Thursday, October 29, 2009

The Reinterpretation of William Burrus

Many postal workers have jumped to the defense of APWU president William Burrus as a result of my article, Mathematically Challenged: Burrus Proposal Doesn’t Add Up for USPS. But if many of these defenders are correct, they should be angry at Burrus for garbling the message and distracting people from the real issue.

Of the scores of comments submitted to this site and several postal-news sites, many stated that what the union president meant to say is that the U.S. Postal Service can sort un-presorted letters for 1 to 3 cents apiece (depending upon the commenter), while mailers get First Class presort discounts of up to 10.5 cents each. But that’s not even close to what Burrus actually said.

It’s also contrary to the Postal Service’s own data earlier this year estimating that the presorting of First Class letters saves it far more than 3 cents. But some of Burrus' defenders have raised legitimate questions regarding whether those cost estimates match current USPS practices.

One of the best what-he-really-meant-to-say defenses of Burrus came from “uncommonsense”, a commenter who acknowledged that Burrus’ challenge to Postmaster General Jack Potter is “a publicity stunt” with “no serious evaluation of costs/benefits”. (Rather damning praise, I’d say.) Here is uncommonsense’s valuable history lesson:

"In 1999 the OCR machines that the USPS was running had the ability to read about 20% of the letters ran through them. They had a throughput capacity of 30,000 letters per hour and required 2 trips through the machine to sort to the first breakdown and only allowed 94 different breakdowns. Most of the mail that ran through this machine had to have images of it sent to human keyers on terminals at one of 55 REC sites to provide the information to barcode the mail piece.

"Today, because of faster computers and more advanced software, the equipment that the USPS is running is able to read and apply barcodes to about 97% of the letters ran through it and sort the mail pieces to over 200 possible breakdowns, ALL on the first pass @ 40,000 letters per hour. An equivalent breakdown in 1999 would have required 3 passes through the OCR @ 30,000 pieces per hour and many human keyers. So in one hour with 2 operators and a keyer, the USPS can now process what used to take 4 hours 2-3 operators and many keyers. The new equipment is also much less expensive to maintain then the old MLOCR was.

"Now, 3% of letter images are sent to human keyers at one of 2 REC sites. Despite adding Flats and Parcel images to keyer duties, technology has allowed the USPS to eliminate 53 out of 55 REC sites.
Since USPS costs for bar coding and sorting letters has decreased so much since 1999 why have the work share discounts not also decreased?"

An anonymous commenter chimed in with this spot-on observation: “Presort mailers prepare the mail to the exacting specifications of the USPS Domestic Mail Manual in order to claim any worksharing discounts. If there is an issue with the mail handling once received - we are doing what the Postal Service told us to! If it needs to change, it's a Postal Service task.”

Another commenter added, “It's the silly Post Office rules that waste money. I have seen carrier-routed mail returned to the SCF in order to co-mingle it with other mail. Why can't the carrier just case it in? Think of the man-hours, and transportation costs that this takes. All of this is because management has the silly notion of having all the carrier's mail ready to deliver when they arrive in the morning.” (In defense of the Postal Service, using otherwise underutilized mail-sorting operations to do work that takes a burden off the delivery operations might actually make economic sense.)

The picture that emerges from these comments is that, while the size of presort discounts on letters might once have been justified, they are no longer consistent with the Postal Service’s costs. Put another way, the commenters are indicating that automation and excess mail-handling capacity have shrunk the cost difference between handling presorted and un-presorted letters.

I’m not saying they are correct. Some of their cost calculations are clearly wrong – for example, some included only salaries while ignoring benefits, facility costs, etc. But the arguments of “uncommonsense” and some of the other commenters have far more logic and plausibility than Burrus’ vague, ill-conceived proposal.

Burrus muddied the waters by proposing an obvious money loser for the Postal Service – remove First Class presort discounts that average 8.9 cents per letter and pay APWU members 10.4 cents instead to do the sorting. He tried to make his sloppy math more favorable to the Postal Service this week by throwing in a bonus – free sorting of parcels – along with his usual big-mailers-are-vermin bluster.

But the proposal is still too vague to be taken seriously. Among many flaws with the Burrus plan is that it would decrease the demand for First Class mail by raising prices. So with APWU members getting their 10.4 cents on fewer letters but still having to sort parcels for free, would they end up having to take a pay cut?

In an attempt to move the discussion away from character assassination and conspiracy theories, Dead Tree Edition offers these observations and ideas:

• Rather than trying to keep excess employees busy by incenting mailers to mail in a less efficient manner, which is in essence what Burrus is proposing, downsizing the workforce via meaningful early-retirement incentives would be more productive. (See What the Postal Service Left Out of the Early-Retirement Deal.)  Presorting address data before a mailing occurs is inherently more efficient than sorting the actual mail pieces.
• Presorted letters are highly profitable for the Postal Service, as evidenced by its eagerness to offer the Summer Sale on Standard mail and the Fall Sale on First Class. USPS’ problem is not that it doesn’t charge enough for letters, it’s that it doesn’t have enough letters to charge for. Anything that would reduce the volume of letter mail, as would Burrus’ proposal, would be counterproductive for the Postal Service and its employees.

