2006 Emails re UBC VFM

[This is a selection of highlights from the VFM Committee's correspondence with Mark Latham. For privacy and brevity, email addresses, some signature contact info, and appended echoes of emails received are omitted.]

From: Mike Thicke
Sent: Friday, August 04, 2006 2:25 PM
To: Mark Latham
Cc: Tristan; ncrompton
Subject: Turbo Democracy Comments

Hi Mark,

I spoke with you briefly after your presentation to the AMS. I've just read through your paper and I thought I'd offer some comments. (Turns out this is pretty long, sorry.)

It looks to me like what you're trying for is a system that in a sense runs itself. There is no need for regulation, and no opportunity for corruption. Voters determine both who gets money and how much, so narrow interest groups aren't able to channel money towards their own interests against the public good. This makes sense. However, my feeling is that this ideal needs to be comprimised in the name of voter accesibility.

I sit on the executive of the Social Justice Centre at UBC, which gives several thousand dollars in grants each year. We average something like a proposal every week or two. It is very hard to determine how much we should grant to each request. We have to go over their budget, look at other sources of funding, and decide if the amount they are asking for is reasonable for the project they are working on. That is on top of just evaluating if the project's aim is compatible with ours, and if the project's scope makes it appropriate to be funded by students. For large proposals deliberation can last weeks as we ask questions, have discussions, and consider alternatives. It can be a daunting task. I cannot see how an average voter is supposed to do this in any responsible way for up to a dozen proposals, while just looking at a single ballott.

The reply is, I suppose, that most voters will simply vote "Yes" or "No"

without considering the amount requested. This is probably true, but it is not responsible voting. There is a big difference between $5000 and $2000, especially for a group of students. This difference will be hard for a voter to appriciate on voting day, while spending 30 seconds filling out a ballott. The result is that there will be some sort of game where proposers play amateur psychologist in deciding an amount to request that will maximize their expected profit. They don't want to request so much that the voter's stomach lurches, but they also don't want to request less than they would be able to receive. Some proposers could even be expected to short-change themselves in the name of caution. They may decide that a 60% chance at $1000 is better than a 40% chance at $1800, even though the latter has a higher expectation.

I would propose simply having voters rate each media provider on a sliding scale, and have a committee evaluate each budget along with this data to allocate the prize money appropriately. This is not as iron-clad against corruption as your proposal, but there can be checks and balances and formulas that minimize the potential for abuse. This would allow voters to vote on what they are qualified to vote on - what media they found useful - while allowing another group to do the hard work of finding appropriate funding levels for the groups that voters direct them to fund.

My next concern is a broader philosophical one. Corporate media is how it is today because the measure of success is not quality of reporting but amount of profit. Thus, media is rewarded for attracting and keeping the attention of readers, not for informing them. The result is that it is more appropriate for a newspaper to be run by the marketing department than the editorial department, or for the marketing department to simply take the place of the editorial department. The measure of success under Turbo Democracy is also not quality of information but amount of profit. By bringing in a substantial monetery prize, media sources are encouraged to behave in a way that will maximize their monetery returns. Now ideally the way to maximize returns would be to provide good information, but ideally the way to maximize returns for traditional newspapers would be to provide good information, and hopefully we can agree that this isn't what happens. Why should we expect it to work for Turbo Democracy when it doesn't work for the rest of the media?

My substantial worry is that competitors in this experiment would have a large incentive to PROMOTE THEMSELVES rather than provide information.

When a student invests $200 for a chance to get $2000 they are in essence investing in a small business. It is in their best interest to do everything they can to ensure a profit. I see this effort coming at the expense of being good information providers to students.

Having the $200 entry with a high expectation of return and high variance will, I think, result in the field being mainly larger institutional media than single intrepid reporters. Larger institutions can stand the variance much more easily than an individual, because they have a larger pool of money to start with. The Ubyssey with its $350,000 (this surprised me, btw) annual budget wouldn't mind another $5,000 but it won't break the bank if they just lose $200.

My feeling is that to combat this both the variance and expected return need to be reduced. Good reporters with minority viewpoints should not be discouraged from participating because they don't expect people to like what they have to say. Media providers should not have to labour under such uncertainty that they might be getting reimbursed for thousands of dollars, or nothing. Rather, I think it would make more sense to make the probability of winning some portion of the purse higher, and make the maximum prize available lower. I would also like to reduce the entry fee required to be in the contest, perhaps replaced by some other hurdle, such as a full budget proposal and statement of intent - ie. discourage joke / selfish entrants by making it not worth their trouble. Then make the maximum award something more reasonable, like $500. There is still some incentive to do a good job, but covering the election won't replace your day job. Thus you are more likely to get contributors who care primarily about the democractic process.

About "Turbo Democracy" - I share what seems to be the majority view that this isn't a great name. As I mentioned to you before, it sounds like "Sound Bite Democracy" to me. Personally, I like Slow Democracy, deliberative democracy. Turbo Democracy sounds like a way to vote by cell phone or something, so it doesn't take time out of my TV watching.

On your card you have the phrase, "Media reform by majority vote funding." Why not just call it "Majority Vote Funding"? MVF? People have expectations about what something is by how its name sounds. Majority Vote Funding sounds academic and responsible. Turbo Democracy sounds like a gimmick. For example, do you think STV would have gotten more or less support if it was called Easy Vote?

Anyways, I'm not a marketer either. We'd probably get more traffic at The Knoll's website if I was!

However it ends up being implemented, I'm sure this will be a very positive experience for the AMS and I'm really looking forward to seeing what will happen. As things develop The Knoll will probably want to cover it more in detail and perhaps interview you or something.

And I'm more than willing to have an involved discussion about this if you want to respond to my comments, but don't feel duty-bound to.

-Mike

From: Mark Latham
Sent: Monday, August 07, 2006 10:44 PM
To: 'Mike Thicke'
Subject: RE: Turbo Democracy Comments

Thanks Mike for taking the time to read my paper and write some comments.  It's so helpful when people think seriously about how to implement this.  I'm looking forward to working with UBC students on it, to anticipate problems and figure out how best to handle them.  Am digesting your comprehensive feedback.

I'm open to many design changes.  Waiting to hear from the AMS executive as to the next step, but it sounds like a committee will be formed to work on this.  If you're interested in being on that committee, you could let Ian Pattillo or Kevin Keystone know.  (Ian's out of town now for a week, by the way.)

I appreciated your explanation of why Turbo Democracy is not a good name.  The message is getting through to me.  I'm surprised I didn't get much criticism of the name during the almost two years I've been using it.  Maybe now that people are seriously planning to announce this to voters, they take a harder look at how it's going to sound.  Will think more about better names.

More responses to your comments later...

-- Mark

From: Mark Latham
Sent: Monday, August 14, 2006 12:12 PM
To: Ian Pattillo
Subject: Renaming turbo democracy

Hi Ian –

How was your trip?

I think it would be good to choose a new name soon to replace “turbo democracy”, so I’ve given it some thought and would like to run some ideas past you.  Please let me know what you think.

1. Turbo Democracy -- Minuses: sounds gimmicky; many people at the AMS Council meeting August 2 didn't like it.  Pluses: distinctive, so would serve well as a kind of brand name.  (By this I don't mean a commercial trademark, but just a clear identifier when people talk about it etc.)

2. Voter-Funded Media -- Pluses: not gimmicky; it's straightforward and self-explanatory.  Minuses: one reaction I've had to it so far is that it lacks pizzazz, ring, poetry; it's not as distinctive, so not as good for a brand name.  ALSO: it’s too long for a web domain name.  So shorten it to:

3. VoterMedia.org – this is easy to spell when you hear it spoken.  Followup slogan: Media for voters, funded by voters.  Compare website names FairVote.org, MoveOn.org, FreePress.net, DemocracyNow.org

There are many uses/contexts for the name:

- AMS Council election; then other universities

- larger democracies

- corporations

- website

- business card

- my article I hope to publish in a law journal

We don't have to use the same name in all contexts, although clearly there are benefits from using a consistent name.  Two-part names also possible -- now I'm thinking of entitling my article "Voter-Funded Media: Governance Reform for Democracies and Corporations".  My website and business card both say "TurboDemocracy.org" and on the next line "Media reform by majority vote funding".  My corporate governance reform website www.corpmon.com says "Corporate Monitoring" then "Power to the Shareowners!"  There, the same voter-funded media proposal is called "Proxy Advisor Proposal" because the organizations that advise investors on voting shares are called proxy advisors.

