Playing To Win

The Impossibility Theorem

When There is No Overlap Between Required Capability and Willing Candidate

Roger Martin
8 min readJan 31, 2022


Source: Roger L. Martin

Work with a client last week reminded me that I have been meaning to write a piece on decision rights allocation, a key part of the Enabling Management Systems (EMS) box of the Strategy Choice Cascade. So, this week’s Playing to Win/Practitioner Insights (PTW/PI), my 14th Year II piece, is on The Impossibility Theorem: When There is No Overlap Between Required Capability and Willing Candidate. You can find all previous PTW/PI here.

The Importance of Decision Rights

While I argue that in Playing to Win strategy, the heart is a paired Where-to-Play/How-to-Win choice, I also argue that the heart means nothing unless it passes the reality test, by which I mean you identify and put in place a set of Enabling Management Systems (EMS), which build the Must-Have Capabilities (MHC) that enable you to win, where you have chosen to play, to meet your Winning Aspiration. To build and maintain those MHC, your EMS need, in part, to allocate decision rights to people who are indeed qualified. Capabilities don’t build themselves; empowered and qualified people build them. Hence, allocation of decision rights to the people tasked with building your MHC is an important task. Who needs to do what, in coordination with whom, in what structure?

As with most important things in business, there are lots of frameworks around to help leaders think about allocation of decision rights. The RACI model has been around for many decades. It lines up decision rights across four categories: who is responsible, who is accountable, who must be consulted, who needs to be informed. The strategy houses each seem to have one. Boston Consulting Group has OVIS — own, veto, influence, support. Bain & Company has RAPID — recommend, agree, perform, input, decide. Presumably, McKinsey has one, but I couldn’t quickly find it. I have a hard time arguing that one is superior to another. Each helps the user think through the task of figuring who does what with/to whom.

But I would argue that the blind spot of each framework is that it implicitly assumes that a capable individual occupies the chair to which each decision right is assigned. I.e., there is a capable recommender in the chair assigned to the recommender task, and a capable decider in the decider chair, etc. It is assumed, again I think largely implicitly, that there will be some other process, typically performed by other people (e.g. the talent acquisition people within the HR function), that will ensure that there is an appropriate person in each chair to which decision rights are assigned. If there isn’t currently a person in a specified chair, the decision rights allocation process generates the basis of a job description for recruiting for the position.

The Schism

However, I observe a frequent schism between the system for assigning decision rights to chairs, and the system for getting a capable person into each chair. And I was reminded of that last week at a client with whom I was doing strategy work and at the same time they had a consulting firm doing operating structure work that involved a major rethink of the decision rights allocation. As part of the new structure, they created an important new position (repeated across each major geography) — which at one level made a lot of sense in that it filled in a decision-making gap that was currently problematic. But when I looked closer, it was clear that the position required person in the chair to be highly adept and capable of framing problems and coming up with clever recommendations — but to defer to someone else to make almost all the important decisions on those issues. And I realized that the decision rights allocation appeared to make sense, it violated the Impossibility Theorem.

The Impossibility Theorem

You have a case of the Impossibility Theorem in action when decision rights are assigned in a way that makes it impossible to fill the chair with a person who is capable of carrying out those decision rights as specified and assumed. That is to say, there is no overlap between the universe of people who are capable of fulfilling the requirements of the role and the universe of people who would accept the job of fulfilling the role.

Let me provide a real and very important example. After the meltdown in 2000–2001, The Financial Accounting and Standards Board (FASB) came to the determination that many investors were fooled by wildly overstated intangible assets and goodwill on the balance sheets of companies that flamed out in the meltdown. So, it promulgated FASB 142 which completely changed the treatment of intangible assets/goodwill. Rather than being placed on the balance sheet at cost and amortized over a fixed schedule, these assets would no longer be amortized at all. Rather, each firm’s audit provider would declare, as part of its annual audit, whether or not the intangible assets/goodwill were impaired. If impaired, the auditor would force the company to write them down to their impaired value — which could be as low as zero. FASB’s fundamental theory was that more accurately stated assets on the balance sheet would drive more sensible/realistic stock valuations.

