Low standards in high places

by Henry on September 20, 2003

Technical standards are dull stuff for the most part; engineers or techies talking to other engineers or techies about the appropriate ways to implement this or that. While the politics of standard-setting is interesting in its own right^1^, it usually isn’t a very _political_ kind of politics. Here comes a prominent exception.

Via “Cory”:http://boingboing.net/2003_09_01_archive.html#106401808835764965 at BoingBoing: the Electronic Frontiers Foundation (EFF) has issued a “call to arms”:http://www.eff.org/Activism/E-voting/IEEE/ over voting machine standards. According to the EFF, various vested interests are trying to push through a weak standard for voting machines in the Institute of Electrical and Electronics Engineers (IEEE). If the EFF is right, this isn’t just an argument over technical issues; it has potentially serious consequences for politics and vote-fraud.

So what exactly is going on here?

To understand the politics of this, it’s necessary to start with a bit of background information. The IEEE, like other standard setting organizations in the US, is a private organization. As a consequence, unless you’re a paid-up member of the IEEE, you’re not going to have much input into IEEE standards. This said, once you’re a member, you should have a reasonably good chance to influence proceedings; discussions among members are usually driven by the desire to find a consensus that everyone can live with. Finally, even if the IEEE has no official ‘power’ to make others comply with its standards, it possesses quite considerable informal clout. Not least, standards set by private organizations such as the IEEE may very substantially influence government regulations.

So what’s happening with voting machines standards? According to the EFF, a committee within the IEEE is putting forward a standard for voting machines that doesn’t properly address security risks.

bq. the standard fails to require or even recommend that voting machines be truly voter verified or verifiable, a security measure that has broad support within the computer security community.

Further, the EFF suggests that the usual procedures of consensus-building did not take place. Indeed, the committee in question had

bq. serious procedural problems … including shifting roadblocks placed in front of those who wish to participate and vote, and failure to follow basic procedural requirements. We’ve heard claims that the working group and committee leadership is largely controlled by representatives of the electronic voting machine vendor companies and others with vested interests.

The draft standard has now gone out to vote among IEEE members; the EFF is trying to mobilize them against it. If the standard goes through, there’s a real risk that it could be adopted across the US; the IEEE

bq. sits on an advisory committee to the forthcoming Election Assistance Commission established by the Help America Vote Act (HAVA).

In other words, even if the IEEE standard doesn’t have the force of law, it may very easily come to exert a compelling influence on future government regulation.

I’m not familiar with the specifics of this story, but the EFF is usually pretty reliable on issues of this sort – it has intimate and deep connections with all levels of the tech policy community. If the EFF says that there’s something fishy going on, it’s very likely right.

That said, I doubt that there’s any vast, right-wing conspiracy behind these moves to create a toothless standard. More likely, this is just the usual politics of regulation; companies with dodgy products watering down the standards so that they can stay in the marketplace. Certainly, there’s evidence that Diebold, an extremely prominent manufacturer of voting machines is making products with “gaping security holes”:http://avirubin.com/vote.pdf. One may reasonably surmise that companies with similarly flawed products are going to oppose any standard that would declare their products to be unsafe to use. Whether they’ll get away with it is another matter..

^1^ Or at least, interesting to political scientists; see the work of Walter Mattli, Tim Buethe, Raimund Werle, Philip Genschel and Christoph Knill among others. This use of the word ‘interesting,’ is, of course, quite a restrictive one .



Cory Doctorow 09.20.03 at 8:34 am

I agree — the conspiracy here isn’t to “deliver an election” to any politician (despite Diebold’s senior execs promising just that in respect of GW Bush). Rather, this is a consensus among vendors and electoral officials to bless the existing, fielded designs for voting machines so that when the next election rolls around, everyone will be able to point to the “standards compliant” voting machines at the polls. Nevermind that these machines are universally viewed as wholly inadequate by compsec experts who’ve studied them: if anything *other* than these machines are adopted as a baseline for the standard (which has only being proposed in response to the inadequacy of the very designs it is about to bless!) then there’s going to be a mad (and possibly unsuccessful) scramble to design, build, certify and distributed machines that comply with a meaningfully secure standard.


Jim Miller 09.20.03 at 3:14 pm

it is curious that potential fraud or error from voting machines draws so much attention, and actual fraud draws so little. For example, there was just another report from St. Louis about a mass of fraudulent voting registration cards, turned in by the leftist organization, ACORN. There was much suspicious activity in Florida during the 200 election, mostly in Democratic counties, and it drew almost no attention from the national media. One exception: Fox–but I think only Fox–reported on the voter tampering in Haitian precincts in Miami.

