- Project Title:
- Accuracy of IOCS Data
- Start Date:
- Wednesday, March 1, 2017
- Estimated Report Release Date:
- August 2017
The In-Office Cost System (IOCS) is the Postal Service’s primary probability sampling system used to distribute the labor costs of clerks, mail handlers, city carriers, and supervisors related to the handling of mail of all classes and rate categories. The data is used to distribute about $20 billion annually in volume-variable costs to mail products. Approximately 550,000 IOCS samples are taken annually.
IOCS data is collected in one of two ways: in-person or over the telephone. An IOCS in-person reading consists of observing and recording the activity that a selected employee performs at a designated time during a specific workday. In addition, the data collector records the characteristics (e.g. indicia, special services, weight, etc.) of any mail or mail transportation equipment that the sampled employee is handling.
Telephone readings are conducted when total time to (1) travel to the facility, (2) locate the employee, (3) take the reading, and (4) return to the normal duty station are estimated to be in excess of 60 minutes.
The Data Collector Technician (DCT) conducts and calls the facility, identifies themselves, and asks the postmaster/station manager/supervisor or person in charge to conduct the IOCS telephone reading. If the primary or backup is not available, the DCT asks the person to recommend someone to help conduct the reading. The DCT conducting the telephone readings follows a specific set of instructions.
As noted within a previous audit, 70 out of 163 (43 percent) of the IOCS telephone readings reviewed were inaccurate. The telephone readings were inaccurate because data collectors and supervisors did not follow policies and procedures, data collectors may have falsified the reading, or respondents may have provided inaccurate information. In addition, an analysis of complete IOCS telephone readings indicates that some data collectors adjust the timing of their morning and afternoon IOCS readings which could alter the accuracy of the results.
- How can the Postal Service improve its IOCS telephone readings to obtain the data needed to effectively and efficiently distribute labor costs?
- Are there any obstacles in the sampling process that would prevent the collection of accurate and complete data?
- Can telephone sampling be automated?