From: <spm2101@columbia.edu>
  To  : <cpc@emoglen.law.columbia.edu>
  Date: Thu, 12 May 2005 13:34:20 -0400

Second Paper

Paper II:  One Possibility for Curbing Identity Theft

By Steve McBride

	Many problems with identity theft arise because information is so
easy to misappropriate.  Information generates so many benefits to
both companies and consumers no one is willing to restrict its
exchange.  In this paper I theorize a system that gives individual
information more security while keeping its benefits.
	My idea is to centralize sensitive databases.  Only commercial data
brokers can maintain these databases.  The data brokers can only
sell the outputs of searches (such as name and contact information)
as opposed to inputs such as income level or purchasing habits.  The
data broker alone has access to sensitive information.
	Under this regime, Choicepoint customers could no longer buy
databases of information from Choicepoint.  Instead, clients tell
Choicepoint what they want. They negotiate the records and
algorithms that are useful in creating the output list. Then,
Choicepoint generates the customer a list with individual names and
contact information.  The sale of any extraneous information would
be illegal.
	Companies would be required to register as data brokers to maintain
databases with sensitive individual information.  Restricted
information would include sensitive information such as credit card
numbers and social security numbers.   Other valuable information,
like shopping habits or driver’s license numbers, should probably
be restricted as well.  There would have to be exceptions.  Banks
for instance will always need access to their customer’s account
information.  However, exceptions can be given case by case.
	There are several advantages to this scheme.  First, data brokers
would be liable for security breaches, so they would be encouraged
to minimize theft in order to minimize the threat of litigation.
Data brokers would be compensated for this loss because they would
get all of the data mining business that private corporations had
been doing. The transition would be relatively smooth since the
major players would still satisfy more or less the same roles, and
could adjust to their new roles relatively painlessly. Finally,
sensitive information would be located in fewer places, meaning
fewer people have access to it.  This would mean data thieves would
have less chance to access valuable private information.
	Registration would impose strict liability on the data brokers for
any damages arising from the lost information.  This would include
attorney’s fees and other incidental consumer costs.  The
possibility of punitive damages should also be included to increase
the risk associated with litigation.  The goal is to force data
brokers to create an internal procedure to quickly deal with
security breaches.  Perhaps the data broker would internalize some
of the costs and create a more efficient market. If not, at least
victims of identity theft would be fully compensated.
	The data broker could never give sensitive information like social
security or credit card numbers out to clients.  If merchants need
unique identifiers for the purpose of verifying individuals, they
can assign individual records random unique identifiers.  These
identifiers must be truly random and not based on individual
information.
	Obviously data brokers will not like the extra liability, but
that’s the trade off for opening up a large new market.   Data
brokers would get all of the business that private corporations
currently perform in house.  Corporations won’t mind outsourcing a
department that doesn’t generate revenue.  The data brokers will be
more efficient data miners than the corporation, so the corporation
could possibly save money by using the data broker.  Transaction
costs wouldn’t be prohibitive; long term contracts would minimize
the cost of negotiating for each new database query.  Data brokers
would take over the operation of programs like rewards clubs for
customers.  Instead of the super market collecting and analyzing
customer buying habits, data brokers would.  The information would
still be collected at the supermarket, but it could be uploaded to
a data broker instead of the supermarket’s headquarters.
	For example, Kroger wants to know its customers better.  So Kroger
sets up a rewards club.  Customers fill out a form with personal
information and the form is sent directly from the supermarket to
the data broker.  For Kroger or any non licensed entity to store
this information electronically would be illegal; the data broker
must store all information.  When Kroger decides it wants to know
which customers regularly buy mangos, it asks the data broker and
the broker gives Kroger a list of mango buyers without any other
information.  The data broker is free to feed this information back
into its data base, allowing the broker to create and maintain
individual customer profiles.  Kroger, however, is prohibited from
doing this.  Kroger must go through the data broker to get this
information.
	The industry itself won’t be fundamentally changed by this system,
so this system would be easy to implement.  Corporations still get
the data services they have come to rely on and data miners still
provide basically the same function.  All that changes is that some
services now being performed in house would be performed by data
brokers.
	Since sensitive information is stored in fewer places, fewer people
will have access to the information.  Fewer people have the ability
to steal the information.  Also, fewer companies have the ability
to lose the information.  For instance, the Choicepoint theft had
identity thieves posing as legitimate companies to steal
information.  This could no longer happen since Choicepoint would
not give out the information in the first place.  Inside jobs would
still be possible, but would be much easier to trace given that only
a handful of companies have the information.  In addition, the
threat of litigation would make data brokers take more precautions
with employee access, making inside jobs more difficult to pull
off.
	 In summary, by giving data brokers exclusive jurisdiction to
sensitive personal data and then imposing liability on them for the
data’s safety, we would be able to deter identity theft.  Personal
data would be in fewer locations and would have more safeguards,
while the interests of all involved parties would be preserved.



Paper II  One Possibility for Curbing Identity Theft.doc
-----------------------------------------------------------------
Computers, Privacy, and the Constitution mailing list