One of the authors has several other papers on his site [link|http://gsbwww.uchicago.edu/fac/steven.davis/research/|here]. One that caught my eye was [link|http://gsbwww.uchicago.edu/fac/steven.davis/research/Evolution%20of%20the%20PC%20Operating%20System%20(with%20Jack%20MacCrisken%20and%20Kevin%20Murphy),%20June%201999.PDF|this] one (81 page .pdf) on the evolution of PC OSes from 1999. It might as well have been written by Microsoft:

p.75-76:
The integration of new features and functions into the PC operating system benefits PC users and facilitates innovation in computer applications and technologies through several channels. OS integration makes PCs easier to use and thereby opens the door to new, non-integrated PC products. Integration also simplifies the task of software development, which allows applications developers to concentrate on their areas of expertise and leads to an increase in the number and variety of specialized software applications. OS integration helps promote a standard computing environment and reduce customer support costs. The bundling of software applications and utilities with OS software leads to their wider and cheaper distribution among PC users. Bundling also adds to consumer welfare and economic efficiency by stimulating the development of software applications that would otherwise be unprofitable.

These positive consequences of OS integration generally serve to enlarge the market for PC products, in particular, and computer products, in general. Because scale economies -- declining average costs in development, production and distribution \ufffd are important in the computer industry, these market-enlarging effects of OS integration imply that the full benefits of integration exceed the direct benefits.


An economic study that doesn't consider the impact of bundling on competitive products seems to me to be flawed from the outset.

Also, I think their analysis of the costs of containment are suspect. E.g. [link|http://www.historyguy.com/no-fly_zone_war.html|The History Guy] says:

Since American and British forces carried out Operation Desert Fox in December 1998 against Iraq, this "forgotten" war in the Middle East has only become more intense. According to the New York Times in an article on August 13, 1999, American and British forces have escalated the continuing war against Saddam Hussein and Iraq. Since the beginning of 1999 through August 1999, Allied pilots launched over 1,100 missiles against 359 Iraqi targets. That number equals nearly three times the amount of ordnance used in the four-day Desert Fox strike. Also, the pilots in the Iraq War have flown two-thirds the number of missions as NATO pilots in the 1999 Kosovo War. By all accounts, Iraqi forces continue to target their radar and fire missiles at Allied warplanes despite the punishment inflicted from the air. The estimated, unofficial cost of this war to U.S. and British taxpayers is around $1 billion per year. As of August 1999, over 200 military planes, 19 naval ships and 22,000 American military personnel are committed to enforcing the "no-fly zones" and to fighting Iraq. In addition, reports indicate that the death rate for small children has doubled in Iraq over the past decade. These child deaths are attributed to the continuing war and economic sanctions on Iraq and Saddam Hussein\ufffds unwillingness to live up to the 1991 cease-fire agreement.


Emphasis added.

I vaguely recall posting something here earlier (from the NYTimes?) that said the cost was ~ $2B/year to enforce the no-fly zones, but I can't find that now. I can believe another factor of 2 or 3 (meaning 2-3 $B/yr), but not a factor of 19.

FWIW.

Cheers,
Scott.