Mindless Philosopher

A few weeks ago, I attended an amazing wedding in Jackson Hole. The ceremony took place on the banks of the Snake River and the bride rode in on a horse-drawn carriage. What impressed me even further is that the couple chose excerpts from A.A. Milne and Taylor Mali as poetry readings; I’m big fans of both authors.

The Mali excerpt was from “How Falling in Love is Like Owning a Dog,” a very clever poem that I hadn’t heard before. (Word to the wise, the rest of this post won’t make sense unless you click that youtube link.)

Needing something clever myself for an entry in the guestbook, I thought I would try my hand at Mali’s analogy with a slight twist:

How Falling In Love Is Like Owning A Cat
A. Strauss

One day, unexpectedly, Love shows up on your doorstep ready to be a life partner.

Circumstances may be trying, but Love always lands on its feet.

Push the wrong buttons and Love scratches. But one caress behind the ears, and the claws retract.

Sometimes Love needs to be alone. And sometimes it jumps on you from above — Love is full of surprises.

Most importantly, Love always finds its way home.

Happy Marriage, Libby & Dave!

A few weeks ago, I wrote the following nerdy and somewhat-cryptic tweet:

Abromowitz polarization variable: if(yr<96,0,if(is.Inc | app>0,1,-1)) counts as 4 degs of freedom? http://j.mp/SbHyzx

The “degrees of freedom” that I reference is the statistical term, which represents the difference between he wealth of data at your disposal and the complexity of your model. Allow me to explain as best I can:

Let’s say that you want to forecast the 2012 result based on the 16 presidential elections since 1948. In this example, you start with 16 degrees of freedom because those 16 election results could be anything. And in the simplest model, you can retain all 16 of those degrees of freedom:

The simplest model would be to guess that the incumbent party will garner 50% of the vote. (A plausible reason for this guess is the Median Voter Theorem, which says that parties move to the center and votes will generally be split equally.) Because your guess stays at 50% no matter what the 16 presidential elections could possibly be, the 50% model costs you zero degrees of freedom, and you retain all 16 degrees.

But, perhaps a keener idea would be to actually examine the historical trends in an attempt to improve accuracy. In the 16 elections since 1948, the average popular vote percentage for the incumbent party’s nominee is 52.1% — suggesting that the Median Voter Theorem by itself may be missing an inherent, incumbent party advantage. This new model, displayed as the horizontal line in the chart below, costs one degree of freedom for the averaging parameter.

Here’s why that model costs one degree of freedom. Imagine that you only had one election result to work with: the 2008 campaign in which McCain (the incumbent party nominee) won with 46.3% of the two-party vote. Averaging the historical data is trivial: the sole election of 2008 averages to 46.3%. The fact that your modeled forecast exactly equals the historical data means that you have used up all of your degrees freedom left. One data point minus one average equals zero degrees of freedom. In general, each additional data point adds degrees of freedom, and each additional variable to your model subtracts a degree.

A similar thought process controls for two data points. Since one line (defined by a slope and intercept) can always exactly intersect two any two points, the lines’ variables of the slope and intercept cancel out the two data points, and there are no degrees of freedom left. Adding a data point will (usually) cause the line to start missing points, even if it fits the data well — that mis-estimation is the degree of freedom.

In general, if you have lots of degrees of freedom, but your model still estimates the data well, then you’re in good shape. Nate Silver has written extensively on forecasting, saying “A general rule of thumb is that you should have no more than one variable for every 10 or 15 cases in your data set.” This keeps your degrees of freedom from edging toward zero add complexity (e.g., variables) to your model.

Abromowitz’s 2008 forecasting model, which was based on 15 data points, included three variables: presidential approval, the economy (specifically, GDP), and whether an incumbent president is running. Throw in the implicit historical average, and his model’s complexity costs him four degrees of freedom, leaving him with 11. With so few presidential elections to learn from, he was already pushing the limits of the data. Yet, after the 2008 election, he noticed another trend in the data:

The unexpected closeness of all four presidential elections since 1996 suggests that growing partisan polarization is resulting in a decreased advantage for candidates favored by election fundamentals, including first-term incumbents. … In fact, the last four presidential elections have produced the closest victory margins and the smallest inter-election vote swings of any four consecutive elections in the past century.