More name-brainstorming material follows...

-- Mark

turbo democracy

voter-funded media

voter media

votersmedia

democratic media funding

democratically funded media

democratic media

media democracy

infomediary democracy

informed democracy

enlightened democracy

insightful

understanding

discerning

enhanced insight

perceptive

empowered

opposite of superficial – deep, profound, meaningful

checks & balances

integrity

public journalism

consensus

balance of views

multiple perspectives

voter sovereignty

media responsive to voters

political news media in public interest

political insight

evaluation

analysis

election info award voting

Voter-Funded Media: Governance Reform for Democracies and Corporations

Media reform by majority vote funding

Corporate Monitoring

Power to the Shareowners!

Proxy Advisor Proposal

Compare existing names for related ideas:

Single Transferable Vote – STV

Deliberative Democracy

Deliberative Polling

Citizens Jury

Citizens’ Assembly

Planning Cells

Lay Citizen Deliberations

Consensus Conferences

The Battle for the Soul of Capitalism

From: Mark Latham
Sent: Wednesday, August 16, 2006 10:33 AM
To: Ian Pattillo; Kevin Keystone
Subject: Turbo democracy & AMS Council meeting Aug 23

Hello Ian and Kevin –

I’m planning to attend your AMS Council meeting again next Wednesday August 23, so that I can be available for any followup to the turbo democracy discussion from the previous meeting.  Would that be helpful?  Also I’d like to get some feedback from council members about possible names to replace “turbo democracy”.

I emailed Ian a long list of possible words to use, but so far my favourite alternative is:

“VoterMedia.org” for the website name;

“Voter-funded media” for a descriptive term;

“Media for voters, funded by voters” for a slogan or subheading;

“Voter-Funded Media: Governance Reform for Democracies and Corporations” for the title of my article that I’m submitting to law journals.

Would it make sense for you to send this to council members as part of their August 23 meeting materials?

By the way, let me know if you’d like to have some council discussion on this without my being there – I can step out of the room for a while or whatever.

See you next week...

-- Mark

From: Ian Pattillo
Sent: Wednesday, September 06, 2006 9:58 PM
To: Mark Latham
Cc: AA
Subject: RE: Turbo democracy & AMS Council meeting Aug 23

Hi Mark,

The committee has been struck. We will meet in the next few weeks, and I will let you know when that is to happen. I am trying to get an EA hired quickly so that they can assist. You should really just continue to think about how this could be administered with an AMS election. Advertised, too.  

Ian Pattillo

From: Mark Latham
Sent: Thursday, September 07, 2006 10:21 AM
To: 'Ian Pattillo'
Cc: Kevin Keystone
Subject: RE: Turbo democracy & AMS Council meeting Aug 23

Thanks Ian for the update.  I’m looking forward to getting going on this project.

As I outlined in my August 2 presentation to the AMS Council, a successful first implementation of turbo democracy will require a lot of publicity and education, especially for voters and for potential media contestants.  Here are the main steps I think we need to follow:

1. Choose a name to replace “turbo democracy”.

2. Decide the rules for the media competition, including ballot format.

3. Get AMS Council approval to proceed with implementation.

4. Publicize the competition to potential media contestants.

5. Media sign up, start covering AMS issues and building their websites.

6. We publicize the new info system to voters, with media help.

7. Media gradually gain reputation in the eyes of voters and other media.

8. Potential candidates decide to run for AMS positions, partly because of this new info system that will help voters elect good candidates.

9. Election campaign period begins.  Media cover debates, interview candidates, write evaluations and endorsements.

10. Media review and critique each other.

11. Voters read media websites, vote for election candidates, and vote funds to media contestants.

12. Candidates are elected; media awards are paid out.

All these stages take time.  That’s why I recommended phasing it in over the next two elections.  I was happy to hear Council’s enthusiastic response favouring a faster full implementation in January 2007, but for that to be successful we would have to move fast.

I think it’s important for voters, media and candidates to develop their understanding of how this new info system works, before the election campaign starts.  That education process is in steps 5, 6 and 7 above, and can be expected to take several months.

Ideally, I would have liked to be at step 5 by now as the semester begins.  How fast can we play catch-up and get the first four steps done?  As each week goes by, the chance of success diminishes for a full implementation by January 2007.

Can we do step 1 (choosing a name) by email and phone, instead of waiting for the first committee meeting?  This is urgent for marketing – e.g. the Ubyssey will publish an article about this any day now; also I can get going on building a new website.

Please let me know your thoughts on these plans and the timing challenge.  Let’s talk soon.

-- Mark

From: Mark Latham
Sent: Wednesday, September 20, 2006 6:01 PM
To: Ian Pattillo; Matthew Naylor; Mike Thicke; Stephanie Ryan

Subject: Turbo democracy committee

Thank you for working on this project to make democracy work better.  We are going to change the world!

We have many things to talk about at our meeting tomorrow.  Here’s one key item:  Based on feedback so far (especially from Mike Thicke), I’ve revised my suggested design.  Thought it would be helpful to sketch it out for you to think about before we meet:

1. To encourage a diverse range of media, cap each award at e.g. $3000.  That way one dominant media organization won’t get most of the pool.  At least four awards will be given out.

2. I had suggested a ballot design where each media contestant requests an award amount, then students vote “Yes” or “No” or write in another amount.  One problem with this is the strategic gamesmanship involved for each contestant to choose their requested amount.  I’m OK with either leaving the design that way, but now think it’s better to eliminate the Yes/No part and just have each voter choose a dollar amount for each contestant.

3. Mike pointed out that choosing award amounts may be difficult for many voters.  He suggests instead that voters rate each contestant e.g. on a scale of 1 to 10, and then a committee use those as input to deciding award amounts.  I’m afraid that runs contrary to my goal of creating a check and balance to government – too much government discretionary control over media funding.  As a compromise, that government-appointed committee could itself be one of the media contestants.  Contestants are expected to give voters guidance not only on voting for electoral candidates, but also on voting funds for media.  This committee-contestant could specialize in advising on media budgets, and not give advice on electoral candidates.

4. My previous design of awarding the median voted amount had some nice properties (like being based on majority vote) but would almost certainly end up awarding either more than $10,000 and have to be adjusted down, or awarding less than $10,000 (maybe zero!).  Media contestants would prefer to have more assurance as to how much award will be given out.

Here’s how to solve both those problems at once.  We base our awards on the consensus level that allocates the full $10k:

- If there is a Yes/No option, “Yes” is the same as writing in the requested amount, and “No” is the same as writing in zero.

- If a voter doesn’t vote for a given contestant, that’s the same as voting $0 (we could change this, but it’s probably the simplest fair design).

- For each contestant, arrange the voted amounts for that contestant in increasing order from $0 to $3000.  Suppose there were 1000 voters.  Then each contestant would have 1000 voted amounts, where #1 is probably $0 and #1000 is probably $3000.

- For each rank #1 through #1000, sum the voted amounts across all contestants.

- Find the two consecutive ranks #n and #(n+1) where the #n summed amount is below $10k and the #(n+1) summed amount is above $10k.

- Find the weights so that the weighted average of those two amounts is $10k.

- Award each contestant the similarly weighted average of their #n amount and #(n+1) amount.

- This system virtually guarantees that the full $10k will be awarded.  Only in the unlikely event that the #1000 sum is less than $10k, the awards would be the #1000 amount for each contestant.  I think you’ll agree this is an extremely unlikely case.

5. This new design is much more generous.  If there are only 4 contestants, each one is guaranteed at least $1000 just by casting one vote for themselves!  So we can be sure more than 4 will enter the contest.  Maybe then $10k is too much award, or $200 is too low an entry fee?

Let’s talk about it – see you tomorrow!  (Thursday, September 21, at 5:15 pm in SUB Room 266J)

-- Mark

From: Mark Latham
Sent: Sunday, September 24, 2006 7:14 PM
To: Ian Pattillo; Matthew Naylor; Mike Thicke; Stephanie Ryan; Steven Klein
Subject: Simplified Award Voting Design

During our meeting last Thursday, I thought up a new simpler design for award voting which I’d now like to write out and run past you.  This is partly in response to our concerns about the complexity of determining award amounts by vote, and partly because we expect that Webvote may not handle write-in amounts.

