Longer-Term Effects of the Better Care Reconciliation Act of 2017 on Medicaid Spending

JUNE 2017

Longer-Term Effects of the Better Care

Reconciliation Act of 2017 on Medicaid Spending

In the Congressional Budget Ofﬁce’s assessment,

Medicaid spending under the Better Care Reconcilia-

tion Act of 2017 would be 26 percent lower in 2026

than it would be under the agency’s extended baseline,

and the gap would widen to about 35 percent in 2036

(see Figure 1). Under CBO’s extended baseline, overall

Medicaid spending would grow 5.1 percent per year

during the next two decades, in part because prices for

medical services would increase. Under this legislation,

such spending would increase at a rate of 1.9 percent per

year through 2026 and about 3.5 percent per year in the

decade after that.

CBO and the staff of the Joint Committee on Taxation

do not have an insurance coverage baseline beyond the

coming decade and therefore are not able to quantify the

legislation’s effect on insurance coverage over the longer

term. However, the agencies expect that after 2026,

enrollment in Medicaid would continue to fall relative to

what would happen under the extended baseline.

On the basis of consultation with the budget com-

mittees, CBO’s just-released cost estimate for the bill

measured the costs and savings relative to CBO’s March

2016 baseline projections, with adjustments for legisla-

tion that was enacted after that baseline was produced.

For consistency, this longer-term analysis uses CBO’s

extended baseline published in July 2016.

CBO ana-

lyzed these longer-term effects at the request of the

Ranking Members of the Senate Budget Committee and

the Senate Finance Committee.

1. Congressional Budget Ofﬁce, cost estimate for H.R. 1628, the

Better Care Reconciliation Act of 2017, an amendment in the

nature of a substitute [LYN17343] as posted on the website of

the Senate Committee on the Budget on June 26, 2017 (June 26,

2017), www.cbo.gov/publication/52849.

2. Congressional Budget Ofﬁce, The 2016 Long-Term Budget

Outlook (July 2016), www.cbo.gov/publication/51580.

CBO’s Extended Baseline

The ﬁrst 10 years of projections in CBO’s extended

baseline match the agency’s 10-year baseline projections,

which are based on a detailed analysis of the Medicaid

program. Beyond the coming decade, however, pro-

jecting federal spending on Medicaid becomes increas-

ingly difﬁcult because of the considerable uncertainties

involved. A wide range of changes could occur—in

people’s health, in states’ decisions about Medicaid

eligibility and covered beneﬁts, and in the delivery of

medical care—that are almost impossible to predict but

that could nevertheless have a signiﬁcant effect on federal

spending on Medicaid. Therefore, for the projections

beyond 2026, CBO has adopted a formulaic approach—

one that combines estimates of the number of enrollees

with fairly mechanical projections of growth in federal

spending on Medicaid per enrollee (adjusted to account

for demographic changes in Medicaid enrollees). That

straightforward approach, which was designed to help

make long-term projections of federal deﬁcits and debt,

can be usefully applied only when analyzing proposed

changes in law that, like this bill, would affect spending

in a similarly straightforward manner.

The agency’s estimates of per-enrollee growth in spend-

ing combine projected growth in potential gross domes-

tic product (GDP) per person and projected excess cost

growth for Medicaid, which together average 4.3 percent

in CBO’s extended baseline during the 2027–2036

period. Potential GDP expresses an estimate of the max-

imum sustainable level of growth in the economy. Excess

cost growth is the growth rate of health care spending

per person (after the effects of demographic changes are

removed) relative to the growth rate of potential GDP

per person. (CBO uses potential GDP rather than actual

GDP in its estimate of excess cost growth to limit the

effect of cyclical changes in the economy on its estimate.)

The concept of excess cost growth and its phrasing

are not intended to imply that growth in health care

2 Longer-Term effecTs of The BeTTer care reconciLiaTion acT of 2017 on medicaid spending June 2017

spending per person is necessarily excessive or undesir-

able; the term is used simply to describe the extent to

which the growth in such spending exceeds the growth

in potential output per person.

