Postal advocate testifies against reducing USPS service standards

SteveBlog, Slideshow

By Douglas Carlson

Editor’s note: The Postal Service’s plan to reduce service standards is now being reviewed by the Postal Regulatory Commission for an Advisory Opinion. The Postal Service presented five witnesses to make its case. Last week, four witnesses submitted rebuttal testimony. Below is the rebuttal testimony of long-time postal advocate Douglas Carlson, the registrar and assistant vice chancellor for student information at the University of California, San Francisco.

On Wednesday, June 9, at 10 a.m., Mr. Carlson will conduct cross-examination of three of the USPS witnesses. You can watch a livestream of the hearing on You-Tube at

You can find all the testimonies as well as the rest of the PRC’s docket here and on our N2021-1 dashboard.

Biographical Sketch

I have been employed as the registrar and assistant vice chancellor for student information at the University of California, San Francisco since 2004.  Previously, I was employed at the University of California, Santa Cruz, and the University of California, Berkeley.  I received a bachelor’s degree in economics from the University of California, Berkeley, in 1990 and a law degree from the UC Berkeley School of Law in 1994.  I live in San Francisco, California.

I provided testimony to the Postal Regulatory Commission on post office box service in Docket No. MC96-3.  In Docket No. R97-1, I provided testimony on the rate and fee for stamped cards, problems with return receipt service, and problems with post office box service.  In Docket No. C2001-1, I provided testi-mony on problems with collection services on holidays and eves of holidays.  Finally, in Docket No. C2001-3, I provided testimony concerning changes in First-Class Mail service standards that the Postal Service implemented in 2000 and 2001.

I am providing this testimony in my role in this proceeding as an individual citizen.  Although I may refer at times to my work at UC San Francisco, my testimony reflects my own personal views only, whether concerning my work or otherwise.  My testimony may or may not reflect the views of my employer, and my testimony should not be understood to represent the views of my employer.

Purpose of Testimony

According to 39 U.S.C. § 403(a), “The Postal Service shall plan, develop, promote, and provide adequate and efficient postal services at fair and reason­able rates and fees.”  Throughout my life, I, along with most other American citizens and businesses, have depended on the Postal Service to fulfill this statutory mission.  When service problems exist, mailers expect the Postal Service to fix them, not to reduce service levels and declare that the diminished service is better than the previous service.  In submitting the proposal in this docket, contrary to the statute, the Postal Service is affirmatively planning and developing postal services that are not adequate to meet the needs of modern-day customers.

My testimony will demonstrate that the Postal Service’s proposal in this docket would not meet the needs of customers.  My testimony also questions whether the Postal Service is properly calculating the effect on volume of the changes in service standards that its proposal would cause.  And my testimony questions whether the Postal Service is properly repre­senting the preferences of customers.


As background, since late 2012, I have been placing Intelligent Mail bar codes on nearly all my outgoing single-piece First-Class Mail, both personal and business, and have used the Postal Service’s free Informed Visibility service (and the predecessor Mail Tracking and Reporting service) to monitor mail processing and likely delivery times of this mail.  Letters comprise the vast majority of my outgoing mail, but I do send some postcards and flats as well.  Although for many decades I have known the service standards for First-Class Mail, Informed Visibility has allowed me to understand and analyze actual service performance.  My observations of my Informed Visibility data have largely tracked trends in the performance data that the Postal Service reports to the public, including the decline in on-time delivery of three-day mail after the Postal Service eliminated overnight delivery for single-piece First-Class Mail in January 2015 and the poor service performance since the COVID-19 pandemic began.  Since I began my Informed Visibility monitoring project, I have observed service performance for over 42,000 pieces of mail.

In 2019, a typical year, I oversaw the mailing of approximately 4,500 pieces of First-Class Mail.  While this number may seem small in the context of the mailing industry, my insights represent a perspective that the Postal Service has not adequately considered in developing the proposal in this docket.  Nearly every piece of mail provides a service to the sender and the recipient.  The service changes that the Postal Service proposes will affect nearly the entire population of the United States.  I send and personally monitor delivery times for far more mail than most individuals in the United States do.  Moreover, much of the business mail that I send is at least somewhat important, if not very impor­tant, to senders and recipients.  My annual mailings represent approxi­mately 9,000 customer experiences — small to the mailing industry, perhaps, but more than the Postal Service has represented in this docket. 

