Thursday, November 17, 2016

Peter Turchin — I Use the Science of Predicting the Rise and Fall of Societies. What I Discovered Will Alarm You.

Cliodynamics is a new “transdisciplinary discipline” that treats history as just another science. Ten years ago I started applying its tools to the society I live in: the United States. What I discovered alarmed me.
My research showed that about 40 seemingly disparate (but, according to cliodynamics, related) social indicators experienced turning points during the 1970s. Historically, such developments have served as leading indicators of political turmoil. My model indicated that social instability and political violence would peak in the 2020s (see Political Instability May be a Contributor in the Coming Decade).
The presidential election which we have experienced, unfortunately, confirms this forecast. We seem to be well on track for the 2020s instability peak. And although the election is over, the deep structural forces that brought us the current political crisis have not gone away. If anything, the negative trends seem to be accelerating….
Evonomics
I Use the Science of Predicting the Rise and Fall of Societies. What I Discovered Will Alarm You.
Peter Turchin | Professor in the Department of Ecology and Evolutionary Biology at the University of Connecticut, Research Associate in the School of Anthropology, University of Oxford, and Vice-President of the Evolution Institute

48 comments:

MRW said...

Oh jesus. His model tels him this, tells him that, and he claims his model is scientific because it contains all the tenets of social science.

So he writes stuff like this: From what I have seen so far [in his model], it seems unlikely that the Trump administration will succeed in reversing these negative trends. And some of the proposed policies will likely make them worse.

Pulls it straight out of his ass, and his model.

He continues in the same graf to say: For example, drastically reducing taxes on the wealthy Americans will hardly strengthen fiscal health of the state.

Hunh? Clueless.

I could have told Turchin seven years ago what his model is telling him now, but I got that from just listening to people. I knew nearly eight years ago when Obama refused in Jan 2009 to prosecute any bankers for the crisis, and when he dismissed what they did as unfortunate but still legal that it would lead to the social discontent Trump capitalized on. it was a fucking nobrainer.

This guy Turchin thinks you can model human behavior and that it's predictable. I'll bet you 10-to-one his model didn't predict the 2008 Financial Crisis.

(Thank god I got my degrees before 1985.)

GLH said...

I've read too many profits over the years, enough to know that none of them know what they are talking about until after the event.

MRW said...

It's clueless because (1) it's a drop in the bucket, and (2) it won't make his model's "negative trends" worse one way or the other.

Matt Franko said...

8 years of ZIRP looks this way....

Ignacio said...
This comment has been removed by the author.
Ignacio said...

Human behavior is perfectly predictable IMO. With enough data, you do it on a daily basis with a very high success rate, I'm pretty sure, a much better success rate than predicting the weather, for example. If you don't, is because you had a failed socialization. This is what you have been trained for since you were a kid, predicting others behaviour, we do this full time 16/7.

What is not predictable is aggregated human behavior, but it's not because "we are oh so unpredictable and unique!" (bullshit from radical individualism ideology propagated during the last centuries), but because lack of enough information and combinatorial explosion. Is a problem of COMPLEXITY and information intractability, is not a problem of uncertainty.

Yes, I'm fully determinist and materialist ;) Stochastic approaches are just replacements for the lack of data or computational power. All the heuristics (in form of abstract concepts) we use to explain social behavior are stochastic approaches to explaining aggregated behavior.

I think the classics understood this way better than us, as did religions created in the past (so both atheists and theists understood this much better than current generation). Spreading the "unique snowflake" feeling for marketing and ideological reasons has done much damage in this regard.

Matt Franko said...

Ignacio they often use figures of speech with nothing behind them...

This is good:

"Yes, I'm fully determinist and materialist ;) Stochastic approaches are just replacements for the lack of data or computational power."

No need to put the ' ;) ' in there for me anyway... our bodies are material the earth, etc...

At core, all of this is about 'eating and pooping' when you get down to it... and we have been under surplus conditions since the flood... we only F it up to create shortages...

Tom Hickey said...

Human behavior is perfectly predictable IMO. With enough data, you do it on a daily basis with a very high success rate, I'm pretty sure, a much better success rate than predicting the weather, for example. If you don't, is because you had a failed socialization.

Act of us arrives at a computational approximation (stochastic) based on our experience with others. Some are better than others at doing this and this reflected in outcomes based on these approximations, along with other factors.

However, this is not rigorous. There are no random controlled trials, for example, since even if it were possible the transaction cost would not make it worthwhile in most cases.

But humans are creatures of habit and societies are based on customs, traditions, and institutional arrangements. Where there is regularity there is predictability.

