31 Replies to "Never mind Iowa - OUR caucuses are just weeks away"

• TheVelvetBulldog January 3, 2008 (8:05 pm)

  Thanks for the info – I’ve decided I’m going to my first caucus!

• WSB January 3, 2008 (11:47 pm)

  They are a “don’t miss” event. We always bring The Junior Member of the Firm, dating back to his infancy – he also comes along when we vote; trying to set an example. (And it’s paid off; he is paying attention to politics and government at quite a young age – if we had a personal blog, we would have had a field day with his observations and rants during tonight’s Iowa TV coverage tonight!)

• Jan January 4, 2008 (12:09 am)

  When you go to the 34th DD Caucus link in the article, there is a menu on the right side of upcoming events. In particular is the event on Jan.23, 2008…drinkingliberallyburien at Mick Kelly’s Irish Pub….sounds good to me ;-)

• Bob January 4, 2008 (12:24 am)

  Averaging the current head-to-head polls shows McCain as the strongest Republican candidate now, beating Clinton but not necessarily Edwards or Obama.

  Political machines are working overtime to anoint Clinton, who has the keys to more of the machines, but possibly not to the White House if McCain is the Republican nominee.

  All this could change. But it does matter who the Washington state caucuses choose. More people turned out in Iowa than usual this year. Let’s hope the same is true of Washington.

• SA January 4, 2008 (9:00 am)

  As an independent I must say I feel disenfranchised by this whole caucus/primary process. I have two candidates I’d like to support (one Dem, one Rep) yet because of my independent status I’m prevented from participating in either caucus. Moreover, if I were to vote for a Dem in the primary that vote will be completely ignored. To protest I will only vote for 3rd party candidates in November.

• JE January 4, 2008 (9:39 am)

  SA–This is why we need RCV (Ranked Choice Voting) or IRV (Instant Runoff Voting). Pierce County will be using it this year for county offices, so for those offices they won’t require a primary (saves money on elections, too!) What they’ll be able to do is rank their choices for an office (1st, 2nd, 3rd, etc.) for all candidates for the office, not just those in a single party. It lets a voter vote for his or her favorite candidate without “spoiling” the election, and opens up elections to third-party candidates who may appeal to voters in both of the major parties. And, it lets you, SA, vote for whatever candidates appeal to you, regardless of party.

  So, SA, I suggest you declare yourself a Democrat (the Republican machine appears less likely to support this. The Dems at the top don’t like it (they don’t want to give away power), but there’s quite a lot of interest at the lower levels.) Write up a resolution, introduce and argue for it, and if enough people do this at enough precinct caucuses, the issue gets pushed to the next level. If it comes up year after year, it helps eventually get the party leadership to take it seriously (as in Pierce County).

  Also, the King County Charter is under review now. King County could use RCV or IRV, just as Pierce is doing. Lobby your council members–now’s the time to make it happen. I know that’s just county races, but when enough counties use it, the state will have to consider it (and ditch the Pick-a-Party primary–yay!), and when enough states use it, the federal races will be forced to. And then you can rank all the candidates you like, from whatever party.

• Kayleigh January 4, 2008 (10:14 am)

  Not to be petty and cranky but it would be nice if my precinct were named something I could actually remember, rather than a meaningless (to me) number. Now I have to find my voter registration card to look it up…sigh.

• WSB January 4, 2008 (11:03 am)

  I’m trying to find an easy search utility for that online. So far, just got some unnavigable maps at King County Elections. Can’t believe there’s not a “type in your address, get your precinct” somewhere … if we find it we’ll post it here (or maybe someone else knows of one).

• RS January 4, 2008 (11:15 am)

  There is a search utility for King County. It’s how I found my polling place.

  https://www.metrokc.gov/elections/pollingplace/voterlookup.aspx

• Charla Mustard-Foote January 4, 2008 (11:28 am)

  https://www.metrokc.gov/elections/pollingplace/voterlookup.aspx

  This tells you much more than you want to know, but I thought it was kind of neat. And it does include your precinct number. (I’m still looking for the version that allows you to type in an address and get a precinct.)

• SA January 4, 2008 (11:37 am)

  I used this to find my precinct: https://www.metrokc.gov/elections/pollingplace/voterlookup.aspx

• WSB January 4, 2008 (11:51 am)

  thank you all. we pride ourselves on generally killer search skills but somehow couldn’t find that to save our lives. will add to the main post too.

