By Lambert Strether of Corrente
Readers had many excellent comments and suggestions on this post’s companion piece of two weeks ago on election eve (“What ‘Our Democracy’ Should Look Like When Voting: A Simple Plan“). So I thought I’d revise the “simple plan” to incorporate some reader suggestions, and also explain why I rejected others.
How we vote is becoming an increasingly inflammatory topic, sadly ignited in our day — among Republicans — by The Former Guy’s “Stop the Steal” campaign. (While it would be foolish to deny the election theft in the United States has occurred in living memory, I am not persuaded that election 2020 was stolen, and don’t @ me (and not on RussiaGate, either)). However, it would foolish to deny that our balloting system, regardless of individual cases or desired outcomes, lacks legitimacy (“Why I Am Worried About the Legitimacy of the 2020 Election Balloting Process,” January 2020). Republicans, being Republicans and therefore more serious or at least aggressive about their politics than Democrats, are working hard both to rejigger existing electoral systems to their advantage and to remedy real problems. Into the latter category falls Republican advocacy of paper ballots, the worldwide “gold standard.” I would hate for Democrats to be able to force “hand marked paper ballots, hand counted in public” into the “election denial” frame, and so I’m writing these posts in an attempt to prevent that. I also want to put voting as an act of civic engagement, and not a reflection of partisan fealty or “team spirit.”
I will first present a recap of the postulates, principles, and plan from the earlier post. Then I will present additions, rejections, and some “cutting room floor” material I decided not to add to the plan (but readers may think differently). Readers are, of course, free to make additional suggestions!
Recap: The Simple Plan
From election eve:
Postulates: Of elections and election technology, I postulate:
(I). Digital = hackable;
(II). The financial stakes for any election are enormous;
(III). Phishing, in essence, is the proposition that if fraud can happen, it will already have happened. (Phishing is ubiquitous, especially in a financialized economy. See Shiller and Akerlof on this point.)
(A). Every citizen should have an equal chance to vote.
(B). Every voter should have the ability to vote from same fact set.
(C). Every voter’s ballot should be marked and counted using the same process.
Item (1) is modified. Other items are added. See below under Additions.
Declare election day (this year, November 8, tomorrow) a national holiday (i.e., paid). No early voting. No drop boxes. Mail-in only for those physically unable to travel to the precinct; nursing homes, the military, overseas voters, etc.
(2). Mandate that the default voting system for all precincts must be hand-marked paper ballots, hand-counted in public (modulo accomodations in point (3) below.
(3). Accomodations (disabilities; language; transport) should be Federally mandated and funded (by principle (D)) at the precinct level.
(4) Election resources should be evenly distributed across precincts, and remediation funded (by principle (D)) if need be.
(5) Counting, and ballot-handing generally, should not be performed by party members.
(6) Assistance for voter IDs, where mandated, should be Federally mandated and funded (by principle (D)).
Three-day weekend. alert reader aj wrote:
Election Day being a holiday wouldn’t mean much for a lot of people, especially if it’s still on a Tuesday. People in the service industry work every day, holidays or not. We work on Thanksgiving and Christmas, what makes you think I wouldn’t have to work on election day. I think a better solution to satisfy both principles A and B would be to have voting over a few days or maybe up to 7, some of which should span the weekend.
After discussion, we converged on a three-day weekend. (In my view, seven days is long enough to concoct a “November Surprise,” gin up a oppo attack or a Twitter dogpile, etc. This would be much harder over a weekend.) Hence:
(1). All voting takes place over the first three-day weekend in November. The Monday is a national holiday (i.e., paid). No early voting. No drop boxes. Mail-in only for those physically unable to travel to the precinct; nursing homes, the military, overseas voters, etc.
None Of The Above. Alert reader C.O. wrote:
I think there should be a formally printed “none of the above” option required on the ballot instead so that each voter can clearly show they have not spoiled their ballot by accident or because they are so uncaring or uninformed they wrote in something stupid. Then a sensible goal would be to strive to do a decent enough job to reduce the number of votes for “none of the above,” and it could then be plausibly argued that something close to no vote being wasted is happening.
An added element to the plan:
(7) Each set of candidate choices on the ballot must enable “None of the Above” (NOTA). NOTA votes shall be tabulated, but shall not affect the outcome.
Airborne infection mitigations. Alert reader Giuseppe wrote:
[Y]ou would require my wife and me to stand in line with strangers who refuse to wear masks in the middle of a triple viral pandemic in order to exercise our right to vote.
An added element to the plan:
(8) Voting precincts shall mitigate against airborne tranmission of disease by minimizing wait times, installing ventilation according to ASHRAE standards, and providing masks.
It might be that, given the givens “going forward” I should add an additional Principle:
(D). No princinct should give rise to superspreading events.
Standard exit polling. Alert reader mrsyk wrote:
I would like to add standardized exit polling as a consideration.
