futurehead

House of mirrors: Complexity and the presidential campaign

Clearly, our presidential campaign system is broken – dysfunctional, busted, out of order.  The fact that we might have 32 straight years of rule by Bushes and Clintons is just one symptom.  The ascendance of Jimmy Carter, Bill Clinton, and George Bush are also symptoms – whatever their qualities or lack thereof, these are not people who built reputations within the political system or nation, that threw them into the ranks of world-class leaders, the way Roosevelt, Eisenhower, Johnson, Nixon, and probably even Reagan did.  (Kennedy was famously a lightweight one-term Senator, but arguably won through a gradual & linear accumulation of popularity due to his extremely unusual popular appeal.  Truman was somewhat random – the handpicked successor of a uniquely entrenched leader.)

I believe that several aspects of the problem can be captured through the lens of complexity theory.

One of the fundamental findings of complexity theory is that you get chaotic complexity and nonlinearity when thing are – how to say it? – circular, reflexive, turned in upon themselves, self-referential. This is true of the formulas behind the Mandelbrot Set and many other complex mathematical objects. 

To take another example, one interesting area where complexity manifests itself is in so-called "networked industries," where positive feedback loops or "accelerating returns" kick in.  In a linear market, people choose a product – cereal, say – based on what they like best.  End of story.  If more people like your cereal, you will sell more of it.  It's a linear relationship.  But in a networked market where your product connects with others, you can't just go by what you like – you must also ask yourself, "what does everyone else like?" 

The most famous example of a networked market is the market for VCRs when Sony Betamax and VHS were battling for dominance, but there are many more, including computer operating systems, instant messaging programs, even languages (English is winning).  Networked economies are highly unstable in the early stages, and then they often snowball and settle on a winner that enjoys "lock-in" and can be impossible to dislodge.  Importantly, the identity of that winner is highly unpredictable at the beginning, and not dependent on any kind of rational factors – the usual kinds of things such as better service, superior products, etc. – due to its chaotic 'sensitivity to initial conditions' (aka the Butterfly Effect). VHS won even though Betamax was better.

The reason such markets are unstable and unpredictable is that they are self-referential.  Think about how circular it is: everyone wants to buy what everyone else is buying.  Tiny, unpredictable factors early in a "standards war" can determine who the winner is.

That's just in case you thought Bill Gates was a billionaire just because he was smart or a good businessman.  He may be smart or a good businessman, but that's not why he's a billionaire.  DOS was never the best operating system on the market. 

But this is all by way of background for my main point today:  our presidential campaigns have become a self-referential cesspool of circularity and have thus become chaotically unstable in some unmeasurable but real mathematical sense.

What are the factors that give the system its circular and thus chaotic and thus seemingly random character?

First there is the money factor – that it takes so much to run.  Much of that money is driven not by conviction, but like an investment.  Companies in our corrupt system want to buy influence, and you don't get none if the guy you back doesn't win.  Even earnest individuals do not want to "waste" money on someone who doesn't have a chance of winning.  When campaigns center around money, they become governed by some of the characteristics and incentives of capitalism – in particular, as with any networked market, the money flows in the direction that things are already tipping, thus tipping things further.  Think of trying to carry a full bathtub full of water on your back.  For a while it sloshes around wildly, then it tips beyond recovery in a certain direction, and the water all flows there, making it tip further, which makes more water flow there, etc, and then it's irrecoverable.  Same in the presidential race – first you have chaotic instability, then you have lock-in. 

Second there is the media.  In particular, two things about the media make it complex in a mathematical sense:

1. Pack journalism.  Because the media all know and watch and read each other and talk at parties, they tend to synch up their views.  Plus, like any community, they collectively adopt certain conventions and parameters – that some candidates are not "real" for example.  That means that the conventional wisdom is more likely to lurch in particular directions – and diversity within the media is less likely to serve as a counterweight.
2. The horse race.  As people have been complaining about for years, the media tends to focus on the horse race side of campaigns – who's ahead and who's not.  This makes people keenly aware of who's on top and who's going down and amplifies the effect of the other factors I've discussed.
   