Suppose we offer a predetermined sequence of media awards, such as:

1^st prize: $2,500

2^nd prize: $2,000

3^rd prize: $1,500

4^th prize: $1,000

5^th prize: $500

Voters simply click to vote “yes” on as many media contestants as they like.  Thus voting would be formally similar to the senator elections where voters can choose multiple candidates, but without the limit of five that a voter may select.  I’m assuming senator elections have this limit of five, and we could impose it here too, but I don’t think that’s necessary or desirable.

Then the 1^st prize goes to the contestant with the most votes, 2^nd prize to the one with the second most votes, etc.  Tie votes would be settled by awarding the average amount.

Desirable features of this design:

1. Webvote can surely handle it.

2. The voting format is easy to understand, and familiar to students.

3. The connection from votes to awards is easy to understand, and would seem reasonably fair to most people, being straightforward and similar to many other voting contests.

4. This award system would give a clear incentive for media contestants to appeal to as broad a consensus of voters as possible.  This should discourage extremism, while encouraging a variety of coverage that would not neglect any substantial voter subgroup.

5. This array of awards would induce a minimum of five contestants to enter.  The total number of entrants would also be greatly affected by the entry fee.  I suggest a fee of about $200, which I hope would result in about 10 or so entrants – enough for variety, competition and cross-checking, but not too many for busy students to read.  The entry fee funds could be used for award system infrastructure and advertising costs.

6. These awards could give reasonable incentives to both large and small media groups, from the Ubyssey to individual bloggers.

Voter media awards are designed to empower voters, so I generally prefer to give voters as much control over award allocation as possible.  That’s why my previous designs let each voter write in an allocation if desired.  If we choose the above simplified design, voters would lose the option of expressing their degree of consensus separately from their desired award amounts.  For example, with write-in amounts, voters could express an 80% consensus for awarding $1000 to contestant A, and an 80% consensus for awarding $3000 to contestant B.  In the simplified system however, % consensus is treated as synonymous with award size ranking.  That is the price we would pay for simplicity.  I view this as “training wheels” that we would plan to remove in the future, when participants are more familiar with how voter media awards work.

Looking at the pluses and minuses, at this point it seems to me that the above simplified system is preferable to others we have considered, even if Webvote could support the other, more complex systems.

What do you think?

I’ll send you this email now so you can start digesting it.  Probably tomorrow I plan to send a followup email outlining the designs “A, B and C” that we discussed at our Thursday meeting.

Meanwhile a few notes on the total award amount.  I’m sure you did the math to see that my suggested awards above total to $7,500.  Whichever design we choose, we’ll have to decide what we think the best award amount and entry fee are.  I suggested what I did here for these reasons:

1. My original proposed design offered $10,000 total, but required minimum 50% consensus vote for any contestant to get an award.  There would have been some chance that no one would qualify.  Here, the design would guarantee that these 5 awards totalling $7,500 would be given out (except for the extremely unlikely case where most contestants get zero votes).  This guarantee makes it easier and more attractive for contestants, so the total potential award pool need not be as big.

2. We can reasonably expect 10 or so entrants, which seems to me about what we want.

3. It’s getting a bit late to launch the competition, so contestants won’t spend as much total time prior to the campaign period covering AMS issues as I had thought.  I’m still in favour of launching, but with this smaller award pool.

4. I’m starting to think that I will probably offer to sponsor this again for the January 2008 AMS Council election, so a smaller pool this time makes it easier for me to afford doing that.  We can expect our voter media system to evolve in various ways, including removing some “training wheels”, getting AMS Council to liberalize the restrictive election rules, improved understanding by voters and media, etc.  So a followup year would be very desirable if it goes well the first year.

I look forward to seeing what you think of all this.

-- Mark

From: Mike Thicke
Sent: Monday, September 25, 2006 4:25 PM
To: Mark Latham
Cc: Ian Pattillo; Matthew Naylor; Stephanie Ryan; Steven Klein
Subject: Re: Simplified Award Voting Design

This might be a little scattered.

Concerns:

1) Does this encourage strategic voting? That is, a vote of "yes" for 10/10 candidates is the same as voting for none at all. A vote of "yes"

for just one candidate is the best option if what you care about is that candiate getting money, and each subsequent "yes" vote makes your overall contribution smaller.

2) Before it seemed we were trying to fairly compensate people for good work. It's possible that every single entrant would get paid, if they all did a good job. The pool would be spread thin, but it seems likely that all of the good efforts would have gotten a reward. So it didn't feel like so much of a contest. Now it is explicitly a contest.

3) In the old system, was an abstain the same as a "no"? I'm unclear reading your paper. I think this is bad, because it means that institutions with greater access to the voters (eg. The Ubyssey) have a very large advantage. In fact I think that there is a very high (like

90%) chance that the Ubyssey would win first place under the new system, whether or not it deserved it. That the AMS pays the Ubyssey to cover the elections makes this even more unpalatable.

4) Do we really want to have a system that rewards "centerism"

specifically? I don't believe that centerists have any natural claim to being right, so a system that is designed specifically to reward them seems biased. There is a difference between having, say, an ethno-centric perspective and having an unpopular political perspective.

I am happy with a system that does not reward very narrow interests, but I am not happy with a system that excludes unpopular opinions.

5) The current media system seems to be very concentrated on personal attributes and competancy. Did Bush really desert from the national guard? Things like that. Competancy issues intrinsically have a wide appeal, as everyone responsible should care about electing someone who is at least capable of doing the job. However they should not, in my opinion, drown out political debate and ideologies. This is what seems to happen - campaigns focus on people, not issues. Does this system

(either) fall into the same trap of encouraging media to focus on universally-appealing issues like competancy rather than on trying to evaluate the issues? I would like it if there could be media contestants actively setting the agenda for candidates and forcing them to take positions, rather than on some sort of muckraking activity.

6) I feel that $200 is a very large hurdle for students to clear before entering the contest.

7) I support reducing the overall prize pool (not that it's really any of my business). I'd also like to suggest making the distribution flatter.

That's probably more than enough for now.

-Mike

From: Mark Latham
Sent: Monday, September 25, 2006 10:37 PM
To: Ian Pattillo; Matthew Naylor; Mike Thicke; Stephanie Ryan; Steven Klein
Subject: Previous Award Voting Designs

Previous Award Voting Designs

Last Thursday we were comparing three award voting system designs, which we designated A, B and C.  Below I’ve written what I think we meant by those three designs, to help get us all on the same page so that we can better discuss and choose one.  Let’s call the new simplified system I emailed you yesterday Design D.

Date: The date the design was proposed.

Y/N Vote: Do voters have a yes/no voting option?

Yes means: If a voter votes Yes, what does that mean?

Write-In $: Can voters write in the amount they want to award?

Award determined by: The formula or method that determines the award.

Design A:

I emailed you this the day before our last Thursday meeting.

Could be designed with or without a Yes/No vote option.  Voters can write in award amounts.  If there is a Yes/No option, each media contestant requests an award amount, so “Yes” would mean voting that amount.  If there’s no Yes/No option, voters would have to write in amounts.

The awards are determined by finding the number of votes n such that the nth highest amount voted for each candidate just uses up the award pool.

Design B:

I presented this design to AMS Council on August 2.

I defined this as having a Yes/No option and a write-in option.  It could also be defined without the Yes/No option.  The other alternative of having a Yes/No vote with no write-in option is also possible, but would substantially change how it works.  “Yes” means voting to award the contestant’s requested amount.

Each media contestant is awarded the median of its voted amounts, subject to proportional scaling down if those medians sum to more than the award pool.  If the medians sum to less than the award pool, then they are not scaled up (though we could change the design to scale up if we want).  The idea of that is to allow voters the option of choosing not to award the full pool if they felt it was not warranted.

Design C:

This was proposed during our meeting, I think by Matt.  I’m not sure I got exactly what the design was, so please let me know if you have a different understanding of it.

Voting is just Yes or No on each contestant.  “Yes” means voting for a $3,000 award?  Also not sure what the rule is for determining which media get awards.

Design D:

I emailed you this yesterday.

Voting is just Yes or No on each contestant.

Contestant with the most Yes votes gets 1^st prize, etc.

