When the Postal Regulatory Commission issued its advisory opinion on the Retail Access Optimization Initiative (RAOI) last year, it said the plan to close 3,700 post offices wouldn’t optimize anything.  On Friday of last week, the PRC issued its advisory opinion on Mail Processing Network Rationalization (MPNR), and while it doesn’t come right out and say so, the Commission suggests that the Rationalization plan won’t really rationalize the network.

The mountain of testimony and legal arguments submitted by the Postal Service just rationalize a decision made long ago, and postal management doesn’t seem to care if the PRC thinks the decision was rational or not.  The Postal Service didn’t even wait for the opinion to come out.  The changes in service standards were implemented in July, the consolidations began in August, and while the plan now calls for a phased implementation, there are few signs that the Postal Service is interested in changing course.

The advisory opinion nonetheless aims to shift the direction in which the Postal Service is headed.  The opinion is long (240 pages), thorough, and extremely technical, but as the Commission states in the Introduction, its advice can be “succinctly summarized” as follows:

“The Commission views positively the network rationalization actions planned by the Postal Service through January 31, 2014, and recommends that the Postal Service take into account the considerations outlined in this Advisory Opinion before proceeding further.  Specifically, the Commission encourages the Postal Service to make every attempt to retain overnight delivery in keeping with the analysis presented in the subsequent chapters.”

That summary of the Commission’s findings contains several important points.  First, the Commission is basically approving the first phase of the Postal Service’s two-phase implementation plan, which was announced back in May.  According to this revised plan, some 140 plants will be consolidated as part of phase one, starting this past August and running through January 2014.  Phase two will begin after that, and include about a 90 more plants.  (A list of the plants is here.)

The Commission expresses approval for phase one because it preserves overnight delivery for most of the mail that currently enjoys it.  The Commission is much more skeptical about phase two because it requires ending overnight delivery for almost all First-Class mail, and it would cause most of the damage in terms of revenue losses.  The Commission advises the Postal Service to consider the opinion’s recommendations carefully and to review what happens in phase one before proceeding with the second phase — if it moves on to phase two at all.

A large part of the advisory opinion reviews the detailed cost-savings analysis provided by the Postal Service.  It concludes that “the Commission’s range of potential net savings estimates is lower than that projected by the Postal Service.”  The Postal Service says that the gross savings from the plan (before deducting lost revenue) will be $2.1 billion.  The Commission, on the other hand, says the gross savings could be as low at $54 million.

As for how much business the relaxed service standards may drive away, the Postal Service claims that its market research projects a net loss of $500 million (later corrected to about $430 million), bringing the total savings to $1.6 billion.  The Commission, however, found the market research to be so problematic that it could put no credence in the bottom-line estimates.  The Commission would not offer an estimate of its own, but the advisory opinion leaves one with the distinct impression that the lost revenue will be greater, perhaps much greater, than the Postal Service is anticipating.

All in all, then, the Postal Service’s estimate of $1.6 billion in net savings comes under serious scrutiny in the advisory opinion. The Network Rationalization plan will probably save much less than the Postal Service says, and there is the very real possibility that rather than saving anything, it will actually lose money, drive the Postal Service deeper into its downward spiral, and, as one of the expert witnesses put it, “herald a death knell” for the country’s postal system.

 

The great unspoken

For the most part, the advisory opinion presents a technical analysis of the Postal Service’s plan and the alternative models that were developed by outside experts.  It was written by a staff well versed in the technicalities of postal operations, postal law, and the principles of statistics.

The advisory opinion mentions the word “jobs” only once, in a sentence about how the Postal Service is “a vital component of a mailing industry that supports millions of American jobs.”  There’s not one reference in the advisory opinion to the jobs of postal workers, even though nearly all of the savings the plan purports to achieve will come by eliminating their jobs, and lots of them — 28,000, according to the USPS press release.

The advisory opinion doesn’t discuss how closing 229 large processing plants will affect postal employees or their families and communities.  There’s no discussion of the economic impacts of shutting down a large plant, nothing about how some workers will have to commute two hours to their new jobs, nothing about how families will be split apart when the postal worker in the family gets transferred to a plant in another town, nothing at all about the 28,000 workers who won’t even have a job to go to.

