Three Irrational Policies?

Bounced Checks:

Let A and B be banks and a,b be members of banks A and B respectively and that a writes a check to b.  Now suppose b attempts to deposit the check at their bank, then bank A charges bank B some fee m when B attempts to transfer the funds.  In general this is an investment for B since it serves as an incentive for customers of B (customers can deposit from checks–rather than just cash).  But if the check bounces, the deposit is not made, and the investment is not vindicated.  So B charges b a fee (typically much greater than m).

From the perspective of A, B is simply attempting to transfer funds from customer account a, and consuming some time from A (A has to grant permission for the transfer) in the process.  So it makes sense for them to charge B.

Clearly a is the only one at fault (assuming A has made no clerical errors).  The simple solution seems, in the event of a bounced check, for A to bill a for the transfer charge instead of B, and apply a penalty p to a.  This way B has no reason to charge b.  Typically B will return the bounced check to b, so in theory b could attempt to deposit many times and rack up charges for a under this policy.  To counter this abuse, A could implement a policy (to be written on checks) that imposes a cap on deposit attempt frequency of a given check.  Or the industry could cooperatively agree to not return bounced checks to the depositor.  Both options are fine for A,B, and b, but the latter also protects a (even though it is the fault of a anyhow) from further charges until they have sufficient funds to write a new check.

NOTE:  I was assuming B finds out immediately if a has sufficient funds, which tends to happen iff A=B, rendering the policy not quite as irrational for when they aren’t equal.

Receiving Text Messages:

Let a and b be customers of cell phone providers A and B respectively.  When a texts b, a uploads the message to a server owned by B–using A‘s bandwidth.  Then b can choose to receive the message–downloading from the server and using B‘s bandwidth.  Correspondingly A charges a, and B charges b.  But who likes paying to receive messages?

Of course if you want to do away with this policy, you must be willing to pay double for sending. B could preemptively charge A for b‘s download, and A would in turn have to double the charge of their customer a for sending.

So it’s really an issue of whether you send or receive more.  If you send more, then the 50/50 policy (paying for both) is cheaper for you.  But if you receive more than you send, then the 100/0 policy (pay double for sending) is clearly better.  If you send and receive equally, then the policies are the same cost-wise.  I wouldn’t be surprised if statistically most users sent more than they received, which would be a good explanation for the conventional policy.


If person a bribes person b and both get caught, typically both are punished with severities p_a and p_b where p_a<p_b.  a, the principal/solicitor, solicits a task from b, the agent, who has some power.  The basic idea is to punish both.  Consider instead increasing p_b and setting p_a=0 (i.e. the person who solicits walks away free).  As the solicitor has nothing to lose, the agent cannot be sure of the solicitor’s intentions on top of the fact that they have even more to lose now with the bigger penalty.  Plus if the intent of anti-bribery laws is to dismantle bribery networks, then going after the agents (who are hubs in the network since they could have many solicitors due to their power and are fewer in number) is far more critical than going after solicitors.

One problem in allowing the solicitors to walk is that it would open the door to criminal entrapment–and could even hinder the legitimate work of the agents as they are bombarded with illegal solicitations, although I’m not convinced it would be that serious.  The mere fact that the agents are all by themselves in the realm of punishment may make a solicitation seem futile and a waste of time anyway.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: