Background. The Temporary Assistance to Needy Families, enacted under the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, is a reality for many working families. As public policies are enacted, unintended consequences for infants/children must be minimized. Child advocates in Wisconsin, leading this nation in reforming Aid to Families with Dependent Children (AFDC), are concerned about supporting eligible infants/children as safety-net programs are unlinked.
Objective. This study reviews the enrollment status of 4 linked programs over time in Wisconsin, from January 1995 to August 1998. Eligible infants/children in programs, such as Medicaid/AFDC, Medicaid/Healthy Start, and Food Stamps, were analyzed and compared with enrollment in Special Supplemental Nutrition Program for Women, Infants/children (WIC), a nonlinked program.
Design. A cross-sectional analysis of monthly enrollment for infants/children was subdivided into 3 periods: prewelfare reform or AFDC (January 1, 1995 to December 31, 1995), the welfare reform pilot or Pay For Performance (January 1, 1996 to August 31, 1997), and welfare reform better known as Wisconsin Works (W-2), (September 1, 1998 to August 31, 1998), periods 1, 2, and 3, respectively.
Participants. Infants/children in Wisconsin from birth to 18 years of age enrolled in W-2 and/or other safety-net programs were monitored: AFDC or W-2, WIC, Food Stamps, Medicaid/AFDC, and Medicaid/Healthy Start.
Results. The average number of infants/children removed from AFDC and Medicaid/AFDC during periods l and 2 were −1210 increasing to −3128 per month, respectively, almost tripling the rates of decline during the pilot period (see Fig 2). By the end of this study, >100 000 (111 198) infants/children were removed from AFDC/W-2 enrollment and 51 559 fewer infants/children benefited from Medicaid. This rate of decline slowed during period 3, averaging −687 per month, while W-2 enrollment continued to decline significantly at a rate of −2692 per month. In contrast, Medicaid/Healthy Start enrollment, targeted to infants/children <6 years of age, increased significantly over all periods by +332, +1327, and +266, respectively. Food Stamps enrollment also declined throughout all 3 consecutive periods, −603, −2462, and −1450, respectively. However, enrollment in the WIC program did not decline significantly to the same degree as other certification-linked programs with AFDC or W-2, as indicated by the consecutive slopes of −60, −111, and −183, respectively.
Conclusion. Wisconsin infants/children were rapidly removed from welfare rolls in unprecedented numbers during the periods January 1995 and August 1998. Comparisons of periods before W-2 implementation and 1 year after implementation support the fact that certification-linked programs, such as Medicaid and Food Stamps, were sufficiently aligned to AFDC/W-2 to significantly impact infants/children enrollment. Historically, WIC certification in Wisconsin has not been linked to AFDC, and infants/children traditionally eligible for Medicaid and Food Stamps are also eligible for WIC. Yet, contrary to the AFDC-linked safety-net programs, declines in WIC enrollment were not statistically significant during all study periods. Statewide and local interventions within Wisconsin, such as outreach activities, targeted to Medicaid/Healthy Start and more recently Title XXI (State Children Health Insurance Program), slowed the reductions of Medicaid enrollment for Wisconsin infants/children.
These findings support that altering safety-net programs can result in unintended consequences if not carefully transitioned as demonstrated in Wisconsin welfare reform.
Health insurance status has been well-established as the most fundamental mediator for access to health care.1,2 Historically in Wisconsin, eligibility for government cash assistance under the entitlement program, Aid to Families with Dependent Children (AFDC), automatically qualified families for other government assistance programs, such as health insurance—Medicaid or Healthy Start, food stamps, and child care support. Children living in AFDC-subsidized households, (both single-parent and 2-parent households) generally rely on Medicaid for health coverage.3 Successful employment in families with children having chronic health conditions can be critically impacted by their access to health coverage. Consequences in Wisconsin with the delinking of Medicaid from the cash assistance program can result in limited resources for working families unless measures are taken to ensure continued health insurance, nutritional support, or supportive services for infants/children. A study conducted by Moffitt and Slade4 between 1989 and 1992 found that states with the most generous Medicaid expansions have the highest employment rates. According to the Urban Institute, Medicaid coverage is particularly important for children on welfare because families on cash assistance programs generally have other factors that hamper their successful employment, including lack of education, the presence of other children, and limited opportunities in the local labor market.5 Moffitt and Slade4 cite that limited health insurance options for working families faced with health problems constrain employment options for some families and expose others to greater health risks because of lack of health insurance coverage. Other studies have shown that a significant relationship exists among inadequate Medicaid benefits levels, above average medical expenditures, and poor health status for families on AFDC. Darnell and Rosenbaum2 confirmed that in situations in which separate enrollments were required for medical coverage and other social programs, individuals delayed their enrollment in health insurance plans until catastrophic care was required. Thus, these families decrease their likelihood of using primary and preventive care.1,2 Again, these associations between successful employment and Medicaid emphasize the value of securing a predictable linkage for access to health insurance and successful employment.