The advisory opinion doesn’t discuss the social costs of the Network Rationalization plan.  It examines the Postal Service’s plan in the same terms with which the plan was presented — as a matter of sorting machines and square footage, of maximizing efficiencies and improving productivity.  It’s all about numbers, not people.  If money can be saved by consolidating plants, the issue is how many and which ones, not the value of providing jobs, sustaining families, and maintaining communities.

Perhaps eliminating 28,000 postal jobs is a necessary evil if we are to save the Postal Service — the lynchpin in a mailing industry that employs millions — but it seems odd to cut jobs in order to save them.  Yet that’s what network “rationalization” is all about.

 

The ever-shifting savings estimates

Much of the advisory opinion is devoted to analyzing the Postal Service’s cost-savings analysis and explaining why the savings are probably overstated by a significant amount.

When the Postal Service initially presented the plan, a press release said it would save $3 billion.  The USPS website still has a page on Area Mail Processing that says, “The U.S. Postal Service has proposed sweeping changes designed to save the organization up to $3 billion a year.”

When it presented the Request for an advisory opinion in December, the Postal Service said the plan would save $2.6 billion in operational costs, minus $500 million in lost revenue, for a net savings of $2.1 billion.

In February, the Postal Service released the Area Mail Processing studies on the plants being studied for consolidation, and a few weeks later USPS VP David Williams testified to the Commission that since not all of the AMPs were approved, the Postal Service would submit revisions to its cost savings estimates.  The revised estimate was a $2.1 billion savings in operational costs, minus $500 million in lost revenue, for a net savings of $1.6 billion.

The USPS website, however, continues to say that the plan will save $2.1 billion.  That little piece of misinformation is made possible by the way the press release is worded.  It says that the plan, when fully implemented, will “generate approximately $2.1 billion in annual cost reductions.”  This phrasing conveniently omits reference to the $500 million in lost revenue that needs to be subtracted from the $2.1 billion.

The advisory opinion adds to the confusion about the savings numbers.  The first sentence of the Executive Summary states, “The Postal Regulatory Commission has analyzed the Postal Service’s Mail Processing Network Rationalization (MPNR) initiative, a plan to provide forecasted net savings of $2.1 billion by consolidating its processing and transportation networks to better match estimated mail volume.”

That statement makes it sound as if the Postal Service is still forecasting a net savings of $2.1 billion.  Why use an earlier estimate in this context without clarifying that this number was subsequently revised?  It just perpetuates the USPS website’s misleading claim that the plan will save $2.1 billion.

The plan won’t save $2.1 billion, and it probably won’t save anything like $1.6 billion.  The advisory opinion is clear about that: “The Commission’s range of potential net savings estimates is lower than that projected by the Postal Service.”

That’s putting it mildly.  The Commission’s estimates of gross savings range from $1.97 billion all the way down to $45 million.  A savings of $45 million is not simply “lower” than the Postal Service’s estimate of $2.1 billion.  It’s not even in the same ballpark.  It’s not even on the same planet.

As for lost revenue, the Commission says the market research was so problematic there was nothing to work with, and it wouldn’t even venture a guess about how much revenue the plan might drive away.  The Commission obviously viewed the Postal Service’s estimate of $500 million with considerable skepticism.

If you consider that the plan may save as little as $45 million in operating costs and lose significantly more than $500 million in revenue, Network Rationalization starts to look like it has the potential to lose a significant amount of money, maybe more than a billion dollars a year.  The Commission refrains from discussing this possibility, but there is substantial evidence in the docket and the advisory opinion to justify such speculation.

 

Overstating the cost savings

During the advisory opinion process, critics of the plan said the savings in operational costs were considerably overstated for a variety of reasons.

For example, the estimates should not have included savings that aren’t directly associated with the consolidations.  Using more non-career workers who don’t earn as much as career employees would be possible without doing the consolidations, but those savings have been embedded in the Postal Service’s analysis.  The plan also calls for outsourcing to more non-postal contract workers, but converting Postal Vehicle Service (PVS) to Highway Contract Routes (HCR) can be done without the Network Rationalization plan and shouldn’t be included in the savings.

