Progress Daily

Place a rat in a cage with an electrified floor and subject it to repeated shocks; it will show many signs of stress, at first flinging itself against the walls with each shock. But after a while, it just sits there apathetically, showing no inclination to escape from its painful prison. When autopsied, it will be found to have oversized adrenal glands and, frequently, stomach ulcers, both indicating serious stress.

Now repeat the experiment, but with a wooden stick in the cage alongside the rat. When shocked, the rat chews on the stick, and as a result, it can endure its experience much longer without burnout. At autopsy, its adrenal glands are smaller, stomach ulcers fewer.

Put 2 rats in the electrified cage. Shock them both. They snarl and fight. At autopsy, their adrenal glands are normal, and, even though they have experienced numerous shocks, they have no ulcers.

Recently physiologists have uncovered the hormonal basis for such behavior. Animals and people subjected to attack or threat experience “subordination stress,” as a result of which their adrenal hormones go up, along with blood pressure and the probability of developing ulcers. But when given the opportunity to “take it out” on someone else, victims show no sign of stress.

“The Targets of Aggression,” by David P. Barash

The 2007 Statistical Abstract of the United States lists death rates per 100,000 population.

The 1951 Statistical Abstract of the United States lists death rates per 1000 population.

In 1949, the death rate of 1-4 year olds per 1000 population was 1.5 (itself a huge improvement over 1900’s 20). In 2003, the death rate of 1-4 year olds per 100,000 population was 32.

If the denominator hadn’t been changed, modern tables would be hard to read because the numbers would not be whole numbers.

“Julian Simon Moment of the Day,” Bryan Caplan

For a couple of years, researchers (Ray Fisman, Sheena Iyengar, Itamar Simonson, and Emir Kamenica) ran a speed-dating experiment at a local bar just off the Columbia campus.

Subjects were asked to rate their partners’ intelligence, looks, and ambition after each meeting. Each event had between 10 and 20 daters of each gender, and in the course of the evening, every man met every woman and vice versa.

The researchers collected data on thousands of decisions made by more than 400 daters from Columbia University’s graduate and professional schools.

They found that men put significantly more weight on their assessment of a partner’s beauty, when choosing, than women did.

Intelligence ratings were more than twice as important in predicting women’s choices as men’s. Men tended to prefer women whom they rated as relatively smart and ambitious but they avoided women whom they perceived to be smarter and more ambitious than themselves.

When women were the ones choosing, the more intelligence and ambition the men had, the better.

Women of all races (white, black, East Asian, and Hispanic) revealed a strong preference for men of their own race. But a woman’s race had little effect on the men’s choices.

The researchers found no evidence of a white male preference for East Asian women. But did find that East Asian women did not discriminate against white men (only against black and Hispanic men). So the white man-Asian woman pairing was the most common form of interracial dating. Daters of both sexes from south of the Mason-Dixon Line revealed much stronger same-race preferences than Northern daters.

“An Economist Goes to a Bar,” by Ray Fisman

If the decision to vote is based in part on genetics, James H. Fowler reasoned, identical twins should behave more alike than fraternal twins, because identical twins share all of their DNA, whereas fraternal twins share only half on average.

The researchers matched data from the Southern California Twin Registry with publicly accessible electronic voter registration and turnout records from Los Angeles County. Their analysis of voting histories for 326 identical and 196 fraternal twins suggests that genetics was responsible for 60% of differences in voting turnout between twin types.

Fowler also investigated a larger, more nationally representative database from the National Longitudinal Study of Adolescent Health, or Add Health. This study not only asked if participants voted but also inquired about participation in other political activities, such as whether they contributed to campaigns or attended political rallies or marches. The researchers’ data on 442 identical and 364 fraternal Add Health twins indicate that genetics underlies 72% of differences in voting turnout and roughly 60% of differences in other political activity. Preliminary results from the Twins Days festival in Twinsburg, Ohio, also support the findings.

Fowler hypothesizes that the drive to vote or participate in politics may be linked with genes underlying more ancient behaviors, such as innate dispositions toward cooperation.

Robert Plomin analyzed Fowler’s data, and concluded that genetics was responsible for 40%, not 60%, of differences in voting turnout between twin types. 40% is also the average estimate of heritability seen in twin studies of personality.

Evan Balaban cautions that about two thirds of identical twins share the same bloodstream while fetuses, so greater similarities between twins could be attributable not only to sharing genes but to sharing more similar levels of hormones and other compounds each fetus produces during development.

“The Genetics of Politics,” Scientific American

Studies have shown that the IQs of breastfed children are, on average, about 6 points higher than those given baby formulas.

Terrie Moffitt and Avshalom Caspi suspected the involvement of a gene called FADS2, which regulates the metabolism of long-chain polyunsaturated fatty acids — which are important for the growth of nerve cells and are abundant in human milk. FADS2 comes in two varieties, known as C and G, and the researchers wondered if these two varieties interacted differently with breast milk.

To find out (see “Moderation of breastfeeding effects on the IQ by genetic variation in fatty acid metabolism“), they drew on data from two groups of people, one in New Zealand (born in 1972 and 1973) and one in Britain (born in 1994 and 1995).

The researchers found that the increase in intelligence associated with breastfeeding only happened to people who had inherited at least one copy of the C version of FADS2. (Most genes are present as two copies, one inherited from the mother and one from the father.) The effect did not depend on parental social class or IQ, nor on birthweight (low birthweight has been linked to lower IQ). And the difference in IQ was preserved into adulthood.

