For further comment on Durbin, see my previous posts .

The panel was asked to discuss a particular provision in the Durbin Amendment (Section 920(a)(5)) that allows debit card issuers to recover some of their fraud costs through higher interchange. In its draft rules, the Federal Reserve more or less threw up their hands, asking the community to tell it which of two options would work best:

1) A prescriptive rule, where the Federal Reserve picks a list of fraud-prevention technologies, and allows issuers to increase their interchange rate by some amount to cover the cost of implementing one or more of those technologies;

2) A non-prescriptive rule, where issuers document their actual costs of preventing fraud, and are allowed to increase their interchange rate by some amount to cover that cost.

Reading the law as written, it is clear that some variation of #2 was what Congress intended; however (as the Fed frankly states) this is poor policy, for it shifts the costs of fraud management onto the merchants, without ensuring that the merchants share in the benefits. Also, option 2 would be very expensive and difficult to administer, because each bank would have to document their fraud costs (which could be an epic project in itself), and bank examiners would have to decide which of the fraud costs to allow, and what the resulting interchange rate would be.