Accountability measures and executive sponsors

Accountability for equity work should fall on commitments and systems, not on scoring individual people. This program does not rank staff, and committees should not build tools that do. Measures that tally which employees completed the most equity activities, or that rate units against each other in a public list, tend to reward appearance over change, invite gaming and can expose staff who already carry more of this work. They also shift attention away from the question that matters: did the process get better for the people it serves?

Useful measures fall into three types. Process measures show whether the committed work happened: the share of policy proposals that received an equity analysis before submission, or whether notice templates were tested with readers. Outcome measures show whether the process performs better: time to arrange an interpreter, the rate at which people respond to a notice by the deadline, or the gap in approval rates for a service across groups, where data quality supports it. Experience measures come from the people affected, gathered with consent and without requiring identity disclosure. For each measure, state the source, the baseline, how often you will look and who will interpret it. Report the limits of your data along with the results.

The executive sponsor is the committee's link to authority. A good sponsor attends often enough to understand the work, carries recommendations to the leadership table, secures time and resources, removes barriers, and tells the committee plainly when a recommendation is declined and why. A sponsor is not the committee's supervisor, spokesperson in every setting or a substitute for formal offices. Agree on how the committee will bring decisions to the sponsor: a short written recommendation stating the change, the evidence, the options considered, the owner proposed and the decision needed by a date.

When a decision stalls, escalate respectfully and in writing. Remind the sponsor of the decision needed and its effect on people served. If the matter belongs elsewhere, such as a legislative proposal, a contract change or a statewide policy, ask the sponsor to route it. At least once a year, report what changed, what did not and why, both to leadership and to the staff and community partners who contributed. Honest reporting about slow progress builds more trust than a polished summary.

Fictional case: An equity committee wants to show progress to division leadership. Jamal proposes a public chart ranking each unit by how many staff attended equity trainings. Wren suggests instead tracking two things for the notice-redesign action: whether revised notices were tested with readers before release, and whether the rate of on-time responses changes across language groups. The sponsor, Mei, has asked for a report next month.

For one committee action, write one process measure and one outcome measure. For each, name the data source, baseline, review frequency and one limit. Draft a one-paragraph recommendation to your sponsor naming the decision you need and by when.

Privately recall a measure at work that changed behavior in a way nobody intended. What did it reward, and what would a better measure have looked at instead?