[EM] What is PR?

Steve Eppley seppley at alumni.caltech.edu
Sun Feb 25 11:01:22 PST 1996

Marking up the CVD faq revised by Rob...
First, my general comments:

CVD should be about more than PR.  Single-winner reform is always
given short shrift, even though every 4 years there's a great cry in
the land about the lesser-of-evils and spoiler dilemmas in the
Presidential and other single-winner elections.  Instead of the "PR
movement", it should be the Electoral Reform movement.

The winner-takes-all phrase should be routinely accompanied by the 
vote-for-only-one phrase.  Denying the voters a way to express their 
true will is another major factor underlying the two-party system.

- - -

>Under PR, representatives are elected from multi-seat districts

It's not obvious that "multi-seat districts" includes at-large PR,
in which there is one large "district".  I think at-large PR is better 
thought of as districtless, not one district.

>In contrast, in the United States we use "winner-take-all" single
>seat districts, where votes going to a losing candidate are wasted,
>even if that candidate garners 49.9% of the vote. This leaves
>significant blocs of voters unrepresented. Voters sense this, and so
>often we do not vote for a candidate we like, but rather the one who
>realistically stands the best chance of winning -- the "lesser of
>two evils." Or, all too often, we don't bother to vote at all.

This paragraph is so important it should be expanded, with clear

>   * more diverse representation
>   * cleaner campaigns run on the issues, not mud-slinging
>   * reduced effects of big money

* increased voter turnout
* satifaction that one is represented by someone with similar views, 
  instead of the turkey elected by the votes of others.
* dynamic coalitions formed around issues, instead of big-tent 
  electoral coalitions just to win power

>A parliamentary system is a type of governmental system, while PR is
>a type of voting/electoral system. One is about the structure of
>government, the other about how votes are counted.

Deceptive: PR *systems* also specify structure.  What needs to be
pointed out is that in parliamentary systems, the executive branch is
chosen by the legislative branch.  Even with PR, this reproduces the
winner-takes-all need for big-tent "executive electoral" coalitions
of parties (or legislators, in non-party-centric forms of PR) to
gain executive power.

>   * Mixed Member System (MM) -- This PR hybrid elects half the
>     legislature from single-seat, "winner-take-all" districts and 
>     the other half by the List System. MM smoothly combines 
>     geographic, ideological and proportional representation.

Except it's less proportional than other forms of PR.  How about a
paragraph attempting to explain the alleged value of geographic
organization, so the assumptions can be made explicit and exposed if

>   * Preference Voting (PV)

Since it's also commonly called STV, this alias should be included.

* Proxy PR
* Direct Democracy
Both of these are proportional systems.  It's wrong to omit mention 
of them.

>What About The President? We Can't Divide Up The Presidency, Can You?

We or You?

>No, we can't.

Sure we can.  A small committee could govern, and emergencies could
be handled by one of them, maybe a rotating chairperson.  The
committee could be elected in many ways, including PR.  The only
justification for having one person in charge is speed of response
to emergencies.

>However, there are much better ways for electing officials
>such as president, mayor, or governor than what we use today:

Single-winner legislative districts offices like Senators
Representative, Assemblyperson, etc. can also be elected with better
voting systems than plurality.

>   * Majority preference voting (MPV) -- related to PV.

This description is inadequate.  How about:
-- a simple form of PV (see the PV prop rep description above)

>   * Condorcet's Method -- Like PV and MPV, the voter simply ranks
>     candidates in an order of preference (ex. 1. Perot 2. Clinton 
>     3. Bush). Unlike MPV though, several two-way races are simulated 

Not "several".  All the pairings are checked, since the ranked
ballots provide enough info to determine which of each pair is
preferred by the voters.

Include the alias Pairwise.

Maybe add:
* Condorcet+Approval -- Like Condorcet's method, but each voter can 
  also indicate which candidates are disapproved.  Candidates 
  disapproved by a majority are disqualified from winning.

>All of these methods give voters a greater voice in how their vote
>is used, and promote compromise among all voters. Our current
>winner-take-all system promotes candidates who blame those who would

If this paragraph is about single-winner reform, than the problem
here isn't winner-take-all, it's vote-for-only-one.

>Could PR Help Break The Political Impasse In The U.S. Over
>Important Issues Like Health Care?
>Yes, it could. PR allows small parties to be a credible alternative to
>voters, giving them a national audience for their views to advance new

Another important point:  It's the big-tent electoral coalitions of
the winner-take-all system that block change and lead to false
claims of mandates.  Example: if the Republican big-tent coalition
of wealthy, fundamentalists, and Libertarians split into several
parties, the fundy party might be interested in a health care
coalition with progressive parties.  The wealth and Lib parties
wouldn't be interested in prayer in schools.  Coalitions would be 
based on issues, not on getting elected.

>But I Like Having A Representative From My Own District. Won't I
>Lose Out Without It?
>A representative from your own district is nice,

Why?  I don't take this for granted.  It produces geographic-based 
pork, and representatives of other districts don't bother themselves 
with the concerns of non-constituents.

