The Holacracy Experiment in Washington Government

Michael DeAngelo
9 min readOct 4, 2017

A lot has changed for the Office of the Chief Information Officer since we started learning and experimenting with Holacracy nearly 16 months ago. The OCIO of Washington merged with two other state IT organizations to create the new agency called Washington Technology Solutions (WaTech), which numbers about 550 mostly long-time state employees. As part of the structural merger, we created a division called e-gov, which has the purpose of transforming state government by being an internal disruptive force that drives change. Encouraged by our initial results, we continued the Holacracy experiment in e-gov.

Now, as our experience running Holacracy gains momentum from increased visibility, curiosity, and excitement, we are seeing improved team performance in all metrics. The empowerment metric continued to climb, and seems to have plateaued at about 90% — a 50% increase from the original 60% measured at the beginning of the experience. The speed of processing operational issues has also plateaued at about an average of two minutes to raise, discuss, and decide on an action, a significant improvement from the original 20 minutes at the beginning of the experience. We regularly process 20–24 issues in a one-hour Tactical meeting. Does this prove we are becoming faster and more adept at practicing Holacracy or does it demonstrate improved organizational outcomes? To determine this, a lot of questions need to be answered.

As a first step, we are continuing to grow the coalition of parties interested in the concept of self-management and interested in a real test of its ability to support employee needs at work. In the blog post, “Holacracy in a represented workforce — (part 1 and part 2) “, I wrote Holacracy and the potential implications with organized labor after meeting with the state’s HR Policy Director and an attorney from the Office of the Attorney General. Since then, we met Union leaders on self-management and Holacracy. The union leaders at the Washington Federation of State Employees (WFSE) were excitement by the concept and thanked State leaders for having the courage to explore self-management as a possible organizing model.

We need to get beyond the hype

As a government agency, the idea of self-management creates a lot of anxiety and fear. There is the usual fear that comes with any company making change that challenges the long-standing cultural norms and beliefs on how to best organize. There is also the fear of how a self-management system will change the traditional role of a manager and challenge their perceived source of power. We’ve seen justifiable concerns around the best way to lead employees through a radical change that rattles the inertia of the status quo. In government there is also the fear that comes with being in a fishbowl. How will the public perceive this type of change? Will they cheer the change on and support a government that is trying to innovate and transform? Or, will they criticize leaders for being too bold?

Holacracy and other self-management models lack real scientific data that prove they are better than a hierarchy. (One could also make the observation that hierarchical models lack real scientific data that prove they are better than self-management, too, but I will touch on that in a moment.) What we do know is there is a lot of anecdotal information from companies that have adopted Holacracy, or other self-management models, and there is a high degree of hype in the media about the teal organizations which have taken this path. This has been enough for many companies to take the leap of faith as they struggle to be more adaptive with the increased pace of change and competition in their sectors. But government often lacks the same competitive pressures to change and has far more incentive to stay the same.

Companies have largely been relying on hype to create the force needed to break the inertia of the status quo. Hype is useful, but data is better. For government leaders to propose bold changes and confront the typical concerns surrounding those changes, they will need data. Since rigorous data doesn’t exist, we’ll have to create it. There is risk that the results of the experiment will not support the hype, but the potential upside for improving organizational and employee outcomes is worth testing.

Holacracy in Washington: Act 2

We have shown that Holacracy has improved outcomes for the relatively small group of people included in our initial test groups. The modest amount of data we’ve collected thus far gives us a strong reason to believe that a self-management system like Holacracy creates better employee and organizational outcomes. However, success in a relatively small team of about 20 people doesn’t mean such success can scale, or that there would be success in significantly larger organizations.

The next act in evaluating whether self-management should be a part of an EVP for state government is to expand Holacracy practices to more teams and construct an experiment that scientifically tests Holacracy.

“The most difficult thing is the decision to act, the rest is merely tenacity.” — Amelia Earhart

The Holacracy Experiment

To have a scientific experiment you need a hypothesis to test, an experiment design that adequately probes the hypothesis, partners to run the experiment, a plan for measuring results, data collection, and a plan for executing the experiment. WaTech has been creating a group who is interested in assisting us with these objectives to better understand if our initial hypothesis is correct.


“Holacracy improves employee and organizational outcomes.”

