SB0897: SUMMARY OF BILL REPORTED FROM COMMITTEE (Date Completed: 4-18-18)

The exemptions would apply to a medical assistance recipient who was any of the following: the caretaker of a family member under six years old; pregnant; receiving temporary or permanent long-term disability benefits; a full-time student; designated as medically frail; the caretaker of a dependent with a disability who needed full-time care; the caretaker of an incapacitated individual; a recipient who had proven that he or she met a good cause temporary exemption; a recipient who had a medical condition that resulted in work limitations; a recipient who had been incarcerated within the last six months; a recipient of unemployment benefits; or a recipient under 21 years old who had previously been in a foster care placement in the State.

The bill also would enact Section 107a to define terms used in Section 107b and state the purpose of adding workforce engagement requirements "is to assist, encourage, and prepare an able-bodied adult for a life of self-sufficiency and independence from government interference".

"Able-bodied adult" would mean an individual at least 19 to 64 years of age who is not pregnant and who does not have a disability that makes him or her eligible for medical assistance under the Act.

Proposed MCL 400.107a & 400.107b Legislative Analyst: Suzanne Lowe

FISCAL IMPACT

If the bill were enacted and if the necessary Federal waivers were approved, the legislation would likely lead to an indeterminate but likely net marginal reduction in State expenditures. There are numerous elements in this estimate: 1) administrative costs, both one-time and ongoing, to implement the work requirement; 2) reduced Medicaid expenditures due to individuals, subject to the work requirement, who either did not meet the requirement and were rendered ineligible or increased their hours of work enough so their income exceeded the Medicaid eligibility threshold; 3) potential supportive services costs, in particular job training, child care, and transportation; 4) indirect revenue effects, including reduced Health Insurance Claims Assessment (HICA) revenue and, to the extent employment increased, increased income tax and sales and use tax revenue; and 5) potential secondary effects tied to changed incentives and increased income.

The magnitude of the nonadministrative costs and savings would depend on how many Medicaid-eligible individuals subject to the work requirement would increase their working hours to comply, how many such individuals would fail to comply and would be cut off from Medicaid eligibility, the number of the newly employed who would need supportive services, and the potential secondary behavior effects.

While various states have discussed implementation of Medicaid work requirements, there is no experience in other states with such requirements, so it is not possible to provide anything approaching a precise estimate of costs and savings. As a result, this fiscal analysis is vague. The key factors in any more precise estimate would include: the percentage of those subject to the requirements who failed to comply and were cut off from Medicaid, the percentage of those subject to the requirements who increased their income sufficiently to leave Medicaid, the potential increased tax revenue from those who increased their income (whether or not they left Medicaid), the percentage of the work requirement population who would require and seek child care and other supportive services, and administrative costs to implement and administer the work requirement.

Administrative Costs

The State has implemented work requirements for the Food Assistance Program (FAP) population, so one may expect that the administration of a Medicaid work requirement could be "grafted" onto administrative and systemic changes that were used to create a FAP work requirement. However, the populations, while they overlap, are handled differently in terms of eligibility and information technology. Eligibility for FAP is a group eligibility process handled by the Department of Health and Human Services BRIDGES system, while Medicaid eligibility is determined on an individual basis and is handled by the Community Health Automated Medicaid Processing System.

While there has not yet been any direct experience with a Medicaid work requirement, a

number of states and cities have estimated the administrative costs of implementing work requirements for human services programs. These estimates include $37.5 million in state funding for systems changes in Wisconsin, $70.0 million Gross (no state amount specified) for Medicaid work requirements in Tennessee, $17.5 million in state funding for Medicaid work requirements in Kentucky, and up to $23.1 million in state funding for Medicaid work requirements in the first full year in Virginia.

Based on these estimates and the fact that Michigan has a larger population than most of these states, the Senate Fiscal Agency (SFA) estimates costs in Michigan of $20.0 million to $30.0 million GF/GP per year for administration of a Medicaid work requirement.

Estimating the Base Population

The legislation would apply to nonelderly, nondisabled adult Medicaid recipients with exemptions built in for full-time students, pregnant women, medically frail individuals as defined by Federal rule, those engaged in job training, individuals receiving unemployment benefits, individuals being treated for substance use disorders, caretakers of children under six years old, caretakers of a dependent with a disability, individuals with work limitations according to a licensed medical professional's order, and individuals recently released from imprisonment. There also would be temporary good cause exemptions for the death of a family member, weather or family emergencies, and temporary illness or injury, as well as limitations on the work requirement if county unemployment were at or greater than 8.5%.

