Cosma has done a little calculatin’ on steel production and the demise of the steelworker.
He concludes
that domestic employment in steel production has collapsed largely because increases in productivity have not been matched by increases in demand.
I think he’s correct, but I don’t think that “lack of demand” is the whole story.1
Specifically, there haven’t been increases in demand for American made steel. The amount of raw (or crude) steel that has been produced in the US2 since 1950 has never been more than about 127 million tons. 3 Worldwide production (a proxy for worldwide demand), on a slow but steady increase in the 1990′s, skyrocketed in the 2000s (until last year) to the tune of 1.3 billion tons per year.
The collapse of the American steel industry in the 1980′s brought about a huge reduction in production capacity. Capacity figures are hard to come by, but I wonder if there is enough physical plant remaining in the US to make 127 million tons of steel.
[brief side trip: Let's see... reading some numbers from Steel Guru, US production in the last week of February 2008 was 2.186 million metric ton, at an estimated capacity utilisation of 91.6%. This means the weekly capacity is about 2.4 mmt, and simply multiplying by 52 (risky, but who cares?) gives a US capacity of 124.8 million metric tons. Yay! What do I win?
For comparison, this position paper from the American Iron and Steel Institute says that China's capacity is over 500 million tons per year, and is expected to soon reach over 600 mtpy. The worldsteel stats say that China produced 500 mmt in 2008.]
Back when I cared about such things, I remember there was a lot of discussion about the rise of “niche” factories — smaller, more computer controlled, less capital intensive plants designed to pump out a limited amount of steel with a minimum amount of labor, simply to avoid the US Steel problem of having to keep great honking machines staffed and running 24/7.
The newer equipment is likely much more productive not only from a labor perspective (takes fewer people to get it going and keep it running), but the control systems undoubtedly do a better job of getting decent steel out the first time. I imagine that scrap and loss is pretty low in a steel mill because everything can be recycled unless some terrible contaminant got in, but if you avoid having to rework stuff, then your productivity goes up for free.
Assuming the economy turns around before all of these newer factories are scrapped, the US steel industry will likely only ever have about 125 mmt of capacity. Factories are not easily built in the US anymore, at least not primary production plants. I could see more finishing plants, where raw steel is turned into stuff, but until the global overcapacity is used up, there’s no reason to build more mills here.
That’s why “demand” is only part of the story — there is (was) plenty of demand for raw steel, but there is (was) no compelling reason to buy US-made raw steel beyond what’s already available, and therefore no reason to build capacity in the US and no need to hire more people.
And that’s the end of my brief analysis of the raw steel industry. By the way, there’s a nice little report at the BLS on the future of steel manufacturing that I found after I’d written most of this.
- Kind of like saying LOTR is mostly about trudging around Middle Earth.↩
- I’m using the same USGS table, with some corroborating figures from worldsteel. Note that the USGS table is “raw steel” whereas the worldsteel data is “crude steel”. The numbers don’t exactly match, but I declare they are close enough.↩
- That was in 1973. Lately it’s been in the mid-90 million tons.↩