The Smartest Guys in the Room

adminStory

It looks like there are some guys over in L’Enfant Plaza having a goof making graphs and charts to show why they need to close thousands of post offices.  If the stakes weren’t so high and the damage to communities so great, it might be amusing watching these pranksters at work.  

These are "the smartest guys in the room," and their graphs and database tables use silly tricks that manipulate the data to make the plight of the post office seem way worse than it is and to make the alternatives seem better than they are.  They're like the Enron nerds who made millions shutting down the California power grid, sitting in front of their computer screens, detached from the reality of the pain they were causing.  (If you haven't seen the movie, here's the trailer.)

These tables come from the testimony Postal headquarters presented to the Postal Regulatory Commission.  They were also used in briefing materials that have been circulating in DC for a few weeks.

Table 1 is a chart that shows “annual customer visits” to the post office.  Given that we’ve been in a recession and the Postal Service has been encouraging people to use alternative access outlets, it should come as no surprise that these visits are on the decline.  Here’s the table:

As you can see, the table makes it look like the slope of the decline is precipitous, and in a couple more years, no one will be going to the post office at all.  (It's not clear what that y-axis is measuring, and in its interrogatories to the Postal Service, NAPUS asks for an explanation, but you get the idea—visits are declining.)  In the presentation materials, this graph is made to look even worse:

The graph uses a trick.  It only shows the top part of the vertical axis (starting at 800), so the decline looks much steeper, as if visits were falling at the rate of something like 15% a year, and heading rapidly toward zero. 

Here's what that table would have looked like if the whole vertical axis had been shown, instead of just the top part.  As you can see, the decline is gradual — about 4% a year — and hardly an illustration that people have stopped going to the post office.

The pranksters play the same trick with Table 2, which shows the proportion of retail revenue from various sources—post offices, retail partners, stamps by mail, self-service, etc.  

As you can see, the table makes it seem as if the revenue from post offices, stations, and branches is on a steep decline, rapidly heading toward nothing, while the revenue from other sources is dramatically increasing and overwhelming the post office. But once again, the smart guys in HQ are only showing the top part of the y-axis, thus exaggerating the proportion of revenue that comes from non-post-office sources.

Here’s a different view of the same data, this time showing the whole vertical axis.  The blue line is revenue percentage from post offices, the red line, from all other sources.  

Here's the same data represented as a bar chart, which shows even more strongly how post offices remain the largest source of revenue.

This graph and the chart show that, yes, revenue from post offices is decreasing while revenues from other sources are increasing.  But these representations show more honestly the fact that post offices still account for a much larger portion of revenue than the other sources (two-thirds), and they show that the changes are happening much more gradually.  And again, this shift is due largely to the fact that the Postal Service is working so hard to get people to use alternatives to the post office.  They have a PR campaign to accomplish exactly that.

Let’s move on to Table 4, which shows the total number of post offices (these are main, independent post offices, and the numbers do not include stations and branches).  Here’s the graph:

This graph comes from the USPS OIG’s “Barriers to Retail Network Optimization,” a document that complains about all the difficulties that are encountered when the Postal Service wants to close post offices.  The graph is intended to show that (1) closing post offices is nothing new, and (2) post office closures slowed dramatically in 1970, when the Postal Reorganization Act made it more difficult to close rural post offices. 

But the graph doesn’t really illustrate what they want it to.  It shows that post offices have closed at the rate of about 450 a year over the past 110 years, which comes to an annual closure rate of about one percent.  I'm sure if the Postal Service said they wanted to close one percent of its 32,000 post offices over the coming year, the country would not be going into the turmoil we're witnessing.

But they want to close something like 4,500 post offices in one year—that’s over 14% of our post offices.  And they want to close half the post offices over the next six years. The graph would have been a little more helpful if it had shown the dramatic drop they’re planning:

Let's move on to Table 5, the Distance Chart.  This one shows how far the nearest retail location is to any given post office with revenues less that $100,000 a year.  HQ had fun with this chart, too. 

