Download File

BEFORE THE
POSTAL RATE COMMISSION
WASHINGTON, D.C. 20268B0001
POSTAL RATE AND FEE CHANGES
PURSUANT TO PUBLIC LAW 108-18
Postal Rate Commission
Submitted 7/15/2005 4:14 pm
Filing ID: 46083
Accepted 7/15/2005
Docket No. R2005B1
REPLY COMMENTS OF THE UNITED STATES POSTAL SERVICE
IN RESPONSE TO RULING NO. R2005-1/38
REGARDING DESIGNATION PROCEDURES
(July 15, 2005)
The Postal Service hereby offers its reply comments in response to Presiding
Officer’s Ruling No. R2005-1/38 (June 29, 2005), on the topic of appropriate procedures
for the designation into the record of written cross-examination. The Postal Service first
raised concerns about partial designations of interrogatory responses on June 27,
2005.1 In this pleading, the Postal Service replies to the initial comments of Valpak and
ABA/NAPM, both filed on July 8 (although the latter set of comments was dated July 7).
Both Valpak and ABA/NAPM take issue with a matter that the Postal Service did
not even consider to be on the table – the opportunity for a party to counter-designate
subparts of a discovery response of its own witness, if those subparts have been
excluded from the designations of other parties. Both sets of comments conclude that
such counter-designations should not be allowed. The Postal Service is in agreement
with them only to the extent that, since the Postal Service is of the view that partial
designations should not be allowed, no need should arise for such counterdesignations. If partial designations are allowed, however, the Postal Service would
1
Request of the United States Postal Service for Ruling to Clarify the Practice of
Designating Parts of Interrogatories (June 27, 2005).
very strongly disagree with any suggestion that counter-designations of omitted
subparts (by the same witness) not be permitted as a matter of right. Failure to permit
routine counter-designations under these circumstances would violate fundamental
principles of fairness and due process.
For starters, it may be useful to discuss the status quo. Valpak is apparently of
the view that partial designations are allowed, because on page 1 of its comments, it
refers to the consequences “if ‘partial-designations’ are continued to be allowed.” For
what it is worth, the collective recollection of postal counsel is to the contrary, that partial
designations have not been routinely allowed in the past. Perhaps both recollections
are correct with respect to different cases at different points in time. Regardless of
whether there has or has not been a consistent practice in the past, however, it does
seem that some clarification is warranted.
Valpak, however, is clearly in error on one point. In an effort to buttress its
position that counter-designations should not be allowed, Valpak argues that the
principal result of allowing counter-designations would be that what are currently filed as
subparts would instead be filed as separate interrogatories. Valpak opines that “[s]urely
this is not a result that should be desired.” Valpak Comments at 2. In fact, however, if
partial designations are going to be allowed, as suggested by the Postal Service’s initial
comments, this would be tantamount to changing the unit of response from interrogatory
to interrogatory subpart. If subparts were to be the unit of response, and could be
individually designated or individually omitted, then the current requirement of Rule
26(b), that the responses to each interrogatory start on a separate page, would make no
-2-
sense. The rule should then be changed to require response to each subpart start on a
separate page. In effect, this would create exactly the situation which Valpak claims to
be undesirable – each subpart might as well be submitted as a separate question, as
the responses would not appear on the same page anyway. In truth, therefore, the
circumstances which Valpak would view as “surely” undesirable would be a
manifestation not of the practice of allowing counter-designations in response to partial
designations, but a manifestation of the problems inherent in allowing partial
designations in the first place.
Valpak indicates on page 1 that it would oppose a “general rule” allowing a party
to counter-designate undesignated responses of its own witnesses. It is distinctly
unclear whether Valpak’s comments in this respect are directed at a rule regarding
undesignated and omitted subparts (which the Postal Service believes to be the subject
at hand), or whether Valpak is referring to a broader rule regarding as well answers of
the same witness to entirely separate interrogatories. If the latter is the source of
Valpak’s concerns (page 2) that the Postal Service could thereby insert “virtually
anything” into the record through the vehicle of counter designation, Valpak is knocking
down the proverbial strawman. To the best of its knowledge, neither the Postal Service
nor any other party is advocating at this time that parties be allowed to counterdesignate responses to anything other than omitted subparts. Nevertheless, as
discussed below, there may be certain instances with respect to other materials
covering the same subject matter in which such a result could be considered consistent
with the minimum standards of fundamental fairness and due process.
