Back to previous post: Open thread 24

Go to Making Light's front page.

Forward to next post: A Houseful of Lords, pt. 2

Subscribe (via RSS) to this post's comment thread. (What does this mean? Here's a quick introduction.)

June 14, 2004

A few more questions
Posted by Teresa at 02:10 PM *

4. If you put a dollar in a change machine and the machine refuses to take it, do you throw the dollar away?

(If your answer is “no”, then regardless of your other political affiliations and opinions, you know that a hand count is more accurate and reliable than a machine count.)

5. When James Baker stood up in front of the mikes at a press conference during the fuss over the Florida miscount, and asserted that machine counts are more accurate than hand counts, how come no one asked him what he does when change machines reject his bills?

Failing asking the obvious question, how come no one asked him how he thought he could get away with that statement, given that it’s a long-established fact that hand counts are more accurate?

(He might as well have announced, “We’re lying, and as long as the U.S. press keeps mum, we don’t care who else in the world knows we’re doing it.”)

6. Why should we believe that these people think it possible that they’ll be out of power following the next election?

Comments on A few more questions:
#1 ::: Ken Houghton ::: (view all by) ::: June 14, 2004, 02:30 PM:

Follow-up, just to throw a spanner into the works:

7. If you have a jar of change, do you count it carefully and put it into rolls, or do you take it to one of those supermarket places that charges you 7.527% and gives you a printout at the end that tells you how much money you gave it?

The answer to "6" is that we shouldn't, and they don't. Sacrifice Ashcroft (who is going away anyway) in late July (taking the thunder from the DNC in Boston), deliver Osama in October, and "announce" the open-secret departures of Rice, Powell, and/or Rumsfeld as needed in the runup.

Otherwise, keep pounding on the "sacraments" issue, and if that stops working, point to the study that shows Kerry to be the BEST Congressman in voting "Catholic beliefs."

If all else fails, you can depend on Diebold. But recommending staying the course and showing some activity (while complaining that your opponents "just won't be satisfied") should make that unnecessary.

#2 ::: fidelio ::: (view all by) ::: June 14, 2004, 02:35 PM:

Was that Howard Baker, or James Baker, perennial water-carrier for the Bush family? I remember JB running loose in Florida at that time, but I don't recall HB being in it up to his neck.

#3 ::: Ken Houghton ::: (view all by) ::: June 14, 2004, 03:11 PM:

Fidelio is correct; it was James. Howard Baker (R-TN) is a relic from the days when My Ancestral Party believed in integrity, small government, and balancing the budget.

#4 ::: fidelio ::: (view all by) ::: June 14, 2004, 03:37 PM:

I kept thinking it had to be; for one thing, Howard Baker is quite old; for another ["What did the President know and when did he know it?"] he's too much of an independent thinker--like Fred Thompson, whom I did not vote for but do miss greatly--to do la famille Bush's bidding.

"the days when My Ancestral Party believed in integrity, small government, and balancing the budget."
Ah, you mean the days when the GOP had gentlemen in it, as opposed to self-identified aristocrats?

#5 ::: Tom Whitmore ::: (view all by) ::: June 14, 2004, 03:58 PM:

And why does the porridge-bird lay her eggs in the air?

These are very good questions, T. They might actually matter to a great many people who are never going to hear them. Getting them out into a Major News Service would be a mitzvah, but I know how much spare time you have for actually writing articles that would hit Salon, the New Yorker, or other places that aren't preaching to the choir. And, let's face it, over 90% of the people who read here regularly are the choir. We agree with you, many of us love you (for our individual values of love!) and we're all somewhat screwed by the current way information spreads.

I push this blog/website continually in person and in small online fora. I think the folks here actually come up with good ideas and talk well about them. But without getting some sort of mainstream attention, will we make more than a minuscule difference? Now, a minuscule difference might have tipped the last Presidential election, and is clearly worth making in those contexts! We've got people here who can write well enough to hit Major Media, people with the savvy to know how to pitch to those folks and how to disguise the message so it might get across -- and we've got your subtle amazing mind to point the way to messages that might actually carry.

Not in a general sense, but in an individual sense: what should I do next? Even knowing it's likely that I'll personally fail, what should I do?

Those of you who don't know T or me might think this is sarcastic. It isn't. I believe T is doing a lot by raising these questions, even while I think that raising them in other places may have better effect. And I'm not promising to take her advice: I'm just asking for it. And hoping that it may point a direction for a few folks here who are seeking an approach.

#6 ::: James D. Macdonald ::: (view all by) ::: June 14, 2004, 04:36 PM:

The answer to what we can do right now is write. For paper-and-ink publications. Letters to the editor. Op-ed pieces. Write to and for anything that has a post office address and a news stand presence.

And write to our representatives, federal, state, and local. Lots of letters. Paper-and-ink letters. Calm, reasoned letters filled with facts.

Flood them with letters.

All of the internet isn't worth the electrons it's painted with. We must go to paper and ink.

