The US has organized its own demise. I can conclude that from reasons of systems dynamics and problems of controlling open, evolving, complex systems. This analysis and line of thought is important as a cross-check to the political, social science view and analyses. Those have the weakness of so many loaded words, euphemisms, circumlocutions, and rhetorical slights of hand most are difficult to assign concrete meaning.
In the systems view, the rapid, haphazard growth of our current socio-econo-political system is both a symptom of the base problem and a cause of further decline.
- Enormous systems of subsystems of subsystems is a useful mental model of the workings of our world.
- Systems engineers’ maxim that “If a system is not designed to be fail-safe, tested under all combinations of extreme conditions, and operated perfectly, it will fail” generalizes to that mental model and through that to the reality.
Add the fact that no technology exists to control open, evolving, complex systems, most important systems in our economic system are not so designed, tested or operated. No entity has the power to do that for more than computer systems or machines that can be replicated.***
In empirical support of that statement appear headlines every hour. Specifically, note that our leaders in banking failed to regulate S&Ls 25 years ago, fail at regulating banks every 3 or 4 generations, and just fixed the last bout of failure by merging banks that were judged to be ‘too big to fail’ with bigger banks. That is not engineering, it is wall-papering to keep out the tsunami.
It is not possible to manage such systems, no group has ever successfully done so for generations. Assuming control of any part of an economy or society is hubris to the power of Dunning-Kruger.
We must conclude that our systems will fail because they were not developed so as to not fail and because the tasks they are doing are not possible with current technology.
Murphy’s Law predicts the failure will do maximum damage. Taleb invests on his understanding that large sophisticated investors assume too-low risks of major losses on investments. Nassim Taleb is wealthy man because of that insight.
Some rate of failure is normal, even desirable when a result of a questing system. Failure cascading to catastrophe is the problem. What can a systems’ view tell us about the scene in the US in 2015?
We know how failures propagate through networks of links in the standard model, that is a convenient descriptive summary of data more than a model, the failure of A very often causes the failure of B by direct linkage of money or products supplied, and many authors have analyzed very many other kinds of linkages.
Do we see failures of a system, or normal failures that happen in all systems of interacting entities in evolutionary competitions?
All failures can have cascades, so that is not diagnostic. I believe the two key variables**** are the rate of the arrival of new failures vs the rate of adjusting to them and the proportion of those adjustments that cause additional, future problems. Both the recognition of new problems and the effectiveness of fixes are proportional to accuracy and sensitivity of sensors and effectors.*****
By those measures, our world socio-econo-political systems are in a cascading failure. There are many people who seem to me to be rational, knowledgeable and analytical who say, that for the elements of the world they understand, in their models of those, they do not see how to avoid catastrophe if trends continue, and they see no way to change those trends. Or worse, their proposed definitions of the problem and solutions would make things very much worse, by historical evidence. There are very few proponents of returning our economic ecosystems to natural states, they have little power.
My conclusion is that the only question is how low and how long the dip in our various measures of fundamental human values such as food, water, life, health, education.
You should be skeptical of this mental model, of course, one of “we live in enormous systems of subsystems of subsystems”. That is map not territory. Even if I got that right, experts can be pretend (I am) and make bigger mistakes as they are more confident.****** The model and argument are independent checks, a different view. Everyone thinks his view is fundamental, so I won’t claim that.
The best one can do, I think, is get the major trends right and a first-order of interaction between them. That works in ‘predictable times’, where the relationships between subsystems are only changing locally, linearly, and entities have time to adjust to take advantage of the trends or avoid the consequences.
‘Predictable’ until the next Black Swan, which ‘comes from nowhere’, by the definition of a Black Swan, and often begins with very small changes changing exponentially.
We can hand-wavingly sort-of validate the model by checking other trends that would be predicted by the trends and model. For example, if the system itself is broken :
- Candidates and office holders should increasingly appear ineffective and be ineffective.
I think the current leadership of both parties is at an all-time low. Old and ugly, a geriatric kleptocracy. But the class just before them was the worst to that point, and recursively into the past as long as I can assess it. Normal people in a broken system don’t look good, and our leader-selection system is as broken as the rest of the system.
- The average size of problems increases with time faster than other measures of growth.
One reason for this during bubbles is that cheap money and economic stress leads to consolidations. That is a symptom of failure, as the average size of the biggest 5 companies at every scale in every part of US business is a higher proportion of the total industry every year.
Empirically true, I think. The last bailout, as a proportion of GDP, was by far the largest in history. TBTF banks, US corporate and government debt has accelerated every administration. Size and cost of wars, military, pensions, medical procedures, scientific projects, etc.
- Degradation of Sensors, Effectors, Operations
OMB was assigned the problem of checking people for security clearances, a basic sensor function. Those inputs to the rest of government were quite bogus for some years. The FBI’s Crime Lab, Translation Service, … have also had problems of generating incorrect inputs to other parts of government.
Our government is OK at complete routine, cutting checks and keeping records. At huge expense per operation, the military can manage logistics and basic operations. They even evolve under a good commander, but little of that learning appears to propagate or stick.
