Insurers take on the hospitals

1. In California, the number of effective hospital competitors varies widely.

Thompson treated one organization that controlled more than one hospital in a rating region as a single effective competitor.

He found that a few Covered California gold plans operate in regions with 16 to 18 effective competitors, a significant minority in regions with nine to 11 effective competitors, and the majority in regions with two to six effective competitors.

See also: 10 wild places to replace a knee

2. The correlation between high hospital market concentration and high premiums was not perfect.

In California, the cheapest gold plan in the most competitive market was only about $200 per month, and the most expensive was close to $300 per month. That cheapest gold plan was the cheapest of all the gold plans Thompson studied, but the most expensive cost almost as much as the average gold plan in the least competitive markets.

In the moderately competitive markets, the cheapest plans were almost as affordable as the cheapest plan in the most competitive market, and the most expensive plans were about as expensive as typical plans in the least competitive markets. 

See also: Insurers go for the bronze buyers.

3. Efforts to consider factors that might throw off market-to-market comparisons did little to change the effect of hospital market concentration on premiums.

Thompson says he tried to adjust the data using measures such as average local housing costs to see if that would change the picture. He says including that variable and other variables did not appreciably reduce the estimated dependence of premiums on hospital market concentration.

See also: 19 PPACA exchange gorillas: 9 are pushing rates higher in 2015.