Experiment Design

The experiment is constructed as a control experiment with a set of teams practicing Holacracy and another set of teams remaining unchanged by continuing to operate in a status-quo hierarchy. The first step was to find out which teams wanted to participate in the experiment by allowing all teams, nearly 600 people, the choice to “opt in” or “opt out” of the experiment. To opt-in required that the manager and a majority of team members agreed to participate. Nearly half of the organization opted into the experiment. The expectation of teams opting in is that they actively participate in the data collection activities of the experiment. This decision effectively split the WaTech agency into an opt-in group and an opt-out group.

The next step was to assess the teams in the opt-in group to further bifurcate them into the groups running Holacracy and groups acting as the comparison or control. To create the two groups, baseline data was collected to assess all teams on multiple measures to ensure there was balance between the two groups. The teams didn’t know, at the time they opted in, whether they would be in the Holacracy group or the comparison group.

Experiment Partners

To structure a rigorous experiment and avoid organization bias, WaTech knew it required help. To assist with building and running the experiment we’ve put together a team of organizational partners who are interested in transforming government and understanding self-management’s potential for organizational improvement.

  • Michael Lee, Harvard Business School (HBS) researcher, helped construct the experiment design and will be collecting the data during the experiment as well as performing the analysis at the conclusion
  • Evergreen State College (Evergreen) Master of Public Administration (MPA) program and students are helping by providing on-site data collection of qualitative data at the direction of HBS researchers and are supplying operational support during the experiment
  • HolacracyOne (H1) and Thoughtful Org Partners (TOP) are providing expert Holacracy coaching, training, and facilitation support throughout the experiment
  • Washington State Federation of State Employees (WFSE) is advising on labor issues and boundary management during the experiment
  • Office of Financial Management State (OFM) HR office is advising on classification and compensation issues and boundary management

Data Collection

During the twelve month experiment, the HBS researchers with help from Evergreen MPA students will be collecting both qualitative and quantitative data. Quantitative data will be collect through a variety of sources including employee survey data from the Holacracy group and the comparison group at the beginning (baseline data) and at future intervals of six and twelve months.

Qualitative data will include monitoring team meetings and performing in-person interviews for both the group practicing Holacracy and the business-as-usual comparison group. Information will be collected on a regular basis. While eventually specifics about the data we’re collecting, general categories of observation, and other data collection particulars will be disclosed and available for scrutiny, at this point of the experiment we are not supplying further detail, so as not to influence the experiment or its practitioners. We recognize the potential for experimental disruption, and we want to avoid that as much as possible.


Since this is the first time a Holacracy Experiment is being conducted, an agile approach is being used to develop the plan. This allows us to better manage the risks and quickly adapt to change and learning. The complete high level plan was created early in the project and consists of three releases or Phases:

  1. Launch
  2. Run and grow capacity
  3. Reporting on results

The details of each Phase are being built as we get closer to starting the phase by defining the major Epics in order to accomplish the goals of the Phase. The Launch Phase is broken into four Epics:

  1. Setup, communication, and identifying teams
  2. Collecting baseline data
  3. Launch
  4. Providing initial on-site training.

We Launched!

In April, 2016, we launched our scientific experiment. The experiment includes roughly fourteen teams of about 130 people who will operate within a Holacracy. Another fourteen teams of about 130 people are acting as our comparison group, and they will continue to operate without changes from their existing processes. The experiment will run for twelve months.

In May an on-site Holacracy Practitioner Certification Training was conducted to nearly 50 of the participants concluding Phase 1 of the plan. The experiment has moved into Phase 2 — Run and grow capacity, and this phase will encompass the duration (12 months) of the experiment.


Inside of the experiment, there is a lot of work to properly learn Holacracy, build proficiency and expand capacity among employees practicing it. Outside of the experiment, there is mainly waiting — waiting for the experiment to run its course, so that we can all see the data and the results of testing the hypothesis.

State government is full of people who are passionate about making a difference in the community and creating a positive impact in Washington State. The experiment will create the scientific rigor needed to test the hypothesis that Holacracy improves employee and organizational outcomes. It’s not yet clear whether it will make the improvements we are hoping to make, but by their embrace of the challenge, state employees have shown their huge energy and heart. It’s also clear that there are numerous organizations impressed by Washington State’s commitment, courage and boldness to lead and to innovate.

Although there are some who fear the possibility of change, the vast majority of the people I’ve shared this story with are amazed by the emerging story of Washington. I am energized and excited to continue to find like-minded leaders inside and outside of state government who are just as passionate about transforming State Government as myself. One thing I did not expect was to see and hear people express a renewed faith in their government because of WaTech’s willingness to take a deeper look at changing the way we work together to accomplish more for our organizations, our employees, and the citizens who need us to succeed.