/DataStory_global-abortion

Data Storytelling and visualization project: How anti-abortion laws in US states compare with other countries around the world

DataStory_global-abortion

In-progress:

(Please excuse the mess. A more concise version of this ReadMe would be updated upon project completion. See uploaded ouline PDF to get a quick summary of the project.)

Groundwork completed (Phase 0: found data, decided on scope, wrote a pitch and an outline for the write-up for this project).

Data Cleaning/Storytelling/Visualization project:

Motivation

How do anti-abortion laws and regulations recently passed in certain US states compare with those from other countries around the world?

(Hint: it does not look good for the US.)

Inspired by this article: https://www.cnn.com/2019/05/17/world/how-alabama-abortion-compares-to-world-intl/index.html

"Alabama passed a near-total ban on abortion this week, strict enough to rival abortion rules in countries like Brunei, Guatemala and Syria."

So far I have found some good resources and dedicated groups doing research on women's reproductive health. However, none of the visualizations I've come across have put the abortion regulations by world countries and US States in direct comparison. This is what I'm hoping to achieve with this project.

There are quite a few challenges including finding and integrating the data (with matching topics and years), reducing the scope and choosing only a handful of key factors in this very complex global and socioeconomic issue, as well as choosing a type of visualization that displays verbal and geographical information well, and allows for some flexibility in scaling (can actually zoom in and see the US States on world map, or might need multiple maps for that), among others.

Notes

My goal is to generate two sets of visualizations to compare progress (or the lack thereof) among US States and other nations of the world, over the span of the past x years (depending on what data I can find - 10 years seems the most reasonable).

Phase 1 -- Data Cleaning and Preparation (time-consuming! getting different forms of raw data from NGOs, government websites, and international organizations like the United Nations)

Phase 2 -- One set of visualizations: US States and World Nations

  • How many variables to include? The main one is under what circumstances can a woman in the state/nation get an abortion. But might also need to consider the ease of access (mandatory waiting periods), cost (healthcare coverage and income - can vary by ethnicity)...)

Either choose 3 key factors to visualize for Phase 1, or build into a ranking system which would make plotting easier

--Note: Phase 1 and 2 make up a self-contained project, sufficient to write one article on

--Phase 3 and 4 are more social science research questions: dig deeper into the human rights issue out of personal interest (but might need to curb curiosity here to move to machine learning projects in the short term)

Phase 3 -- Add a second set of visualizations: How policies changed over time in the world and in the US (US landmark cases)

  • Include more variables during comparison: dig deeper into the possible reasons behind the contrasts; how do the different factors correlate?
  • also consequences of anti-abortion laws (combine with, or separate from phase 1: infant mortality rates (Status of Women report data)
    • New angle (from New Yorker article, March issue): induced abortion rates worldwide (2018 Guttmacher data) stayed about the same, suggesting anti-abortion laws lead to unsafe underground practices and affect women's health (might be hard to find data to support the second part)
    • another possible new direction to explore: Abortion restrictions and crime -- Freakonomics podcast; paper (need crime data)

Phase 4 -- Explore more variables on a global scale for a more thorough study (ie. youth crime rates and homelessness, public education, healthcare, cost of raising a child), compare with US policy change over time: more progress? less progress?

Data Acquisition and Preparation

So far I noticed there are a lot of US data for 2012 and 2013, but few recent data from State governments - the only ones on data.gov are from Illinois from 2012, and all the URLs are broken.

I was able to find World data for 2017 and 2018 (as recent as it gets). I plan to add the new anti-abortion policies from US states manually, to produce a visualization that is more up-to-date than the ones I found. I also plan on monitoring the news and updating the visualization.

(A bonus feature in the long run (after writing up a blog about the conlcusions) would be to include web scraping of news sites, or creat customized news alerts when the visualizations need to be updated.)

(The original goal of finishing Phase 1 and 2 as well as article by the end of August was delayed by health issues)

Data Visualization

Geographic map, by state/country, interactive features

August 2019

Project paused due to applications and FT program

(Pick up from: abortion-worldwide-2017.pdf pg 59, Evaluating Priorities August 2017.pdf pg 11)

July 25 Update

Installed packages for making choropleth maps with Plotly.

(Keep using Folium for now to get preliminary analysis results for article, then try Plotly?)

