For several years I have wondered what it would be like to have a local, state, or federal government run with principles and methods from quality improvement. Until now I had no way to know whether improvement principles could be as effective in governing as they have been in the private sector. What has changed is that a good friend and longtime colleague of mine, Don Berwick, is running for governor of Massachusetts. Don is a pediatrician and former CEO of the Institute for Healthcare Improvement. He is also a master leader of improvement.

In a recent TV interview in Boston the interviewer asked him what differentiated him from the other candidates. He replied:

“What I bring to the table, what excites me, why I want to be governor is thirty years’ experience on innovation and improvement. … How you get things to improve. How you bring innovations to scale. That is my forte.”

To help Don refine this concept API organized a webinar featuring Don to which we invited colleagues and friends who were knowledgeable of improvement. The webinar had two purposes: investigate how improvement methods can be adapted to governing a state and to generate financial support for Don’s campaign. The webinar was free.  We used chat features (online comments from participants) to get input from attendees on some of the initiatives that are emerging as Don’s priorities for his administration and how improvement methods could be used to make progress on them. We covered three topics: Using measures to communicate goals, spreading good ideas from one part of the state to another, and using the Model for Improvement to test ideas and solutions.

MEASUREMENT

System level measurement to communicate what we are trying to accomplish and to provide feedback for learning is fundamental to improvement. Few on the call would be satisfied with an improvement initiative that did not contain key outcome measures plotted over time. Shouldn’t we require the same of our public officials and public initiatives? We discussed some measures related to the issues described on the campaign website. For example, total prison population and number of repeat offenders were proposed as measures of success for the Corrections initiative. A time series of the unemployment rate for Massachusetts from 2006-2013 was discussed. The state data was further divided to display the identical time series for eight regions comprising the entire state. Time series graphs of this type would be common in the private sector.

The chat box  contained some insights on considerations in the public sector if graphs such as these were to be used effectively. Don was encouraged to test out some measures displayed over time in some of the small meetings he is holding with constituents. This could be a very informative test of a basic improvement method. One participant suggested that a debate could be structured around a set of measures displayed over time. Each of the debaters could talk about their ideas in the context of the facts and the current situation.They could be prompted by the moderator to predict how long their ideas would take to change the measure for the better. Some recalled that Ross Perot used charts effectively in debates and speeches. Several of the commenters encouraged the use of stories to complement the measures and connect emmotionally with people.

Recognizing that many of the measures that we were discussing were already publicly available, it was suggested that a third party could set up the set of measures and post them on the web to infuse this thinking into campaigns and governing. Don suggested that in Massachusetts this might be done by the Massachetts Budget and Policy Center http://www.massbudget.org/


SPREAD

Methods for spread of improvements and rapid scale up of promising prototypes are at the leading edge of improvement science. Don told the group that he had encountered numerous ideas or working solutions in one part of the state that could be spread to other locations. One he mentioned was the development of the Assembly Square neighborhood in the city of Sommerville as an example of smart growth that creates jobs. He stated that an adapted approach could work in other parts of the state.

(http://www.bostonglobe.com/opinion/2014/01/18/somerville-little-city-that-could/Oa7nklkmzYp3PPjArSCRCP/story.html)

The use of collaborative projects consisting of multiple organizations working cooperatively on a common aim was developed and popularized as a method for spreading evidence based practices in health care by Don and colleagues at IHI and API. He pointed out that he has already committed on his campaign website to using this approach in state government: “And I will personally help lead a collaborative project among willing communities in the Commonwealth to construct world-class supports to every child under five.”

The concept of spread is found in other areas of the website: “Massachusetts’ charter schools have demonstrated achievements that can serve as laboratories whose results inform mainstream public education." “Our 15 Massachusetts community colleges are especially important in aligning education and workforce needs. The potential for strong public-private partnerships for community colleges is enormous, and we already have some fine prototypes in the state, and they should be available to all.”


