Voting - Part II
Secure, easily-accessible voter ID
Welcome Back to Win-Win Democracy
Let’s continue our discussion of achieving win-win solutions to problems in voting. In this newsletter, we’re going to focus on a win-win over voter ID. If you missed last time, you can either skip it and read the TL;DR below or (re)read it now:
TL;DR from last time
State Republican leaders want to use voter ID to solve voter impersonation fraud. Democratic leaders oppose voter ID requirements as a form of voter suppression.
There’s evidence neither of significant voter impersonation fraud nor of voter ID suppressing voter turnout, but it seems commonsensical to many that voter ID is beneficial without being an undue burden, and commonsensical to many others that voter ID is a solution in search of a problem, which makes it too difficult for some people to exercise their voting rights.
Emotions on both sides have been inflamed, battles rage in courts across the country, and politicians solicit funds based on the issue.
A win-win policy seems obvious: Implement a nationwide ID that would be easy-to-obtain yet secure, and which would, by law, be acceptable for use in all elections.
Republicans would “win” their battle to prevent voter impersonation fraud. Democrats would “win” their battle to prevent voter ID being a form of voter suppression.
A Request for Help: A newsletter like this will work best if we have participants from across the political spectrum holding a variety of opinions on important issues facing America. Most of my friends lean left, so many of the initial subscribers lean left. If you have friends or relatives who lean right, please encourage them to subscribe and participate.
The Freedom to Vote Act
In 2021, Senate Democrats (including two independents) introduced The Freedom to Vote Act, which would substantially strengthen access to the ballot box, increase election security, including use of voter ID when states so desire, ban partisan gerrymandering, and limit the role of “dark money” in campaigns.
Passage would be a leap forward for effective democracy; it would also make the discussion in this and the next newsletter issue moot!
But, despite Senator Manchin’s efforts to find Republican support for the bill, it is almost certainly doomed. As NPR put it, “Senate Democrats' most forceful and perhaps last push for major voting rights legislation this year was blocked by a Republican filibuster on Wednesday afternoon.”
So, let’s press forward on the narrower discussion of voter ID.
How Big is the Voter ID Gap?
Let’s do a back-of-the-envelope estimate of how many potential voters don’t have IDs that would be accepted for voting if all states required voter IDs.
The most commonly used voter ID is a state driver’s license. The Census Bureau’s Citizen Voting Age Population tabulation for 2019 estimates that there were about 231 million citizens of voting age nationwide, and the Federal Highway Administration estimates that in 2019 there were about 229 million licensed drivers in the US, with about 3 million of them under age 18 and hence ineligible to vote. Thus there are about 226 million voting-age people with drivers’ licenses.
So, about 5 million people are eligible to vote but don’t have drivers’ licenses. This could be off for a number of reasons: Some of the drivers may not be eligible to vote, which would make the gap larger; and, some non-driving potential voters will have other forms of voter ID, like a passport or state identification card, making the gap smaller.
But, back-of-the-envelope, we’re talking about a voter ID gap of more or less 5 million potential voters, or around 2%.
Does the Voter ID Gap Matter?
Two percent is small but potentially significant in close elections. Some recent examples:
The 2020 race for Chief Justice of the NC Supreme Court was decided by 401 votes out of 5,391,501 ballots cast, just .007%!
In 2000, Florida delivered victory to George W. Bush in a race that was decided by .009%.
And, of course, the 2020 presidential race was decided by a number of squeaker state races: Pennsylvania by 1.2%, Wisconsin by .6%, Arizona by about .3%, and Georgia by about .25%.
An analysis by the Pew Research Center concludes that in “the 50 presidential elections that have taken place since 1824, the first election for which statewide popular votes were both determinative and reasonably reliably reported, there have been 187 instances in which a state was decided by less than 2 percentage points.”
Important elections decided by margins within the voter ID gap are not rare.
Moreover, many people in the voter ID gap are people too poor to afford a car and therefore don’t have a current driver’s license, and taking off time from work and traveling to a DMV office to get a state identification card they don't otherwise need is too costly. Given this, people in the voter ID gap are not broadly distributed across the population, but concentrated among the poor, potentially depriving that group, particularly, of their voting rights.
How does this square with my earlier statement that there’s no evidence that voter ID actually suppresses turnout? Unless we know that people in the voter ID gap would otherwise turnout and vote, we don’t know that voter ID suppresses turnout. It is plausible but not proven that voter ID suppresses turnout.
