Remember when the Obama administration promised that the Affordable Care Act’s expansion of Medicaid would be awesome?

The idea behind this part of the law was that the expansion of the pool of people eligible for Medicaid would help reduce the number of uninsured and that these new enrollees would be relatively cheap when compared with those already in the program. Well, that’s not what happened.

A colleague of mine at the Mercatus Center, Brian Blase, recently reported a shocking statistic: The Department of Health and Human Services just “found that the ACA’s Medicaid expansion enrollees cost an average of $6,366 in (fiscal) 2015 — 49 percent higher than the $4,281 amount that the agency projected in last year’s report.” That’s quite a mistake.

If you’re thinking this type of mistake is nothing new — because government officials always tend to underestimate the cost and overestimate the benefits of the programs they’re pushing — you’re right.

In this case, the error comes from having assumed that the state officials who oversee the Medicaid program are good stewards of our money. It turns out they aren’t, however, because they simply respond to the terrible incentives built into the law.

Under the ACA, states that expand Medicaid have to expand eligibility beyond the previous subset of low-income Americans — e.g., pregnant women — to cover all adults with incomes of up to 138 percent of the federal poverty level.

The expectation was that the cost for these new enrollees would be much lower than the cost for the people already in the system, precisely because the original enrollees were poorer and likelier to be chronically sick.

In truth, the states didn’t care much about the cost because the law gave the states an incentive to expand by offering to pay for 100 percent of that expansion for a few years. That share would drop thereafter but would still be at least 90 percent.

In fact, the states that have expanded were so insulated from the cost, thanks to the enhanced funding, that they “set outrageously high capitation rates — the amount government pays insurers — for the ACA Medicaid expansion population,” Blase explained.

“The rates are much higher than the amounts for previously eligible Medicaid adult enrollees,” he added, “and suggest that states are inappropriately funneling federal taxpayer money to insurers, hospitals and other health care interests through the ACA Medicaid expansion.”

The damage for taxpayers in this one simple underestimation error alone is roughly $20 billion. However, that’s just the tip of the iceberg — an error made for just the cost of the Medicaid expansion.

Back in March, the Congressional Budget Office increased its projection of overall federal spending for Medicaid significantly, by $146 billion over the 2016-25 period.

This time, the error was the dramatic increase in the cost per existing enrollee.

Congress could do something about this new ACA scandal. The only acceptable solution would be to radically reform a program that — according to new research by academics Amy Finkelstein, Nathaniel Hendren and Erzo Luttmer — only returns 20 to 40 cents of value for each dollar spent on new enrollees and fails to demonstrate that it provides real health benefits to those enrolled.

Ending the ACA’s Medicaid policy would be a big improvement. Blase also favors changing the financing structure to incentivize value instead of higher and higher spending.

He concludes, “We should then allow states a great deal of additional latitude to set eligibility, premiums, copayments, benefit structure and more.”

— Veronique de Rugy is a senior research fellow at the Mercatus Center at George Mason University, a columnist for Reason magazine and the Washington Examiner, and blogs about ecomomics for National Review. Click here to contact her, and follow her on Twitter: @veroderugy. Click here to read previous columns. The opinions expressed are her own.