TOEFL, technology, and trade

Some reports are saying that Chinese TOEFL-takers (the Test of English as a Foreign Language) have declined precipitously in recent years. According to the article, at peak, in 1999, 100,000 Chinese students a year took the test, most of whom hoped to study abroad, in particular in the U.S. That number has supposedly declined 90%, with only 10,000 Chinese students taking the TOEFL this year. The article blames post 9-11 visa restrictions.

If true (but see below), the decline has many implications. First, and most importantly, it means far fewer Chinese graduate students. And that, in itself, is a big deal for suffering U.S. universities. Foreign graduate students pay among the highest tuitions on campus, and they make up more than half of many graduate science & technology programs. Less Chinese graduate students means less university income, and less income means commensurate cost cuts — or more income from somewhere else, like tuition increases or goverment transfers (taxes).

Fewer incoming graduate students also means fewer outcoming graduate students. Given the disportionate numbers of such graduates employed at U.S. technology companies, from Intel to Cisco and so on, that is an implicit tax on U.S. technology firms. In effect, they are forced to bid for other students, thus raising the cost of doing business, while reducing the overall depth and quality of their workforce.

Finally, while I said at the outset this decline is thought to have to do with post-9/11 restrictions, any real decline would almost certainly also have to do with the slow but steady maturation of the Chinese economy. They simply don’t have as many students who feel that they have to go elsewhere to succeed — and that would be more noteworthy news for the rest of us than whimsical U.S. visa policies.


[Update] Despite the claims to the contrary I cited above, the statistics from the TOEFL folks themselves don’t bear this decline out. See this chart for what ETS says is the number of people in mainland China taking its TOEFL each year since 1999. Sure doesn’t look like decline to me. It actually looks more like a fairly healthy increase. While there may be regional patterns that are missed in this data, overall it certainly looks like 9/11 visa restrictions haven’t cut the Chinese ardor for taking the TOEFL.

Dean Kamen on patents, innovation and the new-idea glut

Fascinating Gartner interview with dystopically optimistic Segway inventor Dean Kamen. Here is his way of looking at the idea glut:

“When I get a new patent these days, the patent number is 6 million something. When I got my first patents, they were numbered 3 million something. So in my career, the number of patents has doubled, just in the US, and I can assure you, there have not been 3 million innovations since I started working.”

Sex as disruptive technology

Don’t let anyone kid you about how the Internet destroys incumbents’ business models. Case in point: General Media, the publisher of Penthouse magazine. It filed for bankruptcy protection in August of 2003, and it now says it has a reorganization plan and expects to re-emerge in February of 2004.

What did the web do to General Media’s business? Well, the publisher’s revenues dropped by nearly half between 1998 and 2002 — gosh, wasn’t that Internet thing growing like ponzi during said period? — while it racked up $40 million in debt. At the same time, its flagship magazine’s circulation dropped from nearly 1 million in 1998 to about 565,000 last year, almost a 50% drop in four years.

You want disruption? We’re talking destruction.

A telephony inflexion point

And the free-phone beat goes on. Tim Bray discovers the wonders of Internet telephony (sans even would-be service providers like Vonage):

“I can have a free videophone call home, that goes on as long as I need to and nobody’s counting minutes or running up a phone bill. Let’s see; free telephone with video, or pay-for-it telephone with no picture. Costly and voice-only, or free with a picture. I think this is what an inflexion point smells like.”

It is increasingly hard not to imagine how the regional Bell operating companies (RBOCs, in telephony parlance) are not in desperate trouble. Disagree? Try having part of your business that represents 60% of your profits go to zero in five years and see how you fare.

The essence of political economics

From Marginal Revolution:

Political writers have established it as a maxim, that, in contriving any system of government, and fixing the several checks and controls of the constitution, every man ought to be supposed a knave, and to have no other end, in all his actions, than private interest. By this interest we must govern him, and, by means of it, make him, notwithstanding his insatiable avarice and action, cooperate to public good…. It is, therefore, a just political maxim, that every man must be supposed a knave; though, at the same time, it appears somewhat strange, that a maxim should be true in politics which is false in fact.

David Hume, “Of the Independency of Parliament”, in Essays Moral, Political, and Literary. Rpt. Oxford: Oxford University Press, 1963 [1741], pp. 40-47.

