Medicaid Capitation Payments by State.

Congress has repeatedly proposed changing Medicaid from an entitlement to a block grant. Each state would receive a fixed amount instead of a Federal payment influenced by state decisions on eligibility, coverage, and pricing. This paper uses existing data series to simulate redistributing the annual $353 billion Federal payment among Medicaid's 56 state (and territorial) programs. Capitation by general population would shift $52 billion, mainly from large Northeastern and West Coast states to large Southern and Mountain states. Capitation by population below the Federal Poverty Line (FPL) would shift $60 billion in a similar pattern. Policymakers should understand likely state-to-state effects when considering Medicaid legislation. States could then prepare for possible changes in their Federal payment for Medicaid.

What do we already know about this topic?

Since the 1960s, multiple Administrations and Congresses have proposed converting Medicaid from an entitlement to a block grant. Capitation would let the Federal and state governments limit their Medicaid expenditure. It also would permit states to choose whom to enroll and what services to deliver. Previous research has described the national effects of capitation.

How does your research contribute to the field?

This article investigates the state-by-state outcomes from capitating Medicaid either by general population or FPL population.

What are your research’s implications toward theory, practice, or policy?

Either form of capitation would shift $52 to $60 billion among states. Certain states would need to prepare for changes in their Federal payments.

Background

Medicaid pays for the health care of low-income US citizens and legal permanent residents.[1] It enrolls about 23% of the US population and funds about 17% of national health expenditure.[2] Medicaid operates as a Federal–state partnership. The Federal government sets the rules for enrollment (persons covered) and benefits (services covered). It pays part of total expenditure based on each state’s Federal Medical Assistance Percentage (FMAP). States with lower average per capita incomes have higher FMAP rates. States deliver services via shifting combinations of fee-for-service, managed care, limited-benefit carve-outs (eg, behavioral health), competitive bidding, and so on.

Like other “entitlement” programs (Medicare, Social Security, veterans’ benefits, unemployment compensation, Food Stamps, agricultural price supports, etc.), Medicaid has no limits on its expenditure.[3] It must provide benefits to any person meeting its eligibility criteria. The open-ended benefits of entitlement programs make it difficult to predict or control their costs. In contrast, a “discretionary” program (eg, defense, highways, education) cannot spend more than its annual Congressional appropriation.

Despite Medicaid being an entitlement, Federal rules currently give a state indirect influence over its Medicaid expenditure.

Enrollment: Statute requires coverage of “mandatory eligibility” groups (low-income children and pregnant women, aged and disabled receiving Supplemental Security Income [SSI], selected diseases, etc.).[4] A state may elect to cover “optional eligibility” groups (higher income children, pregnant women, aged, and disabled; medically needy; etc.). States vary in their enrollment of optional groups.

Benefits: Statute requires coverage of “mandatory benefits” (inpatient, outpatient, laboratory, etc.).[5] A state may also choose to cover “optional benefits” (dental, vision, prescription drugs, hospice, etc.). States also differ in their coverage of optional benefits.

Medicaid expansion: The Affordable Care Act (ACA) expanded Medicaid enrollment to low-income, nondisabled adults.[6] The Supreme Court subsequently ruled this “Medicaid expansion” to be noncompulsory.[7]

Price: By setting the price per service, a state can indirectly affect the volume of services.[8]

Congress intermittently proposes changing Medicaid from an entitlement to a discretionary program.[9] A block grant based on general population or FPL population would permit the Federal government limit its share of Medicaid to a fixed appropriation.[10,11] States could also set limits on their own expenditure. States would gain flexibility setting their own priorities for enrollment and benefits (eg, a limited package of services for a large population vs extensive services for a narrower population). Finally, reduced Federal control might allow states to operate their Medicaid services more efficiently (eg, by substituting integrated care for fragmented services).

On the other hand, converting Medicaid to a block grant could reduce coverage by eliminating mandated enrollment and benefits.[12] If a state can decrease its own Medicaid expenditure without affecting its Federal matching payment, then reductions of enrollment or benefits accrue entirely to the state. Also, states would likely widen their differences in the populations served and benefits provided.

This analysis investigates the effect of capitation on individual states. Which states gain or lose Federal payments under capitation by general population versus FPL population? Which states gain or lose as a dollar amount versus percentage change? Does capitation penalize states that previously expanded benefits via optional eligibility, optional benefits, ACA expansion, or pricing?

