>Doug, do you by any chance have a link to the Pew center polls?
The front pages is <http://www.people-press.org/>. The NY Times figures seem to be drawn from an April poll, reported at <http://www.people-press.org/kosovorpt.htm>. A more recent poll <http://www.people-press.org/may99rpt1.htm> shows a less callous U.S. public:
<quote> Worries Change
At the same time, public worries about the fate of U.S. troops in Kosovo have declined in recent weeks. Some 56% of Americans say they are very worried that U.S. troops might suffer casualties, down from 66% in April. Even fewer (49%) are very worried that U.S. troops could be involved in Kosovo for a long time, down from 63%.
The public now worries as much about civilian victims in Yugoslavia as about U.S. troops. Some 50% of Americans say they are very worried that innocent people are being hurt or killed by the air strikes. And 37% are very worried about Serbian civilians, a figure that is unchanged from April (40%).
The drop in support for the air strikes has been sharpest among those most concerned about these civilian casualties. In April, people who were very worried about Serbian civilians supported the air strikes, 58%-33%. Today, those very worried about Serbian civilians oppose the air strikes, 43%-54%. People who express less concern about civilians continue to support the strikes. The drop in approval of the way Clinton is handling foreign affairs is also especially large among those most concerned about civilian victims in Yugoslavia. </quote>
With polling numbers like that, I guess the war just *had* to stop.
There's also an interesting short analysis of the effects of question wording on poll outcomes <http://www.people-press.org/99watch2.htm>:
<quote> Polls on the use of ground troops in Kosovo remain hard to read, because small differences in question wording lead to significant differences in results. Majorities support the use of ground forces in polls that ask about deployment "to end the conflict in Kosovo" (ABC). When this phrase is not in the question, national surveys find majority opposition (Gallup, NBC/Wall Street Journal, Los Angeles Times).
A split sample experiment by the Pew Research Center tested the impact of this phrase. Half of an April 15-18, 1999 sample divided evenly (47% favor and 48% opposed) when asked about sending ground troops "if air strikes do not stop Serbian attacks there." When the other half sample was asked the same question reinforced with the phrase "to try to end the conflict," it found a 51%-42% majority favoring the proposal.
Differences are even more dramatic when the public is asked about the use of ground troops as "peacekeepers," rather than peace makers. A Los Angeles Times poll found 68% of the public favoring the use of ground troops once peace is achieved.
When Americans are asked about the ongoing air campaign -- a real rather than hypothetical issue -- there is generally majority support, with some growth in support during the first few weeks of the air campaign. Majorities have consistently supported the air strikes, compared to the 29% to 56% variation in support for ground troops. </quote>
Obviously this is of more than technical interest to pollsters - it's also of interest to politicians marketing policy decisions.
Doug