Supply Sider Scott Sumner

With substantial consequences to my growthism, Sumner suggests that the reason that Europe has leveled off at a 25% lower standard of living than the USA is that their higher taxes, and higher regulation mean that Europe can't catch up.  Summarizing:

Both continents seem to have plateaued at a mellow, low growth rate, but since Reagan, the US has been growing faster, besides being ahead to begin with.  And as Tino pointed out some time ago, that's really crazy.  Convergence happens in all modern economies.  But the problem is that the European model capped a lot lower than the US model, or indeed other countries with low-tax, low-regulation systems that have already surpassed European standards (Singapore).  Indeed, ultra-low tax Singapore is still growing at unreasonable rates.  It will be interesting to see how long they can keep it up. 

Software Quality Improvement

I just tossed together a paper for work because a boss asked about how to improve our software quality.

A little research and a lot of memory gave me a summary that looks something like this:

• The standard answer is more layers of testing.  No single method is very good, and every additional method used finds some percentage (mostly between 15% and 50%) of remaining defects.  The most cost effective approach is to stack cheap methods as high as you can.  For instance, don't use code review meetings, but rather checklisted pass-around reviews, or a even code-review tool that assists with partially automating the pass-around reviews.
• The biggest improvement available overall seems to be higher modularity in code.  Moving to more object-oriented, and more service-oriented systems.  Small-unit, decoupled systems are better for quality.  Slow improvement.
• Closely linked to the modularity factor, the ability to do unit testing in an automated fashion is my personal crusade at work.  Setting up automated unit tests, automated integration tests, and getting the ability to send emails to managers, leads, and individual developers when their code breaks the build is substantial, and idiotic to be missing.
• The fastest cheap activity that I'm aware of that makes for high quality finished products is early prototyping.  An end-to-end prototype of a software system, completed at t=20% (when waterfall projects are still gathering requirements) along the project and then given to (especially) users and the performance-testing group,is worth a substantial net quality improvement over the same prototype at t=40% or later.
• Hardcore statistical, process improvement methods FTW...the deep, long term win.  I'm personally primarily a fan of Toyota Production System, Lean and Agile (for software), and I suppose I can tolerate 6-Sigma, even if they're all cattywhumpus with their slow, top-down orientation instead of lean's fast bottom-up CI version.  PSP, as per Watts Humphreys seems to be the closest thing to a serious process improvement approach for software that doesn't require my getting into a shooting war with the Waterfall Nazis at my job. 

All y'all software smarties out there.  How good is this list for a fair summary of what we actually know?

Actor summarizes libertarianism

Looking for Stein's law, "if something cannot go on forever, it will stop", I foucnd actor Ben Stein.  Despite any minor disagreements we may have politically, he (unintentionally) sums up my view of politics beautifully in this rant about Nixon:

[Nixon] was a peacemaker. He was a lying, conniving, covering up peacemaker. He was not a lying, conniving drug addict like JFK, a lying, conniving war starter like LBJ, a lying, conniving seducer like Clinton—a lying, conniving peacemaker.

People in power have "good" motives, and because of these "good" motives, they excuse their other actions.  Problem is that if a politician has enough "good" that they're pursuing, they'll excuse everything to that end.  And usually their followers will too.  All politicians are roughly the same...they're all lying and conniving, charismatic, honestly convinced they're doing good, and causing seriously bad consequences.  Which bad consequences they cause varies from politician to politician.

Left-Libertarianism -- Health-care edition

Robin Hanson discusses a health care research study that seems to say:

Industry funded studies are significantly biased, but government funded studies are even more so.

de Soto Watch -- Tragedy of The Commons edition

Tyler Cowen links to a Elinor Ostrom interview.  In it, she argues basically the de Soto line...and she got her Nobel prize in econ for this work.

The Tragedy of the Commons is generally caused NOT by a narrow lack of property rights, but by lack of the rights of the people to negotiate an agreement between themselves.  The legal structure causes the problem.  Remove the legal barrier, and the people solve their own problem.  As per Coase, the primary problem is legal transaction costs.

Random Useless Sampling

In the middle of my walk to work, I abruptly decided to count facial hair and hats on men (because I have both).  Quick summary.  Out of about 150 men spotted (I made up rules for who to count),

I found that ~30 had facial hair, split moderately evenly between mustaches, small-ish goatee-like things, and fuller beards.  Incidentally, I have substantially more & thicker facial hair than any man in the 150 I checked.  My impression at the end, not checked well, was while <1/3 the men I saw were black,  >1/2 the facial hair I saw was on black men. 

I found ~10 that were wearing hats.  Of those hats, 2 were not baseball caps, and 1 had a brim rather than a bill.  My wide-brimmed australian outback/cowboy hat is substantially unusual. Again, black men were somewhat to substantially more likely to wear hats. 

Confirming All my biases

In email conversation, Grigs has the best line about the new bit on magnetic disabling of non-harm based ethics:

Dance, meat puppets, dance!  HAHAHAHAHAHAHAHA

http://news.bbc.co.uk/2/hi/health/8593748.stm

 Jon Haidt should be thrilled with this research, as it separates harm-based reasoning from other types of ethics.  All of us multi-module brain-ers, meat-robot-ists, and compatibilists are also awful pleased by such a confirmatory study. 

Tyler quips about this as well, and I saw it first on slashdot.

Trusting your Government

1. Megan McArdle has the most complete discussion of the Energy-Star fraud.
2. Megan also notes the Congressional war on Accounting
3. Robin Hanson bets that with lots of money pledged, the government won't redo the large study that demonstrated that most (>50%) medical spending has Zero or Negative net health value.
4. Steve Sailer (not for sensitive sensibilities) points out that while ~98% of people worldwide believe, 100% of traditions suggest, ~70% of studies support, and the basics of evolutionary theory require that there are (a) cognitive differences between people, (b) passed by inheritance, and that there are (c) statistical differences between races, our government nonetheless begins from the assumption that this cannot be true.
5. Commenter Mark Horning sends me a link from Jerry Pournelle's place that suggests positive things for the radiation hormeisis hypothesis.  In cattle, substantial radiation (Bomb testing nearby) not only increased health/lifespan substantially, but also was effectively covered up by USGov.  It's not clear whether the cover-up was intentional or simply incompetence.  Usually, I bet on incompetence.

Authority without Responsibility

Having spent several days learning from the Moldbugians, I think there's a stable position wherein we all agree.  One of a small number of core problems with government is Authority without Responsibility.

The snarkier folks in the audience might ask: Isn't that an awful lot like taxation without representation?  Not quite. Taxation without representation is Authority without Voice. 

The modern problem is that the personal cost to individuals in government of damaging the country is minimal.  So they try stupid experiments like "Health Care Reform" that don't solve any problems, and put us substantially closer to having big problems later.  Ditto the Iraq War.

The big dispute between the Modlbugians and me has come down to the best way to reduce Authority without Responsibility.

My assertion, contra Moldbug:  The only thing that has historically worked is a combination of 3 factors.  I claim that the industrial revolution was primarily a flowering in 3 locations: Holland since 1581, England since ~1660, and America since ~1700.  Further, I claim that the  growth rates caused by the industrial revolution were due to:

1. Killing an oppressor or two, thereby putting the fear of god into rulers.  In England, Charles 1 went headless followed by Cromwell's posthumous execution.  Proper fear was instilled in the rulers.  The Netherlands kicked out the Spanish in 1581.  Proper fear was instilled in the rulers.  The Americans fought the British, and thus instilled fear in the subsequent rulers of America.  Violence by the people was the core check on government power.
2. Balanced interests, thereby making divided government.   Holland in the 16th century had the Union of Utrecht, a republican, distributed power center.  England in the 17th Century had Parliament, the King, and a dozen other moderately inviolable power centers.  See my link to Nick Szabo.  Madison expressly built the structure of US government to balance power against power (Courts, Legislature, Executive -- States, Locals, Feds)
3. Ability to exit:  Europe in the 16th-18th centuries was a patchwork of small-ish countries, all somewhat at odds with one another.  This means that if you got in trouble politically somewhere (England), you had the ability (if you were educated enough to get in trouble) to decamp to Holland, Spain, Portugal, or Venice.   America had a federalist system and a frontier.

The Moldbugian response, as best as I can make it (Thanks to Devin, Foseti, and Andrew):

1. Distributed authority makes it near impossible to shoot the person responsible.  
2. Due in no small part to the rise of the university/media/civil service as a unitary entity (no discernable differences of opinion between the 3), there is no longer any capability to build balanced power, as the interests are the same.
3. There is no frontier and America is a hegemon.  Exit is impossible.
4. Therefore, regardless its historical merits, the Madisonian solution is no longer possible.

