Friday, October 24, 2014

Epistemologically without value and yet irrisistibly alluring

Sometimes you come across something that is indisputably meaningless but also indisputably fun. For example, Your Name on a Liberal to Conservative Spectrum from Crowdpac.

From the article, Here Are The Most Conservative And Liberal Names In America by Katherine Miller.
Crowdpac, a nonpartisan group, focuses on how money and policy work in politics. The group scores all donors who have made two or more campaign contributions since 1980. Using that algorithm — and a cut-off of names with at least 1,000 donations made to avoid unusual names and outliers — Crowdpac built a tool to show how conservative or liberal first names are.

[snip]

You’ll notice there is a very stark gender divide! The liberal names generally sound like a group of women in their late 20s; the conservative names sound like the members of a large bluegrass band from the 1930s. This makes sense: Women are more likely to be Democrats, for one thing, and the liberal names also generally represent younger Americans, while the conservative names generally skew older.
A somewhat murky methodology or algorithm generating a likely meaningless list that has no apparent use. But you, or at least I, can't help but run the names of your nearest and dearest through the site to see how "liberal" or "conservative" they are.

Political statistics

Very interesting. It has long been a staple belief in some conservative circles that Democrats manipulate vote counting at the margin. This attitude is reflected in the adage "You can't just win. You have to beat them by the margin of fraud." It has also been assumed that Democrats' objection to voter ID and other efforts to shore up the integrity of the voting system were motivated by Democrats' desire to maintain some flexibility for winning narrow contests. I have always assumed that while this is not an unreasonable argument, it was likely simply a function of observer bias. That Republicans see the close races they lose but don't focus on the close races the Democrats lose.

The accusation took on a little more substance in the past ten years. I think it was a governor's race out in Oregon or Washington where the Republican won by a narrow margin. After three recounts, he had lost. Something similar happened with Al Franken up in Minnesota. Still, it seemed reasonable that it might simply be sour grapes.

Or not, as it now emerges.

From Do Democrats Always Win Close Statewide Elections? by Dan McLaughlin.
To get a sense of the answer, I took a look at all the statewide Senate and governor’s races from 1998 through 2013 (thanks to Sean Trende of RealClearPolitics for a big assist with the data) as well as all the statewide results in the presidential elections during that period. Let’s begin with the very closest races, those decided by less than one percentage point. There have been 27 such races since 1998, and Democrats have won 20 out of 27
You would expect the outcomes to be 50:50 but they are 74:26 in favor of the Democrats. So maybe there is an issue.

There are all sorts of rationals you can come up with to explain such a disparate impact but they all seem fairly improbable or anemic. And maybe it is still just a numerical quirk. It is a small population of data points. What would happen if you added in contested House seats, would you get the same results?

Almost every large firm or corporation has a person inside the IRS

From Whistleblowers: IRS Officials Behind ‘Fraudulent’ Multi-Billion Dollar Corporate Tax Giveaways by Paul Caron. Who knows whether this is an issue with merit or might merely be a storm in a teacup. It is certainly consonant with the behavior of the IRS in other recent scandals.

Regardless of the merits of the accusation, it is interesting as an example of good intentions gone astray in complex systems, (as virtually all human systems are). I am currently reading Doing Bad by Doing Good by Christopher Coyne which is a pretty rich catalogue of examples and a reasonable explication of why good intentions so often lead to bad outcomes.

The IRS is accused of avoiding bringing big cases against large corporations while at the same time mercilessly pursuing small businesses and private individuals. The reason is ascribed to a well intentioned change in the late 1990's.
[T]he private sector lawyer and ex-IRS attorney explained that since 1998, IRS restructuring has focused on bringing in “outside people.” This led to the employment of an extra layer of executives who were previously “partners from big accounting firms.” Citing active IRS criminal agents, the ex-IRS attorney said: “Almost every large firm or corporation has a person inside the IRS. It’s a revolving door, with the top two or three management layers all from big accounting and law firms, and this is why they won’t work big billion-dollar cases criminally. Private bar attorneys are, in effect, controlling the IRS. It’s a type of corruption – that’s the word used by one IRS agent I’m in touch with whose case was shut down by higher ups without cause.”
I vaguely recollect this issue. I think the intended change was meant to bring more cutting edge talent to theIRS as well as a stronger awareness of real-world accounting and tax issues. The hoped for outcome was better cases against large corporations, more wins, higher yields, and more revenue to the Government. Not a bad argument or intention but in hindsight you can see why the good intention might have led to an unintended and bad outcome.