1^st prize: $2,500

2^nd prize: $2,000

3^rd prize: $1,500

4^th prize: $1,000

5^th prize: $500

Notes:

For designs C and D, voting is only by Yes/No on each contestant.  Normally (e.g. in Webvote I believe), that is done by clicking a single box next to each contestant’s name.  There is no “No” box to click, only a “Yes” box to either click or leave unclicked.  So I guess that would mean that there’s no way to distinguish “No” from a blank non-vote on that contestant.

But if we offer both a Yes/No option and a write-in option (e.g. design A or B), I guess we would have a Yes box, a No box, and a write-in box.  Then it would be possible to distinguish between an active “No” (same as write-in $0) meaning “Dislike – give $0!” versus a passive leaving that contestant blank meaning “Don’t know or don’t care”.

Let’s discuss by email.  Thanks Mike for your detailed thoughts on design D.  I’ll digest and contribute to the discussion.

-- Mark

From: Mark Latham
Sent: Thursday, September 28, 2006 6:33 PM
To: Ian Pattillo; Matthew Naylor; Mike Thicke; Stephanie Ryan; Steven Klein
Subject: RE: Simplified Award Voting Design

We have a range of possible contest designs, and concerns about them.  Here's a suggestion for how we can bite off a smaller first step for our decision process:

Let's start by figuring out which design we would choose if Webvote limits us to voting just by clicking (or not clicking) a single box next to each contestant's name.  Let's limit this further by supposing we wouldn't have a different suggested or requested award amount next to each name.

Thus for example, this effectively postpones Mike's concern #3 to a later decision stage, when and if we can also have a "No" box and/or $-amount box.

There is no perfect design.  The design we choose will reflect tradeoffs.  When we look at an imperfection, we need to ask what alternative design we can think of, see if it compromises some other aspect, then weigh the pros & cons.

So what contest designs are possible with single-box voting?  Out of the A B C D designs we’ve looked at so far, I guess only design D qualifies here.  But there are others.  Besides changing the entry fee and award schedule, we could choose:

Design E

Every contestant getting over 50% Yes votes receives $1500 [or $2000], unless more than 5 contestants qualify, in which case the $7500 [or $10,000] pool is divided equally among them.  This resembles designs B and (I think) C, and would be more cooperative / less competitive than design D.

Design F

Voucher system: Divide the total award pool by the number of voters.  Then for each voter, divide the per-voter award equally among the contestants selected by that voter.  Each contestant’s award is the sum across voters of these amounts.  [I oppose voucher systems for reasons I’ve described earlier, e.g. in my “Turbo Democracy” paper.  I’m just listing this as another possible design example.]

No doubt there are many other possible designs under single-box voting.  If you can think of any that we should consider, please describe them.

Here are my thoughts on Mike’s concerns, numbered as in his email below:

1)  I agree that some types of vote counting rules encourage strategic voting in a way that seems socially undesirable.  An example would be a first-past-the-post election with three or more candidates (e.g. Bush vs Gore vs Nader in 2000), in which you might not vote for your favourite one (e.g. Nader) if you think he has no chance, so you vote Gore.  A better system would let you express your honest preference by ranking Nader 1, Gore 2, Bush 3 and then count votes using a Condorcet method (http://en.wikipedia.org/wiki/Condorcet_method).

But the design D we are considering here doesn’t seem as strategic as that to me.  Each voter has a different insight to contribute to the group decision.  If you only like one contestant, just vote yes for that one.  If you like nine of them equally well, but detest the tenth one, vote yes for your nine even though only five of them can win prizes.  The other voters’ insights will decide which five of the nine are better (or might outweigh you and reward the tenth).

For typical voters with more graduated preferences, it won’t be so clear where to make the cutoff of which fair-to-middling contestants to vote yes for.  But that doesn’t seem too serious or socially harmful to the consensus decision process.

Another aspect that could be considered strategic is trying to compensate for the one-dimensionality of yes-bpx voting that I mentioned in my Sept 24 description of this design.  (Maybe this is what Mike had in mind.)  Design D treats consensus as synonymous with award size rank.  What if there is a blogger that everyone likes, but doesn’t really need to be paid $2500?  $500 or $1000 would be enough.  If you think most voters will vote yes for her, you might not vote yes even though you like her ideas the best, because you want the bigger awards to go to media groups that have put more time and effort of more people into their work.  That would be a voting strategy that I hope many voters would factor into their decision, to help compensate for the limitations of yes-box voting.  I doubt that other yes-box designs can avoid this issue (can they?), but we should revisit it if and when we can consider richer voting options like write-in amounts.

2) I agree design D is more explicitly like a contest than design B, and that may have some negative effects.  I think the interaction of media contestants with one another will be a crucial component of the voter media system.  We need the media to review and critique each other.  We want that critiquing to be balanced, with reviews sometimes favourable, sometimes unfavourable, based on balanced intelligent opinion rather than factional competitive mudslinging.  We should think about what flavour of interaction our system design would induce.

Even with a competitive system though, I hope and expect that voters would not reward contestants that descend into factional bickering and knee-jerk criticism of competitors, so such behaviour would be minimized.  And competition can have the positive effect of inducing better service to voters.

I think design D is more competitive than design E.  This goes hand in hand with the aspect of whether the entire award pool is guaranteed to be paid out.  If so, then one contestant getting more implies others must get less.  But design E has a significant chance of not paying out the full pool.  That implies the disadvantage of uncertainty as to how smart/generous/coordinated the voters will be in giving out award funds.  We can’t have it all – no perfect system – so must choose which tradeoff we want to make.

3) As I mentioned above, it looks to me that with only a yes-box there’s no way to distinguish abstain from no.  Do you agree, or am I missing something here?

I agree there is something unpalatable about a big Ubyssey advantage leading to an easy first-place win.  But the Ubyssey’s competitors will make sure that voters know how much student money the Ubyssey is already getting.  I believe in informing voters well, then letting them decide.  On a closely related issue, you plan to recommend to Council that they allow all newspapers to compete, including those with AMS funding, right?  You would probably also recommend that if AMS-funded groups are banned, then the Ubyssey should be banned too, right?  So it’s kind of a negotiation point.

That still leaves the design question of whether to reward contestants who get more votes.  It may give the Ubyssey an advantage, but still I think it’s so important to encourage information with broad public benefit.  This seems to be an issue where a concrete alternative proposed system would help focus our discussion of pros & cons.  I invite Mike or anyone to suggest one (or say that you like design E) if you would like us to pursue this question.

4) Here too I think a specific suggested alternative design would help us focus the discussion.  Design D rewards votes, regardless of where in the left-centre-right spectrum they come from.  So does design F (vouchers).  But F would reward many narrow-interest media whereas D would reward only broad-interest media.  My economic analysis indicates that our existing media fail us especially in not providing insights of broad voter community interest.  Letting voters allocate collective funds by voucher invites them to spend the money on media that give less social benefit than the social cost.  I would rather spend my $5 voucher to create a $1 benefit for a group of 100 that I belong to (1 cent for me), rather than a $10 benefit for a group of 10,000 that I belong to (1/10 of a cent for me).  So I think design F would create less benefit than its cost, whereas design D would create more benefit than its cost.

From my outline of design D:  “This award system would give a clear incentive for media contestants to appeal to as broad a consensus of voters as possible.  This should discourage extremism, while encouraging a variety of coverage that would not neglect any substantial voter subgroup.”  So I don’t think monority views (ethnic or political) will be excluded.  I expect to see media appealing to a broad range of views, often providing political insight that would be useful to any voter regardless of their stance.

5) I agree that existing media focus too much on gossipy superficial stimulation.  I think this is driven by the economics of channel-surfing viewers getting entertained.  Those same viewers (or rather, the ones that vote) also want deeper political insight, but aren’t sufficiently willing to pay for it by spending much time on such a serious channel.  But they would vote public funds for it, then use that info briefly before they vote.  So I predict that our new voter-funded media will quiz candidates on issues more effectively than existing media.

6) I agree $200 is a substantial hurdle, and might well discourage some worthwhile potential contestants.  The purpose of the hurdle is to limit the amount of “noise” that voters have to wade through to get at the “signal” of helpful political insight.  “Noise” includes partisan campaigns (which may well be disguised as news), and other types of advertising – single-issue cranks who want to push their issue to all student voters (hmm, that sounds like me!), perhaps even commercial advertisers (again, mixed with news).  A simplistic but helpful way to think about it is that we only want entrants who think they have a real chance at winning some prize money.  Award voting is what sorts out signal from noise.  As you know, I don’t favour having an oversight committee that outlaws some types of media content (e.g. for being partisan or commercial).  I count on the other media to critique such content, and I count on voters to reward good content.

