The Future of Pay for Performance: Can the Membership Model Work For Us?

In their session on the next generation of pay for performance at last month's WorldatWork Total Rewards conference, Mercer consultants Brian Levine PhD and Haig Nalbantian presented the case that companies have overstepped in their move to variable pay during the downturn, as they have worked to leverage limited resources to the greatest productive advantage, and that - given employees' strong preference for base pay as a reward - we should return to more salary-centric model of compensation.  With this, they proposed two "alternative" pay for performance models (which focus on base pay as the primary reward) for our consideration:

The “Tournament” Model
- Pay rates need only vary with hierarchical level, not current performance
- Pay rates in levels above motivate those below
- Relative performance evaluation determines who advances

The “Membership” Model
- Pay rates exceed market
- Partial monitoring, supervision is difficult
= Employment discontinued when employee found not to meet threshold performance

I haven't yet given great consideration to the tournament model; that may be the topic of a future post.  On the surface of it, though, it strikes me as an approach which relies on the existence of substantial organizational and pay level hierarchy to have any motivational power.  Many of today's flatter, leaner organizations may not have the number of striations and upward growth slots required to make this approach ultimately effective.

For this post, however, I wanted to take a closer look at the membership model.  Interestingly, although he doesn't call it by this name, this is the pay model that Dan Pink advocates when he advises us, rather than relying overly on incentives, to 1) Pay people enough to take the issue of money off the table, and 2) Pay people fairly.

The membership model, in the words of the Mercer team, can motivate if premium pay - as a substitute for direct supervision - is coupled with threat of termination.

Call me vision-impaired, but I get stuck when I try to envision above-market pay as a substitute for supervision.  Particularly because I suspect that little supervision also means little performance management, direction or coaching.   Does this approach rest, then, on the assumption that employees who are "purchased" at a higher salary level - by definition - innately understand the link between company strategy and priorities and their own role ... and can be turned loose to collaborate and generate success on the required terms, without "interference" by leadership? 

To my way of thinking, the membership model must rest heavily on performance management - perhaps even more so than other approaches.  How else do we ensure that our high investment in talent delivers value for us?  And how else do we reliably, decisively, accurately and compassionately identify and terminate those individuals who aren't delivering on that investment?  After all, the threat of termination is the cornerstone of this approach, is it not?  Otherwise, don't you risk simply building an overpaid and entitled workforce?

In what I can't help seeing as a particularly ironic note, it often seems as though many of the calls for this model of compensation - the cries of "just pay them fairly!" crowd - come from the same group who insists that performance appraisals must be killed off.

Perhaps it ultimately hangs on what you see as the end goal.  The Mercer consultants reference employee turnover (or its reduction) as a key measure of reward plan success.  For sure, premium pay will likely reduce employee turnover.   I would argue that business results are at least as important a goal as reduced turnover; in the end, a successful and economically viable organization is in the best position to serve everyone's interests - owners, leaders and employees.

Help me here.  What's your take?  Does the membership model represent the best "next generation" of pay for performance?  Is it truly workable and, if so, for what kinds of organizations?  Where are we seeing it work effectively now and what have been the conditions for its success?

Image courtesy of aigasf.org

Comments

[Amy]

I look forward to hearing more about the "tournament" model, as this is a concept being requested by and researched for our IT division. As a non-profit hospital deeply set in the groove of providing ATB wage increases, we are far from pay for performance but look at it longingly from afar.

The IT proposal includes 5 levels of progressive career ladder (newbie, analyst, senior analyst, grand poobah . . . etc.) and each member is ranked based on 10 criteria, only one of which is job knowledge; many are behavior related. The ranking slots you into a pay range. There might be 4 or 5 pay ranges for each of the 5 levels in the career ladder.

I'm fascinated with the idea of doing something completely new, but fret over ranking people using criteria that isn't easily measureable and subject to interpretation.

Thanks for always keeping us noodling the possibilities.
Amy

[E James (Jim) Brennan]

I would characterize those two options as the Darwinian model and the Elite Team model. Of course, that is overly broad and general, but they do exhibit similar traits.

Darwinian = Survival of the fittest, internal rivalry contests, warfare in which you compete with your co-workers sometimes more than the outside competitors, bowb your buddy King of the Hill stuff, etc. Elite Team = better TREATMENT (the Total Reward concept) but not always the highest cash, exalted status for self-directed and self-monitoring fanatics, extremely high minimum standards enforced by the team itself or by intense direct feedback.

Neither approach is new or a panacea. Sales organizations tend to be hysterically Darwinian. Public entities and bureaucracies like universities can also be Darwinian, but in a slow-motion mode. Both elite uniformed forces and entrepreneurial start-up enterprises tend to feature the Membership model where being on the team is a reward which carries a high price willingly paid voluntarily rather than externally imposed.

