Mandate, legitimacy and decision rights · Lesson 9

Evaluating board work candidly

16 min read

Lesson objectives

  • Compare board self-assessment with deliberation and delivery evidence and select a traceable improvement.
Reading tools

Board members rate their meetings highly, yet half the decisions return unfinished every month. How should you interpret the contradiction? Members may appreciate respectful discussion and punctuality while papers, authorizations and follow-up remain weak. Define what is being evaluated before collapsing everything into one satisfaction score called governance effectiveness. This lesson compares member experience with decision records and selects an improvement that can be checked later. It separates the board's collective work from the chair's conduct or a particular committee's performance. That distinction supports cause-specific repair rather than polite reassurance or personal blame. The evaluation becomes useful when it produces an improvement decision and evidence of its effects, not merely when the questionnaire is complete.

Choose a bounded scope with observable evidence: decisions on seasonal contracts, a nominations committee or welfare-action follow-up. Ask what that work should achieve for the institution. For the whole board, purposes might include understanding alternatives, reviewing constraints and assigning an identifiable implementer. For a committee, they might be a defined examination and timely recommendation. Write one observable test for each purpose, such as alternatives in a paper, evidence of recusal or an implementation owner and deadline. These operational tests derive from organizational duties, not a questionnaire's prefabricated score. Begin with expected work, then choose questions that reveal whether it happened. A smaller evidence-rich review is often more useful than an ambitious survey unable to connect answers to decisions.

Use three evidence types deliberately. A survey captures members' perceptions, minutes review shows what the body recorded, and short interviews help explain gaps between them. Each has limits. A member may withhold criticism if anonymity seems fragile; discussion may be good while minutes are excessively brief; an interview may reflect only one person's experience. Connect evidence to one specific question rather than forcing agreement. If members say alternatives are clear, inspect actual papers and ask about contrary cases. This triangulation prevents favorable opinion being mistaken for implementation evidence, or a missing record automatically being interpreted as bad faith. Contradictory evidence can be the most useful starting point for understanding how board work actually occurs.

Ask about a specific behavior and period. 'Is the board excellent?' mixes feelings and standards. 'For the last three financial decisions, did you receive a comparison of two alternatives before the meeting?' is closer to observable events. Separate paper clarity, discussion management and freedom to challenge. Papers may be strong while the chair closes debate early, or discussion may be open while information is missing. Include an option for not attending or lacking information so knowledge is distinguishable from guessing. Decide interpretation before collecting responses. Treating 'unable to judge' as medium satisfaction inflates the indicator and hides an information gap. Each question should help identify a repair, not merely produce a pleasing average.

Select evidence according to the question and retain its denominator. Six minutes chosen merely because they are available cannot establish the whole year's compliance rate. To examine large-contract follow-up, begin with all decisions meeting the stated large-contract definition, then disclose selection or inspect all if there are few. Include troubled decisions rather than successful files alone. Record who selected and examined them, protecting personal details appearing in records. Reviewers may need identities internally, while the wider report can describe the defect category. Separating working evidence from the published account supports useful criticism without exposing details irrelevant to the improvement decision. Sampling should illuminate the failure mechanism rather than make the organization look uniformly successful.

Examine role-related bias before interpreting satisfaction. Members who prepare papers may find them clearer than those reading them for the first time. New members need procedural orientation different from long-serving members. Compare explanations without turning tiny subgroups into identifiable people. In a small board, reporting 'the sole new member's result' effectively removes anonymity. Group comments by theme without identity clues, or use an independent reviewer to listen and provide a protected synthesis. Explain confidentiality limits at the outset. Promising complete anonymity among people familiar with one another's experiences can damage trust when a recognizable sentence appears later. Safe evaluation design considers what readers can infer, not merely whether names were removed.

Distinguish an individual's failure from work-design failure. If every committee's actions stall at the same employee, follow-up capacity may be allocated impossibly. If one committee repeatedly lacks acceptance criteria, it may need training or clearer terms of reference. Write two causal hypotheses and seek evidence distinguishing them. Did the implementer receive a clear decision on time, and did they have resources and authority? These questions examine cause before proposing replacement personnel. Accountability still matters for evidenced neglect, but institutional evaluation should also reveal conditions making failure recurrent. Reallocating resources while retaining weak follow-up criteria may reproduce the problem in another form. Diagnosis should therefore guide a repair specific enough to be tested, rather than defaulting to blame or generic training.

Convert findings into one or two changes with expected effects. Too many recommendations can create an ownerless review project. If ownership is missing, change the minute template and specify who returns items lacking assignments. If challenge opportunities are weak, test discussion management and invite objections before voting. Keep improvement measures close to the cause: new decisions with owners and dates, or documented responses to material objections. Add an outcome test such as timely implementation. Form completion can improve without work being delivered, so the board should review both procedure and effect. A narrow, well-owned trial is more informative than an elaborate list of aspirations nobody can connect to an observed result.

Before selecting a reviewer, examine their relationship to the scope and ability to understand it. A secretary can collect minutes, but sole judgment of their own minute quality creates an interest. A chair can describe their experience, while review of their chairing needs a route through which others can speak safely. Independence should fit sensitivity and resources. Unconflicted cross-committee review may suffice in one case; a high-impact issue may need external expertise. Record who commissions the reviewer, permitted access and reporting recipient. Ask about actual and potential conflicts before starting. Someone unfamiliar with the subject does not become an effective reviewer merely by being distant from the board. Independence and competence both matter.

Define judgment criteria before reading case identities. Categories might distinguish complete documentation, a correctable omission and an omission preventing review of authorization. Give an example of each, then trial the classification on two files with another reviewer. Discuss disagreements and refine definitions before examining the remaining sample. This guards against changing the standard after discovering who made a decision. It also makes differences explainable rather than merely numerical. Refining categories during a pilot is sensible; changing them after an embarrassing result requires a documented reason and consistent reconsideration of every affected case. Review design should expose its own uncertainties instead of presenting subjective category boundaries as unquestionable facts.

Avoid judging a decision solely with hindsight. A poorly documented project may profit through luck, while a carefully examined one may suffer an unforeseen shock. Ask what the board knew when choosing, which alternatives were available and whether it considered material risks within reasonable capacity. Then examine how it responded to new information. This separation assesses judgment quality without excusing neglect or treating every loss as incompetence. A sound process can need a revised plan; a favorable outcome can coexist with an approval defect. State both judgments separately so members know which behaviors deserve repetition and which weaknesses need repair. Evaluation should improve future decisions rather than reward luck or punish unavoidable uncertainty.

Lesson summary

Board evaluation combines views, decision records and follow-up with clear denominators. Explain discrepancies, protect privacy and choose cause-based repair, distinguishing hindsight outcomes from information available at the original decision.