Introduction

In 1966, the Los Angeles Times published an editorial typical of how reform advocates talked about abortion policy prior to Roe v. Wade (1973). The editorial was atypical only in its length—12 paragraphs instead of the usual three or four—and touched on every prominent argument that reform advocates were advancing in the 1960s. It discussed the risks of illegal abortions and how prohibition increased those risks. It described the suffering of women in need of medical care, and the cruelty of prohibiting professional physicians from providing it. It referenced victims of rape and incest, congenital disorders, pregnant teenagers, and unwanted children. And it discussed the nearly 10,000 American women dying each year from illegal, unsafe abortions. The editorial argued that legalizing abortion would be humane, that it would not supplant birth control or promote promiscuity, and that it would save lives (Los Angeles Times, August 1, 1966).

What the editorial did not mention was privacy. The word did appear 6 years later, however, in the newspaper's opinion page, on the day after the Supreme Court's decision in Roe, when the editorial board wrote,

More important is the right of privacy, “surely.” So surely that in the 11 editorials the Los Angeles Times published in the decade prior to Roe, not once did the word appear in reference to abortion. In 1973, the editorial board of the Los Angeles Times declared a right to privacy to “surely” be the basis of an abortion for the first time. It would not be the last.

A second transformation in how Americans talked about abortion paralleled the rise of a politics of rights. In 1970, the New York Times published an editorial in support of an abortion reform law being debated in the New York state legislature. The three-paragraph editorial read, “reform … would place responsibility where is should be in a modern and free society—in the decision of a woman and her physician” (New York Times, April 9, 1970). Another editorial, published later that month, read, “the decision to have an abortion is one properly to be left with the individual woman and her physician” (New York Times, April 29, 1970). Three years later, in their editorial endorsing the U.S. Supreme Court's 7-to-2 decision in Roe v. Wade, the Los Angeles Times wrote, “[t]he decision will not satisfy those who had argued that the mother should make the decision. It insists that the decision be made with the personal physician” (Los Angeles Times, January 23, 1973).

In the years immediately following that Supreme Court ruling, the way advocates discussed women's abortion decisions changed in a subtle but profound way. By the end of the decade, editors would write of “a woman's right to freedom of choice,” that “a woman's right to personal privacy includes her decision whether to end pregnancy,” and that “the state has no business intruding into the individual woman's abortion decision” (Los Angeles Times, July, 1, 1980). In the years following Roe, the abortion decision, previously talked about as medical decision made between a woman and her doctor, was transformed into a purely personal decision, devoid of any obvious medical significance.

By the end of the 1970s, the way reform advocates talked about abortion laws had changed. Before Roe, mainstream liberal opposition to prohibition was diverse and loosely centered around a framework of health and safety. That kind of advocacy had won significant victories in a number of states, but had yet to deliver national change. Roe transformed the nature of public advocacy, shifting the center of the debate to a once peripheral conception of abortion as a constitutional right.

Before the Supreme Court released its decision upholding a personal right to an abortion, opinion elites in government and the news media talked about a medical procedure performed by a doctor on a patient in need. By the end of the 1970s, Americans were engaged in a seemingly intractable debate that pitted a personal right to choose against a fetus' right to life. As talk about abortion in the United States became increasingly consumed by a language of rights, something else receded from prominence. What receded were the experiences of women in need of medical care and of physicians legally barred from providing it.

In this article I argue, through a critical reading and a quantitative content analysis, that by upholding a constitutional right to an abortion in Roe, the Supreme Court played a unique and critical role in bringing about a transformation in the social conception of abortion. I argue that this role reveals, contrary to Reference RosenbergRosenberg's (2008) view, a significant judicial influence over society. I develop a theoretical account of this phenomenon and present and contextualize texts spanning 50 years of public debate about abortion. Through the qualitative analysis of these texts, I develop an interpretive account of how the social conception of abortion changed after Roe, from a medical to a moral decision and from a politics of practice to a politics of rights. I then present a quantitative analysis confirming the nature, timing, and magnitude of these transformations. A census of opinion page coverage in four major American newspapers offers substantive evidence of change in the language of abortion politics consistent with the account of the Court's influence over the way we talk and think about abortion politics.

Supreme Court Skepticism and Abortion Politics

In his discussion of the limits on the Court's ability to influence society, Reference RosenbergRosenberg (2008: 228–46) focused on changes in material access to abortion in the short to medium term and rejected the possibility of other kinds of judicial influence. Subject to less empirical scholarship have been changes in the social conception of abortion. Despite good theoretical analysis that Americans no longer talk and think about abortion as they did in the pre-Roe era (Reference GlendonGlendon 1993), empirically oriented scholars have only begun to explore these complex social changes. I argue that the Court's opinion in Roe presaged a profound shift in how opinion elites discussed abortion politics. My analysis runs against the skeptical tradition in judicial politics, typified by Reference RosenbergRosenberg (2008), by pointing to rhetorical influence as a location of judicial influence over society.

