In its yearly projections of the effect of the ACA on insurance coverage, CBO has always had a category called "nongroup and other" that I found frustratingly opaque. Included were off-exchange individual (nongroup) market enrollees, people on disability Medicare, and a miscellany including student health plans and Indian Health Service coverage.
In new projections released today, CBO breaks out the major component parts of this grab-bag -- for the first time, as far as very close watchers whom I checked with can remember. The breakout is illuminating in a couple of respects. Here are the numbers for 2016:
These tallies shed some light -- more or less by negation -- on CBO's past overestimates of the number of exchange enrollees.
In 2010, CBO projected that the exchanges would have an average monthly enrollment of 21 million this year, 17 million of them subsidized. In 2013, the forecast had risen to 22 million for this year, with, again, 17 million subsidized. In fact, as of the end of open enrollment on Jan. 31 of this year, 12.7 million were enrolled, 10.5 million of them subsidized. CBO now forecasts average monthly enrollment this year of 12 million, with 10 million subsidized (a conservative estimate of attrition, compared to past experience).
As has been pointed out before (e.g., by the Kaiser Family Foundation), one major factor in the overestimate was CBO's estimate of how many fewer people would have access to employer sponsored insurance as a result of ACA enactment.* In 2013, CBO estimated that ESI enrollment would be 6 million less in 2016 than it would have been under then-current law and a million below the 2013 tally. Last year, CBO forecast that in 2016 ESI would be 7 million below the without-the-ACA benchmark, a sudden drop from just 2 million below in 2015. The 2016 drop has not happened (though CBO is projecting a decrease of 3 million over the next ten years, which is nine million below what CBO estimates would have been the total if there were no ACA). In fact ESI has remained nearly constant, or perhaps upticked, since ACA enactment. Kaiser's 2015 employer survey found that 57% of firms offered coverage in 2015 compared to 55% in 2014, a difference Kaiser dubbed statistically insignificant. As for CBO, last year it estimated that ESI enrollment in 2015 was 156 million; its new estimate for this year is 155 million. The pool of potential marketplace enrollees may be smaller than forecast if ESI growth has not been inhibited as much as CBO envisioned.
Another possible factor was CBO's estimate of shrinkage in the off-exchange individual market. The breakout sheds some new light on that. And here, CBO was not far off. The 2013 projection had the "nongroup and other" category shrinking by 4 million. It was estimated at 25 million in 2013, with a counterfactual "prior law coverage" (i.e., with no ACA) estimate of 26 million for 2016. In the current report, add together Medicare (for disabled adults under 65), "Other," and "outside marketplaces," and you have 23 million. That suggests shrinkage of 3 million rather than the forecast 4 million -- a pretty negligible difference. If the Medicare and "other" categories were predictable and on-target, the off-exchange nongroup total of 9 million is about a million larger than anticipated -- perhaps a negligible difference with these rounded estimates.
CBO also only slightly overestimated the percentage of on-exchange buyers who would be unsubsidized -- 23% in their 2013 forecast for this year, compared to 17% in actuality. Because they thought the whole market would be bigger, they anticipated 4 million unsubsidized exchange enrollees this year, while in actuality there are 2.2 million at present (2 million in their rounded-to-the-million current estimate for this year).
The failure of the drop in ESI to materialize is thus the only category in CBO forecasts that explains a large chunk of the smaller-than-expected marketplace population. The rest of the shortfall would seem to be explained by a) fewer eligible people than expected enrolling, because of some combination of cost and continued ignorance of what's on offer, and b) an overestimate of the eligible population. And in fact, CBO is now forecasting 19 million exchange enrollees by 2019, compared to a 2013 forecast of 25 million by that year.
How much of the shortfall is due to slower-than-expected takeup rather than smaller-than-estimated population size? Well, CBO estimates total nongroup enrollment (on and off-exchange) at 22 million this year and 27 million in 2019-20, falling back to 26 million in 2021-24. So that suggests about 83% capacity at present (splitting the difference between 26 and 27 million). In 2013, CBO's estimate for the nongroup market in 2019 was perhaps 32 million, depending on where a forecast shrinkage of 3 million the "nongroup and other" category compared to 2013 was concentrated.**
CBO's reduced projection may be a combination of a down-sized estimate of the potentially eligible population and an assumption that more eligible people will reject the offering than previously assumed. CBO is now assuming that among the uninsured, "35 percent would not purchase insurance to which they had access through an employer, through the marketplaces, or directly from insurers." I don't know whether they've published comparable estimates in the past (and will see if I can update here).
