LILA ~ Learning Innovations Laboratory at the Harvard Graduate School of Education

Looking for content and documents from our Gatherings? Login

Transparency & Unlearning with Ethan Bernstein

Posted by

Bernstein 2014 LILA SummitKey Questions/Themes: Though transparency at work seems like it is always preferable, giving employees privacy from monitoring can enhance performance.

Summary of Session Content

Transparency – there is a gospel of transparency in leadership in organizations. The concept went from ‘being able to see’ to a broader definition of openness and freedom of information. But there was an unanswered question in this assumption of transparency as good – does transparency increase productivity? Ethan conducted a field experiment in a factory in China that was, at the time, following all best practices.

He found that when observed, employees did what were ‘supposed’ to do. But when unobserved, they engaged in lots of local problem-solving. Management walking around and monitoring was counter-productive. Workers saw lower productivity as keeping managers happy and meeting quota. So, conducted field experiment on two lines on the factory floor. In this initial experiment, the control and experimental line were too close to each other so he put up a curtain to keep the original experiment controlled. This turned into a new experiment when a worker suggested putting up curtain around entire line. The line with the curtain around it had higher productivity (than the control line right next to it) because employees:

  1. Had more ability to do real-time adjustments
  2. Could experiment with new knowledge before sharing – prototyping
  3. Avoided interruptions by management – interruptions had negative consequences

Transparency paradox

We achieve privacy in two ways

  1. creating visible boundaries
  2. encrypting exchange of information

This second one is key. We’ve forgotten workers have agency – they can remain private in ‘plain sight’ and encrypting take energy and resources?

Outcomes of providing visible boundaries

1. Boundaries matter – observer to observed

2. As visible barriers go down, human behavior creates invisible ones. We can try to fight it but people are very creative. This creativity consumes time and resources.

3. Without effective zones of privacy, orgs undermine capacity for performance

4. There is a U-shaped curve from low to high transparency and productivity.

To apply these finding in organizations, Ethan provided four ways to think about transparency

  1. To whom?
  2. For what purpose?
  3. Used how?
  4. For how long?

Organizations should ask these questions as they think about when and where transparency will serve productivity.

Synthesis of Small Group Conversation

Key themes:

Considering how and when to apply transparency in organizations is an unlearning process that challenges current narratives and assumptions.

It seem important to consider transparency at the local and organizational level separately because trust and fear differ along these lines.   Workers saw themselves as managing the attention of others down the line because thought could do better than managers (observers). They monitored each other and let each other know when they needed attention and keep each other from being distracted. However, the organizational narrative matters. How much does evaluation factor into it? Privacy can be viewed as ‘not ok’, you are hiding something. This will affect behavior at the local level.

Reference: Daniel Solove – Nothing Hide. Talks to the danger of implying motivations behind privacy.

Privacy is an easier sell than autonomy but still respects employees’ agency. People came to line just as frequently so didn’t necessarily have more autonomy but more visible privacy.

This work can be applied to knowledge workers. So far, there is supporting evidence with engineers and consultants. Allows for greater ability to conduct experiment, try things when not so closely watched. Managers are not actively inhibiting but, by the very nature of the manager employee relationship, privacy allows for greater risk-taking.

Optimal privacy also has limits. For example, when considering handoffs or transfer of knowledge. Boundaries need to be both explicit and flexible. Minimal structures but clear boundaries. Can experiment locally but must still share and transfer information. This is a design approach to create conditions under which people are more likely to share relevant information. Making everything visible creates conditions where people seek to hide. Productivity comes from trading off entire audience to have honest smaller conversation.

Conclusion | Key Take-Aways & Lingering Puzzles

There remains a tension between privacy and knowledge sharing. The open space design that is gaining popularity in workplaces increases transparency and likely underscores the narrative that needing privacy is equivalent to hiding. However, there seems a more fundamental issue at the heart of the transparency paradox. Managers must relinquish control over the process – whether it be a production line or a problem-solving process.

Especially in areas where productivity is more difficult to measure, how do we know if privacy is beneficial?

The organizational narrative and the local sense of trust will likely drive the decisions made around privacy. The connects strongly with topics discussed this year about unlearning. We have to change the story and challenge assumptions about what the optimal role of management is, and we have to trust that workers want to do good work and be productive. One way to make this easier is to consider the insight that, whether granted or not, employees maintain privacy even ‘in plain sight’, through encryption. Replacing encryption with explicit visible boundaries frees organizational resources and changes the narrative from one of monitoring to one of trust.

How do we know the optimal level of privacy for a given situation/group?

Harvard Graduate School of Education