Black American Populations and Drug Overdose Death Rates in the United States

Introduction

The opioid crisis is affecting countless Americans nationwide as drug overdose death rates significantly increase. Opioid intake disproportionately impacts people of color from low socioeconomic backgrounds. For instance, according to the National Center for Drug Abuse Statistics, among men, Black or African Americans are the most likely to die from an overdose (NCDAS, 2023). 

This paper expands on that body of literature by utilizing state-level data from 2020 to evaluate the association between the number of Black Americans and drug overdose death rates. Ultimately, this report found a positive correlation between Black American populations and drug overdose death rates and statistical significance with data excluding West Virginia, a state with a unique dynamic in the opioid epidemic.

Methodology

This analysis relies on a dataset containing the 50 U.S. states (n=50). The independent variable is the percentage of Black Americans living in each state per 100,000 people. Black Americans made up 12.4% of the total population of Americans in 2020 (U.S. Census, 2020).  For the purposes of this analysis, Black Americans are defined as solely Black, and have no mixed identity with other races or ethnicities. The dependent variable, drug overdose death rates, includes drug categories such as heroin, morphine, codeine, semisynthetic opioids, methadone, fentanyl, tramadol, cocaine, and psychostimulants (which includes methamphetamine) (Ahmad et al., 2023). Both variables are scaled per 100,000 people for each state, and the dataset’s variables did not contain any missing or erroneous data. For this report, a descriptive and distribution analysis as well as a bivariate regression were conducted.

Description of Data

Table 1 displays the distribution of variables for the sample. Data for each variable was taken from each of the 50 states. The average number of Black Americans across all 50 states in 2020 was 10 per 100,000 people, with a range from 1 in Montana to 37 in Mississippi. For drug overdose deaths, there were an average of 29 overdose deaths in 2020, ranging from 10 in South Dakota to 81 in West Virginia. West Virginia, however, is an outlier in this data. The mean for both variables is higher than the median, demonstrating a positively skewed distribution.

Figure 1 contains a histogram displaying the frequency distribution of the dependent variable, drug overdose death rates. As mentioned previously, the data has a positive skew and, due to high overdose death rates in West Virginia, Kentucky, Delaware and elsewhere, the mean is being pulled above the median in the overall distribution.

Bivariate Results

The scatter plot in Figure 2 illustrates a positive association between Black Americans and drug overdose deaths. A notable outlier in the upper left quadrant is West Virginia, which has the highest number of drug overdose deaths but only 4 Black Americans per 100,000 people in the state. Nevertheless, this data is consistent with sources demonstrating the disproportionate effects that drug overdoses have on Black Americans in the United States.

While it is easy to see the positive association in the scatterplot above, the bivariate regression results were not statistically significant at the 10%, 5%, or 1% levels. However, the regression results reveal that one more Black Americans per 100,000 people in a state is associated with a 0.255 increase in the drug overdose death rate. The R2 of 0.033 signifies that the proportion of Black Americans in a state explains about 3.3% of the variation in the drug overdose death rate, a relatively small percentage. The data excluding West Virginia, however, tells a different story.

To analyze this data further, this report also features bivariate results excluding West Virginia, due to its presence as a notable outlier. The bivariate regression results were revealed to be statistically significant at the 10% and 5% levels. One more Black American per 100,000 people in a state is associated with a 0.346 increase in the drug overdose death rate, a 0.091increase compared to results containing West Virginia data. The R2 of 0.090 signifies that the proportion of Black Americans in a state explains about 9% of the variation in the drug overdose death rate, a larger percentage than reported previously with WV data.

Conclusion

To conclude, this analysis found that the strength of the relationship between Black Americans and drug overdose deaths with all 50 states is weak. As mentioned, this model did not generate a statistically significant result at standard levels. In other words, there was not enough evidence to reject the null hypothesis. But, when excluding the state of West Virginia, the strength between the two variables became relatively stronger and yielded a statistically significant result. The model containing 49 states was statistically significant at the 10% and 5% levels and had enough evidence to reject the null hypothesis. 

That said, this weak association could be the result of omitted variable bias. Some independent variables that could be controlled for that would improve the robustness of this analysis include education level (no high school diploma), income/wealth levels, and access to healthcare. Additionally, state-level factors that affect drug overdose rates and should be considered when determining associations between Black Americans and drug overdose rates. It is also crucial to recognize outside factors such as historical and contemporary legislation that have disenfranchised Black Americans and other racial minority groups, and thus likely made them more susceptible to drug overdoses.

Suzzy Ndiforchu

+ posts

This author does not have any more posts.