Rethinking poverty, the brain, and education

Note from author: This post was transferred from my Medium page on April 19th, 2020. All future blog posts will be posted on my GitHub page.

Recently, I read Dr. Sendhil Mullainathan’s and Dr. Eldar Shafir’s 2013 book titled “Scarcity: Why Having Too Little Means So Much”, which explores how limited resources — with an emphasis on poverty and finances — affects our behavior. Initially, I thought poverty contributed to “fixed” adverse consequences: that the detriments of impoverishment would fester brain development that would cement into (less than optimal) human behavior necessary to thrive in today’s society. Additionally, I felt confident that if these behaviors unique to low-income communities were static, then these differences would become apparent in my statistical analyses during my graduate studies at Columbia. However, Dr. Mullainathan and Dr. Shafir’s book refined my understanding and forced me to rethink how to explore the delicate relationship between poverty and behavior.

One of the arguments that the authors make is that there are no intrinsic differences that occur but rather that the psychological effects of scarcity can encroach into our decision-making at the moment. For example, they introduce a study done at the mall in which participants were asked two scenarios: (1) how would they come up with $150 for paramount car repairs or (2) requiring $1,500 instead. Following their response to one of these questions, participants answered various Raven’s Matrices questions (i.e., an IQ test). In the former scenario, they found no significant differences between the two groups ¹. However, the introduction of a financially-aggravating state — forcing low-income subjects to think about the barriers to come up with $1,500 in a short amount of time for a single expense — resulted in statistical differences between the two groups: low-income households had lower IQ scores.

The significance of this finding for someone like me — someone who aims to study poverty’s impact to learning in the classroom — is that it provides a paramount explanation for the life differences among individuals of different economic classes — including any nuances between families within the same income range. It supports the notion that the rich and the poor are not fundamentally different, but that they will behave differently in specific circumstances (i.e., when money becomes involved). Thus, depending on the range of financial adversity — and access to mitigate monetary hardships — a household experiences, how one behaves as a response to their reality will vary notably. As a result, this indicates that the ways in which I think about how to statistically illustrate these behavioral patterns (and the subsequent influence on my experimental design) require more rigorous studying.

For example, even among low-income households, there is significant variance in backgrounds that manifest behaviorally in everyday life. One household may have no one who attended college; another may have a parent with a Ph.D. — but from another country and is not recognized in the U.S.; perhaps another household strongly emphasizes a college-bound mindset; while another household has high financial literacy skills; ad infinitum. Each of these factors are extremely important because all of them will influence the framework that the individuals operate under, which will undoubtedly affect their behavior. Due to this, it is crucial I explore the differences amongst these subgroups rather than grouping them all together.

So as I prepare to undertake this field of research, I thank Dr. Mullainathan and Dr. Shafir for providing me with two new insights:

1. Not to assume that any communities are fundamentally different based on socioeconomically class alone. This is not to say that, with regards to behavior, they are not neurally homogenous (but is anyone??). Rather, if there are similarities in low-income communities compared to a higher SES group, this can exist without assuming that these indicate permanent differences. Additionally, this can indicate that these are not innate to one’s genetic markup, race, gender, or culture, but rather a response to one’s (immediate) environment. Perhaps, as a result, may even insinuate that altering one’s environment — and identifying elements that may emerge in learning centers (e.g., classrooms) — is sufficient to reverse any negative effects on optimal human behavior.
2. Statistically, I need to contextualize the aforementioned nuances in low-income households in any of the studies that I conduct. So far, in the literature that I have reviewed, much of the data collected to differentiate is based solely on race and gross income, but I think that for the purposes of my studies, the following factors are all extremely important to incorporate: the total number of educational attainment in the immediate household; culture of learning and education; quality of schools students had attended; and current financial stressors (e.g., debt, medical expenses). Essentially, I need to capture the extent by which poverty had affected preferred behavior because income alone is not sufficient to conclude that the impacts of poverty are congruent across individuals.

I am curious to know if there are differences among individuals in those sub-categories: I really want to know how different my life could have been if my parents had inculcated a more college-bound mindset. How would more intellectually-infused conversations at the dinner table influence aspects of my brain development? How exactly does the quality of one’s instruction interact with our learning mechanisms, and are those patterns consistent across families of different socioeconomic backgrounds? I can’t know for sure just what patterns exist (if any!), but I know that as I engage in this body of research, I want to disentangle as much as I can to get to the root of how poverty affects our development.

1. While they found significant differences, I will try to locate the original paper and verify the results and include any caveats here. The first noteworthy limitation is obviously the selection bias: looking at mall-goers (if someone is at the mall, they may have the financial bandwidth to shop.) Additionally, I would argue that the likelihood that someone on the extreme end of poverty would not be spending their free time at the mall (but this is merely speculation and I will verify if more context regarding the income of the two groups is reported in the original paper). ↩