Coyote Blog

Great idea, and consistent with my growing skepticism of all published research given a general bias towards positive results.

If you’re a psychologist, the news has to make you a little nervous—particularly if you’re a psychologist who published an article in 2008 in any of these three journals:Psychological Science, the Journal of Personality and Social Psychology,or the Journal of Experimental Psychology: Learning, Memory, and Cognition.

Because, if you did, someone is going to check your work. A group of researchers have already begun what they’ve dubbedthe Reproducibility Project, which aims to replicate every study from those three journals for that one year. The project is part of Open Science Framework, a group interested in scientific values, and its stated mission is to “estimate the reproducibility of a sample of studies from the scientific literature.” This is a more polite way of saying “We want to see how much of what gets published turns out to be bunk.”

I wrote about problems in research publishing here.

Most of you likely remember the state settlements with tobacco companies.  The settlements were set up to pay states a percentage of future tobacco company earnings and sales.  But just like a profligate homeowner borrowing against his paper equity in his home after housing prices increased, governments wanted to spend the money NOW, not over 20 years.  So they borrowed against future settlement payments.  Except that now, given lower smoking rates (incentives work) the settlement payments are less than they were forecast, and states must find a way to make up the difference and pay their creditors.

The tobacco settlement has created funky incentives for state governments form the very beginning.  Formerly adversaries, the settlement effectively made large tobacco companies partners with state governments, and states have had substantial incentives to promote the business of large tobacco companies and sit on their rivals

Big tobacco was supposed to come under harsh punishment for decades of deception when it acceded to a tort settlement seven years ago. Philip Morris, R.J.Reynolds, Lorillard and Brown & Williamson agreed to pay 46 states $206 billion over 25 years. This was their punishment for burying evidence of cigarettes' health risks.

But the much-maligned tobacco giants have subtly and shrewdly turned their penance into a windfall. Using that tort settlement, the big brands have hampered tiny cut-rate rivals and raised prices with near impunity. Since the case was settled, the big four have nearly doubled wholesale cigarette prices from a national average of $1.25 a pack (not counting excise taxes) in 1998 to $2.10 now. And they have a potent partner in this scheme: state governments, which have become addicted to tort-settlement payments, now running at $6 billion a year. A key feature of the Big Tobacco-and-state-government cartel: rules that levy tort-settlement costs on upstart cigarette companies, companies that were not even in existence when the tort was being committed.

I commented here:

The government has found over time that it is able to sell higher taxes to the voters on certain items if they can portray those items as representing some socially unwanted behavior. These are often called "sin" taxes. The justification for the tax in its beginning is as much about behavior control as revenue generation.  Taxes on cigarettes, alcoholic beverages and even gasoline and plastic grocery bags have all been justified in part by the logic that higher taxes will reduce consumption.

However, a funny thing happens on the way to the treasury.  Over time, government becomes dependent on the revenue from these taxes.  The government begins to suffer when the taxes have their original effect — ie reducing consumption — because then tax revenues drop.  The government ultimately finds itself in the odd position of resisting consumption drops or restructuring the tax so it no longer incentivizes reduced consumption so that it can protect its tax revenue collections.

Hey, Obama Administration!  The evil speculators are moving oil prices again.  Time to get after them.  Hello?  Anyone there?  Where did everyone go?

Pretty definitive.  

This has already made the rounds but I can't resist mocking an HBS professors whose classes I assiduously avoided when I was there.  Her house was hit by lightning.  Apparently, this was not the fault of poor lightning protection for her house, but was due to your SUV:

I am not a climate change scientist, but I have come to understand that I am a climate change victim. Our daughter took the lead investigating destructive lightning in Maine. She found that the NASA Goddard Institute estimates a 5-6% change in global lightning frequencies for every 1 degree Celsius global warming. The Earth has already warmed .8 degrees Celsius since 1802 and isexpected to warm another 1.1-6.4 degrees by the end of the century. Maine's temperatures rose 1.9 degrees Celsius in the last century and another 2.24 degree rise is projected by 2104. I learned from our insurance company that while the typical thunderstorm produces around 100 lightning strikes, there were 217 strikes around our house that night. I was shocked to discover that when it comes to increased lightning frequency and destructiveness, a NASA study concluded that eastern areas of North America like Maine are especially vulnerable. Scientists confirm a 10% increase in the incidence of extreme weather events in our region since 1949.