• JanS January 4, 2008 (11:52 am)

  from your post on Dec. 19…click on this link…put in your name and birthdate, and get your voter information, including precinct # :)

  https://www.metrokc.gov/elections/pollingplace/voterlookup.aspx

• WSB January 4, 2008 (12:04 pm)

  Oh great, not only are we losing our search talents, we’re also suffering amnesia. Going to pull away from the computer now for some in-person coverage, maybe that will clear the mind.

• Kayleigh January 4, 2008 (12:08 pm)

  Thank you…that worked (of course, I had to enter the usual misspelling of my name…)! See all y’all at West Seattle High.

• SA January 4, 2008 (12:29 pm)

  Thanks, JE. While it does appear RCV or IRV would increase interest in 3rd party candidates (which I personally would favor) I just don’t see the Dems or Reps willing to relinquishing any power.

• Ken Davis January 4, 2008 (12:58 pm)

  JE has a good point. But resolutions, I have heard, will not be haggled over this year in the precinct caucus. Get ahold of the Dem PCO for your precinct and they might even help you word a resolution that they or the delegates elected at your caucus can carry and submit at the district level.

  RCV has pros and cons but the grassroots of our local party are a lot of greens and progressives who often buck the more conservative state party on these kinds of issues.

  But choose Dem this year for the caucus if you want to have a voice at all in presidential politics. We get ignored in WA when the general rolls around.

• Ken Davis January 4, 2008 (2:07 pm)

  Ok some more useful links:

  Precinct maps for south seattle (pdf)Hint, zoom in.
  http://www.metrokc.gov/elections/gis/maps/cities/seapg3.pdf

  34th Dem caucus page
  http://www.34dems.org/news_caucus.htm

• Todd in WEstwood/S.Delridge January 4, 2008 (7:16 pm)

  (said like Bevis would) “heh, heh, heh, He said caucuses, heh heh heh…..ummm welcome to Burger world”

• Bob January 6, 2008 (11:51 am)

  About instant runoff voting, as in Pierce County – there’s an interesting (although a bit dorky and mathematical) explanation of range voting at rangevoting.org. Approval voting (approvalvoting.org) is the same but simplified.

  Both are mathematically a little better than IRV if some voters actively try to outwit the process. IRV, though, does look a lot better than the present primary voting mess.

• Clay Shentrup January 6, 2008 (7:14 pm)

  Bob,

  Range Voting and Approval Voting aren’t “a little” better than IRV, they’re enormously better.

  http://rangevoting.org/vsi.html
  http://rangevoting.org/CFERlet.html

• Bob January 7, 2008 (3:29 am)

  Clay,

  This is interesting. But I did say that the comparison I gave was true when only “some” voters gamed the system, not when all or nearly all of them do so. Your conclusion appears to be increasingly true when a great majority, or all, of the voters play to game and subvert the election.

  My limited understanding of this comes from such analyses as Brian Olson’s at

  Brian Olson: Voting Graphs with 10,000 Voters – Average Happiness.

  His first chart in that series compares voting methods when no one attempts to subvert the process (OK, let’s skip that one). The third chart shows the results when 17% of the voters try to subvert the election (that could be about right). The fourth chart shows the results when 71% of voters subvert the voting (but you’d have to prove to me that 71% of voters really do cast their votes primarily to subvert the honesty of the election process).

  If I’m interpreting Olson’s graphs correctly (and I may not be), then with 10 candidates in a race and 17% of the voters trying to be evil, one-vote voting produces about 45% of the average voter happiness of range voting, IRV about 75%, and approval voting about 85% of that of range voting.

  If you assume that every or nearly voter casts votes evilly, then range voting looks better and better compared to IRV. But at lesser levels of voter misfeasance, the difference seems to me (though I am far from a mathematician) to be much less pronounced.

• Clay Shentrup January 7, 2008 (11:46 am)

  Bob,

  Thes pages look at asymmetric strategy:
  http://rangevoting.org/StratHonMix.html
  http://rangevoting.org/ShExpRes.html

  Here are some of the “Yee” diagrams:
  http://rangevoting.org/IEVS/Pictures.html

  Warren Smith’s simulations are more extensive than Bolson’s I believe. (Also I think Smith has more expertise.)

• Rob Richie January 7, 2008 (12:44 pm)

  IRV has been used in many very important, hotly contested elections –for president of Ireland, for parliamentary elections in Australia, mayor of London, a growing number of US elections and many organizational elections. The alleged problems with gaming the system just don’t emerge. The basic problem is that it’s next to impossible to know how to try to “beat the system” without knowing how everyone else is voting

  Meanwhile, approval voting has very little history to point to, yet that limited history is not a good one for its advocates. The problem is that it’s OBVIOUS to figure out to help your candidate — you truncate your vote to one candidate. If you vote for two, you may hurt your first choice. Approval voting typically devolves into plurality voting or allows some insincere voters (who only vote for one) to defeat sincere voters (who vote for candidates in addition to their favorite).