My riposte:

1. I have advocated before, and continue to advocate the following...which appears thoroughly Modlbugian.  People who make and enforce decisions should sign their names to them, and should be responsible (read liable) for them.  My boss made me do it is NOT a defense.  True in military, corporation, and government.  Government should not be allowed to indemnify.  Insofar as Moldbug's line is that we need to be able to identify who to hang for each bad decision, I agree.  Insofar as he's saying more than that, I am less sanguine about it.
2. If one were to create an amendment, even in the current system, wherein TOTAL government expenditures (State, federal, local) were limited to X% of GDP (start at 30, decrease to 5 over 25 years) , it would create a system wherein the interests of Moldbug's Cthuloid monster would no longer be cohesive cohesive.  Rather than being able to extract money from the private sector, Government would be in competition against itself.  Arguably, it is this feature of the pre-industrial revolution world along with the separation of powers that built the Industrial revolution.  As Clark's Malthusian view informs us, there simply wasn't enough surplus to extract...and so all levels of government were forced to compete for the same small pie.  Perhaps government pie restriction to the zero-sum game they tend to believe in would solve a problem. 
3. There is indeed no frontier.  America itself now constitutes a monopoly of the type that antitrust regulators hate.  It needs to be broken up, like AT&T.  The federal government should be left with only the national-defense business unit.  Smallness and competition is the only solution.  This may be the key issue.  What I really want is for the meaningful differences between California's entirely disfunctional government, Illinois's hideously corrupt government, and Texas's minimalist approach to become obvious.  With Texas getting better results on every front than California, with fewer regulations, lower taxes, better economic growth, and better social services (though less pretty scenery -- not terribly impactable through government programs), eventually smaller systems under mild competition will gravitate towards successful models.  
4. Madison may indeed no longer be possible.  However, I believe that the odds of a hard-cap on total government expenditures, a real accountability measure (preferably that includes violence), and/or a breakup of the American monopoly are all (as incremental improvements) more likely to succeed than the Moldbugian reset.  It may be that the solution that the MOST successful state has matches Moldbug's.  I'd bet against it though.  Not because I think Moldbug's approach doesn't have value...but because I think that experimentation is better attuned to reality than I am (or Moldbug is). 

History & Moldbug

Question?  How much of Moldbug's theory rests on Carlyle's interpretation of History?  How much of Moldbug is undermined if Carlyle is wrong on his history?

In particular, (similar to a comment I made earlier) I am of the opinion that English Monarchy around the time of the Industrial Revolution...the time Carlyle likes to write about(?) was the weakest Monarchy ever found 'til that point.  Oliver Cromwell had recently been wandering about beheading Charleses , and the Kings were therefore tremendously circumspect in their exercise of power. 

Ability to exit (to Holland, or America) + a government afraid of the population + uncertainty over who had the authority to regulate things led to way fewer regulations than normal, which led to the industrial revolution...and all of modern goodness. 

Different model than Moldbug.  I claim that this has more historical viability than the Moldbugian model.  Incidentally, it also argues that the primary problems with the modern US are that (1) the government is not properly afraid of the population.  (2) the lines of authority are clear (the states do not clearly have the right to resist federal encroachments).  (3) no real variance wherein people can escape to other (different) models.  ALL of those mitigate for federalism.  More jury nullification and some method of incentivizing the courts to side with the states on 9th, 10th amendment challenges seem to be the best prescription.

I'm not claiming that I'm necessarily right...but that there are other models than Moldbug's.  And I'm far less convinced of Moldbug's historically-resting model now that I've seen someone else with a legal history understanding arguing the same thing I've been arguing.

And models are all wrong.  It's just about where they go wrong.

Labor Theory of Value

As part of my continuing series where I argue that people who are both smarter and know more than I do about relevant topics are unnecessarily obtuse:

Tyler Cowen starts a meme "What is the biggest flaw the Labor Theory of Value"
David Henderson follows up.

Since I think two exceptionally smart economists are both wrong, it's probably my failure.  Regardless, I'm not smart enough to shut up.  Here's the problem in 4 words. 

Value is demand side.

Value is a measure of how much someone wants something on the demand side. Labor Theory of Value attempts to move value to be a supply-side issue, rather than a demand-side issue.  Broken epistemology.  End of story. 

Tyler's explanation fails because he's talking about how to salvage LToV, not about the core issue which is that they want to define value wrong.
Henderson focuses on Randomness, rather than definitional issues.

Nick Szabo rocks

His stuff seems much more in line with (a) what I understand, (b) what I've been arguing recently.

First post I found, as directed by Devin says:

From the Justinian Code we get two totalitarian superstitions: first, that there must be a locus of power, a "sovereign", somewhere in any political system. In Justinian's Empire, this locus was the emperor himself, whose word was law. From this philosophy came the view of the sovereign king or dictator espoused by Bodin and Hobbes. In this model the king is the "head" and the rest of the "body politic" is controlled by the king, just as our brains control our bodies. Under Rosseau and Bentham, this locus was switched to "the people" or to, in practice, a parliament that supposedly represented "the people." Under the extreme sovereigntist view, separation of powers, federalism, and political property rights are all an illusion -- all power is just a revocable delegation from a supreme locus of sovereignty.

Actual English law and political structure were very different. Under this law, royal power was actually divided among the King, the King's counselors, Parliament, and justices. None of these entities was the "locus" of power but all played crucial roles. Furthermore, much of this power had been granted to other entities -- nobles, lords proprietor, municipal and colonial and church corporations, guilds, and so on -- in the form of largely irrevocable political property rights. Under the sovereigntist view, taught in universities, all such property grants were merely revocable delegations. But under the actual common law, taught to the actual lawyers and judges of the time in the Inns of Court (institutions completely independent of universities, and thus largely uninfected by Roman Law), these delegations were property rights forfeitable back to the grantor -- and the original grantor was the Crown -- only under extreme breach of grant conditions under a quo warranto proceeding.

As I've said before...it was only under extreme conditions of limited powers that capitalism ever got going...and it will only be under extreme limited powers that it ever continues.  The Moldbugian pro-authority superstition is just that.

I now officially owe Devin a beverage of choice if I ever see him.

What's natural?

While I tend not to think much of normative naturalism, I do follow the Bacon in that obedience is necessary to command.

In social systems, there seems to be an interesting dichotomy.  An awful lot of machiavellians suggest that power dymanics is inherent in human nature.  An awful lot of other folks suggest that power equalization is the only stable circumstance.  Best citation on that recently is Robin Hanson:

This famous novel suggests that only our “civilized” rules and culture keeps up from the fate of our “savage” ancestors, who were violent dominating rule-less animals.  But though this may be true regarding our distant primate ancestors of six or more million years ago, it is quite unfair slander regarding our face-painting forager ancestors of ten thousand or more years ago.

 Alternatively:

Isolated nomadic forager bands today are “fossils” with crucial clues about our distant ancestors.  Anthropologists who study them report that overt dominance is rare, and long distances make war rare (as 4 million year old fossils suggest). Foragers live in tight quarters and use language to express and enforce social norms on food sharing, non-violence, mating freedom, communal decision making, and norm enforcement.  Anger, bragging, giving orders, and anything remotely resembling dominance among men is punished by avoidance, exile, and death as required.  Human’s unusual hidden female fertility also limits male dominance temptations.

Here's the question...if hatred of inequality is deep in our genes...how much success is it reasonable to expect of systems that rely on massive inequality (autocracies).

Madison vs. Modlbug

As an experimentalist, it's probably worthwhile to take a look at the last grand experiment: The USA, do a bit of analysis.  As the last fabulously working (for whatever reason) experiment, this should be the model off which the next model is tried. Less likely to fail badly than a "new" thing.

1. The relatively laissez-faire capitalism/rule of law adopted for most of the 19th century did more for our wealth/happiness/all good things than our system of government did.
   
2. The ability to shift between states, and even out of the states to the frontier was enormous in our success thus far.
   
3. The setup as a republic with some democratic involvement failed.  Democracy seems to have grown like a cancer, with  precisely the results that Plato expected (voting themselves bread).
   
4. The federalist system seems to have failed.  It has been net beneficial to everyone in power to move repsonsibility from the locality to the state to the national government.  This is a catastrophic failure mode...needs to be fixed if anything like our current system is to be tried again.  Local decisions, and competition between jurisdictions is tremendously important.
   
5. Size matters.  Smaller units simply function more effectively.  Since this is true of darn near everything else, from code to corporations, this isn't shocking.
   
6. Representative democracy as implemented here and in Europe seems to have failed.  The public choice forces are too strong for any silly idealism to resist. So far direct democracy, only really managed by the Swiss, seems to have been a fabulously higher success over time than the representative democracy or autocracies.
   
7. The historical Scottish/English enlightenment's aristocratic character of restraint seems to have been instrumental in getting anything to work at all.  Any further work in government should rely more heavily on the character of the leaders.
   
8. The success that it did have came entirely from proper balancing between the powers of government.  Any further work in the direction of limiting government should rely even further on balancing powers&interests, and less on paper restraints.
   
9. Accountability is an issue.  Pelosi/Reid/Obama has no accountability for the fact that their health-care plan almost certainly advances the federal entitlements crisis by 15+ years. 
   