This is one of those paragraphs that is so bad, I put off writing about it because I could write a book about all the errors.

• The 5-6% lightning strike estimate comes from one single study that I have never seen replicated, but more importantly comes from running a computer model.  Though it may exist, I have found no empirical evidence that lightning activity has net increased with increases in temperature
• The world has warmed about 0.8C over the last century or two. Congrats.  Infinite monkeys and Shakespeare and all that.
• We could argue the forecasts, but they are irrelevant to this discussion as we are talking about current weather which cannot be influenced by future warming.
• Her claim that Maine's temperature rose 1.9C in the last Century is simply absurd.  Apparently she got the data from some authoritative place called nextgenerationearth.com, but its impossible to know since in the few days since she published this article that site has taken down the page.  So we will just have to rely on a lesser source like the NOAA for Maine temperatures.  Here story is from 2009 so I used data through 2009

Annual Averages in Maine:

Oops, not a lot of warming here, and certainly not 1.9C.  In fact, there has not even been a single year that has been 1.9C above the average for the century since the early 1900s.  And 2009 was a below average year.
Well, she said it was in summer.  That's when we get the majority of thunderstorms.  Maybe it is just summer warming?  The NOAA does not have a way to get just summer, but I can run average temperatures for July-September of each year, which matches summer within about 8 days.

Whoa!  What's this?  A 0.3-0.4C drop in the last 100 years.   And summer of 2009 (the last data point) was well below average. Wow, I guess cooling causes lightning.  We better do something about that cooling, and fast!  Or else buy this professor some lightning rods.
And you have to love evidence like this

I learned from our insurance company that while the typical thunderstorm produces around 100 lightning strikes, there were 217 strikes around our house that night

What is this, the climate version of the Lake Wobegone Effect?  If all our storms are not below average, then that is proof of climate change.  Is this really how a Harvard professor does statistical analysis?  She can just look at a sample and the mean and determine from that one sample that the mean is shifting?

Finally, she goes on to say that extreme weather in her area is up 10% from some source called the Gulf of Maine Council on Marine Environment.  Well, of course, you can't find that fact anywhere on the source she links.  And besides, even if Maine extreme weather is up, it can't be because of warming because Maine seems to be cooling.

This is just a classic example of the observer bias that is driving the whole "extreme weather" meme.  I will show you what is going on by analogy.  This is from the Wikipedia page on "Summer of the Shark":

The media's fixation with shark attacks began on July 6, when 8-year-old Mississippi boy Jessie Arbogast was bitten by a bull shark while standing in shallow water at Santa Rosa Island's Langdon Beach. ...

Immediately after the near-fatal attack on Arbogast, another attack severed the leg of a New Yorker vacationing in The Bahamas, while a third attack on a surfer occurred about a week later on July 15, six miles from the spot where Arbogast was bitten.^[6] In the following weeks, Abrogast's spectacular rescue and survival received extensive coverage in the 24-hour news cycle, which was renewed (and then redoubled) with each subsequent report of a shark incident. The media fixation continued story with a cover story in the July 30th issue of Time magazine.

In mid-August, many networks were showing footage captured by helicopters of hundreds of sharks coalescing off the southwest coast of Florida. Beach-goers were warned of the dangers of swimming,^[7] despite the fact that the swarm was likely part of an annual shark migration.^[8] The repeated broadcasts of the shark group has been criticized as blatant fear mongering, leading to the unwarranted belief of a so-called shark "epidemic".^[8]...