Another issue that came up repeatedly was the fact that the AMP studies for the 229 plants approved for consolidation added up to about one billion, not $2.1 billion.  The Postal Service never explained this huge discrepancy in a satisfactory way, and it’s very likely that the AMP-study estimate is more reliable than the one projected by the Postal Service’s modeling.

In the advisory opinion, the Commission was very concerned about yet another problem in the Postal Service’s cost-savings analysis — the issue of productivity.  The Postal Service says that expanding the operating windows and shifting the workload to fewer plants will lead to significant increases in productivity.

The Commission determined that to obtain the savings projected by the Postal Service, productivity would need to improve by over 20 percent at all plants and by over 25 percent at the receiving plants.  “These productivity improvements,” observed the Commission, “are highly optimistic and may not be achievable.”

According to the witnesses sponsored by the PRC (Matz and Weed), the “most likely result” of the consolidations will be a 5 percent improvement in productivity, and it’s possible that there won’t be any productivity gains at all, in which case processing costs could actually increase by $503 million.  “The Commission views this as a worst-case scenario,” says the advisory opinion, but it’s not outside the range of possibilities.

 

The two phases of market research

Things don’t get much rosier when one looks at the question of revenue losses.   There were two sets of market research, and they showed widely divergent estimates of lost revenue.  (Previous posts about the survey can be found here, and the numbers are here.)

The phase-1 research, conducted in August 2011, showed huge annual losses — a total volume loss of 7.7%, a revenue loss of $5.3 billion, and a net contribution loss of $2 billion.  Those kinds of losses are nearly twice what the Postal Service is currently experiencing, and they would have wiped out almost all the operational savings from the consolidations.  The Postal Service had the market research firm re-do the survey.

The phase-2 research, conducted in October 2011, produced more palatable results.  It showed a total volume loss of 1.7%, a revenue loss of $1.3 billion, and a net contribution loss of $500 million.

The Postal Service says it tossed the phase-1 research because it was flawed and irrelevant to the case at hand.  The concept statement read to customers as background before the questioning began mentioned closing post offices and eliminating Saturday delivery, so respondents were supposedly thinking about those factors when they answered questions about the change in service standards.  Since responses reflected the cumulative effect of several causal factors, the Postal Service says the survey was redone to focus solely on the change in service standards.

The advisory opinion describes the two surveys and relates the Postal Service’s explanation for why it rejected the first survey.  The Commission spares the Postal Service any criticism about hiding the first round and dancing around mentioning its existence during interrogatories — until the questions became so direct it was impossible to continue with the charade.

The advisory opinion also does not do a close reading of the two survey instruments, which would have shown how the second survey was crafted to elicit responses more amenable to the Postal Service’s plans.  The revised concept statement, for example, didn’t just avoid mentioning closing post offices and eliminating Saturday delivery; it actually made it sound as if the proposed changes in the processing network would benefit some mailers.

The advisory opinion also refrains from suggesting that the phase-1 results may be a more useful barometer of potential losses than the phase-2 results.  Instead, the Commission accepts the Postal Service’s characterization of the phase-1 research as “All Causes” research because the concept statement mentions the post offices and Saturday delivery.  But in fact, the survey questions are obviously about service standards, and the phase-1 survey actually provides a very good picture of how much money Network Rationalization may lose.

 

Flaws in the phase-2 research

Rather than focusing on the phase-1 research, the Commission addresses the problems in the phase-2 research, of which there were several:

1. The “likelihood of change” factor: The market research firm asked customers how much mail they would send (more or less) after the changes in service standards were implemented, but they were also asked how likely they were to make these changes.  If someone said he’d send 20% less mail, and he was 50% likely to do so, the two numbers were multiplied, so the result was the person would send 10% less mail.  Many of the participants in the advisory opinion process said using this probability factor was inappropriate in this context.

