By Kevin Meyer

In this edition of Fun With Statistics we explore the gender ratio and a couple of surprising impacts.  This morning's Wall Street Journal had a review of a recent book on the issue:

Mara Hvistendahl is worried about girls. Not in any political, moral or cultural sense but as an existential matter. She is right to be. In China, India and numerous other countries (both developing and developed), there are many more men than women, the result of systematic campaigns against baby girls. In "Unnatural Selection," Ms. Hvistendahl reports on this gender imbalance: what it is, how it came to be and what it means for the future.

I'll warn you – the first three-quarters of the review are great, but the last quarter wades into some political swamps that I dare not enter.  But if you're interested in a discussion of "choice" and "gendercide" and how some very surprising activities of organizations like Planned Parenthood helped create this global problem, dive right into it.  I know better so I won't – and as a libertarian I really don't care that much.

So here's the fundamental problem:

In nature, 105 boys are born for every 100 girls. This ratio is biologically ironclad. Between 104 and 106 is the normal range, and that's as far as the natural window goes. Any other number is the result of unnatural events.

Yet today in India there are 112 boys born for every 100 girls. In China, the number is 121—though plenty of Chinese towns are over the 150 mark. China's and India's populations are mammoth enough that their outlying sex ratios have skewed the global average to a biologically impossible 107. But the imbalance is not only in Asia. Azerbaijan stands at 115, Georgia at 118 and Armenia at 120.

For second births [in South Korea], the male number climbed to 113; for third, to 185. Among fourth-born children, it was a mind-boggling 209. Even more alarming is that people maintain their cultural assumptions even in the diaspora; research shows a similar birth-preference pattern among couples of Chinese, Indian and Korean descent right here in America.

Our intuition as to why this happens is probably correct – boys are more valuable in developing societies as they can contribute more immediately to the income and welfare of the family.  So here comes surprising fact number one:

But oddly enough, Ms. Hvistendahl notes, it is usually a country's rich, not its poor, who lead the way in choosing against girls. "Sex selection typically starts with the urban, well-educated stratum of society," she writes. "Elites are the first to gain access to a new technology, whether MRI scanners, smart phones—or ultrasound machines." The behavior of elites then filters down until it becomes part of the broader culture. Even more unexpectedly, the decision to abort baby girls is usually made by women—either by the mother or, sometimes, the mother-in-law.

Lopsided gender ratios are not exactly a good thing.

Ms. Hvistendahl argues that such imbalances are portents of Very Bad Things to come. "Historically, societies in which men substantially outnumber women are not nice places to live," she writes. "Often they are unstable. Sometimes they are violent." As examples she notes that high sex ratios were at play as far back as the fourth century B.C. in Athens—a particularly bloody time in Greek history—and during China's Taiping Rebellion in the mid-19th century. (Both eras featured widespread female infanticide.) She also notes that the dearth of women along the frontier in the American West probably had a lot to do with its being wild. In 1870, for instance, the sex ratio west of the Mississippi was 125 to 100. In California it was 166 to 100. In Nevada it was 320. In western Kansas, it was 768.

So now we have not-so-surprising fact number two:

Unmarried men with limited incomes tend to make trouble. In Chinese provinces where the sex ratio has spiked, a crime wave has followed. Today in India, the best predictor of violence and crime for any given area is not income but sex ratio.

You would think that a shortage of women would then increase the value of women and eventually create a rebalancing.  It does – but not necessarily in a positive way.

And to beat the "marriage squeeze" caused by skewed sex ratios, men in wealthier imbalanced countries poach women from poorer ones. Ms. Hvistendahl reports from Vietnam, where the mail-order-bride business is booming thanks to the demand for women in China. Prostitution booms, too—and not the sex-positive kind that Western feminists are so fond of.

The economist Gary Becker has noted that when women become scarce, their value increases, and he sees this as a positive development. But as Ms. Hvistendahl demonstrates, "this assessment is true only in the crudest sense." A 17-year-old girl in a developing country is in no position to capture her own value. Instead, a young woman may well become chattel, providing income either for their families or for pimps.

