U.S. flag

An official website of the United States government

Medicaid Per Capita Expenditures


VISIT THE PILLAR: National Context
DOMAIN: What are Annual Expenditures for Medicaid & CHIP?

State per capita expenditures provide information about each state’s Medicaid program and all the populations they serve. The following estimates provide average annual Medicaid expenditures per enrollee for calendar years 2018 and 2019 by state for five eligibility groups:

  1. Children
  2. Adults: non-expansion, non-disabled, under age 65
  3. Aged
  4. People with disabilities
  5. Adults: ACA Medicaid expansion

The estimates rely on total spending reported by states to the Medicaid Budget and Expenditure System (MBES) and the number of enrollees and their expenditures reported by states in the Transformed Medicaid Statistical Information System (T-MSIS). The calculations exclude all Children’s Health Insurance Program (CHIP) spending and enrollment through both Medicaid CHIP and separate CHIP programs.

The validity of the results presented here is directly related to the accuracy of state-reported data. For this reason, data quality assessments across four domains and an overall assessment of data quality are included to evaluate the usability of each state’s data to produce the per capita expenditure estimates. The data quality assessments are similar to the approaches used in the 2020 Scorecard, but are now drawn from data quality analyses in the DQ Atlas. The criteria for assigning states a level of overall data quality concern are:

Low overall level of data quality concern: the state’s data was assessed as low data quality concern across all four domains.

  • The number of Medicaid enrollees in state-reported T-MSIS files was within 10 percent of enrollment numbers in Performance Indicator (PI) data, as shown in the second column of Table 2.
  • 5 percent or less of all Medicaid and CHIP enrollees were missing eligibility group information, as shown in the third column of Table 2.
  • The volume of state reported claims data was greater than or equal to 10 percent of the national median for inpatient, other services, and prescription drug claims, as shown in the fourth column of Table 2.
  • Total claim expenditures (fee-for-service plus capitated managed care and other monthly payments) were within 5 percent of CMS-64 expenditures states reported in MBES, as shown in the fifth column of Table 2.

Moderate overall level of data quality concern: the state’s data presents a moderate level of data quality concern in at least one of the four domains, but does not present a high data quality concern in any domain.

  • The number of Medicaid enrollees in state-reported T-MSIS files differed by between 10 and 20 percent from the enrollment numbers in PI data, as shown in the second column of Table 2.
  • Between 5 and 10 percent of all Medicaid and CHIP enrollees were missing eligibility group information, as shown in the third column of Table 2.
  • Total claim expenditures (fee-for service plus capitated managed care and other monthly payments) were between 5 and 20 percent of CMS-64 expenditures states reported in MBES, as shown in the fifth column of Table 2.

High overall level of data quality concern: the state’s data presents a high data quality concern[1]   in one or more domains.

  • The number of Medicaid enrollees in state-reported T-MSIS files differed by more than 20 percent from the enrollment numbers in PI data, as shown in the second column of Table 2.
  • More than 10 percent of all Medicaid and CHIP enrollees were missing eligibility group information, as shown in the third column of Table 2.
  • The volume of state reported claims data was less than 10 percent of the national median for inpatient, other services, or prescription drug claims, as shown in the fourth column of Table 2.
  • Total claim expenditures (fee-for service claims plus managed care capitation payments and other monthly payments) differed by more than 20 percent from the expenditures the state reported on the Form CMS-64, as shown in the fifth column of Table 2.

Multiple factors drive variation in state Medicaid spending. Some of these factors include, for example, differences in:

  • Enrollee characteristics and health status
  • Eligibility rules that determine who qualifies for Medicaid
  • Scope of benefits covered
  • Mix of delivery systems, such as managed care and fee-for-service
  • Payment levels and arrangements for providers and health plans
  • Local health care costs and market features

In addition, an array of factors beyond T-MSIS data quality and completeness may impact state performance on the data usability assessments, some of which may be outside of the control of Medicaid agencies, for example, differences in:

  • Organization of the data sources being compared, e.g. T-MSIS expenditures for each calendar year are grouped primarily based on service date, while CMS-64 expenditures are reported based on payment date
  • Guidance to states for reporting information to each data source

The following estimates do not adjust for these differences (Table 1). See the data usability section and Table 2 for more detail about the assessment of state T-MSIS data usability.