  Approval voting advocates like Clay Shentrup tend to be very zealous in their advocacy of approval voting, but until they can show it can be won for anything significant, they essentially act as agents of the status quo by attacking that’s winning and working. See http://www.fairvote.org/irv and http://www.instantrunoff.com for more if interested.

• RJ January 7, 2008 (6:34 pm)

  While almost any change will be an improvement to our current system, RCV seems to be better than range. The problem with range is that all the supporting evidence is based upon unrealistic computer simulations, not the real world. Here’s a real world scenario that range can’t handle. I’m a progressive Democrat so I really dislike Clinton among the Dems but I like her better than all the Republicans. If Clinton is the front-runner then how do I cast my vote? I’ll give Kucinich and Edwards 10 points but then I don’t know what to do because it depends on what everybody else is doing, which I don’t know. I’d like to give her, maybe, 1 point. But then I’d worry that I’m helping elect a Republican. So do I have to give her 10 points to prevent this? In other words I won’t be able to vote honestly using range. I’ll be forced to express support for someone I don’t like just to prevent someone I really don’t like. Sound familiar? It’s the problem we have now.
  RCV is better. I’ll go Kucinich 1st, Edwards 2nd, and so on.

• Bob January 7, 2008 (7:57 pm)

  Clay,

  In checking Warren Smith’s comparison between range voting and IRV, on his page entitled

  WHY RANGE VOTING IS BETTER THAN IRV (Instant Runoff Voting),

  I admit to being troubled by how easily he is willing to assign winning ratings to lesser known candidates who have smaller followings. In the example Smith gives there, two candidates who received a total of 540 and 471 points from all nine voters were defeated by a candidate who received only 372 votes from six voters, but whose substantially smaller number of followers were on average slightly more favorable toward him.

  This shows up even more on some online review sites, where various little known items receive five stars from one or two devoted adherents, and thus appear atop the Best Rated lists while other items with 500 votes and an average of four stars appear far down the page in the ratings.

  Even though I don’t understand all of the “happiness math” surrounding range voting, after reading Prof. Smith’s comments I believe it might be prudent to introduce a dash of “uncertainty math” or “risk math” regarding rating candidates who receive votes from fewer voters, especially from substantially fewer voters. Prof. Smith in this example simply fills in every blank vote with the average of all of the other votes received by a candidate, whose name a sizable chunk of voters left a blank for some reason. That’s increasingly less certain to be a valid measure, as the number of voters voting for a candidate decreases.

  Some of the better online review sites apply a scaled statistical moderation factor to lesser vote totals, so that a five star rating from one or two voters counts far less than a four and a half star average rating from hundreds or thousands of voters. Some sort of simple correction for the increased risk involved in electing candidates with smaller coteries of adherents would seem worthwhile to consider.

  In other words, I like the general idea of range voting, but want to see its results clearly laid out and compared with one-vote voting and IRV when applied to real world elections such as are happening this year in the pesidential primaries.

• Clay Shentrup January 7, 2008 (10:39 pm)

  Bob,

  The problem you speak of is fixed with a quorum rule, so that a candidate can only win by getting half as much total score as the candidate who got the most. That way you can’t win by getting high scores from an extremely small following.
  http://rangevoting.org/RuleD.html

  Also, I don’t know how you got to Warren Smith’s home page instead of the Center for Range Voting, http://rangevoting.org/

  Smith simulations used 720 combinations of 5 fundamental parameters (“knob” settings), and hundreds of thousands of trials were averaged for each of those combinations. Range Voting won in all of them, hugely surpassing IRV. Out of all those models, some of them surely had to closely approximate our actual political reality.

  You can see lots of real world “funny” elections here:
  http://rangevoting.org/FunnyElections.html

• Bob January 7, 2008 (10:46 pm)

  Clay,

  Let me simplify this and present a graph that illustrates it. Draw a graph with the horizontal axis representing the percentage of voters who voted for each candidate, from 0 to 100%. Let the vertical axis represent the range score for each candidate, from (say) 0-5. These are both linear scales.

  No candidate should finally win an election who gets less than half the vote. With range voting, this could be either 100% of the voters voting halfheartedly for the candidate (a range score of 2.5 out of 5), or 50% of the voters voting with complete enthusiasm for the candidate (a range score of 5.0 out of 5).

  On the vertical line at the far right of the graph, representing 100% of the voters, plot a point halfway up, at 2.5. This would be the score of the most mediocre candidate who could win if everybody votes.