MODEL for IMPROVEMENT

The Model for Improvement was developed by API to serve clients in a variety of industries and sectors. It is composed of three questions about aims, measures, and changes that set up the learning and the Plan Do Study Act, PDSA, cycle to test changes. We discussed how the model might be used in governing a state.

What are we trying to accomplish? Clear goals that articulate the priorities for the administration.

How will we know a change is an improvement? Transparently displaying set of measures that allows citizens to judge the progress being made.

What changes can we make that will result in improvement? Good ideas for accomplishing the goals that can come from any location or political persuasion.

Testing and learning: Learn which of these ideas should be adopted, adapted, or scrapped through testing first on a small scale and then spreading the ones that work to other locations.

Attract individuals and organizations in any political or economic sector to work to accomplish the goals.

There was some agreement that the model would be as appropriate for internal state government processes as it is for organizations in other sectors. However, participants noted that more public improvement efforts would face some obstacles not present or at least not as formidible in other sectors. The chat contained some comments on how to define improvement for a general audience: "Improvement is changes that make things better for residents of the state." Provide examples from daily life that are easy to relate to such as more reliable trains, easier renewal of driver's licenses, or more  timely snow removal.

The concept of testing and learning would be different than the prevailing approach of a candidate instilling confidence by assertively giving the solution to any issue or problem.  For example, failed tests might be equated by the media or political rivals as failed policy or ineptitude, rather that opportunities to experiment and learn. One participant recalled that Obama in his first acceptance speech said that there would be mistakes and they would learn ... he was not sure that people remembered or built on that. Many participants expressed belief in the need for an improvement framework in government with proper adaptation to the political environment. Don is in a unique position to make these adaptations because of his knowledge and experience with improvement. Several participants suggested that an agency within the state government could be assigned responsibility for resourcing, educating, measuring and coaching for improvement in the priorty areas.

After the webinar I received personal communication from some participants with some thoughtful comments. One participant wrote the following.  "I was surprised there was no mention of "debt" or fiscal responsibility.  One of the first thoughts that crossed my mind was anything can be improved with enough money.  What might be helpful for Don would be to show how improvement often reduces total costs and how one doesn't necessarily need more money to fund improvement." Regarding cost reduction, the same participant wrote: "Yes, I believe productivity combined with design and outcome thinking could be useful.  However, the idea of productivity solely focused on reducing waste does not resonate with me.  I recall "reducing waste" being connected with productivity during the meeting.  Doing more with less sounds better than doing the same with less although both represent improved productivity."

Two participants had some strong views on communication.  "I agree that Don needs an overall rally cry: e.g.  PEOPLE---PROCESS--PROGRESS.  ...start with empathy...what are people feeling.  Apply Don's special talents in leadership for improvement (process for improvement); results in real progress as measured by the model. My guess is there is a far more powerful theme for Don...just a quick thought. The other suggested: "My political gut says Don could craft his message into pragmatic everyday language, speaking WITH his fellow residents of Massachusetts and decrying that Massachusetts deserves better and then repeatedly describe how the state must make changes that are supported by data that is fully available to all at any time at www..."

Another participant provided some useful advice about how improvement must be imbedded in the right context to be effective. This is the job of leadership - the governor and his or her staff. The following is an excerpt from an article she had written. "Stretch goals have been a topic for disagreement among people interested in improvement.  Some argue that stretch goals are a good thing in that they help people understand that business as usual will not be sufficient and something extraordinary must be done to solve a problem, achieve an improvement, or stay in business.  Others argue that stretch goals are damaging in that they may introduce fear, they may be seen as outrageous and introduce cynicism, and they may introduce gaming, competition and conflict into an organization.  Perhaps the two sides have been talking about two different worlds". (Cultures and contexts)

NEXT STEPS

This summary has been posted on the public API website and I hope that other insights will be put into the comments section. Don was very thankful for the time and insights the participants gave to him. He is considering which ideas from the webinar could be tested in the short term in his campaign. We plan to have another webinar in future as new developments emerge.