Even $200 may be too small.  Suppose 4000 students read those web pages.  For a $200 advertising fee, that’s only 5 cents per person you can reach.  Thinking about it this way, I’d be nervous about a lower fee.  The fee should be high enough to screen out most noise, and then the awards should be high enough to attract public-interest media, who must expect to win enough award to pay back their entry fee plus compensate them for journalism work.

7) For reference, here are the design D awards (call this schedule D1):

1st prize: $2,500

2nd prize: $2,000

3rd prize: $1,500

4th prize: $1,000

5th prize: $500

I encourage any of you to comment on how much the total prize pool should be.  As for flatness, let’s compare specific alternatives.  Here’s one (D2):

1st prize: $2,000

2nd prize: $1,700

3rd prize: $1,400

4th prize: $1,100

5th prize: $800

6th prize: $500

Here’s another (D3):

1st prize: $1,800

2nd prize: $1,500

3rd prize: $1,200

4th prize: $1,000

5th prize: $800

6th prize: $600

7th prize: $400

8th prize: $200

These also add to $7,500.  Keep in mind the tradeoff of how much the average busy student voter is willing to read.  How many more contestants will enter than the number of prizes?  What do you all think?  Other suggestions, comments etc?

I think we need to make progress on these decisions soon, so please share your views with an eye to reaching your committee consensus.

-- Mark

From: Mark Latham
Sent: Sunday, October 08, 2006 4:56 PM
To: Ian Pattillo; Matthew Naylor; Mike Thicke; Stephanie Ryan; Steven Klein
Subject: Followup on our 20061005 meeting

Here I’ll try to write down some of what we decided at the voter media committee meeting last Thursday, and make it more specific.

We reached consensus on having an entry fee of $100.

One major debate we had was between two contest designs:

Design D:

VoterMedia voting appears in the ballot somewhat like the student senator race, but without the five-vote limit.  Each contestant gets just one line, with a single box to click indicating voter approval for award.  Each student can vote for as many contestants as desired.  Contestant with the most votes gets first prize, second most votes gets second prize, etc.

Design E:

This design from Matt may be the same one he suggested on Sept 21 which I called design C and didn’t really understand.  But to be safe I’ll call it design E now.

Each media contestant gets their own ballot race, with perhaps four possible award amounts, each with a voting click-box next to it.  Each voter clicks one amount-box for each media contestant.

Ian is going to check into whether Webvote can accommodate design E.

In case we end up choosing Design D, we reached a consensus on what the award sequence should look like.  We didn’t completely specify the award sequence, so I’ll try to do that now.  As I understood it, we decided on having 8 awards ranging from $500 to $1500.  If we space those evenly, the difference between consecutive awards would be $1000/7 = $142.86 .  So I suggest rounding the step size off to $150 and lowering the 8^th prize to $450 – call this schedule D4:

1st prize: $1,500

2nd prize: $1,350

3rd prize: $1,200

4th prize: $1,050

5th prize: $900

6th prize: $750

7th prize: $600

8th prize: $450

These sum to $7800.

Is that close enough to what we agreed on?

Next I’ll write an email about possible ways we could specify Design E.

-- Mark

From: Mark Latham
Sent: Monday, October 09, 2006 1:26 PM
To: Ian Pattillo; Matthew Naylor; Mike Thicke; Stephanie Ryan; Steven Klein
Subject: VoterMedia Design E options

If we go with Design E, we have several choices to make in specifying it.  First, here’s the general description again:

“Each media contestant gets their own ballot race, with perhaps four possible award amounts, each with a voting click-box next to it.  Each voter clicks one amount-box for each media contestant.”

Choices – we would need to decide:

1. Award options – e.g.

(a) $1500, $1000, $500 – This would be similar to the range of our design-D awards.

(b) I think Matt was suggesting something like $3000, $2500, $2000, $1500, $1000, $500.

(c) A compromise intermediate offering could be $2000, $1500, $1000, $500.

Here we are trading off precision of voters’ wishes versus length of the ballot.

2. Non-vote = $0 vote?

If a voter leaves one contestant’s ballot section blank, is that the same as voting for $0?

Closely related question: should we have a $0 option?

If blank is different from $0, then presumably we should have a $0 option. How would blank i.e. non-votes then be treated?

If blank is same as $0, we don’t need to have a $0 option; in fact, it may make this point clearer to voters if we don’t have a $0 option.

3. How votes are tallied to determine awards –

(a) We could award the median voted amount for each contestant. But we’ve seen the problems with that:  If that adds to more than the prize pool, scale them all down proportionally.  If it’s less than the award pool, then the full pool is not paid out and the remainder goes back to the sponsor.

(b) Instead of the median (50% point), find the level of consensus that exactly allocates the full award pool.  This is similar to my “Design A” that I emailed you on Sept 10.  For a specific example, I’ll suppose here that we chose award option 1(c) above, and non-vote = $0 vote:

- For each contestant, arrange the voted amounts for that contestant in increasing order from $0 to $2000.  Suppose there were 1000 voters.  Then each contestant would have 1000 voted amounts, where #1 is probably $0 and #1000 is probably $2000.

- For each rank #1 through #1000, sum the voted amounts across all contestants.

- Find the ranks #n where the summed amount equals the award pool total.

- Award each contestant their #n amount.

(Because the award options are in $500 steps and the pool would be a multiple of $500, it’s unlikely we’d have to average two ranks #n and #(n+1), but we’d write that possibility into the official rules.)

-- Mark

From: Mike Thicke

Sent: Monday, October 09, 2006 3:23 PM

To: Mark Latham

Cc: Ian Pattillo; Matthew Naylor; Stephanie Ryan; Steven Klein

Subject: Re: VoterMedia Design E options

Would it be too complicated / undesirable to have a "no opinion" option that counted as half a vote for $0? I see this as a playing field levelling device where people who have seen the media source and dislike it are more powerful than those who just haven't seen it. This would make self-promotion less of an issue.

I am somewhat unclear as to the practical differences in payouts that would result in the simple scaling vs. the median adjusting methods.

Could you perhaps give a concrete example where we can see what each system would result in?

From: Mark Latham
Sent: Tuesday, October 10, 2006 11:34 AM
To: 'Ian Pattillo'; 'Matthew Naylor'; 'Stephanie Ryan'; 'Steven Klein'; 'Mike Thicke'
Subject: Non-votes vs $0-votes

I agree that passively not voting on a contestant is a weaker negative signal than actively voting $0.  So counting a non-vote as half of a $0 vote seems reasonable to me, even desirable as a good way to interpret voter preferences.

Some people might think this looks weird and perhaps unfair in some vague way, having half votes.  They might be more comfortable with a clear-cut rule in either direction -- non-vote counts the same as $0 vote, or non-votes are ignored.  Some might even contest the "legality" of half-votes.

I would be OK with half-votes, or with counting non-votes the same as voting $0.  As for ignoring non-votes, I think it would depend on the specific vote counting system we use.  I would be afraid of cases where a contestant was ignored by most voters but strongly supported by a small group of friends/allies/members of that contestant.

We would have to decide this non-vote / $0-vote issue if we choose Design E, but not if we choose Design D.

As for Mike's second point (email below), I'm making an example of scaling versus consensus threshold adjustments, and will send it in a later email.

-- Mark

From: Mark Latham
Sent: Tuesday, October 10, 2006 3:57 PM
To: 'Ian Pattillo'; 'Matthew Naylor'; 'Stephanie Ryan'; 'Steven Klein'; 'Mike Thicke'
Subject: Scaling versus threshold adjustment

Scaling down medians versus raising consensus threshold

This responds to Mike’s question:

“I am somewhat unclear as to the practical differences in payouts that would result in the simple scaling vs. the median adjusting methods.

Could you perhaps give a concrete example where we can see what each system would result in?”

