The economics major in me is drooling over a new site called Swivel whose founders describe the service as “YouTube for data.” Users upload data sets via a clean, intuitive interface, add metadata (description, source, tags), and publish for the world to parse as it wishes. I’m excited because a) data sets can be notoriously difficult to dig up (and either time or cost prohibitive), b) the field could use a nice injection of creativity (a la Steven Levitt) which scores of data sets in a single place are bound to generate, and c) Swivel makes data fun.
Last night I uploaded a set of data that I generated my senior year in college for my thesis on Wal-Mart’s economic impact post-entry on a metropolitan area (or “MSA” in geek speak). This was tough data to collect. In order to run regressions against employment, retail sales, and number of retail establishments, I needed Walmart store opening data and Walmart wasn’t terribly forthcoming. After a brief stint calling stores directly, I ran the numbers and recognized that I’d be at it for months so I needed a plan B. I discovered with some trial and error that if I took the company’s branded Rand McNally Road Atlas, which contained store lists by state, I could call Walmart customer service and ask them to look up store info by store number in their database. Two long weeks of calling customer service yielded what I wanted - data on nearly 1600 Walmart store openings across 28 states in the Midwest and South which represented Walmart’s core area of growth. Persistence pays off. I couldn’t look at a phone for a month (and that’s okay because I needed to run my regressions and write my thesis!).
Here’s the link to my Walmart store opening data on Swivel. I hope the company becomes “Wikipedia for data” - a much used service with watchdog groups hounding the accuracy of uploaded files. I’m imagining professors asking, “You got your data set where?!”
____________________
The following is from Appendix D of my thesis (dated 4/9/03), a surprising aside to my core work:
1,043,970 workers are employed by Wal-Mart in the United States. The Census Bureau’s estimate of the population on July 1, 2002 is 288,368,698.[1] A simple calculation reveals that about 1 in 276 Americans is a Wal-Mart employee.
A more detailed analysis proves insightful. According to the Current Population Survey run by the U.S. Department of Labor’s Bureau of Labor Statistics, the employment-population ratio (for people 16 & over) is 62.4% for February of 2003.[2] Though the Census Bureau has not yet released age-specific population estimates for 2002, the 2001 numbers may provide a base upon which to extrapolate.
According to the Census’ estimates, the population on July 1, 2001 was 284,796,887 with 220,309,468 (or .7736) 16 & over.[3] Applying this ratio to the 2002 estimate, we find that on July 1, 2002, there were approximately 223,072,503 people 16 & over in the United States. Of those, 62.4% (from the CPS data), or approximately 139,197,242 were employed (admittedly, there is a time discrepancy between the population estimate from July, 2002 and the employment-population ratio from February, 2003 but for the purposes of this derivation the two numbers may simply act as current data).
Therefore, if 1,043,970 workers out of 139,197,242 were Wal-Mart employees, approximately 1 in every 133 workers in the United States is a Wal-Mart employee.
[1] http://eire.census.gov/popest/data/national/popbriefing2002.php
[2] http://www.bls.gov/cps/home.htm
[3] http://eire.census.gov/popest/data/national/tables/asro/US-EST2001-ASRO-01.php
Add New Comment
Viewing 5 Comments
Thanks. Your comment is awaiting approval by a moderator.
Do you already have an account? Log in and claim this comment.
Do you already have an account? Log in and claim this comment.
Do you already have an account? Log in and claim this comment.
Do you already have an account? Log in and claim this comment.
Do you already have an account? Log in and claim this comment.
Do you already have an account? Log in and claim this comment.
Add New Comment
Trackbacks