For Medicaid, the rate of excess cost growth is projected

to be 0.7 percent in 2027 and to rise over the subsequent

decade. In 2036, the rate is projected to be 0.9 percent,

close to its 1985–2014 average of 1.0 percent. That tra-

jectory of excess cost growth reﬂects competing pressures

that are expected to affect the program. On the one

hand, states are likely to face pressure—stemming from

physicians’ practice patterns, new technology, and other

factors in the broader health care system—to increase

payments to health care providers so that they continue

to treat Medicaid enrollees. On the other hand, as health

care costs rise, states are also expected to face pressure to

slow the growth of spending for the program through

actions—such as delivering services more efﬁciently,

constraining payment rates for providers and managed

care plans, limiting the optional services that Medicaid

covers, or restricting the eligibility of certain groups—

that would reduce both state and federal expenditures.

Effects of the Legislation

The largest effects on spending under the Better Care

Reconciliation Act of 2017 would be for Medicaid. Most

of those effects would stem from three major provisions:

O Upon enactment, the legislation would eliminate

penalties associated with the requirements that most

people obtain health insurance coverage and that

large employers offer their employees coverage that

meets speciﬁed standards.

O Starting in 2020, the growth in per-enrollee payments

for nondisabled children and nondisabled adults

enrolled in Medicaid would be capped at no more

than the medical care component of the consumer

price index (CPI-M) and for most enrollees who are

disabled adults or age 65 or older at no more than the

CPI-M plus 1 percentage point. Starting in 2025, the

rate of growth in per-enrollee payments for all groups

would be pegged to the consumer price index for all

urban consumers (CPI-U).

O Starting in 2021, the bill would reduce the federal

matching rate for funding for adults made eligible

for Medicaid by the Affordable Care Act (ACA);

that rate would decline 5 percentage points per year

through 2023 and then fall to equal the rate for other

enrollees in a state in later years.

Overall, including all provisions affecting Medicaid,

CBO estimates that spending for the program would be

reduced by $160 billion in 2026 compared with projec-

tions under current law.

Although it is generally not possible for CBO to provide

detailed estimates of the effects of changes in the nation’s

health care and health insurance systems beyond the

10-year projection period used for cost estimates, the

agency has developed a rough outlook for the decade fol-

lowing the 2017–2026 period by grouping the elements

of the legislation into two broad categories and assessing

the rate at which the budgetary impact of each of those

broad categories is likely to increase over time:

O CBO separated out the portion of Medicaid spending

in 2026 that would be affected by changes proposed

by the bill. For that portion, CBO approximated

spending growth, accounting for the changes to the

expansion of Medicaid eligibility authorized by the

ACA and for the per capita caps.

Figure 1 .

Changes in Medicaid Spending Under the Better

Care Reconciliation Act Compared With CBO’s

Extended Baseline

Percent

2026 2036

-40

-30

-20

-10

-26% -35%

Source: Congressional Budget Oce.

These estimates are for the Better Care Reconciliation Act of 2017, a

Senate amendment in the nature of a substitute to H.R. 1628.

Estimates are based on CBO’s July 2016 extended baseline.

3June 2017 Longer-Term effecTs of The BeTTer care reconciLiaTion acT of 2017 on medicaid spending

O CBO approximated the remainder of Medicaid

spending using the growth rate of such spending in

the agency’s extended baseline.

As always, CBO has endeavored to develop budgetary

estimates that are in the middle of the distribution of

potential outcomes. Such estimates are inherently inexact

because the ways in which federal agencies, states, insur-

ers, employers, individuals, doctors, hospitals, and other

affected parties would respond to the changes made by

this legislation are all difﬁcult to predict.

