Perverse Incentives and the Fox Watching Over the Henhouse

Erik W. Charles
Erik W. Charles
In General
Erik Charles is the Vice President, Product Marketing, at Xactly Corporation where he is responsible for driving the product strategy, defining the product vision, and developing a strong team of product owners and designers.

The past few news cycles have been full of reports about Veterans waiting for health care at several Veteran’s Administration hospitals across the nation.  Some of the news has focused on the bonuses paid out to senior staff, while at the same time Vets were dying without being seen.

I wish I could say that this is a surprise.

Bonuses and other incentives work, the same way that any other tool works. If I bang a hammer with a nail, it works perfectly, and gets the job done properly. If I used it on a screw, however, I destroy what I was working on. Incentives need to be measurable, trackable, and controllable. Anything else is just pretending to have variable pay while instead holding a popularity contest.

Incentives also need to be tied to company objectives. The goal is to convince the employee that their goals are in sync with the firm’s goals.

So how did the VA do? Well, the Government Accountability Office had this to say, “no clear link between performance pay and providers’ performance.” Instead we find gems like “self-directed goals”. A physician set his own goals, only hits one, but got $7,500! I don’t know what is worse, a success rate of under 8% or doing so poorly on goals you set yourself! Not to mention that apparently nobody was holding him to task for any of this along the way.

Let’s dive deeper into the VA story that has received so much attention – long waiting periods combined with altered records to achieve a (questionable) bonus.

  • Measurability. Patient waiting time is an excellent measure – although it is only one part of what a patient wants. Adding in a patient outcome measure would also help. We not only want to see the patient, but we want to be able to know if we are helping.
  • Trackability: If you are measuring it, you should have dashboards open to everyone. Stream the data to a webpage showing your performance. Given that this is a governmental agency, where the information is open and owned by the general public – let’s see the leaderboards for the best VA hospitals in the nation.
  • Controllability: Can the person being measured actually impact the measurement? Health care is a complex business, and the physician does not always have control over all factors. We don’t want our doctor ending our session prematurely just to hit a number of appointments goal after all. Nor do we want an outcome measurement to be thrown by patients refusing to follow doctor’s orders.

 VA Scandal Fast Facts

  • Government Accountability Office last year reported that 80% of the VA’s 22,500 medical-care providers received $150 million in performance pay in 2011, though there’s no “clear link between performance pay and providers’ performance.”
  • Another physician who was disciplined for refusing to see ER patients—thereby causing six-hour delays in care—was awarded $7,500 because he met one of his 13 self-directed goals. He failed to meet the other 12, including attending staff meetings.
  • In 2013, Phoenix VA Director Sharon Helman received the No. 1 bonus out of 3,170 employees (at Phoenix): $9,345. The number of avoidable deaths was higher than at nearly any other hospital.
  •  In Chicago, $1 billion in salaries were supplemented with $4 million in bonuses over a three-year period.
  • At the seven most troubled VA facilities, almost $9 million in bonuses were paid out to 13,000 employees.
  • A scheduler at the New Hampshire VA testified to that effect in 2012 before a Senate subcommittee, saying that employees there submitted false reports of wait times for veterans to help reach performance goals that triggered bonuses for executives.

So lets assume we can find the 2-3  most appropriate measures (not 13 like the physician in my earlier example) for the VA system. What else needs to be done? How about not putting the measurement management under the control of those being paid. If it’s discovered that people are falsifying data, they should be fired. You wouldn’t let your sales team determine if they deserve a commission – you put management of the transactions under the control of the folks in finance. That way you keep a separation of controls, reducing the incentive to falsify data.

Pay bonuses for great care to our veterans, but don’t write plans that fail those same veterans with poor measures, and no controls.

Additional Reading:


Recent Posts

Thank You for Your Interest!

Please Share This Resource!

Perverse Incentives and the Fox Watching Over the Henhouse