Friday, March 22, 2013
The Complexities of Poverty
One of the tropes regularly offered in support of this illusion is an example of how one family in poverty outperforms another when simply offered the same chance at success. This argument is ancient among conservatives, the classic Horatio Alger story of someone pulling themselves up by their own bootstraps as proof that if they can do it, anyone can, thus disproving structural inequality as an oppressive force. Similarly, education reformers point to the success of certain charter schools, saying essentially, "if they can do it, why can't other schools?"
Yet the problem with these arguments is that they rely on faulty comparisons, ignoring the complexity of large data sets in favor of narrow, cherry-picked results. Both examples take a view of poverty, which no one can argue doesn't correlate with disadvantage, that misses the specific qualities of poverty that actually cause disadvantage to begin with. By excluding these causal relationships, they are able to emphasize cases in which the causal relationship doesn't exist, and people are able thus to free themselves from its shackles. For instance, a Horatio Alger who while poor, has not been molested since the age of 11 is a very different young man than the Horatio Alger who has not. Or grew up without a father. Or grew up with a mother who didn't know how nurture him. Or was bullied because he was gay, etc.
In the reformist film Waiting for Superman, a portrait was painted of a community of poor parents whose children's success was determined merely by whether or not they won a lottery to get into a neighborhood charter - ostensibly a "good" school. Yet the reality of good vs. bad schools is that they are overwhelmingly determined by SES. That is, the families themselves. This seems contradictory. These are all highly motivated parents, who want the best for their children, so what gives? The simplistic answer is that they are being held back by "bad" public schools. There is some truth to this. But the bigger picture is more complex.
In any given classroom that draws from a heterogeneous population of poor students, you are going to have a range of family stories. Statistics bear this out, but I can personally attest to witnessing this in my own classrooms. The higher performing students are well behaved, come from relatively stable homes, and are motivated to learn. These are the students whose parents show up on parent night, and are generally involved in their children's education. The middle-ground students are less stable, often experiencing a variety of stresses in their lives that impact their academic engagement. Their parents are harder to get a hold of, and generally less available. The lowest performing students have a variety of behavioral issues, and their home-lives tend to be the most unstable. Their parents might have substance abuse problems, issues with the law, and are frequently very difficult to get a hold of. These students take up most of the teacher's time, and generally place a heavy burden on the rest of the class. Some of these students will of course be successful despite their backgrounds, but it is a daily struggle. Just the other day, I overheard one of my students tell a friend that, when she told her mom about another student who was causing problems with her, that she should "beat that bitch's ass". "But Mom, I said, "as the student went on to describe the conversation, "I can't get in trouble again for fighting because of my probation! I don't care, she said - that bitch deserves it!"
In the film Waiting for Superman, you can guess which parents are clamoring to get into the exclusive charter schools.
Too often, we see such a simplistic model of how SES relates to a student's human capital, or their total abilities to be successful. SES is a powerful general predictor, but it is just that - general. Too often it is assumed to represent more than it should. As it is mostly used in education, it refers only to parent income. But that is only one measure of what is a much more complex picture of SES.
A proper measure of SES would include not only family income but things like parent education, whether the family is intact, are there health issues both physical and psychological, neighborhood safety, substance abuse, etc. The number of factors is almost endless, and there are limits to what can reasonably be measured when designing policy. But that doesn't mean that confining ourselves to a simplistic model frees us from being limited by our data. Just the opposite, it means our data is superficial, and any conclusions we derive from it will be limited.
So, by looking only at parent income, two families could look similarly disadvantaged on paper, while in reality having wildly different levels of disadvantage.