A measure of our governments’ abilities is the incessant stream of propaganda. This is an admission of failure, they need to cover up reality. It is also a great danger in itself, as people eventually believe their own propaganda. Garbage Out, Garbage In arrangement of information flows in control systems is bad for systems.
- Increase in politics
Failing human systems increase the effort devoted to politics because the pie has grown so large that has been a good investment to do so. They redouble that effort in the contractions that follow attempting to maintain their position.
Empirically true, I believe I read this election cycle will cost $2B. I think it is doubling every presidential cycle.
Thus, the US is nearing, even past, peak power and spending. With that power and spending have come our leadership’s capacity to completely wreck the economy and society. Which they are doing because they can’t allow better concepts into their head, they would hurt too many existing interests. They are hemmed in by the complexity of the organizations, rules, regulations, shifting powers.
Even as kleptocrats, they are losing because they cannot organize themselves to make wise decisions. It cannot be in the interests of the Status Quo to cascade to failure, but the scene is politics as usual. I see failure everywhere, no political ripples.
This game is just starting and they are on the the failing side.
As I see this playing out, we 99% citizens will have been failed, will not have failed, will have have been hurt, some badly. That produces an entirely different psychology then being defeated or otherwise producing equivalent hurt, one often leading to bad politics.
Bad politics is the future I would think we need to avoid. This transition can be a very positive thing for us and the entire world, but everyone has to think hard. This is not an easy optimization, a re-division of power and wealth, one appropriate to this scale and productivity of society yet avoiding civil war. In 2016, in my judgment, the judgment that the re-division is necessary has been made by the solid middle of America. Obviously, our political establishment has not found a way to reconcile that with their dependence on campaign contributions and the fealty that buys.
Models of type ‘process+attributes <-flow-> process+attributes’ are used in all of the social sciences, games, control systems, many other areas. They are often expressed as sets of iterated simultaneous equations. Not useless, they sometimes reveal constraints on how systems could have worked.
Very sophisticated versions are being made and tested and updated continuously by the best modelers of social systems, and estimates of different kinds of problems updated weekly. That ‘big data’ analysis and synthesis is the latest version of the historian and modelers helping understand the lessons of history. There are many varieties of tools to record the elements, properties and interactions. Big teams are needed to estimate properties.
Not so clear to me that anything has improved. Big and precise it may be, but it is still correlation data, deriving cause and effect is not possible. But worse than that, your model either has nonlinear feedback, therefore has mathematical chaos and can be correct, or it doesn’t and calculates some average and you don’t know how far that will be from ‘real’.
If chaos, your computation is going to take a very long time relative to simpler models AND you will never get the same answer twice for even slightly different inputs. In fact, two sets of data differing by any one digit can produce totally different results after enough iterations. These are built into the math, and so the guesstimates of initial values and strengths of relationships defined in the equation define the result. The linear model isn’t better, the equations and constants are no more reality than the first, estimates all. Maps of maps of territory, navigating is hard. Validating the model is determining its usefulness, and that is hard to do for the real world.
Perhaps the weakness is the wetware model not the software model or data?
There is much handwaving in all that above, imprecision of words, ‘assume something that pleases you’ relationships, but it is as close to a non-political, engineering-model and -thinking argument as I can make, given my time and brainpower limitations.
Suggestions and corrections are welcome, politically-based arguments aren’t. Those are how we got here.
*Yes, that competition wasn’t just one-way, but Germany and Russia could have decided to conquer the world and then argued about whose version was better. Victor Suvorov says in his “The Chief Culprit” that both Hitler and Stalin intended war with the other, Germany attacked first. It seems like a good argument to me because Suvorov was far more inside the system and archives than the historians who critique him, and an intelligence person who understands all of the nuances of the evidence. I thought he made many telling points, had a strong case. The critiques I read didn’t, it seem to me, deal with the evidence, but that was a year ago.
**That will work, those dumb Danes will never catch on that our oligarchy is giving them mere paper, claims on future real wealth, which gives our oligarchs a big head start in the scrap with their oligarchs for any remaining value. Such a fine example of current solutions causing future problems.
***Every year shows that another company’s systems fail unexpectedly. Even the top technology companies, Google and Facebook and Amazon have failures that stop some services for significant time. Not often, but they have them.
****I had thought that ‘connected’ vs ‘random’ is important, but once past obvious money and material flows, it is hard to trace connections that matter. These ideas are not at all new with me, rather me working through the process to understand it.
*You do realize how powerful negative sells are? People are so overwhelmed by spin they appreciate truth. At least smart people do. YMMV
Also, there are more than a few academics discussing this, I have read articles, so no doubt all the above is stale thinking, but someone will tell me.
Apologies to early readers. I accidentally hit publish (wordpress uses the same blue button for ‘update’ and ‘publish’ in different screens), decided to edit in the published state, and that took many more revisions and 2 days, still isn’t quite right. So early readers have read poor and confusing versions of this.