Table 1. Per capita Expenditure Estimates for States and Data Quality Assessment (2018)

State Data quality assessment Total Children Adult: non-expansion, non-disabled, under age 65 Aged People with disabilities Adult: ACA Medicaid expansion
Alabama Medium concern $5,329 $1,891 $1,595 $12,336 $12,148 -
Alaska Low concern $10,020 $5,779 $7,315 $22,560 $32,852 $9,170
Arizona Low concern $6,556 $3,119 $4,531 $9,247 $21,090 $7,055
Arkansas Medium concern $7,166 $4,542 $8,551 $21,520 $18,358 $1,423
California Medium concern $6,442 $2,678 $2,953 $13,498 $23,685 $5,819
Colorado Medium concern $7,486 $2,889 $4,875 $21,998 $28,393 $5,437
Connecticut Low concern $8,904 $3,692 $5,867 $16,746 $29,425 $7,672
Delaware Low concern $9,315 $4,409 $8,337 $20,441 $24,217 $7,724
District of Columbia Low concern $11,031 $4,447 $6,252 $22,772 $31,786 $6,749
Florida High concern $5,119 $1,087 $2,422 $10,871 $18,187 -
Georgia Low concern $5,339 $2,633 $4,575 $9,804 $11,702 -
Hawaii Low concern $6,673 $3,053 $4,804 $13,644 $23,034 $5,923
Idaho Medium concern $7,165 $2,731 $7,023 $13,004 $19,419 -
Illinois Medium concern $6,523 $2,548 $4,756 $17,275 $15,044 $5,766
Indiana Medium concern $8,400 $3,006 $7,733 $17,872 $17,212 $9,423
Iowa Medium concern $8,355 $3,038 $7,144 $21,842 $23,509 $6,326
Kansas Medium concern $9,001 $3,706 $14,422 $21,042 $15,391 -
Kentucky High concern $6,515 $3,384 $5,836 $10,033 $13,556 $5,642
Louisiana Low concern $6,520 $3,337 $5,824 $10,563 $13,604 $5,958
Maine Medium concern $10,593 $4,433 $4,918 $13,317 $21,772 $4,959
Maryland Low concern $9,129 $3,581 $7,392 $18,973 $24,973 $8,663
Massachusetts Medium concern $10,174 $3,713 $4,098 $24,282 $20,046 $6,712
Michigan Medium concern $6,838 $2,821 $5,468 $18,502 $16,853 $5,246
Minnesota Medium concern $11,591 $3,204 $5,588 $37,695 $47,844 $7,338
Mississippi Low concern $7,553 $3,910 $5,115 $12,793 $13,571 -
Missouri Medium concern $9,435 $4,229 $6,316 $19,369 $24,668 -
Montana Medium concern $7,159 $4,182 $5,749 $17,758 $15,988 $6,630
Nebraska High concern $9,641 $4,481 $9,123 $24,522 $19,349 -
Nevada Low concern $5,878 $2,732 $5,381 $8,172 $16,665 $6,391
New Hampshire Low concern $9,905 $3,987 $14,945 $24,832 $17,811 $9,478
New Jersey Medium concern $9,213 $3,048 $9,621 $23,675 $30,406 $5,953
New Mexico Low concern $6,381 $3,695 $3,829 $10,418 $21,873 $5,790
New York Medium concern $11,694 $4,094 $7,990 $26,799 $33,114 $7,468
North Carolina Medium concern $6,472 $2,877 $2,713 $14,022 $17,953 -
North Dakota Low concern $13,557 $4,707 $5,240 $34,710 $32,967 $14,383
Ohio Low concern $8,168 $3,314 $5,626 $21,194 $18,533 $6,988
Oklahoma Medium concern $6,942 $3,752 $4,744 $14,658 $16,863 -
Oregon High concern $9,437 $5,426 $12,209 $18,390 $17,957 $8,710
Pennsylvania High concern $11,540 $3,243 $5,673 $36,512 $25,821 $3,936
Puerto Rico Medium concern $1,807 $2,101 $2,160 $727 $959 $2,288
Rhode Island High concern $11,606 $5,807 $8,475 $21,822 $18,428 $10,428
South Carolina Medium concern $4,986 $2,561 $2,693 $11,950 $14,294 -
South Dakota Low concern $8,296 $3,164 $6,566 $17,701 $20,902 -
Tennessee High concern $6,419 $3,350 $3,844 $14,111 $14,755 -
Texas Medium concern $8,531 $3,899 $8,099 $16,360 $23,229 -
Utah Medium concern $8,908 $3,851 $7,160 $17,885 $26,237 -
Vermont Medium concern $9,276 $6,084 $7,430 $14,453 $22,998 $6,426
Virgin Islands Medium concern $3,221 $1,211 $3,201 $2,789 $6,216 $3,329
Virginia Medium concern $8,169 $2,936 $4,347 $18,739 $21,573 -
Washington Medium concern $6,911 $2,554 $7,596 $19,087 $20,298 $6,071
West Virginia Medium concern $7,294 $3,134 $5,918 $20,450 $13,773 $5,402
Wisconsin Medium concern $8,136 $2,648 $4,631 $15,209 $22,642 -
Wyoming Low concern $9,894 $4,111 $7,942 $20,702 $24,287 -
Min - $1,807 $1,087 $1,595 $727 $959 $1,423
Median - $8,168 $3,337 $5,749 $17,872 $20,046 $6,391
Max - $13,557 $6,084 $14,945 $37,695 $47,844 $14,383

