Commentary on Sir Michael Dummett's book on Principles of Electoral Reform

Sir Michael Dummett is a well respected British philosopher who has published influential articles on the theory of voting.  He has written an excellent book on "Principles of Electoral Reform" (Oxford University Press, 1997) which ought to be a must read for anyone wanting to make an intelligent decision on the upcoming British referendum on introducing the alternative vote into British parliamentary elections.

For those unable to read Sir Michael's book because of lack of availability or other considerations, there are two quotes from the book which I want to bring to their attention along with two issues outlined below.

First, in his book Sir Michael makes the following observation which, although stated in a slightly different context, is certainly relevant in judging some of the support in favor of the upcoming referendum on the alternative vote, or IRV in American parlance.

    "the [UK] Electoral Reform Society, in particular, is not
    a neutral body impartially investigating rival systems and
    the theory of voting in general, but an organization devoted
    to disseminating propaganda..." (p.91)

In my judgment, this would also be a very apt description of the American Center for Voting and Democracy (FairVote) which has had over the years a close relationship with the UK's Electoral Reform Society.
Both societies have pushed for the introduction of instant runoff voting, the alternative vote in British parlance, into our voting systems to open them up to third party participation.

The problem with this in terms of opening up our political processes to third parties is that they do not tell you that the system empirically has done little in this regard.  When I was doing the research for the first chapter of my book on "Third Parties and Voting
Reform: The American Dilemma" (published on this website), I looked at the Australian data for recent elections and it was clear that this is the case.  If you look at Gavin Thompson's article in the September
2010 issue of Significance (published by the Royal Statistical
Society) titled "Keeping things in proportion: how can voting systems be fairer?" he points out that, "The Australian House of Representatives has been elected using the alternative vote since 1918.  Electoral outcomes have proved only slightly more proportional than in the UK..." (p.130)

Second, proponents of the alternative vote do not seem to recognize that the system suffers from a serious behavioral characteristic--in technical terms its lack of monotonicity.  When a voting system like IRV, or the alternative vote, is non-monotonic "a shift in public opinion toward a candidate can cause the candidate to lose, and a shift of public opinion away from a candidate can cause the candidate to win."  The upshot of this that in casting their votes, voters cannot know for sure whether they are helping or hurting those they wish to support.  Even first past the post, or plurality voting, with all of its flaws does not suffer this defect .  Casting a vote under first past the post in favor of a candidate set to win will always help that candidate and taking votes away from a candidate set to lose will always assure that that candidate loses under FPTP.

Proponents of the alternative vote, or IRV, like those associated with the UK's Electoral Reform Society and the American Center for Voting and Democracy say that the problem is much too rare to worry about in real elections.  On this score it is useful to cite a comment that Dummett relates in his book from Charles Dodgson (Lewis Carroll), a mathematician who did widely regarded work on voting theory.  The comment by Dodgson is as follows:

    I am quite prepared to be told with regard to the
    cases I have here proposed, as I have already been
    told with regard to others, 'Oh that is an extreme
    case: it could never really happen!'  Now I have
    observed that this answer is always given instantly,
    with perfect confidence, and without any examination
    of the details of the proposed case.   It must
    therefore rest on some general principle: the mental
    process being probably something like this-- 'I have
    formed a theory.  This case contradicts my theory.
    Therefore this is an extreme case, and would never
    occur in practice.'  (p. 149)

One person responding to my June 22, 2010 Comment is Free Guardian piece on this asserted that this problem would only occur once in a century!  Unfortunately the evidence simply does not bear this out.
On this score it is useful to cite a statement that Anthony Gottlieb made in his July 26, 2010 article in the New Yorker on voting systems.

    Supporters of instant-runoff voting
    say that the problem is much too rare
    to worry about in real elections, but
    recent work by [Professor] Robert
    Norman, a mathematician at Dartmouth,
    suggests otherwise.  By Norman's
    calculations, it would happen in one in
    five close contests among three candidates
    who each have between twenty-five and
    forty percent of first-preference votes.
    With larger numbers of candidates, it
    would happen even more often.  It's
    rarely possible to tell whether past
    instant-runoff elections have gone
    topsy-turvy in this way, because full
    ballot data aren't usually published.
    But in Burlington's [VT] 2006 and 2009
    mayoral elections, the data were published,
    and the 2009 election did go topsy-turvy.

IF WE WERE TO INTRODUCE INSTANT
RUNOFF VOTING, OR THE ALTERNATIVE
VOTE, IN AMERICAN OR BRITISH ELECTIONS,
IT SHOULD BE CLEAR FROM THE ABOVE
THAT THIS WOULD BE THE EQUIVALENT
 OF MANDATING UNDER OUR CURRENT
SYSTEM OF PLURALITY VOTING, VOTE
STEALING IN SOMETHING LIKE A FIFTH
OF OUR ELECTIONS.

Those seriously interested in voting reform in the UK would do themselves a favor by reading and bringing to the attention of others the book by Sir Michael Dummett, "Principles of Electoral Reform"
(Oxford University Press) a copy of which I sent to Gordon Brown to bring it to his attention.

                                John Howard Wilhelm, Ph.D.,
                                Economics
                                Ann Arbor, Michigan USA