Beginning next year, people without health insurance through their employer will be able to buy it monthly on "exchanges." They are encouraged to buy the insurance, or else face a penalty for each month and family member not covered. Let's suppose for the moment that the penalty is enforced (even though law limits how the IRS can enforce it).
By law, people can let their insurance lapse for three months with no penalty. Moreover, they can choose when the three month lapse occurs, and accelerate and/or delay medical procedures to fall outside the lapse interval, without concern for being denied when they restart insurance because of "pre-existing conditions."
Arguably, essentially everyone without employer health insurance would find it in his interest to purchase insurance for only 9 months per year, and cram all of their medical procedures into the nine months of their choosing. Medicaid already experiences this phenomenon (namely people enroll if and when they get sick), and Medicaid participants do not pay premiums. Exchange participants will be paying really money.
There is no free lunch in the aggregate, so insurance companies will have to increase their monthly premiums by 12/9 so that they collect the same annual revenue from participants. When this happens, do we say that health insurance costs went up by 33 percent? Or that they were constant?
If you chose the latter, then you are saying that health insurance costs should be measured by premiums per person, where "persons" denominator includes people who are not currently insured. But then the historical trend for health insurance costs should be measured the same way, and would be less steep due to the fact that uninsurance rates have increased over time.
Another problem with choosing the latter is that exchange participants (suckers?) who do not play this game -- they pay their premium all 12 months of the year -- will pay more for health insurance on both a monthly and an annual basis.
Both proponents and opponents of the ACA will want to know whether the new law fulfilled its promise of "bending the cost curve," but my guess is that an army of economists will be needed merely to do the measurement -- to measure health insurance costs before and after 1/1/2014 in ways that are comparable.