A new study of 13 federal organizations exposes what board effectiveness scorecards are not measuring, and what that costs when conditions deteriorate.
By Tim Glowa
Most of what board effectiveness frameworks measure has nothing to do with whether the board can do its job when it counts.
Composition. Attendance. Committee structure. Independence ratios. Evaluation processes. These are the things that show up on board scorecards. They are also, almost without exception, compliance metrics. They tell you whether the board looks like a well-functioning board. They tell you nothing about whether it actually is one.
The test of a board is not how it performs in calm conditions. It is how it performs when conditions deteriorate. And the standard frameworks have no way to measure that.
What 13 Federal Organizations Reveal About Board Effectiveness Under Stress
A new analysis of four cycles of Canada’s Public Service Employee Survey, 2019 through 2024, examined 13 federal organizations matched on mandate and workforce type. Six are governed by boards or governing councils that sit between the executive and the responsible minister. Seven are minister-direct departments with no board in between. (For the narrative treatment of the CIHR case, see When the Chair Sits Empty on Smart Board Governance).
Before COVID, the two groups produced identical results on employee ratings of senior leadership. After the pandemic, a four-point gap opened on a 100-point scale and held. Board-governed organizations lost less ground than minister-direct departments, and the difference survived every robustness check the analysis could throw at it. The buffering effect was invisible in calm conditions and decisive under pressure.
That is the finding. Boards do something during stress that you cannot detect during calm.
When the Buffering Function Fails: A 15-Month Vacancy at CIHR
Inside the board-governed group, the variation tells the rest of the story. Two of the three federal research granting councils, NSERC and SSHRC, kept their executives in place through the pandemic and produced the strongest resilience numbers in the sample. The third, the Canadian Institutes of Health Research, lost its president in September 2023, ran without permanent leadership for fifteen months including the entire 2024 survey window, and produced the worst.
Same governance form. Same statutory independence. Same insulation from direct ministerial control. Different leadership-continuity outcome.
At CIHR, ratings of senior leadership collapsed by 17.8 points on a 100-point scale during the vacancy. Ratings of immediate supervisors barely moved. That asymmetry, where leadership questions move dramatically while supervisor questions stay flat, is the empirical fingerprint of a leadership vacuum. It is also something that no standard board effectiveness framework would have caught in advance.
Why Standard Board Scorecards Cannot Detect Resilience Risk
The board effectiveness frameworks in widest use were built for TSX-listed companies operating under capital market discipline. They emphasize composition and process: enough independent directors, the right committees, attendance records, term limits, evaluation cadence.
None of those things measure stress performance. They measure whether the board is configured the way a well-functioning board is supposed to be configured. They are necessary. They are not sufficient.
What is missing is any instrumentation of the buffering function. How does the board actually engage with senior leadership? Is succession depth real or theoretical? When did directors last interact with the executive layer below the CEO? How quickly could the organization fill an unexpected CEO vacancy? These are testable in calm conditions. The standard frameworks do not test them.
The gap matters most for organizations that do not look like TSX-listed companies. Crown corporations, health authorities, post-secondary boards, regulated utilities, federal and provincial agencies, indigenous governance bodies, municipal governments. The frameworks were built for one set of conditions and applied to another. The mismatch is invisible in calm periods. It is decisive under stress.
Three Layers of Board Effectiveness That Standard Frameworks Miss
The CIHR pattern points to what predictive board effectiveness measurement should look like, and the federal data show why each layer matters.
First, baseline assessment of how employees actually perceive senior leadership, segmented by tenure, function, and geography. The standard frameworks do not collect this. The federal data show that this perception is the leading indicator of trouble, not a lagging report on it.
Second, a stress-test review that identifies where the buffering capacity is thin before conditions deteriorate. Succession depth, executive-team relationships, board institutional memory, the speed at which the board can act when a transition arrives. These are knowable in advance. They are not measured.
Third, board development that builds the specific capabilities that matter during turbulence. Signal interpretation, communication cadence with the executive layer, the board’s own visibility to the workforce. None of this is in the standard framework. All of it is buildable.
What This Changes for How You Buy Board Effectiveness Research
For board chairs, CEOs, and the people who commission board effectiveness research on their behalf, the implication is sharp. If the methodology you are paying for cannot detect the difference between a board that will hold under stress and one that will not, you are buying a compliance report dressed up as governance analysis. The next crisis will not test the report. It will test the board.
This research is the basis for a board effectiveness diagnostic built specifically for institutions where the standard frameworks do not fit. The work splits two ways. If you need the governance diagnostic that measures stress performance directly, that is at Governance Research. If you need the workforce analytics that turn your existing survey into a leading indicator, that is at Analytics. Most boards facing the next decade need both. None will have time to build either when the next crisis arrives.