Signal Two: “I’m Just a Bill” and science can prove if I will become law
Nathan Caplan
Out of the 70,000 bills that were introduced in congress between 2001 and 2015, not even 3000 were passed(1). How can we ensure congress doesn’t spend so much time on bills that won’t pass?
Researchers have recently developed, using natural language processing (NLP), a model to score bills to see how likely such bill is to pass. It could lead to low scoring bills not getting face time and encouraging politicians to work towards passing bills with high scores. Not only that, but this technology also reports out succinct summaries of bills, increasing transparency and educating constituents(2).
For instance, this technology was used to determine the ACA repeal only had a 15% of passing(3). And, yet, congress has tried to repeal the ACA over 70 times and is still counting to try. Will it happen? Not according to NLP. Using this technology, congress should focus on bills that are proven to become law.
While these models are far from perfect, a newer generation of data scientists could expand upon these models with new advancements in NLP and machine learning tools. I believe we are a part of that cohort and should learn these tools, become educated on issues, and vote.
- https://www.quora.com/About-how-much-does-it-cost-US-taxpayers-each-year-to-fund-the-House-and-Senate-staff-Senators-Congressmen-printing-travel-healthcare-and-retirement-benefits
- 2) https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5425031/
3) https://engineering.vanderbilt.edu/news/2017/john-nay-ai-predictgov/