But even RCT's cannot be generalized beyond the population involved without making the assumption that other populations are the same. They may not be.

Similarly, we often generalize from out limited experience of others in our social environment to all humans, regardless of differences in habit patterns among different groups and different traditions, customs and institutional arrangements.

For example, most of the vulgar analysis by Americans other nationals falls victim to this type of thinking. American "experts" were convinced that humans prefer freedom so just liberate the masses in Iraq and they will spontaneously adopt a Western liberal democratic form of government and quickly learn how to operate it.

The New York Time and Washington Post tell us how Putin is thinking. Or they report that the Chinese are "inscrutable," which they are to Westerners but not to themselves with whom they interact daily.

Analysis starts with the sample as assumptions about it. This where a lot of the unpredictability (uncertainty) lies. Either the sample is not representative of the population or the population is not as generalizable as assumed.

Malmo's Ghost said...

Ugg, more pseudoscience.

The real deal is that Half the electorate didn't vote. They r presumably content with their lives and societal stabily to the point of not voting. 25% of electorate voted for Trump. Therefore apprx 75% of electorate is happy Trump is president or aren't bothered by it enough to vote. Looking pretty stable for 75% of us. So much for the 50-50 divide meme.

Not sure about the prospects of society falling being right around the corner.

Six said...

Malmo:

Your premise "They r presumably content with their lives and societal stabily to the point of not voting" is one possibility, but it is most likely false. The most worse off tend not to vote. Perhaps due to the heads I win, tales you lose nature of our elections (re: screwing the "lower" classes)? But I could be wrong.

Tom Hickey said...

People that think their vote won't count don't vote. Not voting is also a vote for none of the above. People that are alienated from the system don't vote either. Lots of reasons not to vote other than apathy.

Malmo's Ghost said...

Six, Tom,

Both of you are correct, at least in part. There are obviously a myriad of reasons why people don't vote.

I'd guess there isn't a small number who don't vote because they simply don't give a shit about politics. Tens of millions of good people do many other interesting things with their lives besides being obsessed with politics.

But still, no mater the reason, almost half of the country doesn't vote--even in an election like this where we had the starkest contrast in memory. These people are no more disgruntled than they were before the election. In fact my guess is whether they be rich, poor, young, old, male, female they are simply emotionally detached from politics, partisan or otherwise. That's a fairly stabilizing force in our so called divided polity.

Malmo's Ghost said...

...They're like the Jimmy Stewart character in the movie Shenandoah

André said...

"Is a problem of COMPLEXITY and information intractability, is not a problem of uncertainty."

That's a very interesting and philosophical subject...

I have a different view. Although I believe that more data would help a lot, somethings would he highly unpredictable even with more data.

For example, more than once I decide on who to vote in the last few weeks before election, and a lot of people do that. Long term predictions in this case are very imprecise.

And also, in last presidential elections, a popular candidate died in an airplace accident few weeks before the election. Who could predict that?

If he had survived, there were real chances of winning. Who would he choose as Finance Minister? Who would run the Central Bank? Would this person be more prone to raise or lower the interest rate? I don't know, but the intetest rate could be very different now.

How a model would get that even with more data?

MRW said...

Tom wrote this:
Act of us arrives at a computational approximation (stochastic) based on our experience with others. Some are better than others at doing this and this reflected in outcomes based on these approximations, along with other factors.

Which is true. (Intelligence, ability to listen, access to info, concern for the issue in the first place, all important vairables.)

Then he wrote this:
However, this is not rigorous. There are no random controlled trials, for example, since even if it were possible the transaction cost would not make it worthwhile in most cases.

Dead right. And the point. Not rigorous. Not reliable day after day, year after year, as a measurable event (say like the movement of Mars, or the phases of the moon). That's the whole point of why it's bullshit. That is what keeps it out of the range of science, no matter the claims made.

Because you apply one process during times of stability, when risk is measurable, but you cannot apply that process to disruptive times, when risk is not measurable and there is uncertainty. You cannot model uncertainty, only recognize it in your model. Big diff.

[Great universal examples of disruptive times: WWI, going off the gold standard domestially, Churchill's drunkenness leading to WWII and terrible world war, Russia's first satellite in space, death of Kennedy, etcetera, etcetera, etcetera.]

As Keynes wrote, assessing for risk during times of stability is like a deck of cards, variable, but measurable. It is useful to estimate your chances of something occurring. This is what Turchin is assuming is scientific. It's not.

But. As Keynes equally made apparent--and everyone ignores him about this because none of the quoters have ever read him--risk management means boo-fuck-all during times of uncertainty. You cannot estimate whether we will be at war ten years from now. You cannot accurately predict the price of oil in 10 years. You do not have the basis to know. It's incalculable. You can only do what Tom suggested in my first quote above.