When attempting to estimate the nonelderly, nondisabled adult Medicaid caseload, one finds about 690,000 individuals who are enrolled in the Medicaid expansion program, the Healthy Michigan Plan (HMP), and about 300,000 individuals who are generally referred to as "TANF" (Temporary Assistance to Needy Families). That leads to a maximum population of just under 1.0 million who could be subject to the work requirements.

Due to the exemptions, especially those for full-time students, caretakers, and pregnant women, the 1.0 million figure should be considered to be a maximum well in excess of the actual number of people who would be subject to the work requirement.

The bill would exempt most individuals in counties with unemployment at or above 8.5%. In February 2018, 26 Michigan counties had unemployment rates at or above 8.5%. These counties are all small counties located in the northern Lower Peninsula and the Upper Peninsula. The relevant (nondisabled) Medicaid caseload in these counties is about 5.0% of the total State Medicaid caseload, so the SFA projects that 5.0% of the population subject to the workforce engagement requirements would be exempted due to the unemployment provision.

The population covered by the other exemptions is more difficult to estimate. For instance, the exemption for medically frail is tied to the definition in Federal rule 42 CFR 440.315(f), which mentions "individuals with chronic substance use disorders, individuals with serious and complex medical conditions, individuals with a physical, intellectual or developmental disability that significantly impairs their ability to perform 1 or more activities of daily living". Many of these people would be eligible for Medicaid in the aged, blind, and disabled category and so would already be exempt from the provisions of the bill. Others with often less severe conditions such as diabetes could potentially be exempt but the determination would likely be on a case-by-case basis and tied to diagnoses or an attestation from a physician, so it is difficult to estimate the population subject to the medically frail exemption.

The bill contains an exemption for those enrolled in postsecondary education related to employment. It appears that about 6.0% of the adult population in Michigan (not just the Medicaid population) is enrolled full-time at a community college, college, or university. However, this estimate reflects the percentage of the entire population enrolled full-time, not the percentage of the adult nondisabled Medicaid population.

The bill also contains an exemption for one parent in families with children under the age of six. It appears that there are about 100,000 single-parent families with children under the age of six and about 170,000 two-parent families with children under the age of six in Michigan. This combined total of 270,000 families represents about 8% of the households in Michigan. This estimate is based on overall population, not Medicaid population, so is certainly not precise.

There are a number of other exemptions whose impact is even more difficult to estimate. What is clear is that a significant portion of the nearly 1.0 million who would potentially be subject to the provisions of the legislation would be exempt. It is possible that a third or even over half the population would be exempt, depending on how medical exemptions were handled.

The exemption of a large portion of the population would reduce both the costs and the savings tied to the bill but would not have a major effect on the net fiscal impact as both the increases and decreases in expenditures would be reduced in magnitude

Medicaid Costs for the Base Population

An examination of Michigan's Medicaid physical health and behavioral health capitation rates for nonelderly adults in the Healthy Michigan Plan (HMP) and TANF populations, weighted for caseload, indicates that costs to cover these people average about $4,300 Gross per person per year. Services to the TANF population are reimbursed at a roughly 65.0% Federal match rate. Healthy Michigan Plan services are reimbursed at a match rate that is currently 94.0% but will drop to 90.0% on January 1, 2020. Because this population is so heavily weighted to HMP, the average GF/GP cost would be far less than the 35.0% or so GF/GP cost of the regular Medicaid population; in fact, the GF/GP cost for the population potentially subject to the work requirement, on average, would be about $600 per person per year.

Caseload Reduction: How Many Would Leave Assistance under the Work Requirement?

This question is by far the most difficult to address, although it is key in doing a fiscal

analysis of the legislation. There have not been any long-term studies on work requirements for Medicaid as no such work requirements have yet been implemented on any significant scale.

States have implemented work requirements for FAP benefits. In January 2017, Michigan implemented FAP work requirements in four counties. During a time that the FAP caseload in non-work-requirement counties dropped by almost 5.5% in Michigan, the caseload in those four counties dropped by 12.0%. One could assume that the 6.5% difference was due to the work requirement, with some recipients increasing their working hours and no longer being eligible due to income and others not meeting the work requirement and no longer being eligible due to violating the work requirement.

