(Archived document, may contain errors)
September 6. 1995
THE IMPACT OF NEW JERSEYPS FAMILY CAP ON OUT-OF-WEDLOCK BIRTHS AND ABORTIONS
Robert Rector Senior Policy Analyst New Jersey is the only state in the nation that has instituted a family cap policy: denying an in- crease in cash welfare benefits to mothers who have additional children while already receiving welfare. The evi- dence currently :7= ._;z:' available from New Jersey indi- Births to Women on AFDC in New Jersey: cates the family cap has resulted Before and After the Family Cap in a decline in Pre-Fami@ Cof Average births to women 1,500 '33 1.460 Post- Famik Cap Average on AMC but 1.43 (9193-11194) not an increase 1,450 1.396 1,197 in the abortion 1,400 1,366 1 1,352 1.351 1,338 rate. 1,350 1.295 In all states 1,300 1,317 1.267 59 except New Jer- 1,250 1,298 .231 1.223 1,226 1.219 1.224 sey, AMC 1,200 1,219 1.248 1.208 1.228 mothers who 1,197 have additional 11150 1.170 1,150 children while 1,100 Family Cap Enacted 11110 on welfare re- 1,0SO August 1"31 ceive an auto- matic increase in welfare bene Sourve: Administrative Siobility and Benefit Client Databases, Division of Family Development (NIDFD). New Jersey Department fits. But in of Human Services Data not available on the birth rams prior to August 1992 or past November 1994. 1992, Demo- cratic Assembly- man Wayne Bryant won passage in the New Jersey legislature of an innovative welfare reform 1,43
*396 3 2 '066 W1.338 .350 1. 5 1 317
231 1.298 8 1@" @1@.2191.224 1.24 1.2 .228
known as the family cap. Under the family cap, mothers already enrolled in AFDC no 7fl#-0195
longer receive an auto- Chart 2 . . . . . . . . . . . . . . . . matic increase in AFDC benefits after Birth Rate for Women on AFDC in New Jersey: giving birth to addi- tional children. The Before and After the Family Cap family cap went into 13 Births per I = Woffm effect in October Pre-Family Cap Average: 1992 with a ten- 10.96 11.89 Pm- Family Cap Average month grace period. 12 112 (9/93-11/94) 1136 1@@ 9.72 Thus, the limitation 1132 on benefits applied to households with chil- 11 @@V moA 11.09 10.49 dren born after July 10.74 10.32 9.97 9.97 '9 1993. 10 VVIO. 12 NS@.02 "".A A 9_79/N@@9.97 9.5\\1/ V pact of the family cap 9.44 928 The financial im- 9 9.79 9-89 9.06 -7 on the value of wel- FamUy Cap Enacted 829 8.93 fare benefits for AWst 1993 AFDC mothers was quite small. The $44 J1P e benefit increase elimi- Source: See Chart I source- nated by the family -1. 1 . . . I cap constituted only 4 percent of the total monthly welfare benefits received by the average AFDC mother in New Jersey. Chart I shows the actual monthly number of live births to AFDC mothers before and after imple- mentation of the family cap. The chart indicates that in the 16 months after the cap went into effect (August 1993), them was a significant decrease in out-of-wedlock births. In the year before the cap, the average monthly number of births was 1,33 1; in the 16 months after the cap went into effect, av- erage monthly births among AFDC mothers fell to 1, 197. The cap appears to have caused an aver- age decrease of 134 births per month, or 10 percent. Chart 2 shows the same data as Chart 1, expressed as a birth rate (monthly births per 1,000 AFDC mothers). The chart shows that in the year before the family cap, the monthly birth rate was 10.96 per 1,000 AFDC mothers. In the 16 months after the cap went into effect (August 1993), the aver- age monthly birth rate fell to 9.72 per 1,000, a decline of over 10 percent. The data in the charts represent the information available from the Administrative Eligibility and Benefit Client Databases maintained by the Division of Family Development (NJDFD) of New Jer- sey's Department of Human Services in Trenton. Data on birth rates prior to August 1992 or after November 1994 have yet to be compiled by NJDFD and were unavailable for this study. Critics claim that the family cap has not caused an actual reduction in the number of illegitimate births, but merely a delay in welfare mothers reporting births to the welfare office. They assert that since mothers subject to the family cap no longer receive higher AFDC benefits upon the birth of an additional child, they are less prompt in notifying the welfare department of births. However, under the family cap, AFDC mothers still have a strong financial incentive to notify the welfare department of any birth. Only their AFDC benefits are limited; mothers on AFDC in New Jersey and subject to the family cap still receive increased food stamps and Medicaid benefits for each additional child born. Therefore, they still have an incentive to notify the welfare department of a childbirth.
