In my book I
take the position that a bipartite voting system is necessary to get a good
sense of what the people in any group want. And I offer up something like the Mixed-Member Proportional Representation system used in New Zealand and elsewhere. It is a
bit different from MMP, however, since I specifically call for single-member elections
using Approval Voting and proportional representation using the Single
Non-Transferable VOTE (SNTV). I don’t discuss to what extent (if any)
elections utilizing my proposals should be partisan or non-partisan, however, and to the extent that strong parties
are making exclusive lists of candidates in the PR elections, SNTV voting--where everyone votes for one favorite candidate--becomes equivalent to an open party
list system in which the parties set forth the only candidates that anyone may
vote for. I am agnostic on this matter, but have been thinking about it quite a
bit lately. I therefore put these musings—which are a bit out of my wheelhouse-- on this blog with the hope that those with more knowledge in this area may comment and
enlighten me.
“Parties are intermediaries that connect the
public and the government. Parties also aggregate these diverse
interests into a relatively cohesive, if typically compromise, platform, and
they articulate these varied interests by representing them in government. The
result, in this view, is that partis parlay those compromise positions into policy
outcomes, and so they—a ruling if nonhomogeneous and shifting government
majority—can be held accountable to the public in subsequent elections. The
diversity of actors in the party lead to an equally diverse set of party
arrangements….These diverse structures make possible the key concepts of the
party in this view: interest articulation and aggregation and electoral
accountability.
This idea is fleshed out a bit in the paper “A Theory of Political Parties:
Groups, Policy Demands and Nominations in American Politics” (2012), by Kathleen Bawn, et al. The
authors there consider a society in which there are various interests such as those who want to protect local wool production, those who want to spend money
on school buildings and teachers, those who want to enforce blue laws, and several other groups—including some that coalesce (like of low-taxers, consumer groups or bar owners) specifically to
defeat the other groups. Then the authors try to make the case that without deal-making
between these interest groups, nothing like a coherent public policy is likely
to result. Obviously, their little society greatly simplifies the real world. I
believe, however, that it is still too complex to fathom. In this entry, I will
try to provide an even simpler scenario and attach a level of quantitative support to each
interest in order to see whether it’s really the case that every sort of voting
scheme requires a fairly robust party system in order for a coherent policy
that is responsive to the desires of the voting public to be produced.
Assumptions:
The jurisdiction has a population of 1,000,000
and has three seats for representatives.
500,000 are generally uninterested in matters
of public policy or “politics.” The remaining 500,000 do have first priorities they
want their representatives to focus on and may also have second priorities, but
no one has any public matters they care about, beyond a second priority.
105,000 are Pro-Choice (These care most about a woman's right to have an abortion without interference)
100,000 are Protectionists (These care most about putting tariffs on goods coming from outside this jurisdiction)
100,000 are Laborites (These care most about union rights and raising minimum wages)
75,000 are Anti-Protectionists (These care most about preventing tariffs)
75,000 are Anti-Laborites (These care most about stopping unions and preventing minimum wage hikes)
45,000 are Anti-choice (These care most about the right to life of unborn prenates)
Of
the 105,000 Pro-choice Advocates:
21,000 are secondarily pro-protection
21,000 are secondarily anti-protection
21,000 are secondarily pro-labor
21,000 are secondarily anti-labor
21,000 have no secondary public interest
Of
the 100,000 Protectionists:
20,000 are secondarily pro-labor
20,000 are secondarily anti-labor
20,000 are secondarily pro-choice
20,000 are secondarily anti-choice
20,000 have no secondary public interest
Of
the 100,000 Laborites:
20,000 are secondarily pro-protection
20,000 are secondarily anti-protection
20,000 are secondarily pro-choice
20,000 are secondarily anti-choice
20,000 have no secondary public interest
Of
the 75,000 Anti-Protectionists:
25,000 are secondarily anti-labor
25,000 are secondarily pro-labor
10,000 are secondarily pro-choice
10,000 are secondarily anti-choice
5,000 have no secondary public interest
Of
the 75,000 Anti-Laborites
25,000 are secondarily anti-protection
25,000 are secondarily pro-protection
10,000 are secondarily pro-choice
10,000 are secondarily anti-choice
5,000 have no secondary public interest
Of
the 45,000 Anti-Choice Advocates:
10,000 are secondarily pro-protection
10,000 are secondarily anti-protection
10,000 are secondarily pro-labor
10,000 are secondarily anti-labor
5,000 have no secondary public interest
************************************************************************************
Let us start by assuming that the
electorate will vote for all and only those candidates who back what is
most important to them (this assumption will be changed below). If the three candidates for the three seats in this
imaginary district are to be elected via a system that both (i) allows voters
to vote for as many candidates as they like, and (ii) gives seats to the three
candidates receiving the most votes, then Pro-choicers (with their 105,000
backers will win all the seats. If electors may vote for no more than three
candidates, the Pro-choicers will be certain to win all three seats only if
they put up no more than three candidates. Obviously, one way of assuring that
result in an election of this type is for the pro-choice advocates to organize
a party with the power to determine which candidates with its priorities may
pull papers. If those with other first priorities understand the situation,
they may try to arrange a merger with any pro-choice party that forms, in order
to pick up one seat by a member with a different first priority who favors choice
as a second priority, or to at least make sure that at least two of the seats
are taken by candidates who share their views as a second priority.