Thursday, October 23, 2014

Only 20% of American jobs are even mildly strenuous

There has been a marked increase in national obesity in the past forty years, not only in the US but worldwide. You see documentaries from the 1940s-1960s and ordinary people on the street are slim. What is the cause for the rise in obesity? Changed governmental dietary advice, access to fresh food, preferences for carbs, loss of the tradition of family meals, over reliance on fast food? There are lot's of popular theories. Everyone agrees we are more sedentary than in the past but that observation is offered without qualification.

From Working hours: Get a life—or get fat from The Economist.
So why do long hours result in weight gain? Only 20% of American jobs are even mildly strenuous, compared to 50% in 1960. In 1960 a tenth of the American workforce was involved in agriculture, but today it's more like 1%. More time at the desk means less movement. Busy people may have less time to prepare good meals, instead choosing a take-away. (Management consultants, in my experience at least, tend to be rather knowledgeable about fancy restaurants near them that also deliver). They exercise less. And workaholics sleep less: inadequate shut-eye is associated with weight gain.
50% to 20% of people with jobs that are mildly strenuous - that's a pretty big drop.

Parenting matters much more than parental income.

A marvelously brief summary of the state of play as to child development and life outcomes from Skills and Scaffolding by Nobel laureate James Heckman.

His list of what we know (and my translations in brackets):
1. Character skills matter at least as much as cognitive skills.
2. Important skills are not innate "traits" solely acquired by genetic inheritance.
3. For skill development, timing matters. (There are periods in their development when children are especially amenable to trait acquisition)
4. The early years are the most effective period for investments in both cognitive and non-cognitive skills.
5. Successful adolescent interventions largely operate through promoting character skills.
6. Skills beget skills. (The Goldilocks effect/the Matthew Effect. There is a cascading effect on early skill acquisition)
7. The development of skills takes place within a vital "scaffolding." (You have to modulate delivery to what a child is able to absorb)
8. Credit constraints are not very important.
The last item is probably worth quoting in full as it is a hoary standby for all SJWs. The belief that income determines outcomes (as opposed to parenting practices) is one of those seductively logical assumptions that is so appealing that people rarely look at the actual evidence. In addition, it plays to the near universal inability to distinguish between correlation and causation.
8. Credit constraints are not very important. There is a strong empirical relationship between educational attainment and parental income. However, parental income is a proxy for many attributes of the parental environment. The causal evidence of an importance role for credit constraints is weak. Parenting matters much more than parental income.

Wednesday, October 22, 2014

Suddenly, an earsplitting roar of thunder rends the air.

A learning vignette.

Back in the late 1960s, when I was about seven or eight, my father had business in Norway and took us along with him. I think it must have been early spring as there was still deep drifts of snow when we were up in the mountains. We visited Stavanger, Oslo and other places, saw the Viking ships and the Kon-Tiki, were exposed to a cuisine more dominated by fish and cheese than we were accustomed to.

I recall with some clarity, being shown around Oslo harbor. I don't know who the Norwegian friend of my father's was, but he was recounting Norway's experience during World War II as a proud, independent nation far outgunned by the invading Germans. He drew our attention to a stone fort on a small island in the harbor and related the following story as best I recollect.
The Germans invaded several places along Norway's long coast, not just here in Oslo, but this was the main invasion point. Their fleet was led by a battleship and dozens of other ships. We didn't have a chance. Our air force was out of commission and there were hardly any Norwegian troops. We did have that old fort in the harbor from the turn of century (circa 1900). The guns were ancient and the fort was more for show than a real military installation. The artillery had been plugged with cement long ago.

The night before the German fleet arrived, a small group of Norwegian patriots rowed out to the fort. They unplugged the cannon and searched around and found some old shells. They knew they would likely only be able to get off one shot, that their aim was likely to be poor, and the shells likely to be duds. But they were patriots and even if there was not much they could do, they wanted to do the little they could.