Below I give a specific example to illustrate the difference between these two systems.  Although I’ve tried to keep it as simple as possible, it’s still pretty complicated.  So first, here’s the executive summary:

Fitting the awards to the total pool by changing the consensus threshold (rather than scaling the medians) allocates the awards where more voters want them allocated.  Scaling the medians may shift the allocation away from where most voters think it should be.  We give voters the power to express their choices about how much to award each contestant, so I recommend that we follow their choices as much as possible by changing the threshold instead of scaling.  However, it may not make a big difference so I’m not dead set on this decision.

To focus on these two systems, I have made a very simplified example: only two contestants, 5001 voters (easy to find the median), a $10,000 pool (easy to see proportions), and $1 voting increments instead of $500 (easy to see subtle effects).  I designed the numbers to make the results symmetrical – neither system gives more “inequality” than the other.  You’ll see what I mean below.

In asking voters to choose award amounts for each contestant, we are assuming that voters have some meaningful insight into what levels are appropriate for each contestant.  I believe voters will have this kind of insight, especially with the help of media critiquing each other and advising on award levels.

To illustrate most clearly how voter preferences would work differently in these two systems, I will suppose that the array of voter opinions is quite different for contestant A than for contestant B.  For contestant A, suppose voters unanimously agree that A should get $5100.

For contestant B however, let us suppose that there is a wide range of disagreement.  To get the symmetrical result that I think will help us focus on the main effect, suppose the voter with the lowest contestant-B vote chooses $2808, the next lowest chooses $2809, and so on by $1 steps so the highest contestant-B vote is $7808.  The median contestant-B vote is then $5308.

So the two medians are $5100 for A and $5308 for B, totalling $10,408 -- more than the available $10,000 award pool.  We could scale the awards down proportionally by dividing each by 1.0408, giving an award of $5100/1.0408 = $4900 for A, and $5308/1.0408 = $5100 for B.

Alternatively, we could raise the consensus threshold higher than the 50% median, until the awards sum to $10,000.  Since voters were unanimous in choosing $5100 for A, raising the threshold won’t change A’s award; but B’s award would gradually drop until it reaches $4900, which happens at the 2093^rd-lowest voter = 2908^th-highest voter = 58^th percentile.  So the threshold would be raised from 50% to 58%.

So how can we make sense of this difference?  Which system is better?

Here’s one way to approach the comparison:  Both systems agree that A and B should get at least $4900 each.  The systems just disagree about who should get the last $200.  (That’s the symmetry I designed for.)

Note that this example wouldn’t be affected if 42% of the voters wanted A to get less than $5100.  The final threshold was 58%, so A only needed 58% of that unanimous support.  For the argument that I’ll make next, it makes more sense to say A has 58% support at $5100 rather than 100% support.

Note also that if we use scaling and give $5100 to B, that’s B’s 2293^rd-lowest voter = 2708^th-highest voter = 54^th percentile.

So now compare giving that last $200 to A versus to B.  More specifically, compare giving the last $1 of that $200 to A versus to B.  58% of voters said A should get that $1.  54% of voters said B should get that $1.  Scaling gives it to B.  Adjusting the consensus threshold gives it to A.  The threshold method systematically allocates the awards according to where the most votes are.  (This principle also applies for thresholds below 50%, compared to scaling median awards up.)  This seems to me more respectful of the voters’ decisions, and encourages the media to strive for the broadest consensus support.

A more realistic example would not have such strong unanimity.  But the difference between the two systems would still be driven by the degree of unanimity about A versus about B.

[We would have to decide this scaling-versus-threshold issue if we choose Design E, but not if we choose Design D.]

As my aunt would say, “That’s as clear as a glass of muddy water!”

-- Mark

From: Mark Latham
Sent: Wednesday, October 11, 2006 3:36 PM
To: Ian Pattillo; Matthew Naylor; Mike Thicke; Stephanie Ryan; Steven Klein
Subject: Voter media issues

Here’s an updated summary of where I stand on some of our main issues.  Overall I’m happy with the range of designs we are considering.

1. Lead time:

As you know, I’ve been concerned about the short time frame from now until the January 2007 election.  Nonetheless I’m glad we’re going ahead with it, and we’ll do our best to make it successful.  The challenges will be for potential media contestants to learn about the competition soon enough to build their websites; and for voters to learn about it and form opinions/reputations about media quality.  It would help if the media had time to cover AMS issues (including for example constituency elections) in advance of the main election.  I will most likely offer to sponsor this again for January 2008, when we will have plenty more lead time.

2. Award pool size:

I have some flexibility here, so would appreciate hearing what you think would make sense.  In my August 2 pitch to Council I suggested $10,000.  I am still willing to sponsor that amount if you think anything less would be backing out on my offer.  Lately I’ve been suggesting in the $7500 - $8000 range, having trimmed it a bit because there’s less time for the media to work, we’ve adjusted the rules to make sure the whole pool gets paid out, we’ve set the entry fee to half of what I had suggested, and I’m feeling more likely to offer to sponsor again next year.  You haven’t said much about this issue, so I’m guessing you’re comfortable with the ranges we have been talking about lately.

3. Entry fee:

I’m glad I could at least talk you into $100.  Maybe that will be enough to keep out noise-generating entrants that I would call basically spammers.

4. Design D (one check-box per contestant) versus Design E (four or more amount check-boxes per contestant):

This is a tradeoff.  I like the voter power and flexibility of Design E, which lets them choose award amounts for each contestant.  But I also like the simplicity of Design D – it asks less of Webvote, is easier for voters, and doesn’t look like the media vote is taking more ballot space than the main election.  So I tend toward Design D for this year, and consider something more sophisticated next year.  But if you prefer Design E and can get it into Webvote, that’s fine with me.

5. Secret ballot:

Is Webvote a confidential voting system?  I guess not, if you can vote from any computer connected to the internet.  This opens the danger of vote-selling: you can show someone how you are voting, and receive a payoff for your choice.  Do you know if you can change your vote later?  That would greatly limit the potential for vote-selling.  I guess we’ll have to make do with however Webvote works, but I should at least mention this danger to you.  With cash prizes at stake, this kind of corruption becomes more likely.

6. Election rules:

I support recommending to Council that AMS resource groups and constituencies be allowed to endorse candidates.  I support recommending that slates be allowed.  Both of these changes would would help the VoterMedia competition work better.  And conversely, the VoterMedia competition would help reduce the problems these restrictions were designed to help solve.  But I think we can and should go ahead with the competition even if Council doesn’t make these rules changes.

7. Publicizing the competition:

Although AMS Council won’t make its official decision on launching this until October 25, I’d like to make a preliminary heads-up contact to the Georgia Straight ASAP.  Is that OK with you?  How about other potential media contestants?

Do you plan to launch via an AMS announcement in the Ubyssey soon after October 25?

8. AMS Council meetings October 11 (today) and 25:

It turns out I can attend both of these, so I plan to do so.  I can help answer questions if needed, and keep in touch with Council deliberations on VoterMedia as well as general policy issues.

See you all at our committee meeting, 10 a.m. tomorrow!

-- Mark

From: Ian Pattillo
Sent: Saturday, October 21, 2006 6:26 PM
To: naylor.matthew; mikethicke; stephryan; steven.m.klein; mark
Subject: the report

has been written. Please all have a read through and send me you comments. Do not do this on the document, because I'd rather not read it through that many times. Just make a response and cite the necessary areas. If you rewrite the whole thing then you may send the doument back but please rename it. This goes out to Council on Monday AM. I will be circulating it to Sheldon, Code and the Executive before then, but don't expect them to make any changes at that time (excepting perhaps Sheldon wanting to make the lnaguage more Code-like, even though I said 'Shall' where ever possible). ANyhow, I hope you enjoy and in case that wasn't clear, if you want anything to change you have 24 hours to tell me. Thanks for being the only committee that I have ever seen do anything other than make lousy management decisions. Love Ian

Ian Pattillo

From: Mark Latham
Sent: Sun 10/22/2006 1:04 PM
To: Ian Pattillo
Cc: Matthew Naylor; Mike Thicke; Stephanie Ryan; Steven Klein
Subject: RE: the report

Ian –

I’m impressed that you wrote the report so well in so short a time.

Here are some suggested revisions:

1. I think we’ve now settled on a total award pool of $7850.

2. In the agreement between me and the AMS, I would appreciate some reassurance that if for some reason the funds end up not being used for voter media awards (e.g. Council cancels the contest), they will be refunded to me.  How about adding this to point #1?:

“If the AMS awards a total of less than $7,850 to the Voter Funded Media contest participants, the AMS will refund to Mr. Latham by March 31, 2007 the difference between $7,850 and the total amount so awarded.”