In terms of absolute minutes of television coverage on the three major broadcast networks—ABC, CBS, and NBC—shark attacks were 2001's third "most important" news story prior toSeptember 11, behind the western United States forest fires, and the political scandal resulting from the Chandra Levy missing persons case.^[11] However, the comparatively higher shock value of shark attacks left a lasting impression on the public. According to the International Shark Attack File, there were 76 shark attacks that occurred in 2001, lower than the 85 attacks documented in 2000; furthermore, although 5 people were killed in attacks in 2001, this was less than the 12 deaths caused by shark attacks the previous year.^[12]

A trend in news coverage <> a trend in the underlying frequency. If these were correlated, gas prices would only go up and would never come down.

Because I do not have enough to do, I have joined the blogging team at the Privatization blog.  I am excited, because Dru Stevenson is assembling a group who have very different opinions on the topic and it should lead to a good discussion.  My introductory post is here.

It is amazing the number of goofy ideas folks have generated to try to substitute for prices in matching supply and demand.  And none of them ever work.  David Zetland has a good example in the world of water, where politicians are willing to jump through just about any hoop to avoid matching water supply and demand via prices.

The Left continues to push the myth that government "austerity"  (defined as still running a massive deficit but running a slightly smaller massive deficit) is somehow pushing Europe into a depression.  Well, this myth-making worked with Hoover, who is generally thought to have worsened the Depression through austerity despite the reality that he substantially increased government spending.

It is almost impossible to spot this mythical austerity beast in action in these European countries.  Sure, they talk about austerity, and deficit reduction, and spending increases, but if such talk were reality we would have a balanced budget in this country.  If one looks at actual government spending in European nations, its impossible to find a substantial decline.  Perhaps they are talking about tax increases, which I would oppose and have been occurring, but I doubt the Left is complaining about tax increases.

Seriously, I would post the chart showing the spending declines but I can't because I keep following links and have yet to find one.  I keep seeing quotes about "commitment" to austerity, but no actual evidence of such.

Let's take Britain.  Paul Krugman specifically lashed out at "austerity" programs there are undermining the British and European economy.  So, from this source, here is actual and budgeted British government spending by year, in billions of pounds:

2007: 544.0

2008: 575.7

2009: 621.5

2010:  660.6

2011:  683.4

2012:  703.4

2013: 722.2

Seriously, I will believe the so-called austerity when someone shows it to me.  And this is not even to mention the irresponsibility of demanding more deficit spending without even acknowledging the fact that whole countries already have so much debt they are teetering on the edge of bankruptcy.

Here is the European problem -- they are pouring hundreds of billions of Euro into bailing out failed banks and governments.  They are effectively taking massive amounts of available resources out of productive hands and pouring it into failed institutions.   Had they (or we) let these institutions crash four years ago, Europe would be seeing a recovery today.  The hundreds of billions of Euros used to keep banks on life support could have instead been used to mitigate the short term effects of bigger financial crash.

I have written a number of times before that having only a few page-limited scientific journals is creating a bias towards positive results that can't be replicated

During a decade as head of global cancer research at Amgen, C. Glenn Begley identified 53 “landmark” publications — papers in top journals, from reputable labs — for his team to reproduce. Begley sought to double-check the findings before trying to build on them for drug development.

Result: 47 of the 53 could not be replicated. He described his findings in a commentary piece published on Wednesday in the journal Nature.

I observed:

This is not really wildly surprising.    Consider 20 causal relationships that don’t exist.  Now consider 20 experiments to test for this relationship.  Likely 1 in 20 will show a false positive at the 95% certainty level — that’s what 95% certainty means.  All those 1 in 20 false positives get published, and the other studies get forgotten.

Actually, XKCD did a better job of making this point.  It's a big image so I won't embed it but check it out.

Also, Kevin Drum links a related finding that journal retractions are on the rise (presumably from false positives that could not be replicated or were the results of bad process).

In 1890, there were technological and cost reasons why only a select few studies were culled into page-limited journals.  But that is not the case today.  Why do we still tie science to the outdated publication mechanism.  Online publication would allow publication of both positive and negative results.  It would also allow mechanisms for attaching critiques and defenses to the original study as well as replication results.  Sure, this partially breaks the academic pay and incentive system, but I think most folks are ready to admit that it needs to be broken.