To account for this trend, he added another variable to his model: polarization. At first blush, this new parameter may appear to cancel out only one degree of freedom from the model (leaving Abromowitz with the same number of DoFs as he had going into 2008 because he gained one data point from Obama v McCain). But, the description of the variable demonstrates more complexity than normal:

For elections since 1996, the polarization variable takes on the value 1 when there is a first-term incumbent running or when the incumbent president has a net approval rating of greater than zero; it takes on the value -1 when there is not a first-term incumbent running and the incumbent president has a net approval rating of less than zero.

It the polarization variable took on a value of 1 post-1996 and 0 before 1996, then that addition would clearly cost Abromowitz one degree of freedom — just as approval and incumbency cost him one degree of freedom. To be specific, what costs Abromowitz the degree of freedom is his ability, retrospectively, to examine the data and atheoretically pick 1996 as the transition point. However, his polarization variable does not just take on one value after 1996. Rather, the variable’s post-1996 value depends on both approval rating and incumbency — his picking of those two criteria eliminate another two degrees of freedom, costing him a total of three. (The question mark in my tweet was very appropriate as I miscounted originally.)

Thus, Abromowitz’s model is now down to (16 – 4 – 3) nine degrees of freedom — in other words, he’s used up nearly half of the degrees available from his data. While I believe that the American electorate has become more polarized, this level of model complexity (relative to the data set) reeks of overfitting to me.

Microfinance was a term that I heard bandied about in social situations but was generally unfamiliar with until a few years ago, when a friend gave me a Kiva gift certificate as a birthday present. I found Kiva’s model of change captivating and exciting: I would browse through profiles of people, co-ops, and businesses from around the world and choose whom to lend money to. Once enough people like me had funded a loan, the borrower would receive the money and pay us back, interest free. With my money back in pocket, I could re-loan the principal to another worthy cause.

I became an advocate for Kiva and encouraged my friends to try it out. And since I strive to provide well-informed recommendations, I discovered some details about Kiva that somewhat diminish its illustrious glow. First, a Kiva lender’s loan does not actually go to the person that the user chooses during the selection process. Rather, the loan goes to the microfinance institution that has already funded the business that the lender selects. Second, though Kiva lenders’ loans appear to carry a zero interest rate, the microfinance institutions that interact with borrowers charge a hefty rate. Kiva’s defense of these rates is thorough and passionate.

With that information in hand, I still recommended Kiva, though perhaps not as enthusiastically as I might have. Recently, however, a few of my friends (who also love researching random things) have come down on the anti-Kiva side of the fence — and a friendly debate ensued. My anti-Kiva friends cited Givewell’s negative findings on Kiva, as well as evidence from recent randomly controlled experiments (RCEs).

Givewell’s findings are mainly based on the recipient and interest rate issues that I already knew. More interesting to me are the results of the RCEs. I’m impressed with the level of work it takes to set up large microfinance institutions that purposely skips over (on a random basis) either a set of individuals or communities. And, on top of that structure, these researchers had to conduct large-scale and wide-ranging survey to measure outcomes.

These experiments show that microfinance’s impact on overall well-being is minimal or non-existent (at least one year after the program begins, which is often when these surveys are taken). At first blush, that appears to be a harsh blow. The context, however, softens this judgment. Improving well-being in these communities is very very difficult; my favorite economics book, The Elusive Quest For Growth: Economists’ Adventures and Misadventures in the Tropics is an excellent primer on the history of attempts on increasing well-being in disadvantaged countries. The book’s conclusion is that systematic and institutional issues block nearly all paths to growth and that economists’ ideas, which also come across as fads, have not met their lofty expectations.