The State of Wisconsin started early reforming the traditional welfare system, long before the federal Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) was passed in 1996.6,7 As a result, Wisconsin has rapidly replaced cash assistance with a self-sufficiency employment program called Wisconsin Works (W-2).7 Many policy makers refer to Wisconsin's welfare reform strategies as a model for implementing welfare reform in other states.6,7 Therefore, the impact of these strategies on infants/children is critical to identify and report for interventions to minimize unintended consequences. In this study, the enrollment of infants/children in AFDC or W-2 was monitored over 4 years (January 1, 1995 to August 31, 1998) including the first full year under PRWORA, W-2. Enrollment in 4 additional safety-net programs, such as Medicaid/AFDC, Medicaid/Healthy Start, Food Stamps, and Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), designed to be safety-nets for poor and near-poor families, will be concurrently analyzed over the same study periods.
Congress severed the connections between Medicaid and welfare in the enactment of PRWORA.6,8 Medical Assistance Program (Medicaid) in Wisconsin remains an entitlement program based on AFDC standards as of July 16, 1996 having different eligibility standards from W-2.9 Low-income children may be eligible for Wisconsin Medicaid, including Medicaid/Healthy Starts, even if their parents elect not to participate in W-2. Income-eligible infants/children remain entitled to receive Medicaid and/or Healthy Start within Wisconsin. Healthy Start is a component of the Medical Assistance Program (Medicaid) that pays for medical care for pregnant women, infants, and children up to 6 years of age. Starting in July 1996, persons must have income levels below 185% of the federal poverty level (FPL) to be eligible for the Healthy Start program.
Families in Wisconsin are determined to be eligible for W-2 based on income levels of <165% of the FPL.10 Only the economic support workers of each county are authorized to determine eligibility for Medicaid/AFDC recipients. Welfare policy changes, as of September 1, 1997 associated with the implementation of W-2, resulted in eligibility for Medicaid being determined through a separate process for many eligible families based on income levels or special health care needs. Children up to age 14 years are covered in Medicaid as long as the family income is no >100% of the FPL. Conversely, families not eligible for Temporary Assistance to Needy Families (TANF) benefits under W-2 may still be eligible for Medicaid. In fact, as the national enrollment for Medicaid lags behind the number of infants/children eligible, many experts anticipate that in the long-term as the wage-earning capacities of parents increase, infants/children will become dependents of their parents employment-related health insurance, thus fewer infants/children were eligible for Medicaid. In the short-term, we cannot assume that the pool of infants/children eligible for Medicaid will be reduced or that their enrollment into employer-provided insurance as dependents of working families will increase. Public officials, health professionals, and child advocates for working families in Wisconsin had voiced concern about the declining Medicaid enrollment even before the enactment of W-2.2,5,6,11,12 These concerns turned into realities for many families and providers as Medicaid enrollments dropped sharply with growing numbers of uninsured children reported in Wisconsin during the first year of piloting welfare reform.13,14
Another safety-net program, Food Stamps, provides low-income households with food assistance by dispersing coupons to families living at ≤130% of the FPL. The relationship between TANF and the Food Stamp program has been significantly altered and allows states much discretion in its implementation. In 1992, studies conducted by Devaney et al14 revealed that 95% of income-eligible preschool children and 86% of income-eligible older children and adolescents participated in the Food Stamp program. Under PRWORA, the Food Stamp program freezes the standard deductions from an applicant income at the fiscal year 1996 level and counts the state and local energy assistance as income.15 States are allowed to use food stamps benefits for wage subsidies. As a consequence of welfare reform, Wisconsin now counts the dollar amount, specifically for foods stamps received by a family toward their income when determining eligibility for W-2. Also under the new W-2 guidelines, children under 21 years of age, including those with children of their own, must apply for food stamps using their parent's income levels.