An added element to the plan:
(9). Exit polling shall be conducted according to a plan devised, carried out, and published by the United States Census, within a month of the Monday of election day weekend.
Photo IDs. Alert reader KD wrote:
There should definitely be a national photographic ID. Using drivers licenses is crazy, not everyone has a license, not everyone can get a license and not everyone can get a “nondriver ID.” Its not like homeland security doesn’t have your information anyways.
There’s no requirement for this; voter fraud is minimal. In addition, as far as national photographic IDs go, I’m a small-c conservative. I hate RealID, and I shudder to think what the organs of state security will do with it.
Use party members to count the votes. Alert reader Rasmus wrote:
Using party members to count the votes not only solves the practical problem of finding volunteers who care enough about politics to volunteer a day as an official, it also prevents fraud and builds trust in the result. When officials are party members you know their bias and can take it into account, so ballots are always handled bo officials from different parties, thereby making fraud harder as you are always being watched by your opponent. Having the parties themselves take part in the count also helps them trust the results as they had their own people at every polling station who would have told them is something shady was going on. Officials are paid a reasonable remuneration for their trouble and municipal authorities responsible for organising polling stations will make sure they get something nice to eat and drink as well.
See item (4); lack of volunteers is a resource issue, to be remediated by money. This isn’t a hill I want to die on, but I think we want to reduce, not increase, the power of parties in the electoral system.
Bar-coded individual ballots. Alert reader Tom Pfotzer wrote:
I want a way to take my receipt, go to some public facility (office or website), plug in my ballot number, and be able to verify that my votes were counted.
I think this could work without violating Lambert’s principles. Your paper ballot would have a unique ID (UID) on it and you could keep a record of this UID for the kind of after-the-fact checking that you mention. I think that the hand-counting of the paper ballots could include the creation of a record of all the UIDs of ballots cast for each candidate (this record would have to be digital in order to subsequently be searchable with less than great effort, but provided that the actual vote count was by hand, I don’t think the existence of this digital supplementary data would compromise the count). The total number of UIDs of ballots cast for a candidate would equal the hand count of votes for that candidate.
This would allow voters to after-the-fact verify that their votes were credited to the intended candidates.
It would, however, compromise the anonymity of one’s vote if one’s UID became known to someone else.
That’s the way Singapore’s voting is done. Each ballot has a unique ID number. Ballots are stored for six months before being incinerated in the presence of ruling and opposition party members.
I don’t see a requirement; I think that the checks of a public count are sufficient to make sure that the count is correct (“ambition must be made to counteract ambition”). Let’s not fetishize digital engineering over social engineering. Conor points out that the count could include recording the UID, but that’s either digital or time-consuming. Further, I think the whole notion of “receipts” — introduced by the Ballot Marking Device people, and implying that there is a second authority superior to the ballot — is flawed. The hand-marked ballot is the ballot. In principle, the ballot is secret. If there’s a receipt for the ballot, it’s no longer secret.
Further Topics for Discussion
Mail-in ballots introduce three forms of complexity that should not exist: Ballot curing, mail tracking, and weird validation rituals to inspire public confidence:
Ballot curing. This from House Speaker Nancy Pelosi crossed my Twitter feed:
Our voice is our vote and voters have the right to be heard in our elections.
We're grateful to the tireless Volunteers In Politics helping voters cure ballots in too-close-to-call House races to #CountEveryVote as cast.
Please join us!-NPhttps://t.co/pVdGLClzdm
— Nancy Pelosi (@TeamPelosi) November 18, 2022
NPR has an explainer on ballot curing:
During big U.S. elections, hundreds of thousands of mail ballots are typically thrown out and left uncounted. In 2020, for instance, more than 560,000 ballots were rejected (that’s nearly 1% of the total).
Experts say ballot rejections are largely the result of relatively minor voter errors, often associated with security measures that are designed to verify a voter’s identity.
That’s why about half of states have a process in place to help voters fix their mail ballots if they do make a mistake. It’s known as ballot curing.
Because [genuflects] Federalism, ballot curing systems vary in the states that allow it:
States vary widely on what disqualifying issues can and cannot be cured, and local election officials often decide themselves on how to implement curing requirements with minimal state guidance.
In some states, voters can cure ballots that have either a mismatching or missing signature; in others, voters can only fix ballots with a mismatching—but not missing—signature. In select states that require witness signatures on absentee ballots, voters can cure misplaced or missing witness signatures in addition to their own. Additional disqualifying issues that voters can cure in some areas include missing Social Security numbers, unsealed envelopes that meet certain requirements, and problems with ballot statements.
While ballot curing is an important step in making sure each vote is counted, it does not provide a complete solution to ballot rejection. The lack of standardization and other disqualifying errors cannot be fixed through existing ballot curing procedures. In addition to streamlining and easing the ballot curing process, these issues and others could be further examined to improve the procedures for counting mail-in ballots.
