DS 2500
"Always carry your Green Book with you, you may need it!" was a popular saying among the Black community during the 1930s. The "Negro Motorist Green Book" informed Black travelers about hotels, restaurants, barber shops, and other things African Americans needed when traveling from state to state. It originated in New York City by Victor Hugo Green from 1936 to 1966. This book was revolutionary in that it allowed African Americans to maneuver around the country with dignity and less fear as it connected and strengthened the community. The story of Black travel in America runs deep and is rooted in generations of discrimination, oppression, and fear. The mobility of this community has been heavily policed since before the actual creation of automobiles. Bookmark eras in history, from slavery to Jim Crow to the present day have all seen the suppression of Black movement through laws, social codes, and social norms. These continue to be fought against, as history saw Black people making a way out of none. Since the rise of the Black travel movement, African Americans spend roughly $109.6 billion annually on travel. This community of people would really benefit from a modernized "Negro Motorist Green Book" of sorts that will allow them to visually track various spots around the United States that celebrate them and will always welcome them with open arms.
Through analyzing multiple datasets consisting of Black population rates, and Black-owned businesses among others, this project will attempt to quantify a more interactive estimate of states in America that are most safe for Black people to travel to. The goal of this project is to first track the movement of Black people over time in relation to the rise of "sundown towns" (all-white municipalities or neighborhoods in the United States that practice a form of racial segregation by excluding non-whites via some combination of discriminatory local laws, intimidation or violence) in order to lay a visual foundation of the importance and necessity of this project, and then be able to provide users insight to the best and safest locations for them to visit.
This project aims to use data science techniques to analyze different factors such as Black-owned businesses per state, false incarceration statistics, and overall Black population per state among others. The project will provide a useful tool for providing Black people with ideal travel solutions through a modernized, interactive "Negro Motorist Green Book".
National Geographic: Black Travel Across America
The Travel History of Black America
Why Thes Professors Helped Create a "History of Black Travel Timeline
We will use the Percent Black Population for Every State in the USA dataset from Kaggle, which contains the percentage of populations in the USA that identify as Black from 1790-2018 at 10-year intervals. This dataset will be useful in visually tracking the movement of Black people in America, which we can directly compare to the presence of sundown towns over time.
Features: years from 1920-2018 in ten-year intervals
| State/Territory | 1920 | 1930 | 1940 | 1950 | 1960 | 1970 | 1980 | 1990 | 2000 | 2010 | 2018 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| United States of America | 9.9% | 9.7% | 9.8% | 10% | 10.5% | 11.1% | 11.7% | 12.1% | 12.3% | 12.6% | 13.4% |
| Alabama | 38.4% | 35.7% | 34.7% | 32% | 30% | 26.2% | 25.6% | 25.3% | 26% | 26.2% | 26.8% |
| California | 1.1% | 1.4% | 1.8% | 4.4% | 5.6% | 7% | 7.7% | 7.4% | 6.7% | 6.2% | 6.5% |
| Florida | 34% | 29.4% | 27.1% | 21.8% | 17.8% | 15.3% | 13.8% | 13.6% | 14.6% | 16% | 16.9% |
| Massachusetts | 1.2% | 1.2% | 1.3% | 1.6% | 2.2% | 3.1% | 3.9% | 5% | 5.4% | 7.9% | 8.9% |
black_pop_data_dict = {"1920-2018": "years in ten year intervals with the Black population percentage"}
black_pop_data_dict
{'1920-2018': 'years in ten year intervals with the Black population percentage'}
We will also use the Police Killings US dataset from Kaggle, which contains every fatal shooting in the United States by a police officer in the line of duty since January 1, 2015, prepared by The Washington Post. This dataset will be one of the metrics used to analyze the "safeness" of a particular state in the United States specifically for Black people by tallying the number of Black people killed by police.
Features:
| Name | Date | Manner of Death | Armed | Age | Gender | Race | City | State |
|---|---|---|---|---|---|---|---|---|
| Tim Elliot | 02/01/15 | shot | gun | 53 | M | A | Shelton | WA |
| Lewis Lee Lembke | 02/01/15 | shot | gun | 47 | M | W | Aloha | OR |
| John Paul Quintero | 03/01/15 | shot and Tasered | unarmed | 23 | M | G | Wichita | KS |
| Matthew Hoffman | 04/01/15 | shot | toy weapon | 32 | M | W | San Francisco | CA |
| Michael Rodriguez | 04/01/15 | shot | nail gun | 39 | M | H | Evans | CO |
police_data_dict = {"Name": "name of the person killed by the police",
"Date": "date that the killing took place on",
"Manner of Death": "how the deceased died",
"Armed": "what the deceased was armed with if applicable",
"Age": "age of the deceased at the time of the event",
"Gender": "gender of the deceased at the time of the event",
"Race": "race of the deceased at the time of the event",
"City": "city that the killing took place in",
"State":"state that the killing took place in"}
police_data_dict
{'Name': 'name of the person killed by the police',
'Date': 'date that the killing took place on',
'Manner of Death': 'how the deceased died',
'Armed': 'what the deceased was armed with if applicable',
'Age': 'age of the deceased at the time of the event',
'Gender': 'gender of the deceased at the time of the event',
'Race': 'race of the deceased at the time of the event',
'City': 'city that the killing took place in',
'State': 'state that the killing took place in'}
One issue that might indicate a problem with the dataset's reliability is its heavily weighted focus on more recent police killings. The dataset does not provide any information on police shootings prior to 2015, which can negatively skew the data. It is also important to note that the dataset does not explicitly distinguish if the killing was or wasn't "warranted", not grossly unjust, or an abuse of power. Furthermore, the different metrics that we are using (the number of Black-owned businesses, the number of fatal police killings, and the overall Black population) to distinguish if a state is safe or not may not be the best holistic view to answer that question.
This format lends itself to machine learning, as we could write a program that will allow users to input metrics about themselves (age/gender) in order to return the best states for them to travel to along with recommended black businesses in the state. We will cluster states into a "safeness" scale based on user inputs. Doing so will allow us to rank states and properly recommend safer traveling spots for users.