The last safety-net program concurrently reviewed in this analysis was WIC. WIC is a nutrition program targeted for pregnant and lactating women and infants and children at nutritional risk who live at or under 185% of the FPL. The objective of this nonentitlement program is to raise the nutritional status of participants during critical development periods of pregnancy, lactation, infancy, and childhood. Strategies to accomplish this objective are through the coordination of supplemental foods, nutrition education, and appropriate referrals to health and social services providers.15 Also, the certification process for families eligible for WIC requires an income assessment; to date, this program has not been linked to cash assistance programs or welfare programs in Wisconsin.
For this study, AFDC, W-2, and Food Stamp enrollment numbers were provided by the State of Wisconsin, Department of Workforce Development. Monthly, Medicaid enrollment data, stratified by Medicaid/AFDC and Medicaid/Healthy Start, were supplied by the State of Wisconsin, Department of Health and Family Services, Division of Health Care and Finance. The State of Wisconsin Department of Health and Family Services, Division of Public Health also provided WIC data. Enrollment by specific safety-net programs of infants/children was counted monthly, and children could concurrently be counted in one or more of these safety-net programs.
The study design was a retrospective cross-sectional analysis subdividing the study periods into 3 consecutive periods. The baseline study period 1 occurred between January 1995 and December 1995, during the time the traditional AFDC program was in place and before the enactment of Wisconsin pilot welfare reform. Study period 2 occurred between January 1996 and August 1997, representing a prephase of W-2 for selected counties in Wisconsin known as the pilot and referred to as Pay For Performance. Throughout the second study period, welfare recipients anticipated future changes that would affect their government cash assistance. Study period 3 was designated between September 1997 and August 1998, representing the implementation phase of W-2. By the end of August 1998, all of the AFDC families were terminated or transferred into the W-2 program.
We hypothesized that with the implementation of welfare reform, W-2, in the short-term no significant change would occur for infants/children of working poor families as enrollees in Medicaid, WIC, and Food Stamps. Multiple linear regression analysis was used to measure the rate of change in infants/children enrollment for each safety-net program throughout each designated study period. In other words, no significant difference in the slopes would be anticipated when comparing periods 1 versus 2, periods 2 versus 3, and periods 1 versus 3. The 3 study periods were incorporated into this regression model by using 2 dummy variables and by allowing an estimation of the slope study period 1 to be used as the baseline. Changes in slopes for consecutive periods for both 2 and 3 as they compare with the baseline were considered. The hypothesis was tested for each program period andP values were calculated using SAS (SAS, Cary, NC).16
The enrollment of infants/children in AFDC started at ∼148 792 infants/children in January 1995 and declined to 37 594 by August of 1998. This is a 75% reduction in enrollment by this targeted population (Fig 1). Using AFDC as an independent variable, the reduction in the number of children enrolled in Food Stamps, a program linked to AFDC across all study periods, was highly significant. The Food Stamps program enrollment started at 175 844 infants/children in January 1995 and declined to ∼104 200 by August 1998, representing a 41% reduction in enrollment. However, the WIC program did not experience a significant decline between any of the compared study periods, revealing only a 6% overall reduction between January 1995 enrollment of at 87 037 statewide and 82 152 infants/children in August 1998. Since the health insurance coverage by both Medicaid and Medicaid/Healthy Start was delinked from TANF in September 1997, we used the cash assistance programs (AFDC, Pay For Performance, and W-2) as independent variables, and we analyzed Medicaid/AFDC and Medicaid/Healthy Start for changes in infants/children enrollment across all 3 periods. Medicaid/AFDC infants/children enrollment across consecutive periods declined from 187 173 infants/children starting in January 1995 to 100 336 in August 1998, a 47% reduction by participants. In contrast to the drastic unanticipated Medicaid/AFDC reductions, Healthy Start enrollment experienced a statistically significant increase over the consecutive periods, starting with 42 659 infants/children in January 1995 and increasing to 77 937 participants by August of 1998.
