According to the 2006 Nilson Report, net (not gross) industry losses on bankcard (debit and credit) fraud and theft amounted to $990 million in 2006 in the U.S. The report does not break down how much of the fraud occurs at retail points of purchase, but for the purpose of this estimate, I will assume that ALL of it occurred at a face-to-face transaction.
As of 2006, there were 984 million bank-issued Visa and MasterCard credit card and debit card accounts in the U.S. (Source: Visa USA, MasterCard International)
I will ignore Amex and Discover due to their lower market shares. If you include them, the cost per card drops.
So, on average, bankcard fraud costs approximately $1 per card per year. (Inclusive of all types of fraud - Retail, Internet, gas pump, etc)
I could not easily find cohesive statistics as to how much of the fraud occured outside of a face-to-face retail transaction. We know that plenty of fraud is present on the Internet and at automated locations (gas pumps, etc), so retail fraud and theft cost less than $1 per card per year.
I could not easily find cohesive statistics as to how many bankcard transactions occurred in the United States. For the purpose of this post, we will use very rough numbers to demonstrate the magnitude of the resources required to check ID on 100% of face to face transactions.
Assume:
-A proper ID check takes 5 seconds, assuming the cashier is fast and the customer is ready. (In reality, not everyone is fast or prepared)
-A retail employee costs $8 per hour (wages, employer payroll taxes, unemployment contributions, etc)
Therefore, it costs a merchant 1.11 cents to check an ID
If the average consumer uses a card once per day over the course of a year (a very, very conservative assumption), then the annual cost for checking that consumer's ID will be $4.06, based on 1825 seconds spent checking ID. (well over a half of an hour per person!)
If there are 100 million daily active bankcard users in the United States (a conservative number, once again), then the annual cost of checking all those IDs amounts to $406 million in employee time alone. That does not include customer time wasted standing in line, or customer time wasted when an ID check is required but the customer does not have ID present, or has differing names due to legal name changes.
$406M is rougly 41% of the $990M total fraud.
If ID checks stop 100% of fraud, then the net gain to society by implementing ID checks can be NO HIGHER THAN $0.59 per card.
I used the numbers I did to establish an upper limit on how effective ID checks can possibly be. Of course, in the real world, there are always non-face-to-face transactions and cashiers in on the fraud. ID checks can't stop that. There are always professional criminals who have fake IDs, who cannot be stopped by ID checks. Criminals will always skim and clone cards, which ID checks can't stop, unless the clerk takes additional time to compare the ID to what prints on the receipt. The total bankcard fraud is not 100% retail use of stolen plastic, either.
Depending on these other parameters, it is very possible that ID checking would produce a net tangible loss to society.
These are all hard dollar figures - they do not include such intangibles such as privacy, dignity, avoiding ID theft from showing ID, not being robbed by thieves who know your address, customer time spent waiting in line, etc, that are all lost by the implementation of ID checks.
Given the large amount of frauds that cannot be stopped by ID, and a MAXIMUM THEORETICAL GAIN to society of $0.59 per card per year, why do people continue to think that ID checks are essential to keeping prices low?
This involved all of about 15 minutes of Google, and is hardly scientific. The networks have the real numbers. They have hired people to analyze them over and over, and have allowed the current policy to stand for a reason.