Methods

This analysis compares the effect of capitating the Federal share of Medicaid by the general population or FPL population in the 56 administrative divisions of the United States: 50 states, one Federal district, and five island territories. For brevity the tables refer to the 56 jurisdictions as states. The analysis assumes budget neutrality and no statutory FMAP rates (territories receive fixed rates not based on average per capita income: 55% for the island territories and 70% for the District of Columbia).

The data come from existing statistical series. The Census Bureau (CB) enumerates general population[13] and FPL population by jurisdiction.[14] The Centers for Medicare & Medicaid Services (CMS) report enrollment,[15] total Medicaid expenditure, Federal payments,[16] and payment per service.[17] The Kaiser Family Foundation (KFF) compiles jurisdictions’ adoption of optional enrollment groups, benefits, and uptake of the ACA expansion.[18] The Medicaid and CHIP Payment and Access Commission (MacPAC),[19] US Government Accountability Office,[20] and KFF[21,22] publish additional data on the island territories. These institutions make their data available as public use files. The analysis uses 2017 data wherever available (information about the island territories sometimes runs late).

Excel 365 software[23] converted the general and FPL population frequencies into proportional distributions by jurisdiction. Holding the total Federal payment constant at $352 billion and multiplying by the proportional distributions produced capitated totals. Subtracting them from the actual Federal payments quantified the dollar changes in Federal payment to each jurisdiction. Dividing these changes by their current Federal payments gave the percentage changes, highlighting capitation’s effect on smaller jurisdictions (Table 1).

Table 1.

Medicaid Population, Expenditure, and Change Under Capitation by Jurisdiction, 2017.