In a system where voters may pick only
one favorite (with the top three vote-getters winning seats), it will again be
difficult without the intervention of strong parties to ensure either the victory of more than one representative with any
particular first priority, or even the victory of more than one representative having
either a first or second priority of any one (or two) particular interests. Securing
such results would seem to require the intervention of entities making possibly quite complicated deals involving
both what members may and may not publicly support and who will be allowed to run for office.
What happens if we vary our assumption requiring that each member of the electorate will vote for all and only those candidates
who back what is most important to her by assuming instead that each elector
will vote for all the candidates of whom she minimally approves. Now, everyone will
vote for all candidates espousing EITHER their first or second priority. Would we still need parties? Given this changed assumption, if electors may vote for as many candidates as they want, and
everyone who is interested in politics goes to the polls, we can expect the
following results: 1 Pro-choice winner, 1 Laborite winner, and 1 Protectionist
winner. So long as each interest group has at least one candidate in the race sharing its first priority, it won’t
matter how many additional candidates having that first priority also run. For simplicity then, let us just assume that each of the groups listed above has only one candidate with that first priority on the ballot.
Approval Election Results
The Pro-choice candidate will
receive 105K + 20K + 20K +10K + 10K + 165,000 votes
The Protectionist candidate will
receive 100K + 21K + 20K +25K + 10K = 176,000 votes
The Laborite candidate will receive
100K + 21K + 20K + 25K + 10K = 176,000 votes
The Anti-Protectionist candidate
will receive 75K + 21K + 20K + 25K + 10K = 151,000 votes
The Anti-Laborite candidate will receive
75K + 21K + 20K +25K + 10K = 151,000 votes
The Anti-Choice candidate will receive
45K + 20K + 20K + 10K + 10K = 105,000 votes
Will we therefore have a government that enacts Pro-choice, Protectionist, Laborite policies? Not necessarily. In spite of the first priorities of
these three candidates, we cannot assume that the policies created by the three representatives will end up being Pro-choice, Protectionist, and Laborite. Why not? Without party intervention, the second priorities
of the winners will be randomly distributed, so we cannot simply assume that,
e.g., the Pro-choice winner will be amenable to labor or protection proposals. Similarly,
we cannot simply assume that the Protectionist representative will be pro-choice
or pro-labor. Thus, if the three-member legislative/executive representatives
require a majority to do anything, they may be unable to go forward on any
front. Alternatively, one (or two) of the voter-supported positions may
move forward…but there may be no way to tell which one(s) ex ante. Perhaps
detailed polling and interviews with all the candidates would be helpful here,
but it seems they'd be so only if these candidates are completely forthcoming—and it’s not clear what their
interest would be in getting into secondary matters if they don’t have to, since such disclosures may hurt their electoral chances.
This quite simple scenario seems to me to suggest that with no pre-election coordination, it may well occur that there will be no
movement on any issue post-election. But I do not know how or whether the coordination required for program enactments can occur absent the construction of parties and the subsequent presentation of
party candidates. Deal-making of the required sort would seem quite difficult to obtain in any setting in which interest groups--parties--are not important players. It thus seems to me that if one believes (as I do) that democratic governments must be authentically responsive to the sort of electorate imagined in this example, a relatively
strong party system may be required.
I'm not sure about this, though. And, again, I hope those with more experience in this field will comment and correct me where they believe I need it!
No comments:
Post a Comment