Perhaps, microfinance and Kiva is one of those fads. But I’m not looking to Kiva to solve the world’s ills. My hope is that by donating to Kiva, I’m helping. According to the RCEs, microfinance helps grow businesses and use of institutions. Specifically (study number from this list is in parentheses):

• Entrepreneurship, hiring (3), number of business (5 but not 2), and profits increased (1,2,3,4,5) especially among low-education recipients (4)
• Consumption on luxury goods, such as festivals, cigarettes, and alcohol declines (1,4,5)
• Increase in buying assets, such as livestock (3), or household appliances (4)
• Household members spend more time in education, less in wage labor (2 but not 5 [for 16-19 yr olds])
• Savings decline — presumably to help pay back the loan (5)
• In case of a financial shock (e.g., health issue), recipients are more likely to borrow from institutions rather than friends (1)
• Increased stress and worry (1,2) — again presumably because they have to pay back the loan.

It’s not a panacea (well, what is?), but microfinance helps. I’ll keep my Kiva money circulating, even as I donate to other non-profits.

Watching the BBC’s Sherlock Season 2 Finale has me in a crime-solving mood, so I devoured WaPo’s article on the Guggenheim assault. Here are my notes:

• Despite the paper calling the crime a “howdunit, a whydunit, a whendunit,” it’s highly likely that the crime was committed Thursday evening. That’s when communication from Svetlana Guggenheim stopped. She didn’t pick up the parcels that were delivered Friday. And the evidence in her room (and clothing) indicates that she was settled down for bed.
• Oddly, the paper does not provide the husband’s alibi on Thursday night. The story intimates that he has evidence that he was out of town for the entire Wednesday through Saturday stretch, but fails to mention his whereabouts at the time of the attack.
• (Too bad the husband isn’t left-handed as that would also rule him out — Guggenheim’s injuries are on the left side of her face. I trust the paper that he has an alibi though.)
• The building manager doesn’t remember whether the deadbolt was locked or not when she delivered the parcels. If it had been locked, that would be an incredible clue that the intruder likely had a key. Even if the assailant entered the home through a window, the culprit would probably take the easy way out through the front door — and the only way to lock the deadbolt would be if the criminal had a key. Given the husband’s likely alibi, I’m guessing the deadbolt was unlocked.
• Clearly Guggenheim did not let her attacker into the apartment because she was found in the bedroom wearing bedclothes. If she had let the attacker in the front door, she likely would have been clothed and attacked in the foyer.
• I’m no expert in home break-ins, but it seems odd that Guggenheim would not have heard her attacker enter if she had been awake. Perhaps she had gone to sleep, but why hadn’t she called her daughter as arranged earlier that day? Maybe her recent trip had tired her out (though she came home from Chicago — only one time zone away). It would be nice to know who was supposed to call whom and when.
• Assuming the husband’s alibi holds up (easy to check), then I’m left with three reasonable possibilities: (1) Guggenheim took something to help her sleep soundly, (2) the attacker is very stealthy, or (3) Guggenheim expected the person who entered her apartment to come to the bedroom.
• Other items left unmentioned in the article that would help me weigh the likelihoods of those avenues:
  • Did no one hear anything? The lack of a scream would be interesting since the bruising on Guggenheim’s arm indicates that she was awake when attacked.
  • What did police make of the crime scene? Did they see signs of a struggle? Or forced entry? (We only get the husband’s opinions on these matters in the article.)
  • What was her nearest means of communication? A landline by the bed? A cell phone in the other room?

I’ll be keeping an eye on how this turns out.

Short version: I made this coaster out of scrabble tiles. The eight words (across and down) are all obliquely about beverages and/or water.

Long version: I was visiting my brother and his expecting wife recently and part of the weekend was spent at the local fair. While touring the booths, we saw two knick-knacks that caught our collective eye.

The first was a hand-crafted, double-layered puzzle for kids. We discussed how to cut the pieces out of the wood while leaving a border. (I offered that that the puzzle could actually be separate layers glued together, but some challenges to crafting the pieces still remained.)