Source: CMS calculations using calendar years 2018 and 2019 MBES expenditure data and T-MSIS data on enrollees and expenditures.

Note: Some states report significant capitation dollars as service tracking claims that might distort the distribution of per capita expenditure across eligibility groups. For example, Arkansas's Medicaid per capita expenditures for the adult expansion group are lower than other states because Arkansas submits Qualified Health Plan premium payments for this group as lump sum payment records on service tracking claims. This type of claim record cannot be linked to individual beneficiaries and were therefore not used to allocate CMS-64 expenditures across eligibility groups. The distribution of those expenditures across eligibility groups may appear to be distorted due to this data limitation.

Notes

To conduct this analysis, CMS used two data sources: (1) MBES expenditure data and (2) T-MSIS data. MBES expenditure data (reported by states on Form CMS-64) are at the state level and do not include expenditures at the enrollee or at the eligibility group levels; therefore, CMS used T-MSIS data to classify enrollees, allocate expenditures into eligibility groups, and construct the denominator (number of enrollee years) for each eligibility group.

To compute the denominator, CMS first mapped the 75 eligibility categories in T-MSIS to the five eligibility groups and then determined the count of enrollee months by eligibility group in each state. As the last step, total months were converted to total years.

To compute the numerator (total spending for each eligibility group) for each state, the calculation used total state-level expenditure information from MBES. To distribute MBES expenditures across the five eligibility groups, the calculation used T-MSIS data to determine the percentage of total Medicaid expenditures accounted for by each eligibility group, then applied that percentage to the total net Medicaid expenditures for medical assistance services that each state reported to MBES. For each state, these total net expenditures excluded spending categories not linked to individual enrollees, such as administrative expenses and disproportionate share hospital payments. This methodology largely aligns with methods that CMS’s Office of the Actuary employs to estimate national-level Medicaid spending per enrollee.

The approach used to calculate the per capita expenditure estimates is almost entirely the same as that used in the 2020 Scorecard. One notable exception is how we ensured each beneficiary was enrolled in Medicaid (and not CHIP) in the month in which a fee-for-service claim occurred. In addition, we allocated CMS-64 drug rebates expenditures across eligibility groups using the overall distribution of TAF expenditures rather than the specific distribution of TAF prescription drug expenditures.[2] Per capita expenditures for 2018 in this version of the Scorecard will differ from 2018 values displayed in the 2020 Scorecard because the data has been updated and because the per capita expenditures methodology has changed. See the methodology document for further details. 

Data quality assessment

To determine the completeness and accuracy of states’ T-MSIS data used to calculate per capita expenditures, CMS relied on four DQ Atlas assessments of key data elements. For some states with T-MSIS data that present a high level of data quality concern,[3] differences between their T-MSIS data and other CMS data sources may be due to a number of factors, including reporting issues that are not currently captured in the methodology and may be outside the control of states. In addition to the methodology, there are a range of state-specific factors and external reasons why a state may not have met the data quality requirements. Table 2 summarizes the results of the data quality assessment for 2018 and 2019. 

