Excessive Complexity = FAIL
Remember this ?
Now try this:
Simple, focused and profound trumps ridiculously complex systems designed from technocratic hubris. Even if everything here worked according to plan in these charts, the intrinsic “friction” is a colossal waste of resources.
ADDENDUM:
In the interest of evenhandedness – and because it doesn’t matter in terms of the point of my post – here’s a Democratic take on the House Democratic plan for health care provided by Curtis Gale Weeks in the comments section. While this was not the same as the final Obamacare bill, there’s some congruence. If anyone has a link to a Democratic-produced chart of the final bill, I’ll post that:
August 3rd, 2010 at 2:49 pm
This diagram is not useful.
Imagine what a diagram could look like that showed the complexity of the actions of me reading this post on my computer at home and typing a response. If someone was motivated it could include a diagram of the Intel chip etc. I type these words here at my kitchen table and, in spite of the complexity, they appear.
August 3rd, 2010 at 3:03 pm
hi MM,
.
I think there is still a very large difference in the efficiencies of the charts above and the hypothetical chart you describe.
.
Aside from the greater political character above, unlike those diagrams you did not begin by attempting to design a system to read and type on a computer screen that had yet to exist but would be working backward in a linear fashion to describe the your actions as a "system". That’s not what happened with either COIN strategy or Obamacare. Not descriptive but prescriptive.
August 3rd, 2010 at 3:04 pm
This is a political graphic–for that I am sure it will be quite useful for Republicans this fall. Other than that it isn’t an accurate portrayal of the new system. That’s not to say it isn’t complex, but the visualization is purposefully misleading.
August 3rd, 2010 at 3:22 pm
Hi Bill,
.
For my purposes, magnitude of complexity, the the 2nd chart doesn’t have to be perfectly accurate. Or Republican produced. I’ll happily post a Democratic version underneath for a comparison, then partisans can argue about the legal authorities of office X under subset B until the cows come home (The first chart does not correspond to reality either, it models "intent", not execution).
.
At the end of the day, still excessively complex as a system.
August 3rd, 2010 at 3:25 pm
Right, it’s purposely misleading – data used to confuse rather than clarify. I’m sure you could make a top-down branching chart that shows everything accurately and in an organized fashion, but who’s to say that it would be closer to reality?
As Charles Blow says, "they’re using this chart like a drunken man uses lampposts – for support rather than for illumination."
August 3rd, 2010 at 3:38 pm
http://www.flickr.com/photos/robertpalmer/3743826461/sizes/l/ — another take, made in response to an earlier GOP chart.
August 3rd, 2010 at 3:38 pm
Sorry, bad cop no doughnut. This is Republican talking points with a made-up graphic for some Texan to stir up his base. Equally matched by Democratic talking points on the other side. This is Nero playing while Rome burns. China and India are turning out 7 engineers to every one the US does. China has 70% of the wind turbine market. China is the worlds biggest car market. We are fooling with things like health care and have no energy policy and no industrial policy at all to pay for it. Schools here turn out lawyers and web developers and "communications majors." And the band plays accompaniment to the emperors.
August 3rd, 2010 at 4:45 pm
On the narrow issue that MMaineiac raises, that’s not a good comparison. The diagram explaining how your comment is posted would be a series of mechanical/electrical/algorithmic/etc processes. The diagram for the health care system is a series of human interactions. Some of those humans will be motivated and competent. Others will be lazy, uninterested, forgetful, etc. Those systems will continually change according to considerations less related to efficiency than to politics and union rules.
August 3rd, 2010 at 4:47 pm
Zen – You are right, what matters is the nature of the system, this diagram offers no clues in that regard. It is just propaganda.
Thanks for the reply, it would have been funny if you had just deleted it.
August 3rd, 2010 at 6:04 pm
Simple, focused and profound trumps ridiculously complex systems designed from technocratic hubris.
Agreed.
Now may we please see the wiring diagram of the Pentagon’s latest smart bomb?
Or a graphic design for how the latest Wall St. derivative is supposed to work?
August 3rd, 2010 at 6:07 pm
Schmedlap,
I agree, a real world example is the utility company the provides my home with electrical power and the one that provides internet service. A diagram of each utility might look similar but when my power goes off I don’t even bother calling because I know from experience that my power will be restored in a mater of hours (usually). On the other hand if I lose internet service, the service will stay off until I have to called my provider, negotiated the phone menu, reached an operator, who then explains to me that the problem is on my end etc etc . Typically I can get service restored after two or three calls.
In each case both are each regulated utilities but because of various factors (ie the meter) the default for the power company is power on and the default for the internet service provider is (once it goes off) is off.
It is the nature of the system, including the underlying dynamics that matters, not how confusing a chart someone can produce.
August 3rd, 2010 at 6:19 pm
Schmedlap, Your observation is correct; the unmanageable complexity of the second is compounded by the fact there are people in the nodes—and misses the point that the who scheme will not stand-up to constitutional scrutiny. Our Commonwealth Attorney General is proceeding with legal action, and not a moment too soon.