-3-
Valpak asserts (page 1) a belief that the problems which the Postal Service
claims will occur if partial designations are allowed are “overstated.” To illustrate why
Valpak is in error in this belief, it is necessary to go into the details of specific
circumstances in this case, a step that the Postal Service would have preferred to have
been able to avoid. The circumstance which triggered the Postal Service’s instant
request for clarification was designations by GCA of certain subparts of responses by
witness Thress, and omission of other subparts. Although the problems were not
limited to GCA/USPS-T7-3, that question will suffice for purposes of discussion. (In
order to make it more convenient to follow the discussion, the full response submitted
by witness Thress to that question is attached to this pleading, and also appears at Tr.
3/301-04.)
GCA 3 inquired about the matter of stationarity, which is a fairly academic topic,
even for econometricians. The question and response consisted of five subparts, a
through e. In response to subpart a, witness Thress had to explain the limits of a
general principle regarding stationarity postulated by GCA. In subpart e, witness Thress
explained exactly why there was no evidence of a problem with stationarity in his work.
In between, in subparts b and c, witness Thress confirmed that, if present, nonstationarity can create econometric issues. In subpart d, witness Thress presented the
numerical results of statistical tests regarding stationarity that he had conducted, but
presented no interpretation of those results. Instead, his interpretation was contained in
subpart e.
On June 27, however, GCA chose to designate only subparts b-d of GCA 3, and
-4-
to omit subparts a and e. In practical terms, therefore, GCA designated those portions
of the response confirming that, in theory, non-stationarity can create problems, but
omitted those portions of the response demonstrating why, in practice, non-stationarity
was not a legitimate basis to question the results provided by witness Thress. These
circumstances illustrate exactly the type of unsavory situations to which the allowance
of pick-and-choose partial designations can lead, and demonstrate forcefully that the
concerns expressed by the Postal Service are not overstated.
Examining the full response of witness Thress to GCA 3, it is beyond cavil that, if
GCA were to be allowed to omit portions of that response, even under the most
rudimentary notions of fairness and due process, the Postal Service would need to be
allowed to counter-designate the other portions. GCA opened the door on stationarity
by asking the questions it presented in GCA 3. Witness Thress did nothing more than
answer the questions he was asked. If GCA wanted to close the door on stationarity, it
could have simply omitted any designation of interrogatories on the topic (e.g., GCA 3).
But having left the door open by designating witness Thress’ responses to certain
questions, GCA would be in no position to complain if the Postal Service wanted the
record on stationarity to be complete and counter-designated the response to the
questions presented in the other subparts. (And to its credit, when the Postal Service
did counter-designate those subparts, GCA neither complained nor objected.)
On the other hand, if the Postal Service were prevented from counterdesignation, in accordance with the views stated by Valpak and ABA/NAPM, there
-5-
would be a clear violation of the due process rights of the Postal Service.2 A record in
which a witness is required to confirm that something might be a problem in theory, but
is prevented from explaining why it is not a problem in reality, does not meet the most
basic objectives of fairness.3 In fact, as a hypothetical, the Postal Service would be
prepared to argue that, even if subparts a through e of GCA 3 had instead been
submitted as five separate interrogatories, due process would militate in favor of
allowing counter-designation of any of the five responses that were omitted, if any were
designated.4 As long as the omitted questions were on the same subject matter as the
Valpak (page 2) erroneously claims that the Postal Service’s ability to file rebuttal
testimony is somehow relevant to this discussion. First of all, contrary to Valpak’s
assertion, the ability to submit rebuttal testimony is open to all parties, not just the
Postal Service. Valpak can file a second round of testimony rebutting the testimony of
other intervenors to the exact same extent as can the Postal Service. Second, rebuttal
testimony is not a way “to insert potentially extraneous matters into the record,” as
rebuttal testimony is limited by the scope of intervenor direct testimony. Thus, to return
to our stationarity example, if GCA could open the door on stationarity as an abstract
concern, but prevent the Postal Service from counter-designating responses that
demonstrate its inapplicability in this instance, and if GCA were not to file any direct
testimony on the matter, the Postal Service would have no opportunity to file rebuttal
testimony covering the same ground as the previously submitted responses to subparts
a and e. The ability to offer rebuttal testimony in no way obviates the legitimate
concerns of the Postal Service, Valpak’s gratuitous comments notwithstanding.