Moreover. And this is the kicker. You cannot predict a disruptive time.

So all the methods of assessing the times (whether stable or disruptive) only apple after you have accurately identified the times. "Where the fuck is my tail?"

Excuse all typos.

MRW said...

[BTW, this is the problem with the climate models I discovered when I looked into it starting seven years ago. Far too byzantine to go into here. Wow. Have we been sold a load of shit. And then we're subject to same "scientific" arguments coming out of the poli-sci/social science environmental departments--where most 'climate science' courses reside once it was made a univ. discipline in 1979--as this author claims for his quasi 'end of the world' scenario, which he thinks should alarm you.]

Tom Hickey said...

Some are better than others at doing this and this reflected in outcomes based on these approximations, along with other factors.

The other factors is the kicker. Past success is no guarantee of future performance. There are plenty of people who have been considered experts owing to past success, but when conditions change that they are either unaware of or in denial about, they blow up.

The cognitive-affective bias factor is also huge subjectively in addition to the "buzzin' bloomin' confusion" of the so-called objective world to quote Wm. James.

Tom Hickey said...

This is not an argument against critical thinking but rather a caution. Take medicine. A number of RCTs on the same subject can vary pretty significantly, because of "other factors" not recognized and taken into account. It’s not possible to generalize from one study, as the newspaper headlines regularly do about the next new study. Or if they qualify it, it only at the end of the article "in fine print."

MRW said...

@Tom,

This is not an argument against critical thinking but rather a caution.

Not an argument against critical thinking? Shit, Tom, I thought what you wrote was an example of critical thinking!

I have grown really fed-up with what passes for learning in universities today, or the horseshit I read coming from our intelligentsia.

And I have degrees up the ying-yang. I'm Ivy League, and they were the top schools in the world. If I were to do it all over again today, and knowing what I know now, I wouldn't go. I'd get a good vocational training, along the lines of the perfection demanded in German trade schools (not the 9-month jobbies here), and I would educate myself until the scurvy and disgraceful scholarship in our institutions gets eliminated.

MRW said...

A spiritual guy told me years ago. First, there is the mountain. Then there is the M. O. U. N. T. A. I. N.

Then, there is the mountain. In all its simplicity and completeness.

Tom Hickey said...

The problem with rigor is several fold. Data, choice of approach, application of method, transaction cost.

All involve transaction cost, which is perhaps the biggest reason that we most often use heuristic thinking rather than rigorous. It's a tradeoff between efficiency and effectiveness.

Sometimes efficiency prevails, e.g.., when effectiveness is not all that important, and at other times effectiveness prevails to the degree that transaction cost is deemed worth paying in time, energy and resources, when not being right or only being approximately right is much more costly, like in putting a man on the moon or combat strategy and tactics where winning is everything.

Behavior is measurable on a scale of a heuristic/rigorous ratio. Getting the decision criteria right and applying them correctly is often tricky, too. In some cases where very rigorous thinking is mandated by the criteria, there is not enough time or resources. This is "budgeting."

This is a reason that life (evolution) is "experimental" and nature seems so wasteful of resources in conducting so many failed experiments. But getting things right under uncertainty often requires exploring a lot of options before promising ones emerge from feedback, or successful ones discovered.

Even in the big picture from the POV of perennial wisdom, each individual, being unique, conducts many experiments on the way to realization of wisdom. In a system based on feedback, in which freedom is a factor, failure is a prerequisite to success.
"Nothing ventured, nothing gained."

MRW said...

The problem with rigor is several fold. Data, choice of approach, application of method, transaction cost.

And the statistical methodology which is a rigorous field all unto itself. Most researchers are not trained in it properly; hence, the dreadful medical and climate papers that some journals had to recall. it's a far more serious issue than most suppose.

All involve transaction cost, which is perhaps the biggest reason that we most often use heuristic thinking rather than rigorous.

Ever watch this? Gerd Gigerenzer's talk: "Simple Heuristics That Make Us Smart."
He's the head of some part of the Max Planck Intitute, or maybe the whole thing.
New Economic Thinking's (INET) Paradigm Lost Conference in Berlin. April 12, 2012
https://www.youtube.com/watch?v=PIsNt_7sah4
(this version is from a different conference, or you can use the nakedcapitalism link:
http://www.nakedcapitalism.com/2012/04/gerd-gigerenzer-on-how-decisions-are-really-made-versus-how-economists-say-they-should-make-decisions-and-why-the-folks-in-the-real-world-often-have-it-right.html)


Tom Hickey said...