State	Population			Medicaid expenditure		Capitation by general population				Capitation by <100% FPL population
	General	Medicaid	FPL <100%	Total	Federal payment	Federal payment		Change		Federal payment		Change
	Million persons	% of general population	% of general population	$ billion	$ billion	$ billion	Percent distribution	$ billion	% change	$ billion	Percent distribution	$ billion	% change
US	328.9	23	13	572.24	352.47	352.47	100.0	0.00	0	352.47	100.0	0.00	0
AK	0.7	29	14	1.96	1.37	0.79	0.2	−0.58	−42	0.83	0.2	−0.54	−39
AL	4.9	18	16	5.56	3.92	5.23	1.5	1.31	33	6.35	1.8	2.43	62
AR	3.0	28	15	6.36	4.94	3.22	0.9	−1.72	−35	3.86	1.1	−1.08	−22
AS	0.1	73	19	0.03	0.02	0.06	0.0	0.04	223	0.09	0.0	0.07	373
AZ	7.0	24	15	11.82	8.97	7.56	2.1	−1.42	−16	8.66	2.5	−0.32	−4
CA	39.4	30	13	82.38	48.60	42.23	12.0	−6.37	−13	43.44	12.3	−5.16	−11
CO	5.6	23	8	7.81	4.51	6.02	1.7	1.51	34	3.80	1.1	−0.71	−16
CT	3.6	24	10	7.40	4.29	3.83	1.1	−0.46	−11	3.10	0.9	−1.19	−28
DC	0.7	37	15	2.78	2.06	0.75	0.2	−1.31	−64	0.87	0.2	−1.19	−58
DE	1.0	26	10	2.13	1.31	1.03	0.3	−0.29	−22	0.83	0.2	−0.48	−37
FL	21.0	20	13	23.17	14.19	22.48	6.4	8.30	58	23.30	6.6	9.12	64
GA	10.4	17	14	10.11	6.88	11.16	3.2	4.28	62	12.44	3.5	5.56	81
GU	0.2	23	23	0.08	0.05	0.18	0.0	0.12	244	0.31	0.1	0.26	514
HI	1.4	23	10	2.34	1.53	1.53	0.4	0.00	0	1.17	0.3	−0.36	−24
IA	3.1	22	9	4.07	2.55	3.37	1.0	0.82	32	2.47	0.7	−0.08	−3
ID	1.7	16	11	1.82	1.30	1.84	0.5	0.54	41	1.64	0.5	0.33	26
IL	12.8	22	12	15.05	9.34	13.70	3.9	4.36	47	12.28	3.5	2.94	31
IN	6.7	22	12	11.11	8.02	7.14	2.0	−0.88	−11	6.45	1.8	−1.56	−20
KS	2.9	13	13	3.21	1.81	3.12	0.9	1.31	72	3.14	0.9	1.33	73
KY	4.5	27	15	9.53	7.40	4.77	1.4	−2.63	−35	5.51	1.6	−1.89	−26
LA	4.7	31	21	10.91	7.71	5.01	1.4	−2.70	−35	8.12	2.3	0.40	5
MA	6.9	23	10	17.12	9.35	7.36	2.1	−2.00	−21	5.79	1.6	−3.56	−38
MD	6.0	22	8	11.16	6.78	6.46	1.8	−0.32	−5	3.77	1.1	−3.00	−44
ME	1.3	19	12	2.57	1.65	1.43	0.4	−0.22	−13	1.37	0.4	−0.28	−17
MI	10.0	23	12	16.71	12.13	10.69	3.0	−1.43	−12	9.92	2.8	−2.21	−18
MN	5.6	19	9	11.35	6.49	5.97	1.7	−0.52	−8	4.19	1.2	−2.31	−36
MO	6.1	15	12	10.10	6.40	6.55	1.9	0.15	2	6.12	1.7	−0.27	−4
MP	0.1	35	46	0.03	0.02	0.06	0.0	0.04	255	0.21	0.1	0.20	1174
MS	3.0	21	20	5.46	4.08	3.20	0.9	−0.88	−21	4.92	1.4	0.84	21
MT	1.1	26	11	1.77	1.42	1.13	0.3	−0.29	−20	0.94	0.3	−0.48	−34
NC	10.3	20	14	13.34	8.94	11.01	3.1	2.07	23	12.01	3.4	3.07	34
ND	0.8	12	11	1.22	0.75	0.81	0.2	0.06	8	0.71	0.2	−0.04	−5
NE	1.9	13	10	2.04	1.06	2.06	0.6	1.00	94	1.60	0.5	0.54	51
NH	1.3	13	7	2.06	1.22	1.45	0.4	0.22	18	0.73	0.2	−0.49	−40
NJ	8.9	19	9	14.74	8.87	9.53	2.7	0.66	7	6.68	1.9	−2.18	−25
NM	2.1	35	18	4.80	3.71	2.24	0.6	−1.47	−40	3.18	0.9	−0.53	−14
NV	3.0	21	12	3.53	2.65	3.19	0.9	0.54	20	2.95	0.8	0.31	12
NY	19.6	33	13	76.40	37.26	21.00	6.0	−16.27	−44	20.62	5.8	−16.64	−45
OH	11.7	23	13	23.06	15.89	12.50	3.5	−3.39	−21	12.86	3.6	−3.03	−19
OK	3.9	20	14	4.63	2.85	4.22	1.2	1.36	48	4.47	1.3	1.61	57
OR	4.1	23	11	8.31	6.20	4.44	1.3	−1.75	−28	3.81	1.1	−2.39	−39
PA	12.8	23	11	28.08	17.11	13.71	3.9	−3.40	−20	11.97	3.4	−5.15	−30
PR	3.3	50	44	2.32	1.54	3.56	1.0	2.02	131	12.08	3.4	10.54	685
RI	1.1	29	12	2.62	1.55	1.13	0.3	−0.42	−27	1.04	0.3	−0.51	−33
SC	5.0	20	15	5.96	4.26	5.38	1.5	1.13	26	6.21	1.8	1.95	46
SD	0.9	13	12	0.85	0.50	0.94	0.3	0.43	87	0.90	0.3	0.40	81
TN	6.7	21	13	9.09	5.91	7.19	2.0	1.28	22	7.40	2.1	1.48	25
TX	28.3	15	14	35.64	20.11	30.36	8.6	10.25	51	32.18	9.1	12.06	60
UT	3.1	9	9	2.45	1.72	3.33	0.9	1.61	94	2.23	0.6	0.51	30
VA	8.5	12	11	8.99	4.50	9.07	2.6	4.57	101	7.64	2.2	3.13	70
VI	0.1	17	22	0.05	0.03	0.11	0.0	0.08	238	0.20	0.1	0.17	491
VT	0.6	25	10	1.60	0.94	0.67	0.2	−0.27	−29	0.52	0.1	−0.43	−45
WA	7.4	23	10	11.89	7.57	7.96	2.3	0.39	5	6.45	1.8	−1.12	−15
WI	5.8	18	10	8.05	4.74	6.21	1.8	1.47	31	4.89	1.4	0.15	3
WV	1.8	29	18	4.00	3.17	1.95	0.6	−1.22	−38	2.69	0.8	−0.48	−15
WY	0.6	10	12	0.59	0.30	0.62	0.2	0.32	104	0.56	0.2	0.26	85