3. Trivia and "perceptions."  The media does not focus on broad policy positions, political context, and the constituencies at work – the things that are likely to determine how a president actually governs.  As Paul Krugman put it a few years ago:

Somewhere along the line, TV news stopped reporting on candidates' policies, and turned instead to trivia that supposedly reveal their personalities. We hear about Mr. Kerry's haircuts, not his health care proposals. We hear about George Bush's brush-cutting, not his environmental policies.

The fact is, taken as a whole, the media in its coverage of presidential campaigns is a smarmy schoolyard bully, out to tease and mercilessly denigrate every candidate it can, fixating on every weakness it can find, or finding none, on random qualities which it then uses its power to define as weaknesses.  (Funnily enough, this matches some descriptions of our president's personality I have read.  But I don't know him personally.)

The media's justification for this focus on personality and perception – and this is the key point – is that such perceptions are what in fact determines a politician's success, and are therefore relevant and newsworthy.  Hopefully the circularity of that statement leaps right out at you.  As a matter of fact, it's circular in two ways.  First, the media is saying that perceptions and personality are determinative, so they cover that, but the fact that they cover it is what makes it determinative.  Second, it creates a situation where all the voters, even earnest policy-focused wonks – are constantly trying to gauge everyone else's perceptions in order to figure out who they will support.

Third, there is the candidates themselves.  The candidates themselves – especially the Democrats – also intensify the feedback effects of the campaign dynamics.  That is true in two ways:

1. Desperation. The fact that Democratic voters want above all to win the White House – and this has especially been true in 2004 and 2008 – means that they want a candidate above all who will win.  This means that many are voting based not on who they themselves like – as with breakfast cereal – but who they think everyone else will like, as with a networked product.  I bet a decisive number of Iowa primary voters in 2004 didn't like John Kerry for who he was, but because they thought the rest of the nation would like him best, and he would therefore be best to beat Bush.
2. Shaping.  The fact that the media focuses so relentlessly on perception, combined with its schoolyard bully unpredictability, creates a field of candidates who are themselves focused intensely – even for politicians – on the perceptions of themselves held by others.  They become highly, extremely self-conscious – not only personally, but institutionally, as they build entire staffs of imaging experts who are constantly peering into the house of mirrors for guidance on how the candidate should behave.  That is inimical to candidates who are calm, steady, and consistent, and instead breeds candidates who are constantly monitoring and creating and recreating themselves, adding further to the instability of the whole situation. 

The result of all these factors is a presidential selection process made up of highly unstable, self-referential dynamics that result in a process that swings wildly and nonlinearly between chaos and lock-in.  Symptoms of this include the sudden, personality-based destruction of Howard Dean's campaign in 2004; the sudden emergence of John Kerry the same year, the making or breaking of the fortunes of some presidential candidates in 2007, long, long before any voters have cast any ballots for the 2008 race; the erratic 1992 race in which Bill Clinton teetered on the edge of destruction while Paul Tsongas and Ross Perot went up, went down.  And as of this writing, it is starting to appear that Hilary Clinton is achieving lock-in already.

Mosquitoes & chemicals - update

So I had an extensive conversation the other day with the head of Arlington's Environmental Health Bureau, Richard Cole,  about the spraying of pesticides near my house in reaction to the threat of West Nile Virus.  So far in the years we've lived in Arlington I've had a very favorable opinion of all the interactions I've had with Arlington county government, and Mr. Cole was no exception.  He was open and professional, and responded patiently if not always with crystal clarity to the pointed questions I posed to him on this matter.

