USPS OIG Report: Accuracy of In-Office Cost System Data

Objective

Our objective was to assess the accuracy and reliability of In-Office Cost System (IOCS) telephone readings. The IOCS is the primary probability sampling system used by the U.S. Postal Service to attribute the labor costs of clerks, mail handlers, city carriers, and supervisors related to the handling of mail of all classes and rate categories.

Samples are collected using in-person observations or by telephone readings. Telephone readings are acceptable when it is not possible for a data collector technician (DCT) to reach an employee at his or her work location at the scheduled time of the sample. If an employee is not available for an in-person observation or telephone reading when the sample is scheduled, the sample must be rescheduled according to policy. In fiscal year (FY) 2016, the Postal Service conducted 546,650 IOCS readings: 305,800 (about 56 percent) were performed by telephone, and 240,850 (about 44 percent) were performed in-person.

The Postal Accountability and Enhancement Act of 2006 requires the Postal Service to file an Annual Compliance Report (ACR) with the Postal Regulatory Commission (PRC) within 90 days of the end of each fiscal year. The report analyzes cost, revenue, rates, and quality of service for all products. Further, it reports whether revenue for each mail class and service type covers its attributable costs, which are costs directly or indirectly caused by products.

What the OIG Found

While the DCTs observed by the U.S. Postal Service Office of Inspector General during selected site visits generally followed policies and procedures, opportunities exist to improve sampling procedures and controls to enhance the accuracy and reliability of the data. Specifically:

  • Discrepancies existed between FY 2016 IOCS data set filed with the PRC and the FY 2016 IOCS records contained within the Postal Service’s IOCS data entry system. Specifically, the data set filed with the PRC was missing 14,596 completed readings, 6,298 rescheduled readings, and 121 canceled readings. We also found 17,439 delinquent reading records within the final IOCS data set that were not complete and, therefore, unreadable. Management stated these discrepancies likely occurred because reading records were not transmitted to or approved in the IOCS data entry system by the end of the quarter, or were not properly processed for inclusion in the final IOCS data set. These factors resulted in the exclusion of reading records from the final IOCS data set.
  • Discrepancies existed between IOCS telephone reading data and Time and Attendance Collection System (TACS) data. IOCS telephone reading records indicated 94,929 carrier readings (31 percent) and 771 clerk/mail handler readings (0.25 percent) were not located in the facility or in the immediate area of the facility (“off premise”) during the scheduled reading times. From those identified as “off premise,” we selected a statistical random sample of those readings and determined that an estimated 15 percent of carriers and 59 percent of clerks/mail handlers were clocked into office operations at the facility during the scheduled reading times.

These issues occurred because the Postal Service did not have systems in place sufficient for DCTs to validate employee availability during telephone readings or for management to reconcile discrepancies after the fact between IOCS and TACS. Additionally, supervisors may have miscommunicated the actual location of the sampled employees to DCTs during telephone readings. If employees are potentially misidentified as being “off premise” and unavailable during the sampling process, it hinders data collectors from collecting comprehensive data. The data collected during IOCS readings are used to attribute costs to products. Therefore, if DCTs are missing opportunities to gather data, it could impact the accuracy of cost attribution.

  • Discrepancies existed between FY 2016 IOCS data set filed with the PRC and the FY 2016 IOCS records contained within the Postal Service’s IOCS data entry system. Specifically, the data set filed with the PRC was missing 14,596 completed readings, 6,298 rescheduled readings, and 121 canceled readings. We also found 17,439 delinquent reading records within the final IOCS data set that were not complete and, therefore, unreadable.

Management stated these discrepancies likely occurred because reading records were not transmitted to or approved in the IOCS data entry system by the end of the quarter, or were not properly processed for inclusion in the final IOCS data set. These factors resulted in the exclusion of reading records from the final IOCS data set.

In addition, there were no controls in place to verify completeness of the final data output. The Postal Service uses the final IOCS data set filed with the PRC for cost attribution. Incomplete sampling data impacts the accuracy and reliability of reported costs for mail products and services. This issue resulted in $31 million in misallocated costs among products, representing about 0.09 percent of the total labor costs distributed to FY 2016 products.

What the OIG Recommended

We recommended management:

  • Establish automated controls in the IOCS data entry system that validates IOCS telephone readings are rescheduled according to policy.
  • Verify employees’ activities and location at the time of IOCS telephone readings and report discrepancies with TACS to supervisors if corrections to clock rings are needed.
  • Verify all IOCS readings are transmitted to and approved in the Postal Service’s IOCS data entry system by the end of the quarter.
  • Confirm the completeness of the final IOCS data set filed with the PRC for cost attribution prior to submission to the PRC.

Read full report


Source: USPS Office of Inspector General

Leave a Reply

Your email address will not be published. Required fields are marked *