As par tof the "discovery" phase of the PRC proceedings, the Public Representative (PR) can ask for more information and explanations from the Postal Service.
First Interrogatories, #1 – #5[The original document is here.]
The PR refers to witness Boltd’s statement, “It should be emphasized that postal management is not pursuing the RAO Initiative in order to achieve any predetermined operating cost savings target in the postal retail network.”
The PR asks the following: Does the Postal Service expect to achieve cost savings if the RAO Initiative is implemented? What is the estimate of the cost savings? And please confirm that your testimony does not include an estimate of cost savings.
(This may be a reference to the fact that the Postmaster General and others are going around saying that they hope to save $200 million by closing these post offices. If there is a particular goal like that, it would suggest that post offices are being closed in order to meet the goal, rather than on the basis of an evaluation of the post office itself.)
The PR also asks if the Postal Service currently maintains a database of all retail facilities, and if so, to provide it, with a full explanation of all the data fields, for fiscal years 2008, 2009, and 2010. (The Postal Service is known to have such a database, and in an article a few months ago, which I can’t locate right now, there was reference to the way someone could call up the info on a computer in seconds.)
The PR then refers to another passage in Boldt’s testimony, where he states, “the Postal Service has examined earned workload data for each of its retail facilities and has identified all Post Offices for which earned workload amounted to less than two hours per day and annual revenue was no greater than $27,500. Over 2,800 candidate offices were identified using data collected from the Small Office Variance tool.” In response, the PR asks how the Postal Service determines the workload and EAS classification of each Postal Facility?
Next, the PR refers to Boldt’s statement that“retail locations with annual walk-in revenue of less than the relatively low figure of $100,000 generally tend not to be as geographically isolated as some might assume,” and he refers to Table 5, which provides data for 13,494 retail locations for which latitude and longitude data were available. The PR askshowmany postal retail locations averaged less than $100,000 in yearly walk in revenue in FY 2008, FY2009, and FY2010, and she asked why data was not available for all of the postal locations, and what the implications of incomplete data might be in terms of calculating impacts of the RAOI.
The PR also notes that the file “FY2010 Offices under $100,000 by Technology.xls” contains location and revenue data for postal retail locations using different reporting technologies. So the PR asks for confirmation of the accuracy of the database, wonders why some retail postal facilities have negative yearly revenue, and asks for more discussion about the different reporting technologies used to estimate walk-in revenue.
The PR continues, noting that the Postal Service has stated that “approximately 85 percent of postal retail revenue is generated from the sale of postage,” and she asks for more information about data this as well.
Second Interrogatories, #6 – #9[The original document is here.]
The Public Representative proceeds to ask about the non-public files on the proximity of postal retail facilities to other postal retail facilities, the “Nearest Neighbor graph data” table. The PR focuses, as an example, on several post offices in Arkansas that are on the closure list: Red Devil, Crooked Creek, and Steelmute. The PR has used the post office locator on the USPS website, and found that the closest retail facilties to Red Devil are Steelmute (12.7 miles) and Crooked Creek (31.6 miles), as well as Chuathbaluk (not on the list, 37.8 miles from Red Devil).
The PR then asks if the information she has found on the USPS website is accurate, and why this information differs from what appears in the database provided by the Postal Service. She also asks if the Nearest Neighbor pool contains post offices that may also close, how are “nearest neighbors” identified and their distances calculated, and what steps the Postal Service will pursue to ensure that accurate information is available regarding the distance to alternative access points in event of closure. (The point here is that the Postal Service may be measuring distances in a flawed way, and they may be identifying a nearby post office as an alternative, yet that post office itself may be closed.)
The PR goes at the issue from another angle, this time focusing on the post office in Goodsprings, Alabama, on the closing list. The Post Office locator on the USPS website says the closest postal retail facility is the Parrish AL location, which is 5.5 miles away. The PR notes that in the Nearest Neighbor data sheet, the nearest neighbor to the Goodsprings post office is not recorded as 5.5 miles away, so she wants to know why there are these inconsistencies.
Boldt Replies to Public Representative, #1 – #9 (August 8, 2011)[The original document for Responses # 1 – 3, here; Responses #4 – 9, here.]
1. On page 13 of your testimony, you state: “It should be emphasized that postal management is not pursuing the RAO Initiative in order to achieve any predetermined operating cost savings target in the postal retail network.”