Note. FPL = Federal Poverty Line.

R-Project software[24] cross tabulated Federal payment changes against selected state coverage choices (Table 2).

Table 2.

Effect of Capitation on Selected Medicaid Optional Eligibility Groups, Optional Benefits, ACA Expansion, and Price, 2017.

Number of states	Capitation by general population		Capitation by FPL <100%
	Increase	Decrease	Increase	Decrease
Pregnant women
>200% FPL	15	18	11	22
≤200% FPL	15	8	14	9
Dental coverage
Yes	24	24	19	29
No	6	2	6	2
Vision coverage
Yes	20	19	15	24
No	10	7	10	7
ACA expansion
Yes	8	24	3	29
No	22	2	22	2
Price per service
>$400	15	16	14	17
≤$400	15	10	11	14

Note. Statistically significant relationships in bold. ACA = Affordable Care Act; FPL = Federal Poverty Line.

Enrollment: Medicaid had mandatory enrollment of pregnant women with modified adjusted gross incomes (MAGI) <133% FPL, optional enrollment for 133% to 185% FPL, and optional Children’s Health Insurance Program (CHIP) enrollment for >185% FPL. Over half of jurisdictions covered >200% FPL.

Benefits: States elected to offer a variety of dental and vision services. This analysis counted whether a jurisdiction offered any services but did not quantify their scope and value.

ACA expansion: This variable counted adoption of coverage for nondisabled adults but did not consider administrative barriers or special conditions (eg, work requirements).

Price: The jurisdictions had a mean cost per claim of just under $400 and a median cost just over $400.

These cross tabulations tested whether the coverage decision involved a sufficiently large sum to be affected by capitation. They implied nothing about causation.

Results

Population

Half the general population resides in only 9 states (in order: California, Texas, Florida, New York, Pennsylvania, Illinois, Ohio, Georgia, and North Carolina). Half the Medicaid population enroll in 8 states (California, New York, Texas, Florida, Pennsylvania, Illinois, Ohio, and Michigan). Medicaid enrollment ranges from 9% of the general population in Utah and Wyoming to 73% in American Samoa (which does not determine Medicaid eligibility on an individual basis).[25] The 56 jurisdictions average 21% ± 10% (SD) of their general populations enrolled in Medicaid (Table 1).

Similarly, half of the population below the FPL lives in 9 jurisdictions (California, Texas, Florida, New York, Ohio, Georgia, Illinois, Puerto Rico, and North Carolina). The FPL population ranges from 7% of the general population in New Hampshire to 46% in Northern Marianas with a mean of 15% ± 7%.

Expenditure

Half the $572 billion in total Medicaid expenditure occurs in seven states (California, New York, Texas, Pennsylvania, Florida, Ohio, and Massachusetts). Total expenditure ranges from $30 million annually in the Northern Marianas to $82 billion in California. Dividing total expenditure by Medicaid population, cost-per-enrollee goes from $820 annually in American Samoa to $13,354 in North Dakota with an average of $7,437 ± $2,556.

Federal payments cover 62% of total Medicaid expenditure. Half of this $353 billion flows to 9 states (California, New York, Texas, Pennsylvania, Ohio, Florida, Michigan, Massachusetts, and Illinois). The Federal share ranges from 49% of total expenditure in New York to 75% in Mississippi with a mean of 58% ± 8%. The Federal payment varies from $17 million for the Northern Marianas to $49 billion for California.

