Attract, retain and demotivate

As I've blogged much in the past, I think the effectiveness of performance-related pay should be properly reviewed by shareholders and others involved in the executive remuneration debate. In particular I think the question of motivation needs further analysis. In passing, I was pleased to note that in the Hermes discussion doc on exec pay there was a brief nod to this

They should consider how best to motivate the executive directors and top management, using other tools at their disposal as well as those provided by pay. We note that there are many studies, from Maslow onwards, that demonstrate that pay for performance is not always the best motivator, particularly for those in senior positions and we expect boards to ensure that they take account of this.

OK, it's brief & basic, but at least a mainstream asset manager has noted that there is a debate starting here.

In fact, I think when you dig into this stuff it becomes fairly obvious that both current and planned practice in exec pay doesn't sit well with different perspectives on motivation. As I've said before, in essence performance-related pay appears rooted in a behaviourist view of human psychology - the consequences of behaviour affect whether that behaviour will be repeated. So, from a behaviourist viewpoint, bonuses, share awards etc are a positive reinforcer. And by positively reinforcing we expect more of the desired behaviour to occur. Do X and you'll get Y.

But why do behaviourists, or those that think like them, think positive reinforcement is required, and how do they think it can be achieved most effectively? Here's Aubrey Daniels in Performance Management (all about the use of positive reinforcement) on why:

Four basic reasons explain why people don't do what we want them to do on the job: 1) they don't know what to do; 2) they don't know how to do it; 3) obstacles in the environment discourage or prevent them from doing it; or 4) they don't want to do it.

How much of this applies to board directors? Whilst I might agree that, actually, board directors don't know what to do or how to to do it (because in very large organisations it's sometimes really hard to see which combination of factors works) I think that they think they know what to do and how to do it. It is to be hoped that, having got to the board, they do want to do the job. So it's not immediately obvious to me that positive reinforcement is needed.

But even if we concede that it is, what's the best way to do it? Aubrey Daniels is interesting here, as he stresses the need for 'consequences' to follow the thing we want to see happen:

Behaviour is affected by the consequences which follow it. Consequences either strengthen or weaken behaviour. Because a consequence affects the behaviour that precedes it, it is critical that the consequences be timed to follow the behaviour we want to be affected.

There's not much point praising, or rewarding, someone months after the behaviour you want to reinforce has occured. It needs to take place soon after to have the desired reinforcing effect.

I apologise to people with a background in psychology, for whom this will be behaviourism 101, but it's an important point. In recent efforts to 'reform' executive pay, much emphasis has been put on the need for reward to take account of long-term performance, and, increasingly, to be paid out only over the long term too. This is all very laudable on one level, since we trying to avoid a) incentivising short-termist decision-making and b) paying out rewards on the basis of performance that turns out to have been illusory. But - if you take behaviourism seriously - you have to doubt that reward of this nature will have an effective positive reinforcement effect. In fact exactly this point was made (arguably for different reasons!) by PwC last year.

Maybe you think this is still ok, after all the payouts from incentive schemes are rewards for hard work, it's a fair reward more than anything. That's a valid point, even if I don't necessarily agree. But then let's accept that the idea that long-term incentive schemes (as in those that pay out over the long term) aren't particularly 'motivating'. Maybe they 'retain', but not motivate.

Remember too that this is looking at things from a behaviourist perspective (which I believe is implicit, though unacknowledged by most corp gov people, in remuneration design as it stands). There are, as I've bored on at length about, a whole host of other reasons why we might not think it's a good idea. I think what we need to tease out here are the assumptions people are making about the interactions between incentives, behaviour, motivation and performance. These overlap but are not the same thing, which is why there is sloppy thinking in this area. Incentives might get more of a certain behaviour, but not extra motivation or better performance. (No matter how much you pay me I'm never going to be a great dancer, or want to do it!) This is why performance pay may only be really effective at uninteresting, straightforward and measuraible tasks.

If we can make explicit what we think about performance-related pay, we may find that we don't feel it does the job (or well enough to be such a huge part of directors' reward) and therefore will want to scale it back. That in turn can feed the desire on the part of almost everyone in the executive pay debate for simplification.