Marginal Revolution: Where does Wal-Mart get its products?

November
S	M	T	W	Th	F	St
1	2	3	4	5	6	7
8	9	10	11	12	13	14
15	16	17	18	19	20	21
22	23	24	25	26	27	28
29	30

« Resale price maintenance | Main | Questions about Sen »

Where does Wal-Mart get its products?

Click here for a map, via Kottke.

Posted by Tyler Cowen on April 20, 2007 at 12:21 PM in Data Source | Permalink

Comments

Turkey I wouldn't have guessed, and more from Ireland than Japan? Weird.

Posted by: adrian at Apr 20, 2007 1:46:11 PM

The absence of Japan made me curious about the methodolgy used to select the products. It turns out that the sampling methodology was "guy wanders around in Walmart for a few hours while looking at arbitarily chosen labels." Sigh - this would have been very interesting if it used a methodology that constructed a representative sample.

Posted by: MattXIV at Apr 20, 2007 2:52:19 PM

MattXIV, is the methodology really that bad. It seems pretty representative to me considerign that there are more than 700 data points drawn from multiple stores. what might be bias here that can significantly skew results?

Posted by: sa at Apr 20, 2007 3:09:27 PM

I don't think number of products is a very good sample either. I think it would look different if you broke it down by the total value of all products from one country. You might see Japan and Korea move up with the high unit value of their electronics. It would also be more representative of the origin of country for Wal-Mart's overall revenue.

Posted by: Ben at Apr 20, 2007 3:54:41 PM

sa,

-Different products are allocated different amounts of shelf space and promenance of location.
-Product picks may suffer from confirmation bias regarding where Walmart is believed to get its products
-Products are location clustered by purpose
-Walmart store layouts are somewhat standard
-Path travelled in store was defined in part by personal agenda (he was shopping for stuff he needed too)
-He didn't inspect high price items kept locked up like video games

See here for a description of the method: http://www.benjaminedwards.net/Writings/walmart.htm

It doesn't matter how large the N is; if it's not representative (ie properly randomized or the complete population), you can't draw conclusions about the population. The correct methodology would be randomized picks off an inventory list.

Posted by: MattXIV at Apr 20, 2007 3:54:53 PM

Naught from Africa.

How sad.

Posted by: alphie at Apr 20, 2007 5:07:29 PM

When I look at this discussion I am reminded of what happened many years ago when WMT loved to claim it was buying American and advertised that 90% of its merchandise was American. Of course they defined it as buying from an American firm. Tennis shoes from NIKE were counted as American even though NIKE did not manufacture any shoes in the US. Moreover, they had foreign suppliers sell to US middlemen firms that WMT could buy the good from so they could claim they bought the merchandise from an American firm.

Posted by: spencer at Apr 20, 2007 8:51:49 PM

there was Kenya...

Posted by: Neal at Apr 21, 2007 6:32:06 PM

It just makes me sick who makes the most money off of one pair of Air Force Ones. The shoes cost only $8 or so to make and sell for more than $200 for some models. These shoes have more of a profit margin than even Air Jordans!

In case you want a pair, here are the Jordan release dates.

Posted by: matty at Jun 11, 2007 8:44:02 AM

liqingchao 07年8月17日

google排名
 google排名
 wow gold
wow gold
powerleveling
powerleveling
wow gold
wow gold
powerleveling
powerleveling
wow powerleveling
wow powerleveling
wow power leveling
wow power leveling
wow power level
wow power level
world of warcraft powerleveling
world of warcraft powerleveling
world of warcraft power leveling
world of warcraft power leveling