« Cafe Classic: Reward Design is a Balancing Act, Baby! | Main | The Wannabe Business Partner »



Feed You can follow this conversation by subscribing to the comment feed for this post.

Thanks for veering into difficult and dangerous ground, Dan, to discuss those supposed "insights." The initial points you offered seem more theoretical contentions than indisputable facts. Various hypotheses about the uncertainty of performance management techniques need verification via the legitimate scientific method before being universally accepted.

Alas, poor implementation of fragile concepts applied outside their potency area will always produce unpredictable or negative results. Water can extinguish flames but only exacerbates oil fires. Some cures kill, under the wrong circumstances. Conclusions should derive from the circumstances. Assertions should be qualified judgments. Overgeneralization is hazardous, especially when based on controlled student behaviors or other foreign contexts.

In addition, shareholders/owners find it much easier to justify sharing their wealth when the potential beneficiaries have skin in the game. No risk, little reward. Mutually beneficial conditional guarantees make more sense.


Great post, but why is it "just too hard to have enormous swings in salary each year"? Would it really be so difficult to introduce a partially "variable salary" program at the executive tier?

Also, if "paying for results are difficult to predict and model", doesn't this fact negate the causal assumption underlying the pay-for-performance at the executive level -- that company performance can be largely attributed to the actions/decisions of the executives, rather than exogenous factors?

Can't have it both ways.

Thanks for the comments Ted.

I think the difficulty in having a variable salary is mostly in execution and somewhat in communication. As a profession we have spent the past three decades or more focusing on "market data" to defend pay levels and practices. With a strong base like a fixed salary (base pay) survey data would be so variable that it would likely no longer be useful. Of course, that may abetter result than current practices, but it is something that is truly unlikely to happen with so many professionals (and service providers) who have learned the current paradigm.

And, as soon as you have a fully or partially variable salary, you essentially are just creating a new angle on Short Term Incentives. Since shareholders, and most companies, would like their executives to focus on "Long-Term" (even though most could argue three years is, at best, mid-term), the thought of aligning pay only to short periods would never gain traction. This is likely why the authors of the original article focused on simply making fixed pay (Salary) consistently higher. Variability defeats their conclusions that pay for performance doesn't work.

As for paying for results Vs paying for actions and decisions that should align with success....
Paying for results, when those results are often driven by outside factors, seldom works in the long run (well, maybe in the very very long run).

Paying for the actions and decisions that should drive results generally makes more sense. If your executives are doing what you want them to do and the market, or just their industry simultaneously experiences hard times then your company should be outperforming its peers or better positioned to recover after the downturn. Simply cutting their pay because the results were poor doesn't make much sense.

Of course, determining the correct metrics and goals for driving success is more difficult (and sometimes more contentious) than simply statins that "TSR must be above X" or "EBITDA must be above Y".

Sort of like a sprinter. It is important to win the race. It is also important to set world records. But, if there is a tremendous tailwind, you may run incredibly fast and win the race, but the WR wil be disallowed. If there is a tremendous headwind, you may win the race with a very slow time, but it may still be impressive. Lastly, someone else may simply be faster than anyone in history and second place may prove to be an incredible achievement.

Many moving parts, Some are in the executives control, others are not. Pay for Performance only works if there is balance. Too often we let executives get away with "world record" pay when the tailwind was factor in their success. Too often we blame executives for terrible performance, when the headwinds were unusually strong.

It ain't easy (as we all know.) Until the entire world agrees on a pay scale and measurement system for executives and their success, we will need to keep working on the ongoing experiment that we call "pay for performance.

The comments to this entry are closed.