More Smoke Signals: Identifying Education’s Relationship with Tobacco

Published in

Information Expositions — Spring 2024

9 min readMay 7, 2024

Tobacco consumption remains a significant public health concern, with implications for individual health and broader societal well-being. As an extension of a previous data analysis project aimed at understanding the relationship between tobacco imports and education spending, this research study seeks to explore the intricate dynamics between these variables while considering additional confounding factors such as median income and population, asking the question: What drives tobacco imports and how can they be controlled?

The initial hypothesis argued an inverse relationship between education spending and tobacco imports, grounded in the assumption that higher education funding would lead to a more educated and wealthy population, thereby reducing tobacco consumption. However, the analysis produced unexpected results, revealing a positive correlation between tobacco imports and education spending. This finding prompted a critical reevaluation of the underlying mechanisms driving it and underscored the need for a more comprehensive study.

One plausible explanation for the observed positive correlation may stem from the presence of confounding variables, specifically median income and population. By incorporating these variables into the new analysis, this study aims to disentangle the complex relationship between education spending, socioeconomic factors, and tobacco consumption. Furthermore, a time dimension adds another layer of complexity, as the effects of policy interventions and societal changes unfold over multiple years. The overarching goal of this research endeavor is twofold: first, to recognize the causal mechanisms driving the observed correlations; and second, to inform evidence-based policy interventions aimed at reducing tobacco consumption and promoting public health.

A major point of interest is in the potential implications for education funding policies. While the positive correlation between education spending and tobacco imports may seem counterintuitive at first glance, it underscores the interconnectedness of public health and education. By investing in education, policymakers have the opportunity to empower individuals with knowledge and skills, but also to foster healthier lifestyles and communities. As such, this research advocates for increased education funding as a means to curb tobacco consumption and improve public health outcomes.

Methodology

In this study, we adopt a longitudinal research design, looking at the same variables over a period of time, to examine the relationship between tobacco imports, education spending, median income, and population across eleven states and over a seven-year period from 2012 to 2018. This design allows us to capture trends and patterns over time, providing a more comprehensive understanding of the dynamic relationship between these variables. The decision to incorporate median income and population statistics as confounding variables stems from their potential influence on both education spending and tobacco consumption. Median income can significantly impact purchasing power and lifestyle choices, while population dynamics can affect resource allocation and public health initiatives. By including these variables in our analysis, we aim to uncover the nuanced relationships at play.

In terms of metrics, we will employ correlation, R-squared, and F-Statistics/P-Values to quantify the strength and significance of relationships between variables and assess the goodness of fit for regression models. These metrics will provide valuable insights into the underlying dynamics driving our analysis. Moreover, our chosen visualizations will further complement our analysis by providing intuitive representations of the data. Firstly, correlation matrix will allow us to visualize the strength of many relationships at once. We will then perform a multivariate regression to prove our model is a good indication of the dependent variable, tobacco imports. Finally, we will study a variety of scatter plots that will help inform the conclusions we can draw about the overall relationships.

Results

The correlation matrix heatmap generated for our analysis reveals several noteworthy relationships among the variables. Notably, the strongest correlation observed is between education spending and population, with a correlation of 0.9. While neither of these variables is our dependent variable, the strength of this relationship paints its significance in the overall picture. The second most prevalent correlation is between education spending and tobacco imports, with a correlation coefficient of 0.49. This relationship, of moderate strength, exhibits a positive value, indicating a tendency for education spending and tobacco imports to increase or decrease in tandem. This finding is the key to our study, requiring extended examination to understand why education spending appears to be positively correlated with tobacco imports. One plausible explanation for this positive correlation could be the indirect effects of education spending on societal factors and lifestyles within larger, more populous schools.

The third most prevalent correlation observed is between population and tobacco import value, with a correlation coefficient of 0.31. While population dynamics may indeed influence tobacco consumption patterns, it is worth noting that education spending demonstrates a stronger correlation with tobacco import value. This suggests that while population size plays a role, education spending emerges as a more prominent determinant. Overall, the result of our correlation matrix provides a stepping stone for further analysis, indicating that education spending and population size are the driving factors of tobacco import value. In order to further understand these socioeconomic factors and dive into a causal argument, let’s employ the use of a multivariate regression analysis.

Our multivariate regression reveals important insights into the complex relationship between tobacco imports and key socioeconomic variables like, education spending, median income, and population size. The model exhibits a strong overall fit, as evidenced by the R-squared statistic of 0.433 and the statistically significant F-statistic of 18.58 (P-value <= .001 for all three independent variables). These metrics indicate that approximately 43.3% of the variance in tobacco imports can be explained by the independent variables included in our model, and the regression model as a whole is statistically significant.