Table 2. Assessing State T-MSIS Data Quality (2018)

State Overall data quality assessment Enrollment benchmarking[a] Missing information on eligibility[b] Low claims volume[c] Expenditure benchmarking[d]
Alabama Medium concern Low concern Low concern Low concern Medium concern
Alaska Low concern Low concern Low concern Low concern Low concern
Arizona Low concern Low concern Low concern Low concern Low concern
Arkansas Medium concern Low concern Low concern Low concern Medium concern
California Medium concern Low concern Low concern Low concern Medium concern
Colorado Medium concern Low concern Low concern Low concern Medium concern
Connecticut Low concern Low concern Low concern Low concern Low concern
Delaware Low concern Low concern Low concern Low concern Low concern
District of Columbia Low concern Low concern Low concern Low concern Low concern
Florida High concern Low concern Medium concern Low concern High concern
Georgia Low concern Low concern Low concern Low concern Low concern
Hawaii Low concern Low concern Low concern Low concern Low concern
Idaho Medium concern Low concern Low concern Low concern Medium concern
Illinois Medium concern Low concern Low concern Low concern Medium concern
Indiana Medium concern Low concern Low concern Low concern Medium concern
Iowa Medium concern Low concern Low concern Low concern Medium concern
Kansas Medium concern Low concern Medium concern Low concern Low concern
Kentucky High concern Medium concern Low concern Low concern High concern
Louisiana Low concern Low concern Low concern Low concern Low concern
Maine Medium concern Medium concern Low concern Low concern Medium concern
Maryland Low concern Low concern Low concern Low concern Low concern
Massachusetts Medium concern Low concern Low concern Low concern Medium concern
Michigan Medium concern Low concern Low concern Low concern Medium concern
Minnesota Medium concern Low concern Low concern Low concern Medium concern
Mississippi Low concern Low concern Low concern Low concern Low concern
Missouri Medium concern Low concern Low concern Low concern Medium concern
Montana Medium concern Low concern Low concern Low concern Medium concern
Nebraska High concern Low concern Low concern Low concern High concern
Nevada Low concern Low concern Low concern Low concern Low concern
New Hampshire Low concern Low concern Low concern Low concern Low concern
New Jersey Medium concern Medium concern Low concern Low concern Medium concern
New Mexico Low concern Low concern Low concern Low concern Low concern
New York Medium concern Low concern Low concern Low concern Medium concern
North Carolina Medium concern Low concern Low concern Low concern Medium concern
North Dakota Low concern Low concern Low concern Low concern Low concern
Ohio Low concern Low concern Low concern Low concern Low concern
Oklahoma Medium concern Low concern Low concern Low concern Medium concern
Oregon High concern Low concern High concern Low concern High concern
Pennsylvania High concern Low concern Low concern Low concern High concern
Puerto Rico Medium concern N/A* Low concern Low concern Medium concern
Rhode Island High concern High concern Low concern Low concern High concern
South Carolina Medium concern Low concern Low concern Low concern Medium concern
South Dakota Low concern Low concern Low concern Low concern Low concern
Tennessee High concern Low concern Low concern Low concern High concern
Texas Medium concern Low concern Low concern Low concern Medium concern
Utah Medium concern Low concern Low concern Low concern Medium concern
Vermont Medium concern Low concern Low concern Low concern Medium concern
Virgin Islands Medium concern N/A* Low concern Low concern Medium concern
Virginia Medium concern Low concern Low concern Low concern Medium concern
Washington Medium concern Low concern Low concern Low concern Medium concern
West Virginia Medium concern Medium concern Medium concern Low concern Low concern
Wisconsin Medium concern Low concern Low concern Low concern Medium concern
Wyoming Low concern Low concern Low concern Low concern Low concern

Notes

* Puerto Rico and Virgin Islands do not submit PI data and therefore were not assessed for enrollment benchmarking.

