START WITH WHY
Monetary compensation for work has been a key component of any people system for quite some time now, ever since the vast majority of people realized that slavery is not the best of ideas…
It had gradually evolved in structure and purpose: from a purely mechanistic structure using time-based components (hourly wages) to incorporating more humanistic components such as loyalty (tenure) and individual contributions (pay-for-performance); and from a pure enabler of sustenance, to an enabler of more refined drivers of motivation aimed at accomplishing higher needs in Maslow’s hierarchy. Monetary compensation is a powerful tool in an organization’s toolbox for improving the organization’s ability to make progress on its mission. In this piece I’d like to make the case for why many so-called progressive compensation systems end up missing that mark, and propose an alternative approach for designing more effective compensation systems.
MEET “INCENTIVE PAY”- “PAY FOR PERFORMANCE”’S EVIL TWIN
There are very good intentions behind the “pay for performance” idea: since individual performance in a given role can vary significantly, this variation should also be reflected in the compensation system. However, in many progressive compensation systems, this idea has devolved into a more simplistic notion called “incentive pay” — a “sticks and carrots”-based approach for improving motivation by offering a monetary reward in return for completing a (short-term) objective such as closing a deal, finishing a project, hitting some KPI or more broadly “crushing it” in a certain role. Unfortunately, research has a few things to teach us about “incentive pay”’s ineffectiveness (at best):
“The failure of any given incentive program is due less to a glitch in that program than to the inadequacy of the psychological assumptions that ground all such plans. […] Do rewards work? The answer depends on what we mean by “work.” Research suggests that, by and large, rewards succeed at securing one thing only: temporary compliance. When it comes to producing lasting change in attitudes and behavior, however, rewards, like punishment, are strikingly ineffective” — Why Incentive Systems Cannot Work, Alfie Kohn
External incentives were found to be ineffective in driving long-term behavior change. In some cases they were even found to diminish productivity. They tend to crowd-out intrinsic motivation, crush creativity, foster short-term thinking, and become addictive. If true behavior change is what we seek — external incentives are not the answer.
Increases tolerance of low performance:
“In a firm which makes extensive use of individual performance rewards, it is often the case that the firm’s reaction to a mediocre or average performer is to say “That’s ok, you can stay — we’ll just pay you less.” This is hardly a recipe for excellence! […]Firms with individual performance-based reward systems often end up tolerating wide varieties of performance” — True Professionalism, David Maister
Improving the part doesn’t improve the whole:
“If you give a manager a numerical target, he’ll make it even if he has to destroy the company in the process.” -W. Edwards Deming.
A system purely based on individual performance lead to attempts to game the system, take shortcuts, and more troubling unethical behavior. It erodes relationships and reduces collaboration. People will sometimes throw this peers under the bus (only figuratively, let’s hope) just to meet their personal objectives.
Fosters a critical attribution error: “no man is an island” cannot be truer in our modern work environment. If you’re a door-to-door salesman, selling cleaning supplies out of a catalog, perhaps you can argue that your contribution to the business can be fully isolated and separated from the contribution of your peers. Therefore, some incentive scheme based on your contribution can be considered as “fair”. But can you make the same argument if you’re selling software to enterprise clients? When each deal is unique and the sales process typically involves more than a handful of people directly and dozens indirectly? Can a successful sale truly be attributed to one person and not the rest? Would the contribution of each of them be identical in each of the deals they contributed to?
FAIRNESS IS THE NAME OF THE GAME
Hopefully by now we are in agreement that pay-for-performance, as implemented in most progressive compensation systems has a significant, undesirable adverse effect. So how can we take a step forward without having to take two steps back in designing an alternative system?
We first need to take into account that, yes, individual performance varies, and the compensation system should reflect that. But we also need to take into account that, as discussed in part 1, the work shifts from shallow (algorithmic) to deep (heuristic), intrinsic motivation is the only effective motivation, so external incentives and “sticks and carrots” approaches are simply ineffective motivators.
We cannot ignore the fact that compensation will have some impact on motivation.
But in order to ensure that is has the desired effect, we need to stop thinking about compensation as a direct driver of motivation, and start thinking about it as an indirect driver of intrinsic motivation, through the values that are reflected in the compensation system itself. To use the framework we introduced in part 1, compensation should be thought of as a lever for driving Behavioral Cohesion rather than a lever for driving Intellectual Alignment. First among those is fairness. When we feel that we are compensated unfairly, our motivation takes a big hit. Let’s unpack what this means.
