Don Norman writes about accidents, and lays them at the feet of a profit-driven culture. I take him to task and he replies. Esther Dyson chimed in via private email to suggest that Don was talking about a short-term focus. Note that Don was not talking specifically about workplace accidents, but I am, just to keep focus.
Workplace safety is a more complicated topic than it first seems. Anyone might say "What could be wrong with increasing safety?" An economist would reply that it might be too expensive. I want to address this in three directions. First, that some safety is free, and some is not. Second, that there are four sources of money to pay for safety. And third, that it should be workers who demand safe workplaces, or not.
The cost of safety
Increasing the safety of the workplace might or might not cost money. For example, locking down a circuit breaker when you're working on the circuit it controls costs almost nothing. You had to touch it to turn it off anyway. Putting safety instructions on a device is the cost of a label. Other types of safety are going to be more expensive. For example, turning off a circuit breaker when you're working on a piece of equipment that has its own power switch is going to be more expensive. You'll have to stop what you're doing, go to the circuit breaker, turn it off, and go back. Or insisting that someone not use the top step of a ladder (the not-a-step) means that they need to locate a taller ladder.
Providing extra safety might be a capital cost (e.g. adding a railing), or it might be an operational cost (e.g. going the long way around, avoiding the place where a railing is needed). These costs are accounted for differently. Capital might not be available for increased safety. The cost of production might not bear the lessened productivity.
Just as safety is rarely free, so will accidents rarely be free. There is always a cost involved in accidents. Perhaps a trained worker is lost, or idled. Perhaps equipment is broken. Perhaps equipment is idled while the worker's corpse is removed from the equipment. Nobody worries about the accidents that are more expensive than the safety needed to prevent them. The profit motive will cause employers to eliminate those types of accidents. The kind that gets peoples panties tied in a knot are the ones where safety is more expensive than the accident.
For the remainder of this article, I will only speak of the latter kind. The former is boring.
Paying for safety
In the short run a company can pay for safety by raising prices, reducing entrepreneurial profits, by reducing dividends, or by reducing wages. A product doesn't change just because it's been made more safely. Entrepreneurs may have a personal desire to not hurt their employees, in which case the safety becomes cheaper than the accident. Boring! Same thing for the capitalists. Boring! Same thing for the workers, particularly since it's them who get hurt in the accident. If they're willing to do this, then the problem again becomes boring, because the cost of the accident exceeds the safety.
The problem with any of these four sources of safety funding is that they're limited. Customers will not put up with unlimited increases in price with no improvement in the product. Once a (not otherwise productive) capital expenditure on safety has been made, no further source of capital is available unless capitalists want even more safety. Once a worker's wages have been cut because they're less productive because they're being safe, further gains are impossible without further cuts. No worker is interested in working at a perfectly safe job for no pay.
I must note here that it's possible to force all employers to adopt safety measures, through the coercive power of government. That doesn't create money from nothing. What it does is force all employers to pay for safety using the same mix of the same four funding sources listed above. Competition will select the most efficient mix. I think that workers have deluded themselves into thinking that only entrepreneurial and capitalist profits are reduced. That's possible, but not likely. Entrepreneurial activity will move to a less regulated (e.g. a more intrinsically safe) business. Capitalists will make further investments in less regulated businesses. Workers who prefer safety will not. And customers are just "workers" the day after payday.
Workers must demand safety
Coercing safety measures on employers is probably not the right idea, for several reasons. A worker might prefer wages to safety. A worker has a direct interest in maintaining a safe workplace, whereas the government's interest is indirect. Workers can detect unsafe practices and eliminate them. Other workers in the same workplace who are not at the same risk will subvert the safety measures, because they don't want their wages to go down.
The best protection of workplace safety is a robust economy with low costs and high wages. Ironically, this may mean less safety...or more safety. Workers will be able to avoid unsafe workplaces unless they pay well enough to cover the risks. Workers who prefer money over safety will seek out those jobs, e.g. explosives deliveryman.