Rationality test:

On the first experiment, which one has a higher chance of working over the first 100 years?  A rerun of the Madisonian experiment, with a few tweaks of the types discussed above, or a Moldbugian restart?

Hint: I think the answer is obvious.  And I don't think that the answer is identical to "which system would work best if you gave me 20 iterations to work out the bugs."  And I don't think that answer is identical to "which system would have the highest rate of stellar success  ( 3 +/- 3%)".

Against rationalism

Insofar as we're willing to separate the world into rationalists and empiricists, I think we can do a decent job of defining the difference thusly:

• Rationalists think that there is a good chance of succeeding without experimentation.
• Empiricists don't.

This means that we can probably define a range of opinion...and indeed we can define the fact of the matter in most cases (by experiment).  How likely is it to gain a working, good enough system without experiment?

As a new empiricist (my academic background is pure math and philosophy), I am of course a bit evangelical...and the biggest errors made by the smartest people I read all seem to me to be of the same category: Rationalism.

Hence...a question:  Given a new complex product, and no experimentation/iteration, how likely is it that the product succeeds?  How likely is it that it is close?  My assumption is that the probability is roughly 0%.  Given almost as much experimentation as you want...how likely is it that the first prototype you build is roughly good enough?  Again...I suggest 0%.

I know that this is true of complex software (more than hello world --  though I once wrote 200 lines to solve a toy problem, compiled it clean and it ran successfully -- once in the 30 years I've been programming.).  It is also famously true of Battle plans.  It is also true of engineering product development.  And it is mostly true of corporate activity (correcting for delusions).  I believe this to be  true of Artificial Intelligence.  And also of government systems. 

If I'm right as a hardcore evangelical empiricist, then all the theorizing you wish to do is near-useless until you get on the ground and try a few different things. 

Rationalism in government.


Madison and company did an insanely good job.  200 year good run by design, world-leader for at least 100, and growth-driver as well.  Wow.  The Swiss seem to have a fabulous, 700+ year model as well with magnificent stability, if not the same level of growth-driving.  I'm personally fond of Singapore's model, but the country is barely older than I am. 

The 1792 French model sucked.  The 1921 German attempt at "democracy" failed catastrophically.  The 1917 Russian revolution was worse.  The Asian Communists were, if anything, even worse (China, Korea, Vietnam, Cambodia).

On systems that were stable for >200 years, we're rather shy of examples.  Feudalism in Europe?  Roman (+/- Holy) empire.  Chinese Empire.  Islamic Caliphate.  The age of absolutism just wasn't that long.  And none of these escaped Clark's Malthusian trap for the majority of the people.  One is inclined to believe they all encourage stagnation, or at least the bureaucracy that stifles growth.  Only stuff left is the David Friedman / Roderick Long unearthings of Celtic and Norse anarchies, some of which appear to have lasted longer than the States so far, but which didn't encourage growth either.

Is empiricism the wrong approach here?  Or is someone going to argue that .. yes, it's true of all other complex systems known to man, but in this particular case, there's a better than 25% chance that Big Design, No Testing  would succeed for government?

Education

Only Robin Hanson could make John Taylor Gatto and his uber-powerful essay The 7-lesson schoolteacher look like a moderate.   Conclusion:

While our kids are segregated into schools where light monitoring lets them terrorize each other and form dominance hierarchies, forager kids are mixed among forager adults, who enforce their strong social norms against violence and domination.  At school, our kids are rated and ranked far more often than most adults will tolerate, even though this actually slows their learning!

It seems that modern schools function in part to help humans overcome their (genetically and culturally) inherited aversions to hierarchy and dominance.  Modern workplaces require workers who are far more accepting of being told what to do when, and schools prepare kids to accept this more primate-like environment.

Most important post this month

Long post to read, but Falkenstein has put together a Moldbugian-length analysis of one simple idea.  Envy dominates greed.  If true, whole piles of stuff that we currently believe becomes false (utilitarianism is untenable, for instance)...or in Aretaevian terms, much less useful than other models.

Read the whole thing.

Incentives

As a thought-experiment, and just because I knew it would be hard, I went about attempting to define a metric that captures well-being...of the kind that Moldbug would want for his CEO-autocrat:

It has to be a weighted average.
It has to include both economic growth and violence measures
It has to disincentivize Kleptocracy.
It has to be realistic (Ethnic swedes get better outcomes on violence and income than ethnic nigerians, regardless where they live)

Current thinking:

mean  + median + mode are all important
So is success differential vs. world-average per demographic group.

Measure lifespan and income (in whichever way disincentivizes larger families less)


As a Cartesian product, that's an 8-factor measurement that should capture things.

I haven't captured happiness measures.  Autonomy, charity and physical activity, and physical pain (chronic pain is bad juju, and a lot [most?] of medicine exists for this purpose alone) are my best guesses for things to actually measure for well-being beyond wealth, but I'm not sure how to measure any of those well.  (% of hours in a year on charity and activity?  % of decisions felt to be in my control for autonomy? ).

I haven't captured justice measures, and I think that under my measure self-defense becomes bad because average lifespan reduction dominates crime reduction there. 

Do I think I'm right?  No.  I mostly just  wanted to suggest that the problem was hard, and maybe get some of my smart commentariat thinking on it.

Moldbug vs. the Libertarians vs. Aretae

I have been (I hope) well-educated in Moldbuggery due significantly to my excellent commentariat.  Isegoria, Andrew, Devin and Foseti have been keeping me well on my toes and thinking clearly, (even if I occasionally miss a response to a comment).

As I read more, I am more and more convinced that a careful analysis of the problem suggests that the distance between modern libertarianism, Moldbug, and myself is not that far.

Things that all 3 camps seem to agree on...I think these are almost trivially true. 

• Economic growth is the root cause of most human happiness (liberty, work hours, amelioration of poverty, environmental beautification, clean air, clean water, concern for the poor).
• Growth is, itself, almost entirely maximized by private enterprise, and almost exclusively hindered by state interventions.
• State interventions are primarily done to either benefit some group of elites (rich folks, or rich companies) at some net cost to the rest of us, or to signal some positive feature of government or politician (like caring) at some net cost to the rest of us.
• Those elements of human well-being that are not (at root) caused by economic growth are caused by lack of violence (war and violent crime and physical property crime, employer thugs and union thugs).
• Organizations tend to grow without limit.
• The best historical cases of growth have all been cases where the government is largely inactive or heavily limited (17th Century Holland, 18th century England, 19th century America, 20th century Hong Kong/Singapore, 21st century China), with a strong, trading, lets-get-rich culture. 
• The stifling growth of government generally parallels both the growth of democracy and the growth of total wealth
• We have never seen a successful long-term limitation of government growth based on paper constraints.  Constitutionalism has not yet worked long-term.
• People in general don't know what works, and are moderately easy to sway, in predictable, prejudicial, pro-government ways.

The open questions:

• Is it possible at all to generate long-term solutions?  (Moldbug: yes, with incentives; Romer: maybe; try different things; Patri: probably: float away;  Anarchists: Yes: eliminate government ; BBdM: no? It's inherent in government; Aretae: I don't know)
• How best to limit the stifling effect of government?  (Moldbug: CEO-ish incentives & responsibility; Romer: try different things; Patri: New systems (on water); Anarchists: burn it down;  BBdM: The Selectorate matters a lot; Aretae: I don't know)
• How much of government growth is due to democracy, and how much is due simply to the growth of wealth, because it's hard to skim 20% off a subsistence farmer? (Moldbug: democracy;  Aretae: I suspect wealth is most of it).
• Where is the biggest problem in government?  (Moldbug: Democracy.  Libertarians in general: excessive power.  Anarchists: the existence of government.  Aretae:  Our reliable sample set is tiny in quantity and time and quite homogenous.  We don't actually know at all)
• What is more likely to constrain real world solutions (Moldbug: "Imperium is conserved"; BBdM: The selectorate wins;  Aretae: First prototypes always fail.)

I guess here's where I sit.

Moldbug defines problems pretty well...Limitations on government have never worked before.
He is tremendously insightful about the modern clerisy, and their history, but his core, semi-unique belief is that the libertarians mis-state the problem.  While libertarians agree that paper restraints cannot stop the growth of government (just like Madison did in his design of government, attempting to get parts in conflict), Moldbug goes further and believes that NOTHING can stop the growth of government authority.  However, if authority is simply granted, the costs of government can be minimized by aligning incentives properly (same task Madison tried, but failed at)...and our best current guess is CEO-style structure.

