Proportional Representation for MCSG

from the Macalester Public Knowledge Base

Jump to: navigation, search

The proposal to use at-large, Proportional Representation (PR) is part of a ripple of activity aimed at democratizing Macalester College Student Government that includes Common Platform, the successful implementation of Instant-Runoff Voting, proposals to create Autonomous Student Org Budgets, and more.

Not much activity was put into the PR proposal, though a Hegemon article was written and discussion taken up with MCSG's Constitutional Reform Committee (they were not interested).

The main proponent of this proposal was Jesse Mortenson, with the support of Matt Hauck and other Mac Greens.

Contents

[edit] Purpose and Grounds

By Jesse Mortenson, '05

My basic assumptions are:

  • Any community will include individuals with different and conflicting political desires (opinions, views, etc.)
  • These political desires tend, as a result of socialization, to gather around multiple poles of consensus, which might be called "voting blocks" when electoral systems are applied
  • Politics encompasses all kinds of decisions that take place in a particularly social manner: from deciding who gets how much funding, to deciding what issues to cover in the school paper, to deciding NOT to make a decision, to deciding to oppose a war, etc. etc. Basically, politics is ordinary, rather than some realm only occupied by national decision makers.
  • Honest political discussions and decisionmaking are worth having. There are intrinsic and real benefits to identifying, understanding, and discussing differences in politicla desires.
  • Electoral systems are one important category of institutions which facilitate political discussions and decisionmaking

These assumptions lead me to identify major problems with many electoral systems, though particularly the ones used in the United States. First of all, almost all US systems are Winner Take All, meaning that one person or party wins an election, and as a result gets all the representation in power. In addition, almost all US systems use Plurality Voting, in which whoever gets the most votes (a plurality) in a particular race wins, no matter if that candidate achieved a majority or not.

[edit] the United States

Looking at the US Senate and Congress, we can clearly see problems with the Winner Take All and Plurality Voting mechanisms. In each Senate and Congressional race, the block of voters that achieves a plurality takes all the power. Even if the election is 51% vs. 49%, the voters in the "losing" block get nothing. And because of plurality voting, any vote for a third party is wasted unless that third party candidate gets a tremendous and sudden push into the winning spot. As this rarely happens, nobody votes third party. The result is a US Senate and Congress that is narrowly divided between only two parties. Though US voters obviously fall into all kinds of different political perspectives, only two parties have a shot at getting the plurality of votes in each race. They "win" every time, and thus "take all." This has an obvious negative effect on US political discussions over time, as people think of themselves as "Democrat" or "Republican" in a limiting and reductive way. This is evidenced with all the talk of "red" states and "blue" states, even though there is a great continuum of political thought out there.

[edit] What about Macalester?

How does this apply to Macalester as a community? We currently employ Winner Take All and Plurality Voting for the Legislative Body (executive elections were reformed to use Instant Runoff Voting, which facilitated many candidates running last year). (Note: this was slightly modified in a last-minute and temporary change to voting this fall, which can be discussed later.)

At Macalester, we see some of the same problems. The biggest problem is that the plurality, winner take all voting system for LB candidates offers a significant hurdle for minority voting blocks to gain representation. Some minority voting blocks are built into the current system: these are residence halls and major-groupings. However, if a minority voting block emerges that is not tied to dorms or residence halls (the Common Platform is an example), it will have significant difficulty gaining power. Please note, the low-participation of candidates/voters mitigated this barrier for Common Platform candidates this year. However, if participation increases (as we should hope to see), this barrier will become very significant.

How does this work? Let's take a look at a graphical representation. Let's start out with three hypothetical voting blocks that are present across all of the student body. In each block you might find sub-blocks (such as the Doty block, or the Social Studies block), but in this situation these are the three most defining issues for all voters:

Image:pr_vs_plurality4.gif

Under the current system, each individual election for an LB seat looks like this:

Image:pr_vs_plurality2.gif

This is just one race. In most LB districts, more than one person is elected. These are called multi-member districts. However, the plurality/winner take all effect causes the same result for each elected candidate. The same 51% who voted for candidate A in the Social Science I district will be likely to vote for a similar candidate, candidate B for the second seat. Thus both candidates will represent the 51% block, rather than representing two blocks. The same effect occurs over many districts (assuming these blocks have salience across the student body). The blue block will continue to take a vast majority of seats (though some sub-blocks within blue will take individual seats according to major/dorm, and random political distributions in particular districts). One district, just because of chance, may result in a green block victory. Results across districts look like this:

Image:pr_vs_plurality3.gif

As you can see, the Winner Take All/Plurality system gives a large majority of seats to the blue block, even though it only represents 51% of the total student body. I think this is a major problem, which impedes minority blocks from participating in the process and distorts MCSG conversations about campus politics.