  At the midpoint of the horizontal axis, draw a vertical line representing 50% of the voters. At the top of this line, plot a point at 5.0.

  Now draw a straight line between these two points.

  No candidate whose point total falls below and to the left of this line should ever win an election. Any candidate whose point total falls above and to the right of this line could conceivably win an election.

  To determine the winner, one plausible method would be to find the candidate whose vote total (range score) is the furthest above that line.

  Possibly I’ve missed something essential, but this should make my earlier comments clearer.

  Comments?

• Clay Shentrup January 8, 2008 (12:51 pm)

  No candidate should finally win an election who gets less than half the vote.

  We disagree. We say no candidate should win without getting half as much total score as the candidate who got the most. In your system, say there were 3 candidates, if people who would ordinarily give their least liked candidate a 0 would decide to not score him at all, they could easily disqualify him, even if he should have won according to his score. This reasoning is explained at the link I gave you. A lot of people who spend hours of each week discussing and studying this issue have come to a strong consensus about this.

  With range voting, this could be either 100% of the voters voting halfheartedly for the candidate (a range score of 2.5 out of 5), or 50% of the voters voting with complete enthusiasm for the candidate (a range score of 5.0 out of 5).

  On the vertical line at the far right of the graph, representing 100% of the voters, plot a point halfway up, at 2.5. This would be the score of the most mediocre candidate who could win if everybody votes.

  I don’t know what is special about 2.5 points. On a 1-5 scale, the center point would be 3 points. But that doesn’t have anything to do with the quorum rule I described. It says you have to get at least half as many total points as the candidate who got the most. So a candidate who averaged a 3.6 out of 100 voters got 360 points. So in order to be valid, another candidate would have to have at least 180 total points, even if he had a better average.

• Bob January 8, 2008 (5:28 pm)

  Clay,

  You’re right about the center point being 3 points.

  The difference between my preference about where to draw the dividing line between viable winning candidates and minor also-rans and your preference for setting a lower bar for that, is that mine is much closer to existing election policy.

  All the other features of range voting would continue to apply – letting the voter rate multiple candidates, using a scale of 0-N instead of 0 or 1, letting voters leave as many or as few candidates blank as they wish, not forcing them to rank every candidate, and so on.

  Where to set the bar for winning is our main area of disagreement, I believe. My preference for a clearer winner than you’d accept also means that it I would rather have a conventional two-candidate runoff election if no winner clearly enough emerges in the range voting. But that’s not a bad thing, because you can’t always reliably extrapolate a low-vote-count winner to the real winner of a two-candidate runoff election.

  In short, I believe it’s not always possible to successfully winnow a large field of semi-unknown candidates in one instant fell swoop, and if there are no sufficiently high (i.e. trustworthy) vote totals to produce a clear enough winner that enough people have voted for, then a separate runoff afterwards, to make the result very clear, is the best answer to that.

  A less good answer might be to use the rating information within the range vote data to do something similar to what IRV does with its ranking of runoff candidates, without holding a separate runoff election. In that way, range voting would more resemble IRV, but with the added advantages of range voting.

• Clay Shentrup January 10, 2008 (8:55 pm)

  Bob,

  Like I said, a lot of smart people (Princeton math Ph.D. & company) have put a lot of time into this and have found serious flaws in alternatives like what you’re suggesting. For one thing, the “abstain” option is an added feature that is merely there to help reduce the harm of ignorance and to make it “safe” for voters to only rate the candidates they care about (makes voting faster). If you use an abstention quorum rule like the one you propose, serious problems could occur; it could be advantageous to not rate someone instead of give him a 0, even if you knew enough about him to rate him. Or in a race with 3 candidates or more, everyone could fail the quorum test, so there would be “no valid candidate”. That just doesn’t work.

  If you want a runoff when there’s no clear winner with Range Voting, that’s fine. That would only happen if there was a tie, in which case you have to have some kind of tie-breaker anyway. A runoff actually leads to worse results if voters are fairly honest, although it can help prevent certain types of strategic behavior, and so may be a theoretical improvement. Except that the expense of it makes it an overall negative in practice.

  In short, I believe it’s not always possible to successfully winnow a large field of semi-unknown candidates in one instant fell swoop

  Then people are free to research the top 2 or 3 front-runners and score them and ignore the rest if they like. It sounds like you’re making the too many choices argument, which I think is flawed for reasons explained at that link.

  If you’d like to discuss this more with people who have more time to get down to specifics, check out the Range Voting discussion list:
  http://groups.yahoo.com/group/RangeVoting

Sorry, comment time is over.