Insurance Coverage and Abortion Incidence: Information and Misinformation
With a grim inevitability, the issue of restrictions on insurance coverage for abortion moved to center stage as the congressional health care reform debate came to a head in the fall of 2009. What was less predictable, perhaps, was that abortion opponents would misuse Guttmacher Institute research to bolster their assertion that anything short of a flat ban on coverage of the procedure would somehow greatly increase the number of abortions taking place in the United States. According to Richard Land of the Southern Baptist Convention, enactment of the Senate version of the measure—notwithstanding its stringent abortion coverage restrictions— would “lead, as some experts project, to a 30 percent increase in abortions in America.This legislation, if passed, will be the largest expansion of abortion since the Roe v. Wade decision in 1973.”
The Senate bill was, in fact, enacted into law in March 2010 with its abortion restrictions intact (see article, page 2). Since then, however, a vastly overstated link between insurance coverage and abortion incidence—and the misuse of Guttmacher data to support it—has repeatedly been alleged at the state level, including as a prominent feature of model legislation drafted for state legislators by the antiabortion advocacy organization Americans United for Life. The Federal Abortion-Mandate Opt-Out Act, which would block coverage of abortion in the health insurance exchanges to be set up by states, asserts that the Guttmacher Institute "confirms that, based on Medicaid studies, more women have abortions when it is covered by private or public insurance programs."
The referenced Guttmacher "Medicaid studies" do indeed conclude that denial of abortion insurance coverage in the form of Medicaid funding impedes a sizable minority of America's poorest women from obtaining the procedure—and that restoration of coverage would result in an increase in abortion incidence among this population. However, the claim that restoration of federal Medicaid coverage would result in a significant increase in the incidence of abortion nationwide is not supported by the research, and extrapolating from Guttmacher's Medicaid findings to assert that coverage in the private insurance market is strongly linked to abortion incidence is entirely illegitimate.
This is by no means to say that the question of abortion insurance coverage is not important. Even if coverage may not determine whether most women actually obtain a procedure, it may have a major impact on the circumstances under which they do so and on the perception of abortion as a legitimate health care service. And on these matters, partisans on opposite sides of the abortion debate have sharply different views of what the situation ought to be.
Publicly Funded Abortions
Restrictions in place for more than three decades—measures often collectively referred to as the Hyde amendment for their original sponsor, former Rep. Henry Hyde (R-IL)—have sharply limited the use of federal Medicaid funds for abortion services for low-income women, currently to cases of life endangerment, rape and incest. Under the Hyde amendment, states may use their own funds to pay for abortions for their Medicaid enrollees, and 17 states and the District of Columbia do so.
Claims of a strong link between abortion coverage and abortion incidence—both among the population of Medicaid enrollees and among the population at large—purport to be based on studies in five states (Georgia, Illinois, North Carolina, Ohio and Texas) in which neither federal nor state funds for abortion were available. These studies generally looked at what happened when Medicaid funding restrictions were first implemented some three decades ago and found that approximately one in four women who would have had a Medicaid-funded procedure if funds had been available were unable to do so.
Guttmacher Institute researchers—looking not at the past but contemporaneously at differences between states that are now using state dollars to fund abortions for low-income women and states where funding is restricted—nonetheless found a strikingly similar result. In an unpublished analysis presented at a scientific conference of demographers and statisticians in 2007, Guttmacher researchers compared the abortion rate (number of abortions per 1,000 women 15–44) among female Medicaid enrollees in funding states with that among Medicaid enrollees in nonfunding states. From the differences between the abortion rates for Medicaid enrollees in these two groups of states, the researchers concluded that restoring funding would result in a 28% increase among Medicaid enrollees in states where funding is currently restricted.
By maintaining the Hyde amendment, health care reform represents the largest expansion of abortion funding restrictions since Hyde was first implemented.