Needs of Customers

The University of California, San Francisco (UCSF) is training the next generation of health sciences students, including doctors, dentists, pharmacists, nurses, and physical therapists, along with students in master’s programs and doctoral students who will research new treatments for disease and develop new understandings of the life sciences.  Eleven of UCSF’s basic sciences programs are ranked in the top 10 programs in the world.[1]  Various UCSF professional programs are ranked in the top five or 10 nationally, including a #2 ranking for UCSF’s medical school for primary care and a #4 ranking for the medical school for research.[2]  I am proud of my role as registrar in facilitating the admission, registration, graduation, professional licensing, and employment of our students.  (As I explained in detail in my biographical sketch, despite my references herein to my work and my employer, my testimony reflects my own personal views.  My views about my work are my personal views.  My testimony may or may not reflect the views of my employer.)

Although I would like to describe in some detail the types and purposes of the mailings that I perform in the Office of the Registrar, universities, like other institutions, are targets of IT security threats.  Some modern threat actors use social engineering techniques, such as phishing, to gain access to IT systems.  Therefore, placing detailed information in the permanent public record about our practices and communications with students would not be prudent.  I can, how­ever, provide a general overview of my office’s use of the mail.

My office uses the mail for four general functions.  First, we communi­cate important information by mail to some applicants for admission to our profes­sional schools and programs, and we communicate essential information by mail to all new students.  Second, we communicate some information to current students by mail.  Third, we send diplomas to our graduates.  Fourth, we send official transcripts of academic records to our students, our alumni, and the third parties that they designate, including employers, professional licensing boards, residency programs, and other universities.

The Postal Service’s proposal to slow delivery of First-Class Mail would only minimally affect our communication with current students because most students live within a delivery area that would remain two days.  We do have students in online programs living outside our local area, so changes in delivery times would affect these communications.  In some cases, we use the mail for communications with current students as a supplement to e-mail for time-critical communications because busy students do not always read all e-mail in a timely manner.  The loss of overnight delivery of First-Class Mail in 2012 and 2015 significantly diminished the value of First-Class Mail to communicate time-sensitive matters to students and increased our reliance on e-mail to deliver information.

We also send diplomas to students as First-Class flats and parcels.  In addition, we regularly send official transcripts of academic records.  Our students and alumni often require expeditious delivery of transcripts for employment, licensing, and further training or education.  We allow students to pay a fee for rush processing service in our office and another fee for expedited delivery, such as Priority Mail, Priority Mail Express, or FedEx.[3]  If the Postal Service slows delivery of First-Class Mail by one or two days, I expect that students and alumni will spend more money to upgrade the processing time in our office or the delivery method to compensate.  We also offer secure PDF transcripts.  I expect the use of PDF transcripts to rise if the speed of First-Class Mail service declines, providing another example of how slowing the mail will accelerate and broaden electronic diversion of correspondence.

Although all mail delivery delays concern me, I am most troubled about the effect of changes in First-Class Mail service standards on my office and our mail recipients for our communications with applicants and incoming students.  We recruit students from around the country and the world.  The changes in service standards will slow delivery of a significant portion of this mail.  Already, some student affairs staff, applicants, and new students are not satisfied that we send this information by mail instead of electronically.  The problems typically arise when applicants and students do not receive our first letter for some reason and find themselves confronting a deadline and needing a second copy quickly, or when students are admitted late and must rush to complete various tasks as the academic quarter is starting.  Existing international mail delivery times also pose a challenge for use of postal mail for this correspondence.

We have stayed with the current process because postal mail works and is secure.  An electronic process could work better, but it also could raise different support challenges when the process fails.  Also, developing an electronic process would require time from many technical and functional staff to design and implement, and thus far I have focused our resources on building new functionality rather than replacing processes that, while perhaps not ideal, work reasonably well.

A change to four-day and five-day service standards may very well be the tipping point that will drive us to replace the current process with an electronic one.  I likely will need to evaluate the quality of service to our applicants and new students against the backdrop of their expectation of instantaneous or fast communication.  I also need to satisfy my own internal constituents, the student affairs staff in our schools and programs.

For the people that I serve — applicants, students, alumni, and student affairs staff on campus — I can say without reservation that speed of delivery of information is the most important criterion in the conduct of this business.  I do not believe that most customers would consider a consistently or reliably slow service to be a replacement for fast service.  The changes that the Postal Service proposes in this docket will not provide adequate service.

For my personal correspondence to businesses and friends, I also consider five-day delivery to be inadequate.  The turnaround time would be too slow and would not match modern expectations for speed of communication.

