Often we hear the military has the largest budget, usually twice as large as that of any other country. But could the United States government really be so naive as to spend so much money on fancy fighter jets and multi-billion dollar aircraft carriers? When was the last time we heard about a dogfight (a fight between two fighter jets)? Maybe the problem has been the context under which we always hear about the amazingly large budget of the Department of Defense (DOD). Every article we read about the United States budget focuses on how much money is given to each government agency. I decided to take a different route to find the answer to the question of whether the budget of the DOD is truly the largest.
Instead of looking at the overall budget of the DOD and compare it with the budget of other countries, I went back as far as the records found on the Data.gov website would allow me. As you will soon find out that would be 1962. Here I found the expenditure, based on governmental receipts, of each government agency. I thought this would be a much better comparison of government spending as I would be looking at the actual expenditure of the agencies and not the amount given to the agencies. I decided to compare the data of the departments most often spoken of in today’s political arguments: Department of Education, Department of Homeland Security, Department of Labor, Department of Justice, Department of Defense and the Department of Health and Human Services.
Using Tableau, a data visualization software, I input the data into the application and received a visualization that supported the arguments made by every article I have read. The DOD expenditure grossly overshadowed all the other budgets combined. Except for the Department of Health and Human Services which came in at a slightly distant second. The visualization turned out to be inaccurate as I was looking at all of the expenditure from 1962 through 2014 combined. After filtering through the data I ended up with a line graph that showed the DOD as the agency with the most expenditure during the years 1962-1967 and a second line graph that showed the Department of Health and Human Services along with the DOD as the two agencies with similar expenditures for the years 2009 – 2014. With the Department of Health widening the gap. The three graphs led me to create the final visualization which you see here:
This visualisation adheres to Segel & Heer’s (2010) of telling a story with data visualization. First at the top you have a full visualization which represents the expenditure of each of the selected agencies from 1962 – 2014. This visualization at a glance draws the user’s attention. They can immediately see the history of the expenditure of the selected agencies. The story? The government spending has shifted over the past fifty-five years and will continue to do so in favor of the health and services of its citizens. The viewer can then look below the line graph to find two treemaps, one that shows the expenditure in 1962 and the other expenditure in 2014.
The target audience for this visualization is that of users who simply want to know if there have been any changes in government spending, and if so, how has it changed over the years. For that reason the visualization is kept very simple and to the point. There are little to no details at the initial level of presentation. Users who are looking for further information can select particular points in the line graph to get an exact number on the expenditure. Further still users can also turn on the names of the bureaus, which will let the users see how the major agencies allocate their expenditure. Not overwhelming the user with initial information and allowing them to delve deeper into the data if they choose to was a way to avoid Few’s (2006) pitfall of displaying excessive detail.
The treemaps at the bottom of the page were a direct result of wanting to compare 1962 to 2014. Unfortunately with six agencies to compare using a bar graph would have been too visually stimulating. The bargraph would also have taken more space along the bottom of the page and competed for the user’s attention against the line graph at the top of the page. Creating two small treemaps allowed for the line graph to be the main story of the visualization and because of its size take advantage of the viewer’s pre-attentive process. Similarly the treemaps also allow for the viewer to immediately recognize the proportional change that had taken place from 1962 to 2014. The reasoning behind using treemaps and a line graph goes beyond their simplicity. According to Few (2009) it turns out pre-attentive data analysis is more easily done when the data is encoded in boxes and lines as it allows the user to quickly compare and see the differences.
Another reason for selecting only six agencies was the ability to differentiate the agencies. Having only six categories allowed for six very different colors to be assigned to each agency. In turn this allows the viewer to quickly memorize what color belongs to what agency and, because the same scheme is used for all three visualizations, apply it to all the entire visualization. Overall the visualization represents the simplicity of the data which it represents. Using as little embellishments as possible most of the visualization is used in the representation of data. I believe Tufte would have very little to complain about.
Few, S. (2009). Now you see it: Simple visualization techniques for quantitative analysis (1st ed.). New York: Analytics Press.
Segel, E., & Heer, J. (2010). Narrative Visualization: Telling Stories with Data. Retrieved February 28, 2015, from http://vis.stanford.edu/files/2010-Narrative-InfoVis.pdf