#7 ::: mythago ::: (view all by) ::: June 14, 2004, 05:03 PM:

Hear, hear, James. One well-written, physically mailed letter is worth its weight in Internet petitions. Congressthings take such letters quite seriously.

re #6, y'all remember the great wailing and gnashing of teeth when Clinton was elected--not merely because the lib'ruls got elected, but because those people had been in government for twelve years under Reagan and then Bush Sr. They'd persuaded themselves that they had real, solid jobs, not merely positions served at the whim of the party in power, and they felt very entitled to them.

#8 ::: Linkmeister ::: (view all by) ::: June 14, 2004, 05:16 PM:

If Halberstam is to be believed ("War in a Time of Peace"), it was in fact James Baker, who moved back from SecState to run Bush I's campaign in 1992 and got drafted to run the Florida recount "situation."

#9 ::: Kevin Marks ::: (view all by) ::: June 14, 2004, 05:26 PM:

This is a nit-pick, but it is a germane one. The point of Gresham's law is that people don't throw out counterfeit currency, they try to pass it on quickly so they aren't the ones who lose from it - the subtlety is that the velocity of circulation of counterfeit money is higher than that of real money, which is how it accelerates inflation.

So the analogy doesn't quite apply.

On the basic point, I do agree - having seen a UK election count, the 'mark the X, bundle the votes, and line up on a table to give a live histogram' model works remarkably well for a first-past-the-post election. For US stye ballots where you are voting on 50 offices and 20 propositions, it is harder to do this.

#10 ::: Marilee ::: (view all by) ::: June 14, 2004, 06:07 PM:

Ken: "7. If you have a jar of change, do you count it carefully and put it into rolls, or do you take it to one of those supermarket places that charges you 7.527% and gives you a printout at the end that tells you how much money you gave it?"

I take it to my credit union Monday-Thursday, 8am-4pm and hand it to the teller who will put it through their machine and deposit it to my account, free.

Kevin, I don't think TNH is talking about counterfeit bills. Regular bills that are torn, wrinkled, faded, etc., are frequently rejected by change machines.

#11 ::: plover ::: (view all by) ::: June 14, 2004, 06:11 PM:

At the risk of being accused of picking nits from nits...

The point of Gresham's law is that people don't throw out counterfeit currency, they try to pass it on quickly so they aren't the ones who lose from it

Are you sure about this? According to this definition of Gresham's law, it doesn't sound like the principle applies to counterfeit currency (as it is - by definition - not "legal tender").

So whether or not the phenomenon is real, this doesn't seem to be the right characterization of it.

the velocity of circulation of ... radioactive ... money is higher than that of real money
Just 'A Modest Proposal' from Larry Niven...

(Readers are enjoined to imagine the word 'counterfeit' with a line through it just before the word 'radioactive'. Sadly, the blog engine seems to be preventing me from providing them with the physical experience.)

#12 ::: Steve ::: (view all by) ::: June 14, 2004, 06:19 PM:

Gresham's law doesn't refer to counterfeit money as such; it refers to specie minted using less precious metal, either by mixing it with base metal or by shaving the weight -- I believe the formulation refers to "light coin". But the idea that the bad coin circulates while the good coin is hoarded is right; if you can get $1 worth of goods for $.95 worth of currency, that's the way to go.

#13 ::: plover ::: (view all by) ::: June 14, 2004, 06:21 PM:

So whether or not the phenomenon is real, this doesn't seem to be the right characterization of it.

Bleah... too few referents.

The 'phenomenon' in question is the rapid velocity of counterfeit money.

The 'this' that is being labelled a mis-characterization' is the description of the velocity of counterfeit money as an instance of Gresham's law.

Humble apologies for any confusion.

#14 ::: Columbine ::: (view all by) ::: June 14, 2004, 08:49 PM:

Do the people who are not in the choir ever read The New York Times? (No, it's a sincere question. I have no idea who the NYT audience is.)

Because they pointed out on Sunday that voting machines undergo nowhere near the scrutiny and testing that Vegas slot machines do - and Vegas has a lot less at stake, just their reputation and a whole boatload of money.

(Having recently come back from there, I'm with the Times; I think we should have the Nevada gaming board oversee the election.)

Point being that at least one national newspaper is crying that the emperor has no clothes. The story was also picked up by Truthout, but I will readily concede that Truthout is preaching only to the choir.

#15 ::: Randall P. ::: (view all by) ::: June 14, 2004, 09:11 PM:

Just to play the devil's advocate here, but are we just fooling ourselves to think that the American public actually cares about any of this? (And I'll say the Canadian public as well, considering we're having an election and the platform-less Tories are winning the hearts and minds of a populace I thought was more intelligent than it actually is).

To be quite honest, I find that the populace is turning out to actually be quite stupid. They go home to their diets of Fox news and CNN and watch Survivor and go to the movies to see crap like Shrek, and above all, they do their best to keep from thinking. Let me emphasize that again. They do their best to keep from thinking.