  1. Enrollment. To ensure a reasonably accurate number of enrollees, we assessed the average monthly difference between T-MSIS Analytic Files (TAF)-based enrollment counts and Eligibility and Enrollment (PI enrollment counts. States with a difference of 10 percent or less were assigned a low level of data quality concern, states with a difference between 10 and 20 percent were assigned a medium level of data quality concern, and states with a difference greater than 20 percent were assigned a high level of data quality concern.
  2. Missing eligibility information. To assign expenditures to the appropriate eligibility category, we examined the percentage of enrollees with missing eligibility information. States missing eligibility information for 5 percent or less of records were assigned a low level of data quality concern, states missing eligibility information for between 5 and 10 percent of records were assigned a medium level of data quality concern, and states missing eligibility information for greater than 10 percent of records were assigned a high level of data quality concern. Note that the DQ Atlas uses two measures in the Eligibility Group Code topic: (1) Percentage of beneficiaries missing an eligibility group code and (2) the number of large mandatory eligibility groups with no enrollment in the T-MSIS Analytic Files (TAF). The per capita expenditure analysis uses only the first measure because per capita expenditures are reported at a much higher level of aggregation than individual eligibility group codes.
  3. Low claims volume. To ensure CMS based spending estimates on a comparatively complete set of claims, we assigned a low level of data concern to states that had a claims volume of at least 10 percent of the national median of claims per 1,000 enrollee months for each claim type. Any state that did not meet this threshold was assigned a high level of data quality concern.
  4. Expenditures. Since fee-for-service claims and managed care capitation payments and other monthly per beneficiary payments (MBP)[4]  make up a large share of spending, CMS compared these amounts captured in the TAF to the amounts reported by states in the CMS-64 forms. States with a difference between the sum of fee-for-service plus monthly payments in TAF and CMS-64 of less than or equal to 5 percent were assigned a low level of data quality concern, states with a difference between 5 and 20 percent were assigned a medium level of data quality concern, and states with a difference of 20 percent or more were assigned a high level of data quality concern.

For more detail on the per capita expenditure calculations, please see the detailed methodology document. To view the version of the Medicaid and CHIP Scorecard that was published in October 2020, please visit the archived Scorecard page.

Table 1. Per capita Expenditure Estimates for States and Data Quality Assessment (2019)

