Results for “model this” 2739 found
Has Google peaked?
The economic theory of adverse selection suggests that we should be suspicious when companies go public. Here is a summary of some research on the topic. If your big idea has such a stunning future, why let other people in on the action?
Google.com will be going public, and John Gapper at The Financial Times has his doubts. Are they trying to cash out at their peak? Here is part of his critique:
The more pertinent question is whether its business model will retain the lead. To start with, it can no longer rely on others failing to grasp the importance of search. Algorithmic search engines are tough to design and maintain but others such as Teoma, owned by Ask Jeeves, and Yahoo’s Inktomi are catching up.
So is Microsoft, which is developing an algorithmic search engine that may be launched by spring next year – the likely time of Google’s IPO. By 2006, it will be bundled into the next generation of Windows – Microsoft’s usual tactic when faced with superior technology owned by others.
The biggest uncertainty is whether its focus on internet searches to the exclusion of anything else will remain the best strategy. Although it has clearly been popular so far – Google performs 200m searches a day and is responsible for an estimated 75 per cent of all referrals to websites – it could become an Achilles’ heel. It means that Google has no unique content, and no long-term customer relationship with the individuals who use its technology; it is only as good as its last search. That contrasts with sites that have their own databases and customer networks, such as Yahoo, with its 100m registered users, or Amazon, which holds a mass of data about the products that it sells.
The difficulty for Google will come as rivals combine search with other resources in ways that it will find hard to match. The launch of Amazon’s “Search Inside the Book”, which allows customers to search pages on its database for references and information, is one example of how search technology can be applied to data within internet sites.
Yahoo is augmenting internet search with its own information. Its Yahoo Shopping service not only allows users to search for the cheapest outlet for different models of digital cameras but also combines the results with its own guide to buying cameras, and with user reviews. Google’s own shopping service, known as Froogle, also displays the cheapest prices but looks flat by comparison.
My take: I’m not buying any shares. My understanding of the technical issues is weak. But I understand the theory of adverse selection pretty well.
Why don’t the French work more?
If France were to reduce its effective tax rate on labor income from 60 percent to the U.S. 40 percent rate, the welfare of the French people would increase by 19 percent in terms of lifetime consumption equivalents. This is a large number for a welfare gain. This estimate of the welfare gain takes into consideration the reduction in leisure associated with the change in the tax system and the cost of accumulating capital associated with the higher balanced growth path. The reduction in leisure is from 81.2 hours a week to 75.8 hours, which is a 6.6 percent decline in leisure. I was surprised to find that this large tax rate decrease did not lower tax revenues.
That’s the take of Ed Prescott, one of America’s smartest economists, here is the original research paper.
Consider just how radical the flip has been:
Americans now work 50 percent more than do the Germans, French, and Italians. This was not the case in the early 1970s when the Western Europeans worked more than Americans.
According to Prescott, changes in marginal tax rates are the main reason for this shift. The Prescott cite is from www.2blowhards.com, a never-ending source of interesting material.
My take: Prescott’s critics like to squawk about his oversimplified models and his use of the representative agent construct. Having a background in Austrian economics, I have some sympathy for these criticisms. But on this matter, it is hard to deny that Prescott has nailed it.
We should spend More on Health Care
Tyler appears to be growing more skeptical of the value of health care spending (see his posts here and here). A simple model explains most of what is going on and why he and another of my very smart colleagues Robin Hanson, are wrong. In the graph below spending on health care is on the X axis, health outcomes are on the Y axis. Spending shows diminishing returns. We are currently at point Q on the graph labeled T1 – note that at this point marginal increases in spending have little effect on output (Tyler asks, What margin has low value? Answer: The marginal dollar). Even fairly large increases or decreases in spending will not change outcomes very much given that we are currently at point Q.
Why are we spending so much as to push us into the flat portion of the production function? One reason is that out-of-pocket expenses for medical care are much lower than true costs – we typically are spending someone else’s money. A second reason is that the marginal utility of wealth is low if you are dead so spending on health care near the end of life has unusually low opportunity cost. A third reason may be that various psychological factors make the desire to avoid regret particulary strong for health care, as Tyler speculated earlier.
Although the marginal dollar has low return the value of improvements in medical technology is enormous. These gains are illustrated by the shift from T1 to T2. It has been estimated, for example, that increases in life expectancy from reductions in mortality due to cardiovascular disease over 1970-1990 has been worth over $30 trillion dollars – yes, 30 trillion dollars (for this research see: book, papers, summary). A conservative estimate is that 1/3rd of these improvements in life expectancy were due to better medical technology. One third of the annual benefits is $500 billion – this is much more than total government spending on medical research (the budget of the entire NIH is around 25 billion).
The low value of medical spending at a particular point in time and the high value of medical research over time suggest that we would be much better off if we cut back on medical care spending and devoted the funds to medical research. We should spend less on Medicaid, Medicare, Prescription drug plans etc. and use the savings to better fund the NIH (or other methods of increasing medical research such as prizes etc.)
Bickering with Alex over vouchers
Oh, yes, it is time for that again.
Alex thought that Brad DeLong and I should be cheerier over the prospects for vouchers. My previous post had cited a study of vouchers in Chile, showing no real educational improvement over twenty years.
Like Alex, I am willing to give vouchers a try, but I think he is overselling the idea. Why I am not convinced by Alex’s pep talk?