That said, the conversation in the end didn't do much to dispel my doubts about the rationality of these mosquito sprayings.  Here are some of the things I learned:

• He did not dispute my characterization of the level of risk presented by West Nile (as significantly less dangerous than the flu).  He did make the point that there is a vaccine for flu, but not for West Nile.  I'll have to go back and look at the CDC numbers I used; if they are the death rates that exist today even with the vaccine then his point would not hold.
• He said the mission of his bureau was "to protect public health" overall, but specifically through a few means including restaurant regulation, pest and rodent control, and preventing the spread of disease. To me that suggests that what we have is an bureaucracy that is focused on its particular mission (stopping disease) and that there may not be a proper balancing of competing risks taking place during the decisionmaking process. 
• It also confirmed my suspicions that media hype and political reaction rather than sober calculation is driving this.  Cole told me (I took notes and am a very fast typist): "If all of a sudden the news hit that we have a number of people in Arlington that were sick, I'd feel like I wasn't doing my job right, so that kinds of drives our decision process."  There has been one human case in Arlington (he's doing fine).  It is understandable that the person held responsible for protecting the public from a threat will do everything he can to avert that threat, especially when it's a major sexy media interest, but it is up to his political superiors to keep that threat in perspective and balance it against others.  But of course they don't want to answer to the media any more than he does.  And pesticides – its risks are diffuse and hard to pin down.  For a politician, the choice is easy.
• Cole admitted that the sprayings would have to be repeated indefinitely since they would in no way eradicate the mosquito, but just kill those directly hit.  "It's a temporary fix; it's not in the long term going to prevent more mosquitoes from breeding," he told me.  For a long-term fix, you would have to have a program like helicopter spraying of the entire county, and Arlington has chosen not to do that, he said.  Seems to me even if you did do that, the mosquitoes would still flood back.  This is like trying to clear floodwater with dynamite:  wherever you set off a stick, the water will be flung back and you'll see dry ground – but it won't last long.  We are faced with a choice between indefinite repeated applications of local pesticides, and the dangers that brings (the creation of resistant mosquitoes, for one thing), and just accepting the existence and small danger of West Nile as a fact of life.
• The substance that was sprayed was something called Permethrin.  Some online fact sheets, which as far as I can tell are not from industry front groups, say that Permethrin breaks down in the environment relatively quickly and does not pose a threat to groundwater, and is metabolized efficiently by mammals and does not persist significantly in body tissues.
• On the other hand, Permethrin is also a "broad-spectrum" chemical that kills indiscriminately, being toxic not just to mosquitoes but also to bees, mice, fish and many other species.  It is a neurotoxin and is classified by the EPA as a Category C ("possible human") carcinogen – basically what I take away is that really they have no idea how harmful it really may be. 
• On why the school next to my house kept its children inside on the day of the spraying, and what that implied for my children, I did not really get a clear answer.
• He assured me that "all EPA guidelines" were followed in the use and application of this pesticide. I wish I found that comforting.
  

This may be insanity, but it's an insanity much broader than Arlington County.  Nearby counties and many others across the nation are also doing this, Cole said, and I see online that Permethrin is used very broadly in agriculture (on cotton, corn and wheat) and in household bug sprays.

But I don't like having this dumped near my house.  In fact reading this stuff tonight I realize my son recently got a mysterious rash on his legs we at first thought was poison ivy, but which his doctor could not figure out the origin of.  It persisted for almost a week.  Reading about this pesticide, which can cause "redness, swelling, and possibly blistering" of the skin, I bet he got a dose.  Wonderful. 

I know/know of a lot of children with cancer.  I don't know of anybody who's died of West Nile.

Mr. Cole and the Arlington County Board and whoever else, I hope you really know what you're doing and that you're not letting the fear of sensationalist career-damaging headlines, or simply unexamined conventional wisdom and standard practice, distort your judgment.

Stumped and stuck

I wrote the other day about progress in physics – about the assumptions we all make about it and whether it will bring useful innovations. 

There is another possibility I didn’t mention: that, after all the breakthroughs that Einstein and Niels Bohr et al brought us early in the 20th century, and all the fleshing out of that paradigm shift that has taken place since, we get stuck and see no significant additional breakthroughs for centuries no matter how hard we try.  Maybe instead of an ever-acclerating pace of innovation, we have actually punched through some barriers, and are still mopping up from that accomplishment – but are now up against new barriers and things will slow down.