This chart is intended to show that there are plenty of alternative postal facilities nearby, and people are not going to be seriously inconvenienced when their post office closes.  But there are several problems with this chart. 

First, the proximity to the nearest Post Office is based on "geographic coordinate distance," rather than driving distance.  But if you’re not a crow, it doesn’t help to have distances measured this way.  

In its defense, the Postal Service has told the PRC that the data was used only to help identify post offices for the closure list, but it will use actual driving distance when it does a discontinuance review on a particular post office.  However, in the briefing materials, it sure appears that they're talking about distance to another facility in real terms.  Next to the chart it says that "POs are not nearly as geographically isolated as perceived by the public or the media—90% of POs with less than 100k in revenue are within 10 miles of nearest PO, and nearly half are within 5." That certainly makes it seem like they're talking about driving miles.

So, let's see what this means in more concrete terms.  One of the post offices on the closing list is located in Betterton, Maryland.  We don't know what the Postal Service's database says, since it's "under seal" and not available for public inspection, but here's what happens when you look for the nearest post offices on the USPS Services Locator:

The Locator lists four post offices near Betterton (#1), along with the distance from Betterton, but it gives the useless geographic distance.  To find the actual driving distance, you have to go deeper into the Locator and ask for directions.  Here's what you find: 

Location Listed Distance Driving Distance
#2 Kennedyville 5.6 miles 7.2 miles
#3 Worton 6.6 miles 8.7 miles
#4 Earleville 8.5 miles 20.4 miles
#5 Aberdeen Proving Ground 8.8 miles 57 miles

To show just how off the data is, take a look at how you get from Betterton to Aberdeen.  This assumes you're not a really strong swimmer and will be going by car:

The estimated driving time is 1 hour, 23 minutes, and the actual distance almost 60 miles.  And let's not even get into the fact that this is a army weapons base and you probably couldn't get near its post office.  Yet this post office would be considered one of the retail locations classified as within 5 – 10 miles of the post office in Betterton, now being studied for closure.

Another problem with the chart is that it's not about the post offices on the closing list.  Rather, it is based on data for the 13,494 retail locations “for which latitude and longitude data were available.”  There's no way to know the relevance of this data to the the post offices facing closure.  Since many of them are rural, their average distances to other facilities may be far greater than the average.

And what's this about not having geographic coordinates for all their post offices?  They don’t have this information?  Have some of their post offices already fallen off the map?  The Save the Post Office website provides latitude and longitude data for the entire population of 4,500 post offices being studied for closure.  Just click on a post office on the map and then click on its name in the pop-up window, and you'll get to a page with a map, address, and coordinates for that post office.  (You can also use the state-by-state lists in the sidebar.)

Given that all this information is already out there, it's hard to understand why the database on which this chart is based is "under seal" and why the Postal Service is resisting motions to make it public.

There’s another table in the evidence presented to the PRC showing revenue information about the post offices that have been selected for closure studies, and apparently there are problems with this data too, but unfortunately, the Postal Service ain’t sharing it.  It too is "under seal."  The Postal Service says it would “suffer commercial harm” if the information fell into the hands of competitors.  

You can find more about the issues with these table in the PRC Docket — see the summaries of the interrogatories by the Public Representative and Limited Participant, as well as the issue over the documents under seal.

It doesn’t help the credibility of the Postal Service’s argument when they play games with data and graphs like this, and when they withhold information from the public as if it were a private business rather than a public agency. 

(A special thanks to Craig O'Donnell of the Kent County News, who lives near Betterton, Maryland, for pointing out these flaws in the USPS Locator.  Be sure to read Craig's terrific horror story about trying to save his post office in Still Pond, "They're Coming for Your Post Office."  It also gives a lot of practical advice for what to do when you find your post office on the closing list.)

Print Friendly, PDF & Email
(Visited 296 times, 1 visits today)
Facebooktwittergoogle_plusredditpinterestlinkedinmailFacebooktwittergoogle_plusredditpinterestlinkedinmail