Offensive Strategy in IRO

[Hugh Tobin]

Steve Eppley wrote:
[snip]

>But if you're referring to "offensive"
>tactical opportunities in IRO, as I recall they're not a problem.
[snip] 

Consider the recent disclosure that Nixon and Buchanan plotted to
finance an African-American candidate to draw votes from the Democratic
nominee in 1972.  Could this work under IRO?  Well, yes, though it is
not as easy as under plurality and requires some assumptions about
preferences.  Assume most Muskie voters would choose Nixon second and
virtually all Chisholm voters choose Muskie second.  Now if Chisholm
runs third (N45, M30, C25), Nixon has wasted Ross Perot's money
financing her; Muskie wins in IRO.  But what if Nixon and his supporters
are devious enough to throw 10/45 of their first place votes to
Chisholm?  Now we have N35,M30,C35 and Nixon wins the instant run-off.

This is what I meant in prior postings criticizing IRO for creating
incentives to rank one's least favorite first.  If Nixon has a clear
plurality but figures to lose if Muskie runs second, and if Muskie and
Chisholm are reasonably close for second, then some Nixon voters may
well vote for Chisholm even without any coordinated strategy, figuring
the likelihood that enough others will do so to elect Chisholm is
remote. 
But a coordinated strategy involving a select list of voters who will
respond to a signal from the plurality candidate's team on election day
is the most likely scenario.  Remember, Buchanan is still with us.

-- Hugh Tobin