Each scheme can be highly effective in the right places at the appropriate time under the proper circumstances. Also, they can backfire badly when used in the wrong situations or retained beyond their useful lifetime

[Haig Nalbantian&Brian Levine]

Good comments all around. Ms. Bares, we appreciate your interest in our 2011 WorldatWork presentation and the constructive dialogue you have started.

First of all, Mr. Brennan is correct. These are not new constructs and how well each model works is indeed very sensitive to the context in which it is introduced. There is a vast literature in labor economics that speaks to the models we discussed, including the tournament model and what we termed the “membership model” - what economists call the "efficiency" wage. These are well established constructs for which we claim no credit at all.

What is new in our presentation are the conclusions we draw about the context in which each of these alternatives is likely to work better than standard “pay for performance” models. Our conclusions are empirically based. Over almost twenty years, we have had the unique opportunity in work with client organizations to undertake rigorous statistical modeling of their workforce and performance data to estimate the impact of rewards programs and various compensation components on important workforce outcomes (e.g., retention, individual performance) and key business outcomes (e.g., workforce productivity, customer retention, revenue growth). Our assessments are based on a careful reading of the results of this work. And our entire theme about the pitfalls of variable pay has been provoked by a spate of recent findings in our client work showing serious, unintended negative consequences of bonus and other variable pay programs.

Regarding Ms. Bares’ concerns about the Membership model, paying above market is efficient precisely when it substitutes for supervision. Here’s the logic of the model: you hire capable people and give them a very strong incentive to self supervise. That incentive stems from the employee’s understanding that the premium pay or value proposition is just that – a premium, not readily available elsewhere. Lose the “cushy position” and the employee is unlikely to find a similar one; in the least, s/he will spend a long time searching for something equivalent. So, in this case, s/he is more likely to work hard, exert diligence, and deliver work in a way that makes the employer want to continue the relationship.

Of course, this only works if those caught “shirking” are actually let go. In a membership model, terms of membership are strictly enforced. There must be some probability that shirkers will be detected and a credible threat of termination for those who come up short. Does that require some performance management? For sure. But, all else being equal, the higher the premium, the lower the probability of detection can be to maintain employee incentives to perform.

We’ve seen the membership model work very well to attract, motivate and retain talent. But we’ve also seen the model collapse as organizations maintain their reward premia while effectively eliminating any credible threat of termination for those who don’t live up to the terms of membership. In addition, we've seen the membership model fail when companies institute premium rewards while maintaining high levels of direct supervision (e.g., low spans of control.) In effect, they end up paying twice for supervision. As one client remarked: "I see. We hire top talent who are thoroughly capable and motivated to direct themselves, and then we don't let them do it.”

In both these instances, you get precisely what Ms. Bares warns about: bloated cost structures without the benefits of higher productivity and leaner management.

Haig Nalbantian and Brian Levine
Mercer
7/7/11

[Ann Bares]

Haig and Brian:

Thanks so much for taking the time to add information and color to the discussion that your presentation prompted - very much appreciated!

I think you've also reinforced my performance management point. In the scenario you're describing, I would define performance management as the processes and mechanisms by which the terms of membership are communicated and maintained. This must indeed be the linchpin on which the success of the model relies.

[Haig Nalbantian]

Ann, your point is well taken. A membership model could not function in practice without employees clearly understanding the terms of membership and how compliance with them is evaluated.

We would distinguish this kind of performance management from "hands-on" supervision that aims to direct work effort and continuously monitor performance. In our modeling work, we will often capture this kind of "supervision intensity" through such measures as spans of control and/or hierarchical distance/proximity of employees from their supervisors. Neither of these has much to do with "performance management" as you characterize it. It is this idea of supervision intensity that is invoked in the economics literature on so-called "efficiency wages." A basic version of that model is detailed in Chapter 4 of a volume I edited, entitled, Incentives, Cooperation and Risk Sharing (Rowman & Littlefield Publishers). Hope the distinction makes sense to you.

Thanks again for launching this discussion.

Haig

[Ann Bares]

Thanks for the additional information and clarity, Haig - much appreciated!

[Sameer Panjwani]

Some great information from the article as well as the comments. Although I do have a question, when talking about this membership model, are we talking about paying more than what the market would pay or just higher than the employee's expectations and what we know he would consider as good pay. There are instances when we can easily afford to pay much more than what the employee has been getting but knowing fully well that it's not as high as he could get elsewhere. I foresee this problem because if there were 3 to 4 positions that needed to be filled at similar levels, then should all be paid the same level or just based on what they'd be happy with. Instances of comparing salaries among colleagues could cause resentment? But then over-paying some people by a huge margin could spoil them for the future?