The morning of the invasion, as the German fleet sailed in, the Norwegian patriots sighted the cannon on the German battleship. Fully expecting to die for their efforts, they fired the cannon. Imagine their astonishment! Not only did the shell hit the battleship, not only was the shell not a dud, but the shell hit the German magazine. The explosion was massive, the battleship capsized and sank within a few minutes, taking more than a thousand German occupation troops and much of the leadership to the bottom of the harbor.
That's the story I remember from nearly fifty years ago, on a cool but sunny spring day in Oslo Fjord. I was struck by it then and it has remained with me since. The tragedy of Norway, the patriotism, the miracle. It was a variant of the King Bruce and the Spider story - keep trying and no matter how long the odds, you may still succeed.

I have read much World War II history over the years, lived in Sweden for a number of years and know much of their history. I have seen reference to the sinking of the German flagship at the beginning of the invasion of Norway. For all that, I never got around, till recently, to reading a detailed account of WWII in Scandinavia. I am currently reading Battles for Scandinavia by John R. Elting.

So how does that long ago story stack up with the facts? Here is Elting's account.
Admiral Kummetz, leading the Naval units bound for Oslo, was still eyeing the Oscarsborg Fortress in Oslo Fjord at 4:21 on the morning of April 9 when suddenly a lone searchlight blazed out from the mainland on the opposite shore, bathing the flagship in ghostly white. At point-blank range the huge guns of the fortress opened up; 700-pound shells crashed into the Blücher's port side and then into the Lützow just behind her. Simultaneously came another blast from the mainland battery at Drobak, damaging both ships on their starboard sides. The Blücher's steering gear and aircraft hangar were wrecked by the first shots, and she blundered forward erratically while flames fed by aviation fuel raged aft. They burned away the fog, revealing a torpedo battery built into Oscarsborg Fortress. It had been there for fifty years without showing up on the German charts.

In discounting Oscarsborg as a threat, Admiral Kummetz had not reckoned with the character of Colonel Birger Eriksen, its commander. Eriksen gave his gunners one simple order: Fire. Two of the ancient torpedoes ran straight to their mark and exploded on contact, reducing the brand-new Blücher to a floating inferno.

Captain Kurt Zoepffel, who was conning the Blücher at that moment, recorded: "Suddenly, an earsplitting roar of thunder rends the air. The glare of guns pierces the darkness. I can see three flashes simultaneously. We are under fire from two sides; the guns seem only 500 yards away. Soon bright flames are leaping from the ship." In a moment the Blücher's store of bombs and ammunition began to explode, the engines stopped and the ship heeled over. When at last the order was given to abandon, wounded and dead men were already rolling into the water, and many of the landing troops were trapped below deck. Admiral Kummetz and some 1,300 survivors were rescued by the Norwegians, taken ashore and imprisoned. But more than 1,000 Germans died in the explosions and the flaming oil. For days after the incident, their blackened bodies could be seen floating in the fjord.

Just behind the Blücher, Captain August Thiele of the cruiser Lützow, having no idea that the museum-like fortress and its torpedoes could have been the source of the Blücher's explosive demise, concluded that the flagship had hit a minefield. For the safety of his and the remaining ships, Thiele ordered all engines reversed. Under his command, the Lützow and the remaining consorts retired 12 miles south to an alternate landing site at Sonsbukten, on the eastern shore of the fjord. The bid to take Oslo from the sea was lost.
My unknown father's friend's story, or my recollection of it, holds up pretty well all these years later. The Norwegian defense involved more men than I was aware and the fort, while old, had more armament than just a few plugged cannon. The Norwegian attack was more than a single lucky shot. The Blücher was a heavy cruiser, not a battleship.

But other than those details, the substance of the story remains the same. Patriots against long odds. Armaments that are ancient but effective. A forlorn hope attack. Unexpected success. Heavy German losses.

I don't think the differences between my father's friend's account and the historian's account are all that material. And the differences may not be because of my father's friend's recounting. I may not have understood it completely at the time, I may have consolidated the account into a more manageable form for a seven-eight year old's memory, details may have slipped my mind over the decades.