August 3rd, 2010 at 7:03 pm
J. Scott, the problem with the Commonwealth Attorney General’s position is that a diagram for the status quo healthcare system would be at least as complex as that which we now are discussing.
So it’s future prospects are no more viable than those which he would decry.
As for the Constitution, it is a vague document into which people routinely project their various agendas.
August 3rd, 2010 at 7:21 pm
Almost any large system is going to be complex. Provide a chart for the Defense Department, the State Department, or better yet, a chart describing how foreign policy is developed and implemented (which, of course, would include the legislative and regulatory processes). Then, consider that the health care system in the US consumes, what? One seventh of the GDP? And consider that it has grown on an ad hoc basis. I’m surprised that it is this simple.
As for propaganda, here’s a link to the <Republican description of the Democrats’ health care plan and here’s a link to the <Democratic description of the Republicans’ health care plan. I’ll leave it to you to decide which is more accurate. 🙂
August 3rd, 2010 at 9:38 pm
Hi Duncan and P. Graf,
.
"Now may we please see the wiring diagram of the Pentagon’s latest smart bomb? Or a graphic design for how the latest Wall St. derivative is supposed to work?"
.
"Almost any large system is going to be complex"
.
Gents, you are too smart for this.
.
Complexity per se is not evil or bad. This is a discussion of relative efficiency in terms of performance and costs. Preferring complex solutions or strategies when simpler ones suffice is seldom the better choice. When a task cannot be acheived without a higher order of complexity, then that level of complexity is efficient ( oportunity costs would be a normative question).
.
By contrast, "excessive " complexity is that which consumes resources and creates spillover costs without adding functionality, or does not add functionality in proportion to the costs of adding it. Some smart bomb are just complex enough but defense contractors also offer up monstrosities, like the Future Combat System, that are excessively complex, too much so to even function reliably at an affordable cost ( yes, the DoD likes excessive complexity and that is wasteful and inefficient).
August 3rd, 2010 at 10:49 pm
Zen, if simplicity were the objective of the healthcare system, then single-payer would have been implemented decades ago.
As for smart bombs being optimally engineered, it is my general impression that they do tend to go off when you pull the trigger. That there are few duds. But what all their sound and fury may signify perhaps even William Shakespeare could describe.
August 4th, 2010 at 2:40 am
Of course it’s complex. It has to be, it’s goal is to ensure that Americans have access to affordable health care. What would a simpler system look like? How does one set about simplifying the existing system? Can we really determine excessive friction from the Republican chart?This reminds me of the teabaggers insistence on a tax code no longer than the U.S. Constitution. It’s silly to assume that the tax code can be simplified in such a way. The economy is far to complex to be covered by a tax code cut down to ~ 4,000 words. Like Zen said in the recent comment: It requires a higher level of complexity to be efficient. I think healthcare and COIN operations both require a higher level of complexity.
August 4th, 2010 at 5:15 pm
Perhaps we’re asking the wrong questions. Rather than how to simplify a national system, perhaps we should be asking why we need a national system.
August 4th, 2010 at 6:50 pm
Have you looked at the "Art of Design" Text at Army School of Advanced Military Studies? Adam Elkus refered to "design" in this context in his latest with John Sullivan on "Red teaming ideas.
August 4th, 2010 at 10:16 pm
Nassim Nicholas Taleb has a good article up in Huffington Post "The Regulator Franchise" which I think explains a problem with complex regulations.
August 4th, 2010 at 10:16 pm
The hard sciences are not the social sciences, so some of the comparisons made above are not valid. How a computer works is not the same as how a human works. The health care diagram refers to a human system which is meant to "govern" a biological science – medicine. The "unknown unknowns" are far greater.
.
The ACA empowers a variety of regulatory agencies that have not yet been fully staffed and whose staff have not yet written the rules by which physicians, nurses, and others will have to comply. An empirically minded person might ask, "how am I to judge this system which is not yet fully written?"
.
– Madhu
.
(Graham J. – were you the commenter in the following thread at Abu M where I lost it and sock-puppeted myself as "Anon is all you deserve?" Very bad form. Especially since everyone figured it out immediately and someone emailed me about it. This is so typical of my off-line life, it’s funny where it’s not sad. I have gotten into a lot of trouble by saying things in meetings when I should have kept my mouth shut, but I cannot stand bureaucratic wagon-circling and hemming and hawing. It makes me see red.
.
Just thought I’d apologize here too. Honestly, though, the COIN or security blogging community is awfully uptight. Culturally different, too. I started on the Indian and Indian-American blogs and I assure you nobody cared about yelling vigorously at each other. The only thing required was to be sort of stylish during the yelling. Curious cultural difference. Still wrong, still bad form, because I should know better.
.
I do not think the public policy blogosphere is good for me. I am beginning to loathe all politics, all political people, and what I am able to see of the Beltway participants from my outside-the-loop perspective. This is horribly unfair to the participants, because not everyone is a jerk. )
.
http://cnas.org/blogs/abumuqawama/2010/07/scoop.html#comments)
.
