The casualty-aversion paper mentioned on Orbat.com is, imho, right on about the public willingness to shed blood for a cause. Of especial note is this paragraph, shamelessly snagged right out of the article:

MASS VERSUS ELITE OPINION

The poll upon which analysts Peter Feaver and Christopher Gelpi based their assertion of the relative willingness of the mass public to countenance casualties was conducted between September 1998 and June 1999. It addressed hypothetical missions to "stabilize a democratic government in Congo," "prevent Iraq from obtaining weapons of mass destruction," and "defend Taiwan against invasion by China." In each case the public identified a higher level of acceptable casualties than did samples of elite military leaders and civilian elite leaders. Significantly, in each case the number of acceptable casualties to the public was in the thousands. The question even included a description of how many casualties the U.S. had actually suffered in Somalia (forty-three), the Gulf War (383), Korea (approximately fifty-four thousand), Vietnam (approximately fifty-eight thousand) and World War II (approximately four hundred thousand). Results:

Highest number of American military deaths acceptable to . . .		Military Elite	Civilian Elite	Mass Public
	Stabilize democratic government in Congo	284	484	6,861
	Prevent Iraq from obtaining WMD	6,016	19,045	29,853
	Defend Taiwan from Chinese invasion	17,425	17,554	20,172

Polling sample: 623 military officers, 683 nonveteran civilian elites, 1,001 adults from the general public.