AFDC enrollment for infants/children in Wisconsin experienced a −1210 decline per month during study period 1. During the pilot period, Pay For Performance, 2.6 times fewer (−3146) infants/children benefited from Wisconsin cash assistance programs. This enrollment reduction for infants/children during W-2 continued at 2.2 times (−2692) the baseline rates during study period 3. Therefore, with the transfer of welfare recipients from AFDC to employment under W-2, a significant reduction in infants/children enrolled in transitional cash assistance programs occurred in Wisconsin. The compared rates of change between periods 1 and 2, and periods 1 and 3 were significant to Pvalue <.0001, while P value <.0422 between periods 2 and 3 occurred (Fig 2).
Medicaid/AFDC enrollment decreased at a similar baseline rate as AFDC, −1210 infants/children per month during period 1 and before the implementation of welfare reform. The rate of decline per month during period 2 was similar, 2.6 times the rate of decline that occurred in period 1 or −3110 infants/children per month. The reduction during period 3 was at −687 infants/children per month, much less than that experienced during the pilot period in preparation for W-2. These changes were highly significant at P value <.0001, when comparing across consecutive periods 1 versus 2 and 2 versus 3. Reduction in enrollment for infants/children for the compared periods of 1 versus 3 was borderline significant at P value <.0610.
Medicaid/Healthy Start was programmatically targeted for statewide aggressive outreach activities in Wisconsin during the pilot period 2. Baseline period 1 reveals a positive rate of change of +332 infants/children per month and 4.0 times as many infants/children enrolled during period 2, +1327. During period 3, the Healthy Start program increased enrollment by only .8 from the baseline rate. These changes in enrollment for infant and children under the Healthy Start program was statistically significant between the compared periods of 1 versus 2 and 2 versus 3 at P value <.0001. When comparing rates of change between periods 1 and 3, there is not a statistically significant change (P < .787) in enrollment rates for this program. By the end of period 3, Medicaid/Healthy Start enrollment approached the WIC enrollment. Of note is that these 2 programs (WIC and Healthy Start) are targeted to the exact same age populations, birth to 6 years of age.
The Food Stamp enrollment for infants/children during baseline period 1 declined by −603 infants/children enrollees per month. Reduction in enrollment was 4 times (−2462) and 2.4 times (−1450) the rates of change from the baseline period for periods 2 and 3, respectively. Between periods 1 and 2, a significant decline in Food Stamp occurred at P value <.0001. Compared reductions in receiving food stamps for infants/children between periods 1 and 2, 2 and 3, and 1 and 3 were all highly significant at P values <.0001. The actual data for this program were supplied quarterly over the 4 years and points extrapolated for the other months based on a formula derived from the actual numbers to fit the line.
Enrollment in WIC declined by −60 for infants/children per month during period 1, by −111 per month during period 2, and by −183 per month during period 3. These reductions in enrollment when compared across all periods (1 vs 2, 2 vs 3, and 1vs 3) were not statistically significant and seem to not be associated with changes occurring within the cash assistance program.
See Table 1 for comparisons of enrollment of infants/children by programs. In conclusion, Medicaid/AFDC, W-2, and Food Stamps enrollments declined at statistically significant rates for infants/children between periods 1 and 2 and 2 and 3. Food Stamps and AFDC enrollments continuously declined at statistically significant rates across the 3 study periods. Medicaid/Healthy Start enrollment increased at statistically significant rates during periods 1 versus 2 and 2 versus 3, but not when comparing periods 1 and 3. Contrary to trends experienced in other AFDC-related programs, WIC enrollment, which is not AFDC-linked, did not have statistically significant reductions during any comparison periods (see Table 1 for summary).
We have demonstrated that with the implementation of welfare reform in Wisconsin, the number of infants/children enrolled monthly in AFDC/W-2 and Food Stamps continued to decline significantly throughout all periods. Medicaid/AFDC enrollment suffered similar declines but was corrected attributable to other statewide actions to avert unintended consequences. In fact, the number of infants/children enrolled in Medicaid/Healthy Start increased across every consecutive period because of aggressive outreach activities throughout the state. Although WIC is targeted to similar income families as AFDC, Food Stamps, and Medicaid, this program did not experience similar reductions. We speculate that the lack of decline in WIC enrollment might be caused by the long-term lack of linked certification processes between WIC and cash assistance. In Wisconsin the WIC certification process historically operated separately from the other governmental assistance programs that were reviewed in this case study. It is doubtful that the infants/children enrolled in WIC are from different families than those participating in AFDC or W-2, Medicaid, or Food Stamps. WIC program has similar income criteria as Wisconsin Medicaid/Healthy Start, but WIC families use safety-net programs in different degrees, because they are located throughout local community sites and not centralized as the cash assistance certification has been in county government.
