On 29 June, 1997 1:41 PM, David Bateson [SMTP:major@yoyo.cc.monash.edu.au] 
wrote:

Policies of this sort have been used intermittently throughout history,
the Romans regularly executed captured leaders of broken nations,
so did the mongols.  It didn't stop people trying to fight them.

I feel that this is a little different from those times. What I'm suggesting is 
more of a defensive move. It's not like you're conquering a nation and then 
dispatching its leaders. It's more like tell other world leaders that attacking 
you will lead to their deaths. Besides, broadcasting the execution live might 
provide useful discouragement to to other who might attempt the same.

In the modern world it *might* lower the number of wars, but probably
not.  Most despots would gamble on being able to do a "vietnam" and
outlast the US

The US has learned a lot about war since Vietnam, as evidenced by Desert Storm.

or more likely a "Pol Pot" and go into a comfortable
exile in a friendly nation if the tide starts to turn against you.

If I read the novel correctly, that doesn't matter. His death will be 
delivered, no matter where, not matter how long it takes.

How does this differ from the cold war?  The obvious response is for the
other nation to seek the protection of people like Russia and China,

But, when you seek a protector, you must be sure he can and will protect you 
first. This isn't like a simple trade agreement, where you can argue if the 
other party doesn't keep his promises. With this protection, you're likely to 
be dead if your protector doesn't come through for you. So who would you trust 
to provide it?

If the US threatens more
or perhaps plead US bullying
at the United Nations.

I wouldn't call it "bullying". It's more like, "We don't like the way you run 
your country, we won't protect you anymore."

Simple.  You stage provocations against the target nation, until their
government is forced to attack you to retain credibility.

The loophole in THAT, is that it assumes that the intelligence assets of TWO 
countries (the target and the protector) are unable detect the subversion. To 
implement this policy, I see that a nation must have a very credible 
intelligence agency to root out subversion.

Or perhaps you could just infiltrate guerillas and then deny that you are 
supporting the guerillas, or  perhaps you simply claim to be supporting a 

local >guerilla campaign

But supporting or infiltrating guerillas is useful only if you are planning to 
use it as a prelude to invading the target nation or controlling it as a 
puppet. Invasion is out because its a clear sign that you declared war on a 
protected nation- and get's your head blown off. You can use the gurrellas as a 
puppet only if they overthrow the rulling government. But the protector who is 
strong enough to deliver its threat of death is surely strong enough to put the 
legitement government back in power.

There are good reasons the US has a public stance against trying
to kill head's of state - and equally good reasons why they occasionally
ignore it.

Which is why the best time to implement this policy would be after some nutcase 
was foolish enough to directly attack the US as in the case of Executive 
Orders.