Washington Mutual and Perverse Incentives

The causes of the mortgage meltdown are diverse and complex, and anyone who points to a single cause is almost certainly wrong.  But there is also little doubt that perverse incentives, as economists like to call them, played a role in creating the mess.  I've written about how bad incentives can, over the long haul, drive people to do things that are ultimately destructive to themselves and their organizations -- with a little help from a great Dilbert cartoon -- and see this great list of examples on Wikipedia. I love this one:

"19th century palaeontologists traveling to China used to pay peasants for each fragment of dinosaur bone (dinosaur fossils) that they produced. They later discovered that peasants dug up the bones and then smashed them into multiple pieces to maximise their payments"

I am not among those psychologists who argues that money is a weak motivator of human behavior.  Rather, as Jeff Pfeffer and I argue in Hard Facts, the problem with using money as a motivator is that it is very difficult to get the incentive system designed so it motivates the right kind of behavior and discourages the wrong kind.  You could argue that the problem with using financial incentives is that they work too well rather than not well enough -- causing people to focus their attention narrowly on a small number of things and to forget more subtle and long-term issues.  This is especially  true with individual incentive systems -- which have been shown to lead to everything from garbage collectors in Albuquerque driving too fast, driving broken trucks, and missing many collections so they could finish their routes more quickly to schoolteachers in Chicago cheating on standardized tests -- giving students the right answers to tests or changing the answers themselves -- to get performance bonuses linked to student test scores.

Alas, the once great Washington Mutual bank seems to have fallen, in part, because so much emphasis was placed on writing as many mortgages as possible, fitness of the borrower be damned.  Check out this story in the Sunday New York Times.  Here is a excerpt that shows how their reward system -- and misguided culture to supported it -- helped bring down this once great bank:

MS. COOPER started at WaMu in 2003 and lasted three and a half years. At first, she was allowed to do her job, she says. In February 2007, though, the pressure became intense. WaMu executives told employees they were not making enough loans and had to get their numbers up, she says.

“They started giving loan officers free trips if they closed so many loans, fly them to Hawaii for a month,” Ms. Cooper recalls. “One of my account reps went to Jamaica for a month because he closed $3.5 million in loans that month.”

Although Ms. Cooper couldn’t see it, the wheels were already coming off the subprime bus.

“If a loan came from a top loan officer, they didn’t care what the situation was, you had to make that loan work,” she says. “You were like a bad person if you declined a loan.”

One loan file was filled with so many discrepancies that she felt certain it involved mortgage fraud. She turned the loan down, she says, only to be scolded by her supervisor.

“She told me, ‘This broker has closed over $1 million with us and there is no reason you cannot make this loan work,’ ” Ms. Cooper says. “I explained to her the loan was not good at all, but she said I had to sign it.”

The argument did not end there, however. Ms. Cooper says her immediate boss complained to the team manager about the loan rejection and asked that Ms. Cooper be “written up,” with a formal letter of complaint placed in her personnel file.

Ms. Cooper said the team manager told her to “restructure” the loan to make it work. “I said, how can you restructure fraud? This is a fraudulent loan,” she recalls.

Ms. Cooper says that her bosses placed her on probation for 30 days for refusing to approve the loan and that her team manager signed off on the loan.

Four months later, the loan was in default, she says. The borrower had not made a single payment. “They tried to hang it on me,” Ms. Cooper said, “but I said, ‘No, I put in the system that I am not approving this loan.’ ”

My question: Problems like this crop up over and over again. What can we do to stop them?  Should we stop using individual incentives?  I think that is too extreme, but how do we design individual incentive systems that avert a narrow and misguided focus?  I think part of the answer is supplementing them with other motivators, interesting work, a system that encourages pride and praise, and leaders who set the right example -- all that press people to move in the right direction rather than the wrong direction.  But when the money becomes so vivid, people don't even seem to notice long-term dysfunctions, let alone potential ethical lapses. The Enron story was much the same.

I have no magic wand and don't believe that anyone else does.  There are cultures -- I think of P&G and Google -- that seem to have avoided such evils. But it isn't easy.

I once heard a group of CEOs argue that there would be far fewer ethical lapses and they would be much better at doing the right thing in the long-term if the system was changed so that they reported earnings once a year rather than four times a year.  The world is pretty much going in the opposite direction, of course, with people wanting updated information constantly -- but I think it is an interesting thought experiment.

Other thoughts?

Comments

[Alex]

Yeah, it's not surprising that this happened. The incentives setup for governmental budgeting are especially perverse. Take CA's current budget crisis, for example. The Governator focused on short-term goals...times were great for a few years, but now we are strapped for billions. If CA can issue an IOU to me, can I pay my taxes with an IOU?

[Kevin Rutkowski]

Jay, I like your plan for paying trailing commissions as the loan is paid back. It seems to be a good incentive to give solid loans and to stick around for a while at a company. I wonder why a plan like that wasn't put in place.

I would guess that part of the issue is that Wamu was looking primarily at the fees they could generate up front. Perhaps the fact that Wamu sold the loans and didn't service them long term also would make it difficult to calculate that incentive.

Regardless, it seems like if banks could incent people in that manner, it would be good for the long term health of the bank.

[Paul Hebert]

Hate the player but don't hate the game.

Incentives - properly designed incentives - have their place in guiding and reinforcing behavior. Typically, though, the incentive is designed by someone with little experience in incentive program design and based on a too simple an idea.