There's no great insight here, just my surprise that a long ago story should hold up so well. It seems to me an endorsement of that Garrison Keillor quote "Nothing you do for children is ever wasted." Particularly, when it comes to storytelling. So whoever you were, Norwegian friend of my father's, Thank you.

If you have a theory, you must try to explain what’s good and what’s bad about it equally

Richard P. Feynman in What Do You Care What Other People Think? Further Adventures of a Curious Character
The only way to have real success in science, the field I’m familiar with, is to describe the evidence very carefully without regard to the way you feel it should be. If you have a theory, you must try to explain what’s good and what’s bad about it equally. In science, you learn a kind of standard integrity and honesty.

Tuesday, October 21, 2014

They don’t know enough to have an intelligent opinion, as demonstrated by the opinions they do have.

From Instapundit, AS THE ACADEMY CONTINUES TO EAT ITSELF. Reynolds is commenting on this article, Students: Transgender Woman Can’t Be Diversity Officer Because She’s a White Man Now by Katherine Timpf.

Reynolds comments:
Tip: When students go on about social justice, the proper response is to tell them you don’t care what they think, because they don’t know enough to have an intelligent opinion yet. If universities were run on this principle, the 3% of students responsible for 98% of the idiocy would no longer have their destructive impact. Also, it’s true: They don’t know enough to have an intelligent opinion, as demonstrated by the opinions they do have.
In this instance, I think the critical Wellesley student's position is actually quite logical. Distasteful perhaps, but not illogical.
“I thought he’d do a perfectly fine job, but it just felt inappropriate to have a white man there,” the student behind the so-called “Campaign to Abstain” said.

“It’s not just about that position either,” the student added. “Having men in elected leadership positions undermines the idea of this being a place where women are the leaders.”
It has long been my view that Social Justice Warriors were simply assertive barbarians. They do not actually want tolerance, they want to force everyone to share their particular brand of intolerance.

More critically, tolerance requires a recognition of the fact that different people have different goals, prioritize those goals differently, and pursue those goals by various means. Social Justice Warriors want everyone to have the same goals, with the same priorities and wish to pursue them in the same fashion. Most crucially, the SJW wants to be the person setting the goals, priorities and actions.

Monday, October 20, 2014

That's odd

With the latest report that there is actually an Ebola Czar (Assistant Secretary for Preparedness and Response) who is just not present, it struck me that there seem to be a lot of female sacrificial lambs of late. As best I can tell, women make up perhaps 30% of the senior cabinet members and other senior positions in the administration. But they seem to make up a striking percentage of resignations owing to incompetence, malfeasance or other circumstances.
Lois Lerner thrown under the bus and pleading the fifth regarding illegal IRS targeting of citizen based on their political beliefs is one who still remains in the spotlight.

Then there is Kathleen Sebelius, former Secretary of Health and Human Services who "resigned" after the Obamacare debacle, including but not limited to the failed website rollout.

Then there is Nancy-Ann De­Parle, White House deputy chief of staff for policy, who led the initial work on the Obamacare website disaster. She was later replaced by Jeanne Lambrew who in turn was replaced.

Then there is the recent resignation of Secret Service Director, Julia Pierson, taking the fall for the breach of White House security which allowed an unbalanced intruder to race around in the residence until eventually tackled by an off duty Secret Service agent.

Hilary Clinton has resigned as Secretary of State following multiple failures (Arab Spring, Green Revolution, Libyan war, Iraq, rise of ISIS, loss of Yemen, China territorial expansion efforts, Russian invasion of Ukraine, etc.) including the loss of four Americans to terrorist attack in Benghazi.
Those names were just top of mind. Just statistical chance? What about other bungles. I went googling.

I had forgotten about the resignation of GSA Administrator Martha Johnson for a strikingly wasteful Las Vegas extravaganza "training" session held for some of her employees.

Janet Napolitano, Secretary of Homeland Security, resigned, not because of any particular scandal, but certainly under a shadow.

Shirley Sherrod was thrown under the bus for some ill-considered morality tales to an NAACP conference.