This issue has some history, since the Postal Service and the market research firm used the same probability factor in the five-day delivery case.  In its advisory opinion on that case, the PRC determined that using the factor in such a context was inappropriate, but the Postal Service stuck by its guns, insisted the factor was legitimate, and used it again for the Network Rationalization case.  This time around, the Postal Service had the market research specialist produce a lot of academic research attesting to its viewpoint.  The Commission did not seem to buy these new arguments, however, and it continued to question the validity of this factor.

2. The “solely attributable” factor: Customers in the market research were also asked, What percentage of the change in your mailing is “solely because of” the proposed service standard changes?  The point behind that question was to make sure other factors — like closing post offices and eliminating Saturday delivery — were filtered out of a customer’s responses.  As with the “likelihood of change factor,” the “solely attributable” factor was used to reduce a customer’s estimate of volume change.  If someone said he was going to send 20% less mail, and 80% of the decision was solely attributable to the decline in service standards, the two numbers were multiplied, leading to a 16% drop in revenue.

The Commission concluded that there was insufficient justification for using both the “likelihood of change” and “solely attributable” factors to adjust customers’ estimates of volume losses: “No support has been provided for the use of two, cumulative likelihood adjustments.”

3. Confidence intervals: There was yet another issue with the market research.  The market research firm did not produce what’s called a “confidence interval” for its bottom-line estimates. That’s a number that shows the range within which the estimated drop in mail volume would be expected to fall were the market survey replicated.  In the absence of confidence intervals, the Commission said it had “no meaningful way of comparing differences between Commission estimates and Postal Service estimates.”

4. Replicability: The Commission also found it impossible to replicate the data analysis conducted by the research firm.  It asked several times for assistance and information, but there were so many problems with the data set, the Commission staff could not reproduce the results.

In the end, the advisory opinion’s conclusion about the market research is fairly blunt. “The survey,” says the Commission, “is unable to provide information regarding whether or not there will be an effect on mail volume if MPNR is implemented.”  The results of the research are simply “not reliable.”

In other words, the market research tells us next to nothing about how much business the changes in service standard might drive away.  This research, by the way, cost the Postal Service a million dollars.

 

Estimating lost revenue

The Commission was unable to offer its own estimate of lost revenue, but there is plenty of information in the docket on which to base such an estimate.

In the Saturday delivery opinion, the Postal Service, using the same market research firm, estimated that gross revenues would decline by $455 million and net revenues by $200 million.  The Commission challenged the use of the “likelihood of change” factor, eliminated the downward reduction caused by the factor, and came up with a net loss of $587 million — nearly three times the estimate used by the Postal Service in its calculations of total savings.

If the Postal Service’s estimate of $500 million for Network Rationalization represents a downward reduction similar to what happened in the Saturday case, the estimate of lost revenue would need to be revised to $1.5 billion.

That number may seem implausibly high, but it is actually consistent with what was revealed in the phase-1 market research, which showed a net loss of $2 billion.  The Postal Service dismissed the results because the survey mentioned closing post offices and eliminating Saturday delivery.  But it’s possible to “correct” the phase-1 research by deducting the losses associated with these other causes.

As noted above, the market research on the five-day case showed a net loss of $200 million.  The Postal Service claimed that closing 3,700 post offices under the RAOI would cause no revenue decline at all, but let’s assume that it would have caused a net loss of $50 million (the total savings was going to be $200 million).  That adds up to $250 million. These are the only factors mentioned in the phase-1 (“All Causes”) survey that could result in revenue losses.

It’s possible that customers were thinking about the cumulative effects of all three factors, so let’s say another $250 million can be attributed to a “synergy” effect.  That leaves us with a $1.5 billion loss that can be attributed to the changes in service standards.

It is not unreasonable to expect, then, that the revenue losses will be on the order of $1.5 billion, rather than the Postal Service’s extremely optimistic estimate of $500 million. The advisory opinion, however, doesn’t go there, and just suggests that the market research isn’t very useful.

 

The bottom lines

Here’s a table that brings together some of the cost-saving and revenue-loss scenarios discussed above.