Reference RosenbergRosenberg (2008: 228–41) argued that the Court did not significantly influence abortion policy and did not catalyze or otherwise impel any apparent social change. Rosenberg identified specific criteria for what he called “extra-judicial influence,” but found no evidence of such a phenomenon in abortion politics. I argue that Rosenberg's analysis failed to recognize the significant role the Court played in influencing the social conception of abortion, an influence observable in the language and criteria of the public debate that followed Roe. Although this is not the sort of “social change” Rosenberg directly implicated, it is a vital, important part of our political culture, one in which the Supreme Court played a unique and significant role, and it demonstrates that contrary to Rosenberg's general view, the Supreme Court can change society in subtle but significant ways. My research strongly supports Reference RosenbergRosenberg's (2008: 155–56) deeper point, however, that the Court's decisions can have significant negative consequences for democratic politics.

We should expect judicial influence to be limited and contingent. Social movements drive both Court decisions and culture alike. The justices do not invent arguments or frameworks out of a void. The justices must have a candidate constitutional justification that is plausible enough that the justices can offer it with a straight face (Reference SolumSolum 1987; Reference TrubekTrubek 1984; Reference TushnetTushnet 1991). Many justices have grasped innovative ideas advanced by social movements and pulled them to the center of the public sphere (Reference KramerKramer 2001). Abortion politics offered numerous possibilities for both sides. On the eve of Roe, the social conception of abortion was ambiguous and contested. To the Court's advantage, both sides of the liberalization debate were both undergoing periods of change and growth, facilitating a variety of ways of thinking about abortion (Reference GarrowGarrow 1994; Reference GlendonGlendon 1993; Reference JoffeJoffe 1995; Reference LukerLuker 1984; Reference SolingerSolinger 2001).

Because courts are influenced by political pressures, much of their influence on politics is felt indirectly (Reference McFarlandMcFarland 2005; Reference Spiller and GelySpiller & Gely 1992). Nonetheless, there is excellent research demonstrating the conditional role the Court plays in ordinary democratic politics. Reference Baird and HurwitzBaird and Hurwitz (2002) showed that both the Senate and the President respond to Court preferences. The Court has tended to shift alongside public opinion (Reference LinkLink 1995; Reference Mishler and SheehanMishler & Sheehan 1993, Reference Mishler and Sheehan1994), to defer to local political majorities (Reference Brace and HallBrace and Hall 1997), and to balance policy goals against the strength of public opposition (Reference Flemming and WoodFlemming & Wood 1997; Reference McGuire and StimsonMcGuire & Stimson 2004).

There is good reason to believe that the justices possess a limited and measurable capacity to influence public opinion. Evidence generally favors limited influence in salient policy domains. Reference Johnson and MartinJohnson and Martin (1998) found measurable impacts on public opinion following rulings in salient, controversial issues. Others have studied the conditions of influence (Reference Franklin and KosakiFranklin & Kosaki 1989; Reference HoekstraHoekstra 2003; Reference Hoekstra and SegalHoekstra & Segal 1996; Stoutenborough, Haider-Markel, & Reference Stoutenborough, Haider-Markel and AllenAllen 2006). In a novel challenge to Reference Flemming, Bohte and WoodRosenberg, Flemming, Bohte, and Wood (1997) argued the Court has the power to increase national attention on a policy issue in ways that can be long lasting and disturb established political coalitions. Given evidence from research in public opinion emphasizing that expressions about political preferences by ordinary citizens are highly dependent on the content of elite discourse (Reference ZallerZaller 1992), we should expect judicial influence on political thought to be mediated by opinion elites' adoption of judicial arguments and frames.

I work from this framework to develop, through a qualitative analysis, an account of judicial influence over the conceptual foundations of public discourse. The purpose of the qualitative analysis that follows is to interpret Roe's effect on how Americans think and talk about abortion. The quantitative hypothesis tests that follow the qualitative analysis are intended to show that the interpretation is more than a collection of anecdotes, and that it entails real and measurable consequences.