The second eye-catching ware was a variety of 4-by-4 scrabble-tile coasters. These coasters didn’t actually spell out words across and down; thus, the above challenge was born.

While I was in the midst of figuring out how to fit drink-related words together in a 4×4 grid, my brother and his wife were busy showing me up:

Yep, my bro and sis decided to tackle the double-sided puzzle problem head on. I’m still not sure how they cut the pieces out of the wood while keeping the pieces intact. Whatever the answer, I can definitely say this: Their kid is going to have very cool parents.

Poli sci attracts the kids with world domineering obsessions

Then they quit since its just theory plus multivar regressions

The Web does so much for me; today, I want to give back. First, here’s a music video that I think is astoundingly creative and beautifully executed, and which you may not have seen:

(Thanks to Pandora for introducing me to Lisa Hannigan.)

More to the point of giving back, Pandora introduced me to another song — Beautiful Things by Mary Johnson Rockers — and I had trouble deciphering some of the words. Since Pandora didn’t have the lyrics, I Googled and … nothing. I don’t think I’ve experienced such a blatant Google miss since…2003? After listening several times and successfully Googling for flower names, I think I’ve sleuthed out all the ambiguous phrases. The song can be purchased on Amazon; below are the lyrics — my gift to the Web.

Beautiful Things
by Mary Johnson Rockers

Mary said “Slow down, ease off the gas.
We got some time and we’re gonna make it last.
This afternoon, going nowhere soon.

“The day is exhale.
Got an open road
for the story tellin’.
Got a tale or two
You can listen to(o).”

Oh, beautiful things
Oh woah beautiful things
in this tangled up thicket world

Look for Queen Anne’s lace
in regular places:
Tennessee highways
by the exit ramps.

Look for Black-eye Susans
where they have room to move in
and she knows where the wildflowers bloom.

Well Willy asked Mary for her hand;
And Mary asked the Lord “Should he be my man?
He’s a baptist boy–I’m a good catholic girl.
But oh Lord I love him so;
Give me a sign so I can know.”
And it rang clear and true
in a parking lot on Dixie Avenue.

Oh, beautiful things
Oh woah beautiful things
in this tangled up thicket world.

Look for Queen Anne’s lace
in regular places:
Tennessee highways
by the exit ramps.

Look for Black-eye Susans
where they have room to move in.
And she knows where the wildflowers bloom.

Mary said “Slow down, ease off the gas.
We got some time and we’re gonna make it last.
This afternoon, we’re going nowhere soon.

“Oh, The day is exhale.
Got an open road
for the story tellin’.
Got a few more hours
for the wildflowers.”

Oh, beautiful things
Oh woah beautiful things
in this tangled up thicket world

Look for Queen Anne’s lace
in regular places:
Tennessee highways
by the exit ramps.

Look for Black-eye Susans
where they have room to move in.
And she knows where the wildflowers bloom.
She knows where the wildflowers bloom.

At the Supreme Court this week, the Obama Administration will defend its Health Care Reform Law. I have two pieces of advice for the Solicitor General.

1. State a limiting principle. The Administration has been shockingly blasé about defining exactly why a the health insurance marketplace is unique such that non-participation can be regulated. This led to some scary sentences in the circuit court rulings. From the 11th Circuit:
   

   From a doctrinal standpoint, we see no way to cabin the government’s theory only to decisions not to purchase health insurance. If an individual’s mere decision not to purchase insurance were subject to Wickard’s aggregation principle, we are unable to conceive of any product whose purchase Congress could not mandate under this line of argument….Ultimately, the government’s struggle to articulate cognizable, judicially administrable limiting principles only reiterates the conclusion we reach today: there are none.

   And from the DC Circuit Court:

   The Government concedes the novelty of the mandate and the lack of any doctrinal limiting principles; indeed, at oral argument, the Government could not identify any mandate to purchase a product or service in interstate commerce that would be unconstitutional, at least under the Commerce Clause

   While the Government did “stress that the health care market is factually unique,” I really think the Government should close the door on this attack by doctrinally stating a limiting principle.

