Paul Graham, a well-read programmer, venture capitalist and blogger, offers a view of unions filtered through Silicon Valley. It’s worth thinking about, if only because the tech industry has long viewed itself as having avoided the “mistakes” of older industries.
The reasons unions were able to gain so much power, and economic benefits for their members in the early part of the 20th century in the US, he argues, is because the big manufacturing industries — cars, steel, etc. — were essentially in the same position as startups today. They were fast-growth industries, where profits were strong enough, and the cost of not growing high enough, that it made perfect economic sense to pay for expensive but necessary labor. Expensive workers were the equivalent of today’s expensive-but-critical servers for today’s startups.

If you looked in the head of a 1950s auto executive, the attitude must have been: sure, give ’em whatever they ask for, so long as the new model isn’t delayed.

In other words, those workers were not paid what their work was worth. Circumstances being what they were, companies would have been stupid to insist on paying them so little.

Union salaries were the same kind of aberration as overpaying dot-com workers slinging HTML in the bubble days, he argues.

There are a few serious problems with this analysis.

First, it’s worth a look at the history of growth industries. The early industrial revolution in England was a time of unprecedented growth, to the point where people who were making money had to make wildly speculative investments just to find a place to put their cash. Similarly, there was a grow-or-die mentality; all capital owners saw that if they did not keep up, build new factories, find new markets, the opportunity costs would be extraordinarily high.

Where were the workers? If they got a living wage for working 6 or 7 day weeks, 13 hours a day, they were lucky. This was a time of starvation, abject misery and poverty. Not of comfortable workers with benefits.

Second, it seems dangerous to cite the dot-com bubble as a time of conventionally rational decision-making. Tech companies were funded by millions in venture capital — not profits — and had been given a Wall Street-blessed free pass on ordinary economic measures like earnings or business models. These companies were not operating under traditional economic incentives that even companies in growth periods traditionally must meet.

Ultimately it comes unsexily down to supply and demand, I think. Graham’s analysis hints at this — in a time of fast growth, qualified labor inputs can be in relatively short supply, and so their price can go up, or they can apply the relative power gained by their position to creating unions, which lock in some of these price gains.

Graham argues that people trying to explain the decline of unions fall back on a chimerical  idea of our “fallen civilization,” in which the people of today simply don’t have the moral courage of our ancestors.

That’s ludicrous, of course. There are many good, ordinary explanations for the decline of unions, focusing on the increasing mobility of capital (ie, the ability to build a factory overseas), the use of offshore labor, and changing regulations at home. Go back to supply and demand. If a capital owner wants to create a business in America, the skilled labor pool is still relatively small. But there’s no need to do that anymore, she can go anywhere in the world, so the labor pool is dramatically larger.

It may well be that the day of unions as we know them is over. But saying that there is no place for them, no place for groups which can match the naturally greater negotiating power of employers, is dangerous. We’re slipping into an increasingly stratified economic system, better to find ways to ameliorate this problem, rather than justify increasing disparity of wealth.