Space and Time of #nobannowall

Since most of the tweets from last lab criticized Trump’s travel ban policy, I expect that most tweets will be from the U.S., in the most diverse areas and areas where Trump’s support is the lowest. Most tweets will come from the large urban areas of the West Coast (30 to 60 degrees latitude, -140 degrees longitude) and East Coast (30 to 60 degrees latitude, -100 longitude), where I believe more immigrants and liberals live in the U.S. and where Trump received the fewest Electoral College votes. I do not expect any individuals tweeting from my hometown of Syracuse , New York, even though it is a city located at 45 degrees latitude and 100 degrees longitude. With a population of only 150,000 people, I don’t expect to find many people claiming that they are from the area except for those affiliated with Syracuse University. Since most of the surrounding area supported Trump and is homogeneous, I doubt that anyone would use the hashtag which mostly was used to criticize Trump.

I have 166 mappable tweets which means that 0.25% of my total tweets were able to be mapped.

I have a total of 66,250 tweets and 166 mappable tweets which means that only 0.25% of my total tweets contained geolocations. According to Yau, “patterns, trends, and cycles of data is not always a smooth path from Point A to Point B.” For example, just having the total number of car accidents in a year will generate a different data visualization and understand than if you had the total number of car accidents in a year by time (which would tell you what times are you most likely to be in car accident [1]. Most of the tweets are from 36 degrees latitude, -120 degrees longitude which is closest to the state of California. As I already inferred, California is located on the West Coast which would have higher proportions of immigrants and liberals than other parts of the country. The issue of immigration and U.S.-Mexican border wall may relate more to California since it shares its southern border with Mexico and LAX and San Fran Airports, which are among the largest airports in the U.S., dealt with many immigrants and visa holders from the seven banned countries. All of the tweets, except for three (that were written French), were written in English. However, there were users who used German, French, Spanish, and English versions of Twitter who wrote tweets (not retweets) in English which demonstrates that #nobannowall may be an American issue or users are trying to write in English in order to catch the attention of the U.S.

Though I have a dataset of more than 60,000 tweets, I do not think that I have “big data.” Boyd and Crawford described that “big data and whole data are (also) not the same.” The fact that only 0.25 percent of my tweets have geolocations means that my data probably accounts “spritzer” or roughly less than one percent of all public tweets. In addition, Boyd and Crawford note that “those with money – or those inside the (tech) company – can produce a different type of research than those outside,” meaning that TAGS, a free, open-source software, does not collect or have insights to all of the public tweets. Tim, who is analyzing #makeamericagreatagain, only collected 10,000 tweets. I think that Tim has less tweets because #makeamericagreatagain was more of a support hashtag for Trump during the campaign. Since the campaign is now over and Trump’s approval rating is now less than fifty percent, not as many people are using the hashtag. Due the backlash that that Trump’s anti-immigration policy, as demonstrated in my data, Trump supporters may not want to show their support for him or his policy in public in fear of being characterized as racist. Harrison, who is analyzing #2a, was not in class for the lab. I expect that Harrison would not have as many tweets as Tim or I since the second amendment has not been mentioned in the media as much.

Time Map of Geolocations of Tweets with #nobannowall:

https://jennifertrin.carto.com/builder/0db43fb4-f459-11e6-ab78-0e233c30368f/embed

The map shows the locations of the tweets each day from February 3 to February 11. In order to show the location of tweets per day, I choose the “Time” (which was the date) as my widget and “Animated” as my point. I then choose the “130” steps in order to show a more continuous stream of tweets popping up. I chose the Carto, Positron background since it was simple and gray and the point color as yellow to show the contrast. Everyday, there are tweets from mostly the West and East Coasts of the U.S. which means that #nobannowall can be mostly an American issue or most of the world who is using Twitter and are exposed to tweets are from these areas in the U.S. Other tweets came in on different days from colonized countries such as Australia, New Zealand, South Africa, and various countries in Europe. which may indicate that those countries probably have the next most Twitter users or simply, they have more “documented” and “recorded” immigration that other parts of the world (It does not necessarily mean that they have the most immigrants, but there is more data on immigration in those countries as opposed to Lebanon which probably deals with a higher influx of Syrian and Iraqi refugees than we do). I am most interested in the tweets in the rural areas of the U.S. (ex. border of Nebraska and Texas and Indiana, PA) where I assumed that there was less interaction with refugees.

At first, I wanted to show the maps of geolocations in the rural areas, but no data was available for them in the 1970s and 1990s. Did we neglect data from the rural areas until recently? Below are maps from Social Explorer:

Philadelphia, PA (Zip code: 19133, Census Tract: 163, 164, 165, 174, 175, 176)

Legend showing the percentage of people over the age 14-15 who are married (Philadelphia)

Persons 15 years and over: Currently Married (1970)

Persons 15 years and over: Currently Married (1990)

Persons 15 years and over: Currently Married (1990)

San Francisco, CA (Zip code: 94107, Census Tracts: 227.02, 227.04, 226, 607, 614)

Legend showing the percentage of people over the age 14-15 who are married (San Francisco)

 

 

 

 

 

 

 

 

Persons 14 years and over: Currently Married (1970)

Persons 15 years and over: Currently Married (1990)