The data presented in Charts I and 2 in any case have been retroactively adjusted for late report- ing. The data record all births reported for a given month, no matter when the report was made. For example, if an AFDC mother gave birth in October 1993 (two months after the cap went into effect) but did not report the birth until six months later, the birth was still recorded as occurring in October 1993. This makes it very unlikely that the large decline in births shown in Charts I and 2 is merely a result of delayed reporting.
To obtain permission from the U.S. Department of Health and Human Services to institute the family cap, New Jersey agreed to evaluate the effects of the policy with a controlled scientific ex- periment using random assignment. A small sample of AFDC recipients were assigned randomly to two different groups: an "experimental" group subject to the family cap benefit limitation and a 64control" group that was informed it would be exempt from the limitation. This procedure was in- tended to permit a scientific evaluation of the behavioral effects of the family cap by comparing the two groups. I Mothers in the sample control group received benefits according to conventional welfare policy: a net increase of $44 in monthly welfare benefits for each additional childbirth. But mothers in the sample experimental group were subject to the new family cap; they did not receive any increase in AFDC benefits when they gave birth to additional children. An initial analysis of the control and experimental group data was performed for the state by Dr. June O'Neill, now the Director of the Congressional Budget Office. O'Neill's analysis found that, despite its extremely modest impact on the dollar value of welfare benefits, the family cap had a sub- stantial effect in reducing out-of-wedlock births among AFDC recipients. During the first 10 months after the cap went into effect (from August 1993 to June 1994), births among AFDC moth- ers subject to the family cap were significantly lower than births to AFDC mothers in the control group exempt from the cap; 5.46 percent of AFDC single mothers in the experimental group bore children out of wedlock, compared with 6.75 percent of mothers in the control group exempt from the cap. Thus, welfare mothers under the cap had nearly one-fifth fewer illegitimate births than wel- fare mothers in the exempt group. However, a subsequent analysis of the sample experimental and control f roups by scholars at Rut- gers University appears to contradict the conclusions of the O'Neill study. Although the Rutgers scholars have not completed their analysis and have not issued a report, a brief memo has been leaked to the press. It states that the Rutgers group, at least initially, has found no difference in the birth rates of the experimental and control groups. Several factors may help explain the apparent difference between the two studies: Rutgers had later data which were not available at the time O'Neill performed her analysis, and the two analyses may have examined somewhat different subgroups of welfare recipients, based on time spent on AFDC and other factors.3 A clear explanation of the differences will be possible only when Rutgers
I A control and experimental group evaluation is a very useful scientific tool, but it is valid only to the extent that the control group is fully insulated from all possible effects of the policy. 2 The Rutgers study is headed by Dr. Michael Camasso and Dr. Carol Harvey, both of Rutgers State University. 3 The O'Neill study was limited to members of the experimental and control groups who were enrolled in the AFDC caseload in October 1992 and remained on welfare through June 1994.Me Rutgers study appears to contain this group as well as a group of more recent applicants who applied for AFDC after October 1992. The O'Neill analysis thus appears to have a greater focus on long-term welfare recipients. Ile O'Neill study also was restricted to single mothers, while the Rutgers study appears to contain some married couples.
