Business

Why aren't there more women in math and science?

From: Blog

It’s a hard pill to swallow, but it’s the truth: Gender bias is costing women opportunities in math and science careers. Women are still conspicuously outnumbered by men in those fields, even though more women enroll in undergraduate programs.
 
It seems that discrimination based on gender stereotypes is at least partly to blame, as explained in a new paper published in the Proceedings from the National Academy of Sciences, or PNAS. In "How stereotypes impair women's careers in science," three business school professors design an experiment to isolate discrimination's potential effect on hiring decisions. 

What they found is sobering: When a hiring decision is based solely on appearances, men are twice as likely as women to be hired for a mathematical task. When a hiring decision also includes the candidates' self-reported ability, women still lose out (men tend to overstate, women tend to understate, and hiring managers don't adjust for this). 
 
The researchers, Ernesto Reuben of Columbia, Paola Sapienza of Northwestern Kellogg, and Luigi Zingales of Chicago Booth, did find a bit of evidence that preserves hope for aspiring women mathematicians and scientists. When hiring managers are given an "implicit association test," or IAT, and are able to see hard evidence of their math and science bias against women, their prejudice can be reduced. 
 
Some of you may be wondering whether you're part of the problem. Some of you may just be curious. Whatever the case, you can take and IAT to test your own biases—of gender, as well as race, weight, age, religion, and more—and contribute to further research by participating in Project Implicit, a non-profit organization and international collaboration between researchers.
 
Seriously, go take an IAT right now, and brace yourself. When you're done with that, read this Reddit AMA with Project Implicit, and if you want more about the study, see what Shaila Dewan has to say at the New York Times Economix blog
 
—Chelsea Vail Cat:Business, Sub:Economics, Behavioral Science,

Business

Good news about health care

From: Blog

New research shows productive hospitals are rewarded by market

American hospitals aren't cheap, as the Economist has reported. In fact, they’re the most expensive in the world by a large margin (with average costs of $4,300 per patient per day, versus $1,500 in Australia, the next highest, according to the International Federation of Health Plans).

It appears that American hospitals do not compete with each other in the same way firms do in the retail and manufacturing sectors. For one, prices aren’t transparent. (Want 1950s prices for health care? Read this earlier post to find out why you likely don’t.) Patients with health insurance may be charged more and never know it, because their policies cover the difference. And of course, insurance agencies take hospital invoices as a starting point for negotiation. There is little incentive for the patients themselves to shop around for a good deal.

Just as with price, quality of service is often difficult for patients to compare across hospitals. Whether a treatment is successful has no bearing on whether a hospital gets paid—so the emphasis is on quantity. Moreover, the barriers to entry are high and it is difficult for new, potentially more efficient hospitals to enter the market.

Without intense competition, one would expect a wide dispersion in productivity among hospitals, measured, for example, by how long a patient survives after a medical treatment. But new research shows that productivity across hospitals in treating heart attacks is slightly less dispersed than productivity in narrowly-defined manufacturing industries such as plants that made ready-mixed concrete.

This is a sign that hospitals may have more in common with “traditional” sectors than what conventional wisdom suggests, according to Chad Syverson of Chicago Booth, and his coauthors Amitabh Chandra of Harvard, and Amy Finkelstein and Adam Sacarny of MIT.

In addition, hospitals that were more productive in treating heart attacks had bigger market shares and were more likely to expand over time, which is what one would expect in a well-functioning market. A 10 percent increase in productivity resulted in four percent more heart attack patients in five years and almost six percent more in 10 years.

Are these signs of robust competition? Perhaps so, if patients were indeed choosing hospitals base on which one they thought offered the best chance of survival for the treatment they needed. Without actual productivity numbers on hand, patients may have relied on a hospital’s reputation for performing complex procedures such as bypass surgeries and angioplasties to help decide which hospital to choose.

But it’s also possible that the data were driven by factors that have nothing to do with a hospital’s reputation for success. “It could be that competitive market forces reallocate market share to higher productivity hospitals, or it could be that higher productivity hospitals happen to have other features—such as beautiful lobbies or good managers—which separately increase demand,” observed the authors.

Whatever the driving force behind the results, higher productivity hospitals do appear to be rewarded by the market.

—Vanessa Sumo

TAGS: Health care, healthcare, health-care, hospital, productivity, efficiency, survival rateCat:Business, Sub:Economics,