Abortion rights in the united states


Final Projects, Visualization

by Joanna Thompson & Taylor Norton

Introduction

Even though the 1973 Roe v. Wade decision legalized access to abortion in the United States, there have been continuous attempts to partially or fully restrict this access. Most recently, the confirmation of Brett Kavanaugh, who supports the overturn of Roe v. Wade, as well as the passing of the so-called “heartbeat bill”, have brought the issue to the forefront of the political landscape. At the time of writing, there are 47 bills being considered in Congress pertaining to reproductive rights. Our visualization project explores abortion rights and policy in the United States with the aim of identifying threats to access and providing a general understanding of the contemporary abortion rights debate.

As women who recognize the importance of bodily autonomy and support the right for people to make decisions about their own bodies, our goals in creating this project are to showcase the current legislation surrounding abortion rights in the United States and to support abortion rights activists and people who are negatively affected by restrictions to abortion access. We hope to raise awareness about the precarious nature of abortion access in the United States and the threats to abortion access throughout history up to the current day.   

Topically, our project covers significant events in abortion policy and reproductive rights history from the 1800s until the current day. We highlight three major court cases (Roe v. Wade, Planned Parenthood v. Casey, Carhart v. Stenberg), as well as a few specific issues, including mandatory wait times for abortions and the recent “heartbeat bills”.  While there is certainly no shortage of topics that we could cover in a project on abortion and reproductive healthcare access and policy, our topics were often dictated by the data. These topics, as well as limitations and challenges encountered, will be discussed throughout this report.

To see the interactive Tableau Dashboards with qualitative portions, please go to our site and use the password “rights” to access it.

Process

Prior to creating the visualizations, we both put out an open call on our personal Facebook pages explaining our project idea and asking our connections to share topics that they would be interested in learning about or thought were important enough to include in such a project. This initial call led to suggestions such as visualizing the location of former, current, and future clinics as well as fake abortion clinics (so-called “abortion crisis centers”) and co-mapping clinic location and average household income. The participants in this open call were any of our Facebook friends and were not intentionally selected based on their interest or experience with our topic, though it is likely that only those particularly interested or knowledgeable about abortion policy and access would have felt compelled to comment on our statuses calling for thoughts and suggestions.

During and after the visualization creation process, we asked several women people to navigate our site and ask for their feedback overall to see if there was any unclear or missing information. We also gave people specific questions to see if they could find the answers based off of the information on our site. Users selected were both women and men with an interest and investment in abortion policy and access, though none of them identify as abortion activists. With more time, it would be preferable to interview abortion activists, as they would have more meaningful information to give us about significant dates, accuracy of data, and things in our content/data that might be missing or may cause our visualizations to be unintentionally misleading or confusing.

User 1 was our first woman UX participant. She spoke to us near the end of the visualization creation process, when we had a full site page, but not everything was placed on the page. User 1 recommended that each of our graphs (rate, wait times, etc) be paired with a featured Supreme Court case, but because we did not have a Supreme Court case for each issue highlighted in the graphs, and because we are highlighting history, we decided to maintain the loose chronological ordering. The same user also recommended that we include more links and sources at the bottom of the page. We kept the page intentionally simple, and User 1 found the page easy to navigate and understand.  

User 2 was our second woman UX participant. We asked her to do two things: navigate our finished site to familiarize herself with the site and to find information based off of a question we posed. Overall, User 2 liked the downward scroll of the site. When she came upon the Tableau Map, she wanted to double click on the map rather than the “+” or “-” buttons to zoom in and out, which stalled her browsing for a bit. After she looked through the site, Taylor asked her if she could find out out many states had mandatory wait times of 48 hours or more. She were able to navigate to the wait time map and count the states by color; however, she did not realize that she could click on a certain time to highlight a specific time/color. While having no familiarity with Tableau may have spend up her interaction with the site, she were still able to navigate it. The last comment she made was that she really liked the black background with the white text because it made it easier to read with her dyslexia.

Our third user was a partially colorblind male participant. User 3 said that he had no issues with the color gradients of the visualizations except for the blues in the Heartbeat Bill map. At first he thought he could see all the colors but when it was brought to his attention that 5 colors were present, it took a minute to find the last shade. Because of this, we made the lightest blue a gray color and then changed the blue shades to be more distinctive. After this, he was able to discern the 5 different colors. Another thing he noted was that he got lost in the tabs of the Tableau dashboards. Interestingly enough, no other users went to click on the tabs on top. Because of that, we decided to keep the dashboards in place instead of using screenshots in order for there to be an interactive component.

User 4 was also a man. He first commented on the need to make author names more prominent, as he was unclear who the authors were. He enjoyed the black background, noting that the contrast between the black background and white text made it easier to read. He also requested that the text and the visualization be side-by-side rather than on top of one another so that he didn’t have to scroll up and down to both read the description and look at the visualization. Like User 3, User 4 navigated through the Tableau workbook tabs instead of sticking to the single dashboard, and I had to help direct him back to the original graph.

The process of designing our visualizations was dictated by the type of datasets we had. Taylor was able to create the Supreme Court cases visualizations through researching qualitative data through sources such as oyez.org and uscourts.gov. While we reached out by email and telephone to both Planned Parenthood and the Guttmacher Institute for their lists or abortion clinics and other compatible datasets, we weren’t able to get a complete dataset of locations due to privacy and safety concerns. Joanna was able to get the state of residence and occurrence data from the Guttmacher Institute and used that to create three visualizations. We were able to make our own datasets (csv files) for both the wait time visualization and the heartbeat visualization after researching different case histories.

Visualizations were created using Tableau, Timeline, and Adobe Photoshop. While making the visualizations, we experimented several times with background color and the orientation of the graphs and other visualizations to see what worked best and adjusted them as we got more feedback from users. We also considered appropriate colors to use for each visualization and adjusted them after receiving feedback in order to create a stronger association between the data itself and the data with the users.