State Data quality assessment Total Children Adult: non-expansion, non-disabled, under age 65 Aged People with disabilities Adult: ACA Medicaid expansion
Alabama Medium concern $5,582 $1,878 $1,555 $13,017 $13,317 -
Alaska Low concern $10,104 $5,862 $7,231 $22,714 $33,899 $9,298
Arizona Low concern $7,074 $3,232 $4,823 $9,581 $23,415 $7,637
Arkansas Medium concern $7,928 $4,822 $10,443 $20,633 $20,937 $1,647
California Medium concern $7,433 $3,122 $3,300 $16,501 $29,021 $5,806
Colorado Low concern $7,665 $2,947 $4,789 $21,408 $28,651 $5,650
Connecticut Low concern $8,405 $3,502 $5,619 $15,527 $28,234 $7,263
Delaware Low concern $9,634 $4,630 $8,092 $20,858 $24,914 $8,170
District of Columbia Low concern $11,405 $4,577 $6,511 $23,339 $33,567 $6,707
Florida Medium concern $5,387 $2,329 $2,754 $10,788 $14,423 -
Georgia Low concern $5,373 $2,588 $4,411 $9,880 $12,149 -
Hawaii Medium concern $6,890 $3,291 $4,923 $14,252 $23,758 $5,711
Idaho Low concern $7,966 $3,101 $7,525 $12,600 $21,460 -
Illinois Medium concern $7,033 $2,853 $5,100 $17,453 $15,123 $6,237
Indiana Medium concern $8,999 $3,227 $8,833 $18,875 $18,245 $10,072
Iowa Medium concern $8,897 $3,276 $7,667 $22,588 $25,242 $6,666
Kansas Medium concern $9,944 $4,096 $16,109 $23,097 $16,019 -
Kentucky High concern $6,827 $3,556 $6,451 $9,858 $14,454 $6,046
Louisiana Low concern $7,005 $3,523 $6,029 $10,893 $14,366 $6,611
Maine Medium concern $10,692 $4,504 $4,878 $13,718 $23,186 $6,719
Maryland Low concern $9,359 $3,746 $7,566 $19,056 $25,648 $8,876
Massachusetts Medium concern $10,288 $3,564 $3,974 $24,092 $20,119 $6,735
Michigan High concern $7,608 $3,154 $5,754 $20,658 $18,735 $5,628
Minnesota Low concern $11,829 $3,815 $6,731 $32,854 $43,171 $9,921
Mississippi Medium concern $7,954 $4,030 $5,192 $13,198 $14,475 -
Missouri Medium concern $10,607 $4,642 $6,696 $20,325 $26,652 -
Montana Medium concern $7,708 $4,304 $5,995 $17,719 $18,519 $7,097
Nebraska High concern $9,759 $4,615 $9,170 $24,342 $19,295 -
Nevada Medium concern $6,374 $2,834 $5,748 $8,877 $18,325 $6,918
New Hampshire Low concern $9,297 $4,212 $14,746 $24,632 $18,906 $6,551
New Jersey Low concern $10,066 $3,318 $9,678 $25,402 $31,248 $6,709
New Mexico Low concern $6,712 $3,921 $4,227 $11,021 $23,066 $5,931
New York High concern $9,762 $3,432 $6,735 $21,874 $27,242 $6,344
North Carolina Medium concern $6,820 $3,155 $2,653 $14,393 $19,097 -
North Dakota Medium concern $13,811 $4,694 $5,786 $35,124 $33,685 $13,966
Ohio Low concern $8,534 $3,470 $5,854 $21,039 $18,874 $7,456
Oklahoma Medium concern $7,671 $4,187 $5,197 $15,865 $18,494 -
Oregon High concern $9,787 $5,326 $11,872 $18,610 $18,674 $9,492
Pennsylvania High concern $12,261 $3,222 $6,333 $36,881 $26,079 $5,021
Puerto Rico High concern $2,142 $1,446 $2,299 $1,516 $2,001 $2,956
Rhode Island High concern $12,457 $4,558 $7,795 $27,521 $21,372 $9,068
South Carolina Low concern $5,028 $2,587 $2,659 $11,803 $14,594 -
South Dakota Low concern $8,561 $3,257 $6,792 $17,966 $21,385 -
Tennessee High concern $7,154 $3,985 $4,826 $14,624 $16,105 -
Texas Medium concern $9,084 $4,137 $8,536 $16,396 $24,151 -
Utah Low concern $10,322 $4,446 $9,746 $18,748 $28,612 -
Vermont High concern $9,721 $6,103 $7,489 $16,545 $22,933 $6,667
Virgin Islands High concern $8,264 $3,683 $7,402 $7,684 $15,845 $8,942
Virginia Low concern $9,349 $3,647 $7,618 $19,308 $25,136 -
Washington High concern $8,389 $3,223 $8,867 $21,227 $24,803 $7,442
West Virginia Medium concern $7,428 $3,063 $5,703 $21,316 $13,916 $5,344
Wisconsin High concern $8,436 $3,500 $6,838 $11,152 $20,818 -
Wyoming Medium concern $9,901 $3,945 $7,831 $20,227 $23,991 -
Min - $2,142 $1,446 $1,555 $1,516 $2,001 $1,647
Median - $8,436 $3,556 $6,451 $18,610 $21,372 $6,709
Max - $13,811 $6,103 $16,109 $36,881 $43,171 $13,966

Source: CMS calculations using calendar years 2018 and 2019 MBES expenditure data and T-MSIS data on enrollees and expenditures.

Note: Some states report significant capitation dollars as service tracking claims that might distort the distribution of per capita expenditure across eligibility groups. For example, Arkansas's Medicaid per capita expenditures for the adult expansion group are lower than other states because Arkansas submits Qualified Health Plan premium payments for this group as lump sum payment records on service tracking claims. This type of claim record cannot be linked to individual beneficiaries and were therefore not used to allocate CMS-64 expenditures across eligibility groups. The distribution of those expenditures across eligibility groups may appear to be distorted due to this data limitation.

Notes

To conduct this analysis, CMS used two data sources: (1) MBES expenditure data and (2) T-MSIS data. MBES expenditure data (reported by states on Form CMS-64) are at the state level and do not include expenditures at the enrollee or at the eligibility group levels; therefore, CMS used T-MSIS data to classify enrollees, allocate expenditures into eligibility groups, and construct the denominator (number of enrollee years) for each eligibility group.

To compute the denominator, CMS first mapped the 75 eligibility categories in T-MSIS to the five eligibility groups and then determined the count of enrollee months by eligibility group in each state. As the last step, total months were converted to total years.

