This week's guest is Dr. Ed Keating, Deputy Director of National Security Analysis at the Congressional Budget Office (CBO) and a former defense economist at the RAND Corporation.
Ed has spent more than 30 years helping the Department of Defense and Congress weigh tough choices, from whether to extend the life of aging aircraft to how inflation affects shipbuilding. His approach to trust is simple: look at the data, check with experts, and tell the truth about what the data say, even when a sponsor doesn't like the answer.
As Ed put it, "If we were going to hold off decision-making until the data were perfect, then we're never going to make any decisions." So his team presents a best estimate, then shows which assumptions would flip the answer. Sometimes the analysis reveals that a factor everyone worried about could vary by a factor of 10 without changing the result. In Ed's words, "That's a finding unto itself."
New research backs up Ed's approach. A 2026 meta-analysis in Royal Society Open Science pooled 28 experiments on how people respond when a source admits uncertainty. The finding: being candid about what you don't know did not cost the source trust. Two of the studies also suggest that flagging uncertainty early may protect trust when the facts later change.
We also talked about:
Protecting nonpartisan credibility when every number can become political,
Why "the data are bad" is an empty critique, and what his former colleague called the Lollipop Guild,
The biggest myth about how defense decisions get made,
Why trust isn't free and leaders must give their people room to fail, and
Why he encourages more women to consider careers in national security.
These insights are not reserved for national security. They apply to any leader who has to make a call before all the facts are in.
Listen to the full episode on YouTube or your favorite podcast app.
THE BOTTOM LINE
Data will never be perfect, and neither will your team. Leaders earn trust by saying what they know, what they don't, and what would change their minds. The research shows that candor costs little, and Ed's career shows what it earns: credibility that holds up even when the answer is unwelcome.
If you can look back and say, "Here are the data. The data say this," that's your ultimate bulwark, and that's what gives integrity to your work. – Ed Keating
WINNING MOVES
Name your assumptions. When you present a recommendation, state the one or two assumptions that would change your answer if they turned out to be wrong.
Flag uncertainty early. If a plan rests on preliminary data, say so at the start. Early research suggests it may soften the blow if the facts shift later.
Delegate something that could fail. Give a team member a meaningful stretch assignment, just not a mission-critical one, so a stumble becomes a lesson rather than a crisis. Use the Trustworthy Decision Matrix to weigh their track record against the difficulty of the assignment, then agree on what success looks like and debrief whatever the outcome..
Please view the interview on the Win Via Trust leadership podcast for great insights and advice.
If you want to dig deeper into the trust side of leadership, please read Win Via Trust or subscribe below for future Trust Lens newsletters.