Only about 10% of the population is double-G.

“The nature of nurture,” The Economist

The US spends 1/6 of national income on medicine, more than on all manufacturing.In the aggregate (in studies with good controls), variations in medical spending usually show no statistically significant medical effect on health.

For example, a 2007 study (”Health care funding levels and patient outcomes“) found no significant mortality effects of funding variations across 22 U.S. Veterans Affairs regions over six years.

A 2003 study of 18,000 patients (”The Implications of Regional Variations in Medicare Spending“) confirmed a 2000 study (”Associations among hospital capacity, utilization, and mortality of US Medicare beneficiaries“), and a related 1998 study (”How Much is Enough?“), which together used a huge dataset: 5 million Medicare patients in 1989 and 1990 across 3,400 U.S. hospital regions.

Regions that paid more to have patients stay in intensive care rooms for one more day during their last 6 months of life were estimated, at a 2% significance level, to make patients live roughly 40 fewer days, even after controlling for: individual age, gender, and race; zipcode urbanity, education, poverty, income, disability, and marital and employment status; and hospital-area illness rates. The researchers estimated that a region spending $1,000 more overall in the last 6 months of life gave local patients somewhere between a gain of 5 days of life and a loss of 20 days of life (95% confidence interval).

The tiny effect of medicine is in striking contrast to the large effects of other influences. For example, a 1998 study (”Socioeconomic Factors, Health Behaviors, and Mortality“) of 3,600 adults over 7.5 years found significant lifespan effects: a 3 year loss for smoking, a 6 year gain for rural living, a 10 year loss for being underweight, and about 15 year losses each for low income and low physical activity (in addition to the usual effects of age and gender).

These studies look at correlation, not causation, between health and medicine. But not so the RAND health insurance experiment.

From 1974 to 1982 this experiment spent about $50 million to randomly assign over 2000 non-elderly families in 6 US cities to 3 to 5 years of a specific medical price, ranging from free to full price, provided by the same set of doctors. (See “The Effect of Coinsurance on the Health of Adults” and Free for All?.) Being assigned a low price for medicine caused patients to consume about 30% (or $300) more in per-person annual medical spending.

The result: “For the five general health measures, we could detect no significant positive effect of free care for persons who differed by income … and by initial health status.”

There were no significant differences in either severity of diagnosis or appropriateness of treatment between common and extra medicine.

Other studies suggest larger benefits in particular areas, e.g., immunization, infant care, and emergency care. (See “The impact of public spending on health” and “Returns to Local-Area Health Care Spending“.)

“Cut Medicine in Half,” by Robin Hanson

The ultimatum game brings out conflicting impulses: a researcher offers two players a set amount of money and explains that if they agree on how to divvy it up they will keep that money for themselves. If they don’t, neither will get anything. One player then offers the other a split. Most players reject a patently unfair division — such as offering only $4 out of a total of $20. Yet, self interest would argue that even $4 is better than nothing, which is what will otherwise result.

Daria Knoch and economist Ernst Fehr studied 52 young men in the ultimatum game. The researchers specified the amounts that could be offered - ranging from four to 10 Swiss francs out of 20 - and had computers randomly select some of the offers. This helped distinguish between the recipients demand for reciprocity - only applicable when another human being is in control - and a general resistance to unfair offers. The scientists divided the recipients into 3 groups: those who would receive transcranial magnetic stimulation (TMS) to suppress electrical activity in the right side of their prefrontal cortices, those who would receive the treatment on the left side and, as controls, those who would receive no stimulation.

45% of the men who experienced TMS on the right side of their prefrontal cortex accepted the most unfair offers - a split of 16 to 4 - compared with just 15% of those whose left side had been stimulated and 9% of the controls. 37% of those who underwent right side stimulation accepted all unfair offers - judged as any split less than 10 to 10 - whereas no one was so accepting in the other groups. And they made the decision to accept an unfair offer as quickly as a fair one, while their colleagues needed much longer to decide. This marks the first time that brain researchers have controlled a specific behavior by using TMS on a specific region of the brain. It takes at least 15 minutes of direct application to the skull to induce the changes, and they only last a short while.

“Selfish Impulse Set Free by Magnetic Pulse to Brain,” by David Biello

A 1987 study (”Applying the acute ischemic heart disease predictive instrument“) showed that a successful predictive instrument for acute ischemic heart disease (which reduced the false positive rate from 71% to 0) was, after its use in randomized trials, all but discarded by doctors (only 2.8% of the sample continued to use it).

Most doctors do not like decision aids. And patients think less of doctors’ abilities who rely on such aids. (See “Patients Derogate Physicians Who use a Computer-Assisted Diagnostic Aid.”)

Doctors cannot outperform mechanical diagnoses because their own diagnoses are inconsistent. An algorithm guarantees the same input results in the same output, and this maximizes accuracy.

Physicians can find “exceptions” everywhere they look, and, augmenting a decision aid as they see fit, will only end up lowering its overall diagnostic accuracy — because doctors are subject to random fluctuations in diagnosis caused by judgmentally-irrelevant factors including availability, priming, recency effects, inconsistent weighting of information, fatigue, etc., all of which reduce accuracy. (See “Clinical versus mechanical prediction” and “Clinical versus Actuarial Judgment.”)

“How Doctors Think They Think,” by Charles Lambdin