Farewell to Signatures...

Here's what all of the commentary I've read has overlooked.  Signatures are utterly irrelevant to consumers except to the extent that the slow down the transaction. (Ok, they also require those germaphobes among us to touch a shared pen when we were doing just great with a contactless NFC transaction). The signature requirement has ZERO effect on consumer liability.  Federal law already limits consumer liability on unauthorized credit card transactions to $50.  But that $50 liability only applies if (1) it is an "accepted card" and (2) the card issuer has provided a means to identify the cardholder, and those limitations mean that consumers are rarely, if ever, actually liable for unauthorized credit card transactions.  Put another way, the statute says $50, but it is basically saying $0.    

Here's why.  First, an "accepted card" means a card that "the cardholder has requested and received..." Most card fraud today is with cloned cards or digits lifted off of cards. I think it's a stretch to call either situation an "accepted card".  "Accepted card" would seem to refer to the one and only physical card a consumer has received. If so, then there is no liability whatsoever unless the unauthorized transaction is undertaken with stolen physical card. (Strangely, the CFPB's Official Interpretation of Reg Z doesn't get into this.)

Second, the "means to identify the cardholder" requirement means, at the very least, that a signature is not adequate for any sort of on-line transaction. That would mean that there is zero consumer liability for any unauthorized on-line transaction (unless the issuer has done something very unusual like require a PIN).  That's the CFPB's Official Interpretation to the relevant regulation.  For in-person transactions, the CFPB's Official Interpretation says that "This could include, for example, a signature, photograph, or fingerprint on the card or other biometric means, or electronic or mechanical confirmation."  I don't read this as saying that a signature always suffices, only that it "could," which might depend on the particulars of the signature identification given.  

One situation where a signature couldn't possibly serve as a means of identification is if there is no specimen on the card.  There's nothing that requires a consumer to actually sign the specimen block on the back of the card. I don't because doing so can only reduce my legal rights. My cardholder agreements do not require me to sign. Instead, I'm merely requested to sign by other literature from my card issuer.  Without a specimen, a signature on a charge slip isn't means of identifying anyone.  

Even if there is a specimen on the card, though, is it actually a means of identifying someone?  Perhaps it is to a handwriting expert, but in an actual transactional context, a signature is worthless as a means of identification.  This isn't simply because no one looks at the signature, but it's also because of the unusually cramped specimen space on the back of the card, the use of digital styluses that don't work well, and the fact that consumers often sign sloppily and fast on credit slips.  I would say the same thing about a fingerprint on a card (which is another method apparently blessed by the Official Interpretation)--the science on fingerprint identification is that it's not perfect.  Photographs too are problematic--one of my cards has a photo that's over ten years old.  Let's just say this--the Official Interpretation (which was the Fed's and just adopted wholesale by the CFPB) wasn't based on a careful consideration of the scientific literature on identification methods, and I'm skeptical about how much deference it would really get from a court if put to the test.  

So put this together and the only type of transaction that a consumer might be liable for $50 is an in-person transaction using a stolen physical card that has a signature line and where the consumer has actually signed the card. And even then there might not really be liability. 

If I'm right about this (and I am), what is the point of the signature requirement in the first place?  It's not necessary as a matter of contract law.  A signature is just one of many ways to show assent to be legally bound.  I suspect the signature requirement is a matter of path dependence from the days of paper check transactions, where a signature is actually required for liability on the check (as opposed to liability on the contract).  But there's also some interesting behavioral research about signatures--when consumers sign, they seem to be more likely to feel obligated to honor a contract than when they don't.  This suggests that certain old formalities, abandoned as relics of a less enlightened past, such as doing contracts under seal or having formal closings, might actually have had some value that legal realists overlooked.   

That bottom line, here, is that signatures becoming optional (the merchant's option) is not going to have much effect on consumers other than to speed up in-person transactions a small bit.  I haven't yet run down what the change in signature requirements does to merchant liability for unauthorized transactions, but I suspect it has little effect because few merchants ever actually check signatures, which they would be required to do to benefit from a reduction in their fraud liability. So all told, this is a very good development--it seems as if it is really Pareto optimal.  The only thing I mourn about this is that there are increasingly few places where one has to write something in cursive, and that means it's an art that's rapidly being lost.  But perhaps it should become like counting on the abacus or cuneiform.  

 

Comments

[JJ]

In at least one state (Illinois, if I recall correctly), the signature on the credit card receipt is what turns the credit card agreement into a written contract for the purpose of computing the statute of limitations. So removal of the signature requirement altogether likely benefits consumers from an SOL perspective.

[Matthew Bruckner]

I did an experiment and it turns out that, in my anecdotal experience, you're wrong that "no one looks at the signature" on cards.

I wrote in the signature space, "Please ask for ID." Turns out, lots of people looked at the signature space and then asked me for my ID.

Ultimately, I agree though that the signature is/was a waste of time. No merchant ever told me that my signature was unacceptable, despite using a circle with a squiggly line in lieu of signing my name.

[Adam Levitin]

My anecdotal evidence differs, but I've just got a blank box not a direction to request ID. Amtrak is the only merchant to have ever required ID (twice). I regularly sign other names or squiggles without comment.

Fwiw, photos don't seem to do much. I once took a class speaker out to coffee and had him pay with my debit card, which has a photo. Despite our differences in age, race, height, hair, and glasses, there was no issue.

[FJP]

I've also known people who put "Ask for ID" in the signature block, and made it a point to compliment merchants who noticed it and asked. However, I think that technique has become less effective, as more and more merchants have the customer insert their own card into the reader/terminal/whatever you call it. This seems to me to be correlated with near universal use of chip cards. Most of the time now, the merchant never physically handles my card.

The behavioral research cited focuses on defaults on mortgages where the documents were signed via a POA. I'd be interested in seeing broader research on how electronic transactions, eliminating signatures, and auto-pay arrangements affect consumer spending behavior. Personally, I try to avoid automatic payments, and I still make a lot of payments using paper checks, because it makes me stop and think. I find that I am paying much more attention to how much I am spending if I have to write the amount in longhand, versus clicking a "pay this bill" button, or worse yet, having money disappear from my account without having to think about spending it at all.