When examining the specific relationships between tobacco imports and socioeconomic factors, certain patterns emerge. The negative coefficient for median income (-51.95) suggests that higher median income levels are associated with decreased tobacco imports. This finding is interesting, as individuals with higher incomes typically have greater purchasing power. Similarly, the negative coefficient for population (-0.1173) indicates that larger population sizes are loosely associated with reduced tobacco imports. This indicates a potential difference in lifestyle choices between wealthy metropolitan hubs and poorer rural regions. To explore this relationship further, our next analysis will look at median income and population by state to make an assumption on lifestyle factors.

In contrast, the positive coefficient for education spending (0.0900) appears to contradict the previous relationships. While higher levels of education spending might be expected to lead to increased wealth, and ultimately decreased tobacco consumption, this finding suggests a more nuanced relationship. It is important to note that the correlation of education spending and median income was relatively low at .15, leading us to conclude that the effect of education spending on tobacco must be independent of the assumption that funding education leads to a wealthier population. Instead, the education spending to tobacco imports relationship may be better described through the lens of cultural/social pressures, academic and occupational stress, or perceived health literacy.

In order to explore the idea that higher median income is associated with busier metropolitan regions and lower income in more rural states, we’ve created a scatterplot that shows that exact relationship. Why is this important? Well, this proves that tobacco imports may have less to do with purchasing power of commodities and more with lifestyles associated with rural or metropolitan regions. Because population and income are both inversely related to tobacco imports, we can conclude that wealthy metropolitan areas consume less tobacco than poorer rural regions, which brings us right back to education. Where is education funding going to, and could it be the social lifestyles associated with group education that are actually fostering the habit of tobacco?

Our final leg of analysis aims to account for variations over time, thus decreasing the likelihood of the relationship resulting from short-lived changes to life, like a pandemic. The first visualization of Education Spending vs Population seemingly confirms that larger and wealthier metropolitan areas tend to spend more on education. Combining this with our other finding, shown by the correlation matrix and the second visualization, that as more education funding occurs more tobacco imports are present, particularly for metropolitan areas, we can reasonably conclude that it is something about the lifestyle of metropolitan schools that is causing tobacco imports to increase.

Conclusions

So, why does education spending result in more tobacco imports? It all boils down to peer pressure. Our findings show that education spending is much more prevalent in populous states where they have a larger number of students. This population of students, however, creates an environment surrounded by the increased possibility of exposure to certain experiences, like smoking. Given there are more people in the schools, there are more opportunities that an individual in your friend group is partaking in any given experience (smoking). “Individuals develop their concept of self by observing how they are perceived by others, a concept Cooley coined as the “looking-glass self.”(1)

If another student you knew was smoking, a higher likelihood in populous metropolitan areas, it is more likely that you have access and the desire to try it just once. While many may think a little trial may not hurt, it is exactly that trial cementing a lifetime user in turmoil, “factors outside of our control can make all the difference. For many, that first puff paves the way to long-lasting addiction.” (2) The pure exposure to cigarettes is more likely to occur where people meet and gather, like a school, creating a breeding ground for tobacco usage epidemics across the U.S. It’s a fact that, “students who [have] peer pressure from their friends [are] more likely to smoke cigarettes”, so what can we do to stop this absurdly obvious problem? (3)

In the United States, our metropolitan population is consuming tobacco as a result of the increased exposure to smokers in their places of gathering like schools, where “people care about what others think… [influencing] how much they value different ideas and behaviors.” (4) To effectively control the impact of smoking in schools, implementing comprehensive tobacco control policies is essential. This includes strictly enforcing no-smoking zones on school premises, providing education on the dangers of smoking through curriculum integration, offering smoking cessation programs and support services for students, staff, and parents, and fostering a supportive and smoke-free school environment through awareness campaigns and peer-to-peer initiatives.

Additionally, collaborating with local health authorities and community organizations can further reinforce tobacco control efforts and promote healthier lifestyles among students. These proactive measures not only aim to reduce tobacco consumption but also work towards creating a positive peer environment that discourages smoking initiation and supports cessation efforts among students. By addressing both individual behaviors and systemic factors contributing to tobacco use in schools, we can effectively mitigate the impact of smoking and promote a culture of health and well-being among metropolitan students.

In short, as a community, I argue that populous metropolitan states should allocate a larger portion of their education funding to tobacco use prevention and cessation programs. Such an initiative would most certainly provide more tobacco programs to those communities with the largest populations and tobacco usage, creating a dynamic that would, over time, result in the total cessation of smoking from the majority of U.S. citizens by targeting the most influential gathering places. These initiatives would also be beneficial in the workplace, large venues, and any place where people actively gather.

To further fund these initiatives, one may consider increasing the tobacco tax and allocating those funds to anti-tobacco resources in targeted regions like schools. Taxes on tobacco have the potential to do two things: be a source of funding for targeted anti-tobacco resources, and simultaneously “Evidence…shows that price increases on cigarettes are highly effective in reducing demand. Higher taxes induce some smokers to quit and deters others from starting.” (5)