Thursday, October 6, 2011

Oy, Bai

I know, fish in barrels and I was going to let this one pass, but really:
Even the circuitry of the democracy remains essentially unchanged; a nation of voters who can find their cars and pay their mortgages online still can’t envision the day when they can cast their votes from an iPad.
OK, it's true that we don't vote from an iPad -- but actually voting methods are completely different than they were, say, 30 years ago. Back then, the mix of technologies was heavy on old-fashioned voting machines (which were, by the way, absolutely great) and the infamous Harris Votomatic type systems, which worked so well in Florida in 2000.* Now? We have mostly optical scan, last time I checked, not to mention that we have vote-by-mail in some places, and early voting in's basically a completely different system than in 1980.

In fact, a reporter covering electoral politics should know that there's actually right now, going on in a whole bunch of states and in the Obama Justice Department, a huge fight over how we're going to vote in 2012 and going forward. It's sort of a big deal.

Oh, and while I'm at it, there's this:
Either we’re being told that centralized, 20th-century systems can never be changed to accommodate more individual flexibility (like say, decoupling health care from employment), or we’re being told that all federal programs are wasteful and that every American should basically fend for himself. Either we’re supposed to rely entirely on large institutions, or we’re supposed to rely only on ourselves.
Is one of these options supposed to be the ACA, which deals with that decoupling by making health insurance choice feasible for those without employer-based insurance? Really? Or did Bai miss that one, too?

I mean, he is supposed to be following politics and government, isn't he?

*Mr. Harris's money made my graduate school years a lot more tolerable/enjoyable, so I'm obligated to point out that the Harris Votomatic was the greatest thing since sliced bread and absolutely terrific in every possible way.


  1. For those who are wondering what a "Votomatic" is, it's in the Smithsonian:

    I, too, was a passenger on the Votomatic gravy train during grad school in the '80s, in a different sense: I worked as a consultant / organizer helping media organizations compile usable totals from the thousands of these gizmos then operating in every village and hamlet in the great state of Illinois (except a few where they were still marking X's on slips of paper). The mess that was Florida '00 was previewed in Illinois '82, when the governor's race hung for a while on Votomatic ballots that somehow got wet and had to be dried out under heat lamps. But yeah, I know, that was no doubt the fault of the paper manufacturer. ;-)

  2. Hey, that's a much better link than the one I found.

    Here's a bio of Harris:

    I was trying to find an old picture from the Harris Room, but my search skills are not good enough.

  3. IBM has an archive too, featuring a photo of a Votomatic actually in use, in what I would guess from all the hairspray is about 1965:

    Apart from the memories of '00, this is a fun read as well:

  4. Duh, make that '64, when there was, you know, actually an election.

  5. That votomatic that's at the Smithsonian?

    We found that in a closet at IGS while I was there. It was an old prototype. I can actually say with no exaggeration that I have played with one of the very first machines to make George W. Bush president when he actually lost.

    No, not punchcards themselves as a technology, but the Votomatic making a butterfly ballot an attractive option to Theresa LaPore. The punchcards are more questionable, but there's absolutely no way 3000 people INTENDED to vote for Pat Buchanan in Palm Beach County.

  6. Really? I'm pretty sure I remember that machine. Saw it, don't remember playing with it. It's the one in the Smithsonian? How about that.

    FWIW...I've always assumed that the problem was that it was misaligned, not that the butterfly design confused people; it's certainly possible for the arrow to be pointing to the wrong spot on the card.

  7. No, it was the ballot design. There's a picture of the ballot here:

    The explanation given on this site is quite good, maybe definitive:

    "Not only does it appear that perhaps 4.000 people made the error of punching the second hole on the ballet in the mistaken belief that the second hole represented the second candidate, more than 19,000 people made the error of punching more than one hole, since both were directly alongside their candidate.

    "This is yet another disgraceful example of what happens when you don't bother to user test.

    "Why were the people who laid this out unable to see the problem, even without testing? Because they were not users, they were designers. As such, they were interested in all 10 candidates on the ballet (plus space for a write in), and they saw all ten candidates. They viewed the ballot as a 2X6 staggered matrix with a line of radio buttons in between the two sides. Their cute little arrows appeared to be enough to help people choose the right box from this matrix.

    "The voters saw things very differently. They were not interested in 10 candidates. They were interested in one candidate, the one they wanted to vote for. Their entire focus was on finding that candidate and punching the hole next to his or her name. In the case of Gore, that required scanning only two names down in the first column. There was never any reason at all for Gore voters to ever even see the right hand column, and we now have thousands of pieces of evidence that they, indeed, didn't. Rather than a staggered matrix, they saw a single column with a dedicated column of radio buttons adjoining."