(a) Does the Postal Service expect to achieve cost savings if the RAO Initiative is implemented? Yes
(b) Please confirm that your testimony does not include an estimate of cost savings from implementing the RAO Initiative. Confirmed. Please see the USPS response to DBP/USPS-18
(c) Please provide an estimate of the cost savings from implementing the RAO initiative. Please see the USPS response to DBP/USPS-18
d) Does the Postal Service currently maintain a database of all retail facilities? If so, please provide this database in either excel format or as a flat file for FY2008, FY2009 and FY2010. Please identify and define all data fields
Response: There are various data bases that contain some information about retail facilities. There is no one system that contains all postal data regarding all of its retail facilities. The USPS Facilities Database (FDB) is a system that contains some retail facility data and is able to access or link to a wide variety of other data about retail facilities from other data systems However, it does not exist in a form that makes it possible to create an Excel or flat file replica. It is hoped that the technical conference will shed light on what exists.
(e) Does the Postal Service currently maintain a database of all retail facilities with cost and revenue for each facility? If so, please provide this database in either excel format or as a flat file for FY2008, FY2009 and FY2010. Please identify and define all data fields.
Response: Almost. The Facilities Data Base contains or is able to access data by six-digit facility finance number. However, the data for approximately 1900 stations and branches are rolled-up to the finance number of the Post Office to which they report in FDB. Accordingly, FDB cost data for various finance numbers are an aggregate figure that covers multiple associated retail units. With that limitation, the Postal Service will file a Library Reference document responsive to this interrogatory with cost/revenue data for FYs 2008-10.
2. On page 15 of your testimony, you state: “the Postal Service has examined earned workload data for each of its retail facilities and has identified all Post Offices for which earned workload amounted to less than two hours per day and annual revenue was no greater than $27,500. Over 2,800 candidate offices were identified using data collected from the Small Office Variance tool.”
(a) How does the Postal Service determine the workload and EAS classification of each Postal Facility?
The Postal Service assigns standardized quantitative values to various tasks performed at retail locations to determine earned workload. Nationally established standardized productivity targets are applied to the unit workload to calculate earned workhours. Based on its earned workload, a facility is classified as requiring that its Postmaster be of at least a certain postal EAS grade.
(b) Please provide the “Small Office Variance Tool” in either excel format or as a flat file for FY 2009 and FY2010. Please identify and define all data fields.
The Customer Service Variance (CSV) model is used to review Function 4 operations, delivery unit non-carrier bargaining unit activities, in Cost Ascertainment Group (CAG) H to L offices. Cost Ascertainment Group (CAG) classifications are used to identify Post Offices according to revenue generated. CAG H-J offices have 190 to 949; CAG K offices have 36 to 189; and CAG L offices have fewer than 36. CSV integrates locally reported unit workload from various national data systems. However, it does not exist in a form that makes it possible to create an Excel or flat file replica. It is hoped that the technical conference will shed light on what exists within or is accessible via CSV.
3. On page 11 of your testimony, you state that “retail locations with annual walk-in revenue of less than the relatively low figure of $100,000 generally tend not to be as geographically isolated as some might assume. Table 5 below is based on data for 13,494 retail locations for which latitude and longitude data were available.”
a. How many postal retail locations averaged less than $100,000 in yearly walk in revenue in FY 2008, FY2009, and FY2010. Response: FY 2008: 13,432. FY 2009: 13,860. FY2010: 14,243.
b. The file “nearest neighbor graph data.xls” tab “data” contains location title data for a confidential number of postal retail locations. Was the Postal Service unable to obtain latitude and longitude data for all postal retail locations? If not, please explain why some postal locations were excluded from the “nearest neighbor” calculation.
c. Please discuss how exclusion of postal retail locations from the nearest neighbor calculation will impact the retail access optimization process.
Response to b-c:The Nearest Neighbor depiction in Table 5 is meant to present a general illustration of proximity based upon easily accessible latitude/longitude data from the Facilities Data Base. Creation of the Table did not rely on the more robust data source — Electronic Facilities Management System. EFMS provides a means of determining driving distance between postal retail facilities and is routinely used for determining proximity of nearby postal retail facilities as part of the Handbook PO-101 discontinuance review process. Accordingly, the fact that Table 5 provides a general overview based on limited latitude/longitude data has no bearing on how inter-facility proximity is being measured or analyzed as part of the RAO Initiative.
6. The postal retail facility “Red Devil AK” is under consideration for closure according to cell A6 in tab “new_2hrs” in USPS-LR-2, as is “Crooked Creek AK” (cell A5), and “Sleetmute AK” (cell A14). According to the Post Office locator at http://usps.whitepages.com/post_office/AK/99656?d=40”, the closest postal retail facilities to the Red Devil AK location are as follows:
(a) Please confirm that in cell I13442 of tab “data” in file “Nearest Neighbor graph data.xls,” the nearest neighbor to the Red Devil AK postal retail location is neither recorded as 12.7 miles, nor 31.6 miles, nor 37.8 miles.
(b) Please explain why the data for this location in USPS-LR-NP1 does not match the information available on the USPS Post Office locator.