Abortion Politics Prior to Roe

The making of abortion policy cannot be separated from the social conception of abortion. Abortion policy was traditionally the purview of state legislatures, but the reality of abortion access for American women before Roe was determined by socioeconomic status, community mores, and the practices of local law enforcement officials. Access to legal abortion was widely available in the United States prior to the Civil War, but disappeared in the late-nineteenth century before a variety of cultural and legal trends culminated in the 1960s in an organized and diverse social movement to end prohibition. In the years immediately before Roe, a pattern of legislative successes seemed to stall in the face of increasingly well-organized support for prohibition. Then, just as the dream of guaranteed federal protection of legal abortion access began to dim, abortion was proclaimed by the Supreme Court in 1973 to be a constitutional right (Reference DavisDavis 1985; Reference GarrowGarrow 1994; Reference MohrMohr 1978; Reference RosenbergRosenberg 2008). During the period of its widespread criminality, abortion receded not only from open practice but from open discussion as well. As late as 1869, the New York Times published no fewer than 69 different stories concerning abortion (Reference LukerLuker 1984: 268), yet by the turn of the century discussion of abortion had disappeared from newspapers prominent in polite society.

The history of modern abortion policy begins in earnest in broader social changes, emerging from the rise of mainstream feminism and the liberalization of attitudes toward sexuality among the cohort who came of sexual maturity during the 1960s. Public attitudes toward abortion were both coming into being and changing during this time. Public opinion research during this period is of generally poor quality and, as Zaller wrote regarding the surveying of opinions generally, “what gets measured as public opinion is always and unavoidably dependent on the way questions have been framed and ordered” (Reference ZallerZaller 1992: 95). These concerns must be understood in the historical context in which polling of abortion attitudes was conducted. Reference RosenbergRosenberg (2008: 258–65), relying on contemporary accounts, argued that public opinion was moving toward liberalization, citing Gallup polls, the National Fertility Study, the American Council on Education, and the Commission on Population Growth and the American Future (Reference RosenbergRosenberg 2008: 260–62). Interest group polls must be taken with more than a grain of salt. The question formulation employed by the Commission on Population Growth, for example, asked whether respondents agreed with the statement that an abortion decision “should be left up to persons involved and their doctor.”

It is difficult from the poll, which was reported in the New York Times (October 28, 1971), and absent equivalent prior polling, to know what to make of the 50 percent who agreed. The statement is clearly biased; one can be politically opposed to abortion access and still find disagreement difficult, as President George H.W. Bush discovered in an interview with the Times in 1992 when asked to comment on how he would react to one of his granddaughters pursuing an abortion, not only said that it should be her choice but seemed genuinely unable to imagine the alternative. Asked if the choice should ultimately be hers, the antiabortion president responded, “Well, whose else's, who else's could it be?” (New York Times, August 12, 1992).

When Roe was handed down, there was no obvious ideological connection between either of the major parties and abortion policy. The Democratic Party was generally perceived as (marginally) more welcoming to feminists. Although the Republican Party platform had endorsed the Equal Rights Amendment as early as 1940, the Democratic Party had opposed it until 1972, largely out of concern, originally voiced by Eleanor Roosevelt and leaders in organized labor, that the Amendment would invalidate government regulations protecting women from the ravages of the market (Reference MansbridgeMansbridge 1986). Although contemporary abortion attitudes are frequently characterized by the polarization and absolutism of the public debate, actual attitudes in the contemporary electorate reflect a significant, ordinary level of ambivalence and uncertainty (Reference Craig, Kane and MartinezCraig, Kane, & Martinez 2002). No doubt, public opinion toward abortion was evolving during the 1960s, but there is no way to disentangle these results from the rapidly changing partisan and ideological taking of shape going on in the media coverage of the abortion debate, not only during the 1960s but in the two decades following Roe, during which the parties adjusted and assumed distinct positions, a far cry from the ambivalence of Presidents Nixon, Ford, and Carter (Reference AdamsAdams 1997).

Particularly vexing for making sense of public opinion about abortion during the pre-Roe years is the impossibility of knowing what the survey prompts actually meant to respondents. These meanings seem elusive because of the changing conception of abortion itself. The problem is further compounded by social desirability bias, which is common in social and health issues (Reference BerinskyBerinsky 2004; Sudman, Bradburn, & Reference Sudman, Bradburn and SchwarzSchwarz 1995). These methodological concerns do not throw into doubt reasonable inferences that, whatever prior absolute levels of support for various abortion policies, attitudes appeared to begin changing in the 1960s, prior to Supreme Court action. It is impossible to measure the extent to which these changes were caused by elite action, interest group organization, external events, changing methods of measuring public opinion, or diffuse cultural change. Regardless of the interactions of these important causes of aggregate public opinion change, they influenced not only attitude change but policy change as well.