The theory sounds good: society would benefit if balance sheets were more accurately and correctly stated guiding more realistic and sensible valuations. But it begs the question, to what extent is the audit partner performing the impairment assessment (typically from one of the Big Four global accounting firms) actually capable of making a perceptive determination of impairment or lack thereof? It is not a straightforward task. The auditor needs to be able to forecast accurately the cashflows that will accrue specifically to this particular asset over all future years and then discount those cash flows back to the present at the proper risk-adjusted rate, and then compare that number with the number on the balance sheet to declare definitively on the impairment question.

It turns out that even though the audit partners of the Big Four are smart, talented, dedicated, and hardworking, there is nothing in their experience that renders them capable at all performing this task. There are people who demonstrably are capable. Warren Buffett is an example. He has proven himself highly capable of estimating the future value of an asset and has made a large fortune for himself and the shareholders of Berkshire Hathaway by comparing his calculated value to generally prevailing view and buying if his is appreciably higher. John Templeton was good at it before he passed away. Various private equity fund principals are good at it.

But one thing is certain: if you are demonstrably capable of this task, you won’t be working as an audit partner earning a tiny fraction of the remuneration of a talented principal investor. Hence, there is zero overlap of the universe of people who are capable of assessing impairment of intangible assets/goodwill with the universe of people who are willing to judge impairment for a living as an audit partner. There are a small but non-trivial number of the former and thousands of the latter — they just aren’t the same people, at all.

The same holds for stockbrokers. The only thing you know about those who make a living earning commissions from recommending stock buys and sells to clients is that if they were any good (i.e., right more the 50% of the time and therefore more useful than the investor flipping a coin), they would be doing something else. They would be investing on their own behalf and becoming wealthier by far than by being a stockbroker.

The same holds for bond raters. The only thing you know about those who make a living earning a salary from rating bonds for a bond rating agency is that if they were actually good at rating bonds, they would be working at or running a bond fund.

The problem with running afoul of the Impossibility Theorem is that bad things happen when you do. You count on people to do a job that they simply are incapable of doing. But they are most certainly not going to tell you that, because you will fire them if they are honest about their incapacity. We counted on bond raters but in the lead up to the Global Financial Crisis (GFC), they were rating tranches of derivatives as AAA that should have been rating DWH (Disaster Waiting to Happen). By doing so, they contributed greatly to the severity of the GFC.

And the current impairment measurements are way worse than prior to the arrival of FASB 142 in 2001. I was on the audit committee of a large NYSE-listed company at which the auditors insisted on taking an impairment charge on the goodwill that arose from an acquisition that we had made several years earlier. Ever the business nerd, I asked to see their methodology and they happily sent me a detailed doccument their firm’s impairment calculation methodology. You could have knocked me over with a feather. It was so logically bankrupt and nonsensical that I could barely believe it.

The methodology involved comparing the book value of the overall company with the market value of the overall company. If market value was lower than book value, then an impairment charge was mandated to bring book value in line with market value. Recall that FASB wanted a fairer statement of book value to bring market value into line. The Big Four audit firm did exactly the opposite —it used market value to drive down book value. Essentially it punted on making a determination by essentially saying that if investors in general think (implicitly of course) that assets are impaired, they must be impaired. Stunning.

I would argue that transparency and clarity of public financial statements has taken a big hit thanks to FASB violating the Impossibility Theorem.

Practitioner Insights

When assigning responsibility for capabilities that are important to your organization, remember that it is a two-part exercise. One part is to figure out how to divvy up the decision rights. The other part is to determine whether there are people out in the world that are capable of doing the job so-designed and that are also willing to take the job so-designed. If you don’t pay attention to both parts, you will set yourself up for disappointing surprises.

Fortunately, with the client above, we caught the looming Impossibility Theorem violation in advance and redesigned the job to be doable for people wanting the job. Unfortunately for investors in the company with the impairment charge, application of FASB 142 resulted in a balance sheet that was less accurate and realistic that it would have been without the “enhancement.” And that is entirely FASB’s fault for violating the Impossibility Theorem.



Roger Martin

Professor Roger Martin is a writer, strategy advisor and in 2017 was named the #1 management thinker in world. He is also former Dean of the Rotman School.