Amber Smith Analyzes Abortion Care Access as a Social Network

Amber Smith Analyzes Abortion Care Access as a Social Network
Amber Smith is a Trinity senior in Sociology.

In high school, Amber Smith planned to become a physician.

“Before college, I had no intention of studying Sociology and barely even knew what the discipline was about,” said Smith, who is graduating this Spring with a major in Sociology, concentration in medical sociology, and a minor in Global Health. “During my first semester at Duke, I took an introductory class, SOCIOL 110: Sociological Inquiry, with Professor Craig Rawlings. I quickly found myself deeply interested in the material.”

Smith enrolled in more Sociology classes, including SOCIOL 255: Sociology of Immigration and Health with Professor Jen’nan Read. “I became passionate about medical sociology and social determinants of health,” she said. “I learned about all the structures in place that lead to differential health outcomes, and how there’s a lot more to healthcare and health outcomes than just presenting oneself at a clinic.”

Smith’s fascination with the social determinants of health led to her senior thesis, Zip Code Matters: A Social Network Analysis of U.S. State-Based Abortion Care Access Over Time. We caught up with Smith to talk with her about her research, her Duke experience and what comes next. (Spoiler: it’s Harvard.)

Interview has been edited for length and clarity.

Why did you decide to study abortion care access?

Reproductive health is very important to me. Historically, health care access hasn’t been equitable and there are disparities and differential outcomes for marginalized communities. Reproductive health care disparities may be further exacerbated following the Supreme Court’s decision in the Dobbs v. Jackson Women’s Health Organization in June 2022, which overturned Roe v. Wade.

Amber Smith standing in front of Conference poster
Amber Smith at the International Papillomavirus Conference in Washington, DC (April 2023), where she presented her poster “Cross-Cultural Adaptations of Scales Measuring Stigma Related to HPV, HIV, and Cervical Cancer Stigma for Use in a Kenyan Context.” (Photo courtesy of Amber Smith)

In spring 2023, I took Dr. Rawlings’ graduate level Social Network Analysis course, where I began to think about dynamic networks and the transfer of ideas, relationships and diseases. For my final project, I conceptualized the flow of abortion services across state lines as a network relationship where the states are the nodes and the aggregate of patients who are leaving one state and seeking abortion services within another are represented as the ties flowing through this network.

My thesis uses Center for Disease Control and Prevention (CDC) data from the five years preceding the Dobbs decision (2017 to 2021) to observe changing patterns in the transfer of abortion services in order to identify states that are central within the abortion network. Centrality indicates that a state isn’t just important for its own residents, but that many patients from outside the state are seeking services there. This can ultimately suggest which states are most essential to the overall network and are important to protect in order to support the largest number of patients.

What were some of your most significant findings?

States like Florida, Illinois and New York were often found within the center of the networks, indicating that they were significant providers of abortion services for their own residents and out of state members. I observed a trend suggesting that residents from states that met their abortion needs internally at this time — such as Florida, New York, Illinois, North Carolina and Michigan — were not needing to seek abortion services elsewhere, but that patients from Missouri, South Carolina, Alabama and Indiana were largely traveling elsewhere since their abortion service needs were not met within their state.

I also simulated the redistribution of the network assuming a very central node, Florida — which is currently threatening abortion care access — was removed. Just through the redistribution of Florida’s nodes, then Alabama, Georgia and South Carolina, I observed a five-fold increase in abortion services being sought in North Carolina compared to the number the state provided in 2021.

Considering bans being currently implemented in North Carolina that would further disrupt this distribution, this scenario posits an immense strain on the health care system in North Carolina and other states. This redistribution simulation demonstrates the interconnectedness of the network and how abortion bans and restrictions would not simply impact access of abortion services in the restricted states, but also greatly impact the wait times and access to services in more protected states due to an influx of abortion seekers.

Were any of your findings particularly surprising or unexpected?

This was not surprising to me, but I think a rather key finding (that may be contrary to what anti-choice lawmakers intend) is that the number of abortions decreasing within a state does NOT mean that the residents of those states are receiving fewer abortions; rather, these states may be just forcing their residents to seek services elsewhere. For example, despite the sharp decline in abortion services being provided in Missouri between 2017 and 2021, Missouri residents actually saw an approximate 30% increase in abortion services across these five years. Louisiana also experienced a similar trend, where the number of abortions provided within the state decreased within this five-year period while the number of residents leaving the state seemed to increase more than three-fold.

Network map of abortion care access in 2021
The left figure demonstrates the mapping of the United States abortion transfer network within 2021, while the right figure demonstrates this same network of the across-state transfers in 2021 embedded upon a U.S. State map. States with a severe abortion restriction of less than 10 weeks are shown as red, states with an abortion restriction before gestational age of 24 weeks are pink, states that included incomplete reporting information in the data set are grey, and states with no abortion restriction before gestational viability are blue. These networks demonstrate the immense interconnectedness of the abortion transfer with many individuals seeking abortion services across state lines, even before the Dobbs decision in 2022. (graphs courtesy of Amber Smith)

This phenomena of across-state transfers also privileges individuals with more resources and support systems, who are better equipped to seek services than others.

What are your plans after graduation?

This Fall, I am very excited to be pursuing my Ph.D. in Sociology at Harvard University! I’m grateful to all the incredible sociologists at Duke — especially my diligent and compassionate senior thesis advisor, Dr. Hedy Lee — who have fostered my interest in the social determinants of health, particularly recognizing the combination of structural racism, structural sexism and cycles of poverty as they relate to maternal and reproductive health. I’m very passionate about these health disparities and am looking forward to continuing my research.