But making the leap from a finding that restrictions on public funding make abortion unattainable for about one in four women poor enough to be on Medicaid to the assertion that repealing the Hyde amendment would significantly increase the total abortion rate in those states, let alone in the United States as a whole, is entirely unsupportable. This is because only a small proportion of women are poor enough to be enrolled in Medicaid and therefore affected by the restrictions. In fact, according to the Guttmacher analysis, lifting the funding restrictions would translate into only a 5% rise in the total number of abortions in the group of states in which funding is currently restricted. (Nationwide, only 15% of women of reproductive age are covered by Medicaid; see article, page 17.)
And because several of the nation's most populous states, such as California and New York, are among those that use their own money to pay for abortion services for poor women, the national impact of repealing the Hyde amendment would be even smaller: According to the comparison of abortion rates among Medicaid enrollees in funding and in nonfunding states, the number of abortions among Medicaid-eligible women nationwide would be expected to rise by approximately 33,000 if the Hyde amendment were to be repealed—only a 2.5% increase in the total number of abortions performed nationwide.
In sum, two sets of research yield evidence that could be used to get a sense of the potential impact of repealing the Hyde amendment. The oldest studies, conducted at a time when Medicaid eligibility was much more restrictive than it is today (averaging 45% of poverty in those states, compared with a national average of 85% today), found that about one in four women who were denied funding for an abortion might be likely to have one if funding were restored. More recent work using an entirely different approach yielded a similar result. Contrary to the allegations of leading antiabortion activists, however, both lead to the inescapable conclusion that although the impact on Medicaid enrollees in states that have implemented the funding restrictions may be substantial, the impact of repealing the Hyde amendment on the overall level of abortion in the United States would be minimal.
But, of course, health care reform does not repeal the Hyde amendment. In fact, it essentially has the opposite impact. By maintaining the Hyde amendment, health care reform represents the largest expansion of abortion funding restrictions since Hyde was first implemented. This is both because the health care reform law includes a dramatic expansion of the overall Medicaid program to include all individuals with incomes under 133% of the federal poverty level ($24,352 for a family of three) and because the effect of that expansion will be felt disproportionately in states that do not subsidize abortion with their own funds.
According to a 2010 study conducted for the Kaiser Commission on Medicaid and the Uninsured by researchers from the Urban Institute, the Medicaid expansion provision will bring Medicaid coverage to an additional 15.9 million Americans by 2019. More than two-thirds of these new Medicaid enrollees will live in states where Medicaid funding for abortion is currently restricted. Moreover, the impact of the expansion will be more pronounced in states in which Medicaid coverage is currently less generous and in which more residents are uninsured. As a result, the group of states in which public funds are not available for low-income women needing an abortion will see a disproportionate impact from the Medicaid expansions in health care reform: The proportion of adults who are uninsured is expected to fall by 49% in the nonfunding states, compared with 41% in the states where public funds are available.
Private Insurance Coverage
Claims by antiabortion leaders about the impact of Medicaid coverage on the nationwide abortion incidence constitute a serious misuse of Guttmacher data, but any use of those data to make allegations about the impact of coverage in the private market is completely unfounded. The Guttmacher Institute has not studied the impact of private insurance coverage of abortion. In fact, any discussion of the impact of the federal health care reform law on levels of abortion among those with private coverage is speculative at this point.
What is known is that under the legislation, some 16 million individuals who would otherwise be uninsured are projected to have private coverage by 2019, according to the Congressional Budget Office. The legislation signed into law in March will make it extremely difficult for insurers to include abortion coverage in the plans they will be marketing on the health insurance exchanges through which these individuals will be purchasing insurance. Indeed, the statute's coverage restrictions are so stringent that leading insurance experts have suggested that most insurers will simply decline to sell policies covering abortion on the exchanges—and eventually in the broader private market as well. Yet, even if some newly insured women do receive coverage for abortion, there is little reason to think that it would open new doors for those women to obtain abortions that they cannot afford today. These women, by definition, will have incomes higher than those on Medicaid, as the insurance exchanges will be designed for Americans with incomes above the 133% of poverty cutoff for Medicaid. And even when income eligibility ceilings for Medicaid were much lower than they are today, and far lower than they will be in 2019, three in four Medicaid enrollees were still able to obtain an abortion in the absence of coverage.