And maybe that's how they want it. Maybe the majority of people prefer to be spoonfed and as long as they don't have to use their brains or come up with an independent thought, they're just peachy, thank you very much.

Perhaps I'm just being cynical, but I'm finding my beliefs to place me in a solid minority. We can rail all we want about injustice and the pathetic nature of our so-called leaders, but what difference will it actually make?

There was an excellent article in Harpers (I looked for it but couldn't find it) a few months back about how disappointed the author felt that he'd never put himself on the line for a cause. So I guess that's the question, isn't it? How many Americans would be willing to risk everything they have for their cause? How many Americans would be willing to die for a cause? How much do they have to lose before they do this? Because I think that is what it is going to come down to in the future. Not now, but soon.

And so I think that we can bitch and moan all we want, but it will make little difference. We are the educated minority and until the people are willing to make a change and upset the balance of their lives and educate themselves, all the talking we could ever do would be for naught. I mean seriously, we can write letters, we can call our congressional representatives, we can join the protests, but if the majority of the population wants it the way it is, then it shall be that way. guess I should go back to my Entertainment Weekly with my favorite actor, Tom Hanks, on the cover. Ciao.

#16 ::: Randall P. ::: (view all by) ::: June 14, 2004, 09:23 PM:

Okay, so after I wrote that, after I've blown my load by completely giving up hope on people because of their sheer stupidity, I walk into the room to see my beautiful baby daughter sitting in the hallway singing to herself. It was, quite literally, one of the most beautiful things I've ever seen...

...and then I realize that THAT is the reason why we keep railing against these fuckers. My ass if I'm going to leave the world in the sorry state that it's in without trying, especially since I brought that little girl into the world.

And in that sense, there is the cause that I would die for. My daughter. I've answered my own question. I've had an arguement with myself. You all had to listen to it. guess I should go back to my Entertainment Weekly with my favorite actor, Tom Hanks, on the cover. Ciao.

#17 ::: Scorpio ::: (view all by) ::: June 14, 2004, 09:33 PM:

The quickest way for us to go down is to trust them.

CA did the best thing and tossed the Diebolt machines out until they prove they give a good count (or possibly until they have paper verifications).

Places with unverifiable systems need to get on the stick locally. Wonder if the Dean Machine will gear up for this. It may.


#18 ::: Bob O ::: (view all by) ::: June 14, 2004, 10:01 PM:

Up near the start of this thread, Tom Whitmore was asking:
"Not in a general sense, but in an individual sense: what should I do next? Even knowing it's likely that I'll personally fail, what should I do?"

As last week's Reagan-fest was winding down, I found myself thinking along the same lines.

And I was wondering just what I was going to tell my children -- or the Truth and Reconciliation Committee --
when they ask me: "And just what did YOU do to stop them?"

And so I went over to the computer and I donated a day's pay to the Kerry campaign.

I've been involved in politics -- mostly locally -- on-and-off since the McGovern campaign, and this is the very first time that I've been motivated to give money (instead of volunteering my time) in a Presidential race.

It's not much, but at least I'll have something to say when this is all over and I'm asked what I did to help stop them.

If you're looking for some way to help, give it some thought. It's a start.

#19 ::: Teresa Nielsen Hayden ::: (view all by) ::: June 14, 2004, 10:12 PM:

It was of course James Baker. I'll go fix that.

Just got home. Patrick and I have been over to Claire Eddy's, getting her set up with broadband.

#20 ::: Dave Kuzminski ::: (view all by) ::: June 14, 2004, 10:22 PM:

It might be better to make the point at a slightly lower level than a change machine. Instead, I recommend that we remind folks that:

Vending machines have been around for 2,000 years and we still can't trust those to always deliver. Instead, they frequently steal our money and give us nothing. So, as the forerunners of voting machines, what does that tell us to expect now?

You're welcome to change this to make it trip off the tongue easier should anyone find it useful.

#21 ::: Calimac ::: (view all by) ::: June 14, 2004, 11:24 PM:

4. As suggested above, if the machine rejects your bill, that raises a small possibility that it's counterfeit. Not a large possibility, though, especially if it's a worn $1.

5. Is a machine count more reliable than a hand count to determine the quantity of a large number of bills? I use a hand-counter device to count large numbers of anything, and that's a machine.

Gresham's Law in action was graphically demonstrated in Seattle (and probably elsewhere) in the early 80s, as TNH may remember. Up to that time, the US and Canadian dollars were roughly at par, and Canadian change circulated so freely in border areas, at least as far south as Seattle: so much so that vending machines were tweaked to accept Canadian coins, sometimes so much so that they wouldn't accept US coins. This free circulation was actually OK with the US banking system.

Then about 1982 the Canadian dollar began to plummet. For a brief period it was cheaper to go to Canada to buy US paperback books, then publishers began putting separate Canadian prices on the covers. But the really striking effect was the virtual disappearance of US coins from circulation in the Seattle area, as people hoarded them for their greater value. For a while I had to empty my pockets of change every time I visited California, where Canadian change had never been accepted.