Currency quicksand: $12bn (£6.8bn, €9.7bn)

Here is one reason why currencies are consolidating: $12bn (£6.8bn, €9.7bn). The preceding is a snippet from a Financial Times article today about flat-panel displays. Here is the full quote:

In October, Samsung Electronics started constructing a giant $12bn (£6.8bn, €9.7bn) factory in the province to make liquid crystal displays (LCD) for flat screen television sets and monitors. By 2010 the plant is forecast to be the world’s largest producer of flat-panel displays.[Emphasis added]

Notice what happens when your eyes get to the currency info. Your eyes stop, wander around, and generally lose momentum. You try valiantly to figure out which currency is which, but currencies are like quicksand. Which one do you really want to know? And what are all those other funny symbols? Wait, is that the euro symbol, or is that the euro symbol? It all blurs into something like modem line-noise, and you rapidly conclude that maybe you didn’t want to know anyway.

VOIP and the home “hoot and holler”

What happens to distance conversations when you take away the cost constraint? That thought experiment is playing out empirically right now, and I think we’re seeing that people have text/video conferencing wrong. It is not that that instant message and voice-over-IP are merely cheap replacements for phone calls or physical visits. Sure, it is that sort of difference in degree, but it is also a difference in kind.


If you didn’t have to pay, why wouldn’t you leave connections open longer, perhaps for hours? It becomes a sort of personal/professional relationship squawk box, a home “hoot and holler” analogous to what is on the trading floor at brokerage firms. Such firms generally have a persistent audio link among London, New York, San Francisco, and Toronto whereby traders can hear one another working and yammer back and forth about trades and customers. It creates a disconnected connectedness that bridges distances and reduces conversational transaction costs — there is no need to initiate a new conversation when you are always already in one.

On a more personal level, you see this phenomenon in action in John Perry Barlow’s comments today. He talks about using Apple’s iChat AV and leaving the audio connection on after the “conversation” was over. Having done the same thing, it is a little unnerving at first, but the results can be interesting and collaborative.

Fascinating how many people are noodling about this subject. I’ve had a couple of recent conversations with David Akin about iChat, most of which were precipitated by David Pogue’s (excellent) series on video conferencing in his N.Y. Times Circuits newsletter.

Wireless penetration puzzle

Either Canadians responding to a recent survey don’t know what WiFi is, or pollster Decima Research has its data wrong. That is the conclusion I come to after reading David Akin’s precis of a Decima report about wireless “hot spots” in Canada. According to David, the Decima report says that “…one in 10 Canadians — not one in 10 Canadian Internet users, but one in 10 Canadians — have already accessed a wi-fi hotspot”.

As they say in the Valley, horse-hockey. Straight up, less than 10% Canadians have wireless-ready laptops, so the stat is demonstrably implausible. Unless Canadians like to borrow 802.11-enabled laptops from one another just for the frisson of having tried out a WiFi hotspot, something is seriously hosed here. Maybe Canadians misunderstood the question and are including all the times they felt radio waves bouncing off their parka at coffee shops.

Will faux-futures markets fail?

New Yorker writer James Surowiecki asks the right question concerning futures markets in public events and the like. While such markets are oft-criticized for being illiquid and being dens for (largely nonexistent) abritrageurs, Surowiecki points out that doesn’t invalidate the whole idea of betting on everything from terrorism, to the Lord of the Rings’ take last weekend. He writes, however:

“It’s perfectly possible that there is something about these markets — other than a lack of liquidity, which I think is a huge problem — that will guarantee their uselessness in contrast to other exchanges, but I’d like a coherent explanation of what that something is.”

I don’t think there is anything — other than lack of liquidity — that will make it tough for these markets to work. But that’s a biggie. Bifurcated markets are going to be a problem for some time, doubly so given that trade-driven interest in forcing markets to consolidate is nowhere near as high as it is, for example, in share markets like the NYSE and Nasdaq.

Likert scales and the terrorism threat

The verbal gymnastics that Tom Ridge is going through in explaining the weekend increase in the terrorism threat here in the U.S. should be no surprise to sociologists, statisticians, and other users of Likert scales (named after Rensis Likert). There are five points — green, blue, yellow, orange, and red — in the Homeland Security scale, not unlike having a typical five-point Likert scale that goes from Very Unhappy, through Neutral, to Very Happy. The tendency is to sit at the middle-position in such scales — it just seems safer (no pun intended).

And that is what Ridge has done, with the index sitting at Yellow for all but eight days since being created in 2002. We have never seen green, and we have never seen red. The result, of course, is that there is no way for citizens and officials to differentiate among raised threat indicators — why is this increase to orange different than a prior increase to orange? Better would be to lower to blue, and move the indicator around more regularly. At least people would get a sense that there are terrorism threats, and then there are terrorism threats.