The voter ID gap is big enough to matter. What can be done about the gap, while preserving the perceived (and also unproven) benefits of requiring voter ID?
We are awash in government-issued IDs: drivers’ licenses (and similar state identification cards for non-drivers), passports, passport cards, Medicare cards, Medicaid cards, Social Security cards, United States Uniformed Services Privilege and Identification Card (aka US Military ID), Global Entry cards, public university ID cards, gun permits, and no doubt more.
Additionally, many private organizations issue their own ID cards — employee badges, college ID cards, and private medical insurance cards are ubiquitous.
The states decide (or would decide if they implement voter ID laws in the future) which IDs to accept for voting. Among states with existing voter ID laws, which IDs are accepted varies widely. All states accept their own driver’s licenses and non-driver identity cards, as well as US passports. Some states require these documents to not be expired, some accept expired documents, and some accept recently expired documents.
Many states accept other forms of ID, ranging from other government-issued IDs with photos, student IDs with photos, and documents like utility bills, paychecks, government-issued checks, and bank statements without photos. Two states accept Social Security cards.
The common denominator among state voter ID requirements is government-issued IDs with photo. And, to eliminate the voter ID gap, that ID has to be something that everyone already has or is easily obtainable.
An ID Nearly Everyone Already Has
Nearly every legally-resident person in the US has a Social Security number (SSN).
“The SSN's very universality has led to its adoption throughout government and the private sector as a chief means of identifying and gathering information about an individual.” — The Story of the Social Security Number
We’re all familiar with needing our SSN to conduct many aspects of our everyday life: Employers require it. Healthcare access requires it. Banks require it. Getting a driver’s license requires it. Government programs require it. It is difficult to exist in the US without a SSN.
In the early days of Social Security, people had to apply for it, typically when they first became employed. Having a SSN is so essential that today there is an “enumeration at birth” process whereby a parent requests the SSN (and card) when their child is born and a similar “enumeration at entry” process so that non-citizens admitted for permanent residency in the US can obtain their SSN (and card) as part of the immigration process.
Since everyone has a Social Security card, why not use it as a voter ID?
Social Security Card is a Weak ID Card
Everybody has a Social Security number, but the Social Security card is weak as an ID card. Here’s mine:
Notice that it is even marked as “NOT FOR IDENTIFICATION.” Modern Social Security cards no longer carry this restriction, and they have been enhanced over the years to make them more counterfeit resistant. But they still don’t have photos or any means other than signature to connect the card to its bearer.
Enhancing the Social Security Card
The inadequacies of the Social Security Card as an identification card have long been recognized. Immigration and welfare reform laws passed in 1996 required the Commissioner of Social Security to study ways to improve the Social Security card process and to evaluate the cost of replacing all (at that time) 277 million existing cards with improved cards. The report is extensive and examines many alternative potential improvements, with a price tag ranging from $3.9B to $9.2B. The report made no recommendation.
Enhancing the card would effectively yield a national identity number and national ID card. There is a long tradition in the US of opposing any official national identity number as a threat to our civil rights. The ACLU, for example, opposes national identity cards and numbers. Regardless, the SSN has become a de facto national identity number despite being neither designed for nor well-suited for that purpose1.
The bottom line is that no major overhaul of the system has been implemented. In fact, with the passage in 2005 of the REAL ID Act in response to the 9/11 Commission’s recommendation that the Federal government “set standards for the issuance of sources of identification,” we are currently in the midst of a nationwide effort strengthening state driver’s licenses and non-driver ID cards rather than strengthening our de facto national system.
It would, therefore, be unrealistic to assume that we’ll have a national ID card based on SSN’s anytime soon.
Solving the Voter ID Gap Using Social Security Cards
Since everyone has a SSN, could we solve the voter ID gap by issuing a new Social Security card that includes a photo to anyone who needs a government-issued photo ID for voting? Importantly, this would only require issuing new cards to fewer than 5 million people, a less costly task than replacing the approximately 320 million existing, active Social Security cards.
Such an ID card would satisfy the requirements for all states that currently have voter ID laws. And, if it was easy and no-cost to obtain, getting one would cease to be an impediment to voting.
Here are the steps that would need to be taken:
Design a durable, counterfeit-resistant social security card that includes a photo. The technology for doing this is readily available and already used for the production of REAL ID state driving licenses and other ID cards.
Create a process for taking suitable pictures of the holders to be used on their cards. This process would need to be easy and low cost to the holder. It would also need to be possible to periodically update the photos as the holder ages.