After a couple years the Federal Reserve blew the whistle and declared Canadian change no longer acceptable, and Seattle pocket change returned to normality. I just got a Canadian nickel in Juneau, Alaska, though. Rather surprised me. I got rid of it as quickly as possible, proving that Gresham still lives.

#22 ::: James D. Macdonald ::: (view all by) ::: June 14, 2004, 11:32 PM:

A better example of Grisham's Law might be when the US introduced the "sandwich" coins in place of silver coins. Instantly all the silver coins went out of circulation and into people's desk drawers.

#23 ::: Larry Brennan ::: (view all by) ::: June 15, 2004, 12:31 AM:

Calimac - the same dynamic happened with Candian change in Upstate NY. Now, cashiers are so conditioned to look for and reject Canadian change that it's hard to get rid of. I think most of it winds up in the tip cups at Starbucks's. (That's what I did with my Canadian change.)

Amusingly, when the state quarters came out, they were hard to spend in Rochester - they were too shiny and looked, at first glance, Canadian.

#24 ::: Marilee ::: (view all by) ::: June 15, 2004, 12:43 AM:

Randall P., I liked Shrek. I have Shrek 2 on my Netflix list. That doesn't make me any less educated or liberal.

#25 ::: mythago ::: (view all by) ::: June 15, 2004, 12:44 AM:

Interestingly, when I lived in southeastern Michigan, Canadian coins were used interchangeably with American coins, largely because you knew other people would use them. In other words, the system worked only because there was a tacit agreement that you would take other people's Canadian coins and so they would accept yours; nobody would get stuck with them. (People would also swap coins for vending machines on the same principle--"Anybody got an American quarter for this one so I can use the Coke machine?") Of course, we were also right next to Ontario, so there was a possibility of using said coins in their native land.

This did not extend to Canadian paper money.

#26 ::: Bill Simmon ::: (view all by) ::: June 15, 2004, 02:07 AM:

It's fixed at candleblog too. Thanks for stopping by.

#27 ::: Lois Fundis ::: (view all by) ::: June 15, 2004, 02:13 AM:

4. When a machine rejects my bill, I look to see if the corners are bent or if the bill is wrinkled. After straightening things out, literally, I try again. Sometimes three or four times.

5. No one had thought of this analogy. Heck, I hadn't. And how many of these people use change machines, anyway? Do they go to the laundromat or the self-serve checkout line in the grocery store? They have secretaries and such to do such mundane things for them!

6. In God and Diebold they trust. At least one of those trusts might be misplaced.

Calimac: here in West Virginia, I sometimes get smaller Canadian coins (pennies, nickels, occasionally a dime or quarter) in change at stores. Vending machines don't accept them, of course. So far no Loonies have shown up.

A few weeks ago I *did* find a Susan B. Anthony dollar in my change purse. I think someone was out 75 cents, having given it to me in change thinking it was a quarter. Ironically, it did me no good at that particular time because I was looking for quarters for a vending machine.

#28 ::: bad Jim ::: (view all by) ::: June 15, 2004, 03:19 AM:

After the 2000 election, my brother-in-law, an engineer like me, asserted that it had been shown that machines were more accurate at counting ballots than humans. I really wondered how any researcher could arrive at such a conclusion, but I didn't challenge him, since our relationship has been under repair for a decade or so.

(The answer is that human scrutiny remains the gold standard. He has since been disgusted by the war, and though he may vote Republican again someday, it's beyond unlikely that he'd vote for Bush. As goes Marin, so goes ... Contra Costa?)

A key point about electronic voting is seldom raised: what is the error rate? So long as the results can't be audited, the error rate is unknown. The assurances offered by the vendors of the machine are no more than hand-waving.

Quality gurus should be ringing the alarms. (Aren't we all striving towards ISO-9000 certification, six sigmas and all that?) You don't know what you can't measure.

Election officials don't seem to be bothered that much, perhaps because current practice isn't all that reliable, either, and electronic voting promises to alleviate a multitude of difficulties, the variety of languages required and the special needs of the vision-impaired. Fine.

We still need routine audits of the results to find out what mistakes we're making. To do anything else requires more faith than I've had since I lost belief in the tooth fairy.

#29 ::: Tom Whitmore ::: (view all by) ::: June 15, 2004, 08:22 AM:

False positives, false negatives. Machines reject worn bills and some very crisp forgeries. If a bill changer rejected a crisp, unfolded bill that I'd have accepted -- I'd take it down to my local bank and ask what's going on with this particular bill.

Machines may accept some forgeries that I'd just look at and say -- what's Richard Nixon doing on the $1 bill?

Having been involved in a hand-count of votes (of a relatively small number of people) where both I and someone I know is very reliable put votes in the wrong pile, without malice and without any intent other than trying to count as well as possible: I'm personally not convinced that machines are _more_ reliable than humans, but I believe that honest machines are not _less_ reliable.

And I don't believe that I can count on Diebold machines being honest.