As for my focus, specifically, on five-day delivery, the changes in service standards will disproportionately affect senders and recipients living in the western states and other distant geographic regions of the United States.  For mail originating in San Francisco, 38 percent of our current three-day mail will switch to five-day delivery.  The proportion switching to five-day delivery is 69 percent for Los Angeles, 59 percent for Portland, and 46 percent for Seattle.[4]

For First-Class Mail International letters and flats, the Postal Service sorts outgoing international mail from origin P&DCs in the 48 contiguous states plus Alaska to the Air Mail Center at JFK International Airport in New York for proces­sing, regardless of the destination.  Thus, a letter from San Francisco to British Columbia, Tokyo, Mexico, or Sydney first goes to JFK.  Currently, international mail sent from the West Coast is sorted at JFK as soon as two days after mailing, although this time period sometimes extends several days longer.  If the Postal Service starts transporting mail from the West Coast to New York by truck, this change will add two to three days to the already potentially long total delivery time for international mail, rendering international mail less useful and practical for business and personal corres­pondence than it is now.  Five days may pass before international mail from the West Coast even departs from the United States.

Loss of Volume

The Postal Service asked witness Thress to estimate the loss of mail volume that will result from the current proposal, which would slow delivery of approximately 39 percent of mail by one or two days and slow delivery of approximately 10 percent of mail by two days.[5]  The Postal Service estimated that delivery times will increase by 18 percent overall.  Witness Thress used this number in his model to produce an estimate of the loss of volume that the current proposal will cause.

Although I understand witness Thress’s analysis, I believe that it overlooks how customers actually think about mail delivery times and whether to use the mail.  According to USPS-LR-N2021-1/9, the current average delivery time for First-Class Mail is 2.5693 days, and the projected increase is to 3.0508 days.  In all my years as an individual mailer or as the registrar at UCSF, I have never observed people thinking about their mail as taking, for example, 2.5693 days to be delivered.  Instead, they want to know whether it will be delivered in two, three, or more days.  When students and alumni are deciding whether to pay for rush transcript processing or to upgrade from First-Class Mail to an expedited delivery service, or to purchase a secure PDF transcript instead, they are considering discrete numbers of days to delivery, not decimal numbers.

Since most people think in discrete numbers, we are not going to consider the change in service standards as increasing delivery times for our mail from 2.5693 days to 18 percent more, or 3.0508 days.  Instead, we, and the public in general, will need to think of the mail as taking two to five days instead of two to three days.  The focus will become five days, in part because people do not know the Postal Service’s service standards in detail and will have minimal basis for thinking that five-day delivery might not apply to their mail.  People will consider the longest time in the range.  Thus, for our mailings to applicants and new students, we will need to consider whether a process that may require five days for delivery instead of three will still meet our needs.  (Since the fifth day can be a Sunday or holiday, we might need to imagine that the mail could require six or even seven days, such as for a Tuesday mailing to a five-day destination with an upcoming Monday holiday.)  Trans­lated to a percen­tage, our thought process, by focusing on two to five days instead of two to three days, will consider the change in service standards to be a 67-percent increase in delivery times, not an 18-percent increase.  Our consideration of a five-day delivery standard is prac­tical because we recruit students from around the country, and enough of our mail will be subject to a five-day delivery standard to cause us to review our processes.

I am concerned, therefore, that witness Thress is essentially plugging a number into an equation, but the magnitude of the change in delivery times is so large and so unprecedented that an 18-percent overall increase in delivery times does not begin to capture the effect of this change on the behavior of the mailing public.  No historical data exists to predict the effect of a change of the proportion proposed in this docket.  In fact, the 2020 presidential election and the pandemic focused public attention on speed of mail delivery and revealed that the public cares about speed very much.  The public may even care more about speed now than prior to 2020, thus weakening the reliability of historical data relating mail volume to speed.  Extending mail delivery times by two days could very well be a tipping point that prompts changes in mailing behavior that previous changes in service standards did not.  Therefore, I believe that the Postal Service is under­estimating the loss in volume that the changes in service standards will cause — or, at the very least, is underestimating the risk of a major loss of volume.

Customer Preferences

Witness Monteith testified that, if the new service standards are imple­mented, the Postal Service would expect to achieve on-time performance 95 percent of the time.[6]  Witness Monteith further asserted, “Consistent service performance will likely better align customers’ expectations with actual delivery performance.  As a result, these proposed changes may improve customer satisfaction and minimize any financial impact resulting from the changes by increasing the Postal Service’s ability to consistently deliver mail within the customers’ expectations.”[7]  In reality, the changes that the Postal Service proposes in this docket will not improve customer satisfaction for two reasons.

First, in more than 35 years of closely observing the Postal Service and the shipping industry, I cannot recall an instance in which a postal service or shipping company increased customer satisfaction by slowing delivery service.  Quite the contrary, all else equal, speed is important to most shippers and recipients.  As the Internet and various local delivery services increase the speed of life and commerce, the importance that customers place on speed is increas­ing, not decreasing.  The proposal in this docket will not increase customer satisfaction because it violates the central tenet of the shipping industry that faster is better than slower.

