Edwards expansion savings claim seems questionable

In the final analysis, policy-makers should take Gov. John Bel Edwards’ surprising announcement in his state of the state speech that Medicaid expansion would save the state $100 million in fiscal year 2017 with a grain of salt, as well as understand it would predict costs, not savings, in future years.

Only two authoritative, publicly-released studies have occurred on this policy, both by the Department of Health and Hospitals prior to Edwards’ inauguration, and only the first, released in 2013, went into a year-by-year, component-by-component study of the issue. It presented four scenarios that depended upon the rates of enrollment of the uninsured population specifically newly eligible because of expansion, of increased enrollment in other Medicaid programs because of the publicity surrounding expansion, and of enrollment of the insured population from the private sector switching now eligible to switch to Medicaid. A second report a year later confirmed, utilizing data gathered from other states’ experiences, that the most costly scenario most closely fit reality. The extremely detailed report predicted for the last half of 2016, when Edwards wants to start the increased coverage, this would cost the state $38 million; by 2023, the state would pay an additional $373 million that year.

The public could access both reports – until shortly after Edwards took office. Echoing the communist world when a new regime would wipe away evidence from the previous, the reports mysteriously disappeared from the DHH website, and now locatable online only through other unrelated archiving sites.  Two months later came Edwards’ pronouncement that alleges a positive $142 million dollar swing in the forecast impact from the first for the last half of this year.

But the Edwards Administration at this writing has not made this report available to the public. It appears that only information to the media has gone out, in a very rough outline that presents no data analysis or detail. While this does not automatically invalidate any reputed conclusions from it, unlike with the previous two that have sourced data and lay out all assumptions, independent analysis cannot assess the quality of the work.

However, the extraordinarily sketchy information made available through the media provides at least a limited review of it – an exercise that immediately raises questions about the quality of its conclusions. Keep in mind as well that the figures are according to fiscal year, not calendar years, which can confuse as the escalating state shares occur on a calendar year basis.

From what can be derived from it, it says the extra state contribution – none in 2016, then 5 percent in 2017, 6 percent in 2018, 7 percent in 2019, and 10 percent from there on out – from sources not spelled out would be $51 million for FY 2017, presumably incorporating 2.5 percent of the total cost. It also alleges that moving current Medicaid recipients into expansion (meaning lowering reimbursements from roughly 32 percent) would save (that is, the federal government paying more, even as the overall taxpayer burden nationally increases) $24.8 million, that uncompensated care costs (for the same reason) would decline $37.7 million, that extra state compensation for charity hospitals would decline (for the same reason) that saves $89.3 million, and that prisoner hospitalization (for the same reason) balances would benefit with a $3.5 million reduction, for a total of $155.3 million. The net result comes close to $100 million, at $104.3 million.

With few details given of the Edwards Administration assertions this makes comparisons difficult, but one possible concerns uncompensated care. The 2013 report predicted these costs avoided, represented by dollars spent in the disproportionate share hospital program, would be $35 million – essentially the same figure. With less certainty, that report also estimated administrative costs of the program at about $20 million annually, while the Edwards Administration insists the hiring of nearly 250 staff would cost only about 15 percent of that.

The 2013 report seems more credible on that account. The recent figures allocate only a bit more than $11,000 per new person hired, when the salaries plus benefits easily could exceed four times that. Again, with the lack of details released by DHH, it’s unknown what figure was used in its computations, although it claims providers themselves would foot the rather low startup costs figure (who then presumably would pass these along to the public).

But the savings figure most in dispute would be that shifting of the state subsidy to pull up safety net hospitals, because that does not depend upon Medicaid expansion. That would be a byproduct of it, but the state could realize savings simply by declaring itself out of the charity business and not paying that extra to treat the 138 to 200 percent federal poverty limit segment, making those individuals buy their own insurance.

So, when adjusting down or out these figures, it’s entirely debatable that expansion in and of itself “saves” Louisiana any money at all this next fiscal year. Leaving these objections aside, however, even accepting the assumptions of the Edwards Administrations’ numbers, indisputably after next fiscal year expansion will cost the state more than by refusing it.

If we assume the $51 million figure represents the state portion of expansion, next fiscal year it goes from 2.5 percent to 5.5 percent, or an extra $61.5 million, while the alleged savings remain roughly the same. By the time the proportion has risen to 10 percent, the state will pay more than any asserted savings, even using what appears to be the optimistic Edwards Administration assumptions.

Again, without any detailed information from Edwards’ DHH it’s impossible to make meaningful comparisons, but the broad assertions made certainly appear based upon much more optimistic cost assumptions than the previous detailed studies. And that would run against the field of play regarding Medicaid expansion, which has shown for most states costs well exceeded initial expectations.

Policy-makers should show wariness over accepting the Edwards Administration numbers that appear unsourced and unexplained, contradict more detailed if dated analyses, and, even so, show that expansion will end up a money-loser over time (and this does not even consider expansion will not provide better care, if not produce worse outcomes, than presently available to the uninsured). Any debate on this issue should revolve around data and analysis made publicly available, not by asking to take broad, essentially unverifiable assertions at face value.