>                                                 but with
>"winner-take-all" there's a good chance you didn't vote for that
>representative. In the 1994 Congressional elections, only 21% of
>eligible voters helped elect someone. Under PR, you will have, not
>one, but several representatives from a larger district.

Misleading, and significantly undermines the document's veracity. 
Only true in some PR forms, not all of them.  

Is CVD pursuing a particular type of PR?

>And there is a much greater likelihood that at least one of those
>reps will be someone you voted for.

This would be expressed better in absolute terms, not probabilistic:
More will be represented by someone they did vote for.  Fewer people
will be represented--poorly--by someone they didn't vote for.

>                      In South Africa's 1994 PR elections, 86%
>of eligible voters helped elect someone.

Like the above, "help elect someone" presumes more knowledge about 
the mechanics of prop rep than the target audience will have.  
Better: The winners received 86% of the votes; in the U.S. the 
winners received only 55% (or whatever stats are accurate; I made 
this up).

In Proxy PR it would be 100%.  With list pr, threshold=0, it would
be 100%.  Somewhere, the pros and cons of threshold size need to be 
explained.  Also, why not cap wasted votes instead of using a fixed 
threshold?  That's a more direct and robust way of gagging the 
fringe, for any fringe-gaggers out there.

>Also, the MM form of PR used by Germany can give voters the
>benefits of both: a representative from your district, as well 
>as a legislature that proportionally reflects the electorate.

Hey, talk straight.  It's a compromise, not the benefits of both.

>PR doesn't base representation so much on geography but on political
>viewpoint. [snip]
>geographical representation can no longer ensure fair political
>representation for all voters and all political perspectives.

Right.  (And it never could, but in horse and buggy days what could
do better?)  So deemphasize MMP.  Don't say MMP provides extra
benefits when it really detracts from proportionate power.

>What's Wrong With Only Two Parties?

I'd put this section nearer the beginning, combining it with the
winner-takes-all critique at the beginning.  Present an analysis of
a problem before offering proposals.

Lacks an explanation of how the vote-for-only-one winner-takes-
district system leads to two dominant big-tent parties.  A major
omission!  Readers know there are more than two parties, and
may presume these "unpopular" parties don't deserve more power: if
they deserved it, they'd have won it.

>Two parties limit the voters' choices. U.S. citizens would never
>accept an economic system that allowed us to buy cars from only two
>companies, or to choose from only two airlines. Why then, should we
>have to settle for just two options in politics? It's no wonder
>such a large portion of the U.S. electorate decides not to
>participate. They're not buying what the two parties are selling!

Sure they are.  They're voting for them and not the other parties.
Document needs to explain why this happens, even though it's not what 
the people really want.

>But with PR you don't have to come in first to win seats.

[Now we're back to the solution section of the document.]

>Could PR Help In Voting Rights Cases?
>Absolutely. With PR, you actually need less votes to gain a seat
>than in the winner-take-all system, and you can gather these votes
>from a larger area. This makes it easier for racial or political
>minority perspectives to win seats, WITHOUT having to gerrymander

Drop the "racial or".  If a racial minority shares a political
perspective, it will win seats for that perspective, not for that
race.  PR is not about racial quotas or affirmative action, but you
can bet it will be attacked on those grounds.  The document needs to
be clear *how* the race- and gender-neutrality of PR can alleviate

>Does PR Affect The Election of Women?
>Yes, very much so. [snip]

Explain how.  Don't just give stats.
Combine this with the section above.

>So How Do We Change From "Winner-Take-All" To PR?
>In many states it is possible to convert to PR simply by changing
>applicable laws. Amendments to the U.S. Constitution are not

I'd put that second sentence first.

>If the political will could be mobilized, it is possible to convert
>immediately to a system of proportional representation for electing
>representatives to city councils, state legislatures, and even the
>U.S. House of Representatives.

Mention that an act of Congress is needed to undo the portion of the
Voting Rights act, which disallowed prop rep for the House.

A constitutional amendment would be required to make the House
nationally at-large, but lacking this amendment representatives
could still be proportionally chosen in 50 state-sized "districts".

>U.S. Senators could be elected by Majority Preference Voting (MPV),
>giving voters more choice.

MPV?  Is that single-winner method endorsed by CVD?

Mention that a constitutional amendment would be required to change
the "2 Senators per state" nonproportionality.  Is this a long-term 
goal of CVD?

>As a bonus, PR would spare states the torment of legislative
>redistricting, an arduous, bitter and partisan gerrymandering

Sometimes the document speaks of PR as having districts, sometimes 
PR is at-large.  How slippery.

>Where Can I Learn More About Proportional Representation?
>Here's a reading list we've assembled about PR:

maillists.  websites.

>How Can I Get Involved In the Proportional Representation Movement?

Ahem.  The "Electoral Reform" movement.

>You can get involved by becoming a member of the Center for Voting and

Ah, the pitch for money.  This belongs elsewhere.

More information about the Election-Methods mailing list