3 In concluding that a District Court had unfairly abused its discretion by allowing
opposing counsel to question the author of a letter about some parts of the letter, but
refusing to allow his own counsel to question him about other parts of the letter, the
Supreme Court cited both the common-law “rule of completeness,” and Federal Rule of
Evidence 106. Beech Aircraft Corp v Rainey, 488 US 153, 170-72 (1988). That Rule
reads:
2
When a writing or recorded statement or part thereof is introduced by a
party, an adverse party may require the introduction at that time of any
other part or any other writing or record statement which ought in fairness
to be considered contemporaneously with it.
4 The common law “rule of completeness” referred to by the Supreme Court in the
Beech case focuses on content in terms of subject matter:
-6-
designated questions, the designating party would have opened the door on that topic,
and any responsive answers by the witness on that topic should be fair game for
counter-designation. Fortunately, we are not faced with that hypothetical in this
instance, since the questions are all subparts of the same interrogatory, which, under
the current rules as understood by the Postal Service, constitutes the unit of response.
The above example demonstrates, however, how the current process may not
sufficiently protect the due process rights of the answering party, which is most often the
Postal Service. If GCA had posed its five questions on stationarity as five separate
interrogatories, the Postal Service would have had no discernible way to get all of those
responses into the record if some were designated and other were not. Perhaps the
remedy should be to allow a party to conduct redirect examination on any question
posed to its witnesses, if the response to the question is entered into the record as
written or oral testimony. Thus, intervenors who did not wish to pursue a subject upon
which discovery had been posed could shut the door by omitting all responses on that
subject from their designations, but once they chose to designate a written response
[M]erely by reason of the fact that the first party has introduced a part, [the
opposing party] has the right to introduce the remainder of the writing,
recording, statement, former testimony, or conversation relating to the
same subject matter. . . . [T]he common law right applies so long as the
other passage is logically relevant to the same topic as the part the
proponent offers.
McCormick on Evidence, 5th Ed. (1999), Vol. 1 at 252. Both the Federal Rule and the
common law rule constitute implicit acknowledgement of the prejudice that could be
imposed on parties in the absence of a such a policy, and both address concerns which
must be equally shared by this Commission, regardless of whether the Federal Rules of
Evidence or the federal common law formally apply in these proceedings.
-7-
pertaining to a particular subject, the Postal Service could conduct redirect examination
on that subject, just as it can with respect to any subject raised during oral crossexamination.5 It is possible that the current rules would allow redirect examination on
such grounds, but it certainly is not common under current practice.
In any event, the Postal Service does not agree with Valpak’s claim (page 1) that
a right to counter-designation “would shift much control of the process of written crossexamination to the party being cross-examined.” As long as the party conducting the
cross-examination has the ability to determine which of the subjects it has broached in
written cross-examination will be pursued on the record (via designation), it maintains all
of the control to which it is entitled. Under normal standards of due process, a party that
has, on the record, opened the door on a subject with a witness is not entitled to place
limits on the ability of that witness to address that subject. The door can be opened or
shut, but a wedge cannot be inserted halfway to allow the questioning party to get in
what it wants and keep out what it does not want. The stationarity example is a
concrete illustration of the perils of how such an approach could be exploited.
As the above discussion demonstrates, there are significant fairness and due
process issues at stake in this controversy. But there are administrative and practical
5 The potential denial of due process inherent in any contrary conclusion is shown in
sharp relief when one considers the analogous situation of GCA conducting its
examination of stationarity orally instead of in writing. If GCA had orally asked witness
Thress the questions posed in subparts b-d, there is no doubt that the Presiding Officer
would have permitted the Postal Service, on redirect, to explore the topic of stationarity
further and elicit the responses provided in a and e. It would be a clear violation of due
process to prohibit such redirect. It would be no less a violation of due process to
prohibit the Postal Service from providing its complete answers to written
interrogatories.
