Here's a recently announced coal plant closure: http://www.cortezjournal.com/article/20161001/News05/1610099...
...
the local job options could be pretty limited in far-western Montrose County once two of its major employers close their doors, eliminating what are currently 55 jobs at the plant and 28 at the mine.
...
According to the EIA, the Nucla plant generated 416,150 MWh in 2015 for an average annual power of 47.5 megawatts: http://www.eia.gov/electricity/data/browser/#/plant/527 That's an abysmal productivity per employee (or a fabulous job source, depending on your perspective): 0.86 real annualized megawatts per employee at the plant ; 0.57 megawatts per employee if you include the mining jobs.
A well-sited utility scale solar farm like Desert Sunlight can produce an average annualized power of 147 megawatts with just 15 full time employees, for a ratio of 9.8 megawatts per plant employee.
Replacing old coal plant generation with equal MWh from favorably sited solar requires only ~9% of the number of permanent employees, if Nucla is a typical case for an older coal plant. That's a nightmare number if you're relying on the coal industry for your livelihood, or a great opportunity if you're looking for cost reductions in a competitive electricity market. The fixed O&M costs for solar are already the lowest of any utility scale electricity source; once the capital costs get low enough it has winning costs across the board in any reasonably sunny area. That's for instantaneous cost of generation, anyway, which still has plenty of competitive opportunities in most regions before storage has to be considered.
Every industry loves to brag about "creating jobs" when beseeching governments but that's another way of admitting to "creating costs" (that are then passed on to buyers). The explosion in solar and wind jobs is a temporary effect of rapid expansion; most of that employment is in installation rather than running installed plants, manufacturing equipment from raw materials, or mining the raw materials. There's a lot less work in operating a solar plant than building it and the operational phase is a lot longer. If the American solar industry had already reached steady state (just enough construction to keep new capacity matched with retirements), I believe that the full life cycle labor intensity per MWh would be significantly lower than that of a steady state coal electricity system. And the labor intensity is declining further because a few major trends in PV tech (higher module efficiency, longer module life, longer inverter life), though not explicitly about reducing labor inputs, imply significant further labor reductions.