[edit] What does Proportional Representation intend to do?

This is the problem that Proportional Representation solves. It is not an eccentric system; rather, it is used in most democracies across the globe, notably and historically in Scandanavia. The fundamental assumption of Proportional Representation (PR) is that elected bodies should mirror the voters who elected them. In contrast to Plurality/Winner Take All, PR looks like this:

Image:pr_vs_plurality1.png

PR does this by:

  1. Electing people according to proportion of the vote, not plurality
  2. Using multi-member districts
  3. Using sufficiently large multi-member districts (the larger, the more accurate the mirror will be). Some countries treat the entire nation as one district.

PR is also very flexible. It can be used with a party system (party list), without a party system (Single Transferrable Vote), and also in a mix with geographic representation (ie, pre-defined divisions like dorm/major).

The worldwide consensus on representative democracy is that PR is a better system. New democracies being formed in the last decade are introduced with PR. Great strides have been made in terms of multi-racial representation, gender balance, and multiparty represention using PR as a cornerstone.

So the first question for Macalester is: is PR the way to go?

Only after that question is it time to decide which form of PR to use. The questions of whether to encourage political parties or not, or whether to continue some geogrpahic representation or not, are actually secondary questions which can be solved later (and that's not just a rhetorical trick).

[edit] Proposal to Implement Proportional Representation in Legislative Body Elections

Matt Hauck and Jesse Mortenson – 2/10/05

[edit] Introduction

This proposal is submitted in the spirit of campus electoral reform initiated by Andy Haug and Jesse Mortenson two years ago, with the successful constitutional amendment to institute Instant Runoff Voting (IRV) for Executive Committee elections. Though some confusion needed to be worked through in last year's exec elections, the use of IRV successfully allowed a large number of candidates to compete for each of several offices without allowing a victory with a very low plurality of voters; something that would have been certain had IRV not been in place. The result may have been a blow to the credibility of MCSG.

The proposal for Proportional Representation continues in this spirit of empowering voters to express preferences as fully as possible, to empower students of all perspectives to participate as candidates with the certainty that their participation matters to the results, and finally to ensure that power is granted as closely as possible to representatives of a true majority of students' preferences. It also follows in the spirit of the Blue-Green Amendment, which encourages “substantive access by all to community power.”

Proportional Representation (PR), according to the Center for Voting and Democracy, “describes electoral systems in which like-minded groupings of voters will win legislative seats in better proportion to their share of the popular vote than in winner-take-all elections.” In other words, it describes voting systems that derive from the principle that all significant groupings of voters should be represented in proportion to their size; if group A gets 30% of the vote, 30% of the representatives elected should share the preferences of group A. Proportional Representation is used by a majority of democratic states in some form. New and re-organizing states are selecting PR over winner-take-all. It empowers a greater diversity of opinions in legislative bodies than winner-take-all systems, in which one person or party gets all the representation in power as a result of winning an election.

The most common forms of Proportional Representation are Party List voting and Single Transferable Vote (STV). Though PR is often associated with Parliamentary systems of governance, this connection is not inherent to the voting system. Most PR-using countries use Party List at the national level, and STV for local elections. Party List provides simpler voting and vote tabulation, and the most precise degree of proportionality, but requires the maintenance of a party system. Single Transferable Vote offers voters a chance to select individual candidates, but requires a more complicated ballot and vote tabulation procedures.

Conversations with the Constitutional Reform Committee in Fall of 2004 have demonstrated that the creation of a party system is very undesirable to some students. Therefore, this proposal will focus on STV, a system which allows candidates and voters to choose from a plurality of ways to divide up representation, including both party-like activities and geographical-like representation.

