The 10% Rule

I have written before about how approximately 10% of respondents will answer poll questions inconsistently. According to this study by Public Policy Polling (pdf), 11% of liberals believe that ACORN stole the election for Obama. Mindboggling. Of course, 48% of conservatives believe that.

Horrible statistics

Via Andrew Sullivan, Michael Fitzgerald has an article in the Boston Globe with the sub-heading "the curious economic effects of religion". This sub-heading alone is very misleading. While there are many interesting correlations, there is nothing that proves that religion causes these "curious economic effects". It is possible (I would even say probable) that there is another factor that is influencing both the economic effects and religiosity.

The two collected data from 59 countries where a majority of the population followed one of the four major religions, Christianity, Islam, Hinduism, or Buddhism. They ran this data - which covered slices of years from 1981 to 2000, measuring things like levels of belief in God, afterlife beliefs, and worship attendance - through statistical models. Their results show a strong correlation between economic growth and certain shifts in beliefs, though only in developing countries. Most strikingly, if belief in hell jumps up sharply while actual church attendance stays flat, it correlates with economic growth. Belief in heaven also has a similar effect, though less pronounced. Mere belief in God has no effect one way or the other. Meanwhile, if church attendance actually rises, it slows growth in developing economies.

I assume that it is the author of the article (and not the researchers) who has confused correlation with causation. Their results show a "strong correlation between economic growth and certain shifts in beliefs". However, the article then says that "if church attendance actually rises, it slows growth in developing economies". It should say that "if church attendance actually rises, it correlates with a slowing of growth in developing economies".

However, the researchers do not escape blame entirely. In the next paragraph:

McCleary says this makes sense from a strictly economic standpoint - as economies develop and people can earn more money, their time becomes more valuable. For economic growth, she says, “What you want is to have people have their children grow up in a faith, but then they should become productive members of society. They shouldn’t be spending all their time in religious services.”

For Christians (the only one of the four religions studied with which I am very familiar), this seems wildly implausible. An hour every Sunday spent in church causes the GDP growth for the entire country to slow?

Anti-urban bias

David Brooks has a new column about John Thune, a cookie-cutter Republican. According to Brooks, Thune "is down-the-line conservative on social, economic and foreign policy matters". What makes him so remarkable?

The first thing everybody knows about him is that he is tall (6 feet 4 inches), tanned (in a prairie, sun-chapped sort of way) and handsome (John McCain jokes that if he had Thune’s face he’d be president right now). If you wanted a Republican with the same general body type and athletic grace as Barack Obama, you’d pick Thune.

The gist of the column is that Thune is handsome, "unfailingly genial, modest and nice", so he would be a good face for the conservative policies that the country rejected in 2008.

What I found interesting about the article was how it revealed a little bit about Brooks.

His populism is not angry. He doesn’t rail against the malefactors of wealth. But it’s there, a celebration of the small and local over the big and urban.

The opposite of small is big. And the opposite of local is urban. Because nobody lives in urban areas. According to Wikipedia, 70% of the American population lives in urban areas. 30% of these people (21% of the total population) live in city centers, and the other 70% (49% of the total population) live in the suburbs.

Bias again

Matthew Continetti has an interesting post up in the Weekly Standard saying that Palin can be the new standard-bearer for American populism, following in the tradition of Jackson, Bryan, and Reagan. The piece is poorly argued from start to finish, but his arguments on the financial system are scarily weak.

For the banks, a complicated and technocratic regulatory scheme isn't necessary. A few simple rules that separate the solvent banks from the insolvent would suffice.

And what happens when a bank becomes insolvent, a la Lehman? He doesn't really say, although earlier in the piece he says:

Held to the standards of the marketplace, companies like GM, Chrysler, AIG, GMAC, and Citi probably would disappear. They'd be bought and sold, carved up into little pieces, and the overpaid CEOs who made bad bets would lose their jobs.

So I imagine that he thinks that if Citi were to become insolvent (or admit that it is insolvent, depending on your point of view), it would simply be broken into little pieces and sold off. There would be no major systemic collapse, because... well, Continetti isn't very clear why not. I suppose that if your biases tell you that government intervention makes things worse, then all problems can be fixed by less government intervention. Even if almost anyone who has studied the financial system disagrees with you.

Up is the new Down

Have you heard about the two big wins for conservatives in NY-23? Let RedState's Dear Leader Erick Erickson explain:

The race has now been called for Democrat Bill Owens.

This is a huge win for conservatives.

“Whaaaa. . . ?” you say.

There are two big victories at work in New York’s 23rd Congressional District.

First, the GOP now must recognize it will either lose without conservatives or will win with conservatives.

[snip]

Secondly, and just as importantly, there has all of a sudden been a huge movement among some activists to go the third party route. We see in NY-23 that this is not possible as third parties are not viable.

Third parties lack funding and ability for a host of reasons. Conservatives are going to have to work from within the GOP. The GOP had better pay attention.