I don't really want to ridicule Kevin Drum here for thinking out loud.  I really hate partisan Conservative and Liberal team-politics blogs, but I read a few to stay out of the echo chamber, and Drum is smarter and incrementally more objective (a relative thing) than most.

But this is really terrible, awful engineering intuition:

These two things together reminded me about an energy factoid that's always struck me as slightly odd: virtually every form of energy seems to be almost as efficient as burning oil, but not quite.

For example, on either a power/weight basis or a cost basis, batteries are maybe 2x or 3x bigger and less efficient than an internal combustion engine. Not 50x or 100x. Just barely less efficient. And you see the same thing in electricity generation. Depending on how you do the accounting, nuclear power is maybe about as efficient as an oil-fired plant, or maybe 2x or 3x less efficient. Ditto for solar. And for wind. And geothermal. And tidal power.

I'm just noodling vaguely here. Maybe there's an obvious thermodynamic explanation that I'm missing. It's just that I wouldn't be surprised if there were lots of ways of generating energy that were all over the map efficiency-wise. But why are there lots of ways of generating energy that are all surprisingly similar efficiency-wise? In the great scheme of things, a difference of 2x or 3x is practically invisible.

First, we have to translate a bit.  He mentions power to weight ratios for batteries in the second paragraph.  In fact, batteries have terrible power (actually energy storage) to weight ratios vs. fossil fuels, much worse than 2-3x for energy storage per unit of weight or volume.  That is why gasoline is still the transportation energy source of choice, because very few things short of plutonium have so much potential energy locked up in so little volume.  But I will assume he is comparing an entire electric drive system compared to a gasoline drive system (including not just energy storage but the drive itself) and in this case the power to weight ratios are indeed closer.

But here is the problem:  in engineering, a 2-3x difference in most anything -- strength, energy efficiency, whatever -- is a really big deal.  It's the difference between 15 and 45 MPG.   Perhaps this is Moore's Law corrupting our intuition.  We see electronic equipment becoming twice as powerful every 18 months, and we start to assume that 2x is not that much of a difference.

But this is why Moore's Law is so much discussed, because of its very uniqueness.  In most fields, engineers tinker for decades for incremental improvements, sometimes in the single digit percentages.

The fact that alternative energy supporters feel like their preferred technologies are just so close, meaning they are only 2x-3x less efficient than current technologies, explains a lot about why we skeptics of these technologies have a hard time getting through to them.

Proponents of higher taxes and larger government often criticize small government folks in Congress for being "obstructionist" and "not willing to compromise."

But here is the problem:  Coyote's first rule of budget politics is to never trade current tax increases or "temporary" spending increases for future spending cuts, because the future spending cuts never happen.  Ever.  Not once.  In fact, I would not agree to trading current tax increases for current spending cuts, because taxes will stay forever but spending cuts will just be over-ridden in a few months.

Here is a recent example:

Last summer, Republicans in Congress agreed to increase the federal debt limit in exchange for the Democrats’ pledge to cap future spending at agreed-upon levels. The compromise was embodied in the Budget Control Act; discretionary spending was to increase by no more than $7 billion in the current fiscal year. I wrote yesterday about the fact that the Democrats intended to violate the Budget Control Act by increasing deficit spending on the Post Office by $34 billion. The measure probably would have glided through the Senate without notice had Jeff Sessions not challenged it. Sessions insisted on a point of order, based on the fact that the spending bill violated the Budget Control Act. It required 60 votes to waive Sessions’ point of order and toss the BCA on the trash heap.

Today the Senate voted 62-37 to do exactly that. This means that the consideration that Republicans obtained in exchange for increasing the debt limit is gone. Moreover, some Republicans–I haven’t yet seen the list–voted with the Democrats today.

One principal lesson can be drawn from this experience. It happens all the time that Congressional leaders will trumpet a budget agreement that allegedly saves the taxpayers trillions of dollars–not now, of course, but in the “out years.” But the out years never come. Tax increases are rarely deferred to the out years; they take place now, when it counts. But spending cuts? Never today, always tomorrow.