[edit] Proposal Brief

  1. Legislative Body (LB) elections shall be conducted using Single Transferable Vote (STV). Votes shall be tabulated according to the Meek method of STV.
  2. Candidates for the LB shall be elected in one common, at-large district for all of the student body.
  3. Ballots in an STV election shall list candidates in such a way that the voter may rank candidates numerically according to the voter's preference.
  4. Each candidate shall be required to submit a brief identification phrase, no greater than 40 characters, which will be listed alongside the candidate's name on the ballot. Candidates shall be listed on the ballot alphabetically according to the content of this identification phrase.
  5. MCSG election officials shall utilize freely available vote tabulation software to calculate which candidates are elected, such as the pSTV application.
  6. MCSG shall explore investing in a simple, secure and verifiable system to allow students to vote over the Internet using their existing network login information. In the absence of network voting, MCSG election officials shall request that candidates recruit additional vote-counting volunteers as deemed necessary.

[edit] Why Bother Changing Anything?

There are significant problems with the current system, including the discouragement of participation, as candidates and voters, of students who wish to organize a constituency that is not primarily tied to Major or Residence Hall. Proportional Representation rewards additional participation (by any group) with additional influence. This was concretely demonstrated by the Common Platform last Fall.

Proportional Representation will provide a much more accurate expression of students' desires for what actions MCSG should take or avoid. It will create clear mandates for representatives to take policy action and to understand what ideas and beliefs to which they are accountable.

Proportional Representation will ensure fair representation of minority opinions. It guarantees participation in decisionmaking for students who are not in the political mainstream at Macalester.

Participating in a PR system will educate students about best available practices for representative democracies and advance the goal of making all institutions more democratic.

[edit] Proposal Rationale

[edit] Choice of voting system and method of counting

Single Transferable Vote is chosen as the voting system as it affords the maximum flexibility to students to effectively express which candidate characteristics are most important to their campus political preferences, and does not rely on the maintenance of a party system. The development of the Common Platform of allied candidates (most of whom were elected) in the fall of 2004 LB election demonstrated that a significant constituency exists for representation based upon candidate positions on campus-wide issues (which does not correspond neatly to Residence Hall or Major). However, another significant constituency continues to prefer geographically-proximate representation. The current winner-take-all system (modified in 2004 to add a slight degree of proportionality) with geographical districts cannot accommodate both of these positions. A Party List system can accommodate the desire of many students to choose individual candidates who are familiar and geographically-proximate, but requires more complex activity and organization amongst candidates to do so. Single Transferable Vote is thus the most desirable system, as it allows candidates to self-identify as geographically- or issue-motivated in a straightforward manner, and gives no advantage to voters in support of one or the other position (or any third positions which may arise).

Furthermore, any winner-take-all system by definition grants disproportionate representation to the majority opinion, and is thus less desirable.

The Meek method (created 1969) of counting is selected because it is technically sophisticated, recommended by several major electoral reform organizations, and has been demonstrated to be sound in a national system of local elections (New Zealand).

[edit] Formation of one common, at-large district for all LB candidates

Aside from being winner-take-all, the current system enforces an a priori set of political norms upon the student body as a result its geographical-style voting districts, each of which elects very few (sometimes only one) representatives. It predetermines that voters shall wish to be represented according to their Residence Hall or Major (in this case, group of majors), granting advantage to those voters who agree with this and disadvantage to those who wish to be represented according to another principle (for instance, support of increased student organization funding). How? Students who wish to elect more representatives who favor increased organization funding must encourage candidates in a myriad of districts, rather than drawing from their ranks as a whole. In contrast, students who wish to be represented geographically are automatically provided with candidates (assuming somebody runs) who are geographically-proximate. Creating one at-large district for all LB candidates resolves this problem, giving both sets of voters equal opportunity to find and support candidates supporting their position.

Creating a common, at-large district also rewards those who participate. In the current system, large voter turnout in one particular district does not translate into proportionately greater influence for that group of voters. With an at-large STV system, supporters of a type of representation are encouraged to turn out the greatest possible number of voters, as each supporting voter potentially increases the number of representatives of that view.