Persons 15 years and over: Currently Married (2010)

Kansas City, MO (Zip code: 64106, Census Tracts (1970 to 1990): 12, 13, 14, 15, 16, 28.1, Census Tracts (renamed in 2010): 11, 154, 157, 159)

Legend showing the percentage of people over the age 14-15 who are married (Kansas City, MO)

 

 

 

Persons 14 years and over: Currently Married (1970)

Persons 15 years and over: Currently Married (1990)

Persons 15 years and over: Currently Married (2010)

For the neighborhoods in Philadelphia and Kansas City, the percentage of people over the age of 14-15 married decreased by about 30-40 percent. There are many reasons why the marriage rate decreased over time in a neighborhood. If the age and racial demographics of the neighborhood remained the same, then people could be delaying marriage and marriages between people in the younger age brackets (ex. 14 to 18 years) are less common. This could be due to an increase focus on education or a societal shift in their viewpoint on marriage. More unmarried, single, young adults could also be moving to these areas. In urban studies classes, I learned that there is a revitalization of cities and more unmarried, young people are moving back to the cities, maybe attributing to the low marriage rates. In contrast, the neighborhood in San Francisco saw slight decrease in the marriage rate from 1970 to 1990, but the marriage rate remained stable and still relatively high (about 40 to 60 percent) from 1990 to 2010. The married couples that lived in the neighborhood in 1990 could have still been living there in 2010. It could also be known as a place to raise children and married couples continue to be attracted to the area. Also, it would be interesting to see how the high proportion of males compared to females in Silicon Valley have played a role in the marriage rates: Are men more likely to “settle down” or get married because the longer they wait, the less likely, they will find a partner? Since San Fran is known for its LGBTQ culture, what implications did the 2008 legalization of gay marriage in California had on the city? Did gay marriage help maintain the marriage rate? Though I do not have much Twitter data, mapping the tweets (in conjunction with mapping other data) can show me what societal changes are happening in the neighborhoods that the Twitter user lives in that may contribute to their thoughts.

In Goldsborough’s article, “Out-of-Control Immigration” in Foreign Affairs, he discussed that there is increased individual hostility towards immigrants during times of economic decline. In particular, government agencies also become worried about the influx of unskilled immigrants and their ability to assimilate to the U.S. during times of economic downturn. This pre-9/11 article notes that California politicians are particularly worried because the state does not have as many (water) resources as it used and it is worried that more immigrants are no longer improving their quality of life [3]. Tamura’s economic paper analyzing the effects of low-wage earning immigrants on a welfare state argues that high-income peoples are more likely to support immigrants that earn low wages compared to native low-wage earners. Even though low-income peoples may not need to contribute to income taxes, their utility or happiness does not necessarily mean that it is high [4]. This is interesting because it questions whether the support of immigration is a privileged ideology. These articles now make me think about the relation of the economy to my maps. Are people in the areas that are tweeting #nobannowall living in areas that are more affluent? Are these people less likely to marry because of their economic situation or because they value their economic freedom? Or are these Twitter users an outlier in their neighborhood? Are they actually living near poorer people who cannot afford to get married and blaming it on immigrants? These questions would be interesting for further analysis. According to Graham, software-sorting geographies is when through the use of code and technology, we determine which services and goods that we provide to some neighborhoods and not to others. He divides his discussion in three types of software-sorting: software-sorting mobilities which refers to the use of code to maximize profit and monitor society (ex. If an area does not have a lot of premium service paying Internet users, then the Internet company will drop speeds or quality of service, limiting further Internet access), software-sorting cities which refers to the use of GIS to make location-based decisions (ex. using Trulia to see which neighborhood has the least crime), and software-sorted streets which refers to the use of closed circuit television systems to recognize and develop a database of people’s faces (ex. track what is abnormal behavior) [5]. My data potentially shows people who do not support Trump and the mapped locations can be potentially used by the administration to track dissenters or learn where dissenters live and what can be done to encourage support. It can also be used to potentially map out in real-time where the highest concentrations of immigrants might be at. Though I did not see any Twitter users who used #nobanowall noting that they were undocumented, it would be interesting to find out if law enforcement does track hashtags like #undocumented #illegal to see if people are confessing that they are undocumented on social media (usually to show solidarity). These terms demonstrate that the use of my data can develop even more social problems (i.e. tracking illegal immigrants through social media) that would lead to even further divisions.

Works Cited

1. Yau, Nathan. 2013. “Understanding Data.” In Data Points: Visualization That Means Something, 1–42. Hoboken: Wiley.

2. Boyd, Danah, and Kate Crawford. 2012. “Critical Questions for Big Data.” Information, Communication & Society 15 (5): 662–79.

3. Goldsborough, James. “Out-of-Control Immigration.” Foreign Affairs 79, no. 5 (2000): 89-101.

4. Tamura, Yuji. “Disagreement over the Immigration of Low-Income Earners in a Welfare State.” Journal of Population Economics 19, no. 4 (2006): 691-702. http://www.jstor.org/stable/20008040.

5. Graham, Stephen D. N. 2005. “Software-Sorted Geographies.” In The People, Place and Space Reader, eds. Gieseking, Mangold, Katz, Low, Saegert, 133-138. New York: Routledge.