#30 ::: Skwid ::: (view all by) ::: June 15, 2004, 09:41 AM:

Amen, Marilee.

Shrek is art. Silly art, undeniably, but so is "How Much for Just the Planet," and I think we'd have little argument on that point, here...

#31 ::: mythago ::: (view all by) ::: June 15, 2004, 10:09 AM:

If a bill changer rejected a crisp, unfolded bill that I'd have accepted -- I'd take it down to my local bank and ask what's going on with this particular bill.

I'd be more likely to tell the owner of the machine "I think this thing is busted." (For one thing, $1 bills are rarely forged, and when they are, it's by amateurs who do a very obvious job.)

big Jim, one of the selling points for Diebold is that there are far fewer "errors" in the sense of spoiled ballots with touch-screen; nobody colors outside the ovals or gets hanging chads with a touch screen, you see. That says nothing about the error rate of the machine itself.

#32 ::: Jules ::: (view all by) ::: June 15, 2004, 10:18 AM:

Electronic voting is a topic that has interested me for a while, and years ago, before I'd ever heard of Diebold, I came to this conclusion:

There must be a paper trail that allows voters to verify that their own votes have been registered correctly, and allows an independent auditor to verify that the machines are counting them correctly.

The way to do this is fairly simple: the machine prints out a slip with the details of the voter's vote, and allows the voter to view this. When the voter is happy, he presses a button and the slip is deposited into a sealed box.

The voting machine counts what votes are supposed to be in the box. A random selection of machines are audited following the close of the election, the contents of their boxes counted manually. If the discrepancies are consistently too high (to be determined based on statistics of how many mistakes human counters regularly make), then all of the votes must be counted by hand, as the voting machines are obviously untrustworthy.

Voting machines _should_ be more accurate than human counting. In fact, they should be close enought to 100% that the only way the wrong vote can be registered is by operator error (e.g. not reading the text on the screen, reminiscent of when in the UK a few years back a number of people voted for a candidate that was described as a "literal democrat" rather than the "liberal democrat" they wanted to vote for).

Anything less than this is bad design, of a nature so bad it becomes hard to ascribe to incompetence.

#33 ::: Richard Brandt ::: (view all by) ::: June 15, 2004, 11:20 AM:

Ah, for the days when Mexico introduced a one-peso coin, which was mechanically indistinguishable from the USA quarter. (And this before the peso plummeted to an exchange rate of 4000 to the dollar.)

#34 ::: Randall P. ::: (view all by) ::: June 15, 2004, 03:22 PM:

Thank you Skwid and Marilee for both missing and illustrating my point in the same post. (Sigh)

#35 ::: Skwid ::: (view all by) ::: June 15, 2004, 03:58 PM:

Randall, if your point was anything other than equating an excellent genre film to sensationalist news channels and insipid reality shows, it was certainly not evident from context.

I'm afraid no amount of breathy exhalations is going to make your point clearer.

#36 ::: tavella ::: (view all by) ::: June 15, 2004, 04:29 PM:

Actually, my favorite version is that the electronic machine prints out a paper ballot which the voter then feeds into the machine that counts it. It combines the good qualities of an electronic ballot -- self checking, prevents double votes, insufficiently filled out circles, et al -- yet keeps unity between the receipt and the ballot. Separating the receipt and the ballot is going to lead to a lot of court challenges; the question of whether a recount of the receipts is valid, to start with. They aren't the ballots, can you really use them to recount?

#37 ::: Calimac ::: (view all by) ::: June 15, 2004, 05:41 PM:

Yog, what is Grisham's Law? "Bad fiction drives out good"?

You cite, however, an excellent example of Gresham's Law.

#38 ::: Dave Kuzminski ::: (view all by) ::: June 15, 2004, 07:50 PM:

Okay, I couldn't resist assisting in the effort by making something available at CafePress even though it meant opening a store. If the link doesn't show up with my name above, then cut and paste this:

I went ahead and made a few other things available so it wouldn't be so bare. If anyone has any suggestions, please be sure to let me know.

#39 ::: Jon H ::: (view all by) ::: June 15, 2004, 09:18 PM:

"If you put a dollar in a change machine and the machine refuses to take it, do you throw the dollar away?"

(practical answer)

I'd look for a vending machine. Some are set up so that, if you put in a dollar, and hit the button to get it back, it'll give you change rather than giving you the bill back.

Another vending machine trick is that, sometimes when your bill isn't accepted, it's because the machine wouldn't be able to make change for it. So if you put in a dime and/or nickel, it'll often take the bill. Then you can hit the button and get a dollar in change, plus the extra bit you put in. Or you can buy something.

But, no, I don't suppose that's what you were getting at.

BTW, and OT: Neko Case is playing at the Southpaw in Park Slope tomorrow night. Dunno if tickets are still available. Excellent music. I'd come down from Connecticut if I didn't have a Calc exam Thursday morning.