Combining CBO's estimate of the size of the off-exchange nongroup market with Kaiser's estimate of potential marketplace enrollment (or nongroup enrollment) indicates slightly less (though comparable) progress in reaching the eligible population. Kaiser currently estimates 27.4 million potentially eligible marketplace enrollees nationwide -- including those who are ineligible for subsidies. Add CBO's estimated 9 million off-exchange enrollees to the 12.7 million enrolled on-exchange,*** and you have 21.7 million enrollees in the nongroup market. That would be 79% of Kaiser's potential market total, if you substitute "potential nongroup market" for "potential marketplace enrollees."
For the most part, you might as well. The off-exchange enrollees are in the same risk pools as the on-exchange customers, with the exception of 1-2 million (Charles Gaba's estimate) in grandfathered and grandmothered pre-ACA plans and a small number of ACA-compliant plans sold off-exchange by insurers who do not participate in the state exchange in question. So the total ACA-compliant nongroup market may be at about 75% capacity by Kaiser's measure. As I've noted previously, by Kaiser's estimates the exchanges have enrolled about 70% of the subsidy-eligible population. Hence off-exchange takeup would appear to be somewhat higher.
CBO is also forecasting that current CHIP funding levels will not be renewed when they expire in 2017, and that CHIP enrollment will accordingly be cut in half, from 6 million to 3 million. That would be terrible policy -- CHIP is a highly successful program, and the uninsured rate among children is thankfully much lower than among adults under age 65. The CHIP drop-off figures into CBO's forecast for increased marketplace enrollment. In raw numbers, the CHIP loss is also forecast to be balanced by an increase in Medicaid enrollment as more states accept the ACA's Medicaid expansion. The total number of uninsured is forecast to remain more or less steady, dropping from 27 million this year to 26 million in 2018, then rising gradually to 28 million by 2024. That's half a loaf, with the uninsured rate remaining in the 10% range for people under age 65, cut by a little less than half since ACA implementation.
---
* Originally I phrased this as if CBO were projected a drop in ESI enrollment. In fact, with the exception of a tiny drop in 2016 compared to 2013, the forecast was that growth in ESI would be inhibited. I picked that up from the commenter Larwitt1512 on Charles Gaba's post, so thanks to this person.
** CBO's 2013 forecast had 25 million buying in the exchanges and 22 million in "nongroup and other," compared to 25 million in 2013. If Medicare and the miscellaneous parts of that category were forecast to be where they're at now, that would leave 8 million forecast in the nongroup market. If they wee forecast to grow by a million, that would leave 7 million in off-exchange nongroup, along with 25 million in the exchanges
*** Kaiser will probably soon have its own estimates of off-exchange nongroup enrollment.
In new projections released today, CBO breaks out the major component parts of this grab-bag -- for the first time, as far as very close watchers whom I checked with can remember. The breakout is illuminating in a couple of respects. Here are the numbers for 2016:
These tallies shed some light -- more or less by negation -- on CBO's past overestimates of the number of exchange enrollees.
In 2010, CBO projected that the exchanges would have an average monthly enrollment of 21 million this year, 17 million of them subsidized. In 2013, the forecast had risen to 22 million for this year, with, again, 17 million subsidized. In fact, as of the end of open enrollment on Jan. 31 of this year, 12.7 million were enrolled, 10.5 million of them subsidized. CBO now forecasts average monthly enrollment this year of 12 million, with 10 million subsidized (a conservative estimate of attrition, compared to past experience).
As has been pointed out before (e.g., by the Kaiser Family Foundation), one major factor in the overestimate was CBO's estimate of how many fewer people would have access to employer sponsored insurance as a result of ACA enactment.* In 2013, CBO estimated that ESI enrollment would be 6 million less in 2016 than it would have been under then-current law and a million below the 2013 tally. Last year, CBO forecast that in 2016 ESI would be 7 million below the without-the-ACA benchmark, a sudden drop from just 2 million below in 2015. The 2016 drop has not happened (though CBO is projecting a decrease of 3 million over the next ten years, which is nine million below what CBO estimates would have been the total if there were no ACA). In fact ESI has remained nearly constant, or perhaps upticked, since ACA enactment. Kaiser's 2015 employer survey found that 57% of firms offered coverage in 2015 compared to 55% in 2014, a difference Kaiser dubbed statistically insignificant. As for CBO, last year it estimated that ESI enrollment in 2015 was 156 million; its new estimate for this year is 155 million. The pool of potential marketplace enrollees may be smaller than forecast if ESI growth has not been inhibited as much as CBO envisioned.