Or, as opposed to further examination, we could solve the problem by eliminating mail-in ballots as far as practicable under rule (1). The most reliable working parts are the ones that aren’t there.
Mail tracking. From Government Executive:
The Vote by Mail Tracking Act (H.R. 1307) would require any ballot sent to voters in a federal election to have a barcode on the envelope that allows the U.S. Postal Service to track each ballot. It would allow USPS to create envelope design standards with which municipal election offices must comply.
In recent elections, USPS has “strongly recommended” that state and local governments use its barcodes and the official election logo on ballots, but there was no requirement to do so. Postal management has ramped up its efforts to coordinate with election offices, establishing teams to focus on election efforts year round and working with local officials to establish relationships and set expectations. The Postal Service’s inspector general has previously recommended the agency create tracking requirements for election mail.
Better ballot tracking would give voters better peace of mind, lawmakers said, as they could determine where in the mail stream their votes were at any moment. It would also ease the process for USPS, which every election cycle implements “extraordinary measures” to ensure every ballot is delivered to election offices by their deadlines. With greater visibility into where ballots are, postal employees would have an easier time identifying pieces of mail to pull out of the normal system to prioritize for delivery.
I agree that there should be better tracking, but there should also be much less tracking to do. (It’s also odd that we can have “envelope design standards,” but not “ballot design standards.” Surely a case of putting the cart before the horse?
Weird rituals. From the Atlanta Journal-Constitution:
With the roll of 20 colorful dice in the Georgia Capitol, election officials launched an audit of a random sample of ballots Wednesday that will be reviewed by hand across the state this week.
The audit will check whether machine counts of ballots match hand tallies, showing whether the outcome was accurate..
Gabriel Sterling, chief operating officer for the secretary of state’s office, said he hopes the ballot review will prove to voters that they can trust the results of elections.
Why not just do it right to begin with, and hand-count the ballots?
In each case, we see odd epicycle-like functionality added to and already creaky system save the phenomenon of ballot marking devices and early voting. “They add functionality by deleting code” is one definition I’ve seen of a good programmer. Let’s apply it!
Finally, there is the question of complex paper ballots. Alert reader marym wrote:
My ballot this year had 97 line items. I’ve probably said before, I personally lack the imagination to envision a manual process, even if this large number were reduced by splitting voting into a few separate elections. It’s not an argument against trying, I’d just be interested to see some ideas.
My ballot contained 48 different races or items
2 Federal races (senate & house)
10 state races
1 local water agency
1 county sheriff
1 city council (multiple votes)
2 school districts
22 judicial offices at various levels
7 state ballot initiatives
2 county ballot measures
The Electoral Knowlege Network has an amazing and exhaustive section on Voting Operations, with sections on Ballot Paper Design, and Printing of Ballots. I looked there so I could get some idea of the scale of this potential problem (because in my experience it doesn’t exist). Nowhere is there any suggestion of design or printing issues involving paper ballots that simply have too many items. And if a ballot can be printed, it can be counted. I’ll certainly keep my eyes open for future evidence, but for now, I don’t see an issue.
There are other topics I might get to at a later date: Banning political polling within a set period before the vote, and banning all forms of political advertising except in print. (This would be an enormous subsidy to local newspapers. I know that.)
I want to leave you with the sense that voting is a civic duty, and should be treated — above all, funded — as such. Thank you, NC readers, for your many excellent suggestions. Also, if there are any lawyers, jailhouse or otherwise, who want to tighten up the wording of the plan, feel free to chime in!
 If ballots have gotten to the point where they must be printed as booklets, I suggest that the problem is not ballots, but election officials (or possibly legislatures). Schedule another election; we shouldn’t be asking voters to vote on a booklet of choices in any case. I’m all for civic duty, but I’m also not for arbitary burdens.
Alternatively, as alert reader Kouros suggests, we could shift to a system of sortition. From the Boston Review, ” to a system of sortition. From the Boston Review, “The Case for Abolishing Elections“:
In a poll conducted in January 2020, 65 percent of respondents said that everyday people selected by lottery—who meet some basic requirements and are willing and able to serve—would perform better or much better compared to elected politicians. In March last year a Pew survey found that a staggering 79 percent believe it’s very or somewhat important for the government to create assemblies where everyday citizens from all walks of life can debate issues and make recommendations about national laws. “My decade of experience serving in the state legislature convinces me that this popular assessment is correct,” Bouricius said.
The idea—technically known as “sortition”—has been spreading. Perhaps its most prominent academic advocate is Yale political theorist Hélène Landemore. Her 2020 book Open Democracy: Reinventing Popular Rule for the Twenty-First Century explores the limitations of both direct democracy and electoral-representative democracy, advocating instead for government by large, randomly selected “mini-publics.” As she put it in conversation with Ezra Klein at the New York Times last year, “I think we are realizing the limits of just being able to choose rulers, as opposed to actually being able to choose outcomes.”
Hmm. Perhaps sortition for some offices?