blog dds: 2010-04-07 — Ron Heifetz on Crisis Leadership

Earlier today I had the privilege to attend a lecture on crisis management by the Harvard Senior Lecturer Ron Heifetz. Here is a list of points that struck me (in the form of slightly edited tweets), and my view of their relevance to software development.

Avoid the trap of responding to the crisis's acute phase and ignoring the adaptive phase. Heart surgery (acute phase) without lifestyle changes (adaptive phase) is worse than useless, because it provides you with false security.
Acute phase response: physical presence, name risks, maintain poise, drive response, improvise, frame key issues.
Key crisis management failure: people treat adaptive problems as if they were technical problems.
Technical problems have a clear definition and implementation; adaptive problems require learning in order to define them and to implement them.
Only 20% of US heart surgery patients follow doctors' lifestyle advice. (Or as his hospital's chief MD put it: 1% follow the advice and 19% are lying).
Heifetz's technical solution to a beaten-up girlfriend arriving at the hospital: treat wounds advise her to leave her boyfriend. Three visits later he realized the adaptive challenges she was facing: tolerate loss, fight loneliness, experience family disloyalty (her mother was also abused by her alcoholic husband), cultural adaption.
Adaptive change requires renegotiating loyalties. Crucially for Greece's government, clientelism and patronage is a system of loyalties.
Technical problems are tough, but can be solved quickly. Adaptive problems require behavioral changes, which take time.
Most successful change is less about change than about preserving the best of an organization's current DNA and making small adaptations.
Nature demonstrates the value of diversity: sexual reproduction is very complex and costly (30% of pregnancies spontaneously abort), but it produces a much higher rate of innovation than cloning. Therefore, try to perform many experimental changes.
Roosevelt in his inaugural speech asked for war powers (terrifying his wife) to handle the depression's economic crisis. He didn't use them, because he knew he didn't have the one true answer and he was willing to experiment and learn. On the other hand, around the same time Hitler took those powers and led his country astray because he thought he knew the solution to Germany's problems.
People are distracted thinking that Greece's problem is the Greek government bond spread and not the way they lead their daily lives, their practices, and the clientelism requirements.
To foster adaptive change, experiments, learning, and innovation you must embrace the possibility of failure.
Authority differs from leadership. You can lead without authority and you can be in a position of authority without leading.
Services of authority: direction, protection, order (role orientation, conflict control, norm maintenance).
Paradox: you must exercise leadership to restore lost confidence in authority relationships.
Dividend of trust: the civil society in prosperous northern Italy is much more advanced than in the poorer south where only family ties matter.
A crisis is not an opportunity to win friends. You must be able to take the heat and stay calm.
Political thinking: find allies, keep opposition close, come clean, acknowledge losses, model new behavior, accept casualties.
Question: How do you renegotiate loyalties?
Answer: You do it in your own heart and you do it through dialog.

Although the seminar was aimed at political leaders, this being mostly a hacker's blog, I can't help but noting the relevance of many of Heifetz's observations in the field of software engineering. Consider the differences between technical and adaptive problems and their respective solutions. Phillip G. Armour has been arguing since 2000 that software development is a learning activity. Or think of Heifetz's advice to try many small experiments; this is clearly something that strikes a chord with those of us who believe in agile software development.