It is hard not to conclude that, for some on the left, including some in the media, vote fraud does not count if the victims are conservatives. Feel free to email me with any counter examples, any cases in which people on the left have protested vote fraud by the left. I would genuinely like to see them.

(For those who want some examples of fraud, see the Q&A on the 2000 election on my web site.)


meg 09.20.03 at 6:01 pm

Well, jim, there was the not-at-all-subtle flyer distributed in black neighborhoods in FL that 1) had a date of Wednesday for the 2000 election and that 2) suggested that anyone with outstanding fines, tickets, or warrants would be ineligible to vote.

That looks like fraud to me.


Laertes 09.20.03 at 8:02 pm

Shorter Jim Miller: “I won’t decide if this is a bad thing until I know which side gets screwed.”


Russell L. Carter 09.20.03 at 8:59 pm

Jim Miller seems to have missed the point that the safeguards requested by the EFF & Co. are designed to agnostically reveal *anyone* committing fraud. And as far as I can tell, the resistance by the voting machine vendors to these changes is rooted in nothing more than self-aware incompetence. It’s very difficult to make reliably secure machines, and the requested auditing capabilities will reveal this. It’s the same urge that leads many technology companies to include hw/sw benchmark publishing prohibitions in their customer licenses.


Zizka 09.20.03 at 9:36 pm

Palast’s book “Best Government Money Can Buy” covers the Florida election pretty throughly. The worst was the exclusion of a large number of mostly-black non-felons from voting because their names resembled a felon’s name.

I’ll put Jim Miller’s stories in my rumor file until he comes up with some substance.

This issue has been discussed a lot. The consensus is that electronic tallying is OK if it leaves a paper record for a recount. Otherwise fraud is too easy. Good technology exists but a lot of bad tech is being sold.

As always, I don’t think anyone should rule out the paranoid explanation too quickly. But most disagree.


Jim Miller 09.21.03 at 12:25 am

My point, which I will try to simplify for the critics above, is that it is more important to stop current, actual fraud, than to worry about potential fraud. As it happens, I am opposed both to the unsafe voting machines and the current fraud through mail balloting, fake registrations, and the like.

That none of the critics even gave a perfunctory condemnation of the attempted fraud by ACORN illustrates my point. I am opposed to all sorts of vote fraud, but think we should tackle current, serious fraud before we worry about potential fraud.

As for Gregg Palast, the fact is that thousands of felons voted illegally in Florida in the the 2000 election, partly because some Democratic election boards refused to purge their rolls, as required by state law. That isn’t me saying that; that’s the result of separate invesitgations by two newspapers that endorsed Al Gore, the Palm Beach Post and the Miami Herald. Since felons vote very heavily Democratic, any narrow Gore win would have come from the illegal votes of murderers, rapists, car thieves, and other similar upstanding citizens. Am I right to guess that Palast didn’t mention that in his book?

Again, I suggest that those interested in the problem of fraud take a look at the brief essay I wrote on the election, which mentions other examples of fraud or possible fraud, nearly all of them benefitting Gore. That is, you oppose all fraud, something that is not apparent in the comments above.


Brian Weatherson 09.21.03 at 12:49 am

Is there any actual evidence that felons mostly vote Democratic? (Where they’re allowed to vote that is.) Most felons are male, and male voters trend Republican, so if we’re using lazy stereotypes the evidence isn’t entirely conclusive.

More importantly, is there a single good argument why felons shouldn’t be able to vote? I can sort of understand why felons currently serving a sentence shouldn’t be able to vote in local elections – in prison towns this might lead to the town being run in effect by the prisoners and this *might* be a bad thing. But how any larger restriction can be squared with democratic principles somewhat escapes me.

This is not to say that election boards shouldn’t enforce the law as its written. Though given a conflict between democracy and the rule of law it isn’t always obvious which should come out on top I guess. Wouldn’t life be easier in America if there was a trusted national electoral commission?


nathaniel 09.21.03 at 1:30 am

Bev Harris at http://www.blackboxvoting.com is a good source for this.


Brett Bellmore 09.21.03 at 2:18 am

I think the general reason given that felons shouldn’t be allowed to vote, is that there’s little reason to believe that a criminal’s conception of the political good is particularly benevolent. “Hm, I think I’ll vote for that guy; I suspect he’ll fire half the police force, and he looks like the kind of guy who’d take bribes.” ;) My personal take on the matter is that felons should be restored ALL their rights upon serving their time, as the alternative, maintaining a system of different classes of citizenship, has too many costs to our liberty. Particularly since the class of “felony” crimes has been vastly enlarged over the years, to encompass acts which formerly were mere misdemeanors, if they were crimes at all.