Download "Simple Heuristics That Make Us Smart" free at archive.org

https://archive.org/details/pdfy-Xi4vv7AaDeTOQeF-

MRW said...

Thanks for the permlink.

Tom Hickey said...

I just glanced at it. Looks very good. Weekend reading. :)

MRW said...

Ah, it's a book at archive.org. Thx.

MRW said...

I wish Gigerenzer would have stated that "heuristics" is an academic name for 'using your nose'.

MRW said...

Or going on your gut.

MRW said...

Be sure to watch his video. It's great. Love the opening anecdote.

MRW said...

Last night, Kanye West praised Trump for 40 minutes during his concert in San Jose. Said that if he had voted he would have voted Trump. (Guess his excuse is that he was traveling.) He ended his tirade saying that he is entitled to his opinion and state it, "Im a human being too, I have feelings, I liked the debates, Im a guy & a rapper, I'm a non political rebel, the debates were entertaining."

This was a disruptive generational event that will carry across racial divides.

There was a lot of booing and clapping. But this was a crack across the bow any way you look at it.

This was a response from the culture at a level that is significant.

http://www.independent.co.uk/arts-entertainment/music/news/kanye-west-vote-donald-trump-speech-concert-san-jose-president-2020-a7424316.html

Malmo's Ghost said...

MRW, you nailed it regarding Kanye West. He has tens of millions scratching their heads over their preconceived notions on Trump.

MRW said...

@Malmo's Ghost,

He has tens of millions scratching their heads over their preconceived notions on Trump.

And that's worldwide, Malmo. He just moved the rudder an inch. We'll see where it goes.

Ignacio said...

How a model would get that even with more data?

Right, the problem is twofold: amount of data, and computational power (in which you can include understanding such data).

If you view the universe as deterministic and a big information transfer problem the problem is not uncertainty, but complexity. Complexity 'appears' uncertain.

But there are a lot of classes of problems which are intractable because they are not computable (combinatorial explosion), even if we understood the data, which we don't.


Regarding predictions, we never can predict with absolute certainty, there are no sciences that do that, but we can predict with a high degree of success based on knowledge. This does not prevent fat tails and black swans, but those are not uncertain in a philosophical sense, but then again, a problem of data.

Not saying in practice we have to model stochastically because intractability, but we shouldn't stop trying to aproach problems deterministically, including human behavior.

Tom Hickey said...

Complexity 'appears' uncertain.

Right.

Epistemological uncertainty results from inability to gather and process data and systemic interaction.

Ontological uncertainty results from emergence.

At bottom they are the same and would be overcome in the unity of subjective and objective in the realization of the unity of subjective and objective.

MRW said...

Regarding predictions, we never can predict with absolute certainty, there are no sciences that do that

Yes there is. Astronomy, the apsidal precession for example...well, all the precessions. Wikipedia has a pretty good entry on it.
https://en.wikipedia.org/wiki/Apsidal_precession

MRW said...

OT--and I am going to mention it on this site a couple of times this week--for those of you who have the slightest interest in climate change, you need to watch Joe Bastardi's Saturday Summary for today. It will be up for a week. It's a real lesson in how weather patterns determine what happens in the Northern Hemisphere. Bastardi doesn't normally make his Saturday Summaries this long. Occasionally, he does, and they're for a reason. They are fun to watch during your first cup of coffee.

This video is what all those who shout "Global warming!" need to see before making further fools of themselves. It's also, imo, fucking fascinating...because he makes it clear. I've been watching his free vids for the past three years, and now I watch his dailies too. He has called major events correctly nine months to a year before they happened that I will hear in the news as an omigod-climate-change-is-happening event when they do occur. No wonder most meteorologists think the snowflakes and their professors are full of shit.

http://www.weatherbell.com/premium

Weatherbell does forecasting for industrial and government clients who depend on these forecasts to assess and determine energy, crop, and maritime conditions under weather stresses, because their lives and businesses depend on it. Like fleets, airlines, farmers, ocean shippers, municipal governments, etc. Most of the weather channels use them as well. Joe Bastardi started Accuweather in the 80s.

Tom Hickey said...

Yes there is. Astronomy, the apsidal precession for example...well, all the precessions. Wikipedia has a pretty good entry on it.

Boils down to criteria. I had a physicist friend that worked at the National Bureau of Standards. The issue is making sure that the standard is stable, like the standard meter and the atomic clock. The criterion is itself is established institutionallly.

Peter Pan said...

The IPCC models will be confirmed or invalidated by temperature observation data. Patience.

Ignacio said...

Is, metaphysically speaking, impossible to predict something with absolute certainty (which is what I meant), goes against the principles that rule our reality. How can you predict something with absolute certainty if you can't prove a truism definitively? That would require omniscience.