A BETTER COMPENSATION SYSTEM
Fairness is a relative term, as it is a determination coming out of a comparison of two or more things. In the context of a compensation system, we can think about fairness across 4 different dimensions / areas of comparison:
Organizations are open systems and people have the freedom to choose to leave one organization and join the other. There is a market for talent, through a rather inefficient one, so prices get distorted. This is not just a problem of asymmetric information. Even if we had a perfect data set of everyone’s compensation information, making a fair comparison will be challenging since neither people nor companies are commodities, so it’ll never be an apples-to-apples comparison, both literally and figuratively. And we haven’t even talked about compensation components whose value cannot be determined with certainty like stock options…
All of this is not to say that market prices should be ignored. The absolute prices should be viewed as important but very loose guardrails. Whichever proxy was chosen for the market price, it’ll be hard to argue that even a ±25% variation is unfair since the measurement error is probably greater than that.
Market prices play a more critical role in the fairness realm when it comes trends. If market prices are changing year-over-year for a certain role, the compensation system should reflect that trend as well.
2. Personal needs and preferences
Different people have different needs and different personal preferences.The extent to which those are reflected in the compensation system is another dimension of fairness, though a very subjective, cultural-dependent one. This is not as an outlandish an idea as it may sound, even in a capitalistic society: our tax code provides incentives for marriage and childbearing; extended paid family leave is becoming mandatory in more and more states; more companies given their employees some control over their base/upside compensation mix to reflect their personal risk preferences; cost-of-living adjustments are becoming a more fairness-driven debate as companies become more global.
3. Contribution/value to the company — This is the dimension with the most lowest hanging fruits w/r/t fairness, if we can replace “incentive pay” with a better solution. The key attributes of such solution are as following:
- Focus on the contribution to the long term success of the company. This goes beyond the short term performance of the individual to also include other supporting behaviors such as driving behavioral cohesion (values fit) or helping others improve their performance. Loyalty, which is often implicitly manifested in the compensation system (paid time off is a good example) should we be evaluated through this lens and either be reflect proactively or left out.
- Quantify the contribution through a set of levels, manifested through a set of ladders for each material function or group of roles. Ladders are likely to differ from one another on the attributes that define performance, due to the varying natures of different roles, but should share the same attributes on the other dimensions of contribution.
- Loosely couple levels with role (some degree of movement up the ladder is possible without a movement in the organogram), and fully decouple levels from management (management = different ladder). This breaks the unhealthy tight coupling between an increase in compensation and movement up the org.
- Consistent “above level” contribution merits a promotion to the next level and an increase in compensation.
This is the flip side of contribution/value to the company. When the company is doing well / succeeds — employees compensation should increase.
WHEN SHOULD COMP CHANGE?
Finally, looking at a compensation system as a driver of fairness, not only gives us good guidance on how compensation should be set, it also, and perhaps more importantly, gives us guidance on dynamics of compensation.
Compensation should change, if and only if:
- Market prices change (or:)
- Individual needs change (or:)
- Career ladders or levels change (or:)
- Company performance changes
I’m going to add here answers to some relevant questions I’ve been asked on this topic.
What about “salary bands”? How do they fit in?
The short answer is: they don’t. Let’s keep in mind how “salary bands” came to be. They were a pressure valve in a world in which base compensation:
- Was not responsive to changes in the labor market
- Was tightly coupled to role
In a world where this is no longer the case, and compensation changes due to the 4 drivers listed above, it’s unclear that they still have a role to play.
Furthermore, often times they were used as a tool to implicitly reward loyalty/tenure: “this person has been in this role for a while and have been doing a good job, so I can give them a bump as long as they are “within band””. But if you value loyalty/tenure — why not make it an explicit component of your compensation formula rather than have it be a part of an opaque, subjective “band”?
This candidate that we want to close demands a significant higher comp than the one we can offer her at this level. What do we do?
The short answer is: we don’t hire her. The more nuanced answer is we should ask ourselves two important questions first:
- Is there reason to believe that there was a significant surge in the market and we should adjust all of our salaries upwards to match it? If the answer is no, then we shouldn’t hire her.
- Are we leveling her correctly? Can we make a strong case for why we have good reason to expect that she will be able to contribute more to the business and therefore we should higher her into a higher level with a resulting higher comp? If the answer is no, then we shouldn’t hire her.
If we reflect back on the 4 drivers that our compensation system consists of, it’s easy to see how different companies would value (and therefore compensate) the same person differently. A company that’s already in the “printing money” stage should offer high comp (equity included). That doesn’t mean that you should match it.