Cost Savings Scenarios
($ millions)	1	2	3	4	5	6
Mail Processing Savings	1,466	1,417	569	569	(503)	(503)
Maintenance – Facility savings	910	585	585	585	585	585
Transportation Savings	270	58	(36)	(36)	(36)	(36)
Total Cost Savings	2,648	2,061	1,118	1,118	45	45
Net revenue loss	500	500	500	1,000	500	1,500
Total savings	2,148	1,561	618	118	(455)	(1,455)

Here’s an explanation of what these scenarios are based on:

1. This is the Postal Service’s estimate for the Network Rationalization plan as initially presented to the PRC in the Request for an advisory opinion, back in December, when the plan included 260 plants.  All the other scenarios, with the exception of #6, are based on the revised plan with about 229 plants.  Scenario #1 is not on the table anymore, and it’s included here just for illustrative purposes.

2. This scenario is the Postal Service’s current estimate of savings in operational costs and revenue losses for the plan as it is currently described.  It uses the revised savings numbers presented to the PRC after the AMP studies were completed in February.  This plan involves two phases, with about 140 plants being consolidated between this past August and January 2014, and the remainder later in 2014.

3. This scenario uses $569 million for mail processing workload reductions savings — the “most likely” scenario developed by the Commission-sponsored experts.  Since the Commission did not venture an estimate for lost revenue, this scenario uses the Postal Service’s estimate.

4. Like #3, this scenario uses the “most likely” scenario for savings in operational costs, but it uses a higher estimate for lost revenue that considers the phase-1 survey results and the flaws in the phase-2 survey.

5. This scenario uses the “worst-case” scenario developed by the Commission-sponsored expert testimony; it’s based on the possibility that productivity will not improve at the gaining plants.  This scenario uses the Postal Service’s estimate of $500 million in lost revenue.

6. This scenario uses the “worst case” for cost savings (as in scenario #5) and a “worst case” for lost revenue (as described above).

The bottom line shows that there is a remarkably wide range of possibilities.  The plan could save $1.6 billion, or it could lose almost that much.  The most likely scenarios show a savings between $100 million and $600 million, a far cry from the $2.1 billion being touted by the Postal Service.

 

A missed opportunity

Several witnesses testified that if the Postal Service wanted to optimize its processing network, there were better tools available for figuring out how many plants to consolidate and which ones would be best to study.  Some scenarios could realize significant savings without requiring changes in service standards and the lost revenue such changes will cause.

The Commission puts it this way: “The analysis that the Postal Service has submitted in this proceeding missed an opportunity to leverage fully the potential of network modeling because it assumed, before modeling work began, that the response to declining volume with the least financial risk is to slow delivery by a day.”

In other words, the Postal Service started by assuming that slowing down delivery times (so it could expand the processing window and use machines more hours per day) was the way to go, and only then went to work on the modeling.  Models that didn’t involve changing service standards weren’t considered.  Once the Postal Service had determined on a course of action, nothing would deter it — not even the phase-1 market research results, which showed huge losses from slowing down the mail.

As the advisory opinion process revealed, there were other ways to save money without changing service standards.  “There is a convergence of opinion on this record,” states the Commission, “that the vast majority of mail processing savings that the Postal Service expects to attain can be captured without significantly changing service.  These savings could be captured without incurring significant risk of lost revenue from reduced service.  Consequently, the Commission concludes that consolidating as many plants as possible without significantly reducing service may involve less financial risk to the Postal Service than proceeding to full implementation of the network configuration and service level changes.”

 

Looking for the sweet spot

The advisory opinion explains how there’s a “tradeoff between cost reduction and service changes.”  The greater the cost-cutting, the greater the risk of lost revenues.  The goal is to find what one witness called the “sweet spot” — the point where the savings from reducing the size of the network starts to taper off because slower service decreases revenues.

The docket on Network Rationalization includes three alternate scenarios developed by outside experts, all of which do not require a change in service standards.  Two are based, as is the Postal Service’s own modeling, on the view that larger plants have lower processing costs than smaller plants.  They show that it would be possible to close 191 or 203 plants and save $1.6 or $1.7 billion (respectively) without a substantial changes in service.