#40 ::: CD318 ::: (view all by) ::: June 16, 2004, 09:56 AM:

Randall P. -- the People do not have a diet of CNN and Fox.
Fox has a daily viewership of

#41 ::: CD318 ::: (view all by) ::: June 16, 2004, 10:00 AM:

Urgh. The system snipped nearly nearly my whole post.

CNN, Fox, MSNBC each have

#42 ::: Trish Wilson ::: (view all by) ::: June 17, 2004, 07:55 AM:

"4. If you put a dollar in a change machine and the machine refuses to take it, do you throw the dollar away?"

I was going to suggest you replace the change machine, and then I saw what you were getting at. I realized in the end that I was right after all.

#43 ::: David Dyer-Bennet ::: (view all by) ::: June 17, 2004, 10:40 PM:

"4. If you put a dollar in a change machine and the machine refuses to take it, do you throw the dollar away?"

No, of course not. And this question actually has nothing to do with the relative accuracy of hand vs. machine counts. The change machine is recognizing, not counting, the dollar, and its main risk is letting counterfeits slip through. That's a completely different issue from counting thousands of ballots, so the machine is built with different biases.

Hand and machine counts have different failure modes. Some are much worse on one side, some are much worse on the other side.

The big problem with machine counts is what Bruce Schneier calss the "class break" -- if you figure out a way to reprogam the machines to give the answer you want, you've hit all the precincts at once. It used to be that vote fraud was mostly retail; that trick elevates it to wholesale.

This has nothing to do with *inherent* accuracy. (It may be an entirely convincing reason to prefer hand counts to machine counts, though).

The machines will actually have a much lower error rate than the humans, in general. If they're working correctly and honestly. And if you have decent workflow, including a provision for humans to analyze the rejected ballots.

I've *counted* large piles of things. Many times. Often the same pile many times, backed up by other people. Usually coming to some kind of consensus on how many things are in it. It's really quite hard to count even a few thousand things by hand with high accuracy. Machines are much better at that sort of thing.

#44 ::: Paula Helm Murray ::: (view all by) ::: June 18, 2004, 11:04 PM:

I'm in a funk. Apparently John McCain is supporting BUSH. I sent him about as scathing an email as it's possble for me to send (ms keep-the-peace here). I give up on his IQ. period. I thought I respected him, but if this is the truth of the matter, no. Because Mr. Bush does not deserve my respect either.

#45 ::: eric ::: (view all by) ::: June 25, 2004, 09:52 PM:

I think Al Gore nearly invoked Godwins law on the neocons, while accusing GW of proof by vigorous assertion.

The Administration works closely with a network of “rapid response” digital Brown Shirts who work to pressure reporters and their editors for “undermining support for our troops.”

from here

#46 ::: ET ::: (view all by) ::: July 02, 2004, 03:45 AM:

I may be misremembering things (especially since I'm not a US resident), but wasn't the main point of contention the confusing forms, and not the validity of their counting? Or am I confusing this with some other case in your voting system?

#47 ::: David Goldfarb ::: (view all by) ::: July 02, 2004, 04:56 AM:

The "butterfly ballot" was a point of contention, but not the main one.

#48 ::: James D. Macdonald ::: (view all by) ::: July 02, 2004, 12:24 PM:

The main points of contention were uncounted "spoiled" ballots, ballots where the mechanism had not cleanly punched out the perforations which allowed machine counting, and improperly postmarked mail ballots. This is without mentioning the voters mistakenly removed from the roles for having the same names as various felons.

#49 ::: Xopher ::: (view all by) ::: July 02, 2004, 01:07 PM:

James, by using the word 'mistakenly' in that sentence you again prove yourself more charitable than I.

#50 ::: ET ::: (view all by) ::: July 02, 2004, 05:49 PM:

Thanks, James. Sounds like a lovely and lively election.

#51 ::: Alex Cohen ::: (view all by) ::: July 02, 2004, 06:24 PM:

For your 2000 election memories: the t-shirt.

#52 ::: Jonathan Vos Post ::: (view all by) ::: July 02, 2004, 09:48 PM:

"mistakenly removed from the roles" should be
"mistakenly removed from the rolls"?

And what ensures that this will not happen again?

Was the old contractor replaced, the rolls recomputed, the felons dismbiguated from others with similar names? Who audited that process?

I've enjoyed visiting Florida since the 1950s, and enjoyed watching CSI: Miami, but I'm glad that I don't live there.

#53 ::: Jonathan Vos Post ::: (view all by) ::: July 02, 2004, 10:42 PM:

And now I see some of my questions answered, and others raised, thy this story:

Eligible Voters Said on Fla. Felons List

And I just received a friendly rejection letter on "War Between the Numbers" from F&SF. Maybe I'll send it to Analog now.

#54 ::: Meg Thornton ::: (view all by) ::: July 05, 2004, 01:47 AM:

Y'know, reading all this stuff about the US electoral system that I've picked up over the years from Usenet, and now from the 'blogging community, I find I'm very glad I live in Australia. Not because of the compulsory turnout, or even because of the Australian Electoral Commission, but because of our lovely system of preferential voting.

