Indeed no deaths, but very expensive.

K-19 is an example of a loss of coolant event due to failure of a pipe, resulting in 22 deaths. (Not civilian, but it's an example of the forces at play- civilian reactors have more layers of protection of course).

The point being, it happens.

> very expensive

So... jobs created, businesses paid - sounds like a functioning economy to me.

Not so great if you have to price insurance against such accidents into your operating costs; while competing against other means of electricity generation.

Three mile island is believed to have caused between 0 and 2 deaths from very low level radiation exposure to the general population. Better cancer treatments means this could very well be 0, but we don’t actually know.

"zero confirmed deaths after 43 years" sounds a lot like "zero confirmed deaths" to me

You couldn’t confirm any deaths if the expectation was 2,000 deaths from cancer either. Even for cancers that are highly correlated with specific causes an environmental risk could have caused the original cell’s mutation(s).