Wednesday, August 29, 2007

What would Joe Sixpack do?

One of the great shared assumptions of the American middle class is that America is a land of opportunity: work hard, save your money, and you've got an excellent chance to have a better life than your parents did. This assumption drives much immigration to the US, past and present. It also drives much American social policy: if you can't get ahead in the land of opportunity, then either you aren't working hard (the conservative view) or you have somehow been deprived of a fair chance (the liberal view), or some combination of both (the moderate view). Among people who share this assumption, working hard is clearly the rational, self-interested thing to do.

Long term poverty is therefore a paradox. Poor people will see a larger improvement in their lives from each extra dollar, and therefore should be even more motivated to save and work hard. Yet study after study shows that poor people are less likely to finish school, more likely to make poor financial choices, more likely to develop crippling addictions. From a conventional economic point of view, this is clearly irrational behavior, which can only be explained by some combination of laziness, poor role models, lack of accurate information, and dependence on government handouts. (Liberals and conservatives are likely to pick different mixes of these characteristics.)

But what if you assume that poor people are as rational as anyone else? Steven Pearlstein explains that a rational model of the behaviors seen in very poor communities might reach a very different conclusion. If the chance of building a better life appears to be zero, then maybe blowing that $100 bill on a night out isn't so irrational after all. Sure, saving it might help pay this month's rent, but it won't help with next month or the month after that. Enjoy it now and it won't get stolen, either.

A rational person model has important implications for social policy, as it implies that trying to "fix" the apparently irrational behavior of people in poor communities is pointless. Rather, the goal of social policy should be to try to align the rational decisions of poor people with those of the middle class. That probably means closing the gap: encouraging poor people to believe that the middle class is accessible to them.

It's an interesting article. Pearlstein makes a far more rigorous argument than most people on either end of the current political debate. Agree or disagree, he at least offers a new way to think about the issues.

Saturday, August 18, 2007

Teaching the Best and Brightest

Education researchers and school systems invest enormous sums in figuring out the best ways to teach students with learning disabilities. Students at the far upper end of the IQ scale tend to get much less attention. They're smart, they'll do okay anywhere, right?

No, actually they won't. It turns out that gifted students have about the same dropout rate that students with learning disabilities do, probably due to boredom. Sometimes they blossom when they reach schools like MIT or Stanford, where they find peers for the first time in their lives, but sometimes they never get that far. Time explains the lost potential these kids represent, and considers ways to keep them from falling through the cracks.

Friday, August 17, 2007

Way down in the mines

The unconfined compressive strength of coal is about 3000 psi. Rock weighs about 160 pounds per cubic foot. Those miners in Utah are about 1500 feet down.

I'm not a mining engineer, but it looks to me like those "seismic bumps" are the mountain telling them that they're too close to the edge. I hope no one else dies before they get the message.

Update: Robert Ferriter of the Colorado School of Mines is a mining engineer. He's surprised that area was still being mined at all.

Monday, August 13, 2007

Yes, really, a word count!

I somehow managed to keep track of my totals through the move and related chaos, which is an accomplishment in itself. The total itself is less impressive, though. 20,500 words since my last report, giving a total of 73,850 for the year.

Wednesday, August 8, 2007

All silicon products not created equal

While it's true that many specifications are more relaxed in solar cells than in ICs, that doesn't mean that tackling the solar cell market will be easy for IC manufacturers. Rebecca Tang, general manager at Mosel Vitelic, warns that process control, product quality, and fab management all require a change in perspective. IC manufacturers and equipment suppliers who assume that their expertise in these areas will continue to apply may be in for a rude awakening.

Friday, August 3, 2007

Time for designers to tighten up?

Design re-spins are bad. They cost money, they delay the product, and they force the group responsible to revisit a project that they thought was complete. No one wants a re-spin.

One way to reduce re-spins is to build more margin into the design. More margin means more flexibility to handle process variations, more tolerance for unexpected conditions in other parts of the design.

Unfortunately, as Cadence VP Steve Carlson points out, all that margin erodes product performance and, ultimately, boosts development costs. Oops!