First, Alex cites a study of Colombian vouchers, which showed improvement from a voucher program. Point granted, but I think that correct conclusion is simply that sometimes vouchers improve schooling, sometimes they don’t. The most convincing Chilean evidence, not cited by Alex, is simply that overall educational performance, on the international scale, did not improve after twenty years of vouchers.
Second, Alex argues that Chile did not have a pure vouchers scheme. Again, point granted, but no implementable vouchers scheme will be pure, let us take this for granted. Have you read about the Washington D.C. voucher proposals, which would force private schools to admit a certain percentage of “voucher students” by lot? Not surprisingly, the good private schools don’t want to participate in the program, if it passes.
Alex overestimates how far the Chilean system deviates from a pure voucher model. True, the Chilean system makes the payment to the private school, rather than to the family. But according to most theories of tax incidence, this should not matter. The private school will lower tuition accordingly, hoping to capture more students, and thus a greater payment from the government.
Overall, what is going on? Education is not just another commodity. Some of it is signaling, in which case subsidizing it doesn’t bring great gains. Another big part of education is selecting peers for our child. A school filled with bad kids is a bad school, whether it is private or public. It is not obvious how much a private school can make bad kids good. As experience in Eastern Europe and around the world shows, certain kinds of education may be a prerequisite for well-functioning markets, the right values don’t follow from markets automatically.
Most generally, educational performance varies with many factors, not all of them depending on the scope of the market. There are, in fact, very many good public schools, whether in the U.S. or abroad. Now Chile is a relatively homogeneous and urbanized country, with 13 million plus inhabitants. The country also has a reputation for discipline, order, and strong family structure. Is it plausible to think that such a region can have reasonably good public schools, so good that vouchers won’t elevate their youth to another level? Yes.
Vouchers would give U.S. urban youths another educational chance, but let us not expect too much from this reform.
New Nobel Prize
I hear it is Robert Engle and Clive Granger, not yet on the major news outlets, more to follow later today.
Addendum: Here is the press release from Stockholm. Here is a short article on cointegration, Granger’s most important contribution. Here is an introduction to ARCH models, a technique pioneered by Engle. Here is Engle’s home page, and Clive Granger’s home page.
My take: Very good picks, economists use their contributions all the time, note that their work is of less interest to the general public than is usually the case.
Spam-onomics
The New York Times has an informative interview with a man who used to earn his living sending “bulk email.” There are a couple of take home points:
1. Spammers don’t make that much money because there is a lot of competition. The price for sending a million e-mails is about $900, and will drop soon. It’s only about $300 for 10,000 e-mails.
2. Spammers cater to other dodgy businesses. Not the kinds of people to be deterred by toothless legislation.
3. The author claims mass faxes were reduced because individuals were allowed to prosecute individual junk faxers for small amounts ($500). Enough to harass junk faxers, but not so large that the plaintiff would have to engage in a lengthy court battle. These smaller fines hurt because most spammers are small time operators with slim profit margins.
The most insightful observation is that legislators have considered both models of controlling spam – prosecute a few large operators for millions in fines, or let citizens go after the small fry in civil suits. The deck is stacked against the second solution – one FTC officer said but that there needs to be “a couple of good hangings.” Conclusion: we probably have the legal and economic tools to curtail unwanted mass emails, but the political process won’t let it happen.
The higher the bill, the lower your tip percentage
Two psychologists studied nearly 1000 tips for restaurants, hair salons and with cab drivers. The larger the bill, the smaller the percentage tip. This is consistent with a reciprocal “payment for service” model. You pay the waiter enough to get the job done, but you don’t feel he has to work much harder to bring you a more expensive entree. Or you might simply be feeling poorer, the larger the bill.
Note that the effect levels off for sums larger than $100. After that point larger bills don’t lead to smaller tips in percentage terms. Servers also get bigger tips when they split the bills for large groups. Read here for more detail. Other research shows that servers get bigger tips if they resemble or can mimic the customer.
Why executives should be paid more than they deserve
If workers are paid their marginal product its difficult to understand why some CEOs are paid such high wages. But think of the CEO’s wage as a prize. Valuable prizes make everyone else work hard in order to become the CEO. With this model, the tournament model (JSTOR) of Lazear and Rosen, it may even make sense that CEO wages go up as profits go down. After all, shouldn’t prizes be set highest when motivation is most required? No doubt, some will see this argument as more proof that economists are just shills for the capitalist class.
How high are transportation costs?
“Cities, Regions and the Decline of Transport Costs”, a 2003 working paper by Edward Glaeser and Janet Kohlhase, provides stimulating reading. They build a model, overturning standard location theory, under the assumption that transportation costs are zero.
Does that sound crazy? They estimate that for machinery, electrical equipment, and transportation equipment, transportation costs are no more than 1.2 percent of the value of the product. 36 percent of all shipments, measured by value, fall into this cost category. Transportation costs for goods have been falling for a long time, and will continue to fall.
It is moving human beings that is expensive, not moving goods. Traffic congestion is an increasing problem. It is now less important to live near natural resources, and more important to live in good weather and under good government. Plus people want to live near other people, leading to greater population concentration in SMSAs. But within metropolitan areas, people are dispersing themselves more, they want to be close but not too close. Don’t buy real estate in Duluth (once a vitally important port) is, I think, the final lesson.