As I recall the “felon” list contraversy in Florida, the list was actually distributed as a list of *possible* felons, who elections officials should look into. It was NOT, (I’ve seen the cover letter.) represented as a verified list of actual felons.

Some elections officials wrongly purged people on the list without verifying their status. Others wrongly failed to purge people who legitimately WERE disqualified from voting. Both acts were wrong, and compromised the election.

Oh, and why would anyone think that convicted felons vote heavily Democratic? Because they’ve been polled on the subject, and that’s what they say they do, naturally! Though you’d expect it just based on the demographics of the criminal population.


Russell L. Carter 09.21.03 at 5:40 am

“That is, you oppose all fraud, something that is not apparent in the comments above.”

“Jim Miller seems to have missed the point that the safeguards requested by the EFF & Co. are designed to agnostically reveal anyone committing fraud.”

Enough said.


Zizka 09.21.03 at 4:15 pm

If Jim Miller hasn’t read Palast on the Florida election, as he confesses, he shouldn’t be going around explaining things to people. What I know about the election is that the rolls **were** purged and that among those purged were large number of non-felons, as well as felons whose rights had been restored in other states.

I did finally look at Miller’s two-year-old 2000 election FAQ. It was all allegations with almost no evidence. For example, Jim says that election fraud “almost certainly” tipped Oregon for Gore. He gives no reason why he thinks that. I live in Oregon and I haven’t heard it, though I suppose some Republican berserker floated a trial balloon at one time or another.

The internet is without referees, and there’s nothing to stop people from standing up and making unsupported accusations. And in my experience, they keep coming back and back again, like punch-drunk boxers who believe that they win if they can stay on their feet. (And of course, besides there being no referee, on the internet the other boxer can’t hit you either, so **of course** they are able to stay on their feet). And in my experience, people who work this way are mostly right-wing hacks.

I imagine that at this very moment he’s back at the bar, telling his buds how he whipped our asses singlehanded.


clew 09.21.03 at 11:09 pm

_The Revolt of the Engineers_, by one Layton, is a history of the non-revolt of the engineers; of their coöption by the possibility of promotion to management, and their occasional organizational support of status quo against obvious, incompetent, lethal bad engineering.

So; the IEEE could be showing its belly for any set of suits, independent of current party or faction.


Nabakov 09.22.03 at 2:42 pm

“The internet is without referees, and there’s nothing to stop people from standing up and making unsupported accusations.”

Oh yeah? Prove it!… and with a paper trail thanks.


Matt Weiner 09.22.03 at 5:00 pm

“The internet is without referees, and there’s nothing to stop people from standing up and making unsupported accusations.”
Oh yeah? Prove it!… and with a paper trail thanks.

During and after the 2002 South Dakota Senate Race, Republicans accused Sen. Tim Johnson’s campaign of massive voter fraud on Indian reservations. Defeated Republican John Thune’s campaign gave SD’s Republican attorney genera, Mark Barnett, 50 ballots alleging fraud.

Barnett found that only three of the affidavits actually alleged illegalities; two of those three affidavits were themselves certainly fraudulent, and the third was suspect (the accuser could not be located).

The notary public who had verified the affidavits had been travelling through the reservations with preworded accusations of voter fruad, looking for people to sign them.

[Two election workers on reservations had submitted false registrations; there is no evidence that they led to any illegal votes, and it seems as though the Democratic Party was the victim rather than the perpetrator of the fraud in this case.]

But if you google on “south dakota senate voter fraud”, seven of the top ten hits imply that fraud may have occurred.

(1) Zizka is right–disproved allegations stick around on the internet.
(2) I’m not believing any GOP accusations of voter fraud till I see good evidence, from old media. In this case, the fraud accusations were an attempt to suppress the Democratic vote.
(3) Mark Barnett is a mensch.


Matt Weiner 09.22.03 at 5:02 pm

Oh, hat tip to Daily Kos and Josh Marshall.


Nabakov 09.23.03 at 4:02 am

Opps, feeble humour alert.

“Oh yeah? Prove it!… and with a paper trail thanks.”

was just me being silly re Zizka’s preposterous claim that people sometimes make unsupported accusations on the internet.


Matt Weiner 09.23.03 at 4:37 am

Oh how embarrassing. Well, the SD story is always worth rehashing–and it does seem on topic that the GOP has cried wolf on fraud before.

Comments on this entry are closed.