[EM] Draft of CVD analysis about IRV vs. Condorcet Voting

RLSuter at aol.com RLSuter at aol.com
Fri Apr 2 20:33:01 PST 2004


I'm posting this to the Election Methods list and BCC'ing it to Rob Richie 
and Terry Bouricus of CVD. I hope they will take a look at the other 
commentaries about the CVD draft on the EM list 
(http://groups.yahoo.com/group/election-methods-list/) that are likely to be posted in the next few days.

In a message dated 4/2/04 3:02:14 PM Central Standard Time, 
election-methods-list at yahoogroups.com writes:

<<  Just became aware of this:
 
 Draft of CVD analysis about IRV vs. Condorcet Voting
 http://groups.yahoo.com/group/instantrunoff/message/1548
 (The message archives are open to everyone) >>


My response to the draft:

This subject needs honest, well-informed, fair-minded arguments, not legal 
brief-type arguments that misrepresent opponents and overstate the positives of 
one's own opinions. The CVD draft, unfortunately, is more like a legal brief.

The first misrepresentation is the draft's description of Condorcet voting as 
"a ranked-choice system favored by some mathematicians." The truth is that 
Condorcet voting is favored not only by "some mathematicians" but by a large 
number of very well-informed and sophisticated political scientists and 
economists and if anything has gained increased favor among voting experts in recent 
years. A much better description is the following one in the Encyclopedia of 
Democracy (1995) by Arend Lijphart, a former president of the American Political 
Science Association and, ironically, a member of CVD's advisory board:

  There is such strong and widespread agreement among
  experts that the Condorcet method is the most accurate
  and fairest majoritarian formula that it has become the
  yardstick against which other formulas are measured.

The draft's misrepresentation of Condorcet is similar to a misrepresentation 
made by one of CVD's chief theoreticians, Douglas Amy, in his book "Behind the 
Ballot Box" (2000). In his brief description of Condorcet voting (p. 187), 
Amy states:

  There is a long history of scientists and mathematicians
  who dabbled in election theory and invented voting systems.
  They include Jean-Charles de Borda, the Marquis de
  Condorcet, and C. L. Dodgson -- better known as Lewis
  Carroll, the author of "Alice's Adventures in Wonderland."

Amy thus begins his analysis of Condorcet voting by describing its author, 
who took his politics so seriously that he lost his life in the French 
revolution (and was one of the earliest advocates of gender equality) and whose 
influence among modern day political scientists is far greater than Amy's ever has 
been or ever will be, as a "dabbler."

Other people on this list are much more qualified to comment on the rest of 
the draft than I am, but what bothers me most about it is it's emphasis on the 
possibility that mediocre candidates who refuse to express strong views will 
have an advantage over other candidates and that Condorcet "punishes candidates 
who take clear stands on controversial issues and rewards candidates who say 
little of substance." This is a major theme of the overall draft, and in my 
view it is extremely theoretical and debatable and probably quite mistaken.

Also, the draft neglects to mention perhaps the single greatest failing of 
IRV, that it undervalues and fails to adequately take into account the 
information provided by voters in their 2nd choice, 3rd choice, etc. rankings. As a 
result, it can easily enable a popular but very divisive candidate in a bitterly 
divided electorate (as the U.S. electorate arguably is today, for example) to 
defeat a strong compromise candidate who is the second choice of a large 
majority of the electorate but the first choice of fewer people than favor two of 
the most divisive candidates. Thus, Condorcet voting is arguably more likely to 
elect candidates that will reduce a society's divisiveness, while IRV is more 
likely to elect candidates that will sustain or even increase a society's 
divisiveness.

-Ralph Suter



More information about the Election-Methods mailing list