I have two sisters. My younger sister was born in Puerto Rico; as such, she was a US citizen from birth. I’ll occasionally tell her that she must run to be the president of the United States, since she is the only one among us who is qualified to run. In response, she glares at me. The very notion that the US population elects a Chinese woman born in Puerto Rico as president is logistically absurd. While she certainly meets the legal requirement, there’s still implicit racial and gender bias that US citizens exhibit.
I became naturalized when my mom was naturalized, since I had yet to turn 18. Ironically, I helped her study for that test by memorizing the questions and answers; I would’ve easily passed the test had I taken it. I remember taking the oath as a teenager. My older sister had already turned 18, so she took the naturalization exam and passed it years later.
With citizenship came the ability to vote. As it happens, Asian stereotypes are among the most innocuous among racial stereotypes. More than once, I’ve heard the expression DWO (Driving While Oriental), but people didn’t look at us as implicit criminals. Perhaps the easiest way to describe people’s impression of us was “different, but mostly harmless.” As such, we didn’t feel the need to fight for our civil rights, since we had the best position among the “others”.
The right to vote
The Constitution grants every United States citizen the right to vote; that right extends to naturalized citizens. I have the right to vote. However, I have pointed out that voter registration is not constitutionally required to vote. Do you want another surprise? Photo voter identification is also not constitutionally required to vote. Should we know who is voting? Absolutely. However, comparing a person’s face to a tiny photo on a card is a profoundly fragile and unreliable way to verify someone’s identity. We can do better.
The United States already has a record of all its citizens. They already know that the person with a particular social security number is a citizen. I don’t need to prove what they already know; I need to prove that I’m the person with that particular social security number. In a perfect world, I should be able to establish that I’m the person with that particular social security number without physically carrying any form of identification.
With that said, let’s consider alternatives. Though first, let’s contemplate the requirements for these alternate forms of identification:
- It must rely on something physical and durable about the person.
- It must be reliable, which means extraordinarily few false positives or negatives.
- It must statistically eliminate large numbers from a positive match. The odds should be at least the US population (1:350M) if not the world population. Something like blood type is not effective.
- It must be quick, cheap, and easy to verify someone’s identity.
- It must not be usable to violate your or anyone else’s rights.
Why DNA identification won’t work
There have been interesting criminal cases involving DNA evidence. The Innocence Project often uses DNA evidence to vacate past convictions and exonerate people. While DNA distinguishes between people in a large population, such as that of the United States, it is not unique. A murder suspect once left his DNA at a crime scene; unfortunately, the suspect had an identical twin with the same DNA. The investigators needed to eliminate his twin to secure a conviction.
The Golden State Killer avoided capture for forty years. Long before law enforcement used or even understood the significance of DNA evidence, this killer committed crimes and left his DNA. Unfortunately, this DNA never matched anything on record. This killer conducted his crimes carefully; he never committed any crime where the authorities would collect his DNA on arrest.
Unfortunately for him, one of his relatives decided to have their genealogy tested through one of the services that tells you where your ancestry is from (like a particular region in Ireland), or if you are a familial match with anyone else in their database. As it happens, the results of this particular DNA test were also shared with law enforcement. They realized that the Golden State Killer is a close relative of this DNA; they were able to narrow it down from there.
I do not weep about the apprehension of this killer. However, we need to ask ourselves if law enforcement’s use of DNA information in this way is ‘fruit from the poison tree’? Is this equivalent of a search conducted without a warrant?
Moreover, it takes too long (at least days) to perform a DNA match to be feasible for voting. It violates rule #4 and rule #5 above.
The problem with fingerprints
The ability to unlock the modern smartphone through a fingerprint has existed for years. We can perform a fingerprint match quickly and cheaply. Fingerprint readers can easily be installed in all polling locations. While people with the same fingerprints exist, though exceedingly rare, they’re not as predictable as the DNA from identical twins.
However, you can leave your fingerprints behind under perfectly reasonable conditions. The water glass at the local restaurant will have your fingerprints (and DNA, for that matter). This is generally fine if there was nothing to compare your fingerprints to. Our government has no reason to collect your fingerprints and enter them into AFIS if you have not been arrested for a crime; furthermore, they can’t issue an arrest warrant without probable cause.
They could keep them in two separate databases and promise not to refer to the generic US citizen database on criminal cases. That principle worked splendidly for the Golden State Killer. If they had a fingerprint for Nancy Guthrie’s case, do you believe that ‘good faith’ would prevent law enforcement from peeking into that database?
Finally, this possibility is fraught with potential misuse. If you think that swatting is a problem, then imagine how easy it would be for someone to collect a water glass from a person they don’t like and put it into a crime scene (like burglary or murder), effectively framing them.
Therefore, building such a national database of all citizens violates rule #5. Authorities need an arrest warrant to collect your fingerprints. To collect our fingerprints implicitly, without a warrant, violates our civil rights.
Facial recognition durability
Many phones and computers can now recognize people by face. This technology boasts being able to distinguish between identical twins. I’m apprehensive about the government knowing the precise shape of my face. The cameras used to establish a good foundation for face identification have special hardware. However, it is simply a matter of time before you can throw enough computing power at regular pictures, or these specialized cameras are installed in more places.
At that point, the government can determine everywhere you go. Think about it, the government will know if you break your diet, go into a strip club, have an affair, and yes, even commit a crime. Should the government have the ability to know everywhere you go? This breaks rule #5.
However, the other problem with face identification is its durability. By the time a photo is six years old, its accuracy drops to 99%. While 99% sounds pretty good, that’s still one failure out of 100. How many people are there in line at the polling location? Facial recognition violates rule #1, rule #2, and rule #5.
The eyes have it
Using retinal scans to verify your identity seems incredibly promising. It’ll continue to work even years later; retinal patterns typically stay the same over a person’s life. Since it relies on the pattern of blood vessels in your eye, it can even distinguish identical twins. We can perform retinal scans for identification in seconds.
You won’t unintentionally leave your retinal pattern on a water glass at a restaurant, the way you would with your DNA or your fingerprints. Public cameras won’t be able to read your retinas while you walk by a storefront, and won’t be able to for a very long time. Even if the government has a copy of your retinal pattern, they need another pattern to compare it to… which they can’t easily get from you without your consent.
It’s not foolproof. Certain types of eye conditions (like diabetes, glaucoma, or retinal degenerative disorders) can affect your retinal pattern. For these people, we can use an alternative form of identification.
How do we do it?
Suppose that we agree that retinal scans are the standard for identifying people casting their vote. The very idea that we can have a voting polling location at a nudist colony, where citizens can walk up without a physical form of identification and cast their vote, tickles my fancy. If we really ponder it, there’s no reason why that can’t work. Humans have become enamored with the idea of identifying someone by photo, even when evidence suggests that it’s deeply flawed.
However, the logistics are monumental. There are currently around 275 million US citizens eligible to vote. We need to scan all of them. Or so close to all of them that it doesn’t matter. The ACLU pushes back on strict photo ID requirements because it disproportionately targets minorities. You know who doesn’t have a driver’s license? Someone who can’t afford a car. The only real way to avoid any perception of bias is to scan everyone; go door to door if you have to. Furthermore, we can’t require this new form of ID until that’s done.
Do you remember all the drama around the REAL ID and its impact on air travel once it was finally enforced in May of 2025? Congress passed the REAL ID Act in May of 2005. It took twenty years before they finally enforced it.
Naturally, I have ideas on how to get everyone scanned and into the database, but that will have to wait until the next post. 😉