The first point is debatable. The second is right, and is completely not a "huge win for conservatives". It is a victory for the GOP. A "win for conservatives" would have been if Hoffman had won and shown that a conservative third party is viable.

Bias

Everyone knows where Megan McArdle stands politically, but it's still funny whenever someone puts their bias on full display.

Either Hoffman will lose, in which case the strategy of policing the party will lose some of its appeal, or he will win, in which case Blue Dog democrats and Republicans in squishy states will probably tack right--a critical win during the health care debate.

Now, try re-writing it slightly...

Either Hoffman will win, in which case the strategy of policing the party will gain some more appeal, or he will win, in which case Blue Dog democrats and Republicans in squishy states will probably tack left--a critical win during the health care debate.

These sentences both rely on the same basic idea: a Hoffman win pushes the country to the right, and a Hoffman loss pushes the country to the left. And yet the two sentences sound so different.

1 in 10 Will Get Everything Wrong (redux)

Another piece of evidence pointing towards the validity of my theory that about 10% of people will give wrong or inconsistent answers to poll questions: a Public Policy Polling study on the birther phenomenon that found that 6% of Americans don't believe that Hawaii is part of the United States, and a further 4% aren't sure.

After thinking more about the issue, however, I've started to wonder how much of this is due to people lying. If a pollster called me up, depending on my mood I might be inclined to answer that Obama was born in France and that Hawaii is actually part of Indonesia. Is people lying a problem in polling?

Deciphering Steyn

Mark Steyn is appalled that a pedophile in Britain is getting Viagra via the taxpayer-funded National Health Service. He quips:

I think this is what those heartless American insurers would call a pre-existing condition.

I cannot figure out what he is trying to say. In America, insurers routinely pay for Viagra. Is he suggesting that insurers would deny his claim because he's a pedophile? Are insurers allowed to deny coverage because of a criminal conviction? Or does he have some other meaning that I am missing?

Iterative Design

When Kerry was running for president, there was a fear that he would win and Mitt Romney (then the Republican governor of Massachusetts) would appoint a Republican to fill his seat. So the legislature passed a law that if there is a vacant Senate seat, it should be filled by a special election to be held between 135 and 160 days after the vacancy. Now there's a worry that Kennedy will die, leaving a seat vacant during the upcoming health care votes. Naturally, the Democrats in Massachusetts want to change the law again to allow the governor to appoint a "caretaker" Senator. Jason Zengerle comments:

I think there's a good lesson here about legislative bodies being careful not to muck around with these sorts of rules for short-term political gain. Now, obviously, that lesson would seem to suggest that the legislature not mess with the rule as it now exists--that the seat remain vacant until a special election is held to fill it within five months--except that rule seems really stupid. Kennedy's proposal--have the governor appoint an interim Senator who pledges not to run in the special election--seems exceedingly sensible.

Matthew Yglesias responds:

But when you have a state whose state legislature is firmly and forever in the hands of one political party, the smart thing is for the legislature to be constantly changing rules based on short-term considerations.

Yglesias is completely wrong. If the situation arose where there was a Democratic legislature and governor, and a Republican Senator dropped dead, Yglesias seems to be suggesting that it would be appropriate for the legislature to change the rules so that the governor could appoint a Democrat to the Senate. Not only is this not small-d democratic, but it also breeds extreme cynicism.

The fact that Yglesias is completely wrong does not make Zengerle completely right. He is half right. The original tweak was intended to prevent a governor from reversing the results of an election via his special power to fill vacancies. While this change may have been "for short-term political gain", it also had the advantage of being just and democatic. A governor should not be able to overrule the choice of the people.

Unfortunately, upon testing in real world conditions, it turns out that the changed law has a major flaw: it leaves Massachusetts under-represented in the Senate for at least 135 days after a vacancy. As Zengerle says, the proposed solution "seems exceedingly sensible". And it is. They should implement it.

In product design, I am a big fan of iterative design. It turns out that it is also applicable to politics.

You Cannot Tax a Corporation

Over at the Daily Dish, a reader is upset about people collecting unemployment insurance:

I have to correct the reader who wrote that since he paid unemployment tax throughout the years, he should not feel guilty about claiming unemployment benefits. The reality is that employers pay unemployment taxes, not employees.

And where do you think the money for the employer to pay the taxes comes from? Let's quickly switch over to an idea that was popular in the summer of 2008, increased taxes on oil companies. According to The Tax Foundation:

The U.S. debate over which group of people — workers, shareholders or consumers — ends up paying the bulk of these corporate taxes will go on forever. But the undisputed and most important point is that individuals pay taxes, not corporations. Therefore, attempts to punish oil companies for "obscene" profits by instituting additional taxes ultimately cause all of us to pay the price.

I would imagine that for a tax that is levied per worker (such as unemployment insurance), the cost will be borne mostly by workers. Even if the workers do not see the effect immediately (in a lower income), they will see it over time in the form of reduced wage growth.