Although Wisconsin has one of the lowest uninsured rates for children in the United States (6.0% in 1997), this study reveals serious attention must be given to the dramatic decrease in Medicaid and Food Stamps enrollment as those programs have been closely paralleled with AFDC and W-2. Because the delinking of government case assistance from public health insurance resulted in more eligible infants/children being without health insurance or other safety-net services, welfare reform efforts can threaten the ability of working families to secure needed resources for their infants/children as dependents. If TANF is to be successfully implemented, officials in the state of Wisconsin and other states implementing similar safety-net programs must aggressively address unanticipated consequences. Of interest to policy makers and pediatricians is the fact that children potentially eligible for medical assistance program, as supported by data in the 1995 and 1996 Wisconsin Family Health Survey, estimated that ∼477 000 to 526 000 children live in poor or near-poor households.17 This survey defined poor as a family below 100% of the FPL, and near poor as families living between 100% and 200% of the FPL level. The Wisconsin Family Health Survey conducted between 1995 and 1997 indicates that children from poor and near-poor households were uninsured at a greater rate than nonpoor children (18% and 13% vs 4%, respectively.17 Officials in the Department of Health and Family Services recognized early on in the implementation process of welfare reform that aggressive outreach to Medicaid-eligible families was needed to offset reductions that occurred in Medicaid/AFDC enrollment. By August 1998, 51 559 fewer infants/children received medical assistance within the state of Wisconsin.
In response to the circumstances created with the delinking of safety-net programs, extensive Medicaid outreach efforts were conducted in Wisconsin at medical clinics and hospitals to ensure that families not eligible for welfare with incomes that leave them eligible for Medicaid were being served. As of July 1, 1999, Title XXI the Children Health Insurance Program in Wisconsin (BadgerCare) was available to working families with dependent children who no longer qualify for cash assistance under welfare reform or Medicaid. BadgerCare builds on the existing statewide Medicaid managed care expansion and bridges the gap between Medicaid and private insurance for eligible working families. Families have access to Badgercare when their income is <185% of the FPL and remain eligible until the family's incomes rise above 200% of the FPL. Copayments only start when families' incomes exceed 150% of the FPL.
In statewide efforts to reach vulnerable populations of infants/children, officials of the state aggressively pursued the implementation of the Healthy Community Initiatives to ensure that all eligible families enroll in Medicaid, private insurance, or the family health plan, such as BadgerCare under Title XXI. Outreach programs are being implemented through local partnerships with the state. These local partnerships improve health care access for working families through community-based efforts within schools, health care providers (ie, community health centers and hospitals), and welfare reform certification sites. These efforts targeted to families eligible for Medicaid/AFDC, Healthy Start, and, subsequently, BadgerCare have slowed reduction trends in enrollment for infants/children eligible for health insurance in Wisconsin. Table 2 on Wisconsin Public Health Insurance Enrollment for Infants/Children illustrates the drastic drop-off in Medicaid/AFDC and then the build-up in Healthy Start and BadgerCare enrollment.
Although welfare reform was deliberately instituted, the associated changes for other government assistance programs need to be appreciated and corrected to prevent undesirable consequences for infants/children. In the state of Wisconsin, the administrative policy of delinking the certification process of AFDC/Medicaid, effective September 1997, has been associated with significant reductions in Medicaid/AFDC infants/children previously eligible for AFDC and automatically certified for health insurance coverage under the Medicaid entitlement program. Outreach activities to enroll children <6 years of age resulted in a statistically significant increase in Healthy Start enrollment. Building on those successes, specific components of Wisconsin's outreach plan include:
System changes to support an efficient information system for eligibility determination;
Statewide public information campaigns to minimize the fallout caused by the delinking of welfare and Medicaid;
Training providers, school officials, public health agencies, W-2 agencies, and community-based agencies;
Regional and local initiatives, for example, out-stationing of certification that can be replicated and coordinated; and
State help desk function for case-specific resolution of problems.
