According to the study, "A reasonable person might wonder...whether the $80 million per year that the smelters seek for the purpose of retaining 1,375 jobs might be better spent on retraining the smelters' displaced workers, who are already skilled."

The study questions the long-term viability of the North American smelter industry in the light of long-term cost pressures affecting the country's coal and electricity industries. Short-term support only becomes viable if there is prosperity in the long-term, but if such support merely postpones an eventual shutdown, then such resources would 'be better spent on more promising long-term prospects'.

Rio Tinto Alcan claims that, while unprofitable, its Sebree smelter provides for 1,834 jobs both directly and indirectly. It claims that these jobs create a total value of $198 million per year. Christensen disagrees, arguing that such benefits are overstated. "We believe that these benefits are substantially overstated, particularly because the indirect impacts are likely to be a fraction of those claimed," the study says.

Christensen has no issue with Century Aluminium's claims for its Hawesville facility. The smelter creates 771 jobs directly plus 471 indirectly and creates a total value of $95 million per year.

For Christensen, there are two big risks for Big Rivers Electric Corp if the smelters are forced to close. First is losing the electricity contract or having to renegotiate them. Second is Big Rivers' dependence upon coal-fired power stations, which are becoming more expensive to run in the light of environmental considerations.

In many ways, Big Rivers is caught between a rock and a hard place. Smelter closure would impact power rates, but taking the discounting route would mean raising rates for other customers.

Chistensen believes the only silver lining for Big Rivers if the smelters did close would be the avoidance of a multi-million dollar spend on retrofitting its generators to meet EPA environmental requirements.

While world production of aluminium has been rising steadily since 1974, in the USA it has dropped from 5.5Mmt in 1974 to 5Mmt in 2011. Market share in North America has dropped from 41% in 1974 to 11% in 2011.

In 2000 there were 24 smelters in operation in the USA. Today that figure is just 10.