THE CASE OF CONFIRM: POSTAL PRICING AND PUBLIC GOODS

 The following is a postal perspective by Joe Lubenow, President of Lubenow and Associates. Joe also serves as a Director of the Association for Postal Commerce, a member of its Executive Committee, and as Chair of its Postal Operations Committee. He is well-known nationally and internationally as one of America's foremost experts on mail addressing and address data quality. The views expressed are the author's and do not necessarily reflect the official policies or views of the Association for Postal Commerce.

The USPS Confirm program, since 2002, has been a feasible, low cost way for some mailers to track outgoing mail and incoming replies, and to use the data to improve delivery consistency.

Confirm also helps the Postal Service move toward an automated environment in which mail pieces uniquely identify the sender, the mail piece itself, and the delivery point.  These are the minimum attributes of mail that deserves to be called “intelligent”.  Such mail is essential for future USPS programs including seamless verification, postage accountability, validation and improvement of address quality, and service performance measurement.  Confirm has also provided the Postal Service with a useful tool for detecting operational detours, redesigning work flows, and providing vital operational information to management.

It would seem that the Postal Service would be doing everything it can to encourage greater use of Confirm.  But that is simply not the case with the proposed Confirm re-pricing in R2006-1.  The stated reason for re-pricing Confirm, with a charge for every scan, and prices varying by mail class, is to make sure it covers its projected costs, but there are other ways to do that.  And we have experience with a different pricing model, which does not tend to inhibit usage, namely, the way Confirm is now priced (see #3 below).

In a broader sense, Confirm provides public benefits that were not taken into account while allocating the costs of the program.  This issue has more to do with public policy regarding new types of ancillary postal services than with the kind of argument best suited to carry the day in a postal rate case.  So we ask members of the postal and mailing community, and not only legal and economic specialists, to consider these three broad alternatives:

1)  Provide an incentive for mailers to use Confirm.  Though this is the opposite of the current policy, it is not as novel as it may seem.  Think of this not as a barcode discount but as an information discount.  In other words, mailers provide the Postal Service with intelligent mail pieces to provide the foundation for an entire new way of conducting business.  This in turn makes possible ancillary services that mailers may or may not utilize on a case by case basis.  Other postal services are considering small discounts for mailers to use Intelligent Mail bar codes (whether 4-state or 2-D) while charging them for optional services.  That way they can get a high adoption rate for intelligent mail and can still produce revenue for value provided.

2) Make Confirm free for basic services.  If any services should be free, mail tracking should be free.  The argument for this is that one cannot separate the intention to mail from the desire for reliable and predictable delivery.  No mailer can say “I produced a mail piece, paid postage, and do not care what happens to it.”  Mail tracking, now that it is technically feasible at very low cost, is becoming an essential component of mail service.  On the postal side, after the anthrax crisis of 2001 and with the irradiation of mail for zip codes 202-205 continuing to this day, sender identification may help the system to keep operating in response to threats.

Should free services also include automated address corrections and address changes?  The USPS gives a nod in that direction by proposing in R2006-1 to make the first two automated address changes free for First-Class mail.  As a general rule, efficient investments in the USPS Department of Address Hygiene can be paid for many times over by savings in operations.  But as things stand today, that cannot happen, since each area must try to meet its own budget goals.

3) Keep the current unlimited scan option for Confirm in place.

Confirm costs are modest and not growing dramatically.  Projected total costs are $1,189,000 for 2008.  Revenue and subscribers are well below expectations, with just 180 subscribers to date.  Revenue for 2008 is projected at $1,517,295, with $900,000 from user fees and $617,295 from consumption of units made up of one or more Confirm scans.

But today Confirm revenue is derived from user fees.  This does not amount to free Confirm scans, because the fees are designed to cover attributable costs plus markup.  Previous USPS testimony on Confirm stated that the costs of individual scans were very low, so that fixed costs predominated.  Internet service pricing models, with limited hours of use available for a lower rate, and unlimited service for a higher rate, are comparable to Confirm pricing today.

The USPS advocated in MC2002-1 that more usage of Confirm would be salutary.  In proposing an unlimited option for subscribers, the USPS had its reasons.  According to witness Kiefer:

Once the Confirm^® hardware and software are in place, the cost of additional scans is extremely small. A transaction based price would accordingly exceed the true marginal cost by a large factor. This would be economically inefficient pricing, and would likely produce several undesirable outcomes:

·       It would lead some potential customers to restrict usage by barcoding only some mailings or by just "seeding" barcoded pieces within a larger mailing…

·       It would also increase administrative costs since each transaction would have to be tracked and billed.

·       It would increase the difficulty of projecting Confirm^® revenues, since they would fluctuate depending on customers' potentially volatile needs.

