdom, although there are many 
who are intuitively realizing some of them, such as I can see in your post.


Step C. Find the low leverage points (LLPs) and symptomatic solutions. - 
I'm still working on this. The chief LLP appears to be market 
instruments, like cap and trade. This is rapidly becoming the world's 
favorite. More obvious LLPs are regulations, quotas, and appeals to 
corporate responsibility. Pushing on these LLPs results in symptomatic 
solutions, which tend to have little effect, and if they do, are 
terribly complex and inefficient. For example, have you ever read how 
the EU implements its GHG targets?

At the meta level, the LLPs are all a form of command and control (C&C). 
Even cap and trade is C&C, because you have to set quotas. Experiments 
are beginning to find a way around this, with continuous auctions. But 
these are not exactly catching on. There's too much push back. Recall 
from history how well C&C economies have worked.


Step D. Find the feedback loops that should be dominant. - Step B found 
that the Super Servant Oppression loop was dominant. It increased the 
stock of Rules Favoring Super Servant to obscenely high levels. The 
super servant is, of course, the artificial life form Homo sapiens 
created: the modern corporation. It's important to think of them as 
servants, whose goal should be serving their masters, but whose actual 
goal has become something else.

In step D we find the loop that should be dominant. It's the Super 
Servant Benevolence loop. The stronger this is, the more super servants 
want to do a good job of serving their masters. Since the main threat to 
their master is currently the sustainability problem, a dominant Super 
Servant Benevolence loop will cause corporations to behave as 
sustainably as needed.


Step E. Find the high leverage points (HLPs) to make them go dominant. - 
There appear to be two HLPs. The first is quality of servant design. 
This is low. If it goes high, then the root cause of mutually exclusive 
goals is resolved.

The second HLP is property rights maturity. I see you have mentioned 
"property rights," so you and others are intuitively leaning toward 
this. The idea is profound. The Consumption Growth loop builds capital. 
This depends on the existence of private property rights. But where are 
the universal rights that would apply to environmental things, like 
pollution, depletion of non-renewable resources, and over use of 
renewable resources? They are nowhere to be found. Let's call these 
common property rights. If they existed, then they would probably be 
managed just as efficiently and effectively as private property. Because 
common property rights don't exist, the property rights maturity HLP is 
low.

You can easily guess what happens to the system if property rights 
maturity goes high. Common property rights (and others?) spring into 
existence. If they are defined correctly (along with possible changes to 
private property rights) then guess what? It's now more profitable for 
corporations to behave sustainably. This resolves the root cause of low 
sustainability profitability. The generality of common property rights 
resolves the root cause of low IRC law generality. And as the stock of 
these rights grows and becomes complete and mature, the root cause of a 
low maturity of super servant role definition for accumulating IRC is 
resolved.

Sorry I'm doing this without a model to show you. But actually, Tom, you 
have a sneak preview in that Vensim model I sent you and Bob, related to 
something else. :-) But wait. My mental model says if I leave it at 
this, I'll get 20 emails in my inbox tomorrow asking for a sneak peak. 
So see this link:
http://www.thwink.org/sustain/work/reference/index.htm

Notice how simple all this is. One reason is it's just a concept model. 
The other reason is that in all reality, the root causes really are this 
simple, once you find them. These are probably not yet the correct root 
causes, which will take much further analysis and experimentation. But 
they may be close, and they probably are this simple.


The Solution Convergence Step: Use the above knowledge and 
experimentation to converge on a way to push on the HLPs. - The above 
root causes, LLPs, and HLPs are a completely different analysis of the 
proper coupling problem. So it should be no surprise the solutions they 
suggest are just as different. Remember now, we would not attempt to 
implement this solution until change resistance was overcome and the 
Race to the Top Among Politicians was dominant. Then, as radical as this 
solution may appear to be, it would be welcomed, if it was supported by 
a high quality analysis and experimental proof.

The sample solution consists of two packages, one for each HLP. The 
Comprehensive Property Rights package pushes on the property rights 
maturity HLP. The Corporation 2.0 package pushes on the quality of 
servant design HLP.

The Comprehensive Property Rights package contains four solution 
elements: Common Property Rights, Environmental Property Rights, 
Reflective Pricing, and Worldism. You can see the first iteration of 
this in:
http://www.thwink.org/sustain/manuscript2/ProperCouplingPackage.htm

The Corporation 2.0 package starts with the idea that corporation 1.0 is 
what we have now. This life form was never engineered. It just happened. 
Worse yet, this life form became so powerful and clever that  it caused 
new laws to be passed that gave it even more power, and even less 
incentive to serve its master well.

The Corporation 2.0 package changes all that, with these solution 
elements: Compatible Goal, Limited Lifespan, Unlimited Liability, No 
Competitive Secrets, No Corporate Slaves, and Self-Improving. These are 
somewhat self-explanatory. 2.0 corporations can still be profit driven, 
but now that profit is highly dependent on how well they serve their 
masters.

Here's how the two packages are related: Only 2.0 corporations would be 
eligible to file claims for zillions of environmental properties, which 
they could then manage for a profit. Examples of such properties are the 
concentration of CO2 in the air, the pollution in a river, and overuse 
of a type of fish off the coast of a certain country. This is similar to 
the way utilities manage common properties today.

Also, once change resistance is overcome, it will not be hard for 
virtuous politicians to persuade the public to only buy from 2.0 
corporations, because that is in the best interest of  the common good.

This is a deep solution, one that strikes at the structural roots of the 
problem.

Will anything not this deep really work?

A huge benefit of this approach is high efficiency. The system is now 
socially tuned to run as efficiently as possible, in the direction of 
the right goals. If we cannot achieve extremely high proper coupling 
efficiency rapidly, the world will remain unable to avoid major 
collapse. How close this may be is becoming frighteningly obvious lately.

(Sorry about the length of this. Hope you are able to follow.)

Now let's address what your next paragraph mentions, using the above 
sample analysis and solution:

>
> It seems to me that two paradigm shifts are needed: one recognizing 
> limits and driving forces, and one that changes the approach to 
> solving the problem. Most current policies strike me as instances of 
> proportional control: identify a problem, and push locally against it 
> to the extent that it annoys victims. The problem may subside, but 
> never disappears (for then the will to control also vanishes). For 
> problems involving long-term stocks (e.g. atmospheric carbon, oil 
> depletion), integral control is needed: ratchet up the stringency of 
> policy until a goal is achieved. Californians might find such a 
> control paradigm shift rather liberating: if, for example, the tax 
> structure were shifted to adaptive control of bads (not goods), they 
> wouldn't need so many micromanaging rules telling industries and 
> individuals what to do (and inadvertently inhibiting them from change 
> in general).

I really like the way you are going at this. Both paradigm shifts are 
needed. One that "changes the approach to solving the problem" is vital. 
There are two part of such an approach: the process used to solve the 
problem, and the results. Your "proportion control", "adaptive control" 
and "micromanaging rules" are various results.

"Most current policies strike me as instances of proportional control: 
identify a problem, and push locally against it to the extent that it 
annoys victims." - This is an example of pushing on a variety of  LLPs.

"they wouldn't need so many micromanaging rules telling industries and 
individuals what to do" - This is pushing on the LLP of command and 
control.

"if, for example, the tax structure were shifted to adaptive control of 
bads" - Nice. Adaptive policies are the way to go. They are mode 
oriented rather than exact prediction or behavior oriented. One of the 
great lessons of SD is that system modes are much easier to analyze, 
predict and manage than exact numerical values, which remains impossible 
except for the very short term. The ultimate mode, the one the above 
analysis led us to, is one that could be called the 2.0 Super Servant 
and Common Property Rights Driven Sustainable Mode. I hope that is just 
as self-explanatory as the mode the system is in now: the 1.0 version of 
that mode.

"The problem may subside, but never disappears (for then the will to 
control also vanishes)." - Loss of "will to control" is handled by 
solving the Model Drift subproblem. (Model Drift is another key standard 
term.)


>
> Evolutionary pressure may work against such a paradigm shift. Any 
> region, firm, or individual with a sensible resource management 
> policy, for example, appears less profitable or successful in the 
> short run. Any time there are difficulties attributing long run 
> success to such policies (i.e., all the time), selection pressure will 
> disfavor the sensible behavior. This is, perhaps, the fundamental 
> reason that #2 is the hardest part of achieving sustainability.
Exactly. And see how our analysis considered this automatically? Perhaps 
this had something to do with the process that was driving the analysis.


Well, my apologies about this overly long message. But I didn't see any 
other way to fully respond in a high quality manner.

I hope this explanation allows you and others to see there is a way to 
"figure out how to act."

In fact, there is even a first iteration, and soon to be a second. If 
organizations out there rigorously applied the System Improvement 
Process, or one like it, they could perform their own iterations. Soon 
these would be well past what has been sketched here.

Thanks for such a penetrating post, Tom. It really got me going. :-)

Jack
Posted by  Jack Harich <jack at thwink.org>
posting date  Fri, 09 May 2008 22:53:39 -0400