However...Engineering, System design, software, process improvement, war, and every other HARD problem we've ever faced tells us that

1. In practice, theory and practice are different (In theory, they're the same).
2. First solutions never quite work
• though iteration 18 does sometimes, and sometimes your first iteration even fails less badly than what you had before.  Usually iteration 1 fails worse than what you had before
3. There aren't really any solutions, only trade-offs.
4. More data often helps
• My favorite real data (and models from the data) is from BBdM and de Soto...and it doesn't seem to me line up with the Moldbugian theory
5. With enough eyes, all bugs are shallow.
6. Detailed structure and incentives matter a lot
• Since the details matter so much, they are usually designed with good intentions, but fail rapidly on a technicality

I don't think any of us are far off on the problem.
I think we're moderately distant on which model we think best explains the problem.
I always think that "all models are wrong, but some are useful".
I don't know how convinced anyone is of their respective models (I think Patri gave himself a 50% chance, which is almost certainly self-biased).  I'm not very convinced of any specific solution, but I am convinced that it's obvious that NO ONE else should be convinced either.  Romer and Patri seem to be 100x more likely to be correct (try different things to find a solution) than Moldbug (try this solution), even if Moldbug were to have the best idea going right now.

QoTD

Patri cites Levitt's confusion about why people get mad at his research on solving Global Warming (geoengineering is a substantially, maybe insanely, CHEAPER solution than anything else).  Levitt is especially upset because everyone else who asked the same question (what's cost effective) seems to have reached the same conclusion.  Patri's Hansonian quip:

Silly economists, asking what the best way is to achieve our purported goals, instead of signaling their concern about our planet and their alignment with the natural world.

Libertarianism Question

This time from an email list:

I guess the only thing I don't understand is where this axiomatic faith comes from that reducing the size of government is always the best regardless of how bad the outcomes are for people.

 There's a zillion answers to this one.  My honest answers, in order of importance go something like this:

1.  Authority.  I personally HATE the very idea of authority.  The greater the authority, the greater the use of the authority, and the less you can opt out, the worse things are.  Government is the ultimate in un-opt-out-able, high consequences of disobedience authority.  No BS please about the ability to opt out of this pile of crap into deeper piles of crap.  Until Patri or Paul gets their thing going, no such thing.  If I were putting intellectual pretensions on this, I'd go all Randian on trade or force, and justify that way...but it's pre-justification.  Hate it.  I especially hate people who think they're smarter than you telling you what you should do, or how something is going to work, and being wrong.  Krugman and Klein, Pelosi, Obama, and Reid talking Obamacare got the precise thing I hate nailed perfectly.

2.  Growthism.  Economic growth is responsible for roughly 110% of the difference between our lifestyle and that of sub-saharan Africa.  Economic growth is entirely driven by private activity, bottom up, with the government able (almost) only to get in the way.  The extent to which government regulates is almost identical to the extent to which government squelches innovation and therefore growth.  As per a previous post, anything that slows or stops growth is net bad for everyone in the future...including (usually) most people living right now...but because of short term stupidities, people make the trade of good now for worse later.

3.  Error.   Everyone's wrong a lot.  Errors in business usually cost the business a lot of money.  Erros in government cost everyone else a lot of money...given how often folks in general are wrong, government is highly unsafe.

4.  Political Realism/Public Choice/Regulatory Capture/Incentives/Econ.  The incentives structure of government is atrocious.  Any reasonable study of organizations at all eventually concludes that government as we know it cannot be otherwise than 90% a parasitic entity that siphons value off of industry and creates almost no value.  The other 10% is up for grabs...but with an easy 90% being complete and total parasitic waste, it's worth opposing all of it, because my BS detectors will fail on significant chunks of what good it can do.  This is especially important when analyzing systems that are supposed to help the poor.  Also, regulation and taxation is economically complicated and usually works in predictable ways, but not the ways that the law-makers say it will.

5.  Feedback.  Aretae's first law is that the quality of a system is defined by the quality of the feedback loop.  Feedback loops in government are, if anything, negative to the nominal purpose.  Government workers working on poverty have real incentives to prolong poverty, and negative incentives to decrease poverty.  Unsurprisingly, poverty hasn't decreased, despite 9-11 digits worth of government money thrown at the problem.  Any system without good feedback systems loses.  Apply this reasoning to any governmental system at all...and the nominal goals of the organization and the feedback systems for the individuals in the organization are at odds.

6.  Crowding.  Not only is Government authoritarian, bad for growth, wrong a lot, public-choice constrained to try to do bad stuff, and hideous on feedback...it also crowds out private activities which would arise in the absence of government.  We have no decent education system because we have a public school system.  We have weak mail delivery because we have a post office. 

Note...none of this says that I am opposed on principle to all programs.  I hate being taxed so that the government can spend money that decreases the standard of living of everyone now, and everyone in the future, and so that they can ask for even more next year in a nearly-guaranteed cycle that has no path to improvement.

I don't generally oppose appropriately incentivized programs for the poor, though I oppose the implementation of all the ones we have. My wife sends me this, for instance...and the big question is ... is this an anomaly, or must you expect this, and be surprised by anything different?

I don't generally oppose support for those who cannot support themselves, though the way we do it now sucks.

I don't oppose Health Care, even government run health care...but the present bill may have been the worst one possible...far worse than single payer, and mind-bogglingly worse than Singapore-style mandated HSAs with government-managed support for the poor, and government-run catastrophic insurance.  And I believe they have less different outcomes between rich and poor than we do...while our current crappy system gives less difference between the care of rich and poor than the single-payer systems (like Canada). 

I heartily oppose discrimination from personal interest.  But I hate current anti-discrimination law, and am worried that it net-damages the prospects of my eldest, who is quite dark-skinned, though whenever I have a personal business, it is nominally owned by my black wife, in order to gain benefits I shouldn't be able to gain.

Kling/Friedman-style flat income tax rate, no deductions but a 20k/person personal deduction to replace all taxes and all welfare programs?  All the way onboard.

Singapore-style government mandated HSAs and support for the poor and catastrophe?  Yesterday, please. 

Federalism, and super-federalism pushing to counties and cities, including tax behavior, so that the socialists can see the results of their policies, and the libertarians can see the results of theirs, as we segregate by preference?  Wow!  When can we start?

But overall...the 6 issues I have dominate my concerns..  Worse, none of the liberal arguments for government ever seem to address any of my issues at all.  And at least most of my issues seem reasonable.

Devin's Question

On my previous post, commenter Devin described a student who was bad at math, had an epiphany @ 15 about the quadratic equation, and went on to a Ph.D.  He then asked the following:

Did you ever encounter students like this? Did they go from being obviously dumb/low IQ to being smart? Or were they always smart and hi IQ - they just have a problem with one particular subject, until the light bulb goes off?

A series of great questions.   First, the snotty fast answer:

It doesn't take a lot of brilliance to get to a Ph.D.  It takes mostly persistence and hard work, and an IQ in the 2+ Sigma range.  Maybe for mathy fields, it takes closer to 3 Sigma, but still, that's not impressive.  Almost all of us know some one who is substantially smarter than an average professor, but without the Ph.D.  Probably that also indicates a lack of academic habits (and thought habits and prejudices), and a lack of a 10K academic articles read in one specialty.  This is not entirely off the cuff.  But it shouldn't be the whole answer.

The slower, fairer version:

This is every teacher's dream.  This is why people go into teaching as a profession.  Finding the kid who didn't understand it...who is having a hard time with a subject...and for whom your 7th explanation finally clicks, and the student begins to veritably glow with understanding...that is the God-moment for a teacher.  If you've never felt it...you don't understand why people are teachers...and if you have felt it, you don't know how anyone can ever want to do something else.   Mild exaggeration?  Sure.  The point stands.

This happens...sometimes.  If a normal teacher teaches Algebra to 6 classes of 30 kids for a (4-month) semester, she may see this once.  Maybe twice.  It's generally worth all the crap for the whole semester for that once...if it happens.  If it doesn't happen for 2 or 3 or 4 semesters, maybe she gets tired of teaching, and quits. 

The Aretae explanation: 
Interest is huge.  Practice is huge.  Cognitive readiness is huge.  IQ is merely big.  A smart enough, but slow-developing student who figures something out, gains an interest, and puts in effort to a Ph.D. 10 (or sometimes 20) years later is a real possibility.  I've personally seen it done: Ph.D. in the real sciences by kids who were academically near dropping out of school but for an art teacher who believed in them, or a sports coach.  And those weren't the 4+ sigma kids dropping out because school was hatefully slow and stupid...but rather kids who just weren't in the whole academic thing at 14. 

Much more often, in subjects...I've seen folks who were more than a bit competent in a subject...but whose motivation and self-efficacy were destroyed by bad teachers.  A particular teacher of pre-calculus many years ago did that to a number of people I knew.  Her bad teaching broke the math spirit of an awful lot of otherwise math-comfortable students who never pursued the math that they would have been perfectly able to do, given a tolerable teacher and a decent book (they had neither).

With multiple thousand hours spent tutoring, I concluded that half my job was to straighten out the messes made by bad teachers.  The other half was to do catch-up that no one else had noticed needed done.  I mean, really...how DO you do fractions in a college level teaching-certificate pre-algebra course (or worse, when studying for the teacher-cert test) if your multiplication tables aren't solid?  And how the hell do you teach the 6th graders when you personally don't actually understand the material at better than test-cram level?  But I digress.