Rationale

For our final project, we wanted to work with several different concepts surrounding abortion rights and policy. Because of this, we decided to create a variety of different visualizations. We ended up with a timeline, two choropleth maps, and several bar charts. Using Squarespace as our platform, we placed our visualizations in chronological order on a vertical-scroll site.

We decided to start with a timeline (Fig.1 ) because we had so much information covering a large timespan. We also wanted users to have more context in regards to the more current information we had as well as to see how long this conversation around abortion rights and access has been occurring.

A screenshot from the starting point of a timeline entitled "History of Abortion Policy".
Fig. 1

There are three bar charts after the timeline (Fig 2-4) showing more detailed information in regards to pivotal Supreme Court cases. They are colored by how they voted on the case with a qualitative section beforehand to explain the cases and the decisions reached in more detail. Because discussions surrounding abortion rights, and more broadly reproductive rights, have often occurred in gendered spaces, we wanted to highlight the gender of each Supreme Court at the time of the cases. We used classic symbols of men and women colored either green or red based on how they voted. We did make note on our site that these symbols were perpetuating even further the binary frame around both gender and abortion rights; however, we also wanted to use this as a jumping off point to start changing how people talk about abortion in regards to gender.

A bar chart of how Supreme Court justices voted on the 1973 Roe v. Wade case
Fig. 2
A bar chart of how Supreme Court justices voted on the 1992 Planned Parenthood v. Casey case
Fig. 3
A bar chart of how Supreme Court justices voted on the 2000 Stenberg v. Carhart case
Fig. 4

The next two charts (Fig. 5-6) are based off of the same dataset depicting how many abortions occur per state and how many residences in that state have an abortion. Because we didn’t have information on who was traveling from where for this visualization, we decided not to do a map. We were more concerned with showing which states had a large number of women either travelling out of or travelling in to a state for an abortion. Because of this, we decided a bar chart would be able to convey that number better than through a map. One potential visualization we could have done if we had clinic locations was a heatmap showing which areas had a large change in women who travelled; however, we weren’t able to get this information in a complete dataset.

Bar chart showing the number of women that have abortions in each state and the number of abortions held in each state
Fig. 5
A bar chart depicting which states have women leaving or coming in for abortions
Fig. 6

Next is the wait time choropleth map (Fig. 7). This visualization is really measuring two things: whether or not a state had a mandatory wait time before having an abortion and if so, how long that wait was. Originally, we had all of the states colored on a scale of blue; however, after UX research, we changed it to grey along with a scale of blue, which helped make a bigger distinction because some states have no wait. This change in color also supported the idea that there was more than one variable being measured.

A choropleth map shaded blue in order to depict how long of a wait time states have, if any
Fig. 7

At first, we thought we would display the heartbeat bill visualization as a barchart. However, after some UX research, we changed it to a second choropleth map (Fig. 8). We did this after considering how more specific information could be gleaned from this format, such as which states did or did not have legislation being passed along through their respective house. We also had all of the states colored a shade of red. However, after presenting the map, it was clear that we should change the color of some states because we were measuring two different things: whether or not a state had any legislature brought to the floor of the  house and if so, how far it got. Because of this, we changed the states who had no legislation to grey and kept the states who had legislation in some form shades of red. We also added more information to the qualitative section in order to clarify what each measure meant in terms of where the bill was in the state legislation.

A choropleth showing if states have tried to pass "heartbeat bills", and if so, where in the process these bills were in each state's house
Fig. 8

Finally, to make our designs more cohesive, we made the sheets on Tableau into dashboards. This allowed us to have black backgrounds that blended in with the site background as well as shifting the legends on to negative space. We were also able to move Alaska and Hawaii on the maps in order to have a tighter frame on the contiguous states.

Findings

  1. Ordering/Format of Visualizations Matters

The ordering or format (particularly the spatial relationship between the text and the visualizations) of the visualizations seemed to be important to several users. User 1 recommended that visualizations be grouped by similar content or concerns. She specifically suggested creating infographics about court decisions that led to the circumstances represented in our bar graphs/maps. Because we felt committed to representing important parts of history, we ultimately decided not to create infographics for court cases other than the ones seen on the final site page, but this could be a good way to expand upon this project.

User 4 didn’t like the fact that he had to scroll up and down in order to read the description, view the visualization, and then scroll back up to refresh your memory of the description (particularly in cases when the description can assist in understanding the visualization). He recommended placing the visualization and description side-by-side so that they could be viewed simultaneously.

  1. Effectiveness of Black Background

The black background allowed us to do two things: to incorporate the different formats of our visualizations in a more cohesive manner with our site as well as to make it more accessible to different types of users. We also had a male who is colorblind look at our site to see if he could determine the difference between shades in the choropleth maps and the bar chart with blue and green bars next to each other.

Recommendations for Revision

If we had more time, we would like to include more data on this site. One topic we were interested in but couldn’t find complete datasets for was the location of clinics throughout the U.S. If we could pair that visualization with the bar chart of states of residence and occurrence, a stronger connection could have been made on how far people have to drive in order to obtain a safe abortion. Additionally, during our final presentation, another student inquired about the possibility of connecting use or availability of birth control in certain areas to the abortion rate. While this was not our focus, it is certainly connected, and would be worth investigating if we could find the proper dataset.

Because abortion policy and access is a controversial topic in the United States, talking about experience with abortion can often put women and their healthcare providers in danger. This risk makes it incredibly difficult to find publicly available data, as most reproductive rights organizations take measures to keep data points (and by extension, women and their providers) safe. An ideal version of our site would work with more datasets as well as to incorporate other sites in order to link users to resources and news sources.