Wednesday, July 12, 2006

Social Networks as Instruments of Corporate Dominance

I first noticed the phenomenon that forms the basis of this post years ago when I was working with supplier data for General Electric. Based on its supply lines and distribution lines, as well as its massively interlocking and tiered systems of subsidiaries, wholly or partially owned, divisions, sub-divisions, branches, plants, and so on, General Electric must have one of the largest corporate social networks in the world.

When I say "social network" in this case, I mean by that all the other companies and organisations to which GE is connected by formal or informal ties as described above.

I got re-interested in the subject thanks to my researches on streetcars (see here and here, and here, for instance), which inevitably leads one into a byzantine web of transit companies owned by holding companies influenced by suppliers and manufacturers. Inevitably, no matter what happens, you're going to come back to another one of those nearly-identically-named titans of US industry, General Motors. I became interested in trying to map the social network of General Motors the corporation because it seemed that everywhere you turn when researching transit and transportation in the last 110 years or so, there they are.

A Brief Aside on Nomenclature: Based on cursory examinations, I think there are grounds for suspicion of any company whose name begins with "General," my prime examples being General Motors, General Electric, General Dynamics, and General Mills (founded by extreme right-wing Christian anti-sex zealots, incidentally). Anyhow...

A couple of days ago, through comments at No Capital, I found an interesting entry by the author of an equally-interesting blog, Memetix. Slomo at Memetix did a series of graphics showing the Social Network of 9/11 (September 11, 2001 terrorist attacks in the US). He has also helpfully provided me with information to get started doing my own analyses of GM's social networks.

To show you some of the scope of what I'm dealing with, in terms of a data set, I've organised my list (and it is basically a list, at this point) by Organisation (Name), Type (ie. Subsidiary, Division, Foreign Office/Plant, etc.), Place (ie. US, UK, Canada, Worldwide), Date of Acquisition/Formation, and Industry (ie. Auto Parts, Automotive, Rail, Transit). I've also got a Notes column, which contains entries such as "dissolved 1919," "liquidated 1922," and so on. Most of my information is taken from GM's Corporate History pages, but I have also used other sources, including other pages on the GM website as well. (Hmm, they have a page on Who Killed the Electric Car?, helpfully captioned "Read GM's Response: Learn what really happened to GM's EV1 electric car program. Get the facts!" Estimations of the factuality or the truthiness of GM's official line on the subject I leave up to the Constant Reader.)

I must say this for GM: Thank you so much for helpfully providing all that information. (Don't bother even thinking about scrubbing your site, should it cross your mind; I've saved copies of the relevant information.)

How much information? My guaranteed-incomplete list of companies falling within GM's social sphere of influence, mostly focusing on the years between 1900 and 1960, has 122 entries right now, and includes companies such as Frigidaire, Vauxhall, DuPont, Eastern Air Transport (the forerunner of Eastern Airlines), and The Hertz Corporation (Hertz Rent-a-Car). It spans industries including the obvious (car, truck, and other vehicle manufacturing) to rail, transit, auto parts, auto sales and service, oil, aviation, farm and heavy equipment, insurance, finance, and education.

Keep in mind that by no means is GM an anomaly; they're just particularly good at it. Most companies of any size are tied into an intricate web of connections to other companies that create certain constraints in the marketplace. Modern "vertical integration" of supply chains, where companies will own or invest in their own suppliers (also a key indicator of a monopolising tendency in a sector) is quite common.

The heavy interdependence of corporations (which is still going on) make me even more suspicious of the seemingly top-down enforced anomie and "rugged individualism" that is preached in modern Western culture (particularly the US, but also in Canada and the UK to a certain degree) as one of the main (if not the only) acceptable ways in which lifestyles can be framed. That is to say, even if you don't believe that "everyone for themselves" is the ideal way to live, you are, in this political discourse, sometimes left without ways of expressing contrary ideas without evoking those notions. (Even many of the hard-core conservatives self-identify, cynically, as "Libertarians" these days.)

That makes me wonder: Divide and conquer, versus consolidate and rule?

3 Comments:

Blogger spocko said...

For a while I was talking to some people creating "knowledge networks" where they basically try and capture the knowledge of the workers in some form so that other people in the company can use it. The employee's didn't like it, didn't want to use and and didn't trust it. "Like I'm really going to make all my great info available in the intranet so they can fire me more easily!?"
Needless to say it was an expensive boondoggle. However I did hear a fascinating story about how people really share information in a company. In one Knowledge Network for fixing printers at Xerox they started tracking who was getting the most help from the network. It turned out that most people weren't really benefiting, with the exception of two people. They went to them and asked, "Why are YOU getting and using this information and no body else is?" They said, "When I need help I go down the hall and ask Bud, the expert. It's a lot faster than looking it up on the intranet and besides Bud didn't type in his entire brain. We tell him the problem and he tells us the answer. He's an expert."

So the multi-thousand dollar expert system was nothing more than a make work program for a bunch of programmers who didn't really track how people learned and shared data.

8:45 PM  
Blogger Interrobang said...

Hm, I think you're confusing knowledge networks and expert systems with social networks. One is a sort of relational database of information, with a context-sensitive tool to use it (the expert system), and the other is a graphic and statistical analysis methodology for determining the nature and degree of relationships between entities.

Expert systems can work really well in a novice-user context where there is no (informal) knowledge network to provide support. However, with the problems in granularity, not to mention parsing, metadata, and all the other difficulties that arise when trying to implement what basically amounts to a single-sourcing solution with AI in it, it's not quite ready for prime time yet, as the saying goes.

What I'm trying to do here is build a graphic that shows the network of connections between GM and its subsidiaries/divisions/plants etc., and use the gleaned information for statistical analysis. One area of statistical analysis it might be good for is quantifying monopolistic tendencies in a sector. Just exactly how dominant are/were they, and where?

Statisticians and various kinds of scientists (like epidemiologists and biologists) do this sort of thing all the time; I'm not so sure it's been applied to any great degree in the social sciences...

12:59 AM  
Blogger slomo said...

Actually, I think the social science people came up with it before the epi/biostat/bio people did. The social science developments have been applied in management settings, but not as much for knowledge systems as for getting top-down views of the social network of employees. See my post on this subject for more information. Biologists have recently picked up the methods to get a sense of gene-gene interactions pre and post transcription. Albert-Laszlo Barabasi, one of the luminaries in the science of network topology, is now a visiting scholar at Dana Farber Cancer Institute, so that tells you a little about how the field is moving.

7:16 AM  

Post a Comment

<< Home