 • Under current law, any significant reduction in First Class presort discounts would require decreasing the price of un-presorted letters (that is, the 44-cent First Class stamp). Otherwise, the inflation-based price cap on First Class would be violated.

• If indeed presort discounts on First Class letters are larger than justified, there might be a way to shrink them without hurting mailers or driving business away from the Postal Service – dropship discounts. Business mailers get 4.3 cents for dropshipping Standard letters to Sectional Center Facilities but nothing for First Class, even though dropshipping clearly saves the Postal Service money. Introduction of First Class dropship discounts (which would have to clear some legal barriers) could compensate mailers for the shrinkage of presort discounts and cause them to mail in ways that are more efficient for the Postal Service.

• For the record, I am not "speaking on behalf of the Far Right" (Burrus' description this week of those who defend business mailers) and have not made anti-union statements (just criticisms of a specific union official's proposals). If I were part of the Far Right, I never would have written articles like the recent one ridiculing Rush Limbaugh or have published the ghost-written piece, How sleepy is the giant?.

• Maybe “uncommonsense” should replace Burrus as the APWU’s spokesman.


Unknown said...

Presorting mail at ANY discount makes no sense. I am an automation clerk for the USPS, and I can't tell you how many times we've had to comingle presorted mail with all other letters so the sum is walk sequenced, which cuts down on manual sortation costs by the carriers. At up to 40,000 pieces per hour, this is by far the cheapest way to get the mail out to the street. Any sortation beyond 5 digit is redundant.

Anonymous said...


So what exactly changed from your previous depiction of Burrus? You claim he is politically savvy and “mathematically challenged” and then you reinterpret him as dishonest and cunning.

Your previous attack on Burrus fixates the “10.4 cents” of his challenge.

If you look clearly you see that Burrus wrote “Discontinue the exorbitant postage that are offered to large mailers - which are currently as high as 10.5 cents per letter - and allow members of the APWU to perform all mail-processing functions at the rate of 10.4 cents for every letter and flat.” (Notice the “as high as 10.5” remark.) Those who attacked Burrus (by various means) missed the point on purpose in order to create confusion about what he IS saying. Today the Postal Service can process the outsourced mail cheaper and more efficiently. Your whole argument rests on ‘calculating’ that the Postal Service will lose revenue because it will discourage large mailers. I understand that, but I think the point is to bring back work that is a primary function of its own internal processing system without loss of revenue.

A brief history (0r let me get this straight):
The Postal Service rolled out workshare discounts at a time when it couldn’t handle the complete preparation of the mail. A sub-industry emerges to complement this need, but somehow it evolves from ONE to over 450 workshare discounts and more and more mail being outsourced. The last OIG audit on this matter summarizes workshare discounts as “rate discounts that the Postal Service provides to mailers to mailers for presorting, pre-barcoding, handling, or transporting mail - duties the agency would otherwise assume. The Postal Service first established workshare discounts in 1976 with a 1-cent discount for presorted First-Class Mail®. Today there are over 450 workshare discount rates that encompass all classes of mail (except Priority and Express Mail) and apply to over 80 percent of mail the Postal Service processes.”

What’s more interesting is the fact that workshared mail volume ratio increased every single year (the study looks at volumes between 1984 and 2006 and only for BPM mail). By 1984 the workshared BPM mail already surpassed the volume the Postal Service was processing in house by over 17 million pieces. As mail volume increased (Yes, even during the Internet age) so did the gap between workshared (outsourced) and non-workshared (in house) mail. In 1984 the ratio for workshared mail already surpassed over 50% of total mail volume, and by 2006 it rose to over 80%.

Now let’s add apples and oranges:
Yes, the OIG audit is a segmented and limited study to be applied mathematically to the whole impact that workshared mail has on the Postal Service as a whole. Postal management has been involved in the twisting business for quit some time. “10.4” is the beginning of holding management accountable for what makes sense. Potter claims overcapacity in mail operations when in effect he has been engaged in a distortion of the facts. The OIG audit also found that postal operations was using obsolete data in calculating the true impact of outsourced mail on internal operations. Sure they followed internal policies but “postal operations are continuously changing and the old conversion and productivity factors (arrival and dispatch profiles) and other inputs used to develop the existing BPM workshare discounts might not reflect the current operational environment”

Anonymous said...


Current operational environment? Let’s forward this to present day. Whatever Potter and his team means and how it trickles down are not matching. Burrus has been around too long for this kind of nonsense that only leads to attacking the Collective Bargaining Agreement and bypassing the responsibilities that management has towards its employees. Let’s see if the facts (data) show how tricky Potter’s bunch are.