Purported agreements about what federal spending will be years from now are utterly meaningless. Congressmen will make a deal, brag about the ostensible savings in the press, and then walk away from it the moment our backs are turned, as the Democrats (and a handful of Republicans) did today.

When folks say, "we just want a compromise" on budget issues, what they are really saying is "we want to roll you.  We are hoping you are stupid enough to trade for future cost reductions that will never happen.  We can get away with this because we have an ally in the press, who always treats promises of future cost reductions as entirely credible and believable and thus paint those who are skeptical of them as radical obstructionists."

One of the complexities of analyzing causes of the financial crash was that there were two simultaneous leveraging events.  Clearly, financial firms were over-leveraging securitized mortgages and their derivatives.  But at the same time, home buyers were over-leveraging their real estate assets:

In 1989, only 1 in 230 homebuyers bought a house with a down payment of 3% or less.  In 2003, the ratio was 1 in 7.  By 2007, it was 1 in 3.

These charts make the case that government policy had a lot to do with this change.

From Zero Hedge:

The hoped-for April spike in personal income tax revenues for the State of California fell once again below theoveroptimistic assumptions used to get the budget to “balance.” Instead of the $9.4 billion that the government had counted on collecting in April, it only collected $7.4 billion, according to the nonpartisan Legislative Analyst's Office. A 21% shortfall! In addition, corporate taxes were $450 million below forecast. After months of “disappointing” tax revenues, the total shortfall in income taxes now amounts to $3.5 billion for fiscal 2012 ending June 30.

The budget, supposedly balanced when it was passed last summer, had been spewing red ink from day one. Tax revenues were one problem. Expenditures were the other. The most recent re-revisions pegged the deficit at $9.2 billion. That was a few weeks ago. Now it’s going to be re-re-revised to nearly $12 billion.

Just how bankrupt does a budgeting process have to be for a budget that is supposedly in balance turn out to be $12 billion overdrawn barely 9 months later?  I have a California state tax refund on my desk -- better cash it quick or else its going to be replaced by scrip again.

The same article has this interesting tidbit about California high speed rail:

The CHSRA plan assumes that it would cost 10 cents per passenger mile (the average cost of carrying one passenger one mile at a given load factor) when international high-speed rail systems averaged 43 cents per mile, according to a report that just surfaced. The low-cost leader was Italy with 34 cents per mile; at the upper end were Germany and Japan with 50 cents per mile; Amtrak’s Acela Express, though not truly high speed, was in the middle with 44 cents per mile. And in California, it’s going to be 10 cents per mile?

The CHSRA correctly assumes that train tickets compete with air fares and the cost of driving, which, despite our incessant complaints, are lower in California than overseas. Thus, the US market requires cheaper tickets. And to make the project appear profitable, and thus more digestible for the taxpayer, the CHSRA lowered its projected operating costs to less than a quarter of the international average.

But if actual operating costs are 43 cents per mile and not 10 cents per mile, annual subsidies of $2 billion to $3 billion would be required just to keep the trains running, according to the report. Yet, AB3034, the California High-Speed Train Bond Act, makes these subsidies illegal. A conundrum that the Legislature, the Administration, and the CHSRA have so far successfully ignored.

Unbelievable.  Obama has declared May 1, the traditional day used by the Soviet Union to celebrate the power of the state through military parades and such, to be "loyalty day."  Not "let's respect what America stands for" day or "celebrate liberty day" but "promise unthinking loyalty to our government masters" day.

This is particularly ironic as the other major competing proposal for Mayday is as a remembrance for the people killed by communism.  So, your choice today.  Salute the victims of out-of-control state power, or pledge your loyalty to the state.

There are certain contradictions involved in attempting minimum-rules radical self-expression on government land via a government permit.  Government employees incentives are all geared towards following procedures.  There are no rewards for results or innovation, just punishments for violations of bureaucratic minutia.  It is not surprising that the government and Burning Man cultures have come into conflict.