-8-
concerns to be considered as well. As noted repeatedly, if the unit of response is to be
the subpart (i.e., partial designations are to be allowed), it makes no sense to continue
to set the requirement for separate responses at the full question level. Yet, we can
once again examine the stationarity example to show that the structure initially chosen
by GCA – separate subparts in a question addressing one relatively discrete topic –
seems to provide the most natural and appropriate way to organize a record. Parties
tend to ask questions on one subject within the same interrogatory, and introduce
questions on another subject in a separate interrogatory. Treating complete
interrogatories as the unit of response provides an adequate (if not perfect) means of
separating questions by subject, in accordance with the above point that due process
rights essentially operate by subject.
Administratively, the current process has worked well, up until the point that
parties began to designate some subparts and omit others. This practice created
problems for the Commission’s Docket Section, and for counsel trying to keep track of
what has been designated and what actually appears in the cross-examination packet.
And as long as the Postal Service is afforded its due process right to counter-designate,
and affirms its intent to exercise that right by routinely counter-designating any omitted
subparts, the counterproductive aspects of allowing partial designations quickly become
obvious.
The comments of ABA/NAPM express concern (page 2) that parties may attempt
to “improve” the direct testimony of their witnesses “by injecting into interrogatory
responses self-serving statements which are not responsive to the interrogatory.”
-9-
Valpak never articulates its concerns so directly, but the tenor of its comments suggest
that its concerns are similar. Neither party cites any concrete examples of this practice,
and it certainly is not the intent of the Postal Service to employ such a transparently
troubling strategy. Even if there were any basis to assert that this type of concern were
well-founded, however, there is no reason to believe that changing the unit of response
from complete interrogatory to interrogatory subpart would likely remedy such behavior.
Witnesses intent on such a practice can just as easily insert nonresponsive material
into each subpart answer as they can into a unified answer to the whole interrogatory.
Ultimately, unless parties are willing to engage in far more procedural wrangling than
possibly seems warranted, they will simply have to rely on the Commission’s ability to
sort out the more relevant from the less relevant, and the more responsive from the less
responsive. There certainly is no basis to assert that, to date, the Commission has not
been up to that task.
ABA/NAPM (page 2) concludes that the Commission should not allow partial
designations if responsive counter-designations are allowed. Given the Postal Service’s
view that due process and fundamental fairness would in fact require the ability to
counter-designate, the Postal Service reaches the same conclusion – partial
designations should not be allowed. Maintaining that policy (or adopting that policy, as
the case may be) allows avoidance of any need to counter-designate. Administratively,
it generates less work for the Docket Section staff, less work for the attorneys in the
hearing room, and less opportunity for confusion and a transcript which does not include
all of the materials which the parties understood to have been admitted. Any
- 10 -
particularly egregious examples of non-responsive answers, which for now appear to be
an entirely hypothetical possibility, can much more efficiently be dealt with on an ad hoc
basis.
Respectfully submitted,
UNITED STATES POSTAL SERVICE
By its attorneys:
Daniel J. Foucheaux, Jr.
Chief Counsel, Ratemaking
______________________________
Eric P. Koetting
475 L'Enfant Plaza West, S.W.
Washington, D.C. 20260-1137
(202) 268-2992, Fax -5402
July 15, 2005
- 11 -
RESPONSE OF POSTAL SERVICE WITNESS THRESS TO
INTERROGATORIES OF GCA
GCA/USPS-T7-3
Please refer to your testimony USPS-T-7, Section III., starting at p. 254.
a)
b)
c)
d)
e)
Please confirm that it is a required condition in your regression
analysis that variables be stationary.
Please confirm that if variables are non-stationary, the results may
be spurious.
Please confirm that if the regression result is spurious, the
estimated coefficients may not be correct.
Please confirm that you have conducted the appropriate tests for
the stationary character of the variables. If confirmed, please
provide those tests. If not confirmed, please conduct the tests and
provide the tests results.
Please confirm you have corrected for the non-stationary character
of the data if present. If confirmed, please explain how you
accomplished that. If not confirmed, then explain on what basis
you conducted your regressions.
RESPONSE
a)
Not confirmed. The properties of Generalized Least Squares should be
satisfied so long as some stationary linear relationship exists between variables.