There are legitimate concerns with using at-large voting in an STV system. The list of candidates will necessarily be large (likely 60-90 candidates). Without corrective action, voters will have difficulty finding a sufficient number of favorable candidates in such a large field. However, line four of the Proposal Brief describes a mechanism by which this problem can be mitigated (see below). Another problem is that the number of seats to be elected (roughly 28) means that voters need to rank a large number of candidates in order to fully participate in the results, with a minimum recommendation of at least 10 candidates ranked. This presents a challenge in voter education, but should also be recognized as an opportunity for each voter to have a much more precise impact on the election of the LB than under the current system. A final problem is presented by the amount of information (many ranked candidates per each vote) that must be entered into the vote tabulation software. This problem is addressed in line 6 of the Proposal Brief.

[edit] The ranked ballot

A ballot which allows voters to rank candidates numerically according to preference is necessary for the operation of STV. The system works by transferring “surplus” votes from candidates who have been elected by large numbers to other candidates (while never allowing the number of votes to increase or decrease), using ranked preferences to determine the direction of transfer.

[edit] Candidates listed by self-identification on the ballot

In order to assist voters in identifying the candidates who match voters' desires for representation out of a long list of candidates, the STV ballot shall list candidates sorted in alphabetic order of their submitted identification phrase. This requires a low level of cooperation and communication among like-minded candidates (and the development of simple conventions, such as listing Residence Hall name or Major as the first word if a candidate believes in geographic-style representation), which I think can be achieved through the strong incentive of increased chances of winning election. It would result in groups of candidates who think similarly being listed as a block on the ballot, while still allowing for a second or third word which could differentiate within each primary block. Voters interested in voting for a block of candidates could do so (and specify preferences within the block), while others could avoid the groupings entirely. An example (partial) listing might be:

[edit] Using free voting software for tabulation

The Meek method (along with other sophisticated methods) of STV requires the use of a computer algorithm to tabulate winners. There are simpler STV methods that are nearly 100% accurate and can be executed with a manual handcount of votes. However, this is unnecessary in a campus situation in which access to computer technology is assured. There are several efforts to create free STV tabulation software. A program called pSTV is a mature project which is free, open-source (meaning that the source code must be distributed with the software and can be freely examined and used) and runs on Windows PCs.

[edit] Recommendation to create network voting system

A network voting system is highly desirable in conjunction with STV because it greatly improves the speed of the vote tabulation process. Though computerized (aka touch-screen) voting is highly problematic when applied to most communities (and certainly to high-consequence races like U.S. President), it could be effectively and securely be used a the small, private school setting of Macalester College. All students are guaranteed access to computers and the Internet, and each student has a unique network username and password (used to access network resources like Lester and e-mail). This makes it very easy to establish an electronic vote database and ensure that each student only votes once. Unique “voting receipts” could be generated for each student to ensure that the system is verifiable should concerns arise. The system could also be secured in such a way to ensure confident oversight for MCSG election officials. It could be tied into the Lester system to provide access in a manner that is familiar to students. It could potentially be extended to faculty and staff for future expansions of democracy.

A network voting system is also desirable because it could increase voter participation by making it more convenient and accessible to off-campus students. Such a system could be connected to a candidate information page.

Such a system could probably be developed for $1000-$1500, which is quite affordable in terms of MCSG's capital fund. The system would last many years without maintenance. Full disclosure: Jesse Mortenson's web design business is interested in developing such a system for sale, but does not currently have such software developed or in development.

However, a network voting system is not necessary to perform the Meek method of STV. It is only required to get all the votes into a digital file. Each LB candidate could be required to recruit two volunteers to enter votes from paper ballots into a database. This would yield over one hundred data entry volunteers (who could be used in shifts over a day or two) to achieve the task efficiently. Once the data is in digital form, it takes only minutes to tabulate results.

[edit] Conclusion

Implementing STV at Macalester College will place additional burden on MCSG officials in order to conduct LB elections. This investment will achieve several important results that are beneficial both to our internal practices of student democracy and the external goal of educating students about best available democratic practices. There is clearly a large participation gap between the student body and MCSG; we cannot afford to ignore any reform that promises a significant improvement in participation. Proportional Representation, whether in the form of STV as presented here, or as Party List voting, is associated historically with high voter turnout. The College should always be an incubator for theories and solutions that are new to the community in which it resides. For these reasons, we support the implementation of PR at Macalester College.


[edit] More PR/STV Resources

[edit] See also

Retrieved from "http://mpkb.net/en/Proportional_Representation_for_MCSG"
Personal tools
macalester