But here comes suprising fact number three – and what really grabbed my attention:

A high level of male births has other, far-reaching, effects. It becomes harder to secure a bride, and men can find themselves buying or bidding for them. This, Ms. Hvistendahl notes, contributes to China's astronomical household savings rate; parents know they must save up in order to secure brides for their sons. (An ironic reflection of the Indian ad campaigns suggesting parents save money by aborting girls.) This savings rate, in turn, drives the Chinese demand for U.S. Treasury bills.

So there you have it – the real reason for China's growth and thirst for stable securities to invest in.  Compliments of Planned Parenthood (if you read the part of the article I warned you about…).

By Kevin Meyer

In this edition of Fun With Statistics we'll tackle the brouhaha over the past couple days regarding the discovery that the iPhone as well as phones running Google's Android software are tracking their users' every move.  This has been entertaining to some, disturbing to others. 

I'm personally not all that worked up as location-based services create a lot of benefits – like being able to find the best Italian restaurant close to me in San Francisco last night.  Besides, I do encrypt the contents of my iPhone just as a standard precaution, and unless you live under a rock you realize that your identity and location are logged indirectly a hundred other ways already – when you use your ATM or credit card, log onto any computer network or wi-fi, fly, etc.

Running the iPhoneTracker app that taps into the no-longer-hidden file with the history of where my iPhone has been over the last 10 or so months was interesting.  Here's the large global graphic:

Several countries and of course my quarterly sojourns to the Big Island.  The app lets you zoom in to a square mile resolution, scroll by date, and the works.  Pretty interesting to watch – sort of a diary of where I've been.  Whoopee.

Today's Wall Street Journal describes the real potential of this data even further.

For almost two years, Alex Pentland at the Massachusetts Institute of Technology has tracked 60 families living in campus quarters via sensors and software on their smartphones—recording their movements, relationships, moods, health, calling habits and spending. In this wealth of intimate detail, he is finding patterns of human behavior that could reveal how millions of people interact at home, work and play.

Through these and other cellphone research projects, scientists are able to pinpoint "influencers," the people most likely to make others change their minds. The data can predict with uncanny accuracy where people are likely to be at any given time in the future.

Uncanny accuracy is right.

Advances in statistics, psychology and the science of social networks are giving researchers the tools to find patterns of human dynamics too subtle to detect by other means. At Northeastern University in Boston, network physicists discovered just how predictable people could be by studying the travel routines of 100,000 European mobile-phone users.

After analyzing more than 16 million records of call date, time and position, the researchers determined that, taken together, people's movements appeared to follow a mathematical pattern. The scientists said that, with enough information about past movements, they could forecast someone's future whereabouts with 93.6% accuracy.  The pattern held true whether people stayed close to home or traveled widely, and wasn't affected by the phone user's age or gender.

93% – incredible.  Imagine how further refined that could be with the combined (presumably anonymous) records of the 75% of the world's population that now carries cell phones.  But where you will be is just the tip of the iceberg.

The data can reveal subtle symptoms of mental illness, foretell movements in the Dow Jones Industrial Average, and chart the spread of political ideas as they move through a community much like a contagious virus, research shows.

Dr. Bollen and his colleagues, for example, found that the millions of Twitter messages sent via mobile phones and computers every day captured swings in national mood that presaged changes in the Dow Jones index up to six days in advance with 87.6% accuracy.

And back at MIT, scientists who tracked student cellphones during the latest presidential election were able to deduce that two people were talking about politics, even though the researchers didn't know the content of the conversation. By analyzing changes in movement and communication patterns, researchers could also detect flu symptoms before the students themselves realized they were getting sick.

