I give up on getting the table straight, <code>, etc. isn't co-operating today.

Software/online companies that benefit from the size of their userbase who then end up with near monopolies (positive feedback) in their space (ebay, craigslist, facebook to a temporary degree) probably produce much less employment per dollar of enterprise value.

The online companies are still part of a deep chain that connects directly to the physical, given that their operations run on servers and their users access their sites through hardware. That chain is still going to be a lot less robust than it is for companies like Ford or GM (I don't know what the term is - one job at Ford produces X jobs elsewhere in the economy, etc). As a wild guess I'll bet a company like Apple is close to on par with Ford in that regard.

Mostly my blood boils fiercely at the idea that the economy is a natural force instead of the collection of choices that it is in reality. And you only have to go about two questions deep down the rabbit hole before you ask why the purpose of humanity is to make and consume widgets...

you are the media you consume.

IBM and Oracle are neither internet nor industrial, but service companies. Oracle shares with Microsoft and the "internet" outfits a pseudo-monopoly on their sector; but O and M are rapidly ageing, and will go the way of IBM (and Ford and GM) : they are lumbering on, overweight and not specially innovative, their pseudo-monopoly arriving at its expiry date.

At first glance, the pseudo-monopoly influences the "enterprise value" thing very strongly.

_{It is rightly acknowledged that people of faith have no monopoly of virtue - Queen Elizabeth II}

One thing to think about in terms of the internet companies is how much they sell things to consumers, or sell consumers to various corporate buyers.

The latter, I suspect, will pretty much always need far fewer employees than the former.  Media content producers are fairly similar in that regard - a few writers can generate quite a bit of traffic, which can then be collated and sold.  Without having to worry about distribution of physical product, you just don't need that many people.

In the software industry a small number of workers can produce an immense IIR.  Three people wrote the iMac operating system and (literally) saved Apple from bankruptcy.  Now the company has something like $100 billion in cash.  No way three automotive engineers could duplicate that 15 year trajectory.

How this happens is complicated.  The quickest example is it comes down to duplication cost of the end-user product or service and the total potential market for the good or service.  Duplicating a piece of software costs around $20 (total) and has a total potential market of hundreds of millions.  Duplicating a car costs thousands of dollars and has a total potential market of several million.  

A further complication is a software company doesn't need all that many people to write a successful program.  In fact, too many people are a hindrance to a successful productization cycle.  

She believed in nothing; only her skepticism kept her from being an atheist. -- Jean-Paul Sartre

Another interesting post, Metatone.

The question I've been tussling with a few years has been somewhat related, but deals with the role of the digital age and the automation of work processes.

While this is a decades (century?) old debate, I'd be interested to learn whether (and if so, how) the process of job creation has been able to keep up with the automation in the digital age, particularly now automation is increasingly extending into the services sector (financials, banking, accounting).

That question popped up reading this and this.

I think you have to compare (and I hate this word...) ecosystems to get an accurate picture.

The Internet doesn't create jobs in the traditional sense, with full-time pay and benefits. What's happening instead - and what Lanier's book is about - is that content companies like Apple, Google and Amazon can crowd-source temporary and part-time pseudo entrepreneurs by grabbing huge tracts of content space and collecting tolls for content sold there.

So suddenly Apple has a legion of app developers and musicians, Google has video makers and bloggers, Amazon has writers (and video makers soon) who are effectively on the content production lines but have none of the security of a traditional job.

And there are hordes of support workers - designers, editors, marketers, and so on - who are either employed or hired by the content people.

Some of the content people are creating real jobs. Some are doing very well, financially. But a lot aren't.

For a fair comparison with e.g. GM you'd have to include all the support workers - mechanics, spares companies, petrol stations, cabbies, other professional drivers - who work(ed) in the space created by the car industry.

I think it's clear the Internet has been worse at creating both kinds of jobs than the old industrial economy. But I also think the biggest change has been a new emphasis on turning everyone into an heroic Galtian entrepreneur with a strong personal brand, etc, etc, and a brutal indifference to those who don't do well in that kind of environment.