Love of Reagan

As someone born after the era of Reagan, I have trouble understanding why Republicans love him so much. I understand why they like him: after all, he implemented a huge swath of the conservative agenda. However, the unadulterated love is bizarre. After all, after cutting taxes in 1981, he hiked them in 1982, 1984, and again in 1987 (source). And Iran-Contra? He negotiated with terrorists. And when it was discovered, he took full responsibility, stating:

"A few months ago I told the American people I did not trade arms for hostages. My heart and my best intentions still tell me that's true, but the facts and the evidence tell me it is not."

It reminds me a lot of "it depends on what the definition of 'is' is."

The reason I am talking about Reagan is that I came across this gem by David Frum:

[...] and President Reagan who allowed himself to be photographed lifting weights. (He looked good at it too.)

I couldn't imagine Reagan lifting weights. And if he were lifting weights, I couldn't imagine him looking good at it. I imagine it would look, in fact, look rather ridiculous. After a quick search, I could only find one picture. It is for you to decide whether or not he looks "good".

Image source: http://commons.wikimedia.org/wiki/File:Reagan_Lifting_Weights.jpg

1 in 10 Will Get Everything Wrong

Over at the Daily Dish, Patrick Appel thinks that it is amazing that "7% of those who voted for John McCain do not believe Hawaii to be a part of the United States". I think that this is very consistent with just about all opinion polls I have seen. In general, if there is an answer (or set of answers) that is logically and/or factually impossible, about 10% of people will answer it wrong.

For example, this Fox News poll taken in 2004 shows that 85% of Americans believe in heaven, but only 74% believe in hell. Assuming that the group of people who believe in hell also all believe in heaven, that leaves about 11% of Americans believing in heaven but not hell. Sooooo.... no matter what you do, you will go to heaven. I seriously doubt that 11% of Americans believe that, but 11% of Americans say they believe that.

Andrew Sullivan and Jesus

Sam Harris has a post up critiquing the irrationality of Francis Collins (the next director of the National Institutes of Health) when it comes to religion. Harris writes that "Collins believes in a suite of canonical miracles, including the virgin birth and literal resurrection of Jesus Christ." This made my mind wander over to Andrew Sullivan, another believer with considerable intellectual heft. I know that he doesn't believe the Book of Genesis to be literal truth. Very few intellectuals do. However, a more interesting question is whether he believes in the Biblical tale of Jesus? In years of reading Sullivan, I can't remember him ever addressing this question. I would be very interested to hear his response.

Julie & Julia

I saw Julie & Julia tonight. It's a good little film. The title of one of the cooking books, the "Larousse Gastronomique", put me back in high school, hating the Petit Larousse dictionary.

One of the interesting scenes is in the previews, where Julia Child spills some food on the stove while attempting to flip it. She promptly picks the food up off the stove and puts it back in the frying pan. In the theatre, a good dozen people made "ugh" sounds as she did that. I wonder how many of those people cook.

Death panels

Sarah Palin says:

The America I know and love is not one in which my parents or my baby with Down Syndrome will have to stand in front of Obama’s “death panel” so his bureaucrats can decide, based on a subjective judgment of their “level of productivity in society,” whether they are worthy of health care. Such a system is downright evil.

In the America that Sarah Palin knows and loves, there are three major ways to obtain health insurance:

• Government: This option reeks of socialism. 'Nuff said.
• Employer: Employer-based health care definitely doesn't discriminate against people based on their level of productivity. Plenty of unproductive paper shufflers have health insurance, while many productive janitors are uncovered.
• Private insurance: I can only imagine what the premiums would look like for someone with a pre-existing condition such as Down syndrome. On the other hand, this doesn't discriminate against people with low levels of productivity. It just discriminates against the poor.

Shorter Sarah Palin: I don't want government bureaucrats to ration health care based on their productivity. I want insurance companies to do it based on their wealth.

Robots and computers

Over in the WaPo, Gregory Clark has an article about the robots that are coming to steal our jobs. Ryan Avent believes that Clark is wrong:

But this is silly. Why? Machine and robotic resources aren’t free; they’re resource constrained just like everything else is resource constrained.

[snip]

Meanwhile, as demand for robot labour increases, the price of robot labour will also increase (since the stuff robots are made of is scarce), making the use of a robot for any given task less attractive.

This seems like basic Econ 101. Demand goes up, price goes up. However, let's see what happens when we apply that second sentence to computers.

Meanwhile, as demand for robot labour computers increases, the price of robot labour computers will also increase (since the stuff robots computers are made of is scarce), making the use of a robot computer for any given task less attractive.

This is, of course, patently absurd. The demand for computers has undoubtedly increased, pushing the demand curve to the right. If supply stays constant, this should increase the price of computers. However, because of incredible innovation in the design and manufacture of computers, supply has not stayed constant. The supply curve has shifted even farther to the right, resulting in increased quantities of computers sold at a lower price.