To compute the numerator (total spending for each eligibility group) for each state, the calculation used total state-level expenditure information from MBES. To distribute MBES expenditures across the five eligibility groups, the calculation used T-MSIS data to determine the percentage of total Medicaid expenditures accounted for by each eligibility group, then applied that percentage to the total net Medicaid expenditures for medical assistance services that each state reported to MBES. For each state, these total net expenditures excluded spending categories not linked to individual enrollees, such as administrative expenses and disproportionate share hospital payments. This methodology largely aligns with methods that CMS’s Office of the Actuary employs to estimate national-level Medicaid spending per enrollee.

The approach used to calculate the per capita expenditure estimates is almost entirely the same as that used in the 2020 Scorecard. One notable exception is how we ensured each beneficiary was enrolled in Medicaid (and not CHIP) in the month in which a fee-for-service claim occurred. In addition, we allocated CMS-64 drug rebates expenditures across eligibility groups using the overall distribution of TAF expenditures rather than the specific distribution of TAF prescription drug expenditures.[2] Per capita expenditures for 2018 in this version of the Scorecard will differ from 2018 values displayed in the 2020 Scorecard because the data has been updated and because the per capita expenditures methodology has changed. See the methodology document for further details. 

Data quality assessment

To determine the completeness and accuracy of states’ T-MSIS data used to calculate per capita expenditures, CMS relied on four DQ Atlas assessments of key data elements. For some states with T-MSIS data that present a high level of data quality concern,[3] differences between their T-MSIS data and other CMS data sources may be due to a number of factors, including reporting issues that are not currently captured in the methodology and may be outside the control of states. In addition to the methodology, there are a range of state-specific factors and external reasons why a state may not have met the data quality requirements. Table 2 summarizes the results of the data quality assessment for 2018 and 2019. 

Table 2. Assessing State T-MSIS Data Quality (2019)

State Overall data quality assessment Enrollment benchmarking[a] Missing information on eligibility[b] Low claims volume[c] Expenditure benchmarking[d]
Alabama Medium concern Low concern Low concern Low concern Medium concern
Alaska Low concern Low concern Low concern Low concern Low concern
Arizona Low concern Low concern Low concern Low concern Low concern
Arkansas Medium concern Low concern Low concern Low concern Medium concern
California Medium concern Low concern Low concern Low concern Medium concern
Colorado Low concern Low concern Low concern Low concern Low concern
Connecticut Low concern Low concern Low concern Low concern Low concern
Delaware Low concern Low concern Low concern Low concern Low concern
District of Columbia Low concern Low concern Low concern Low concern Low concern
Florida Medium concern Low concern Medium concern Low concern Medium concern
Georgia Low concern Low concern Low concern Low concern Low concern
Hawaii Medium concern Low concern Low concern Low concern Medium concern
Idaho Low concern Low concern Low concern Low concern Low concern
Illinois Medium concern Low concern Low concern Low concern Medium concern
Indiana Medium concern Low concern Low concern Low concern Medium concern
Iowa Medium concern Low concern Low concern Low concern Medium concern
Kansas Medium concern Low concern Medium concern Low concern Low concern
Kentucky High concern Medium concern Low concern Low concern High concern
Louisiana Low concern Low concern Low concern Low concern Low concern
Maine Medium concern Low concern Low concern Low concern Medium concern
Maryland Low concern Low concern Low concern Low concern Low concern
Massachusetts Medium concern Low concern Low concern Low concern Medium concern
Michigan High concern Low concern Low concern Low concern High concern
Minnesota Low concern Low concern Low concern Low concern Low concern
Mississippi Medium concern Low concern Low concern Low concern Medium concern
Missouri Medium concern Low concern Low concern Low concern Medium concern
Montana Medium concern Low concern Low concern Low concern Medium concern
Nebraska High concern Low concern Low concern Low concern High concern
Nevada Medium concern Low concern Low concern Low concern Medium concern
New Hampshire Low concern Low concern Low concern Low concern Low concern
New Jersey Low concern Low concern Low concern Low concern Low concern
New Mexico Low concern Low concern Low concern Low concern Low concern
New York High concern Low concern Low concern Low concern High concern
North Carolina Medium concern Low concern Low concern Low concern Medium concern
North Dakota Medium concern Low concern Low concern Low concern Medium concern
Ohio Low concern Low concern Low concern Low concern Low concern
Oklahoma Medium concern Low concern Low concern Low concern Medium concern
Oregon High concern Low concern High concern Low concern High concern
Pennsylvania High concern Low concern Low concern Low concern High concern
Puerto Rico High concern N/A* Low concern Low concern High concern
Rhode Island High concern High concern Low concern Low concern High concern
South Carolina Low concern Low concern Low concern Low concern Low concern
South Dakota Low concern Low concern Low concern Low concern Low concern
Tennessee High concern Low concern Low concern Low concern High concern
Texas Medium concern Low concern Low concern Low concern Medium concern
Utah Low concern Low concern Low concern Low concern Low concern
Vermont High concern Low concern Low concern Low concern High concern
Virgin Islands High concern N/A* Low concern Low concern High concern
Virginia Low concern Low concern Low concern Low concern Low concern
Washington High concern Low concern Low concern Low concern High concern
West Virginia Medium concern Low concern Medium concern Low concern Low concern
Wisconsin High concern Low concern Low concern Low concern High concern
Wyoming Medium concern Low concern Low concern Low concern Medium concern

