Incentives Matter (or Why the Lab No Longer Has a Cafeteria)

Many people I know here in Aiken work at Savannah River National Lab. It's part of the Savannah River Site, an old nuclear weapon manufacturing facility that's currently being cleaned up and repurposed. SRNL is run by both the Department of Energy and Savannah River Nuclear Solutions, a consortium of companies put together to manage various aspects of the site and the lab.

Warning: This article may contain an
unsafe level of acronyms

After reading this article about problems at other national labs, I thought it worth writing about a couple of anecdotes I've heard about SRNL.

As a government-funded endeavor, the site and the lab have to get their funding re-approved every so often. It makes sense to have a set of goals the DOE and SRNS have to meet to ensure taxpayers are getting their money's worth. Theoretically, the success of the site and the lab can be measured by these metrics, and the funding and management contracts adjusted accordingly. Naturally, DOE and SRNS management work very hard to meet these goals and to collect the monetary incentives they get by meeting them.

Creating metrics by which the people managing SRS and SRNL can be judged makes sense, but only if the metrics themselves make sense.

Tear It Down

I know the chicken tetrazzini was bad,
but come on!

WSRC, the consortium which managed the site before SRNS, was charged with removing a lot of the old weapons manufacturing facilities. They were required to remove a certain amount of square footage and would get bonuses for removing more. Unfortunately, the contract did not specify what kinds of buildings they had to remove.

So WSRC bulldozed tool sheds, garages, even the lab's cafeteria - much simpler than removing manufacturing facilities full of old radioactive junk. They met their goal and got their bonus (though it's worth noting they did not get their management contract renewed), and the site was left with a bunch of radioactive junk that still needed cleaning up.

Fire It Up

Admittedly, there were problems
with the old procurement system.

This next anecdote particularly resonates with me, as I help program and install large enterprise IT systems.

The procurement system at the site is a mess. It was a mess before SRNS arrived, and it remained so after they took over. Something had to be done, so procurement was charged with installing a new software system to streamline purchasing and bill paying.

The software was built and installed, and the procurement department had a party to celebrate meeting their goal. Several weeks later, the department was in trouble because no one could figure out how to actually use the new system to pay bills and the past due notices were piling up.

Be Careful What You Ask For

In any large company, it makes sense to have goals and metrics with which to judge success. Upper management can't track the progress of every project and every group - that's why there are levels of management to synthesize things and present the big picture.

But success is only real if the metrics by which it's judged measure things that matter. It's important to spend as much care researching and drafting those metrics and goals as it is trying to meet them.