  8. OK, another comment of mine has posted and then disappeared. It's a good thing I'm not the paranoid sort. :-)

    Anyway, the problem wasn't alignment, it WAS the design of the Palm Beach ballot -- as ably explained and illustrated here (scroll down to the image and the paragraphs immediately following):

  9. Jeff: you probably posted using a Dieblod computer.

  10. Yeah, Matt, I guess I shouldn't have titled the post "A vote for Gore."

  11. Jeff,

    Sorry, for some reason the spam filter got it, but I've restored it.

    On the substantive point: I don't find that definitive at all. It's certainly a plausible explanation, but IMO no more plausible than misaligned machines. I've definitely used misaligned machines, in which the (correct) arrow is pointing halfway between two spots, and had to pull the ballot out and visually find the correct spot to punch. It's entirely plausible to me that people faced with that might punch the wrong hole, or (more foolishly) punch the two holes that appear equadistant from the arrow. IIRC, and I'm pretty sure I do, you can't actually see the printed numbers on the ballot paper when it's in the machine, so if it's misaligned you just have to guess.

    An analysis of the whole ballot could probably determine which of those problems is happening, but I don't believe that any of the late analyses looked at sub-presidential votes.

  12. Jeff, the link you provided for the butterfly ballot is pretty mind-blowing. I've been secretly fascinated by that ballot for some time, and I don't know why it doesn't get more play in discussion of Florida 2000. But I didn't realize until I read your link what happened in 1996.

    The author asserts that Dole, who was in the unenviable Gore position in 1996, was victimized by 14,000 spoiled ballots. What's 14,000 votes among friends? According to some website I just randomly found, Dole got 133,000 official votes in Palm Beach in 1996, which means that voter confusion cost him about 10% of his vote total in that county.

    Think about that. The Bush era is entirely and wholly attributable to a flawed ballot design that, one Presidential election previously, had fucked up fully 10% of those intending to vote for the candidate on the Gore line. That empirical example obviously bothered none of the professional Palm Beach voting officials, or if it bothered them, certainly not enough to do anything about it.

    You know, even if The Entire History of the World didn't hang in the balance of their decision to stick with that crap ballot after the bad 1996 experience, I'd still be gobsmacked that a 10% error rate accruing to one candidate didn't automatically cause them to change course for future elections.

    But when you consider the implications of their inertia? There aren't words. There just aren't words.

    Well, there are a few words - people can be really really really stupid.

    (By the way, Jeff, if its any comfort I lose comments back here from time to time as well. I usually figure I've been unintentionally offensive, which considering my political views, at times poor judgment, and the prevailing sentiment of this audience, I'd expect to happen about 25% of the time at least).

  13. ...and the spam filter grabs CSH's comment, too. Restored now, I hope.

    Actually, I think in the history of the blog I've zapped under half a dozen comments that were not clear spam, and I doubt if I've ever done it for a regular commenter -- they were all drive-bys that were highly offensive in some way.

    The blogger spam filter is really won't mistakenly take anything for weeks, and then suddenly it'll grab a few, as it did for whatever reason in this thread. I have to manually check it to restore mistakes, and I sometimes remember to do it regularly and I sometimes don't. So if anyone sees a comment disappear, just email me and I'll check on it.

    (I've had a couple times where I was negligent and wound up restoring comments several weeks after they were autofiltered out. Sorry).

  14. Guys,

    First, thanks for the spam-reversing. I think what the comments of mine that have disappeared have in common was usually something involving html or a quote from another site or something. So I think it's likely something hidden in the coding.

    CSH, yes, I didn't know about the Dole error either until I read that account. I agree, it's amazing to think that the most influential yet unknown American of this century might be Theresa LaPore, the Palm Beach County Clerk who OK'd that ballot.

    And JB, as to whether it was bad design or misalignment, look, it's bad design either way. You're supposed to be designing a ballot FOR THOSE MACHINES. That means it's supposed to take account of how they function (or might malfunction) and is supposed to work correctly in several hundred of them, even if all the machines aren't in equally good repair, as they inevitably won't be. I don't believe that all or most Palm Beach County Votomatics were simultaneously misaligned in the same way, so a decently designed ballot would not have lost thousands of Gore votes (or thousands of Dole votes the election before). As the design critic on says, it seems the problem was that the designers were thinking about one thing -- how to fit 10 different party lines onto one page -- and voters necessarily are thinking about something else: how to find the one line corresponding to the candidate they intend to vote for. Good designers would be putting themselves in the shoes of the end users. But as any of us knows who has browsed various web sites for even 10 minutes, it's kind of shocking how many of them fail to do that.