On top of all this, there's also the Pygmalion effect.  Learning is massively (It's probably close to a 1 stdv difference) impacted by the teacher's expectations of a class.  Futhermore, in school, IQs are not generally well measured, and it's not too hard to get a +1.5 Sigma kid placed in the slow group due to bored behavior problems.  Once there, it's also often true that they're tracked in, and then you're really screwed.

As an aside...I'm something of an opinion outlier regarding subject-level intelligence.  I think that the whole Howard Gardner multiple intelligences thing is 90% premium grade hokum.  There is g, and there is interest and practice.  I suppose there are (almost necessarily, given evolutionary understandings of the brain) various modules, and that verbal and spatial intelligence vary separately a little.  But basically, g + practice (mostly determined by interest) seems to be far and away the best explanation for the picture we see.  

Rob's question #2

In the comments, Rob asked:

If you start with a good instructional model what level of math should kids be doing at the end of sixth grade?

I answered this once...in the spirit of attempting to answer the question Rob wanted me to answer, rather than following my strong impulse and ranting extensively about how this is the wrong question.  This is the post I wanted to write, now that I've calmed down:

This is the wrong question.  Even asking the question probably causes more problems than it solves.

How about these questions instead?

How much math do people really actually need?
When would the lack of math cause problems for someone if they didn't have it?
How could you generate interest in math?
What matters in education beyond when someone learns something?

NEED?
 
If a person wishes to be an Architect/Engineer...they really should have a couple years familiarity with Calculus (Differential Equations is just a different way to write the ideas in Calculus)...which really means 4 years of comfort with Algebra, or 5-6 total years of math study (after correspondence theory of numbers -- count 1 number per object -- which I've seen missing as old as 7 or 8). 

If someone wants to study any of the sciences (social or otherwise), they ought to understand statistics...which can be done after 2 quarters of college algebra.  Probably also 4 total years of math to understanding (1 year to do, 1 year to settle for each of algebra and statistics).  Physical scientists (and perhaps economists) should also learn the calculus, and statistics with calculus.

If you're not a scientist or engineer, how much use is math in your life?  I say you need to be able to estimate whether your calculator is right, and do basic decimals (for balancing checkbooks), fractions (for cooking and doubling recipes, and for home repair with measurements), and percentages (sales).  And honestly, every bit of this that you need can be done in 1 year, any time before they need it.

Auto mechanics?  Need some ratios for gear ratios...some understanding of what a PSI is for tires...but not much else. 

PROBLEMS

Lack of math causes problems when you need to use it for something else, or when you need to use it in life.  Most of it is irrelevant until you (a) have a job, or sometimes (b) an allowance, though it doesn't matter much then, or (c) are trying to do further academics. 

SHOW COST OF MISSING?


I like math-y games.  Also reality.  Cook with kids, and double (or 1.5x recipes).  balance checkbooks, or give an allowance that they can keep track of.  It's not a cost if they can do everything they want without it...so you get to try to build a life where one of the following holds:

Either math's prevalence is obvious, rather than hidden like it is in normal life when parents keep the math out of sight of the kids.
Or math is a source of games and puzzles that the kid would like to figure out, and can use math to do so.
Social proof: other people doing math, using math,or playing with math at a level above what the kid can do is usually a much better motivator than anything a teacher can do...

Overall..if someone's not going to be a engineer or scientist, I'm not convinced that their life is more improved by knowing Algebra than by knowing something about Art History or about juggling.

And that's me with substantial math-fu, a math degree, and who believes that a good understanding of the world requires both concepts of limit and infinity not encountered until Calculus, and a good handle on probability and statistics.  I'd bet someone without such a math background might suppose that one could get by with even less math than I suggest.

Now the question...supposing that it's only useful to scientists and engineers, why do we teach it to everyone?

What matters?
IQ, Interest, Skill at learning solo, and self-efficacy (how much you believe you can succeed on-topic) are the most important factors in learning...and few of those are impacted (much) by school.

If you don't have interest, game over...no learning.  Sure you may get test-passing...but I've tutored those folks: "test-able" and "learned" are two entirely different things.  This is especially true if test-prep is disallowed.

If you don't learn how to learn, solo, then you lose as soon as you leave school.  Since school does/should only comprise on the order of 1/5-1/10 of a person's life, a focus on school learning is stupid.  Proper schooling should primarily focus on issues of how to learn, more than specific content.

Self-efficacy may be the most interesting, because the costs to self-efficacy of teaching folks when they're not ready are large...and because most learning is normally distributed...you're likely to have some folks in a class that are not ready for the topic.  Is it worth damaging their belief that they can succeed, in order to follow a curriculum?

IQ...well, it's there...and seems not to respond to interventions.




Summary:
Why the heck does anyone ask questions as misguided as "what should a kid be doing at various grades?"  Clearly the question is bass-ackwards, and suitable only for messed up approaches that uses the teacup model of education (You are a vessel, and I will fill you with information).

< / RANT >

Benezet (part II)

First, an aside.  Here's 3 articles on the Benezet approach.  Roughly what is suggested was that actually solving the kinds of problems that might be found in the real world was not impacted by math studies.  Rather...learning to address the world by means of dialogue led you to be able to think about problems, while traditional instruction decreased one's ability to address real world problems, though it did increase folks ability to pass traditional math tests...but only if the problems were structured right.  Change the format of the problems, and the math becomes a detriment, rather than a help, for most folks. 

For what it's worth, this matches my experience well.  When people take math, they learn to follow the procedures...there's the hoses filling a pool problem, the two trains problem, and very simply normal math instruction doesn't teach math as something you use to understand the world, but instead teaches it as something separate from the world. 

The standard answer I know is to ignore word problems until algebra.  Algebra makes them make sense, and before them, there are only formulaic answers that teach you to do problems of a very very narrow type (and discourage you mostly from actually thinking about the others).  But most folks in algebra spend hugely less time than is appropriate turning words into problems.  I think that a decent Algebra would be half translation, and half-symbol manipulation.  The translation back-and-forth from words to symbols is the understanding of Algebra's application to the world...and the manipulation of symbols is the understanding of the nature of math.  As it is..it's about 90-10 in favor of symbol manipulation.

New math, or new new math or whatever's current at least gets this part right in theory.  Now if only the people teaching the math knew enough math to do that, we'd be in good shape.   But that's kinda unrealistic.  What I read in Bezenet's approach is that he basically used discussion to get folks comfy with an intuition about numbers and words, rather than talking about formulas.  Once folks are number-comfortable...then they have room to learn.  Without number-comfort, they aren't.

Where should kids be

Rob asks a question in the comments:

"If you start with a good instructional model what level of math should kids be doing at the end of sixth grade? (I am happy if you break it down by IQ group)"

I'll answer as well as I can, based on experience and reading:

2 Standard Deviations above normal (IQ 130, top 2%)...every single one of them is ready to be taught Algebra at 10, if they get 1 years of prior decent math instruction @ 9. And that's probably conservative.  All the ones who succeed here, though,  have encouragement, and support, and a teacher who actually understands math.

The 4 Stdv kids I've seen (of course, not many...probably a half-dozen) are able to (though they didn't all do it this way) teach themselves algebra from a tolerable book by themselves by 9 or 10, and most of the ones who did it did it at a college level (Algebra, Algebra II, Trig, Precalc all in 9 months).  I've seen it done solo-ish from a book at 8 (me tutoring occasionally), and also solo-ish off decent-to-poor software.

1 stdv kids (IQ 115)?  I'm assuming that Sudbury-attendance, or magnet school in a college town, or unschooling parents constitutes evidence of +1Stdv, and my experience matches Daniel Greenburg's.  Across-section of kids from 8-14 can do all of pre-algebra in 6 months, 1 total hour of instruction (over 2 days) a week, given interest. 

All this stuff is consistent, by the way with the Army research suggesting that IQ 80 takes about 5x as long to learn something as a IQ 120...normalizing via geometric mean, and stretching just a smidge...let's call it a learning speed difference of 1.6x per standard deviation.  2 standard deviations = 3+x learning speed.  3stdv=5+x.  4stdv = 10x

And...given the aretaevian multi-factor model...We need translation factors for interest and for teachers.

My best guess is that typical teachers' variance is worth about half as much as IQ variance: 1 stdv variance in teaching skill ~= 1.25x learning speed.  The data I've seen (90th %ile teachers are about 3x as effective of teachers as 10th %ile ones [in grade level advancement per year]) lines up pretty well with that.  I'll broaden this to constitute effective educational practice, because I think that in many cases, learning proceeds without any structure. 

The interest variant is HUGE, though...I'm inclined to place it at significantly larger than the value of IQ.  I've seen a(n unschooled) 2-3 stdv IQ kid (who was learning programming from me) walk in with weak algebra skills, and negotiate learning complex numbers from me.  In 2 hours of instruction, he learned enough complex numbers to follow up starting to learn of the C++ graphics libraries (solo), and come back in 2 weeks having built some fractal generators.   Interest (as economistically measured by willingness to throw hours off-assignment learning stuff) is worth large numbers.