Quick explanation: On an Australian federal electoral ballot (and indeed on most state ones), you're given a list of the candidates. You're asked to number them in sequential order, using the whole numbers from 1 - n (where n=number of candidates). Then when ballots are counted (and it has to be by a human count, simply because of this) all the ones with candidate X given first preference are in pile X, all the ones with candidate Y are in another and soforth. Once all the votes have been counted, if no candidate has 50% of the vote +1 extra, the smallest pile goes on to second preferences.

For example:
After the first round of counting,
Candidate A has 30% of the vote
Candidate B has 45% of the vote
Candidate C has 10% of the vote
Candidate D has 15% of the vote

Candidate C, having the smallest number of votes, is ruled out. However, as none of the other candidates has 50% + 1 vote, his ballots aren't thrown out. Instead, they're distributed as to who the voters for Candidate C have as a second preference.

So, after the second round of counting:

Candidate A has 36% of the vote
Candidate B has 47% of the vote
Candidate D has 17% of the vote.

Candidate D is ruled out, and his votes are divided according to the next preference available for sorting - so the 2% he got from Candidate C are sorted by third preference, and the other 15% are sorted by second preference.

This is called "going to preferences". It appears to work here. Certainly despite compulsory turnout, and the necessity to do more than just punch a hole or tick a box, the level of invalid ballots tends to be rather low - down below 10% for the majority of the country. (The Australian Electoral Commission has figures to prove it)

Maybe it might be a notion for electoral reformers in the US to consider?

#55 ::: Jeremy Leader ::: (view all by) ::: July 05, 2004, 02:20 AM:

Meg, is turnout (showing up and handing in your card) compulsory, or is actual voting compulsory, that is, are you actually required to rank all n candidates to fulfill your obligation?

Also, "it has to be by human count" is inaccurate. While I'm adamantly opposed to the deployment of electronic voting systems without real audit trails, it's certainly possible to program a computer to carry out the algorithm you describe. In fact, less than a computer, I suspect that old-fashioned Hollerith punched card sorters (such as were used for the 1890 US census) could carry out exactly the procedure you describe, with a little cleverness in the card encodings (at least for n less than 10).

But yes, some form of preferential voting would allow minority candidates to make more positive contributions, instead of being mere spoilers.

#56 ::: David Goldfarb ::: (view all by) ::: July 05, 2004, 03:00 AM:

And indeed we find that the Hugo awards use just such a preferential ballot, and they are counted by computer. In the admittedly unlikely event that the US adopts it, perhaps we could adapt the Hugo-voting software.

#57 ::: Epacris ::: (view all by) ::: July 05, 2004, 07:42 AM:

See earlier comment on this, referring you to an official explanation of the system

In Australia it's compulsory to attend and to put a ballot paper in the box.
Because no-one can see what you've put on it at the time, some people who don't like any candidate in their area just put in a blank one (or write "None of the Above", or more forceful expressions of their opinion).

And ninety-odd years ago, the first Totali[s/z]ator was invented by George Julius, originally as a mechanical vote-counter which could deal with preferential ballots. ( This site - - has the later development of the system when it was swapped over to betting.) From the descriptions & photos, the older models sound like a development of the Babbage style of machine.
Now the betting is computerized - an obvious reference point for security and other systems for computer-counted voting.

In this Year of Elections, you may have heard that Indonesia (world's largest muslim nation) is currently having their first popular election of a President. Like so many other places, they have the system of electoral rounds, where the top couple of people from the first lot of voting go on to another one, if no-one wins a majority of the votes, another thing that the preferential system avoids. Remember that moment of horror in the French elections a year or so back, where the despairing left were reduced to a choice between an extreme right & a fairly right candidate?

#58 ::: Meg Thornton ::: (view all by) ::: July 05, 2004, 11:17 PM:


As per Epacris - it is compulsory to show up at a polling place on election day (all elections in .au tend to happen on Saturdays - I'm unaware whether this is legally mandated), to get your name marked off the electoral roll, and to receive the ballot paper. You are then pretty much entitled to do whatever you fancy to it in the polling booth (within reason... the booths are cardboard, and there's no curtains etc - I suppose this does stifle some more extreme reactions to the candidates) and then you're expected to put it into the ballot box. Ballot papers aren't supposed to leave the polling place, and they tend to be numbered (probably to catch electoral fraud) but there's no real way of linking a particular ballot paper to a particular person.

Oh, and it's good to hear that the whole process can be pretty much automated. Not sure when I got the notion of the necessity of the hand count - probably getting muddled with something else.

Generally, the process here tends to be one where electoral fraud has to be performed at the electoral boundary level - Queensland was notorious for a number of years for a large number of gerrymandered electoral boundaries to keep the ruling party in power.

#59 ::: Epacris ::: (view all by) ::: July 06, 2004, 07:31 AM:

Our Australian system does depend quite a bit on a general assumption of honesty. To save on petrol & storage space some years back our ballot boxes (and cubicles) were changed to cardboard.

