2.5 How do we select results for a given population?

The Short Answer

1. Results are plain language conditions of well-being for children, adults, families and communities.

2. Choosing results is a political process more than a technical process. You are looking for a set of statements which are understandable to the public, say something important about the well-being of a given population, and which are reasonably complete.

3. Create a process which has a highly respected steering committee, is well staffed and has lots of room for broad participation. Make sure the development process is kept on track with a clear and coherent framework.

Full Answer

(1) Choosing results is a political process, and it is important to recognize that at the beginning. So we will discuss approaches to organizing the political process and then the technical process of selecting results.

(2) Political process: The most successful processes for selecting results seem to have the following components:

(a)   A highly respected steering committee made up of public and private sector representatives. For example the Georgia Family and Children’s Policy Council has prominent business leaders, faith community leaders, and high level legislative and executive branch representatives. (GEORGIA, SEE ALSO MISSOURI FIT, OREGON PROGRESS BOARD ETC.)

(b)   A framework: It is essential that the group adopt a conceptual framework for the work to keep it coherent and disciplined. (The RPA framework presented here has served this purpose in many states and counties.)

(c)   Staff. If you pick the right people for the steering group they will not have time to do much of this work. You will need one or more staff people to do the research, staff the meetings and make recommendations.

(d)   The steering committee and staff take the time necessary to get their act together and develop a version of results and indicators that they think is pretty good. This will take some back and forth between commissioners and staff. While this process should not be closed to the public, it should allow for plenty of give and take. The product should be something that could stand as the final best thinking of the group. This product is then offered as a starting point for discussion in the next phase.

(e)   Broad participation. The best processes used many different ways to obtain local input including hearings, focus groups, as well as paper and electronic distribution methods.

(3) There is a debate that often takes place about whether it is necessary or desirable to gather public input BEFORE the steering group develops its recommendations. There is not a right or wrong answer to this question. But agreement on a framework is absolutely essential before broad based input is sought. In other words, the steering committee and staff must have a structured way to hear, to organize and to interpret what people offer as recommendations. In one state where this was not done, the public input process generated recommendations for all the steps in the RPA thiking process (results, indicators, story behind the baselines, what works etc.). It is exactly what you would expect from that kind of process. The group did not have a clear method for sorting the difference between indicators and what works ideas and performance measures etc. and the consequent product was a hodge podge of many different (non-parallel) components. It made the subsequent work difficult and confusing and the process was ultimately abandoned.

(5) Technical How to: Here is an exercise which works with groups to develop results for a population starting from a blank slate. An important thing to remember is that different groups will produce different lists of results (or anything else). You can run this exercise once with a single group and then use the product (with refinements) as part of an external review process. Or you could run this exercise with many groups and craft their many products into something that represents common ground.

TECHNIQUE for Generating Results List from Scatch

Purpose: Create a list of 10 plain language results.

Pre-requirements: all participants must have been trained or instructed so they have at least a basic understanding of the difference between results indicators and performance measures. Preferably all participants will have training in the full results-based decision making process.

Step 1. Ask participants in the whole group to brainstorm endings to the sentences:

We want children who are…

We want families that are…

We want to live in a community that is…

Step 2. Break into three groups (by children, family and community) Have each group pare their list of answers down to 5 for children, 3 for families, and 2 for the community. (Option: have each group designate “emissaries”to the other two groups to discuss potential duplication and overlap and other issues necessary for the work of the groups to fit together.)

Step 3. Bring the large group back together and have each group report. Have the large group make recommendations for changes that will make the list function as a whole.


We want adults/elders who are…

We want an environment which is…

We want an economy that is…


(6) We are now accustomed to the idea of  results for children and families. But the idea applies to almost any condition of well-being you can imagine. For example here are some “non-conventional” results:

Community with adequate affordable housing for all

Community with adequate sustainable water supply (LA, Santa Fe)

Forrests not prone to fire (pick any western state)

Communities without graffiti

(7) An important pitfall:  Starting on one result without a complete list of results creates the tendency to load everything onto that one result because it is the only game in town. So if “healthy children” is picked as the only result the collaborative has identified to work on, then every part of the well-being of children can be made subordinate to health. So, for example, family self sufficiency is part of health because families with higher incomes have better nutrition and better access to medical care etc. The next month a task force on family self sufficiency concludes that health is subordinate to self sufficiency because in order to be self sufficient families must be healthy etc. This is in fact just another version of the old game where the whole world is seen through the eyes of just one agency or one profession. It is probably better to develop at least a complete working list of results before choosing which ones to work on.

(8) A word about the relationship of state and local development of results and indicators: The wrong way tot do this is to develop a set of results and indicators at the state level and then impose them on local folks. There is a legitimate state interest in having a core set of results and indicators which allow different parts of the state to be compared. But this does not mean a monolithic top down process. There are two solutions that have been used in other states that seem to work well. In Georgia, the state has developed a set of core results and benchmarks and then allowed local Family Connection Councils to add to this list. In Oregon, the goals and (approximately 92) benchmarks provide a thorough picture of quality of life, counties, cities and communities can choose which of these to adopt for local use. Mutlinomah county, for example has identified a subset which it calls “urgent” benchmarks. Whatever approach you decide to take, make sure it is  respectful of the legitimate differences which exist between local conditions, values and priorities.

(9) Product: The results list can be presented in many different forms. In a number of states (notably Vermont, Missouri and Georgia) , the results or outcomes list is a well-known communication tool all by itself. The most usual format for presentation of results and indicators is a report card on the well-being of children and families. See 2.10 How do we create a report card and what do we do with it? (on child and family well-being, for other populations, for an entire community quality of life)

Marc2.5 How do we select results for a given population?