Per Capita Caps for Medicaid

The per capita caps under this legislation would con-

strain Medicaid spending in stages. Beginning in ﬁscal

year 2020, the federal government would limit the

amount of reimbursement it provides to states. That

limit would be set for a state by calculating the average

per-enrollee cost of medical services for most enrollees

who received full Medicaid beneﬁts over eight consec-

utive quarters of the state’s choosing between the ﬁrst

quarter of federal ﬁscal year 2014 and the third quarter

of 2017. Those enrollees would be in ﬁve speciﬁed cate-

gories: the elderly, disabled adults, nondisabled children,

adults made eligible for Medicaid by the ACA, and all

other adults. The Secretary of Health and Human Ser-

vices would then inﬂate the average per-enrollee costs for

each state as described—for most nondisabled children

and nondisabled adults enrolled in Medicaid using the

CPI-M and for most enrollees who are disabled adults

or age 65 or older using the CPI-M plus 1 percentage

point. Disabled children would be excluded from the per

capita caps and covered as under current law. Beginning

in 2025, the Secretary would shift the inﬂation factor

for all groups to the CPI-U. The ﬁnal limit on federal

reimbursement for each state starting in 2020 would be

the average cost per enrollee for the ﬁve speciﬁed groups

of enrollees, reﬂecting growth from the base period in

the relevant inﬂation factors multiplied by the number

of enrollees in each category. The amount of spending

subject to those limits would be a large share of total

spending.

If a state spent more than the amount eligible for

federal reimbursement, the federal government would

provide no reimbursement for spending over the limit.

By CBO’s projections for the 2017–2024 period, the

limit on federal reimbursement would reduce outlays

because Medicaid spending, on a per-enrollee basis,

for non disabled children and nondisabled adults under

current law (after the changes to the Medicaid expansion

population have been accounted for) would grow faster,

at 4.9 percent, than the CPI-M, at 3.7 percent. However,

for most enrollees who are disabled adults or age 65 or

older, that rate under current law would be 3.3 percent,

lower than the CPI-M plus 1 percentage point. The per

capita caps would have a small effect on spending for

those groups, even though the caps would not gener-

ally be binding for them, because some shifting of costs

among groups would probably occur, and spending for a

particular group in a particular year could be affected.

In 2025 and beyond, the differences between spending

growth for Medicaid under current law and the growth

rate of the per capita caps for all groups would be sub-

stantial. CBO projects the growth rate of the CPI-U in

those years to be 2.4 percent.

Effects on Spending

Over the next decade, CBO projects, a large gap would

grow between Medicaid spending under current law and

under this bill. In later years, that gap would continue to

widen because of the compounding effect of the differ-

ences in spending growth rates. CBO projects that the

growth rate of Medicaid under current law would exceed

the growth rate of the per capita caps for all groups cov-

ered by the caps starting in 2025.

In CBO’s extended baseline, Medicaid spending is pro-

jected to be 2.0 percent of GDP in 2017 and 2.4 percent

by 2036. The 35 percent reduction in that spending that

CBO estimates for 2036 under this legislation would

result in Medicaid spending of 1.6 percent of GDP.

3. CBO generally presents long-term estimates as percentages of

GDP and not in nominal dollars. In the agency’s judgment,

a presentation in nominal dollars can be misleading. The key

problem is that a dollar today means something very different

from a dollar in the distant future, for at least two reasons. First,

the cumulative effect of changes in prices over a long period can

be quite large, so a dollar amount in the distant future will have

much lower value than the same dollar amount today. Second,

the population, the economy, and people’s incomes will all grow

substantially over time, so a dollar amount in the distant future

will be much smaller relative to the size of the economy or a

person’s income than the same dollar amount today.

4 Longer-Term effecTs of The BeTTer care reconciLiaTion acT of 2017 on medicaid spending June 2017

This document was requested by the Ranking Member

of the Senate Committee on the Budget and the Ranking

Member of the Senate Committee on Finance. Jeffrey

Kling prepared it with guidance from Holly Harvey and

with contributions from Xiaotong Niu, Lisa Ramirez-

Branum, Michael Simpson, and Robert Stewart. Jessica

Banthin, Chad Chirico, Theresa Gullo, Mark Hadley,

and David Weaver reviewed the document, John Skeen

edited it, and Casey Labrack prepared it for publication.

An electronic version is available on CBO’s website

(www.cbo.gov/publication/52859).

Keith Hall

Director

June 2017

Under this legislation, after the next decade, states

would continue to need to arrive at more efﬁcient

methods for delivering services (to the extent feasi-

ble) and to decide whether to commit more of their

own resources, cut payments to health care providers

and health plans, eliminate optional services, restrict

eligibility for enrollment, or adopt some combination

of those approaches. Over the long term, there would

be increasing pressure on more states to use all of those

tools to a greater extent.