Second, the Postal Service misunderstands how customers think of delivery times and service performance.  In all my years of discussing postal issues with members of the public who do not work in or study the mailing industry or the Postal Service, I have never met a person who actually knows the Postal Service’s service standards.  Specifically, most people do not know that service standards exist or, if they assume that standards of some kind exist, they do not know much about the standards.  Some people may have investigated the service standards between a few city pairs for particular mailings, but their under­standing is not broad.  They do not know the nationwide design of the service standards, such as two-day delivery within a six-hour truck drive time between origin and destination processing plants.  When we had overnight delivery of First-Class Mail, most people I met did not know, or underestimated, the geographic reach of the one-day and two-day delivery areas.  While witness Monteith is theoretically correct that meeting service standards a higher percen­tage of the time could, all else equal, prompt an increase in customer satis­faction, his assumption that customer expectations of delivery times align with service standards is unsupported, if not incorrect.

Instead, in my observation, the public has a general idea of how long mail takes to be delivered.  People gain this information from anecdotal experience over the course of many years.  Individual customers’ expec­tations about delivery times thus are based in part on the service level that the current service standards produce — as distinct from the service standards themselves, which they probably do not know.  For example, in FY19, the Postal Service delivered on time 81.4 percent of the mail committed to a three-day to five-day standard.[8]  (Since the three-day standard is the maximum within the 48 contiguous states, and mail within the 48 states comprises most of the volume, this score is largely providing service performance results for mail subject to the three-day standard.)  By the fourth day, 93.9 percent of the mail was delivered, and by the fifth day, 97.2 percent of this mail was delivered.[9]

Under the Postal Service’s proposal, some of this mail will shift to a four-day or five-day service standard, for which the performance target will be 95 percent.  For mail for which the service standard changes to four or five days, customers will notice that the mail has slowed, particularly if the mail shifts from air to surface transportation and never is delivered in three days anymore.  An 81.4 percent chance of delivery in three days is much better than no chance of delivery in three days.  Postal managers might congratulate themselves on a job well done if they achieve slow delivery in four or five days 95 percent of the time.  However, the public will not share this opinion because the public is not thinking in terms of specific service standards.  Also, the public is not closely monitoring delivery times of enough of its mail to perceive a change in on-time performance in relation to a service standard.  For example, I doubt that most customers who mail a bill payment in advance of the deadline check later to see in exactly how many days the letter was delivered and determine whether the delivery time met the service standard.  As long as they are not assessed a late fee, the mail was delivered on time.

However, if the proposal in this docket is imple­mented, the public will notice that some of the mail that previously was delivered in three days is now taking longer for delivery.  The public will perceive the change as a deterioration in service, which is exactly what it would be, even if the Postal Service defined the slower delivery as more reliable.  Indeed, reliability, itself, is a subjective and amorphous term, and whether customers would con­sider five-day service cross-country to be reliable at the conceptual level is an open question, even if the Postal Service consistently delivered in five days.

I understand that Postmaster General Louis DeJoy wrote the following statement to industry mailers on April 19, 2021: “While some mail will be subject to slightly longer service standards, the truth is that we have not been able to meet the current standards for many years and we have no reasonable prospect of ever meeting them.”[10]  From this statement does not follow the conclusion reflected in the proposal in this docket that customers would prefer for mail to be delivered, for example, 95 percent of the time in exactly five days instead of more than 80 percent of the time in three days.  Moreover, this proposal fails to recognize that, for more than a century, mailers in the United States and other developed coun­tries have understood the implicit bargain of a postage stamp: for a modest price, the postal service will do the best that it can to deliver the mail fast; the mailer probably will receive very good delivery service; but no guarantee exists for the delivery date, and the letter could be delayed.  Unfortunately, the Postal Service proposes in this docket to slow the service and prepare to declare that service has improved.  And, in a separate docket, the Postal Service proposes to raise the price.[11]  These actions violate the fundamental trust that Americans have placed in the Postal Service for decades to develop and promote adequate and efficient postal services, as 39 U.S.C. § 403(a) requires.

Worse yet, the Postal Service is advocating for the current plan without having sought the opinions, through representative market research or otherwise, of the hun­dreds of millions of individual household mailers who rely on the Postal Service.  Therefore, the Postal Service does not know whether the public agrees with this plan to slow mail delivery or whether this plan will provide adequate service to the public.  I am startled that the Postal Service would advance such a dramatic plan to reduce service without providing any evidence that the public supports it.  I am confident in my assessment that the public would view a shift of First-Class Mail from planes to trucks as a step backward that is inconsistent with the needs of mailers in the 21st century for speedy communication



[3] Our fees are posted here:

[4] DFC/USPS-T3-10.

[5] USPS-T-3 at 22.

[6] USPS-T-4 at 19.

[7] Id.

[8] SH/USPS-T1-2.

[9] Id.


[11] Docket No. R2021-2.