It is a sufficient condition, therefore, to have stationary dependent variables. The
Generalized Least Squares assumptions may be satisfied even with nonstationary variables, however, so long as a stationary linear relationship exists
between the dependent and independent variables using in the equation. In
such a case, the true residuals in the regression specification should be
stationary.
b)
Confirmed.
c)
Confirmed.
d)
Confirmed. I have performed stationarity tests on the data which I use in
my demand analysis on several occasions in the past. Results of investigating
the stationarity of mail volumes using a Dickey-Fuller test which were conducted
in the fall of 2004 are presented below. A more comprehensive investigation of
RESPONSE OF POSTAL SERVICE WITNESS THRESS TO
INTERROGATORIES OF GCA
possible stationarity and co-integration issues was conducted by my staff in
2002.
RESPONSE OF POSTAL SERVICE WITNESS THRESS TO
INTERROGATORIES OF GCA
Dickey-Fuller Test of Unit Root (Volume variable only)
Starting Date
Constant and Trend
Test Critical
t-Statistic
Values
Constant and no Trend
Test Critical
t-Statistic
Values
No Constant no Trend
Test Critical
t-Statistic
Values
FIRST-CLASS MAIL
First-Class Letters & Flats
-- Single-Piece
-- Workshared
First-Class Cards
-- Single-Piece
-- Workshared
Priority Mail
Express Mail
Mailgrams
1970:1
1970:1
1976:1
1970:1
1970:1
1976:1
1970:1
1977:1
1974:4
-8.7004
-9.8483
-5.5570
-4.8732
-4.6343
-9.3641
-3.1121
-4.2688
-4.1088
-3.4427
-3.4427
-3.4508
-3.4427
-3.4427
-3.4508
-3.4427 *
-3.4512
-3.4480
-5.0723
-4.3528
-6.6716
-2.5708
-4.4955
-3.6390
-1.3124
-7.7846
-0.7004
-2.8824
-2.8824
-2.8877
-2.8824 *
-2.8824
-2.8877
-2.8824 *
-2.8879
-2.8859 *
-1.6284
-4.2629
-7.2127
-0.4168
0.2011
-1.4649
-0.8010
-4.0479
0.9002
-1.9432 *
-1.9432
-1.9437
-1.9432 *
-1.9432 *
-1.9437 *
-1.9432 *
-1.9438
-1.9435 *
PERIODICAL MAIL
Within County
Nonprofit
Classroom
Classroom and Nonprofit
Regular Rate
1970:1
1970:1
1970:1
1970:1
1970:1
-3.3855
-7.0748
-7.9730
-6.8543
-6.0134
-3.4427 *
-3.4427
-3.4427
-3.4427
-3.4427
-0.9144
-2.7774
-6.9952
-2.6355
-3.3244
-2.8824 *
-2.8824 *
-2.8824
-2.8824 *
-2.8824
1.4071
0.3635
-0.1735
0.3980
0.6131
-1.9432 *
-1.9432 *
-1.9432 *
-1.9432 *
-1.9432 *
1970:1
1993:1
1993:1
1979:2
1993:1
1993:1
-3.1386
-4.2457
-3.9919
-5.9155
-2.9320
-6.2134
-3.4427 *
-3.5107
-3.5107
-3.4549
-3.5107 *
-3.5107
-1.3964
-1.8553
-3.6527
-6.6924
-2.5089
-4.3347
-2.8824 *
-2.9266 *
-2.9266
-2.8903
-2.9266 *
-2.9266
-1.0240
-0.8357
-0.5430
-4.3511
-0.4549
-0.3698
-1.9432 *
-1.9481 *
-1.9481 *
-1.9440
-1.9481 *
-1.9481 *
1970:1
1993:1
1993:1
1980:3
1993:1
1993:1
-8.5504
-8.3013
-5.4841
-6.6764
-4.7210
-3.8398
-3.4427
-3.5107
-3.5107
-3.4573
-3.5107
-3.5107
-6.3978
-8.3914
-4.2095
-6.5317
-3.5663
-1.7962
-2.8824
-2.9266
-2.9266
-2.8919
-2.9266
-2.9266 *
-0.5999
-0.1453
-0.0956
-1.7707
-0.9662
-0.4224
-1.9432 *