Installed Basemap toolkit documentation:

The matplotlib basemap toolkit is a library for plotting 2D data on maps in Python. Basemap does not do any plotting on its own, but provides the facilities to transform coordinates to one of 25 different map projections.

Basemap makes it easy to convert from the spherical coordinate system (latitudes & longitudes) to the Mercator Projection-- this projection is commonly used by popular mapping softwares

  • Because basemap uses matplotlib, need to import matplotlib.pyplot into your environment when using Basemap.

July 23 Update

Cleaned a small dataset and made the first choropleth map using Folium and GeoJSON.

Wrote outline for DQ article (data story and tutorial); longer outline draft for blog post

July 12 Update:

Edited the entire README, separated Phases 1 - 4, set shorter-term goals, more emphasis on data cleaning before trying Plotly vs. Folium

Even though Folium works for choropleth, noticed Plotly allows easy addition of hovering data boxes which could make understanding the visualizations earlier;

For example, box next to a state or country could say:

no reason given: N, health of woman: Y, rape and incest: Y -- which would mean for this particular location, an abortion request with no reasons provided would not be permitted, but if it threatens the health of the woman, or if the pregnancy is due to rape or incest, then abortion is legal in this state or nation.

Plotly cheatsheet: https://images.plot.ly/plotly-documentation/images/python_cheat_sheet.pdf

July 1 Update:

Found out the Choropleth class in Folium has multiple issues and may no longer be supported in the new version after 2018. Consider using Plotly instead?

Plot.ly:

https://plot.ly/python/choropleth-maps/

https://community.plot.ly/t/choropleth-map-in-dash/4807

General considerations for making choropleth maps: https://blog.datawrapper.de/choroplethmaps/

Updated to Folium 0.9.1 (Newest, May 26, 2019), now choropleth example works.

GeoJSON and choropleth: https://nbviewer.jupyter.org/github/python-visualization/folium/blob/master/examples/GeoJSON_and_choropleth.ipynb

June 19 Update:

Folium Choropleth Map -- works best when the colour-coded layer represents a column which contains a numerical value (such as %unemployment rate in example) - How to translate text result in data?

Folium documentation: https://python-visualization.github.io/folium/quickstart.html

Data Analysis

Further Work

References:

General

World Data

  • Guttmacher Data Center data:

(unfortunately categories don't match with the ones for US States; abortion data by region not by individual country)

https://data.guttmacher.org/regions

Data Source (table: Abortion and outcomes by region): Abortion incidence between 1990 and 2014: global, regional, and subregional levels and trends The Lancet. Sedgh G et al. 2016 https://www.guttmacher.org/article/2016/05/AWW-levels-and-trends-abortion-incidence-1990-2014 https://doi.org/10.1016/S0140-6736(16)30380-4 (can compare 10 year trend for Phase 3 or 4)

Data Source (table: Unintended pregnancies): Global, regional, and subregional trends in unintended pregnancy and its outcomes from 1990 to 2014: estimates from a Bayesian hierarchical model, Lancet Global Health, 6(4):e380–e389, Bearak J et al., 2018

*44% of pregnancies world wide are unintended!* (https://www.guttmacher.org/infographic/2018/proportion-pregnancies-are-unintended-worldwide) (2018 inforgraphic using 1990-2014 data in source above)

  • Center for Reproductive Health " the only global legal advocacy organization dedicated to reproductive rights"; has world map of abortion laws, updated in-real-time; cited by Guttmacher reports (if found it earlier would've saved time); also has global comparison over time, and state-by-state comparisons

https://reproductiverights.org/worldabortionlaws

News update August 27 - In addition to Alberta, which is traditionally conservative, one Conservative MP from the Province of Quebec is also making claims about running on an anti-abortion agenda for the election in Fall 2019 (even though the Conservative leader Andrew Sheer seems to deny that). This just shows in a country like Canada, which is often leading the way on protecting and advancing human rights, there can still be efforts to erode that progress.

US Data - by State

Installation and Usage

  • Download the .csv data file, as well as requirements.txt (generated after completing notebook, as libraries continue to get updated). Install the requirements for this project using pip install -r requirements.txt. This will install the exact version of the libraries that I had when doing this project.

  • (upgraded to pip-19.2.1 on July 25; also installed geopandas0.3.0, pyshp 1.2.10, and shapely1.6.3 - suggested by Plotly documentation for choropleth map; don't need them if only use Folium to generate visualizations)