Numbers We Can Trust: Why the Postal Regulatory Commission Needs Better From the USPS

April 11, 2019

“There are three types of lies: lies, damn lies and statistics.” Mark Twain popularized the phrase more than a century ago, and today’s pundits still cite it to reject data they don’t like. Twain wasn’t condemning statistics, though — he was pointing out that flawed methods distort real numbers.

Unfortunately, the world of postal is not immune to those bad habits.

The Postal Regulatory Commission has recognized flats processing as an issue. In the FY2015 Annual Compliance Determination, it requested that the USPS identify the flats causing “pinch points” and create a measurement for each point.

The USPS followed orders, and those measurements appear in its required annual compliance report to the PRC. It bases those measurements on abundant electronic data that the mailing industry supplies, and on its own IV scan data.

Below are two examples of the information that the USPS submits to the PRC, along with reasons why they’re in line with Twain’s notion of “statistics.”

1. Flats Sequencing System

The Flats Sequencing System (FSS) is a known issue for the USPS. The Postal Service put together a set of data points that report on the efficiency of the process. Year after year, these numbers keep going in the wrong direction.

The data tells us that the system is inefficient, yet the USPS operations folks say that FSS works. Either these folks are mistaken, or the statistics are wrong.

2. Bundle Breakage

Bundle breakage is another concern for the USPS. There is no direct calculation, so the Postal Service tried to create one. It says that a bundle breaks when processing equipment scans more than three individual pieces from that bundle, as identified in the electronic documentation.

It seemed logical to someone at the USPS that this would be correct. But in reality, there are plenty of valid reasons for this to happen — the reorder and recall process used to make all books, manufacturing observation that bundles as designed will fail — and perhaps a broken bundle.

When asked, the USPS acknowledges that there are other reasons for this data irregularity. Yet they continue to publish their findings to the PRC as “broken bundles.”

These are just two measurements that the USPS supplies to the PRC. They illustrate why it’s not surprising that the USPS still has trouble managing flats processing costs.

The PRC is concerned about the flats processing — unfortunately it claims it has no power to compel the USPS to do the right thing. Right now it can only continue to ask the USPS to supply data. A recent proposed rule (Order No. 5004) seeks to expand the annual request for information.

The question is, why hasn’t the PRC been able to get the USPS to make changes, even when the data clearly says it should?

Remember that the USPS is at the center of a $1.4 trillion mailing industry — if the PRC can’t compel it to act in its best interest, then maybe exposing its shortcomings will eventually bring the issue to light. Unfortunately, time is running out. Waiting for the USPS to acknowledge the issue might not work.

While it’s unlikely that the USPS deliberately manipulates numbers to create a false picture, it’s just not scientific in the way it actually calculates and presents the data. Those methods need to change, and the PRC must set more stringent standards for reporting — then hold the USPS accountable to improve the numbers, not just report on them.