What if what Thomas Kuhn called “normal science” never generates a new revolution, because we simply never are able to gather enough pieces of the puzzle to pull it off?  (That might be, for example, because each new piece requires another $20 billion supercollider.)

Because we've lived at a time of constant scientific/engineering/technical innovation and discovery – in fact, it has been accelerating since the 14th century – we assume that will always be true.  But what if it doesn't?  Is that even imaginable? 

Innovation always moves with fits and starts in any one area; check out this jetliner – you wouldn't look twice if it flew by today, but it's a de Havilland Comet, the first successful jet airliner, which entered service in 1949.  If I had a 1949 computer, there'd be no danger you'd mistake it for something current (to begin with, it would be the size of a small house).  But other technologies such as automobiles and refrigerators are still being made today with only refinements to the same basic design as 1949, and earlier.  The path to land locomotion was long and steep, but not much has happened on that front since the internal combustion engine came along.

In fact it’s not even clear that technological change is accelerating, as measured by its impact on our lives. Jeff Madrick has written critically of the assumption that the computer revolution is bringing greater changes than ever seen before:

What about the steam engine, which reduced transoceanic and transcontinental travel from a few weeks to a week or less in the nineteenth century? Or the telegraph, to cite just one other of many such inventions? In the 1840s, it took ten days to send a one-page message from New York to Chicago. In the 1850s, after the telegraph was invented, it took a few minutes.

Think too about the effect things like electricity and the car had on people’s lives.

On the other hand, I do think the social effects of the information revolution will be long and deep, at least as long as Moore’s Law holds. 

But when it comes to the most basic science, like physics, it could be that we just get stumped and stuck, and that eventually all the engineering and technology implications and permutations of the basic science engine are worked out.  It could be that basic science is the sun that powers everything else and its death leads to a long cold death for the other areas too.

Okay, I don’t really believe any of this.  But maybe I’m just a man of my time who can't shake his deep-seated assumptions.  What if?

More on running injuries, with and without shoes

I have been shocked to learn that some people have read my explanation for why I do not think my stress fracture was caused by my barefoot running, and yet are not convinced!  Apparently I'm not as persuasive as I thought I was.  Also, I don't think most readers actually took a look at any of the studies that I linked to, which back up the barefooting approach pretty solidly.

In any case, time will tell.  And miles.

Meanwhile, here's an excellent piece written last year by an orthopedic surgeon summing up the subject.  See especially his part 2. 

Diminishing Returns on Mastery of the Universe?

I've always enjoyed the history of science, and I'm fascinated in particular with the dilemma of quantum mechanics, which is that the only theories that seem to actually work in explaining the behavior of particles at the subatomic level are also bafflingly messy, counterintuitive, ad hoc-feeling and, as many physicists feel, "not beautiful."  Einstein for his part famously rejected them (and spent most of his life as a bit of a scientific outcast as a result).

Many people feel that the situation we are now in is akin to Ptolemaic astronomy, in which the motion of the planets, which were understood to revolve in perfect circles around the earth, were explained through the construction of an ever-more complicated, ad hoc system of cycles and cycles within cycles (epicycles).  Beauty and simplicity were restored to this messy system through Kepler's laws of planetary motion and Newtonian physics.

Some physicists, feeling that there must be some similar underlying unity behind the ad hoc theories of quantum physics, are searching for a new Kuhnian revolution, a Grand Unified Theory (that will for example encompass gravity, which has kind of been left out of quantum physics). The whole thing is really fascinating and dramatic.  Will some startling new insight emerge that once again restores some order to physics and makes it accessible to human intuition?  Will string theory, upon which many are pinning their hopes, do the trick?  Or will physics just continue to get weirder?

I've been thinking lately that it's funny how there's an assumption that if a new breakthrough is reached, it will have huge consequences.  The quantum revolution, after all, produced nuclear energy and weapons, and the transistor, which revolutionized modern life and still is doing so. 