Similarly, disinterest variance is HUGE as well.  Someone who is committed to not learning math, or not learning history is at a substantial disadvantage compared to someone without said attitude.  Unfortunately, I don't have a unit-wise measure here.  I'll say that I'm pretty convinced that "I want to learn it, let me find a teacher, or a book, and spend 20 hours trying, even if it's hard" is worth on the order of 2-3stdv of IQ.

And finally...there's cognitive development.  I have seen kids who were not cognitively ready for Algebra and formal operations.  I've seen some at 8, and some at 14 and some at 30.


My hypothesis (testable):

I believe that normal kids hit the level of formal operations necessary for algebra at around 10-12 (which lines up well with Piaget's number of 11).   After formal operations, I think that Algebra is a 1-year learning activity...

1 stdv of IQ is worth about 1-1.5 years of time to start, and about 1.6x learning speed.

Interest is worth up to +/- 3stdv of IQ.  Both bad teaching, and lots of failure can cause disinterest.

Effective educational practice is worth about 1/2 as much as IQ (per stdv), 1.25x learning speed,

Teaching proceeds at a fixed rate, and will slow down anyone ahead of the fixed rate.

Having no method (beyond a book) costs about 1stdv of IQ in making it harder (you learn about 0.6 as fast as you would with a well-targeted learning method, but for most (3+ Sigma) smarter folks, this is still faster than what any teacher is willing to teach at)...

My assumption is that self-efficacy and conscientiousness are also factors, that can at least match IQ in learning-speed for 2 stdv...but I don't know that the range is as large. 


Does that help?

Datum

If you are going to understand how things work...it is often good to study outliers that work, and normal cases that don't work, to try to understand them.  Black body radiation explains most of the rocky planets pretty well: Mercury, Mars, Pluto, and the asteroids.  A lot less so for Venus, Earth. To really understand the temperature of planets, you can't just stick stick with the normal ones.  You also have to investigate the outliers.

In the quest for understanding education, one of the more unusual, and therefore worth understanding approaches is the Waldorf/Steiner schools.  Their most interesting oddnesses to me are:

1. A cohort which advances together with a teacher.  8 years together.  Almost family.  Unlike in normal schools...the kids don't really know one another much better than the teacher knows them. 
2. Pursue the arts, and avoid academics for a LONG time (up to 14 for some subjects)

 There appears to be significant success in this approach, at least in the parents' eyes.  Why?

Most recent datum, cribbed from slashdot, is this one.  Summary:

In sum, Benezet showed that kids who received just one year of arithmetic, in sixth grade, performed at least as well on standard calculations and much better on story problems than kids who had received several years of arithmetic training.

Go figure.  It at least supports the Waldorf, and incidentally the Sudbury/unschooling model.  I'm not all the way on board with the approach, though, because I've taught a class of 5th&6th graders algebra...and I've helped tutor kids as young as 8 who did very well in college algebra classes.  Regardless, it indicates that the scope of the problem with our current system is much bigger than you thought it was, unless you already think (like me) that public schooling is one of the horseman of the apocalypse.

Moldbug question

It seems as if Moldbug's primary thrust is identical to that of Plato's Republic

The aristocracy (philosopher kings) > military rule > oligarchy > democracy > tyranny.. with occasional hints of something else.  For instance, in his last post, he seems to have suggested that we build a system to elect our own Hugo Chavez.

My concern, and the concern of an awful lot of thinkers since Plato, is that the distinction between aristocracy and tyranny seems to be a very thin distinction, and not something you can predict beforehand.  Indeed, while rule of a philosopher-king tends to produce superior results to a democracy...rule by a tyrant is MUCH worse for human happiness. 

I wonder how much of Moldbug's data is consistent with or better explained by the simple notion that rule by autarchs has a higher variance, but lower average, than rule by democracies.

I wonder how much of Moldbug's information is narrowly attached to British(and maybe Dutch) data, British law, and specifically from reading mostly about the high (relative) point of the British empire.  Certainly 5000 years of kingly rule by the Chinese, and some other large number by the Indians and Middle Easterners never bought us past the Malthusian trap of subsistence farming with a protective/parasitic aristocratic upper class. This despite basically all the technology being available (at least in the Chinese case).

I also wonder how parochial Moldbug is?  Are we talking just English/American present/history?  How good is his understanding of the 700+ year old Swiss direct democracy (oldest continuous system of law in the world, I believe) ?  How about the Singapore->China strongly limited democracy?   Danish general well-being and economic freedom in the face of every sin of democracy he can name?  How about the Irish/Icelandic anarchies that David Friedman and Roderick Long like to talk about? 

It seems that the hypothesis (Kings are better) has a massively limited dataset of roughly 1 century in the 2-3 nations leading (for unknown reasons) and in the middle of the biggest change in human civilization since the conversion to agriculture near 10,000 BC.

Am I unfair?  He's written a lot, and I certainly haven't read it all.

Growthism 101

So...one of the Aretaevian themes is that growth > almost everything else in importance.

The oversimplified demonstration:

Suppose you have value v.  V may be food, cocaine, being outside in the Sequoias, preventing climate change, or saving African children from malaria. Suppose also that there is a universal currency, say '$'.  By universal currency, I mean that $ may be exchanged for any of the above.  Oddly, this situation holds in the real world.

Suppose also that for each of the above values v1...v5, you can presently purchase 1 unit of v for $1. The question is...should you buy 1 unit now, or should you try to grow your $. 

At a rate of growth of 1%/year, your $1 will be worth $2 in roughly 2045, and unless 2 units of v in 35 years are worth more to you than 1 unit of v right now...Odds are, given time preferences, no.
At a rate of growth of 8%/year, your $1 will be worth roughly $16 in roughly 2045, and unless 1 unit now is worth more to you than 16 units 35 years hence, you should save. 

Furthermore...for most items...at least v1,v2,v4,v5...they are at least partially technologically based...and you can expect that $1 in 2045 will buy 2,4,or 100 units of your value. 

It becomes an important question then whether, for ANY value you hold for society, spending money now (or making laws) on that goal, rather than spending money on growth (or not, or not making regulations) is worth the cost.  It seems to be a real open question whether spending money on feeding the poor now is better or worse than no money on the poor, and focusing on growth, so there's more to go around in 10 years.

But it's worse...if this is a national economy...any money/regulation spent now is money/freedom not devoted towards growth, and thus buying more of v1 now not only decreases the v1 that we can buy later, but also the v2-v5 available as well.

While there are counterexamples (Is Chile's continuing fabulous growth, caused by Pinochet and his secret police, net worth it?) , there should be, in the mathematically literate, a strong presumption against doing things that decrease growth.  On the other hand...if you're advocating "saving rainforests" and it's going to cost growth, know also that you're probably not only saving less rainforest than if you'd waited til you/country were wealthier, but also you're causing less succor to the malarial victims as well.

In the climate change case, it's particularly shocking.
$1 spent now is generally $2, $4, or $8 not available later, when technology will have advanced, things will be cheaper by 50-99%, and we'll know more due to having data instead of highly fallible mathematical models.   Acting on climate change this week/year seems almost guaranteed to be colossally stupid if it decreases growth at all, (as all known measures would)...due to the magnifying effects of $ as time goes on. 

Growing up

When I was a kid, probably from 10 to 25, I found the economist to be the most respectable "newspaper" printed in the English language.  It was literate, economically literate, scientifically literate, and generally fabulous.

Yesterday, I read an article in the Economist on climate change.  I was appalled by the lack of critical thinking shown, as compared to my own attempt at the same topic.  I'm almost embarassed at the weakness of their reasoning...and their conclusion is absurdly simplistic, to the point of having to read it twice to check if it was a joke.  I guess I've grown up.

Why then, I asked, was their paper so impressive for me when younger, but not so much now?  I have a small set of hypotheses:

1. The economist is the magazine of the British educated, pragmatic, economically literate "liberal".  There has been a sea change in the educated classes' opinions in Britain between the Thatcher/Reagan era when actual free-marketeering was respectable and Tony Blair et al.'s 3rd way is substantial.  The paper, as it should, has moved with the educated opinion.
2. I've become either more well informed (true), and less mainstream (true), while the Economist has become at least more mainstream.
3. The Economist has Hansonian status management to deal with.  The serious challenges against Climate science all fly hard in the face of the prejudices of the educated classes, and so few educated folks would write them.
4. The internet has created room for the contrarians.  In the past, the semi-respectable, intellectual contrarian would work with the Economist, as something near the edge of respectability, and end up moderating his own opinions as the he also pulled the economist towards the extreme.  Now, the contrarian has her own website/blog/column in an e-zine, and doesn't write at the economist, and it has become the centrist rag of the economically literate intellectual.

