The impact of state abortion restrictions on minors' demand for abortions.

I. Introduction

In 1973 the United States Supreme Court ruled in Roe v. Wade (410 U.S. 113) that a right of personal privacy exists under the Constitution and this right includes a woman's decision whether or not to terminate her pregnancy. The court also ruled that the right of personal privacy is not unqualified and must be considered against state interests in regulation. The court's decision did not end the public policy debate concerning abortion. At both the federal and state levels, this debate centers on whether taxpayer dollars should be used to pay for abortion services and the extent to which states can regulate women's abortion decisions.(1)

Since 1973, the Supreme Court has permitted enforcement of certain state restrictions on abortion and prevented states from enforcing other restrictions. The enforceable abortion restrictions include parental involvement restrictions and restrictions on public funding for abortion. Parental involvement restrictions or state requirements of parental consent or notification for unmarried minors (women under the age of 18) are enforceable if the state provides a judicial bypass mechanism.(2) By October 1995, 27 states had enacted and begun to enforce parental consent or notification laws for minors.(3)

With respect to the public funding issue, in 1976 the U.S. Congress passed the Hyde Amendment prohibiting the expenditure of federal funds for abortion services except in cases where continuation of the pregnancy threatens the pregnant woman's life. The Hyde Amendment restricts federal funding of abortion services through the Medicaid program; however, it does not prohibit states from paying for abortions. As of September 28, 1995, abortions to save the life of the pregnant woman were the only abortions funded in four states.(4) In 16 states and the District of Columbia Medicaid paid for all medically necessary abortions.(5)

Very little is known about the impact of these enacted, and in some cases enforced, abortion restrictions. To address this issue two models, a weighted least squares model and a fixed-effects model, are estimated using annual, state data on abortions obtained by minors from 1978 to 1990. The paper focuses on minors because the parental involvement laws have a direct impact only on women under 18 years of age. Further, the problem of unintended pregnancies is particularly evident for this age group. By age 20, approximately 40 percent of all teenage women have been pregnant and of these pregnancies, approximately 84 percent were unintended (U.S. Department of Health and Human Services 1990). Abortions obtained by minors represented 12 percent of total legal abortions in 1987 (Henshaw, Koonin, and Smith 1991).

II. The Market for Abortion Services and the Literature on Abortion Restrictions

The U.S. abortion rate or the number of abortions per 1,000 women aged 15-44 has remained relatively constant since the late 1970s.(6) In 1978, 1988, and 1992 the abortion rates were 27.7, 27.3, and 25.9 respectively (Henshaw and Van Vort 1994). However, there is tremendous variation in abortion rates by state of occurrence. In 1992 abortion rates ranged from 46.2 abortions per 1,000 women aged 15-44 in New York to 4.3 abortions per 1,000 women aged 15-44 in Wyoming (Henshaw and Van Vort 1994). Further, state legislative opposition to liberalized abortion policies has varied widely across states. Between 1973 and 1989 state legislators in Connecticut and Alaska did not enact any abortion restrictions, while state legislators in North Dakota enacted 15 abortion restrictions (Halva-Neubauer 1990).

These statistics raise many interesting questions. Are the states with the lowest abortion utilization rates also the states that enforce abortion restrictions? Does state-level abortion legislation simply confirm statewide sentiment toward abortion and thus would abortion utilization rates be lower in the most restrictive states even in the absence of the restrictions? Or does statewide sentiment toward abortion determine the extent of abortion restrictions,(7) and then both sentiment and restrictions affect abortion utilization rates?

Previous research has focused on addressing the first question, and the results are mixed.(8) The results of Singh (1986), Garbacz (1990), Lundberg and Plotnick (1990), and Haas-Wilson (1993) suggest that Medicaid funding restrictions have a negative impact on minors' demand for abortion. However, Blank, George, and London (1994) found an insignificant effect of Medicaid funding restrictions on teen abortion rates.

With respect to parental involvement restrictions, the results of Donovan (1983), Cartoof and Klerman (1986), Haas-Wilson (1993), Ohsfeldt and Gohmann (1994), and Blank, George, and London (1994) suggest that these restrictions have a negative impact on minors' abortion rates. However, the results of Lundberg and Plotnick (1990) suggest that parental involvement restrictions have a statistically insignificant impact on white teenagers' abortion rates.

Three limitations of this previous research on the impact of abortion restrictions on minors' abortion utilization should be noted. First, state abortion policies are continuously changing, and previous empirical analyses of the impact of abortion restrictions cover only a short period of time, in most cases one year.(9) Between 1978 and 1990 11 states enacted and began to enforce parental involvement laws, and 13 states began or stopped public funding of medically necessary abortions for low-income women. Thus, previous research does not take advantage of this "natural experiment." The research reported in this paper is based on 11 years of state data and thus, the change in abortion demand in states which enacted and began to enforce abortion restrictions can be compared to the change in abortion demand in states which did not.

Second, many states have enacted parental involvement restrictions, but never enforced those restrictions or enforced them for only a short period of time. Previous research does not distinguish between enforced and unenforced restrictions. With the help of the legal department of the National Abortion and Reproductive Rights Action League (NARAL), the exact dates of enactment and enforcement of parental involvement restrictions were collected, and thus this paper includes separate estimates of the impact of enforced and unenforced restrictions.

Third, the published research does not take into account unmeasurable taste factors, such as anti- or pro-abortion sentiment, and thus suffers from omitted-variable bias. Previous research attributes differences in abortion utilization to the presence or absence of abortion restrictions; however, it is possible that the state residents' ideological preferences on abortion determine both the restrictiveness of a state's abortion policies and the abortion utilization rate of the state's residents. In other words, there are fewer abortions in states with abortion restrictions; however, this may be because states that enact restrictions are …