Step 2 is the challenge.
Where would a person go to request a photo-bearing Social Security card? The obvious answer is to the Post Office: The US Postal Service has 31,330 Postal Service-managed retail office across the country (plus contract offices). The number of post offices dwarfs the number of DMV offices. Even small rural towns have post offices. Most post offices have some Saturday hours so that there would be some available times outside of the usual workweek.
A person could take their existing Social Security card together with some document that shows their address to the nearest post office, where a clerk would take a digital photo of the person, the card, and the document using an app on a post-office-owned device. The information would be transmitted to a central card-producing facility, which would mail the new Social Security card to the applicant (thus verifying that the correct address was given).
This approach would require setting up appropriate infrastructure at post offices around the country, and people would still need to travel to a post office to do it. In many cases it would be easier and less costly for non-driving individuals to obtain suitable voter ID this way than through their state’s DMV.
Pay Citizens to Get Voter IDs
The Social Security card solution is attractive because of its universality, but it does require creating suitable infrastructure. So let’s consider another approach that requires very little infrastructure.
A subscriber who read the first newsletter suggested that a good way to eliminate the voter ID gap is to “pay citizens to get IDs to eliminate any economic burden (time off work, transportation costs, etc.) that may have a disproportionate impact.” He didn’t offer any specifics, so I’m going to flesh out the idea myself.
The beauty of this approach is that it recognizes that the states have already put in place the infrastructure to issue secure photo ID cards to non-drivers that would be acceptable for voting in that state. The only issue is to remove the economic barrier for non-drivers to get one.
Well, that just takes money! The Federal government could pay any non-driver who obtains a non-driver state ID card for their time and expense of getting the card, removing the economic disincentive to get the ID card. State governments would not be impacted and would not have to change their processes.
I can’t design the process, but here’s a rough idea of how it could work:
A non-driver would request payment by providing the receipt for obtaining the ID card, together with a form with a claim for number of hours missed from work, hourly wage, and transportation costs. The request could be made by mail or online and the money delivered by check or electronically. Reasonable maximum claim amounts could be established, as well as limitations on how often a person could get this benefit.
Alternatively, the process could be handled as part of filing an income tax return, with the money delivered as a fully-refundable tax credit2. This approach would reduce administrative cost but would delay reimbursement to people who might need the money immediately.
Some people might object to making a payment without proof of the expenses. We could require proof, but that just adds mechanism and difficulty. Instead, we could treat this as we do for tax deductions that can be made without submitting proof3.
How much would this cost? It is hard to estimate without knowing what percentage of the potentially 5 million people in the voter ID gap would participate. Say that, initially, half of them participate at a cost of $50 each — that would be $125M. Then there would be a much smaller ongoing cost as new people enter the voter ID gap.
On the scale of federal government programs, this is a small price to pay to ensure that everyone who wants to vote can actually afford to get the ID that they need in order to be able to vote.
We’ve discussed two ways by which the voter ID gap could be closed. Either could work; which is chosen could be negotiated during the legislative process.
Regardless of which approach is taken, the win-win to be incorporated in the bill is essentially the same:
(Win for Republicans) States and localities may impose voter identification requirements for in-person voting. A voter identified by the state’s non-driver identification card (or enhanced Social Security Card if Congress chooses that approach) will be deemed to satisfy voter identification requirements.
(Win for Democrats) The Federal government will ensure that anyone who wants to vote can easily obtain acceptable voter ID.
Republicans could claim that they’ve secured in-person voting across the country, not just in Republican-controlled states. No longer will their attempts to introduce voter ID run into prolonged court battles with uncertain outcomes4.
Democrats could claim that they’ve made it possible for people of limited financial means to easily obtain a suitable voter ID and vote.
I’ll continue the discussion of voting in the next issue, discussing ways to control partisan gerrymandering.
The history of the SSN is interesting, especially if you have any understanding of modern data processing. I highly recommend The Story of the Social Security Number.
With fully-refundable tax credits, one receives the money regardless of whether or not one has any tax obligation.
For example, one doesn’t submit proof of charitable contributions (money, items, volunteer expenses) although one must keep records that the IRS can, in principle, audit.
Consider, for example, the history of the North Carolina Constitutional amendment requiring voter ID: It was ratified in 2018, was ruled unconstitutional in 2019, ruled valid in 2020, and again ruled unconstitutional again in 2021. It seems unlikely that voter ID will be required in North Carolina for the 2022 elections.