I've seen boxes in more suspicious places which are not only very lockable & metal, but are built in a framework so that before voting starts, they are fully opened & held up so that people can see they haven't been pre-stuffed.

There also isn't a necessity to prove who you are, you just rock on up, tell the person with the list your name & some other detail, and get crossed off as they give you your ballot papers. No finger-inking. Since you can vote thusly at any booth within your electorate, one can theoretically go vote several times and they won't find out 'til the end of the day when they reconcile the lists. Of course, they don't know which are those votes, so can't cut them out. I think it would only make a big difference if the voting was close and the loser could challenge the result.

You can also vote 'absentee', i.e. out of your electorate or pre-poll (before the actual date). The time I did this, you put your ballot(s) [Every ballot paper has to be initialled by the local electoral officer] into a blank (initialled by officer) envelope, then into another envelope you filled out your details on, which was sent to your electorate. Now it's one envelope with a detachable flap you fill in. Postal voting is similar, I believe. Means they can check the details on the outer envelope/flap, then remove the outer envelope or flap, pile up the blanked ones and check them anonymously. Again, this is based on trusting the electoral officers.

There are independent State and Federal electoral authorities who organise the basic 'structure' of elections. They are supposed to be quite non-partisan. Every larger party will have scrutineers & other volunteers at counting rooms to watch the counting & argue over doubtful votes, so the idea is that each will keep the others honest.

#60 ::: Epacris ::: (view all by) ::: July 06, 2004, 07:34 AM:

Meg - strictly most of the problem in Queensland (and, I think still in Western Australia) is 'malaportionment' rather than 'gerrymander'. Rural electorates were allowed to have far less voters than urban ones.

Usually each electorate has to be within 10% (I think) above or below the official standard number of voters/electorate. It meant a given number of rural voters elected several members more than that number of city folk.

Because of the way our population is highly centralised, with a small number of big cities, not that many medium-sized towns & some very low densities in the outback, if the voting distribution is equal you end up with huge rural electorates. One (Murray-Darling) takes up most of western New South Wales, and there's another that's supposed to be the largest in the world, taking up a big part of Western Australia, our largest State.

#61 ::: Larry Brennan ::: (view all by) ::: July 06, 2004, 01:29 PM:

Epacris - no signature match? While it doesn't block double voting, it does help prevent impersonation. Every place I've ever voted has had either the original registration signature or a facsimile in the voting rolls for comparision by the poll workers.

As far as malproportionment goes, you've got nothing on us in the States. We have an entire legislative body that's deliberatly malproportioned. The 500,000 people in Wyoming have as much Senatorial power as the 35,000,000 in California. Then again, they do have to live in Wyoming...

#62 ::: Xopher ::: (view all by) ::: July 06, 2004, 02:11 PM:

Larry - Even more, the inhabitants of Rhode Island have grossly disproportionate influence in the Electoral College. One of the many reasons that body should IMO be abolished in favor of direct one-person-one-vote election.

#63 ::: Larry Brennan ::: (view all by) ::: July 06, 2004, 03:03 PM:

Xopher - agreed, but it aint gonna happen. There are enough small states to block the necessary amendment to the constitution. Why would VT, NH, RI, WY, NE, SD, ND, MT, ID, NM, AK, HI, DE, ME, IA, NV or any state with less than about 10 EV's go for such a change?

The big beneficiaries would be NY, CA, TX, FL, PA and IL, states that are mostly disliked by the small states.

#64 ::: Xopher ::: (view all by) ::: July 06, 2004, 03:37 PM:

I agree, Larry. And the people in power benefit from gerrymandering also. There are some injustices that are difficult to change; that doesn't make them less unjust.

#65 ::: CHip ::: (view all by) ::: July 06, 2004, 05:56 PM:

David Goldfarb notes

And indeed we find that the Hugo awards use just such a preferential ballot, and they are counted by computer. In the admittedly unlikely event that the US adopts it, perhaps we could adapt the Hugo-voting software.

Not only are they counted by computer, they've been counted by computer for decades; a recent cleanup at the NESFA clubhouse discovered the punch-card ballots for the 1971 Hugos. But there are a lot of counting programs, most of them probably too inflexible.

#66 ::: Jonathan Vos Post ::: (view all by) ::: July 06, 2004, 08:55 PM:

Maybe there should be RETRO-retro-Hugo Awards, for Hugos that would have been awarded if there had been Worldcons 100 years ago.

Would votes be counted Analog computers, or a Babbage Difference Engine?

Think of it: Guests of Honor H.G. Wells, Jules Verne, and H. P. Lovecraft...

#68 ::: fidelio ::: (view all by) ::: October 19, 2004, 09:23 AM:

Spam at 11 o'clock.

Smaller type (our default)
Larger type
Even larger type, with serifs

Dire legal notice
Making Light copyright 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017 by Patrick & Teresa Nielsen Hayden. All rights reserved.