One possible barrier to keeping the unlimited option in Confirm pricing may be the issue of variable costs.  The USPS proposal in R2006-1 shows help desk costs as the largest component of Confirm variable costs.  In turn the variable costs closely match the amount to be raised from charging for scans.  So the users apparently would pay for scans in order to fund the help desk.  It would be better to build those costs into the subscription fees, or even to separate out fees for help desk services, rather than inducing mailers to ration their use of Confirm, as the current proposal would encourage.

Regardless of whether one of the three above approaches is adopted, or whether the USPS proposal of charging for every scan is used, the following arguments are applicable:

a) Some Confirm costs should be institutional rather than attributable.

In MC2002-1, the Postal Rate Commission stated about Confirm, “The new service is designed to benefit both mailers and the Postal Service” (Opinion and Recommended Decision, p. 3).  As a direct beneficiary, the USPS arguably should also bear part of the costs.  In the long run, it needs intelligent mail to survive, and Confirm is directly on the critical path toward intelligent mail.  But this is hard to argue under current law.  It would not be easy to show definitively what is the exact proportion of costs that should be borne by the USPS.

Not just the USPS, but also the public, benefits from having a postal system in which mail pieces can be tracked.  As a USPS witness in MC2002-1, I testified that the Postal Service should be a “glass box”, that is, transparent, rather than a “black box” lacking internal visibility.  That perspective remains relevant.  Confirm could also be useful to individual members of the public, who care about avoiding credit card late fees, or whether a birthday card was on time.  The USPS had agreed to look into meeting that need with Confirm, but so far nothing has come of that.

Mailers who use Confirm benefit directly from the information, and mailers who do not use Confirm also benefit when the Postal Service and mailers take advantage of the insights gained to fine tune the system and eliminate bottlenecks. 

Many mailers are aware that the concept of ECSI value (educational, cultural, scientific and informational value, as for instance of magazines) is a criterion in rate making.  Looking at this more closely, the statute specifically applies this value to “mailed matter”.  From a public policy point of view, it is possible that this concept could be defined more broadly.  Ancillary services such as Confirm that are closely connected with mail delivery may have a kind of ECSI value of their own, mainly informational.  They increase the value of mail as a means of communication.

b) Confirm prices should not discriminate among mail classes.

The USPS has recently introduced several cases of value-based pricing, with ratios of three to one for re-positionable notes in Standard mail versus First-Class, and four to one for some address changes, though cost-based reasons also apply there.  With Confirm, the USPS has devised a “currency” of units, and without providing either a value-based or cost-based reason, has proposed that scans in other classes will cost five units to one unit for First-Class scans. 

Perhaps the USPS is trying to price mail the way that airlines price first class and coach seats.  The problem is basically that both parts of the plane arrive at the same time, just as automated letters do (and as automated flats increasingly will do) once loaded on the machines. To secure a premium price, the airlines throw in additional “perks” in first class, and then charge more for the same services or make them unavailable in coach.

If it is a marketing objective to differentiate First-Class from Standard, then why not go further?  Rather than Standard mail tracking costing five times more than First-Class, perhaps all the ancillary revenue should be derived from Standard mail.  This may work for private air travel, but the USPS is a public service with a monopoly over an important means of communication that wants to charge users to find out if the service is working properly, and then to charge them more for that in Standard mail, where the service is less predictable and less reliable.  Value-based pricing in this set of circumstances should not even get off the ground.

c) Confirm should be accountable to users for its service deficiencies.

Such defects as invalid operations numbers or erroneous zip codes when scanning Planet Codes and adding event information could be reduced or eliminated if the USPS made improvements as a service provider.  Causes of these defects are likely to be local and can be reduced by management discipline and by software validation.  It would be reasonable for Confirm service providers faced with a per-scan or per-unit cost structure to push for rebates or credits for deficient scans.  This may lead to additional administrative costs.  At the same time it might work to reduce the frequency of deficiencies. 

d) The percentage increase of the proposed rates should be mitigated.

It is not easy to calculate the proposed percentage increases for various Confirm user profiles in R2006-1.  The Postal Service does not even try to do this.  But since it is desirable to avoid, where possible, very large increases in rates that may have a disruptive effect on the industry, the issue should be raised.  Our analysis suggests that an unlimited service provider with an equal mixture of First-Class and Standard mail will face a price increase ranging from 100% to 400%, increasing with the total number of coded mail pieces, and greater still when Standard mail predominates.  Given the high elasticity of demand attributed to Confirm by the USPS itself, this will certainly reduce the usage of Confirm, contrary to previously expressed USPS goals.

The important point is that the USPS proposed pricing for Confirm is fundamentally unbalanced.  It needs to take more fully into account the value of intelligent mail to the institution itself, to the mailing industry, and to the larger society.  It would be helpful to retain some positive features of the current subscription pricing structure, including the unlimited service option.  If that can be achieved, discrimination by mail class when pricing scans will also fall by the wayside.