Resistbot users sent over 670,000 messages to Congress about health care.

It’s hard to believe it’s been just over a month since the dramatic midnight vote on “skinny repeal” ended one of the GOPs best chances of killing Obamacare. Because this is 2017 and literally anything is possible, we wanted to look back at the healthcare debate and analyze it through the lens of Resistbot data before another round of repeal attempts inevitably takes place.

Resistbot launched on March 8th, 2017 to enable U.S. citizens to contact their elected officials via SMS texting. Through the end of July, our users sent over 1.6 million distinct messages that generated over 4.3 million fax pages. We analyzed these messages, sent between March 16th and July 31st, 2017, sorted them by topic, and discovered that 42% were related to healthcare.

The volume and share of messages related to healthcare sent on a daily basis shifted considerably in response to the news cycle. Looking at total volume of messages, we can see that the AHCA passing the house solicited the strongest reaction from Resistbot users. Between May 4th and May 5th, Resistbot users sent nearly 88,000 distinct healthcare-related messages to their members of congress.

It is worth noting that the user base was considerably smaller during this time-period — roughly 320,000 users vs. ~780,000 at the time of writing.

Between the AHCA passing the House on May 4th and “skinny repeal” failing in the Senate on July 27th, approximately 53% of messages were related to healthcare. On June 23rd, a day after Senate Majority Leader Mitch McConnell (R-KY) released the draft text of BCRA, 87% of all messages sent that day were related to healthcare. We noticed that a considerable majority of messages were related to healthcare during the month of July, when the Senate held debates on various Obamacare repeal measures.

In many of the states where Republican senators played pivotal swing roles in voting against Obamacare repeal bills, Resistbot users received targeted reminders from the application at key phases in the legislative process, nudging them to message their elected officials. Likely as a result, users in these states tended to send disproportionately more healthcare-related messages versus nationally (42%).

In the following nine states, healthcare messages comprise a majority:

• Kansas (55%)
• West Virginia (54%)
• Louisiana (54%)
• Ohio (53%)
• Arizona (50%)
• Arkansas (50%)
• Nevada (50%)
• Indiana (50%)
• Alaska (50%)

In eight of the states listed above, Medicaid was expanded under the ACA. Proposed cuts to Medicaid likely played a large role in the failure to pass a repeal bill.

Beyond healthcare, several topics are top-of-mind for Resistbot users, including the Russia investigation, the environment, and education. We hope to analyze these topics in detail soon!

Methodology

Because Resistbot users can choose to send the same message to multiple elected officials at once, only distinct messages were counted in this analysis. A distinct message is defined as a unique combination of user, date, and message text.

To determine which messages were related to healthcare, we first pulled a random sample of 10,000 distinct messages (anonymized to protect user privacy) and fitted them to a topic model via latent Dirichlet allocation (LDA) using the tidytext package in R. The topic model resolved several groupings, including messages related to the Russia investigation, the environment, education, and healthcare. Healthcare messages contained the following keywords, which were used to filter the original set of 1.6 million distinct messages:

AHCA, American Health Care Act, ACA, Affordable Care Act, Obamacare, Trumpcare, Health care, Healthcare, Medicaid, Medicare, BCRA, Better Care Reconciliation Act, preexisting condition, Planned Parenthood, health insurance

(variations in spelling, spacing, punctuation, and capitalization were considered)

To determine states that expanded Medicaid coverage under the ACA, we referenced the Kaiser Family Foundation’s dataset “Status of State Action on the Medicaid Expansion Decision, January 2016.”