The movie Blade Runner is a pretty substantial departure from the Phillip Dick book "Do Robots Dream of Electric Sheep" on which it was based.  Even so, and perhaps uniquely in literary history, Dick seems to have absolutely loved the movie.  It kept the right elements of the book - ie, what makes us human -- and shed the silly, trippy stuff.

I don't remember it being a huge box office success.  Probably too dark, even with the last minute change of ending (the happy notion that Rachael had no programmed termination date was added to give audiences a more upbeat ending.)  But the movie certainly had a huge effect on the look and feel of sci-fi.  After the Matrix and the Terminator, we are used to future dystopias, but in the 1970's most popular sci-fi had cities that were as bright and shiny as a new penny.   I remember seeing it the first time, and Blade Runner was arresting, a whole new category of sci-fi noir.  I still love the movie, and it wears pretty well, but nowadays fan argue endlessly of the merits of the original release vs. the directors cut.  The latter purges the Harrison Ford narration and happy ending that were tacked on to make the movie more audience friendly.  I personally like the narration-- it feels consistent with the noir genre -- though the faux happy ending is lame.

I have been reading of late some histories of Germany in the 1930's, with a particular emphasis on racial laws and policy.   Over time the expanding bans on Jewish participation in the economy and society as well as preferences given to non-Jews for government jobs led to some practical problems, including:

• What percentage of Jewish blood made one Jewish?  The Nazis messed around with this problem a long time, in part because of Hitler's absolute reluctance to get involved in such details.  Was it one grandparent?  Three grandparents?
• How does one test for such things?  In the thirties, there was an boom in geneology research in Germany, as everyone raced around trying to figure out what evidence was sufficient to establish someone's race

It would be nice to think we put this kind of thing to bed, but here we are in the 21st century running around trying to answer the exact same questions

• Does 1/32 Cherokee blood make Elizabeth Warren the member of a protected group?  Does 1/8 black blood mean George Zimmerman is not a racist?
• We are building infrastructures to make rulings on race.  I found this story amazing

This story reminded me of the 1980s case of the twin red-haired Boston firefighters who claimed to be black, based on a photo of a great-grandmother and alleged oral history. While I remembered that they had gotten fired for their alleged fraud, I didn’t remember this detail:

Under current rules, said [general counsel to the state personnel office] Ms. Dale, candidates who say they are members of minority groups are judged by appearance, documented personal history and identification with a minority community. Disputes over claims of minority status are resolved by the Department of Personnel Administration.

 And indeed, there eventually was a two-day administrative hearing, in which the hearing officer determined that the twins failed all three criteria, and thus were not black. A judge upheld the ruling, finding that the twins had claimed minority status in bad faith.I have to admit being under the impression until now that as a legal matter, minority status was an in issue of self-reporting. But at least in the Massachusetts Civil Service system, one can get fired for “racial fraud.”

• Every year, in the name of some sort of racial harmony, I have to sit down and report to the government on the race of each of my employees.  For 364 days a year I can ignore the race of my employees, but one day a year the government makes me wallow in it.  Here are part of the instructions:

Self-identification is the preferred method of identifying the race and ethnic information necessary for the EEO-1 report. Employers are required to attempt to allow employees to use self-identification to complete the EEO-1 report. If an employee declines to self-identify, employment records or observer identification may be used.

Where records are maintained, it is recommended that they be kept separately from the employees basic personnel file or other records available to those responsible for personnel decisions.

Race and ethnic designations as used by the Equal Employment Opportunity Commission do not denote scientific definitions of anthropological origins.

I am told we are trying to create a society free of racism, but the results sure look a lot like racism to me.

I've already told the story of being in Manhattan on 9/11.  Through the day, vehicles could leave the city, but they could not come back (even taxis).  That evening, most people who could leave Manhattan had done so.  We were stuck until the next day.   We ended up finding a restaurant for dinner in Times Square that was open.

Times Square was just totally bizarre.  There were no cars at all.  Perhaps one car would pass every five minutes.  A couple of guys were roller skating around the streets, I supposed just because they could.

I was reminded of this experience by this photograph by Lucie and Simon, who take pictures of cities and digitally remove the cars and people.