releases a full report. Yet, contrary to press reports, the Rutgers analysis did not show that the fam- ily cap had no effect on illegitimate births. While it apparently failed to show a consistent difference in births between the experimental and control groups, it did show a drop in births within both groups after the cap went into effect. Thus, the Rutgers analysis shows a pattern of decline in births similar to that presented in Charts I and 2. It is important to note that both studies rely on the same government data base used in Charts I and 2. However, while Charts I and 2 cover all births within the entire New Jersey AFDC caseload, the Rutgers and O'Neill studies analyze only a small sample of the caseload; most AFDC mothers in New Jersey were subject to the family cap rule but were not included in either the experimental or the control sample groups for purposes of analysis. Thus, it should come as no surprise that the Rutgers analysis shows a chronological pattern of decline in births similar to Charts I and 2. Indeed, if the Rutgers analysis failed to show the same pattern of decline in births that is evident for the en- tire New Jersey caseload, this would indicate merely a lack of representativeness in the small sam- ple of cases selected for the study. The intriguing question, however, is why the Rutgers study appears to show birth declines both for those who were subject to the cap and for those who were exempt from the cap. The most plausi- ble answer is that establishment of the family cap sent a broad moral message that influenced the general AFDC caseload, including women in both groups. 4 By enacting the family cap, the New Jer- sey legislature sent a signal that society did not want welfare mothers to have additional children they could not support, and it backed up that moral judgment with a modest but concrete change in welfare benefits. The family cap and its moral message were publicized throughout New Jersey. Welfare mothers in both the experimental and control groups were bombarded with the clear moral message of the family cap, and it appears that welfare mothers in both the experimental and control groups re- sponded to that message a even though those in the control group were tech- Abortions to Women on AFDC in New Jersey: nicalIy immune from I the specific policy. It is Before and After the Family Cap also likely that welfare MOD mothers in the control Pba- Faadk Cap Avemp (9193-3194) 1261 group, interpreting the Pro-Famik Cap AwW 992 985 family cap as a harbin- 1200- ger of a fundamental 1111 1110 1101 0 1050 and acceptance of wel- 1022 1023 1 979 shift in public support 1000 - N /_\\ [111Z - fare, may have altered \\4 V 3 %0 944 "2 943 V 942 924 their behavior in antici- ;!V V069 883 pation of future retrench- Sao - 814 ments in welfare F-ly Cap E Aqua M leniency even while they remained exempt from the immediate ef- e fects of the family cap Sow= See Chan I sowce. themselves.
It is also possible that members of the control group may not have understood fully that they were exempt from the family cap policy.
The Family Cap and Abordon There has been concern that a family cap in national welfare reform legislation would reduce out- of-wedIock births by increasing abortions. However, the data currently available from New Jersey indicate that while the establishment of the family cap was followed by a clear and significant decrease in the Abortion Rate for Women on AFDC in New Jersey: number of births to welfare Before and After the Family Cap mothers, it did not result in 12''em per I= Women any significant increase in ft-caffWy Cap A-V.* Fbsr, Fan* Cap A--V the rate of abortions among 8.10 (9193-3194) these women. 8.08 10.26 10- Chart 3 shows the number 9.W 9-01 8.97 of abortions performed on &56 839 0.56 7A MA -A welfare mothers per month 8 ;,,., V V &I" Nlo@M in New Jersey. It covers all 7.64 V7.14 7.19 7.53 the months for which data V3 are currently available- 6 - from August 1992 to March Fanl@/ C.V Swted 1994. As the chart shows, Aqm M the average number of abor- 4P tions per month among AFDC mothers in the year Sou= See Oan I sotrce. before establishment of the family cap was 984.75. In the eight-month period after the cap took effect, the average number of abortions per month was 991.5. The difference in the number of abortions before and after the cap is tiny-only 0.7 percent of the monthly total-and well within the expected range of random statistical fluctuation. Chart 4 presents the same data expressed as an abortion rate equalling the number of abortions per 1,000 women on AFDC. As the chart shows, the monthly abortion rate per 1,000 AFDC moth- ers was 8.1 in the twelve months before enactment of the cap, but fell to 8.08 per month after the cap went into effect. The difference between the pre- and post-cap abortion rate is extremely small and not statistically significant. Overall, the available data indicate that the family cap did not cause an increase in either the abortion rate or the number of abortions.