Another possible factor was CBO's estimate of shrinkage in the off-exchange individual market. The breakout sheds some new light on that. And here, CBO was not far off. The 2013 projection had the "nongroup and other" category shrinking by 4 million. It was estimated at 25 million in 2013, with a counterfactual "prior law coverage" (i.e., with no ACA) estimate of 26 million for 2016. In the current report, add together Medicare (for disabled adults under 65), "Other," and "outside marketplaces," and you have 23 million. That suggests shrinkage of 3 million rather than the forecast 4 million -- a pretty negligible difference. If the Medicare and "other" categories were predictable and on-target, the off-exchange nongroup total of 9 million is about a million larger than anticipated -- perhaps a negligible difference with these rounded estimates.
CBO also only slightly overestimated the percentage of on-exchange buyers who would be unsubsidized -- 23% in their 2013 forecast for this year, compared to 17% in actuality. Because they thought the whole market would be bigger, they anticipated 4 million unsubsidized exchange enrollees this year, while in actuality there are 2.2 million at present (2 million in their rounded-to-the-million current estimate for this year).
How much of the shortfall is due to slower-than-expected takeup rather than smaller-than-estimated population size? Well, CBO estimates total nongroup enrollment (on and off-exchange) at 22 million this year and 27 million in 2019-20, falling back to 26 million in 2021-24. So that suggests about 83% capacity at present (splitting the difference between 26 and 27 million). In 2013, CBO's estimate for the nongroup market in 2019 was perhaps 32 million, depending on where a forecast shrinkage of 3 million the "nongroup and other" category compared to 2013 was concentrated.**
CBO's reduced projection may be a combination of a down-sized estimate of the potentially eligible population and an assumption that more eligible people will reject the offering than previously assumed. CBO is now assuming that among the uninsured, "35 percent would not purchase insurance to which they had access through an employer, through the marketplaces, or directly from insurers." I don't know whether they've published comparable estimates in the past (and will see if I can update here).
Combining CBO's estimate of the size of the off-exchange nongroup market with Kaiser's estimate of potential marketplace enrollment (or nongroup enrollment) indicates slightly less (though comparable) progress in reaching the eligible population. Kaiser currently estimates 27.4 million potentially eligible marketplace enrollees nationwide -- including those who are ineligible for subsidies. Add CBO's estimated 9 million off-exchange enrollees to the 12.7 million enrolled on-exchange,*** and you have 21.7 million enrollees in the nongroup market. That would be 79% of Kaiser's potential market total, if you substitute "potential nongroup market" for "potential marketplace enrollees."
For the most part, you might as well. The off-exchange enrollees are in the same risk pools as the on-exchange customers, with the exception of 1-2 million (Charles Gaba's estimate) in grandfathered and grandmothered pre-ACA plans and a small number of ACA-compliant plans sold off-exchange by insurers who do not participate in the state exchange in question. So the total ACA-compliant nongroup market may be at about 75% capacity by Kaiser's measure. As I've noted previously, by Kaiser's estimates the exchanges have enrolled about 70% of the subsidy-eligible population. Hence off-exchange takeup would appear to be somewhat higher.
CBO is also forecasting that current CHIP funding levels will not be renewed when they expire in 2017, and that CHIP enrollment will accordingly be cut in half, from 6 million to 3 million. That would be terrible policy -- CHIP is a highly successful program, and the uninsured rate among children is thankfully much lower than among adults under age 65. The CHIP drop-off figures into CBO's forecast for increased marketplace enrollment. In raw numbers, the CHIP loss is also forecast to be balanced by an increase in Medicaid enrollment as more states accept the ACA's Medicaid expansion. The total number of uninsured is forecast to remain more or less steady, dropping from 27 million this year to 26 million in 2018, then rising gradually to 28 million by 2024. That's half a loaf, with the uninsured rate remaining in the 10% range for people under age 65, cut by a little less than half since ACA implementation.
---
* Originally I phrased this as if CBO were projected a drop in ESI enrollment. In fact, with the exception of a tiny drop in 2016 compared to 2013, the forecast was that growth in ESI would be inhibited. I picked that up from the commenter Larwitt1512 on Charles Gaba's post, so thanks to this person.
** CBO's 2013 forecast had 25 million buying in the exchanges and 22 million in "nongroup and other," compared to 25 million in 2013. If Medicare and the miscellaneous parts of that category were forecast to be where they're at now, that would leave 8 million forecast in the nongroup market. If they wee forecast to grow by a million, that would leave 7 million in off-exchange nongroup, along with 25 million in the exchanges
*** Kaiser will probably soon have its own estimates of off-exchange nongroup enrollment.
No comments:
Post a Comment