3. I understand and agree that it makes sense for point #2 to clarify that sponsoring and advising on VFM gives me no right to interfere in the election. However, you would agree that I would still be free to express my opinions, in the same way as if I were not involved in this at all, right?  To make it clear that expressing my opinions would not be considered interfering, how about adding this?:

“This condition does not, however, restrict Mr. Latham’s freedom to express his opinions.”

I’ve attached a revised agreement with these suggested changes.

I suggest adding the following second sentence (underlined) to Code Article 14 point 5. This is to prevent demands by participants for more space, multiple links etc.:

5. A central site will be located on the AMS elections website that will list the participants, who may also choose to submit a brief statement and have links to their respective websites. The elections committee will determine the length and format of participants’ content allowed on the AMS elections website.

Article 2 Section 8(c):

I suggest this slight rewording (changes underlined) –

“Neither the Student Resource Groups nor the Society’s Constituencies shall spend money or resources on behalf of any candidates, with the exception of any publication dependent on a Resource Group or Constituency for funding that has entered the Voter Funded Media Contest and abided by all procedures outlined in Section IX: A: Article 14.”

-- Mark

From: Ian Pattillo
Sent: Sunday, October 22, 2006 3:04 PM
To: Mark Latham
Cc: Matthew Naylor; Mike Thicke; Stephanie Ryan; Steven Klein
Subject: RE: the report

Thank you Mark. I am taking all of those changes. The prize money i knew was innacurate but I really just did not want to add it all up.

Ian Pattillo

From: Mark Latham
Sent: Sunday, October 22, 2006 6:44 PM
To: 'Ian Pattillo'
Cc: Matthew Naylor; Mike Thicke; Stephanie Ryan; Steven Klein
Subject: RE: the report

I thought of a couple more points:

1. You could mention the reasons why our proposed award total is now $7,850 instead of the original $10,000:

(a) The entry fee is $100 instead of the originally suggested $200.

(b) Media contestants are no longer required to get 50% voting support in order to get an award.

(c) There is limited time for the media to provide coverage of AMS issues in advance of the election.

2. In Article 14 point #8, the 4^th prize of 1050 is missing its dollar sign.

-- Mark

From: Mark Latham
Sent: Sunday, October 22, 2006 10:17 PM
To: 'Ian Pattillo'
Subject: Questions re payment

Ian –

Two questions about our planned agreement, in particular the main sentence “Once the AMS Council has passed a motion instituting Voter Funded Media for the elections of January 2007, Mr. Latham will be required to transfer $7,850 to the AMS within 5 business days.”:

1. The Council “motion instituting VFM” refers to what’s planned for November 8, not the approval in principle planned for October 25, right?  So if all goes as planned, I’ll have 5 business days from November 8 to make the payment?

2. Is a personal cheque on my Royal Bank account OK, or should I do a wire transfer?

-- Mark

From: Ian Pattillo
Sent: Monday, October 23, 2006 12:03 AM
To: Mark Latham
Subject: RE: Questions re payment

Hey,

I think I will make those adjustments, so thanks again for your input. As for the money, the situation I was trying to create was one where you would only be required to pay if we actually made it happen, obviously, but also one that would commit your funds in such a way that you could not excercise control over the process once it is under way. I think this is something council would expect to see. They don't know you like the committee does so I think we have expect them to be suspicious of your plans here. So you are correct, the november 8th meeting is the beginning of the 6 day time period. I will have to change it to explicitly state that institution is Code changes, not support in principle. And I believe a cheque would be just fine, but I will double check for you. See you,

Ian Pattillo

From: Mark Latham
Sent: Monday, October 23, 2006 8:06 AM
To: Ian Pattillo
Subject: RE: Questions re payment

I agree with your strategy for protecting both sides – makes perfect sense.  It will also reassure the media contestants to know that I have already put up the funds, since they don’t know me either.

Are you planning to go into law?   ;-)

-- Mark

From: Ian Pattillo
Sent: Monday, October 23, 2006 10:04 PM
To: Mark Latham
Cc: AA
Subject: RE: Questions re payment

My whole family has been in law for generations, so I always thought that it would be lame if I ended up there. But then, I did enjoy working on this project.

Ian Pattillo

From: Ian Pattillo
Sent: Tuesday, October 24, 2006 7:56 AM
To: Mark Latham; Matthew Naylor; Mike Thicke; Stephanie Ryan; Steven Klein
Subject: Wednesday

All,

Executive response to the report, like most people, has been cautiously optimistic. Everyone is keen to see this thing in action, but there remain some very real hesitations. I attach the primary questions raised in executive committee yesterday. Some of them we have answers to. The last one I do not. We are, in some sense, giving people the opportunity to get around elections spending guidelines and create massive propaganda campaigns. My faith in the voter wanes when leaflets start falling from the sky. Just because my blog says that Stephanie Ryan is on a massive spending spree to become President does not mean that anyone will read it as opposed to the scatter copies of the Underground on the B-Line home. I think that piece of Code will remain. Assuage my diffidence, committee! And then do the same for council. Ian

Their questions:

• Wasn’t the prize money originally $10,000? 
• Who pays the prize money after this coming election?
• What’s the worst scenario that could happen with Turbo Democracy?
• What due diligence has been done in reviewing the proposal?  Have “experts” been consulted other than Latham?
• If Constituency papers can support candidates, then couldn’t constituencies “buy” the election by diverting money into their paper to support a candidate from their constituency?  Would that lend an unfair advantage to constituencies with a lot of money to spend?
• An example: a Constituency paper like the 432 could spend lots of money and put out lots of issues to sway the election, allowing the candidate they endorsed to get around the spending restrictions, and thus affecting the democratic process.  So I guess this amounts to the question: how can we maintain a level playing field for election finances if Constituencies can spend what they like?

From: Mark Latham
Sent: Tuesday, October 24, 2006 11:07 AM
To: 'Ian Pattillo'; 'Matthew Naylor'; 'Mike Thicke'; 'Stephanie Ryan'; 'Steven Klein'
Subject: RE: Wednesday

The Executive’s questions are reasonable and understandable.  Here are some responses:

1. Wasn’t the prize money originally $10,000?

[From my Oct 22 email:]

The reasons why our proposed award total is now $7,850 instead of the original $10,000:

(a) The entry fee is $100 instead of the originally suggested $200.

(b) Media contestants are no longer required to get 50% voting support in order to get an award.

(c) There is limited time for the media to provide coverage of AMS issues in advance of the election.

[From my Oct 11 email:]

Award pool size:

I have some flexibility here, so would appreciate hearing what you think would make sense.  In my August 2 pitch to Council I suggested $10,000.  I am still willing to sponsor that amount if you think anything less would be backing out on my offer.  Lately I’ve been suggesting in the $7500 - $8000 range, having trimmed it a bit because there’s less time for the media to work, we’ve adjusted the rules to make sure the whole pool gets paid out, we’ve set the entry fee to half of what I had suggested, and I’m feeling more likely to offer to sponsor again next year.  You haven’t said much about this issue, so I’m guessing you’re comfortable with the ranges we have been talking about lately.

2. Who pays the prize money after this coming election?

I don’t think anyone would want to commit to future years until we see how the first election with voter-funded media goes.  That said however, I am pleased with how our plans have been developing, and I believe the benefits of VFM will grow with time and experience as everyone learns how it works.  The benefit this year should also be enhanced if participants expect it to happen again in the future.  That gives media more incentive to invest in their reputations, foregoing sensationalism in favour of giving insight that will still look good in retrospect a year later.  So I expect to be willing to sponsor it again for the January 2008 AMS election.  Of course we’ll have much more lead time then too.

After January 2008 however, I would hope to step back and see if the voters like VFM enough to want to pay for it from general AMS funds.  Another possibility could be to find another sponsor.

3. What’s the worst scenario that could happen with Turbo Democracy?

My biggest concern is if apathy is too pervasive for VFM to make an impact.  I don’t think we’ll have trouble getting at least 8 media contestants, but maybe we won’t be able to get enough students to read the media websites and vote on awards.

The scenario about Constituency papers is addressed in #5 below.

4. What due diligence has been done in reviewing the proposal?  Have “experts” been consulted other than Latham?

I’ve had a wide range of feedback on these ideas in the past 10 years.  I’ll outline it in a separate email later today.

