Spatial inequality in veterans benefits

NPR reports this morning on their analysis of veterans benefits availability and generosity across the nation, analyzed down to the county level.  Here's the lede:

NPR, together with member stations WBUR, Lakeshore Public Radio and KUOW, looked at data from 3,000 counties nationwide, and found there's a huge variation in coverage from state to state — and even within a state — on how much the VA spends per veteran. 

We also found there's no obvious pattern. And there's no strong association between spending per veteran and the size or age of the veteran population, or the affluence of a particular area.

Even though the reporters said there is "no obvious pattern," I immediately assumed that rural vets were likely getting less than urban vets when it comes to benefits including "health care, monthly disability checks, home loans, life insurance, and education through the GI bill, among others."  That is supported by this tidbit:

• Spending in Suffolk County (Boston) is $11,286 per veteran while in Barnstable County, on Cape Cod, spending is just $4,560.  

But then there is this tidbit:

Among the states, West Virginia and Arkansas had the highest per-veteran spending in 2013 — just over $7,600. Indiana, New Jersey, Delaware and Pennsylvania had the lowest — less than $5,000. Nationally, the average is just over $6,000. That's after filtering out things like costs to build and operate VA facilities.

Because West Virginia and Arkansas are widely considered "rural states" and have significant rural populations, it would appear that some rural vets are getting their due.  But then, Pennsylvania and Indiana also have significant rural populations, and their numbers are not good.  

Looking just at health benefits, the reporters also found wide disparities "without discernible patterns."  Here's one data point:

[S]pending is nearly $30,000 per patient in San Francisco, and less than $7,000 per patient in Lubbock, Texas. Nationally, the average is just under $10,000. In places where more veterans are enrolled in VA health benefit plans, spending per veteran did tend to be higher.

That last point may be indicative of economies of scale that larger VA hospitals can achieve.  

Journalists Quil Lawrence and Martha Bebinger note many possible explanations for the disparities:

Some are demographic and beyond the VA's control, while others the VA could maybe do something about. Some of the range is because benefits cost more in different places. Other discrepancies are because veterans aren't using all the services they're due.

The story features a cool interactive features, so I looked up the data on my home county, Newton County, Arkansas, population 8,064, and a persistent poverty county.  I expected to find that lots of vets live there, and I was right.  Here's what it showed me:

Number of veterans 
756 in Newton County
250,095 in Arkansas
21,882,153 in the United States
Veterans as a share of the population
13% in Newton County
10% in Arkansas
8% in the United States
Average spending per veteran
$6,935 in Newton County
$7,627 in Arkansas
$6,088 in the United States
Share of veterans receiving medical benefits
42% in Newton County
35% in Arkansas
26% in the United States
Medical spending per veteran patient
$7,331 in Newton County
$10,259 in Arkansas
$9,840 in the United States 

Note:  U.S. Department of Veterans Affairs spending figures include only categories that go directly to veterans (such as medical, compensation or education benefits) and exclude capital expenses (such as facilities and new construction).

Here is another recent post linking veterans to rurality.