It is conceivable that some former AFDC families attained higher economic status or that these declines occurred because of the unanticipated effects of delinking within the certification process. Welfare reform changes and associated consequences should be continuously monitored with caution and diligence to intervene for the protection of infants/children. Earlier trends in enrollment for infants/children in Wisconsin safety-net programs suggest that efforts need to be implemented to ameliorate these unintended consequences. The impact social reform such as welfare reform must be monitored to appreciate where within the health care system will infants/children outside of the health insurance system seek catastrophic care—expensive care. Although the reduction in AFDC and W-2 could have been forecasted, reductions in enrollment for infants/children in Medicaid/AFDC by 46.3% and in Food Stamps by 40.7% were not necessarily anticipated. To avoid additional unintended consequences, other beneficial programs for working families not discussed or analyzed in this study require monitoring and analysis for unintended consequences. Safety-net programs, such as those not examined in this study including subsidized childcare, transportation, subsidized housing, and child nutrition programs (such as, school breakfast and lunch programs) need to be maximized for their benefit to infants/children in our community. Successful welfare reform is integrally intertwined with the maintenance of effective safety-net programs for working families and children.
We thank Richard Aronson, MD, MPH, Maternal and Child Health, Wisconsin Division of Health, Bureau of Public Health; Susan Wood, John Haine, and Wayne Thompson, Wisconsin Department of Health and Family Services, Division of Health, Bureau of Health Care Financing; Jean Rogers, Ingrid Rothe, and Beth Dorschner, Department of Work Development, Division of Economic Support, for providing access to the data for this study; Varghese George and Steve Subichin, Division of Biostatistics, Health Policy Institute, Medical College of Wisconsin, for their expert advice on statistical issues; and John Meurer, MD, MM, Assistant Professor, Medical College of Wisconsin, Department of Pediatrics, for his reviews.
- Received April 10, 1999.
- Accepted June 27, 2000.
Reprint requests to (E.W.) Department of Pediatrics, Medical College of Wisconsin, 8701 Watertown Plank Rd, Milwaukee, WI 53226. E-mail:
- AFDC =
- Aid to Families with Dependent Children •
- PRWORA =
- Personal Responsibility and Work Opportunity Reconciliation Act •
- W-2 =
- Wisconsin Works •
- WIC =
- Special Supplemental Nutrition Program for Women, Infants, and Children •
- FPL =
- federal poverty level •
- TANF =
- Temporary Assistance to Needy Families
- Darnell J,
- Rosenbaum S
- Greenberg M. Participation in Welfare and Medicaid Enrollment. Menlo Park, CA: The Henry J. Kaiser Family Foundation; 1998:1–15
- Urban Institute. Realistic Expectations for Welfare Mothers. Washington, DC: Urban Institute; 1998
- Ellwood KU. Welfare and immigration reforms: unintended side effects for Medicaid. Health Aff. 1998;May/June:137–145
- Rector R
- Personal Responsibility and Work Opportunity Reconciliation Act, Public Law 104-193 (1996)
- Wisconsin Department of Work Force Development. W-2 Fact Sheets. Milwaukee, WI: Wisconsin Department of Work Force Development; 1997
- Dresang J, Stephenson C. Work of Else: Special report: the launch of Wisconsin's sweeping welfare reform program. Milwaukee Journal Sentinal. May 18, 1997
- Dresang J, Manning J. Low Medicaid numbers lead to concerns. Milwaukee Journal Sentinal. February 15, 1998
- Dresang J. Welfare confusion persists over Medicaid, health providers say. Milwaukee Journal Sentinal. July 30, 1998
- SAS Institute Inc. SAS/STAT Users' Guide, Version 6, II. 4th ed. Cary, NC: SAS Institute Inc; 1989:846
- Wisconsin Department of Health and Family Services. Wisconsin Family Health Survey 1997. Madison, WI: Wisconsin Department of Health and Family Services, Division of Health Care Financing, Bureau of Health Information; 1997:15–18
- Copyright © 2000 American Academy of Pediatrics