5. If Constituency papers can support candidates, then couldn’t constituencies “buy” the election by diverting money into their paper to support a candidate from their constituency?  Would that lend an unfair advantage to constituencies with a lot of money to spend?

- An example: a Constituency paper like the 432 could spend lots of money and put out lots of issues to sway the election, allowing the candidate they endorsed to get around the spending restrictions, and thus affecting the democratic process.  So I guess this amounts to the question: how can we maintain a level playing field for election finances if Constituencies can spend what they like?

One possible strategy for avoiding this danger could be to allow Student Resource Groups and Constituencies to favour or endorse candidates only on website pages, and require that each of those pages have a link at the top to the AMS Elections VFM page.  (This link requirement might be a good idea for all VFM contestants.)  They could advertise their web pages in their print publications, but limiting endorsements to those web pages would put them on a more equal footing with all the other VFM contestants’ endorsements.  Web pages are almost costless to put up, and the links to the main VFM page would encourage voters to read the other contestants’ views too.  This way we would get the benefit of those papers’ reporting and writing talents, without giving them an unfair advantage over less well funded media contestants.

If you want to do that, you could revise Article 2 Section 8(c) to become:

 “Neither the Student Resource Groups nor the Society’s Constituencies shall spend money or resources on behalf of any candidates, with the exception that they may enter the Voter Funded Media Contest and, if so entered, favour and/or endorse candidates on worldwide web pages only, provided that the top of each such web page contains a clearly indicated link to the main AMS Elections Voter Funded Media Contest web page.”

-- Mark

From: Mark Latham
Sent: Tuesday, October 24, 2006 2:19 PM
To: 'Ian Pattillo'; 'Matthew Naylor'; 'Mike Thicke'; 'Stephanie Ryan'; 'Steven Klein'
Subject: RE: Wednesday

Here are some responses to the questions “What due diligence has been done in reviewing the proposal?  Have “experts” been consulted other than Latham?”:

Our VFM proposal is the latest evolution of an idea that I have been publishing since 1997.  During that time I have had a lot of feedback on it, some positive, some negative – samples of both are linked below.  Given the wide range of reactions plus the fact that it has not yet been implemented, no one can really be sure whether or not it will be beneficial.  But it has been reviewed by enough smart people to at least give assurance that it is based on reasonable arguments.

Publications: see lists at www.turbodemocracy.org/publications.htm and www.corpmon.com/publications.htm (these overlap quite a bit).  Most of my papers were reviewed by editors and/or referees in order to qualify for publication. For example, Corporate Governance: An International Review and Financial Analysts Journal are well respected journals in their fields.

Presentations: see list at www.turbodemocracy.org/presentations.htm .  I list not only the institutions but also the individuals who hosted my talks, so that people can contact them and ask about the quality of past presentations.  Note in particular the repeat invitations by UBC Sauder School Professor Kai Li (3 times), U C Berkeley Haas Business School Professor Nils Hakansson (5 times), and Santa Clara University business school Professor Atulya Sarin (3 times).

On the negative side, I’ve taken lots of flak from corporate boards and their legal hired guns when I have submitted VFM-type proposals as a shareowner.  A typical example is at www.corpmon.com/OS-debate.htm.  There’s plenty more linked from www.corpmon.com, including in Microsoft’s current proxy (search there for “Latham”).

Further confirmation that this type of proposal makes some sense can be found in related publications by respected researchers.  I review many of these in my paper “Turbo Democracy” (soon to be retitled) at www.turbodemocracy.org/publications.htm.  Prominent among them is the 2002 book “Voting With Dollars” by Yale professors Ackerman and Ayres.  It proposes letting each American voter direct $50 of public funds every 4-year election cycle to political candidates and related organization campaigns.  Of course, I criticize it for being a voucher system; and it focuses on campaigns, not media.  Nonetheless, it provides some experts’ confirmation of the idea that letting voters direct public funds could improve a democracy’s political information process.  The book was so well received that the California Law Review and the University of Richmond Law Review each devoted a special issue to articles discussing it.

In the field of media reform, leading guru University of Illinois Professor Robert McChesney (see www.freepress.net/content/faqs#question15) favours a proposal by economist Howard Baker to let each voter allocate $100 (per year, I think) to nonprofit media.  Detailed reference is on page 10 of “Turbo Democracy”.  Again, a voucher system and therefore subject to abuse and countervailing regulation, but otherwise broadly similar to my proposals.

Neither Ackerman & Ayres’ nor Baker’s proposal has been implemented as far as I know.  We can be the first!

UBC political science professor Yves Tiberghien gave me these helpful comments on an early draft of “Turbo Democracy”:

From: Yves Tiberghien

Sent: Wednesday, October 13, 2004 11:10 AM

To: Mark Latham

Cc: Andrew Eggers

Subject: Re: Turbo Democracy revision attached

Dear Mark,

I actually read carefully the Oct 5 draft in the plane and had some comments that I am writing here below.

Overall, I like the reach, the ambitious agenda, and the very creative

ideas. This is a very exciting research agenda, with some neat thoughts.

On the improvement side, three main concerns came out for me:

1/ the paper mixes 2 different issues: information to improve voting

efficiency/effectiveness, and flows of funds (avoiding capture by private interest groups). Given the core focus on Infomediaries and Turbocharger, it seemed to me that the additional points on funding (eg in pp 4-5) was a distraction (including the Canadian bill on party funding). Information and party funding in elections are very different issues.

2/ the core of your argument is in advocating the creation of infomediaries in different settings and elections. I think that the paper needs a full theoretical section on how infomediaries work. The key concern that you have not yet addressed is how you regulate these intermediaries. In effect, you create a new agent and thus can be caught in a classic principal-agent problem. How can you ensure that infomediaries are not captured by key politicians or interest groups? How do you make sure they do not exploit their position of power for their own private advantage? Intermediary agents in positions of power, typically, end up drifting into capture or private pursuit of profit.

There can also a more general discussion of the creation of additional interfaces in the democratic process. Key theorists have long argued for the need of immediacy, ie. The removal of interfaces in the democratic process, since intermediaries can create noise, transaction costs, or distortions. For example, will infomediaries be representative of the average voter or will they represent an elite with preferences that are different from the median voter? These are classic problems in democratic theory.

3/ Finally, at the outset you jump from corporate governance to the

political field, without pushing further the key difference between the 2 fields, namely, that shareholders bring capital as well as votes. In fact, corporate governance includes 2 separate dilemmas: managers vs shareholders, but often more importantly majority shareholders vs minority shareholders. Medium or large shareholders have means at their disposal that voters in the political system don't. In many corporate cases, the shareholder college is a concentrated one, whereas voters in the political system are more numerous and more fragmented. There could be more discussion of those key differences and how you overcome them in making the case that the similarities are more important than the differences. I was not convinced enough yet.

4/ on details: the abstract is a bit confusing and unclear. The conclusion is good and includes in particular a good definition of turbo democracy. This should be in the abstract and the introduction (first paragraph of conclusion)

Have a great day!!

Cheers,

Yves

---------------------------------------

The paper and the proposal design have evolved considerably since then.  In 2004 my design called for voters to allocate funds to media one year in advance of the election the media were supposed to cover.  In 2005 I got the idea of making the media take the risk of providing coverage first in the hope of getting paid for it afterward – harder for the media, but much easier for voters to decide which media deserved funding.

In large part to respond to Yves’ questions, I wrote section IV of the current version of “Turbo Democracy” as a theoretical section on how these media would work.

-- Mark

From: Mark Latham
Sent: Thursday, October 26, 2006 11:53 AM
To: Ian Pattillo; Matthew Naylor; Mike Thicke; Stephanie Ryan; Steven Klein
Subject: Make it an even $8000?

Seeing how pathetic it sounds to say our award pool is $7850, how about rounding up the three awards that ended in $50 to the next $100?  That would make the pool an even $8000, and each award in round 100s.  With 8 awards, the average award would be $1000.

So instead of:

We make it:

Although this makes the spacing a bit uneven, that seems pretty minor.  This award schedule is what would have resulted from making 8 equally spaced awards from $500 to $1500 (step size = $1000/7 = $142.86) and then rounding each award to the nearest $100.

What do you think?

-- Mark

Contest rules 2006-2007 resulting from the above discussion