But it could be that, as we get better and better at manipulating our universe (however much social-historical arbitrariness there is in the direction of scientific exploration, that's basically what appears to be happening), successive revolutions have less and less impact.  So the next revolution will refine our ability to manipulate the world.  And the next one refine it even further, and so on, in ways that are progressively less directly related to our everyday life.  So, the social utility of sucessive revolutions in physics declines as the subject of study becomes smaller and smaller and actually begins to lose applicability to practical life. Even as it gets more and more expensive to figure out the next smallest level of reality, with multi-billion-dollar supercolliders and other 'big science,' it becomes less important for reducing suffering and increasing the comfort of mankind.

Of course, a new explanatory theme for physics such as string theory could also burst open entirely new doors like interstellar space travel, or bring things as unimaginable to us as the transistor was to Ulysses S. Grant. This suggestion of mine could easily appear laughable in a century in light of events to come, like the famous (but apocryphal) nineteenth century patent office guy who supposedly resigned because he thought that everything that could be invented already had been.  But the possibility that new scientific discoveries could actually get less and less important, is an idea that shocks the contemporary sensibility.  So I like it.

It's time to stop mincing words: our president is a torturer

In honor of today's Supreme Court decision (refusing to review case of CIA "rendition") as well as this past weekend's fascinating New York Times revelation of further Bush Administration perfidy on the torture issue, I finally was inspired to make this bumper sticker I've had in my head for a number of years now:

Our president is a torturer: but hey, at least he's not a serial killer!

Who lost the arctic?

Yesterday I wrote about the possibility of climate changes that are even more sudden and catastrophic than anyone is expecting.  You might think that this would be a national security issue of the first order.  You might think that conservatives and Republicans, who have made national security their top issue, who have yelled and screamed about various threats posed to our nation, who have insisted that absurd sums be spent on the national (military) defense, and who have attacked Democrats without restraint – often by calling them traitors implicitly or explicitly – you might think they would be pushing for immediate action. 

Yeah, right.

The same conservatives who have been cowering in fright at every potential shoe bomber, pushing for expanded defense budgets since 1947 just in case we have to fight two and a half wars simultaneously (which – at least before Iraq – has been our defense preparedness doctrine) and just in case the Soviets took the extremely unlikely step of rolling their tanks across Europe toward Paris – want nothing but more studies when it comes to this particular threat.

Now why would that be?  The obvious conclusion is that we shouldn’t take conservatives seriously when they prattle on about the security of the nation, stability, the need to avoid radical change, etc.; that as I discussed previously they’re just not serious and not sincere.

To flesh this out a little more, I think there are several factors at work:

1. Politically, conservatives are not honest when they pump up the threats to our nation.  They are doing it for partisan purposes, and to distract people from the fact that while they wave the flag with one hand, they are stealing money from the poor to give to the rich with the other.
2. Financially, certain sectors of the wealthy, whom conservatives and Republicans ultimately represent, have been benefiting enormously from the military-industrial complex for many years, and the inflation of perceived military threats helps them, while taking action against global warming would hurt business by leading to regulation.
3. Ideologically, conservative fair-weather anti-statists are permitted to accept an enormous defense and security establishment (even though nothing exemplifies the raw power of the state than these).  But regulations of the kind needed to deal with global warming do not jibe with their ideology.  Why would their ideology be shaped in this particular way?  See #2.
4. Psychologically, conservatives need an enemy to hate, so they can easily get themselves worked up over Reds (1848-1991), the Japanese (early 1990s), Chinese (mid-1990s-2001), “Islamo-Fascists” (2001-?), the French (2003-2004), or other human antagonists – but global warming just doesn’t push their buttons; the enemy is us.
5. Existentially, conservatives are frightened of the possibility that we, human beings, might be stuck here on a planet in space, in charge of our own destiny, with no adult supervision, and no God or other absolute around which to orient ourselves.  They prefer to rest upon absolute foundations, can’t deal with the possibility that the future could be so open-ended, and are frightened of confronting the responsibility implied by the lack of foundation and fixity of our situation.

Apocalypse: eco-collapse

Even with all the attention finally being paid to global warming, I believe most people are seriously underestimating the possibility of sudden, drastic, truly frightening changes in our climate.