The PRC also examined a scenario that was based on actual productivity at large and small plants.  The Commission-sponsored experts, Weed and Matz, did not begin with the assumption that bigger plants will have lower unit processing costs in a reconfigured network.  Rather, they looked at historical patterns and found that larger plants tend to be less efficient than smaller plants.  They observed that with the shift to larger plants, productivities could decrease or be much less than the Postal Service expects.

Weed and Matz concluded that it would be better to optimize the network by closing down the least efficient plants.  In that way, closing just 107 plants could yield a savings of $1.8 billion — without a substantial change in service and therefore with only “minimal” lost revenue.  The advisory opinion contains this table showing the four scenarios:

Savings/Service Trade-Off Using Alternative Cost Driver Assumptions
	Cost Driver Assumptions	Reduction in Mail Receiving Overnight Service	Reduction in Plants	Estimated Savings

($ billions)

 

With the exception of the Postal Service’s plan, all of the models, notes the Commission, “lead to the conclusion that most of the savings that the Postal Service expects to achieve by reducing its capital stock could be achieved without a substantial change in service.”

 

Small is beautiful

One of the Postal Service’s basic assumptions, driven by a conventional notion of “economies of scale,” was that larger plants are more efficient and productive.  The belief that bigger is better seemed to guide the thinking, but it’s not necessarily true.

The Commission examined this assumption back in 2006, when it did an advisory opinion for the Postal Service’s first network optimization initiative called Evolutionary Network Development (END).  The Commission observed that the belief that bigger plants are more efficient is generally true of capital-intensive industries, such as telecommunications and power generation, but it’s not so true of labor-intensive industries like the Postal Service, where only a small portion of costs are capital expenses.

When it comes to mail processing, it’s not economies of scale that are important but economies of density (aka economies of fill), and smaller mail processing plants are as likely to exhibit economies of density as larger plants.  In fact, smaller plants are generally more productive.

The advisory opinion contains the following table illustrating this point:

Comparison of Modeled and Actual Average Cost per Square Foot
Plant Size	Postal Service Modeled Cost Per Square Foot	Average Actual Cost Per Square Foot of Mail Processed on Automation	Pieces Processed per Square Foot	Average Cost per Piece Calculated from Modeled Cost	Actual Average Cost per Piece
450,000 sq. ft. +	$130.25	$200.37	5,372	$0.0242	$0.0373
210,000-450,000 sq. ft.	$198.98	$239.51	7,261	$0.0274	$0.0330
0-210,000 sq. ft.	$238.13	$235.22	8,267	$0.0288	$0.0285

The table shows that while larger plants cost less per square foot, when we look at the actual cost per piece of mail, the smaller plants are less expensive.  When it comes to mail processing, as economist E. F. Schumacher always said, small is beautiful.

 

Playing fast and loose with numbers

On September 18, the Postmaster General gave one of his regular video pep talks to postal employees.  He noted that the consolidation plan was well under way, and he said that phase one save $1.2 billion annually.  That same number is cited in the press release on the USPS website.  This is not an accurate estimate, however, even according to the Postal Service’s own analysis.

The $1.2 billion figure comes from the Postal Service’s May 17, 2012, announcement that it would be moving forward with a modified implementation plan that involved two phases.  This estimate was scrutinized during the advisory opinion process, and it turns out the number was very problematic.

The Postal Service was asked, what portion of the $1.6 billion in net savings would be realized by phase one of the plan, and what portion would be realized by phase two?  The Postal Service couldn’t answer that question because it had not done market research that would allow it to separate the revenue losses into two phases.

But the Postal Service did offer an estimate of how the savings in operational costs would break down.  For some reason, however, instead of breaking out the $2.1 billion in operational cost savings — the savings associated with the revised plan — the Postal Service went back to the original plan, when there were about 260 plants and the cost savings were estimated at $2.6 billion.

In other words, the Postal Service split up the original estimate of $2.6 billion and said phase one will save $1.2 billion and phase two, the remaining $1.4 billion.  But the Postal Service should have split up the revised estimate of $2.1 billion, which would have shown that phase one will save about $0.97 billion and phase two, $1.13 billion.