Political hypothesis

Citing a friend: 
I have a theory, which is mine:

The boring part, that most good libertarian/economist types believe:
The (economic) success of a country is more or less defined by the extent to which it harnesses the ideas of its people.  Some other details

• The fundamental enabling structure for idea-usage is small-/home-/new- business 
• The fundamental motivating structure for idea-usage is positive incentives to risking it: 
• Safety of risk, including the legality of using new ideas
• Dislike of other peoples' rules.  
• The chance to make it much bigger than an equivalently skilled person following a corporate, order-taking job.  
• Untrammeled competition 
•  The fundamental disabling structure for idea-usage is elite capture of the legal apparatus.
• Gov't created monopolies, preventing competition
• Gov't rules favoring large business 
• especially, quantity of laws, causing disproportionate impact on small employers. 
• Harshness in bankruptcy law, benifitting elites 
• Crime/violence is usually really really bad for business/innovation.

Now my idea:

Systems are good insofar as the elite are able to prevent aggression, but unable to impose their preferred law on the masses.   Indeed, I'm getting more and more fond of mob-justice.  All regulatory systems under all systems of government will eventually be gamed by the plutocrats who may or may not be formally inside government.

England was great in her youth primarily due to the fact that there were no real external threats, and the Elites proved unable to regulate the productive masses, and because the law was bottom-up.
America was great in her youth for the same reasons.
Recently Singapore, and now China are operating on low-crime, but a seat-of-your-pants economic legal system where what's ok is uncertain, and the owners of elite status are in tumult, and thus not ossified. 
The American computer industry has been largely (but not completely) like that...significantly un-impeded by the lawyers of the elites. 

What is the solution then?

Grow law like corn...

This of course leads me to suggest that Patri and Romer are onto something more, even, than I had thought.

The Poor.

There's a post up at Isegoria today, which suggests we romanticize the poor.  There is also a post up at Falkenblog that says the same thing.  Of course, Steve Sailer is onboard. 

I find all these things related to the recent study on green consumers (Acting environmentally conscious makes you more likely to cheat on other moral issues).  It's actually a problem with moralism in general, as explained very well here, and indeed, several years ago, I saw a good analysis of this being a big issue with the whole "self-esteem" thing.  Self-esteem raises your basic moral view of yourself, which makes you more likely to do bad stuff. 

Is Original Sin (you have an ineradicable stain on your soul, that you can at best mitigate, but don't you ever think you're actually good) or the hardcore unsugarcoated real-comparisons reach-higher ethic I've seen my Jewish friends' families, and read about in some of Feynman's biographies, (Sure you're smart, but you're no Hillel/Einstein) the right way to go?  Do we need to instill a hunger to be better and reach further in order to not be bad?

Deirdre McCloskey's book also goes here, though it may be the most mellifluous, reference-packed, wide-ranging, hardest to follow book I've ever read (and that includes many philosophers' original work).  From her, I understood (I'd heard it before) that pride is the root sin in Christian ethics, just as much as it was for the Greeks, as I suggested recently.

Back to the poor.   Isegoria points to his new favorite educator, Miss Snuffleupagus, who has recently(-ish) been blogging about the tremendously poor ethics of the inner city youth that she teaches in London.  Indeed, her most memorable recent bit is: send them back, in which she argues that it's sad that while badly behaved foreigners can be straightened out with 2 weeks back home, we can't do that to entitled-feeling folks here, and that's very sad/bad for the kids who never get that dose of reality.

Why are the western poor poor?  Low IQ?  Poor self-control?  Bad ethics?  Bad ethics significantly worsened by government entitlement programs?  I don't know...but it's a real question.

As an aside, I still think that unschooling & sudbury (done right) are the best approaches for teaching true responsibility: It's your life, and you not only reap the benefits of your positive actions, but you also pay the penalties for not doing stuff.  I only worry that too many unschoolers mitigate the costs too much.

Diet

• The current state of affairs is non-ideal (internally) for many/most people.
• There are huge numbers of attempts to solve the problem, many of them published in the popular press, and most of which work at least partially for some group of people.
• As far as we can tell, no diets work well for most/all people, often largely due to willpower failure.
• There are almost as many competing theories, almost all of which explains something successfully, but not everything.
• The reality seems to be that there are LOTS of factors involved, and no Grand Unified Theory (GUT).
• It should be obvious to everyone even marginally paying attention that anyone proposing a complete, for-everyone theory/implementation is blowing smoke with 99.9% certainty.
• There is also a significant (but of unclear magnitude) relationship between diet and health.
• There is also some relationship between human weight/fat percentage and health, but the relationship is not monotonic, nor independent of other factors, nor best where most westerners people prefer their weight to be.

Given first that NO ONE knows what they’re doing, what do we do then?

• The most convincing theory I’ve seen says
• We are evolutionarily adapted for a hunter-gatherer lifestyle/diet (2M years, 100K generations), and not an agricultural lifestyle (6-10K years, depending, 3-500 generations)
• The most convincing papers I’ve seen suggested
• that some folks responded better to low-fat diets than to low-carb ones, but that more people responded better to low-carb diets than low-fat ones
• that modern obesity is due not primarily to eating differently, but to eating more (between meals for instance)
• The most convincing tidbits I’ve seen suggest
• The best first guess for a working diet for most people’s health is to pursue a paleo-style diet:
• Eat a variety of food
• A significant chunk of (most?) calories should come from lean-ish (fished/hunted) large-animal (no squirrel insanity) meat, which is MUCH different from modern farm-raised fatty meat.   
• Most of the rest of the calories should come from fruit/veggies.
• Don’t eat much stuff (<3 meals/week) that a caveman NEVER ate
• Domestication of Milk-animals (Sheep/cow/goat): 10Ky (Kilo-years)
• Domestication of Rice:10Ky China / 500y Europe
• Modern Wheat breed: 2Ky (contemporary with Rome)
• Corn: 400y in europe
• Sugar: 300y for commoners
• Twinkies and other synthetic food are also obviously recent
• More unusual paleo-hypotheses that I buy
• Primitive folks consumed large quantities of bacteria, both from dirt and especially from fermented or fermented/partially rotted food…it’s probably important to health(Seth Roberts/Unami hypothesis)
• Yoghurt, Pickled stuff, Sauerkraut, Soy, Worchestershire, Miso, Kefir, Natto, etc.
• Hunger seems to have been not-unusual. Skip food for 18-24 hours at least once a week.
• Psychological hypothesis that I buy
• Human willpower as a method semi-universally fails.  Don’t rely on willpower for diets. Instead, rely on habits & social support.
• Hunger is the mind-killer. If you can avoid hunger, you win.
• Seth Roberts stuff (calories sans taste) is the most interesting hypothesis here.
• Interesting additional bits:
• Nutritionally sound calorie restriction seems to be the most solid diet research around.  It slows the aging process, and eliminates basically ALL diet-related illness.  Substantial evidence, huge benefit...very hard to do.
• Intermittent fasting (alt-day, ~24h) may have almost identical benefits, without the substantially reduced weight (which may in and of itself be costly)
• Some vitamins, are important (D3, Omega-3 seem especially so).

So... my last run at diet (since new years) has basically failed, and rather than giving up entirely, I'm chasing a restart, based on family history (for genetic diet-matching) and my best guesses. I'm sitting right at the top of the mildly overweight category that was recently listed as "the most healthy", but I'm vain, and would like personally to be closer to the bottom of the mildly overweight category.

Furthermore, since social proof is most success in personal reform...I am posting, so as to see if my fabulous commentariat will provide the social support necessary:

My proposed diet goals, in order of importance/likelihood of success:

1. Alternate day fasting: Dinner to Dinner.
• Goal: Fast roughly 10 days a month, for between 20-24 hours at a time. 
• My target will therefore be Fasting MWF (13 days a month), with the expectation that eating with friends will mitigate that some.
• Easy, because all I have to do is avoid snacking after dinner, avoid eating at work
• Hard, because I will have to avoid caffeine too (I can't drink coffee w/o milk & piles of sugar)
• Maybe can leverage Seth's anti-hunger methods for after-dinner snacks (1h after dinner, most nights). 
2. Fruity vitamin/yogurt Shakes 5x/week with dinner.  This solves:
1. Unami: consume bacteria/fermented food 1x a day -- This is normally hard for me, basically because sour isn't all that good for me.  Yogurt based shakes work, though, with enough fruit/honey.
2. Vitamins: oils and pills are hard for me to eat normally...but a regular, habitual, vitamined shake isn't.
3. Vita-Mix blenders are fabulous.
4. Target every day, miss 1/3, hit 5/7
3. Attempt rough paleo-diet otherwise (avoid rice, wheat, corn, milk, sugar), with <20% of calories (2-3 meals a week) violating this.
4. Monitor for 2 months, before doing anything else.

I'm starting April 1, no joke.

The quest for sane hours

One further step:

Starting school at 10am, rather than 1-2 hours earlier at the crack of dawn, leads to substantially reduced absenteeism.