I've done a lot of reading about the subject that was once popularly known as "chaos theory" and more recently and broadly has come to be known loosely as complexity theory. Among the subjects this encompasses are nonlinearity, self-organizing systems, emergent behavior, fractals, reflexivity, feedback effects, cellular automata, and evolutionary algorithms. (Here is a good introduction, but there are many others books on the subject.  Malcolm Gladwell’s book The Tipping Point seems to be a businessy popularization of some of these ideas focussed on social examples, though I haven’t read it. )

One of the things that you develop when you focus on this subject (I also spent a lot of time back when my life was simpler playing around with fractals) is what has been called a "nonlinear intuition."

To oversimplify greatly, a linear relationship is one that follows a predictable path. If your temperature rises 5 degrees, you get 3 more inches of rain; if it rises 10 degrees, you get 6, etc. A nonlinear relationship enters the picture when, say, there's a mountain range, and if the temperature is above 60, you always get 3 inches of rain no matter what, but if it falls below 60, the clouds can no longer make it over the mountain range and you get zero. So, a 1-degree change at a certain point has drastic effects. That's nonlinear.

I didn't choose that example at random, of course, I read somewhere a few years ago that a relationship like that actually holds true in the Himalayas or somewhere.

And the larger point of course is that weather systems are like that – they are very nonlinear.  Pumping x more tons of CO2 into the atmosphere will not necessarily have smooth, predictable effects.  Someone once put it like this:  if you take one rivet out of an airliner, nothing is going to happen. But if you keep removing them, at some point you will have a sudden catastrophic failure – the airplane will crash.  If you pump a little carbon dioxide into the air, nothing happens, but at a certain point, the ice caps tumble into the sea, or the jet stream flips direction, or – who knows?  Something crashes.

The result is that we're in a much, much more precarious situation than most people, with their linear intuitions, realize. Spontaneous global eco-collapse is a distinct, though probably remote, scientific possibility. Spontaneous much-bigger-than-anyone-expected eco-disruption is not only a possibility but perhaps frighteningly likely.  Global eco-collapse is one of my (slightly tongue-in-cheek) "Seven Paths to Apocalypse" – but this is no joke.

I was hoping Gore would try to capture some of this in his movie, and he did allude to it a little if I recall, but even among that slice of the population that watched it, most, lacking a quirky interest in complexity theory, are not likely to catch the full flavor of the danger we may be facing.

News reports have been suggesting for the past several months that in the arctic we might be seeing just the kind of feedback effects that can lead to catastrophic changes.  The melting of arctic ice is

a)    reducing the amount of solar heat that is reflected back into space
b)    accelerating the release of C02 that had long been trapped in the soil beneath the ice

Both of these effects of melting have the effect of increasing the pace of melting. That is just the kind of self-referential loop that leads to unstable nonlinearity and sudden discontinuities – aka big trouble.  Ironically, what is often called a "snowball effect."  I bet we're seeing it in the arctic because that's where we're looking most intensely – but it could happen in all kinds of ways in all kinds of places.

Just to get away from the ice caps, for example, did you know that the world’s coral reefs are crucial regulators of oceanic carbon dioxide and calcium, and are home to a quarter of all marine life?  Did you know that a quarter of the world’s reefs have been destroyed, for example in the 1997-98 El Nino event, which killed 70% of all corals in the Indian Ocean from Africa to India?

Most of the media attention paid to global warming has been focused on a couple effects of the warming: the rise in sea levels, and the effects on species, such as polar bears.  There has also been some discussion of an increase in hurricanes (since hurricanes are apparently nature’s way of transferring heat from the tropics toward the poles).

These are all essentially linear-minded (if totally valid) concerns.  Warming goes up, sea levels go up, polar bears die, hurricanes increase.  Warming goes up more, sea levels go higher, more polar bears die, more hurricanes hit.

What we should be most focused upon is the very real danger of sudden, unpredictable, radical changes in climactic conditions in some or many parts of the world.