As explained in response to PR/USPS-T1-X, Table 5 is based strictly on data reflecting latitude/longitude "as the crow flies" distances. Postal Locator distances are driving distances between locations.
(c) In general, does the “Nearest Neighbor” pool include postal retail facilities that, in addition to the target facility, may be closed as a result of the RAO initiative within 365 days of the determination to close the target facility?
The question is not clear. Nearest Neighbor facilities are ones for which latitude/longitude data were available in a particular data base that does not reflect the entire postal retail network or the subset of the retail network within the scope of the RAO Initiative. No analysis of the Table 5 facilities has been performed to determine any degree of correlation they may have to RAO candidate facilities, or to facilities that may be closed 365 days (or any other time period) after a facility is closed as a result of RAO.
(d) How are “Nearest Neighbors” identified and their distance from each other calculated when there are multiple (i.e., more than two) concurrent postal retail facility closures within a particular geographic area?
Bear in mind that the Nearest Neighbor Table 5 plays no role in the execution of the RAO Initiative. Surface driving distances will be used in the RAO Initiative to determine the distance between a candidate for discontinuance and the closest nearby retail location. If RAO candidate facilities are in sufficiently close proximity that each is the "nearest neighbor" to the other, this fact will be taken into account in the review process in determining the fate of the two facilities.
(e) Please explain the steps the Postal Service will pursue to ensure that accurate information is available regarding the distance to alternative access points in event of closure.
Electronic Facilities Management System driving distance measurements will be checked against local management knowledge, other driving distance mapping software such as Bing, or odometer tests, if necessary.
7. The postal retail facility “Goodsprings AL” is under consideration for closure according to cell A26 in tab “new_2hrs” in USPS-LR-2. According to the Post Office locator at http://usps.whitepages.com/post_office/35560, the closest postal retail facility is the Parrish AL location, which is 5.5 miles away.
Please confirm that in cellI 1636oftab“data”infile“ Nearest Neigbor graph data.xls,” the nearest neighbor to the “Goodsprings AL” postal retail locations is not recorded as 5.5 miles away. Confirmed
Please explain why the data for this location in USPS-LR-NP1 does not match the information available on the USPS Post Office locator. Please see the response to PR/USPS-T1-6(b).
According to Google Maps, the driving distance from the Goodsprings AL location to the Parrish AL location is 7.6 miles, (http://maps.google.com/maps?saddr=824+Goodsprings+Road,+P arrish,+AL&daddr=5911+Highway+269,+Parrish,+AL+35580&hl=en &ll=33.694352,-87.261658&spn=0.212236,0.140762&sll=37.0625,- 95.677068&sspn=51.222969,31.289063&geocode=FbuUAQId9B3 N-imNkHZqS_OIiDGuiiDCy8Ixhg%3BFXzwAgIdz1nM- inlfxUlZoqIiDHSSKzOd99jwA&mra=ls&t=h&z=12) please discuss the steps the Postal Service will pursue to ensure that accurate information is available regarding the actual travel distance to alternative access points in event of closures.
Please see the response to PR/USPS-T1-6(e).
8. Please discuss how the database used to calculate USPS-LR-NP2 and table 5 on page 12 your testimony differs from the database used to calculate the distance between postal facilities at http://usps.whitepages.com.
Table 5 is based strictly on data reflecting latitude/longitude "as the crow flies" distances.
White Pages is an outside vendor that currently hosts the Post Office Locator on usps.com. It uses the Bing mapping system to create maps and driving directions. As customers use the Post Office Locator on usps.com, White Pages geo codes a customer's input address information, then utilizes geo code data for the facilities in the USPS Facilities Data Base to create the map/driving directions to each postal location.
9.Please provide the database used to calculate the distance between postal facilities at http://usps.whitepages.com.
The Bing mapping system database used to calculate driving distances for White Pages is proprietary to Microsoft. The database is the intellectual property of Microsoft and is not within the custody or control of the Postal Service.
Third Interrogatories, #10 – #13[The original document is here; a summary is on the way.]
Boldt Replies to Public Representative to #11 – #12[The original document is here.]
11. Please confirm that the calculation of walk-in revenue in USPS-LR-NP2 does not include revenue from any First Class Presort Mail, Standard Mail, or Periodicals Mail.
12. Can First Class Presort Mail, Standard Mail, or Periodicals Mail be entered and paid for at any of the postal retail facilities listed in USPS-LR-NP2 file “Basket Analysis by UFN.xls” or “USPS-LR-N2011-1_NP2 Revenue Distribution for LT100K Locations – Pie Chart.xls”? If so, please explain why this revenue has not been included.
It can be entered at many, if not all of those facilities. However, see the response to DBP/USPS-30, which explains why it is not counted as retail revenue.