The United States is facing a crisis of opportunity. Recent studies have shown that economic mobility in America is on the low end when compared to rates of economic mobility in other developed countries[i]. This raises an important question: how does one advance in the American economy? To help answer this question, Child Trends, working in collaboration with the Brookings Institution and the Urban Institute, helped to develop the Social Genome Model. To explain what the Social Genome Model is, and how it works, I explored the nuts and bolts of the SGM with Child Trends’s Senior Scholar Kristin Moore and Research Scientist Vanessa Sacks in a Q&A.
Q – What is the Social Genome Model?
A – The Social Genome Model is a microsimulation model that simulates people’s socioeconomic trajectories over the course of their lives. The model looks at academics, social development, economic and parental status, and a number of other factors to predict what the likeliest pathways for success at six different stages of life. This helps us figure out which policies and programs can be used to improve upward social mobility.
Q – What are the six different life stages?
A – The six life stages are circumstances at birth, early childhood, middle childhood, adolescence, transition to adulthood, and adulthood. At each of the first five stages, there are certain benchmarks that, based on the model, are the best predictors of success in adulthood. These benchmarks are:
If these benchmarks are met, then it is likely that by the time a child reaches adulthood, he or she will have reached middle class – in other words, have a family income of at least 300 percent of the poverty line[ii].
Q – How does this model work?
A – In the simplest terms, the SGM is a simulated population, built from a large national survey, which uses a series of regression equations to predict how factors in one life stage predict outcomes in another life stage. In other words, it uses existing information on a nationally representative sample of people, who have various racial/ethnic backgrounds, income levels, and education levels to forecast what is likely to happen to someone in a similar situation.
Q – What can be done with the SGM?
A – One of the most important features of the SGM is that we can adjust any of the variables in the model to see how it changes the life trajectory for a group of people (we call these adjustments “simulations”). For instance, if we wanted to know how an increase in the number of teen moms who complete high school would affect the success of their children later in life, we could adjust a few numbers in the model, and it would give us an idea of what that might look like. This lets us ask a lot of “what if” questions about socioeconomic mobility and other questions about the possible effects of policies and programs.
Q – Can the SGM be used to help policymakers?
A – Absolutely. The ability to run all these simulations is a valuable tool for policymakers. For example, if a state government has a limited amount of money to put towards social welfare programs, public officials want to make sure the programs they invest in are the ones that are most likely to have long-term benefits. The SGM can be used to estimate each program’s effects, which can then be used by policymakers to help them make decisions. Colorado is actually doing this now, through the Colorado Opportunity Project[iii].
Q – How is Child Trends involved in SGM?
The model was originally developed by the Brookings Institution, with input from Child Trends and other scholars; but since then it has become a joint venture of Brookings, Child Trends, and the Urban Institute. We worked closely with Brookings and UI to update the model with more recent data, and we have been running simulations that give us a better idea of which programs have the best shot of helping kids succeed later in life.
Q – Where do you see the SGM going from here?
Well, right now we’re looking for more funding so that we can find new ways to improve the model. One thing we’re looking into is adding more variables to take the model to the next level and continue to increase its accuracy. Specifically, we’re interested in adding geographic information, which would allow us to run simulations on a state-by-state basis, as well as further enhancing the early childhood and middle childhood modules. Another thing we’re looking into is a user-friendly tool that would allow anyone to run simulations just by going to a webpage and entering a few simple parameters. I think that would really expand the relevance of the model and let it reach a lot more people.
Interview by Max Kenower, communications intern and a student at American University in Washington, D.C. studying film and business