UPDATED: Cheapest Exchange-Based QHPs 29% Less Expensive Than Off-Exchange (BEFORE tax credits)!
Wow! This article is from back in June, but it's a heck of a find; insurance-tracking website HealthPocket ran an extensive study comparing the ACA exchange-based policies against their off-exchange equivalent plans, and their findings were pretty striking (and I'm surprised that this hasn't received more coverage):
Overall the least expensive metal plans from United Healthcare, Aetna, Cigna, and Assurant were significantly more expensive than the least expensive metal plans available on state exchanges. Across the bronze, silver, and gold metal tiers, the least expensive plans offered by the four off-exchange carriers were over 40% more expensive on average than the least expensive plans on the exchanges. This suggests that if these carriers entered new exchanges in 2015, then they would not usually be competitive with the cheapest on-exchange plans unless they substantially lowered their current premiums. It is important to note that these premium costs do not factor premium subsidies, which are only available for on-exchange metal plans.
That's right: This is before subtracting the tax credits, which 85% of 2014 exchange enrollees received.
By the way: For those wondering why I list it as "on exchange = 29% less expensive" rather than "off exchange = 40% more expensive", that's simple math: If the exchange-based policy is $100/month and the off-exchange variant is $140/month, that means the latter is 40% higher (140/100)...but the former is "only" 29% lower (100/140)).
UPDATE: Richard Mayhew at Balloon Juice correctly points out that this is something of an apples/oranges comparison, since the off-exchange policies are often targeted specifically for a higher income-level market in the first place, so likely include better options such as wider networks, etc. than their on-exchange counterparts.
Ironically, this is exactly the same point that I was making back in June when I broke down Avik Roy's "Obamacare made premiums jump 49%" claims.
While I was mostly just passing along an interesting study, it was wrong of me to jump to the conclusion that the policies are identical; my apologies. However, it's still an interesting/noteworthy study, so I'm leaving it up for consideration.