from the cbo letter to charles rangel, on the latest scoring for the new house bill hr 3962:

By 2019, CBO and JCT estimate, the number of nonelderly people who are uninsured would be reduced by about 36 million, leaving about 18 million nonelderly residents uninsured (about one-third of whom would be unauthorized immigrants). Under H.R. 3962, the share of legal nonelderly residents with insurance coverage would rise from about 83 percent currently to about 96 percent. Roughly 21 million people would purchase their own coverage through the new insurance exchanges, and there would be roughly 15 million more enrollees in Medicaid than the total number projected for Medicaid and CHIP combined under current law. (Under the bill, CHIP would no longer exist in 2019.) Relative to currently projected levels, the number of people purchasing individual coverage outside of the exchanges would decrease by about 6 million, and the number obtaining coverage through employers would increase by about 6 million.

Under the proposal, certain employers could allow all of their workers to choose among the plans available in the exchanges, but those enrollees would not be eligible to receive subsidies via the exchanges (and thus are shown in Table 2 as enrollees in employment-based coverage rather than as exchange enrollees). CBO and JCT expect that approximately 9 million people would obtain coverage in that way in 2019, bringing the total number of people enrolled in exchange plans to about 30 million in that year. Roughly one-fifth of the people purchasing coverage through the exchanges would enroll in the public plan, meaning that total enrollment in that plan would be about 6 million.

That estimate of enrollment reflects CBO’s assessment that a public plan paying negotiated rates would attract a broad network of providers but would typically have premiums that are somewhat higher than the average premiums for the private plans in the exchanges. The rates the public plan pays to providers would, on average, probably be comparable to the rates paid by private insurers participating in the exchanges. The public plan would have lower administrative costs than those private plans but would probably engage in less management of utilization by its enrollees and attract a less healthy pool of enrollees. (The effects of that “adverse selection” on the public plan’s premiums would be only partially offset by the “risk adjustment” procedures that would apply to all plans operating in the exchanges.)

not sure how fast the premiums would go up for the public plan, but state high-risk pools have been a poor solution for lack of insurance for the sick simply because their [sick people] costs, and therefore their premiums, aren't exactly tiny. if the public plan essentially turns into a national high risk pool, then yes, expect the premiums to go up, even with risk adjustment.

also, the public plan is going to have pay back, starting in year 1 and amortized over 10 years, the startup money the govt loans to it, and this is going to have to be added to the premiums too.