Any professional who has put together more than a couple of programs will tell you there is always a possible "unintended consequence" - the key is to look for them - no different than a stress test on the heart, you must take the program or plan to the extreme and determine if there is a problem. Very often you only have to push a particular program a little bit before you see the issues - and adjust accordingly.

Citing Kohn as a reason to avoid incentives doesn't really translate, as his work was around intrinsic motivation and focused on school children. Incentives and rewards are required in business. To think you can manage a company culture and performance without them is akin to saying we should eliminate email because some abuse the technology.

Properly designed incentives work and pay big dividends.

[Jay Godse]

This WaMu story has all the hallmarks of front-loaded or transactional commissions. In such a case, the sales person or sales manager makes money on deal-flow, and not on long-term value. The fix here is pretty simple. Only pay trailing commissions on loans as they get paid back. That will encourage behaviour that is good for the bank in the long term.

One of the deeper problems in the North American equity markets is that over the last 30 years, shareholders have consolidated and/or banded together. In the past, individual shareholders held small chunks of equity, and didn't have any real say how the company was run. Professional managers could act in the long-term interest of the company with impunity because that led to improved personal compensation through long tenure. That was the best way to make money because CEO salaries were not as high back then in relation to average worker salaries.

Over time, shareholders consolidated into mutual funds and pension funds. These funds could now put the managers' "feet to the fire" at every shareholders meeting. Over time, they insisted on more and more business performance, and offered management teams more and more money for their services. Of course, business performance is hard to measure if there are no profits, so other metrics were needed. Boards of directors compelled management teams to implement the latest fads such as raising market share, increasing capacity in anticipation of booming markets, improving revenues, improve "operating margins", lower fixed and variable costs, outsource, etc, and measured these with numbers.

So, given that most management teams could not turn lead into gold, or squeeze blood from a stone, they proceeded to ensure that they "made the numbers". Unfortunately, this lead to lots of perverse behaviours such as sandbagging sales, creatively booking revenues, incurring "one-time charges" every quarter while "growing" operating margins, and so on. As the management teams made their numbers, the consolidated shareholder proxies gave them lots of money, even if the companies were not profitable.

Unfortunately, most companies are not star performers such as Cisco, GE, Microsoft, etc. Most only grow at the rate that populations grow. As a result, the average company paid too much for their management teams and didn't have any results to match. It's easy to blame the archetype of a greedy money-grubbing CEO, but in reality, a lot of this was driven by consolidated shareholder proxies such as mutual funds and pension funds trying to get more than was reasonably possible from most companies.

The solution to this problem is harder. I believe that paying management teams from long-term profits will encourage the right long-term behaviour. Other remedies would be to find "numbers" that better reflect company needs such as profits, margins, and long-term growth. That is a hard problem...

[Kevin Rutkowski]

One of the biggest problems seems to be that the incentives all reward short term behavior at the expense of the long term. It is very difficult to measure the long term performance of a loan at the front end.

I agree that many of these types of incentives should simply be done away with. Instead, interesting work and a compelling vision for a company should result in people looking at what's best for the company.

The way companies throw away employees and the way employees leave companies, many people don't feel any reason to help build anything long term at a company.

The latest book from the book club that you wrote about earlier this year is Extreme Leadership by Steve Farber. I think that the principles that he discusses in his short, compelling book would well serve a company for both the short and long term: cultivate love, generate energy, inspire audacity, and provide proof.

[Jan]

I think we are hardwired to game the system. It's a survival mechanism since 100 000 years. We simply must understand what make the system tick, and manipulate those levers. Some of us have more of this in our DNA, some less. So what can we do about it? I'm as lost as everybody else. One thing we could do, is to stop designing systems from singularities, i.e. the extreme cases that will always occur. Design the system from the normal case, and deal with the extreme cases on an individual basis. Does that make sense?

[Bob Sutton]

Rick,

What you say makes complete sense to me -- but it mystifies me why so many smart people fall prey to the same damn thing over and over again!

Rick, Brennan, and John -- thanks for he great comments,

Bob

[John Hunter]

The problems of individual incentives seem to far outweigh any potential benefit. Deming was against them decades ago (along with short term quarterly profit focus) and I agree. Scholtes and Kohn (among others) do a good job of explaining why it is a bad idea. Douglas McGregor was right, managers need to eliminate de-motivators of employees not try to find better carrot dangling schemes to somehow make the carrot dangling incentive produce the desired behavior.

[Brennan]

It's hard coming up with an incentive structure that promotes the long-term goals of a company. This post reminds me a lot of Levitt's Freakanomics.

I think this is a "the medium is the message" problem. Where everything your company does should be congruent with its long-term goals.

[rick]

But in the case above incenting the right behavior would be easy. First, though, you need to decide what that behavior is. In the case of WaMu it could have been a loan amount where the portfolio had certain performance characteristics. And instead of looking at how many loans were written per month, look at it over a year. Basically... what loan policies would have meant that the bank was building a stable, growing business? Once you have those, reward them. But you simply cannot build incentives for good behavior when executives want bad behavior.

Of course, we can ask what was incenting execs to want bad behavior - and in most cases it's short term rewards. Is the stock price up this quarter? Bonus!! Is it off a penny? Your stock might drop 10, 15 or 20%. But there are no incentives to build a consistently profitable, stable company. Too often, the Street rewards current quarter performance with no eye to the medium or long term, executive compensation is structured around that and... if the exec is fired? Here's a few million on the way out. If I'm only going to be there 4-7 years and have the choice between doing things that might mak me millions and others that won't...