– Madhu
August 4th, 2010 at 11:38 pm
Hi Dr. Madhu,
.
"The hard sciences are not the social sciences, so some of the comparisons made above are not valid. How a computer works is not the same as how a human works"
.
Yes. There are different kinds of complexity as well as degrees of complexity, as well as open and closed systems, with any system involving humans (be it a social system or biologically) is going to be very complex and make predicting second and third order effects beyond "rough guess" very unlikely.
.
But as I am having difficulty just communicating my basic point of "don’t make or design systems to be more complex than absolutely necessary", I think I will quit while I am not ahead 🙂
August 5th, 2010 at 12:39 am
Zen,
.
FWIW, I thought you got it across. My addition was simply to highlight that a preliminary question in determining the appropriate level of complexity would be to ask, "is this thing even needed?" When I was designing my website (a horrible example, given its poor design), I initially created XML files that contained the blog post and every comment thereafter. I made this as simple as possible. But then I later realized, that while I did create XML files and means of editing them that were as simple as possible, I had settled upon an architecture that was excessively complex. So, I broke up blog posts and comments into separate files. For reasons that can only be explained through the mysteries of programming, using several files was far less complex than one large one. My error was assuming that I needed everything in one file.
.
I think we have also skipped over that issue with healthcare. It seems we start from the assumption that a nationwide plan is necessary and then we search for a way to make an excessively complex system less cumbersome. Rather than putting everything in one program, perhaps we need a bunch of smaller programs. If only we had like 50 or so smaller political units with which to do this.
August 5th, 2010 at 1:33 am
In terms of planning, while the systems these graphics are trying to describe are complex, I believe alot of it has to do with our unfamiliarity with trying to portray it these terms (won’t go into the the side effects of ppt). A previous commentator mentioned the Constitution. Think if the drafters had felt the need to portray the government described as a graphic. While it would appear to be simple (three branches, checks & balances), it could get very confused to portray all of the back and forth relationships between the branches, the states, the populace, other nations, etc.
As a point of departure, a standard military topographic map with a conventional tactical plan graphically portrayed is a very complex portrayal, except to the practitioner who readily understands all the symbols (most of, those on the map itself portarying terrain) and don’t have to put much conscious thought into the interpretation of the relationships and implications/possibilities.
I think this rage for graphic portrayal has come out of Systemic Operational Design (now transformed in simply Design), and the military prediliction to want to see a picture, rather than having to wade through narrative. We somehow think if we can draw it, we can understand it–and we love solving problems via just the right collection of arrows towards the correct objective (whether in the physical or "cognitive" domains).
In my opinion, there are probably discrete processes that can be successfully visually portrayed in healthcare. However, for the most part, its something that needs to be explained via narrative.
August 5th, 2010 at 2:08 am
"the military prediliction to want to see a picture, rather than having to wade through narrative"
—
Phil, that’s one of the sharpest observations in this thread — even without needing to use the military for an example, except insofar as factional politicos and non-mil types from most walks of life are "militaristic" after a fashion. (I mean, cognitively, since intellectually the love for categorization and dialectics is a love for conflict, in which some ideas and understandings are weeded out while others gain ascendancy.)
—
The trouble is, the real trouble is, that besides this natural aversion to "wading through narrative," the narrative in this case is still being written. So that makes the process even more difficult.
August 5th, 2010 at 4:30 am
I just had a vision of the founders briefing the Constitution with PowerPoint. This nation never would have survived.
August 5th, 2010 at 4:31 am
[…] other knaves to form rent-seeking groups. They can dupe saints into supporting efforts to create institutional complexity to replace simpler tribalism. Saints think they can use such complexity to raise everyone to a […]
August 5th, 2010 at 11:30 am
[…] other knaves to form rent-seeking groups. They can dupe saints into supporting efforts to create institutional complexity to replace simpler tribalism. Saints think they can use such complexity to raise everyone to a […]
August 5th, 2010 at 1:34 pm
– This is perhaps somewhat beside the point but…..I have a passing familiarity with GIS (Geographical Information Systems), when I read Phil Ridderhof ‘s point about narrative vs graphic it struck me that a similar tool would be useful for understanding complex systems. If I could see something like the original graphic on my screen with the ability to take away "layers" and the ability to mouse over and see text combines the utility of both the graphic and the narrative (text) perhaps analysts could dig in deeper. Surely such tools are in use already?
– Post at C. Boyz well worth the read.
– Thanks again Zen for taking the time to respond to my comments.
August 9th, 2010 at 12:41 am
The Lost Arts Of War…
…an interestin post over at . . ….
August 9th, 2010 at 3:36 am
How about a chart of the *current* health system?
August 9th, 2010 at 9:16 pm
[…] other knaves to form rent-seeking groups. They can dupe saints into supporting efforts to create institutional complexity to replace simpler tribalism. Saints think they can use such complexity to raise everyone to a […]
March 24th, 2012 at 12:30 pm
military watches…
[…]zenpundit.com » Blog Archive » Excessive Complexity = FAIL[…]…