Meanwhile, a study published in the March 2010 issue of the New England Journal of Medicine did look at changes in the incidence of abortion in a state that adopted a universal insurance coverage policy but without any of the kinds of abortion coverage restrictions included in the federal legislation. This analysis, by Patrick Whelan of Harvard Medical School and Massachusetts General Hospital, examined the impact of insurance coverage in Massachusetts, a state whose experiment in health care reform is often cited as the model for the federal legislation. Massachusetts enacted its own universal health care plan in 2006. Since the beginning of 2007, the state has provided subsidized coverage to individuals with an income up to 300% of the federal poverty line who are either self-employed or unemployed, as well as to small businesses. In stark contrast to the federal law, abortion is covered for individuals with subsidized coverage, known as Commonwealth Care, as well as for Medicaid enrollees.
Yet, since the enactment of health care reform, the number of abortions in the state fell from 24,245 in 2006 to 23,883 in 2008, a decline of 1.5%, even as the insured population grew by 5.9% over the same period. (The number of abortions to teens fell by 7.4% over the same time period.) According to Whelan, these decreases came during a period of rising birthrates and population growth, which meant that the abortion rate in the state declined from 3.8 per 1,000 Massachusetts residents in 2006 to 3.6 per 1,000 in 2008.
The number of abortions in Massachusetts has reached its lowest level since the 1970s, even though more residents than ever were covered by health insurance and virtually all insurance plans covered abortion. As a result, Whelan noted, "The recent experience in Massachusetts suggests that universal health care coverage has been associated with a decrease in the number of abortions performed, despite public and private funding of abortion that is substantially more liberal than the provisions of the federal legislation."
Underlying Causes and Real Solutions
History, common sense and available data all suggest that insurance coverage for abortion is not a significant driver of the incidence of abortion, any more than insurance coverage of pregnancy-related care drives the number of babies born each year. Even the legalization of abortion nationwide that came with Roe v. Wade in 1973 did not somehow "create" abortion. Indeed, although the U.S. abortion rate rose rapidly in the years immediately following the Supreme Court decision before leveling off in the early 1980s, this was to a considerable extent because of legal abortions replacing abortions that previously had been performed illegally and had gone uncounted.
This reality is entirely consistent with the experience in other countries with generous abortion coverage under their national health systems. In the Netherlands and Germany, for example, where almost all abortions are free to citizens, abortion rates are less than half that in the United States. And in Canada, which has no national restrictions on abortion as well as a comprehensive national health system, the abortion rate is considerably below ours.
A wealth of evidence from around the world confirms that underlying levels of unintended pregnancy are the best predictor of abortion rates. Countries with low rates of unintended pregnancy have low rates of abortion, and vice versa. But what insurance coverage can affect to a much more significant degree are the conditions under which the procedures take place. Research on poor women affected by the funding restrictions under Medicaid shows both the financial obstacles women living at or near the poverty line must surmount and the personal indignities they must endure to obtain an abortion in the absence of Medicaid coverage. Moreover, poor women having an abortion do so more than a week later than do more affluent women, likely reflecting their increased difficulty in securing funds.
Better-off women with private insurance may not have to make the same financial sacrifices that poor women do to obtain an abortion, but they face many of the same indignities. From bogus "informed consent" procedures and waiting periods to unnecessary and costly ultrasound mandates, women seeking an abortion and the professionals providing the service are subjected to a host of restrictions and requirements not imposed on any other legal medical procedure in the United States. In that light, the campaign to end insurance coverage is yet another component of a long-standing and concerted effort by abortion rights opponents to paint abortion and ensure its societal treatment as something other than a legitimate medical procedure. And that is why, however limited the relationship between abortion coverage and actual abortion incidence may be, the eventual success or failure of that campaign will be consequential.