The Family Cap and Welfare Reform The New Jersey family cap was based on the moral principle that the welfare system should re- ward responsible rather than irresponsible behavior. Proponents maintained that it is both iffesponsi- ble and immoral for unmarried women already on the public dole to have additional children and to expect the taxpayers to provide increased welfare to support those children. With the family cap, New Jersey proposed to stop rewarding such irresponsible behavior. Few expected the modest limit on benefits to result in a significant drop in births to welfare mothers. The fact that New Jerse''s limited experiment has caused a drop in illegitimate births, and hence in welfare dependency, en- hances the case for this policy. It is clear the United States must begin to address the crisis of illegitimacy. Today, nearly a third of all American children are born out of wedlock. According to Senator Daniel Patrick Moynihan Ch (D-NY) the illegitimate birth rate will hit 50 percent by 2003.
The rise of illegitimacy and the collapse of marriage have a devastating effect on children and society and seriously exacerbate most other social problems. 5Even President Clinton has declared that the collapse of marriage is a major factor driving up Americ''s crime rate and that the existing welfare system has played a significant role in that collapse. Halting the rapid rise of illegitimacy must be the paramount goal of welfare reform. Since the New Jersey data strongly indicate that the moral message conveyed by welfare policy may be its most important feature, it is essential that any welfare reform legislation enacted by Congress send out a loud and unequivocal message that society does not condone the growth in out-of-wedlock childbearing and that the taxpayers will not continue the same open-ended subsidies for illegitimacy that have characterized welfare in the past. This can be done by: * Instituting a national family cap policy prohibiting the use of federal funds to give higher wel- fare benefits to women who have more children while already receiving welfare; * Prohibiting the use of federal funds to give cash welfare benefits to unmarried mothers under age 18; 4) Requiring mothers of children born after January 1996 to establish paternity of the child be- fore receiving federal cash welfare; * Encouraging marriage by providing a refundable tax credit to low-income married couples with children where at least one spouse is working; and * Increasing funding for abstinence education. 6 By contrast, a "refor''''of the federal welfare system which fails to send a clear moral signal, which remains agnostic on the question of illegitimacy, and which simply allows states to "do their own thing" with federal taxpayer'' funds-or to do nothing-would be a disaster. Wbile the enactment of a nationwide family cap is vitally important, it is equally important to rec- ognize that it is far from a panacea for illegitimacy. The family cap did lower out-of-wedlock births in New Jersey, but even with the cap in place, roughly one in ten of the stat''s unmarried welfare mothers continue to bear children each year. The campaign to restore marriage and reduce illegitimacy in the U.S. will be long and difficult. Establishing a family cap policy within the federal welfare system will be only a first step in that process. But in issues of great moral significance to the nation, it is often the first step which is both the most difficult and the most significant.
5 For an overview of the social problems associated with rising illegitimacy, see Patrick F. Fagan, "Rising Illegitimacy: Americ''s Social Catastrophe," Heritage Foundation F. YL No. 19, June 29, 1994; see also Patrick F. Fagan, "rbe Real Root Causes of Violent Crime: The Breakdown of Marriage, Family, and Community," Heritage Foundation Backgrounder No. 1026, March 17, 1995. 6 See Robert Rector,"The Forgotten Crisis: S. 1120, Welfare Reform, and Illegitimacy," Heritage Foundation Committee Brief No. 18, August 31, 1995. For a sound overview of government programs to combat teenage pregnancy, see Joseph J. Piccione and Robert A. Scholle, "Combatting Illegitimacy and Counseling Teen Abstinence: A Key Component of Welfare Reform," Heritage Foundation Backgrounder No. 105 1, August 3 1, 1995.