-1.9481 *
-1.9481 *
-1.9442 *
-1.9481 *
-1.9481 *
PACKAGE SERVICES
Parcel Post
Non-Destination Entry
Destination Entry
Bound Printed Matter
Media Mail
Library Rate
Media Mail and Library Rate
1970:1
1970:1
1991:2
1970:1
1970:1
1970:1
1970:1
-4.0578
-7.1372
-3.9801
-14.0523
-4.4472
-5.6592
-6.3774
-3.4427
-3.4427
-3.4970
-3.4427
-3.4427
-3.4427
-3.4427
-3.9722
-2.9272
-4.4542
-5.7088
-2.2023
-3.4052
-2.3755
-2.8824
-2.8824
-2.9177
-2.8824
-2.8824 *
-2.8824
-2.8824 *
0.0013
0.3877
-2.1577
-0.8554
0.4633
0.0998
0.4023
-1.9432 *
-1.9432 *
-1.9471
-1.9432 *
-1.9432 *
-1.9432 *
-1.9432 *
Postal Penalty
Free-for-the-Blind
1988:1
1970:1
-6.2140
-10.4610
-3.4794
-3.4427
-4.0298
-6.6492
-2.9062
-2.8824
-0.0206
-0.5139
-1.9458 *
-1.9432 *
-0.0482
-4.0626
-2.2587
-0.2569
-4.6965
-2.2071
-6.6799
-3.1235
-2.4179
-4.0488
-2.8824
-2.8824
-2.8824
-2.8824
-2.8824
-2.8824
-3.0124
-2.9266
-2.8824
-2.9266
1.7457
0.1121
-0.9805
1.8375
-0.1878
0.7204
-3.2418
0.1902
0.6065
-0.0034
-1.9432 *
-1.9432 *
-1.9432 *
-1.9432 *
-1.9432 *
-1.9432 *
-1.9581
-1.9481 *
-1.9432 *
-1.9481 *
STANDARD MAIL
Regular Rate Bulk
Regular
-- Basic ECR Letters
-- Basic ECR Nonletters
Enhanced Carrier-Route
High Density/Saturation Letters
High Density/Saturation Nonletters
Nonprofit Rate Bulk
Nonprofit
-- Basic ECR Letters
-- Basic ECR Nonletters
Nonprofit ECR
High Density/Saturation Letters
High Density/Saturation Nonletters
Registered Mail
1970:1
-3.4053
-3.4427 *
Insurance
1970:1
-5.8137
-3.4427
Certificated Mail
1970:1
-6.1692
-3.4427
COD
1970:1
-3.7933
-3.4427
Return Receipts
1970:1
-4.6968
-3.4427
Money Orders
1970:1
-2.1205
-3.4427 *
Delivery and Signature Confirmation
1999:2
-8.3619
-3.6450
Post-Office Boxes
1993:1
-3.9806
-3.5107
Stamped Cards
1970:1
-3.5371
-3.4427
Stamped Envelopes
1993:1
-4.0725
-3.5107
(*) Unit root hypothesis is not rejected at 5%
Asymptotic critical values for unit root test by Davison and MacKinon:
%5 critical value
t_ct
-3.4100
t_c
-2.8600
t_nc
-1.9400
*
*
*
*
*
RESPONSE OF POSTAL SERVICE WITNESS THRESS TO
INTERROGATORIES OF GCA
e)
The results of the reports cited in my response to d above suggested that
stationarity did not appear to be a particular concern in the analysis of mail
volume demand equations. In addition, Dickey-Fuller tests on the residuals from
my regressions indicate the presence of no unit roots. Hence, to the extent some
of the dependent variables may appear non-stationary, there appears to
nevertheless exist a stationary linear relationship between the dependent and
independent variables in every case here.
REVISED: JUNE 23, 2005
CERTIFICATE OF SERVICE
I hereby certify that I have this date served the foregoing document in
accordance with Section 12 of the Rules of Practice and Procedure.
________________________
Eric P. Koetting
475 L'Enfant Plaza West, S.W.
Washington, D.C. 20260B1137
(202) 268-2992, FAX: -5402
July 15, 2005
- 12 -