On August 12, 2009 Maura Robinson (USPS VP Pricing) presented the following slide as part of her presentation to members of the Mailers’ Technical Advisory Committee entitled Move Update Assessment Pricing.

The presentation shows the Postal Service changing the structure of pricing and how the discount is calculated for those who perform postal duties within the private sector. You see, the only thing that moved was that of the penalty for misorted mail was decreased and made universal. Now private sorters pay less fines for bad workshare. But get this: it is all done in order to improve the overall quality of presorted mail. It doesn’t take a genius to see the misconception being presented: management is luring the private sector with a larger piece of the pie! More sort-mills will jump at the 10.5 cent discount offered for First Class Mail Pre-sorting.

The fact is that the current Postal Service can handle any duties previously “sold” as workshare discounts. Potter says the Postal Service must cut more and more workhours by any means necessary to support postal outsourcing. Burrus is offering to do the same work by its current workforce in-house. If management has gained an advantage in pricing (rates and discounts) they are sure slow to follow that principle when dealing in labor relations. So when the VP of postal pricing offers more profit to private mailers based on Volume but refuse to negotiate fairly with their representative of labor unions … who’s really dishonest and cunning? Burrus is on the right track to bring attention to worksharing discounts and the truth behind it. The Postal Service will survive this economic crisis and still provide “decent” wages for working people. That is something that the workshare industry has not been able to match for the last 33 years.

Reinterpreting Burrus? Yes, please do.


Anonymous said...

doninwny you are absolutely correct. Why presort the mail just to have it resorted with outgoing collection mail?

Think of it like pre-paying a stripper extra money for being in her skin, then putting SOME of her clothes on her, so she can turn around and take them off again.

Anonymous said...

Another, perhaps more accurate, way to consider presorting is that, boiled down to a nutshell, the USPS is losing revenue by granting presort discounts.

There is money to be made in presorting. Ask Pitney Bowes, Alan Richey, and RR Donnelley. All are making record profits from presorting. They are the middle man between the USPS and customer. Any business major will tell you there is benefit in eliminating the middle man.

The USPS should end all pre-sort discounts and offer volume discounts (or customer loyalty discounts) to company who bring their mail directly to the USPS.

In this manner, the USPS generate extra revenue and the mailer still gets a discount. This plan will not cause a drop in volume because the mailer will get the same discount he gets now minus whatever they paid to the presort houses


Anonymous said...

Once again, I caution you not to accept USPS stated figures as accurate. The GAO report I referred to in the previous article clearly indicated that the USPS has no means of measuring savings from outsourcing.

That same GAO report postulates that perhaps it is time for the USPS to bring outsourced work back in-house. During the 80's and 90's mail volume exceeded USPS capacity and outsourcing made sense. Now, the GAO says, with volumes at a more managable level the USPS should recapture that work


uncommonsense said...

-Presorting address data before a mailing occurs is inherently more efficient up to a point. The most efficient way for mail to be presented for delivery is in delivery order. There is more than one mailer providing mail for every route/zip code so no mailer is capable of sorting all mail for all addresses on a route into delivery order. That is why, as the first reply to this story pointed out, any sortation beyond 5 digits provides zero cost benefits to the USPS. It is also why the discounts for high density and saturation standard mail dropped at the delivery unit are insane. Much of this mail gets trucked back to the Sectional Center Facility for processing because it is more efficient than manually processing at the delivery unit.
-Presort First Class letters are highly profitable for the Postal Service. Presort Standard…..not so much. Standard letters are typically slightly more expensive for the USPS to process than First Class. Much of it has characteristics like glossy paper, greater lengths and heights, and unsealed edges that make it less automation friendly. It jams more and runs slightly slower. Yet the USPS receives over 10 cents less per presorted standard letter than it does for a comparable presorted first class letter. Thus profit margins on Presort Standard are slim to none. Profit margins on Carrier route Presort Standard are definitely negative.
-Reducing the First Class presort discount and reducing the cost of the First Class stamp would be beneficial to the USPS because it would encourage more usage of the product that is currently by far the most profitable for the USPS, the un-presorted first class letter.
-Introducing First Class drop ship discounts would be murderous to the Postal Service’s bottom line. Currently, without any incentive, about 40% of first class mail is already dropped within the SCF to which it will be delivered. This means for much of the first class mail the Postal Service would effectively give billions of $s of discounts for $0 in cost reduction.
-For the record I would make a lousy APWU spokesman. I go into too much detail and am no good at making the simple (though not quite accurate) statements that fit into headlines. A politician I am not.

Anonymous said...

Uncommonsense, then you shall replace the PMG - when he retires, of course.

How about lowering the price of First Class stamps in exchange for all the outsourcing?