2. That limiting principle should not focus on how non-participation “impact[s] the cost of health insurance for other Americans” (Kliff) and “raises rates for everyone else” (Cortez). Nationwide price changes does not make the health care market unique. After all, not buying Broccoli lowers aggregate demand and changes the price for everyone else.

   Rather, health care is unique because you can’t avoid participating. We’ve decided, as a society, that hospitals should treat emergencies first, and ask about insurance second. There’s even an amicus brief filed in this case from Massachusetts hospitals because they process uninsured out-of-state skiers who get into accidents and are cared for by the Bay State’s health services. That phenomenon exemplifies the fact that everyone, whether insured or not, is part of the health care marketplace.

   The same is not true for broccoli, and the government needs to make that clear this week.

   (All that said, the High Court may very well take the view that the mandate is a tax to cover the expense of providing automatic emergency services to everyone — if so, this whole discussion would be moot for the time being. Update: Looks like we’ll find out whether the law stands this year.)

In browsing my Personal WaPo, I came across Ed Roger’s anti-debt rant. I call it a rant because rather than reasoned arguments, he tries to throw big numbers at you ($341,288!) and hope you’re sufficiently outraged. It occurs to me that people like Ed (i.e., his “top 25%” that “owe” that much) probably go into more debt than $341,288 when they buy houses. And if they think that debt load is worth it to live in a good neighborhood, then I think the same load is worth it to keep our country functioning.

His article did succeed in getting me to think more about our debt, specifically how long it will take to pull it down to sustainable levels. No doubt several years. We could hasten that day if the Federal Government could refinance its entire debt at today’s incredibly low rates. Rather than paying people at a 4% rate (that investors locked in back in 2003), could we give creditors some money upfront and start paying them back at the prevailing sub-2% rate? I would much prefer to spend money now, before the bulk of the baby boomers start retiring and save on the interest later when we’ll also have to pay for the boomers’ health care.

Perhaps the Fed did something akin to what I’m suggesting in 2009 to pump money into our then-depressed economy. If so, I can’t find it in this Wikipedia article on the subject; if not, they should now.

The New York Times has a head-turning story about a James Ross, whose “taxable income” rate is 102%. Before you take too much pity on Mr. Ross, I invite you to soak in these sentences:

That doesn’t mean Mr. Ross pays more in taxes than he earns. His total tax as a percentage of his adjusted gross income was 20 percent…

Since the article states that Mr. Ross makes more than $1 million dollars a year, he has at least $800,000 each year to do with what he pleases. The quirk that enables the Times to publish a sensationalist headline is that Mr. Ross takes the bulk that that $800,000+ and pays for items that are deductible for non-millionaires, but not tax deductible for the 1%. Namely, real estate interest payments.

Thus, I take strenuous issue with Mr. Ross’s 102% calculation. He would pay more in taxes than his “taxable income,” if he were allowed to take the usual itemized deductions and reduce his taxable income. But he’s not. Thus his true taxable income (i.e., income that is taxable) is much much greater than his “taxable income” (as defined by his accountants and the NYT) — about 4.6 times greater by my calculations. His truthful tax rate? 21 percent, which is lower than mine. (And I certainly don’t make millions of dollars a year.)

The article also mentions that since interest payments are non tax deductible for the super rich, Mr. Ross has to dip into his savings to pay his taxes. This logic is sound and I believe him, but I would hasten to add that rent, groceries, and prescription drug co-pays are also not tax deductible so I’m guessing that the author of the story, James Stewart, could find many people who need to dip into their savings to pay their taxes — none of whom makes nearly as much as Mr. Ross.

In sum: the system is not broken for Mr. Ross as he pays (what I believe to be) a reasonable amount of his income in taxes — the Times just a wanted a good headline.

Note: Apologies to linking to a non-free story; the free version is available through Google.