  15. Wow, another one just disappeared! And it included my theory of what's causing this, which has thus been dispreven. :-S

  16. Forgive the perseveration - but in 1996, the third-most populous county in Florida (Palm Beach) utilized a ballot that cost one major-party candidate 10% of his votes cast there, and 0.6% of his overall votes in the state, which had the 4th most EVs total. Though the county is quite liberal, with the Democrat slated for that line in 2000, this problem did not raise alarm among the Government Administrators responsible for running that county's election.

    Jeff's linked writer put forth a few reasons why those election officials didn't change, mainly having to do with cost. Those may be valid proximate reasons, but the ultimate reason is that they simply didn't care enough, or think this was important enough, to do anything about it.

    And while the razor-thin margin in 2000 might not have been expected over 1,000 elections, given the fact that Palm Beach cost a candidate 0.6% of his state total in 1996, those election officials must have realized that over, say, 20 elections, there was a pretty good chance their bad ballot would on one occasion influence at least state-level results (and given Florida's size, possibly national-level results), but that was clearly not motive enough to fix their voting system.

    Its funny, but this makes me think of the Tea Party, and how the media focuses on the sizable contingent of them that are racist or xenophobic or really dumb, which is good for copy I suppose. If there is wheat among that chaff, its the Tea Partiers who are frustrated that giant government brings out the worst in people, and assuming that government entities will do the right thing, left to their own devices, is often a terrible mistake.

    If tempted to throw the sincere Tea Partiers in with their crazy, more media-friendly racist brethren, spare a thought for those Palm Beach election officials, and the disturbing results in 1996 (and how many elections past?) that didn't prompt those officials to do the right thing, thus specifically changing the face of the Presidency in 2000, and, generally, who knows how much further beyond.

  17. I'm not eager to defend Palm Beach on this, CSH, but:

    > I've dealt with county and city clerks like Theresa LaPore. A lot. Their problem is not likely to be that they were too lazy or uninterested, or too given to bureaucratic inertia, to do the right thing. Indeed, the butterfly ballot design was itself apparently the perverse result of an attempt to do the (or a) right thing, i.e. treat all candidates equally by putting them all on the same page in the Votomatic.

    > County clerks were probably required to treat candidates in this fair / equal way, or thought they were required to, by state law. The Votomatic provides limited options for fulfilling such a requirement when there are 10 party lines, each including the party and two candidate names.

    > The Votomatics were probably old and needed replacing, but the clerk's office didn't have the budget for this.

    > County clerks' offices are also usually overwhelmed, understaffed, and unable to hire top-quality document designers. This is because:

    > Public services like these, the kind that don't involve cool, high-tech ways of slaughtering people, are seldom funded at the levels they need to be. To these problems, the Tea Party says,

    > "Let's make it worse." ;-)

  18. Jeff, I take all your points, and in spite of my rant above, I don't expect a government employee to allow the perfect to be the enemy of the good. Private sector employees don't do that, so we shouldn't expect government employees to do so.

    What's fascinating about the butterfly ballot is that, between double-counts and missed Buchanan votes, it looks like something like 8% of Gore's Palm Beach votes were spoiled. 8%, which is roughly in the ballpark of what Dole lost in '96.

    Its also fair to say that fixing that problem, post-'96, would have seemed like going 'above and beyond' to LaPore and the rest of her stakeholders. It would have required some heavy lifting beyond their normal job descriptions.

    But that's the rub, Jeff. From where I sit, regardless of 2000 craziness, a system that spoils ~10% of a candidate's county votes (and ~1% of the statewide votes) such as the butterfly ballot visited upon Dole in '96, is a massive problem, one urgently meriting something more than business as usual.

    I fully understand why LaPore and her peers saw it differently than I do. I just strongly disagree with them.

  19. CSH, I'd go further -- not only should the local authorities (not just LaPore, but the County Board or whatever) in Palm Beach County have been all over this after '96, but there should be some kind of standard nationwide audit, say by the GAO, after every presidential election aimed at finding glaring problems, like the disappearing Dole votes, and directing resources toward fixing them before the next one.

    I'm also puzzled as to why Republican state officials or local members of Congress didn't squawk about the '96 case. Perhaps our system relies too much not only on local officials, but on the premise that party competition and self-interest will lead to problems being called out and corrected.


Note: Only a member of this blog may post a comment.