So now are you worried, or intrigued with the potential of crowd statistics?  As a glass half full kind of guy who still remembers the folks scared about how the new networked world was going to destroy privacy… back in the early 1990s, I'm intrigued.  And as that tune oft attributed to Ozzy but really by the Swollen Members goes, "… paranoia will destroy ya…"

By Kevin Meyer

In this edition of Fun With Statistics we'll take a quick look at a common problem when communicating statistics: describing orders of magnitude.

I'm still scheduled to be in Yokohama in a month to attend a trade show, so obviously I'm keeping a close eye on the situation in Japan.  I love the country and am looking forward to returning for a few days, but the radiation issue is worrisome.  I haven't changed any plans figuring that in a couple weeks it will probably either be stabilized and under control or really, really bad – making the decision easy. 

A sievert here, a rad there, but at what point is it really a problem.  How does it compare with a chest x-ray… to a plane trip… to Chernobyl?

It turns out pretty small, at least for now.  Sometimes it just takes a visual to help show the difference in magnitudes.  I could improve on this by simplifying it further, but as a start it works great.  Here's a full-size version of the chart below.

By Kevin Meyer

In this edition of our irregular Fun with Statistics series we take a look at an oft-unforeseen pitfall of analysis: inadvertently framing the analysis.  How often do we ask our subservient underlings for some report or analysis, but then skew that analysis by drawing boundaries around potential outcomes?  We may think we're just providing "guidance," but where does guidance end and prejudicial blinders begin?

A good article in The Atlantic points out this potential pitfall using one of our, uh, favorite fellows – Paul Krugman – as an example.

Paul Krugman has an interesting post up on how to read a CBO document. He highlights the fact that the CBO reports using the assumptions it is required to make, not any estimate of what is particularly likely to happen in the real world:

What you need to realize is that the CBO is the servant of members of Congress, which means that if a Congressman asks it to analyze a plan under certain assumptions, it will do just that — no matter how unrealistic the assumptions may be. CBO will tell you what's going on, but it will do so deadpan, doing nothing in terms of emphasis or placement to highlight the funny business.

Now of course the interesting angle to this is that Krugman himself often used CBO numbers to justify some of the nonsense he was bloviating about, even just the prior day in an article on how the Medicare Trustees believes that changes resulting from healthcare reform will reduce the deficit.  As Megan McArdle of The Atlantic continues:

I could wish that he had gone over this topic just a wee bit earlier–like yesterday, when he argued that the Medicare Trustees report shows that we've done a whole bunch of deficit reduction:

In other words, the Medicare actuaries believe that the cost-saving provisions in the Obama health reform will make a huge difference to the long-run budget outlook. Yes, it's just a projection, and debatable like all projections. And it's still not enough. But anyone who both claims to be worried about the long-run deficit and was opposed to health reform has some explaining to do. All the facts we have suggest that health reform was the biggest move toward fiscal responsibility in a long, long time.

There are two problems with this claim. The first, as I noted earlier this week when Health and Human Services Secretary Kathleen Sebelius made a similar claim, is that it requires double counting. If the money is used to extend the program's trust fund, then it can't be used elsewhere.

The second is that Medicare's actuaries aren't confident that the program cuts called for by the PPACA will actually happen. For example, page three of the report's introduction notes that the headline estimates assume that Medicare payment rates to physicians will be cut by a total of 30 percent over the next three years despite "virtual certainty" that legislators will override the scheduled reductions. Indeed, in a somewhat extraordinary caveat, the Trustees report explicitly warns readers not to presume that its top-line cost estimates represent the most likely outcome: "It is important to note that the actual future costs for Medicare are likely to exceed those shown by the current-law projections in this report."

I wish Paul Krugman would do some explaining–about the source of his differential skepticism.

In this case the Medicare Trustees and actuaries actually found a way to insert a bit of their own professional opinion, constrained within the artificial boundaries imposed on them by those that requested the report: that model of fiscal prudence themselves, Congress.

There's a good lesson for us: be careful of constraining analysis by "guidance" or preconceived opinion.  You may get what you want, but it may be incorrect.