There are some top males who have been forced out owing to dubious practices including Eric Shinseki who resigned as secretary of the Department of Veterans Affairs owing to the waiting list scandal; Keith Alexander, Director of the NSA for domestic spying, David Petraeus, Director of the CIA for an affair, General Stanley A. McChrystal for overly frank comments about the Commander in Chief.

It would take more time and effort than I am willing to invest to come up with both an exhaustive list of resignations under scandal conditions and to properly define categories for comparisons. I know there has been some sort of simmering scandal around EPA Gina McCarthy and secret email accounts or something like that.

Still, for only a 30% representation in the Administration leadership, there seems to be an overrepresentation in scandals. Are they just easy political sacrifices to make? More principled in accepting responsibility? Promoted beyond the Peter Principle and therefore more subject to missteps?

Someone, somewhere, sometime might dig in to it a bit, but it seems striking to me.

The claim that the current German dislike of inflation dates back to unique memories of Weimar hyperinflation is dubious

From Germany fact of the day by Tyler Cowen. Cowen points out that Germany has for long periods of time experienced inflation rates above 2% and in some years 6 and 7%.
The claim that the current German dislike of inflation dates back to unique memories of Weimar hyperinflation is dubious. Rightly or wrongly, today’s Germans associate high rates of inflation with wealth transfers away from Germany and toward other nations. More broadly, Germany is a more flexible country than outsiders often think, not always to the better of course.
There is an interesting discussion in the comments section about the historical events and current behaviors and decisions.
Ray Lopez October 18, 2014 at 3:46 am
TC’s claim about history is interesting. How much institutional memory does a country have? Perhaps even if present people don’t remember stuff that happened in the past, can a society function the way it does, due to history that nobody can remember? Sounds metaphysical, but here’s a real-life example. Some sociologists from France or Belgium went to the Congo a while ago, interviewed people, and found nearly none of them remember King Leopold of Belgium and the colonization of the Congo for rubber, at the end of the 19th century, as outlined in the best-seller “King Leopold’s Ghost (1998)” by Adam Hochschild . So a major traumatizing experience in a nation’s history is almost completely forgotten by the people there (most of them ignorant to be sure). Thus how big an impact can this searing event have had on modern inhabitants? Very little some might say, but, like slavery in the USA, perhaps not.

[snip]

Mark Thorson October 18, 2014 at 1:33 pm
I remember hearing one proposed explanation for Kondratieff waves was memory. People who lived through high inflation or a severe recession would be especially guarded about repeating the experience, as would their children. But their grandchildren will have forgotten all about it and repeat the mistakes of their grandparents, leading to a cycle about two generations long.

Wikipedia doesn’t mention this hypothesis in their description. They do say the existence of Kondratieff waves is not accepted by mainstream economists.
As someone points out, it is likely that it is not the memories per se but rather the way those experiences get incorporated into the institutions of the nation and culture.

I think that discussion is quite an interesting one. But I think Cowen's original point is even more interesting.

All my life and in all my studies in economics over the years, there has always been a latent and unquestioned assumption that Germans were 1) highly averse to inflation, 2) Germans prioritized inflation fighting above many other goals in comparison to other countries, and 3) that the source of aversion was the experience of inflation in the Weimar Republic.

Cowen's original point is excellent. This is all great theory and perfectly logical. The only problem is that it doesn't accord with the empirical evidence. If Germans are highly averse to inflation, then you would expect their inflation rates to be very low.

I think we more often than we want to acknowledge, accept logical explanations of history without actually checking the key data that would affirm or refute.

I can think of at least one other example. There is a popular effort in some circles to assert that slavery is the progenitor of negative measured outcomes of African Americans today. It is certainly a logical argument and is attractively simple. The only problem with it, as Thomas Sowell has pointed out, is that it does not appear to accord with the evidence. Sowell has argued, and mustered the data to support his argument, that the trends on many critical socioeconomic variables, African-Americans were on an upward trend through the 1960s: family formation, education attainment, labor force participation rates, income, etc. On a few measures they had already achieved parity with the majority population or were exceeding. Sowell argues that these upward trends are inconsistent with a causal linkage between pre-1860s slavery and post-1960s socioeconomic declines.

Both the Sowell argument and the observation about German inflation rates are a call to always double check assumed facts. Just because the story is logical and reasonable doesn't mean that it is true.