Notes

* Puerto Rico and Virgin Islands do not submit PI data and therefore were not assessed for enrollment benchmarking.

  1. Enrollment. To ensure a reasonably accurate number of enrollees, we assessed the average monthly difference between T-MSIS Analytic Files (TAF)-based enrollment counts and Eligibility and Enrollment (PI enrollment counts. States with a difference of 10 percent or less were assigned a low level of data quality concern, states with a difference between 10 and 20 percent were assigned a medium level of data quality concern, and states with a difference greater than 20 percent were assigned a high level of data quality concern.
  2. Missing eligibility information. To assign expenditures to the appropriate eligibility category, we examined the percentage of enrollees with missing eligibility information. States missing eligibility information for 5 percent or less of records were assigned a low level of data quality concern, states missing eligibility information for between 5 and 10 percent of records were assigned a medium level of data quality concern, and states missing eligibility information for greater than 10 percent of records were assigned a high level of data quality concern. Note that the DQ Atlas uses two measures in the Eligibility Group Code topic: (1) Percentage of beneficiaries missing an eligibility group code and (2) the number of large mandatory eligibility groups with no enrollment in the T-MSIS Analytic Files (TAF). The per capita expenditure analysis uses only the first measure because per capita expenditures are reported at a much higher level of aggregation than individual eligibility group codes.
  3. Low claims volume. To ensure CMS based spending estimates on a comparatively complete set of claims, we assigned a low level of data concern to states that had a claims volume of at least 10 percent of the national median of claims per 1,000 enrollee months for each claim type. Any state that did not meet this threshold was assigned a high level of data quality concern.
  4. Expenditures. Since fee-for-service claims and managed care capitation payments and other monthly per beneficiary payments (MBP)[4]  make up a large share of spending, CMS compared these amounts captured in the TAF to the amounts reported by states in the CMS-64 forms. States with a difference between the sum of fee-for-service plus monthly payments in TAF and CMS-64 of less than or equal to 5 percent were assigned a low level of data quality concern, states with a difference between 5 and 20 percent were assigned a medium level of data quality concern, and states with a difference of 20 percent or more were assigned a high level of data quality concern.

For more detail on the per capita expenditure calculations, please see the detailed methodology document. To view the version of the Medicaid and CHIP Scorecard that was published in October 2020, please visit the archived Scorecard page.


[1] The “High concern” category for per capita expenditures reporting was designed to include both the “high concern” and “unusable” categories from DQ Atlas

[2] Some states had a high percentage of managed care encounters in their prescription drug claims, and because the completeness of T-MSIS encounter payments has not been assessed, we used the overall distribution of expenditures to allocate drug rebates.

[3] The “High concern” category for per capita expenditures reporting was designed to include both the “high concern” and “unusable” categories from DQ Atlas.

[4] Monthly beneficiary payments are all monthly payments reported in the TAF Other claims file (OT) which would be claims with claim type = 2: Medicaid or Medicaid-Expansion Capitated Payment. They include: capitated payments to HMOs, HIOs, or PACE plans; capitated payments for primary care case management (PCCM); premium payments for private health insurance; and capitated payments to prepaid health plans (PHPs).

These data present average annual Medicaid expenditures per enrollee by state for five eligibility groups.