However, there are numerous factors that make that 6.5% figure an approximation with a potentially wide variance. First, the portion of the FAP caseload subject to the work requirement was a subset of the entire FAP caseload in those counties. Second, those counties may have seen better or worse economic growth than other counties, leading to an impact on the caseload for reasons aside from the work requirement. On the other hand, studies in other states on FAP work requirements have shown an impact of similar magnitude.

Caseload Reduction Savings

For the purposes of this analysis, the SFA has not attempted to estimate the impact of a work requirement that has not been implemented, especially on a population that overlaps but is still significantly distinct from the FAP population. Instead, the SFA analysis examines the impact of each 1.0% change in the affected caseload. If one starts with the maximum caseload of almost 1.0 million, each 1.0% reduction in the caseload due to the work requirement would reduce State Medicaid expenditures by an average of $43.0 million Gross and $6.0 million GF/GP. If there were a 6.5% reduction in State Medicaid expenditures, the average estimated savings would be $277.8 million Gross and $38.5 million GF/GP. Given that the population subject to the requirements would likely be much smaller than 1.0 million, the actual savings for a given 1.0% reduction in caseload would be smaller. This would be the case whether the individuals were no longer eligible for Medicaid due to violating the work requirement or due to having increased income.

HICA Offset

The State has a 1.0% tax on paid medical claims known as HICA. For each 1.0% reduction in caseload, the State would see a reduction in HICA revenue of $430,000, which would increase GF/GP costs by the Federal share of this amount (due to actuarial soundness requirements) or about $370,000. As HICA is slated to expire on July 1, 2020, this would not be a long-term reduction in revenue.

Supportive Services Costs

The most significant supportive services cost would be child care for those covered by the legislation who have children and who increased working hours. It is important to note that the vast majority of those covered by the legislation either would not have child care needs because they do not have children under 18 requiring child care, are already subject to work requirements, or have access to alternative child care arrangements. The average State cost of full-time child care is about $5,000 per year. This amount would be far less for children who are enrolled in school and, again, would apply only to a limited subset of the covered population.

Other Secondary Effects

It is possible that some recipients would choose to apply for Social Security Income (SSI) benefits to avoid the work requirement. To the extent that these people are already eligible as TANF cases, the fiscal impact would be on the SSI State Supplementation line - each newly eligible SSI recipient would receive a Federal SSI check along with an average $225 per year State Supplementation payment. That increased cost would be more than offset by ending a GF equivalent TANF payment of about $5,000 per year, for net GF/GP savings on grants of almost $4,800. The cost of Medicaid for these now Aged, Blind, and Disabled-eligible individuals would increase by about $3,000 GF/GP per year, for net savings of $1,800 GF/GP per case per year.

If a person currently eligible under HMP applied for and received SSI, then the increased capitation rate and lower Federal match rate would lead to an average GF/GP cost increase of $4,000.

It should be noted that the magnitude of the SSI payment itself, which is over $9,000 per year (almost all Federal) for those in the independent living category, already provides a strong incentive to individuals to apply for SSI benefits even without a work requirement. The work requirement itself would likely have only a marginal impact on individual decisions to apply for SSI.

To the extent that people increased their working hours due to the work requirement, State income and sales and use tax revenue would increase. If a person increased his or her income by $10,000 per year, the State would gain about $250 in income tax revenue (after adjusting for the personal exemption) and an indeterminate but likely similar amount of sales and use tax revenue. Furthermore, for those receiving Family Independence Program (FIP) payments, increased income would lead to a reduction in their FIP grant.

People who were sanctioned and forced to leave Medicaid would likely not be able to afford health insurance, which could lead to an increase in uncompensated care, especially at hospitals. This would lead to a negative State or local fiscal impact for public hospitals.

Summary

The potential net costs or savings of the legislation, if fully implemented, are tied to the degree to which people would leave Medicaid as a result of the work requirement, whether due to increased income or due to failure to meet the work requirement. There is a highly complex set of factors to analyze because no state has yet implemented such a work requirement.

To the extent that individuals would begin to work or increase their hours, there would be cost reductions as some would migrate off of Medicaid, there would be marginal increases in tax revenue, and there would be supportive services costs, especially for child care. To the extent that individuals were removed from Medicaid due to failure to comply with the work requirements, there would be reduced Medicaid costs but potential increases in uncompensated care for hospitals, with a State and local fiscal impact on publicly owned hospitals. The savings for those leaving the Medicaid caseload for whatever reason would average just over $600 GF/GP per case. The costs, in particular child care costs, would be focused on a much narrower subset of the population. Increased income for those in the FIP