How Do Incentives Really Work?

In many of the write-ups of the financial woes gripping the world right now, it's pretty clear that certain incentive programs were put into place over the past decade without any clear understanding of what their logical outcomes would be. For example, people in the U.S. were encouraged in many different ways to get into the housing market: through income tax breaks relating to their mortgage interest, through low interest rates in general as well as - in some cases - even lower start up rates (deferring the higher rates until later in the life of the loan), and through changes in the banking laws, intended to help lower income families get into their first homes. All of those forces acted at cross purposes to set up a scenario where lots of folks would have mortgages that they couldn't really afford, which was at least one of the catalysts for the larger financial mess that we're in. And no matter how harshly one thinks of the last couple of administrations in the U.S., no one in their right mind would suggest that what we see around us today was the goal of those actions!

So why is it so hard to get incentives to work out in the way that you intended? Part of it, I think, is that you need to have all of the players aligned with the same goal. Going back to the sub-prime lending fiasco, there was a clear disconnect between the objectives of those involved. Many of the financial companies doing the lending, for example, didn't care whether or not the recipients of the mortgages stayed in their homes (they'd already repackaged the debt and sold it off to others to worry about) as their focus, in the form of compensation, was on volume, not stability. The recipients of the loans wanted one of two things (or possibly both): a long-awaited entry into the ranks of home owners, and/or an investment that they believed would net quick profit. Finally, many of the original statutory changes were aimed, in one way or another, at getting citizens who had sub-standard credit rankings but were otherwise viable mortgage holders their own slice of the American Dream. There's very little alignment there. More than that: the notion of "quantity over quality" (on the part of the lenders) was often completely at odds with the goal of housing viable mortgage-holders. Hence the unfortunate results.

In a software organization, the stakes are usually a little less earth-shattering (thank goodness!). A typical case might be a yearly or quarterly bonus program in which the portion realized by each employee is dependent upon how they themselves, their team, their department or the company in general (or any combination, thereof) does against certain objectives. That sort of thing rarely causes the worst case outcome of clashing goals and carnage. Instead, the most common result seems to be a relatively benign disconnect, in that the average employee may feel that he or she has little influence over some of the broader metrics. In that case, the setup has simply been ineffective, rather than counter-effective (whew!).

Better programs and initiatives, however, actually manage to inspire results that might not have come about without their inclusion. While admittedly small potatoes, I've seen and heard of nice results coming from something as simple as establishing rewards for suggestions that end up saving the company money. It becomes less about the money itself - is anyone really that motivated by a $20 gift certificate anymore? - and more about the recognition, the feeling of accomplishment, and the confidence that comes from believing that you're operating within a system that values improvements. It can actually create a cultural paradigm shift, away from "Why bother complaining, since nothing ever comes of it?" and toward "If something's broken, let's fix it!" And that can make a world of difference, although it may happen so gradually as to slide under most peoples' radar.

I think a lot of incentive programs never come into being because there's a decision made, at some point along the way, that it'll either be too expensive, or too susceptible to gaming. The cost issue is always a thorny one, and perhaps the only way to deal with it is to budget a certain amount, broadcast that fact, and then see how it goes. If you end up running out of money before you run out of time (or ideas), then either it was a success and you should attempt to increase your budget for the next period, or you got very little value out of it and it's time to re-tool. Either way, you can adjust and go forward.

The gaming concern is a different problem to solve. First, I came to the conclusion years ago that there will always be someone in any group of ten or more (just to pick an arbitrary number) who will attempt to get something for nothing. And that's what gaming is, after all. The person who hears, "Our bonus is tied to how many test cases we write because management wants our test coverage to improve" and thinks, "I bet I can make more money if I make all of my test cases really trivial and small" is hoping to get something (a bigger bonus) for nothing (doing no work that will actually improve his company's situation). But the number of employees who react that way will, generally-speaking, be inversely proportional to the quality, relevance and clarity of the goal. Most people, given the proverbial "SMART objective" (specific, measurable, attainable, realistic and timely), will enthusiastically put their back into it and work toward it... especially when there's money on the line. Therefore, it seems to me that time is better spent by those in leadership establishing the goals, including the metrics and rewards, rather than worrying about what gaming may occur. Even if a few bad apples work against the spirit of the program, the vast majority will produce more than enough of the good stuff to overshadow them. And eventually peer pressure, as well as the sense of accomplishment coming off of those really doing the work, will transform the culture into something in which gaming isn't abided at any level.

Of course, as always, it's easier said than done. But that should never stop us from trying!