And then there’s the matter of lost revenue.  If we use the Postal Service’s estimate of $500 million and say phase one causes half of the lost revenue (it could easily cause more), $250 million in lost revenue needs to be subtracted from the gross savings of $0.97 billion.  That brings the net savings for phase one to around $720  million — significantly less than the $1.2 billion the Postmaster General referred to in his talk the other day.

 

Looking ahead

Of course, all of these estimates of what may happen are largely academic exercises.  We’re never going to know how much the consolidations save or lose.

The Postal Service will fudge the numbers when it comes to determining how much the consolidations are saving in operational costs, just as it does when it prepares Post-Implementation Reviews (PIRs) after a consolidation takes place.

Plus, it will be impossible to determine how much of the future revenue declines can be attributed to the change in service standards.  Even now, it’s very likely that revenues are falling because of all the talk about declining service.  But don’t expect to hear the Postal Service acknowledge that its own doomsday rhetoric and plans to degrade service by cutting hours at 13,000 post offices and slowing down First-Class mail are contributing to declining revenues.  It’s all about the Internet.

The Commission does not take such a cynical view of the matter, however, and instead discusses how the Postal Service can use its phased-implementation approach to learn lessons about how much money the consolidations are saving.  Rather than the hypotheticals it’s been dealing with in its modeling, the Postal Service will have actual hard data available.

The Commission notes that while it’s extremely difficult to predict how Network Rationalization will affect operating costs and customer behaviors, “phasing in the implementation will create an opportunity to evaluate the assumptions that underlie the Postal Service’s forecasts of these effects.  The analysis of data about the actual impacts of the first phase of MPNR will allow the Postal Service to make informed decisions about whether to proceed with Phase 2, and if so, whether modifications are appropriate.”

In particular, if the Postal Service sees that its expectations about productivity improvements were too optimistic, it can “consider alternative strategies, such as shifting volume from low to high productivity plants.”

The phased approach will also allow the Postal Service to see what happens to mail volumes.  In phase one of the plan, there will be an approximately 20 percent reduction in volume currently receiving overnight delivery.  That “could provide the Postal Service with the kind of historical data needed to undertake an econometric analysis of the relationship between speed of delivery and mailing behavior.”

“With the benefit of better information about the cost, service, and volume impacts of consolidations and the expansion of operating windows,” continues the advisory opinion, “the Postal Service will be positioned to better assess the cost-benefit calculus of the full elimination of overnight service and whether more modest changes may deliver sufficient or perhaps greater savings with a smaller impact on mailers.”

It’s nice to think that the Postal Service will reassess things down the road, but it’s not likely.  The Postal Service has made a commitment to changing service standards and closing half its plants, and if the phase-1 market research — with its forecast of volume declines over 7% — couldn’t persuade management to rethink its plans, there’s no reason to believe that some future cost-benefit calculus will change any minds in postal headquarters.

 

How long is too long?

At the end of the advisory opinion, four of the Commissioners contributed separate statements.  They all refer to how long it took to do the advisory opinion.  Commissioners Taub and Hammond produced a joint statement that goes on for several pages about the matter.

Chairman Goldway also acknowledges the time issue: “I am concerned that the complex issues addressed in this docket, including the several iterations of the Postal Service’s proposal, took almost 10 months to resolve.  The Commission is committed to adopting new rules that should facilitate advisory opinion procedures while protecting due process.”

That’s a reference to docket RM2012-4, which is about developing new rules for advisory opinions.  The docket features a letter from Senator Tom Carper encouraging the Commission to move more quickly or risk having its decisions become “irrelevant.”

While it’s likely that some aspects of the advisory-opinion process could be accelerated, the Network Rationalization case was unusually complicated, and it should come as no surprise it so long.

Just consider the volume of material in the docket: The Postal Service’s Request for an advisory opinion included 13 pieces of testimony, 33 public library references, and 6 non-public library references.  There were some 40 witnesses in the case — 22 produced by the Postal Service, plus 18 others called by the Commission and the interveners.   The Postal Service was presented with over 1,000 interrogatory requests, there were over 170 library references, and there were ten days of hearings, which generated 4,700 pages of transcript.  (The docket, N2012-1, is here.)

