This Contingencies article about age band compression under the ACA is an interesting look at potential future premiums based on the ACA's 3:1 age band ratio rule. [Contingencies magazine is published by the American Academy of Actuaries - it's not a special interest publication, so I tend to trust their articles more than something coming from a source that is either biased for the ACA or biased against the ACA.]
The article looked specifically at premiums for younger Americans in the 20 - 29 and 30 - 39 age groups, and the likely impact of the 3:1 premium ratio. They found that for 20 - 29 year olds who are not eligible for subsidies, premiums are expected to increase by 42% over what they would have been without the ACA regulations. And for people age 30 - 39 who aren't eligible for subsidies, the increase is expected to be 31% (things get much better if you're on the upper end of the age scale: for those age 60 - 64, the increase will be around 1%, and some might see a slight decrease). Currently, many states have no official age band limitations for setting individual health insurance premiums, although 5:1 is a pretty common ratio. The ACA mandates that premiums for the oldest individuals cannot exceed three times the premiums of young people. Although this maybe makes things more equally spread in terms of overall prices that everyone pays, it results in a shifting of burden away from older people and onto younger people. It seeks "fair" at the expense of actuarial justification. There are arguments for and against both ways of calculating premiums (ie, based on actuarial claims use of each age band, using a 3:1 ratio, or even the way it's occasionally done with no age-rating at all). Sometimes it's hard to separate one's own experience from the mix and avoid a personal bias - a person who is 58 might feel a lot more positively towards the 3:1 ratio than a person who is 27.
If we were look at this strictly in terms of "fairness" and pooling risk, we'd probably need to remove age-based claims data entirely from the equation, and just calculate an average price and charge everyone the same amount. In the broadest sense, health insurance would be about pooling everyone's expenses together and dividing them up across the whole population, with each person paying the same premiums regardless of statistically-predicted claims experience. But "fair" doesn't always equal practical, and the reality is that younger people (who generally use a lot less medical care than older people) are more likely than older people to go without health insurance even at the current 5:1 premium ratio. In 2010, 18 - 24 year olds made up 9.7% of the total US population, but 16.2% of the uninsured population. 25 - 34 year olds were 13.6% of the overall population, but 23.7% of the uninsured population (data).
For people with an income under 400% of FPL, regardless of age, the ACA premium subsidies are set up so that out-of-pocket premiums will not exceed 9.5% of income (for those with incomes over 400% of FPL, there are no out-of-pocket premium limits - the premiums are what they are). But even though the limits are only based on income, not age or health status, this sentence in the Contingencies article is particularly telling:
"The difference between young and old at similar income levels is that younger individuals at a given income level are much less likely to find it economically rational to purchase coverage if it takes up 9.5% of their income, while older individuals have a greater expectation of health care cost spending as a percentage of income."
If you're right at the cut-off for premium subsidies (about $46,000 for a single individual) and you're 26, paying 9.5% of your income for health insurance might look like a bad deal. You might be inclined to pay the penalty instead (which may be too weak in 2014 to be effective). On the other hand, if you're 60 years old and earning that same income, paying 9.5% of it for health insurance might seem like a bargain - especially if you've been used to paying even higher premiums in the recent past (some of our carriers in Colorado have said that they anticipate a slight drop in premiums for the oldest age band once the 3:1 rule takes effect).
Another issue that we should probably take into account is how income levels change with age. As with all generalizations, this isn't true of everyone, but for the most part, income tends to gradually increase with age until the late 50s, when is starts to decline again as people begin to retire. So a person who is 49 will probably have a higher income than a person who is 27. Combined with the higher likelihood that the older person will have health insurance claims in the near future, it does make sense that we should use some sort of age-banded premium determinations. Setting the ratio at 3:1 was an effort by the writers of the ACA to bring an element of fairness to the process and protect older Americans from unaffordable health insurance premiums. It remains to be seen, however, whether or not the result will end up being more disparity in terms of the distribution of the uninsured population by age.