Apsidal precession falls on the same prediction as "the Sun will raise tomorrow on the East" (which is based on the same physical laws). It would be a bad idea to bet against that, but you cannot guarantee that will happen all the time exactly like that because you don't have complete knowledge of the governing rules of the universe.

So all those predictions always go with a disclaimer: "with a given set of conditions this will happen", this is a conditional fact (if and only if). This may sound obvious, but this basic fact is something people (including scientists) forgets all the time, because is establishing the necessary conditions where most go wrong and why theories are thrown away ultimately.

At core this is what radical uncertainty is about IMO, is not uncertain because it's unpredictable in a metaphysical way (non-deterministic), is uncertain because unpredictable in an epistemic way (lack of data knowledge and/or computational power).

MRW said...

Alright, Ignacio, barring a meteorite or spontaneous combustion of the center of the earth, the sun is going to come up in the east tomorrow morning. That is from predicting the future on the probability distribution of the past, a stochastic method. There are over 3,000+ years of observational data, however, to draw upon to make this reasonable assumption. The Chaldeans were doing it before little Jesus was born.

THIS IS NOT EQUIVALENT to the bullshit that is being fed us in the MSM about climate experiments today. The MSM analysts employed by news outfits to inform us have neither the mathematical training to read the original scientific docs, nor the scientific cajones to judge the statistical skill of research paper writers to set out a scientific hypothesis accurately. As many journals have had to admit to their dismay over the past three years.

Tom Hickey said...

The North Pole is 36 degrees hotter than it should be right nowThe North Pole is 36 degrees hotter than it should be right now

MRW said...

Watch Joe Bastardi's Saturday Summary to find out why, Tom. This week's issue is particularly interesting.
weatherbell.com/premium

for over the past 200 years, 75% of all global warming occurs during the arctic and antarctic winters. When the temp in arctic goes from -32C to -26C, that 6C rise is significant.

Anonymous said...

”.... the sun is going to come up in the east tomorrow morning”.

Of course it is – or is it?

Am not going to argue about any application of the concept, but it is also a good example of how our concepts push us into a framework where sometimes we lose the door.

There is no ‘east’ and there is no sun rise and sun set (they are in the following sense an illusion) because there is always a sunrise happening somewhere on this earth and there is always a sunset happening somewhere on this earth, contemporaneously and continuously. The earth is revolving to create this illusion.

But we say ‘oh, the sun is rising – in the east’. The earth is flat. Our society is reality.

MRW said...

jrbach,

As long as your feet are on the surface of the earth (or maybe you exist hovering 6 ft above the ground) there will always be an "east" as a point of reference.

Even if you're in a plane 33,000 ft above ground, there is always an "east." Ask a pilot.

As long as you're functioning in earth's atmosphere, there is an "east" as a point of reference. The fact that "there is always a sunrise happening somewhere on this earth" is immaterial, because "east" is directional and based on the orbit of the earth, and always stated relative to the observer.

Anonymous said...

"...Am not going to argue about any application of the concept ...." (?)

Peter Pan said...

Google REM - Stand ;)

Anonymous said...

I find human ‘perception’ a fascinating subject. Tom posted a Wiki link here a while back to a list and excellent chart on cognitive biases. There are so many errors in judgement the mind can make it’s a wonder we make it around the block on Sundays. People think that by training the bias maker, we can overcome the biases. That is not logical to me (like asking the criminal class to fix the justice system – oh, wait, sometimes ....!). There have been universities and education on the face of this earth for quite a while now, and prejudice, problems, hatred and inequality are even more entrenched and inflamed. I watched a camera crew interviewing university graduates (some of them engineers) on the steps of their graduation hall as they emerged from the ceremony – asking them to explain how the phases of the moon worked. High School teachers watching the replies squirmed in their chairs; their jaws on the floor. So watching the sun ‘rise’ is typical: people all around still think in their minds that the sun is rising, even though they know ‘logically’ about the earth’s rotation. Columbus thought native Americans were ‘Indians’ and no one has corrected the misconception since. People want to be free, and the mind goes right on creating the bars and prison cells for them.

An illusion is something that pretends to be real, but is not. It’s the same thing with ‘government running out of money’. What a cruel joke. Once people get it, and see clearly it means all constraints are imposed by humans on other humans, and why should we tolerate that – well: I don’t know what will happen. I know what should happen.

The human heart. Only energy at our disposal that can clarify mind.

Peter Pan said...

We live in purgatories of our own making. Even Spock knew that.

It would be a lot quieter if everyone were required to check their cognitive biases.