The Risks of Rewards
No easy formula for chemist compensation
By Derek B. Lowe
You’ve probably heard the saying, “Anything that can be measured will be managed.” And even if you haven’t, it doesn’t much experience to realize that it’s true. The problem is, of course, that since people seem to have a deep psychological need to manage things, the corresponding desire to measure things will gradually get out of hand. Trying to come up with metrics for things like creativity or ability to collaborate is a job that has defeated many penetrating minds (and the minds of many HR consultants, too, which can be a different category). A look at your company’s performance review paperwork will probably be all the evidence that you need.
But there’s a flip side to that saying. Managers want to encourage some actions (very much) and discourage others (perhaps even more!). We’ll stick with the carrot side of that process, leaving the stick for another day, so here comes that flip side: “If you reward a behavior, you’re going to get it.” The problem is, you might not be rewarding what you think you’re rewarding. In a drug discovery and development operation, you’re going to want to think carefully about what you’re actually asking for versus what you’re really getting.
A perfect example is what happened some years ago during the early days of combinatorial chemistry and parallel synthesis. Now, generally, drug discovery chemists at the bench are evaluated partly by the number of compounds that they produce. (Generally, they’re also assured that they’re not evaluated this way, but that’s another story in itself). That’s one of the few quantities that can be easily measured for a bench scientist, so there’s no way to keep managers away from it. So when some early adopters starting using parallel reaction workups, solid phase reagents, and the other tools of the library-making trade, their compound submission numbers naturally shot up beyond those of their peers. Well, once they figured out how to use those tools, they did, and in the early days, “figuring out how to use them” often meant, “figuring out how to make a whole bunch of amides and sulfonamides.” So in came the piles of compounds, and — if the managers involved weren’t careful — out came the rewards for the people who dumped them there.
The result, naturally, was a further flood of the easiest analogs that everyone could think of. For some projects, that worked out fine, but for a lot of others, well, the hard chemistry was the only thing that could get them through. Combichem always had a hard time delivering on the hard chemistry, though, and the people who were just cranking out things for the numbers definitely didn’t bother extending the field’s range. They were too busy flooding the compound repository with adducts of every acid chloride that they could buy from the catalog. (Or every cheap primary and secondary amine, naturally, if they had an acid to start with. Lazy chemists are nothing if not versatile, in their own special way).
But if you don’t want to reward sheer numbers (and you don’t), then what do you want to reward? How about just rewarding numbers of active compounds? That idea has tempted many managers over the years, but most of them have thought about it and — if they’re honest with themselves — realized that that it’s too much like rewarding people for being lucky. Think about it: if we knew how to concentrate on making active compounds, why would we make so many inactive ones? This may sound obvious, but I worked for a company at one point that floated the idea of cutting everyone’s pay, then having them earn their way back up to what they used to make by meeting their research goals that they filled out at the beginning of the year. The philosophy behind that brainwave wasn’t much different than, “Why don’t you just spend more time on the stuff that’s going to work?” (That experiment didn’t last very long, in case you’re wondering, fortunately for everyone involved).
No, while it’s fine to give a small boost to people for making important compounds, if you make it a big part of the recognition process you’ll probably lower morale in the end. Everyone knows that too many of the best ideas for compounds come out of the blue, and if people see the reward process as being too arbitrary, they’ll tune it out (at best) or be infuriated by it (at worst). Most projects, too, have areas off the main hot series of actives that have to be filled in, for patent reasons if for nothing else, and you’ll have trouble getting people to do that grunt work if they feel that it’s killing their performance reviews to boot.
Moving up to the project manager level, you’d certainly want to reward people for getting across the finish line on time — even more if they can finish off a project early. But you have to look out for the people who will tell you that they’re always just about there, just a couple more bits of data and they’ll be done. That’s a great way to grab hold of resources in some companies, to be seen as the project that’s closest to delivering. But if the first team in line really isn’t the one that’s most likely to deliver in the near term, it can not only take up room that other projects need, it can end up hurting its own chances. The people on it may be getting stuck in a permanent short-term mentality, always trying to push things over the goal line with no time to look at the issues that require planning beyond the next three or four weeks. (The other problem with disproportionate rewards for finishing is that you might be encouraging people to keep even doomed projects alive as long as possible, granted always that “doomed” is a relative term in most research organizations.)
These issues get into what economists call “moral hazard” territory. That’s when you end up with perverse incentives to do the wrong thing, because it’s easier and more rewarding than doing the right one. That’s a good test to use when you’re setting up a reward system, but it’ll show you just how hard it is to do well. Coming up with goals that can be quantified but at the same time can’t be gamed or short-cutted by the unscrupulous is not easy — people can be alarmingly ingenious.
Of course, all this extends way beyond the labs. At higher managerial levels, many companies have been surprised to learn just what they’ve been getting and how it’s been delivered. The culture of, “You make your numbers or I’ll find someone who can,” is a steaming pit of moral hazard. You might want to extend that to some other well-known situations: “Surely this company can earn a better return on its cash by buying somebody rather than sitting on the stuff,” is one that’s gotten more than one outfit into trouble, and you can probably add several others. That will be left, as they say, as an exercise for the reader — but it’s something to think about next time your company releases its quarterly earnings forecasts. Isn’t it?