This is not the first evidence that the farmers are oppressing us night owls.

Comment on Antivirus

Commenter Psyonic left a comment on my recent religion post that was particularly insightful, and something that most smarter people eventually struggle with:

I will say, however, that such thoughts  are an area I battle with myself on constantly. It's probably true, that the masses can't/won't handle bare reality, but it's also a bit condescending and reeks a little bit of a totalitarian/paternal view of society (they can't be expected to make rational decisions, so they need something/someone to do it for them). Which, again, may be needed, but I think it's an area where you must tread lightly.

One of the reasons that many of us on the libertarian side are where we are is that we find the smart progressive answer to this particular problem more than a bit offensive.   Turns out the neo-cons...a variety of liberal by all counts are no better.   And the pragmatic religionists who pretend to religion in order to support/court the masses are about the same on the offensiveness scale.

My inclination is that shifting cognitive focus here is more than worthwhile.  Suppose that instead of human brains being configured to handle reality first, human brains are primarily configured to handle social relations, and specifically arguments and deception. In this case, it's not a question of reality-handling, it's a cognitive arms race...and in all arms races, there are winners and losers.  Suppose instead of suggesting that "the masses can't handle bare reality", we say that given bell-curvy dynamics of IQ and argument-skill (which is practice-based), some 99% (99.9%?) of the population are effectively bringing sticks to a cognitive gunfight.

Under this understanding of the world, the claim is that sufficient persuasion expertise & IQ can sell ice to Eskimos, or convince people that Communism is a good idea.  If this is possible with pure argument, then people need a cognitive defense mechanism.  Also...it is trivially true that most ideas are wrong, and therefore should not be believed.

The best defense (over short periods) seems to be some sort of anti-intellectualism, wherein all arguments are discounted.   However, this isn't sufficient over the medium-term, because there are also back-channels of communication into the brain and simple anti-intellectualism just prevents the overt arguments, not the Reifenstahls.  To prevent Reifenstahl...one needs a positive defense of socially accepted priors, which is usually traditionalism coupled with a body of myth (real history works almost as well) about what the tradition is.  The question then, is what buttresses tradition? 

My claim is be that religion is a generally stronger defense than Hayek for most people. 
And if the goal for the average bear is to believe what is true...the religion approach may be the best defense available, given the cognitive viruses out there.

Robin Hanson

As is not uncommon, Robin has a series of good posts recently.  I think, however, that the last two are if anything more indicative of Robin's quality of thought than many others.

First:
Using "great filter" arguments, and the self-indication hypothesis (which Robin has previously embraced), it now looks especially likely that Humans will indeed kill themselves off before very long.  This, incidentally, is a great example of how to apply Bayesian logic to history: All possible worlds...probability of existing in each. 

Second:
Why are human brains large?  Large brains, and especially large human brains with large craniums, are particularly costly to both the individual and mother (brains eat ~50% of daily calories, fruit fly studies show smarter flies get less food net, human cranium + upright walking don't fit well).  Furthermore, evolution is ruthless...if it's a net survival cost, it dies off statistically; if it's a net survival benefit, it spreads through a species very fast (100s of generations, not 1000s). New operating hypothesis.  The evolutionary benefit of large (powerful) brains is deception and rule-bending...which lines up very well with Robin's bias-focus. 

Incidentally, multi-part brain + this deception focus + consciousness as a rationalizer, not controller predict extensive self-deception in a self-serving fashion as well.

Aretae's 1st Law

Everyone cool gets a law named after them.  While not willing to assert coolness, here's mine anyway:

The feedback system defines the system itself.

Applications:

Super-trivial:

Health Care.
Public Education


Merely trivial: 

Agile
Lean/Toyota/Deming
OODA/Boyd/Warfare

CAS/Econ/Evolution



Less trivial:

Education solutions
Epistemology

Politics (what works?)

Agreement

How can you not love a congress that so completely demonstrates that Public Choice is correct.  The republicans hate the bill, that's clear.  The libertarians hate it...that was obvious long before the debate started.  And also, the progressives hate it.  On the other hand, the insurance companies like it, and given current trends towards HSAs, the insurance industry was facing quite a bit of competition coming up...looks like that's no longer an issue.

QoTDs -- Health Care edition

Randy Barnett (HT: Instapundit):

If the Democratic Party could survive slavery, it should be able to survive the passage of this health care bill. But then again, until the Civil War, slavery was less unpopular in the United States than is this bill.

Tyler Cowen:

[...]I recall the passage from Herodotus, where he writes of a group (I can't remember which one) which debates all major policy issues both sober and drunk, both reasonably and full of irresponsible passion.  That's not my personal model, nor do I wish to see it in the comments.  So I'll wait just a wee bit, and let others get both their celebrating and hand-wringing out of their systems...

Immigration

Disclaimer:
I'm an open-borders libertarian...The enormous costs to 3rd-worlders of living in their home countries is so large that all domestic problems pale in comparison.

To whatever extent you are a liberal, classical or otherwise, a global-villagist, or a humanitarian, you must think that the restriction of Haitians to Haiti, North Koreans to North Korea, and Africans to Africa is the biggest (top 3, say) tragedy facing the world today. 

Haitians in the US make dozens (GDP/capita = $100/month vs. minimum wage~=$1,250/mo), or sometimes hundredfold (A little above the 80th %ile of US Household income= $10,000/month) increases in their own standards of living (income), and oftentimes are able to remit enough back to Haiti to make a family of twelve substantially less poverty-stricken.

North Korea is probably worse, as is Zimbabwe, where not only is there poverty, but also persistent (van Creveldian) low level conflict/war.

The net gain to allowing any 1 person from any of those places to immigrate into the US is phenomenally positive.

Now...The counter-argument, recently taken up by my favorite econ/statistics-jedi, Tino:

• Low skill immigrants are bad for the US, even though high skill ones aren't.
• (Low Skill) Immigrants are bad for freedom.
• Low skill Haitian immigration may be too costly

In real life, this makes sense.  The more public-sector stuff you have, the higher the proportion of costs (of immigration) borne by society, rather than the immigrant...and the higher the total cost (government is inefficient).  Anyone who doesn't believe that $ grow on trees should take this seriously.
 

However...there's a further argument for immigration that I've never heard anyone use.  Similarity breeds trust, which breeds higher government activity.  A significant portion of the Scandanavian/Euro- welfare state exists because (1) countries are relatively small (the smaller in population the more welfare-ish), and (2) they are tremendously homogenous.  Notice that the issue being taken up in the Netherlands, Austria, and causing pretty substantial unhappiness throughout France and Ireland as well.  Heck...the extent to which a US neighborhood is racially/culturally homogenous does a pretty good job (I can't find the study) of predicting the levels of social trust in the neighborhood.


To decrease government activity, decrease social trust by increasing immigration. Also, it's tremendously good for the immigrants.  FWIW, the other no-majority countries (switzerland, singapore, hong kong) all seem to do pretty well for themselves, usually by means of low-government activity.

UPDATE:  I was called out for sloppiness by Andrew.  Fair.

Switzerland is famously multi-lingual, but German is 3/4 of the language spoken.  Partial fail. 

• Defense: the number of immigrants in Switzerland who are not citizens may impact this substantially.

Hong Kong -- Chinese, through and through.  Epic-fail.

Singapore -- 3/4 ethnic Chinese.  Partial fail.

• Defenses: immigrants constitute 1/3 the population, and also, Chinese is a suite of cultures...it's like saying that Italians and Swedes are pretty much the same. 

I think an even better attack on my position, by the way, is to point at the fact that we just don't know much about geographically large (> Alaska), rich(>$20K GDP/C), populous(>100M), multi-ethnic societies.  We've never seen one, and the indicators are rather unclear. 

The Core Political Question

To what extent can leader-driven political change succeed? 

Suppose that one has neither the brutality of a real world communist leader (Lenin, Stalin, Mao, Pol Pot, Kim Jong Il) nor a fascistic autocrat (Hitler, Saddam), and is missing the admiration shown to royalty many years ago.  What can one achieve politically?

There are apparently competing theories here...and I fall rather hard on one side of the the dispute. 

Position 1:  Sufficient authority will allow almost anything to get done.
Position 2:  Without personally controlled secret police and hidden torture cells, the leader is ONLY one player among many, regardless the nominal authority structure.  Further, real (as opposed to nominal) rules of conduct are worked out extra-legally, regardless the legal structure.  The only thing the legal structure can do is make it cheaper to play legally in one set of rules than extra-legally in another.

Interestingly, this applies as well to business culture as it does to political culture.

I think without experience attempting to get things done in large bureaucracies, Position 1 might seem plausible.  Having (a) worked for nearly 3 years in a large bureaucracy, (b) read Bruce Bueno de Mesquita, and (c) read Hernando de Soto, I no longer find Position 1 even nominally believable.

The name of the game is how to play the selectorate, regardless your nominal system.  Heck...that was true even of the Soviets.