Mac vs. PC

I was thinking about whether its good or bad for the world to have so many different operating systems and other incompatible systems. Consider a world with just two incompatible types of computers: Macs and PCs.

Suppose 90% of people use a PC, 5% use a Mac and 5% can use both. Because such a large minority can only use a Mac, every store and organization, especially schools and hospitals, have to run two support offices and publish instructions for every operation with both PC and Mac instructions. Everyone can see that if everyone used a PC society could save money, but no one knows how much.

So the question is, why not force everyone to use a PC?

1. Some people grew up using a Mac and it might be hard to learn to use a PC.

2. Maybe it's very expensive to teach people how to use a PC.

3. Some people might have good reason they prefer the Mac, enough to justify the negative externalities on society.

This is a bit of a stretch--it's not that hard to teach people to use a PC and Mac's and PC's have considerable interoperability. You can network them to each other, for instance.

But consider substituting "English" for PC and "Spanish" for Mac. Now how true is this story? Is the US stuck in a suboptimal two-language equilibrium? How large are the negative externalities to learning Spanish? I bet someone has written a paper on this with a wildly inaccurate estimate.

How much can we trust happiness surveys?

Not much when they disagree with your politics, writes Stephanie Coontz:

Wolf is rightly skeptical of the anti-feminist claim that women were happier in the past, pointing out that historical comparisons of reported happiness overlook the ways that women tamped down their expectations when they had few options for challenging unfair relationships at home or at work. A 1962 Gallup poll, for instance, reported that women were the happiest people on earth because, as one housewife reported "a woman needs a master-slave relationship whether it's husband and wife or boss-secretary." Buried in the article was the astonishing fact that almost 90 percent of these self-reported happy housewives hoped their daughters would not follow in their footprints.

There are a lot of problem's with happiness surveys. I'm working with data on life satisfaction for a paper and there are not a lot of conclusions you can draw. My work is largely in studying the biases and unreliability in the data.

But, as the old saying goes, the facts are the facts, so dismissing the evidence on sight isn't a productive strategy. Justin Wolfers, one of the leaders in subjective well-being research, wrote a great post on this topic a few months ago.

Assorted Links

1. Why did everyone jump on Saints bandwagon?

2. Human interest: Saving BlockBuster

3. Ezra Klein on fairness

4. Dept. of Bizarre: Development Edition

5. Testing the Broken Windows Theory

6. Advice for making good slide shows

7. Journalists Need to Learn Statistics: Move to New Jersey and you'll be happy Edition

8. Good commentary on development policy (HT: Aid Watch)

Estimating upper bounds

Slashdot reports that game piracy is not as big a problem as the industry says. What I'm wondering is why anyone would trust an estimate of "lost sales" from the industry?

It's a good analysis but they went too far when they said 10% is an upper bound for how much business is lost. If people who like games the most are the most likely to jailbreak a phone and pirate games, then they are losing a larger proportion of sales than the proportion of people stealing games.

Take this simple numerical example: there are 2 types of gamers. Regular games buy 1 game a year. Hard core gamers by 10 games a year. Say 95% of gamers are regular and 5% are hardcore. Assume 10% of gamers pirate games, 2% being hard core gamers and 8% being regular gamers.

The percentage of lost sales is 1*.08+10*.02 / (1*.95+10*.05) = 19.3%, almost double the 10% upper bound. (These numbers are only for illustrative purposes.)

Assorted Links

1. Charter Cities, because not everyone has heard

2. Insights from Robin Hanson

3. Best markets-in-everything of all time, of all time.

4. Taller buildings = cheaper rent

5. My UROP advisor's TED Talk

6. Sticking it to rich private colleges

7. Thoughts on Think Tanks - good but ignores the variation in the quality of think tank research

Grade inflation

I saw that the NYTimes is taking questions about grade inflation and it got me thinking about how to compare GPAs between schools. A 3.5 at Harvard is not the same as a 3.5 at Pasco-Hernando Community College can't be right, so how do you adjust the GPAs so they are measured on the same scale?

I think one natural thing to do is to decompose the grades into humanities, social sciences, and sciences. We know science and math classes are graded on a worse curve (in general) than humanities and we could gather data on the average difference and correct for it. We'd also like to correct for selection bias but I'll leave that for another post.

After fixing for the composition effect we'd want a way of comparing school Z to school Y. If we assume both schools have roughly normally distributed GPAs and normally distrbuted intelligence, and then use SAT as a proxy for intellect (just because its easy to get those data!) we could put the GPAs on a same scale. Specifically, we can find or estimate the mean and standard deviation for each college's GPA and SAT score. Then we calculate z-score for the GPA and convert that z-score to an SAT.

(GPA_i - mean_gpa,s)/SD_gpa,s = z then (z*SD_sat,s)+mean_sat,s = GPA_equiv,i

where s is an index for the school and i is an index for the individual.

I did this to compare an all-science GPA of 3.9 at the University of Florida to a 3.5 at Harvard, using some, admittedly rough, means and standard deviations. I found that in equivalence terms they are a 1478 and 1512 respectively. So you should probably prefer a 3.5 Harvard grad to a 3.9 Florida grad.

Price Gouging

Earlier this week some water pipe broke in Boston, leaving a few million metro area residents without water. Everyone responded, naturally, by buying bottled water, which created a shortage. In Econ 101 students learn that the natural way to resolve a shortage is to raise the price--then only the people who need the water badly enough to pay the higher price will still be willing to buy it. You continue to raise the price until there is no longer a shortage. This op-ed sums things up nicely. (HT: Greg Mankiw)

Of course, raising the price like that is price gouging which is unfair.

I like this story because it brings up a few important points about economic analysis.

1. Most people don't understand some basic reasoning in economics.

If you don't raise the price for water, then you can end up with people who desperately need it without and those who don't care so much with. That is what we call inefficient (technical term, not in the ordinary sense. Also, if you don't let the price rise, there is no incentive to find more bottled water, say by buying up the stock in New York, transporting it to Boston, and selling it at a mark-up to cover the transport cost.

2. Fairness is a cardinal value in distributing resources

Fairness is very important to people's conception of how we should divide up our resources. Akerlof and Shiller discuss fairness as one of the five Animal Spirits in their recent book.

3. Economists tend to use theories that are based on ranking preferences as if they were direct measures of happiness

This one is a not-quite-dead horse I'm flogging. But it's a fair complaint that doesn't get enough discussion. The fact that an allocation of water is inefficient doesn't mean it doesn't make net happiness greater than some efficient distributions.