A working model to attack “wicked” social problems
Three years ago I moved from Kansas City to Charlotte, North Carolina to lead the startup of Read Charlotte, a community-wide initiative with the Big Hairy Audacious Goal (BHAG) of doubling third grade reading proficiency from 39% in 2015 to 80% by 2025. We don’t operate programs, but rather lead a collective impact effort that seeks to coordinate, integrate and align people, resources and data to improve children’s literacy from birth to third grade.
From the beginning I intentionally sought to develop an approach that would reduce the odds of failure to reach our 2025 goal. I believe we operate in a world of probabilities. Literacy is a socially distributed good. Even though we can’t see the probability distribution, I know factors such as parental education, household income and economic segregation significantly determine which children learn to read well by the end of third grade. However, I was certain we’d be tilting at windmills like Don Quijote if we tried to attack these factors head on to get 80% of third graders reading proficiently within a single decade.
Instead, I made the assumption that we could find specific leverage points in the system where we had a greater chance of influence such that ceteris paribus we could bend the probability curve to give more children a fair chance to learn to read by third grade within a 10-year window.
“You’ve got to have models in your head. And you’ve got to array your experience — both vicarious and direct — on this latticework of models…What are the models? Well, the first rule is that you’ve got to have multiple models — because if you just have one or two that you’re using, the nature of human psychology is such that you’ll torture reality so that it fits your models, or at least you’ll think it does.”
— Charles T. Munger, “A Lesson on Elementary, Worldly Wisdom As It Relates To Investment Management & Business”
One of my favorite Chinese proverbs is “to cross the river by feeling the stones underneath.” (I majored in Chinese as an undergrad at the University of Kansas.) We did not start with a grand strategy, but rather intentionally felt our way into what I believe is possibly a general model for attacking “wicked problems.”
Inner core: research-based predictors. Use high quality research (what the Every Student Succeeds Act calls “Tier 3” evidence) to identify predictors (correlates) of the outcome of interest. Realize you are not trying to address all of the factors that impact your outcome, but rather seek to bend the probability curve by selectively targeting the most critical few. I have a strong preference for proximal over distal predictors. I find it easier to identify predictors by mapping backwards from the outcome. The predictors will correlate to your outcome either directly or through each other. When you move the needle on any of the predictors you increase the odds of improving your outcome of interest (at the individual- and population-level). These are your indicators. Work outwards from the core.
Inner ring: Evidence-based interventions. Find interventions with evidence to move the needle on your indicators. At this stage, your individual indicators are most likely outcomes in research studies for individual interventions. As much as possible, use experimental or quasi-experimental research (Tier 1 or Tier 2 evidence under the ESSA rubric). There are three basic types of evidence: evidence-based practices, evidence-based programs, and practice-based evidence. Practices scale faster than programs; I would look for these first. I recommend you prioritize program models, which can be implemented within existing organizations, over standalone programs. Most likely you won’t have a long list of evidence-based interventions. Use whatever you do find as strategic anchors, and make an effort to stack and align them. After you’ve done this work, next look for practice-based evidence — local “bright spots” that are outperforming on one or more indicators on your outcome of interest.
Middle ring: Execution strategies. Strategy without execution is just theory. Build your execution skillset by learning from three interrelated disciplines — implementation science, improvement science and systems thinking. Implementation science will provide you with tools to choose and implement the most appropriate evidence-based interventions for your issue and community. Improvement science will help you get better at getting better (i.e. “continuous improvement”). Variation in performance is the key problem to address. Both implementation science and improvement science call for “seeing the system” that is producing current outcomes. Programs alone won’t solve the problem. You need to bring the system into the room and look for the places where small shifts can lead to outsized changes in your outcome. These changes will happen either through your evidence-based interventions or independently (sometimes systems solutions produce outcomes on their own).
Outer ring: Everything else. Whatever “wicked” social problem you’re working to solve, it’s doubtless connected to many other issues. Some of these are complementary strategies. Although not directly related to your outcome, they nonetheless increase the overall welfare of the population you seek to serve. Others will be reinforcing strategies. They can help to further support your core strategies in the inner ring. These types of strategies often are easier to identify and understand than the inner ring, which can take considerable time and effort to develop. As a result, it’s really tempting to start in the outer ring. But if there is no strong empirical evidence that these strategies can move the needle on your indicators, you should save them until you’ve fully defined your core strategies. (If there is strong evidence they can move the needle then you need to move them into your core.) Keep in mind that not all strategies in the outer ring are helpful. Some are non-complementary and even potentially destructive with regards to your outcome of interest. Be careful.
For Read Charlotte, we started with the inner core. Our first step in early 2016 was to define a set of research-based predictors of third grade reading proficiency.
Next, we moved to the inner core. First, we identified a set of evidence-based practices. Afterwards, we then looked carefully for evidence-based interventions. Evidence-based interventions is the first factor in the formula for success. We benchmarked proven interventions, looking to meet or exceeded “industry” benchmarks. (The average elementary literacy intervention has an effect size of 0.08, i.e. improves outcomes for 3 out of 100 children. We found 7 that do better.) We then stacked and aligned a set of evidence-based programs. (We are now working to include practice-based evidence to empower families to support literacy at home.)
Finally, we built our internal capacity to execute. We focus our staff professional development on the middle ring. The second and third factors in the formula for success are “effective implementation” and “enabling context”. For the former, we continue to work to master core strategies of implementation science with a particular focus on key implementation drivers. For the “enabling context”, we used two strategies. First, we invested in ourselves (and local partners) to better understand systems and systems change. Second, we are working to master the tenets of improvement science so that we can continuously get better at getting better.
What about the outer ring? Well, the first step is to know whether a strategy is in the outer ring or not. The second step is to decide how much energy to put there relative to the inner ring strategies.
We have seven years left to reach our BHAG of 80% third grade reading proficiency. It’s too early to judge the effectiveness of this model. The children who will take the North Carolina state third grade test in 2025 are currently one years old. I sleep well knowing that we didn’t throw spaghetti on the wall and wait to see what sticks. We’ve got a core group of interventions that are aligned on a key set of indicators that predict third grade reading proficiency. Now our job is to execute them well and get the system working better.
I think this approach might apply to areas besides literacy that have enough good research to identify predictors and a sufficient number of evidence-based practices and programs to choose from. Based upon my cursory review, I think this model possibly could apply to a number of other social issues: early childhood education, youth behavior, high school graduation, crime, youth development, juvenile justice and youth prevention, HIV prevention, community health, child welfare, substance abuse and homelessness.
Let me know what you think. Can this model work for your “wicked problem”?