The advisory opinion took ten months, which is somewhat longer than opinions in the past, but not totally out of line.  Since 2006, the average length of time for an opinion is over seven months.  The 2006 opinion on Evolutionary Network Development (END) took nine months.  The 2010 opinion on the Stations and Branches Optimization & Consolidation initiative (SBOC) took eight.  The 2011 opinion on eliminating Saturday delivery took a year.  The 2011 opinion on the RAOI took five months.  The POStPlan opinion, just completed, was the fastest — just about three months — but it was a relatively simple case, largely unopposed, and an extension, in some respects, of the two previous cases on closing post offices. (A spreadsheet showing how these and previous opinions went down is here.)

As long as the opinion on Network Rationalization took, the process would have been even longer if the Commission had not squeezed in extra steps to deal with several “intervening events” caused by the Postal Service, like the completion of the Area Mail Processing (AMP) studies and the introduction of a modified implementation plan.  If the Postal Service hadn’t kept changing the plan — if it had simply waited, for example, until the AMP studies were completed before submitting the Request — the process might have gone more smoothly and been completed more quickly.

 

What’s the point, anyway?

The advisory opinion contains a wealth of useful information that may very well have a salutary effect on the Network Rationalization plan.  It’s possible that the PRC’s process contributed to the Postal Service’s decision to implement the plan in two phases, and the advisory opinion may discourage the Postal Service from proceeding with phase two, though that will probably depend on other events, like postal reform legislation and the state of the Postal Service’s finances in 2014.

As it is, the Postal Service does not seem very inclined to listen to its regulator.  The Postal Service determined that it did not need to wait for the advisory opinion, and it implemented the changes in service standards on July 2, 2012, almost three months before the opinion came out.  The APWU tried to get the Postal Service to wait, and it filed a Complaint and a Motion for an Emergency Order asking the Commission to stop the Postal Service from implementing the changes before the opinion was done.  Both requests were turned down by the Commission.

The problem is not simply that the Postal Service didn’t want to wait for the opinion before proceeding with implementation.  The Postal Service doesn’t seem to think that the Commission has any business questioning the judgment of postal management to begin with.  The Postal Service expressed its attitude about the advisory opinion well before it came out.  In its initial brief to the Commission, the Postal Service sums up its views very succinctly:

“In the face of trends largely beyond its control, Postal Service management must implement measures that preserve the long-term viability and relevance of the national postal system.  Simply put, this is no time for strategies founded on half measures and hope.  Rather, it is a time for expeditious action that strikes a reasonable balance between historical service levels and the need to eliminate the excess mail processing and transportation capacity that is fueling the imbalance between revenues and operating costs.”

Management thus abrogates all responsibility for the plight of the Postal Service and accuses others of proposing strategies based on “hope,” even though its own plan is based on just that — wishful thinking about improved levels of productivity and overly optimistic estimates about how much business the plan will drive away.

The Postal Service’s brief reminds the Commission that an advisory opinion is “non-binding,” and that section 3661 of Title 39 (which establishes advisory opinions) should not be seen as “an invitation for the Commission to second-guess the judgment of postal management.”

Moreover, says the brief, the Commission shouldn’t be spending months “reducing the cost or savings implications to the PRC’s own best estimate.”  According to the Postal Service, it’s up to management and the Board of Governors to look at the numbers and make decisions.

Only in a footnote does the Postal Service acknowledge that if there were fundamental flaws in its financial analysis, it might be appropriate for the Commission to point them out, but in this case, the Postal Service maintains, there are no flaws of such magnitude.

A close reading of the advisory opinion leaves one with quite a different impression.  The opinion shows that the Postal Service’s case for Network Rationalization is indeed filled with fundamental flaws, and it offers a lot of reasonable advice about how significant savings can be realized without risking huge losses in revenue.  If the executives in L’Enfant Plaza weren’t so sure of themselves, they might be willing to listen.