Capitation by General Population

Multiplying the $353-billion Federal payments by the proportional distribution of the general population and subtracting from their current FMAP payments, 30 jurisdictions would receive larger Federal payments versus 26 with smaller payments. Half of this $52 billion in interstate transfers would accrue to 4 states (Texas, Florida, Virginia, and Illinois). Similarly, half the transfer would come from 4 states (New York, California, Ohio, and Pennsylvania). The change in Federal payment would range from −$16 billion for New York to +$10 billion for Texas. Geographically, Northeastern and West Coast states generally would transfer funds toward Southern and Mountain States.

Assuming no statutory rates, the island territories, Wyoming, and Virginia would each receive an increase of at least 100% in their Federal payments. Conversely, the District of Columbia, New York, Alaska, and New Mexico would suffer the greatest percentage decreases. The percentage change would range from −64% for the District of Columbia to 255% for the Northern Marianas.

Capitation by FPL

Multiplying $353 billion by the proportional distribution of the population <100% FPL and subtracting from their current Federal payments, 25 jurisdictions would receive larger Federal payments versus 31 with smaller payments. FPL capitation would transfer $60 billion with half the increase going to 3 states (Texas, Puerto Rico, and Florida). Half the transfer would come from 4 states (New York, California, Pennsylvania, and Massachusetts). The change in Federal payment would range from −$17 billion for New York to +$12 billion for Texas. The geographic pattern largely tracks capitation by general population.

As a percentage change, the 5 island territories, Wyoming, South Dakota, and Georgia would receive the largest increases in Federal payments. Conversely, District of Columbia would again lose more than half its Federal payment, followed by New York, Vermont, Maryland, and New Hampshire. The change would range from −58% for the District of Columbia to 1174% for the Northern Marianas.

Selected Coverage Options Under Capitation by General Population

States make different choices about optional coverage. For example, 33 jurisdictions enroll pregnant women with MAGI >200% FPL. However, the states’ MAGI limit does not associate with higher versus lower Federal payments under capitation by general population. Similarly, neither the 48 jurisdictions with dental coverage nor the 39 jurisdictions offering vision benefits would suffer systematic decreases in their Federal payments. Conversely, the 32 jurisdictions adopting the ACA expansion would receive lower Federal payments (χ² = 24.5, p < 0.000001). Finally, price per service above or below $400 would not affect Federal payment (Table 2).

Selected Coverage Options Under Capitation by FPL

The FPL population distributes somewhat differently from the general population. Jurisdictions that enroll higher income pregnant women would tend to receive lower Federal payments under FPL capitation (χ² = 4.2, p < 0.05). Provision of dental or vision services would still not relate to Federal payments. FPL capitation would significantly penalize jurisdictions embracing the ACA expansion (χ² = 38.6, p < 0.000001). Price per service ≥$400 does not influence Federal payments.

Discussion

Capitation by general population would transfer 15% of the $343-million Federal payment between jurisdictions. FPL capitation would increase the shift to 17%. These proportions seem relatively modest considering the jurisdictions’ different decisions on coverage and pricing. Indeed, Table 2 suggests that only the ACA expansion and income limits involve enough expenditure to be affected by capitation.

Populous jurisdictions account for most of the shift in Federal payments under either form of capitation. Texas, Puerto Rico, Florida, Virginia, and Illinois receive most of the increase, while New York, California, Ohio, Pennsylvania, and Massachusetts account for most of the decrease. Coverage decisions smaller than the ACA expansion largely cannot offset these states’ large populations. Tax effects outside the scope of this article should exacerbate this trend. Puerto Rico, Virginia, Florida, Texas, Pennsylvania, and Ohio receive more Federal benefits than they pay in taxes, while New York, California, Illinois, and Massachusetts pay more than they receive.[26]

Conversely, small jurisdictions can experience substantial percentage changes despite minimal dollar shifts in Federal payments. The island territories would enjoy the largest increases, and District of Columbia would suffer the largest decrease under either form of capitation. These findings highlight the uneven effect of Congress setting statutory rates for the non-state jurisdictions. The District of Columbia receives a larger Federal payment than warranted under either FMAP (per capita income) or capitation, whereas the island territories get less. Their respective economies may have changed since 1965, when Congress fixed these rates. Consistent policymaking suggests using the same formula for states and territories, whether FMAP or capitation.

Forces external to Medicaid largely drive the recurrent debates over capitation versus entitlement. This analysis generates data to inform the discussions. If Medicaid policies change, governments can prepare to buffer the transitions.