The New York Times did not publish an editorial about abortion politics in the twentieth century until 1965. The Washington Post published its first on November 10, 1951. That editorial concluded that abortion is inherently more dangerous than childbirth and thus should be illegal to protect the public health. The editorial got facts wrong, did not consider whether the purported danger of abortion procedures could be a consequence of their illegality, acknowledged the need for access to abortion is ubiquitous, and claimed compassion for women as its motive. It is an artifact of the assumptions, language, and culture of a pre-Roe conception of abortion.

In the more than two decades between the publication of that editorial and the Supreme Court's ruling in Roe, the New York Times, Washington Post, Chicago Tribune, and Los Angeles Times published 74 editorials about abortion policy. By 1973, all four newspapers supported liberal abortion reforms. The Chicago Tribune went from opposing liberalization to protect women to supporting decriminalization in the interest of public safety. Even the conservative Los Angeles Times and Wall Street Journal supported liberalization on the eve of Roe. Most mainstream national opposition to abortion access emerged later, as opinion elites traversed vast ideological distances.

The liberalization of abortion policy in the United States began in the mid-1960s. The California legislature first debated decriminalization in 1961 and the New York legislature followed in 1966. They were followed by a wave of liberalization efforts in 1967, when 28 states introduced reform bills and Colorado passed the nation's first. Later that year Governor Reagan signed California's liberalization bill and North Carolina liberalized its abortion laws. Over the next few years, several other states enacted bills ordering various levels of reform, while national political elites remained largely indifferent or disinclined to engage in the emerging issue (Reference RosenbergRosenberg 2008: 183–89).

During this period public health professionals, who repeatedly pointed to illegal abortions as threats to women's health, exercised considerable influence in the public debate. Consider this example from the Los Angeles Times, published July 28, 1967, titled “Abortion Laws,” reproduced below:

This editorial has four distinct claims and a fifth, less distinct, potential claim. Typical of editorials of the pre-Roe era, it features no constitutional claims; all its claims are situated in the language of ordinary politics. The first of the four appears in the second paragraph, where the authors cite the trend across various states toward liberalization. This argument proceeds from a claim about public opinion and is repeated more directly in the last paragraph. The second claim appears in the third paragraph. Here the authors note that medical care has advanced considerably, suggesting, albeit implicitly, that laws should be not merely changed but modernized. The third claim is a reference to a German measles outbreak of recent years, but is not fully developed until the fifth paragraph, in which the authors reference a “defective” child, promoting abortion as a method for preventing congenital disorders.

The last claim is for compassion. Significant here is the use of imagery of a woman unable to access a legal abortion, invoking “death or grave impairment.” The lack of any reference to the general health of women, the vivid language, and the presence of a singular “woman” is more an appeal for compassion for another human being than just an argument about health and safety. There is a potential fifth claim in this editorial, hinted at in three places: the seemingly approving mention of the requirement for unanimity of a three-physician panel to carry out the procedure, that use of the procedure does not constitute “birth control,” and the rule-based limitations on the use of the procedure. Taken together, these statements suggest that liberalization is a moderate proposal.

Constitutional claims were offered occasionally in the pre-Roe era, usually in discussions about possible legal strategies and in praise of the wisdom of judges. These cases exemplify the diversity of the discourse. The next text, an editorial unusually rich in its variety of argument, was printed on November 10, 1969, in the New York Times. Titled “Changing the Abortion Law,” it is reproduced in full below:

This is an unusually long editorial that gives a good sense of how some peripheral claims were used in the years prior to Roe. The reference in the first line to “New York's archaic abortion law” is an appeal for modernization, an argument for abortion reform as a path toward a widely shared goal, while the next paragraph makes a nonrights-based argument from privacy. Two constitutional claims are then offered, that the abortion restrictions in question violate due process and equal protection guarantees.

The penultimate paragraph includes a claim that fell out favor following Roe, but which was prominent in the early abortion debate of the 1950s and 1960s. Here the board argued the restrictions violated “a doctor's right to practice medicine according to the highest possible standards” and further impinge on “a patient's right to safe and adequate medical care.” This sort of argument, positing a right to health care, faded alongside medical privacy in the years after Roe. Although the argument from medical privacy was prominently transformed into the three major post-Roe claims for abortion rights—privacy, choice, and autonomy—the argument from a right to health care was peripheral.

One early example of organized advocacy for abortion rights is a quarter-page advertisement in the New York Times, placed by a coalition of social activists called the Women's National Abortion Action Coalition. The ad, which ran April 30, 1972, almost a year before the Supreme Court recognized a right to abortion, declared “the right to abortion is under attack in New York! … Abortion must be a woman's right to choose!” The ad invoked a language of choice and rights as well as a tone of immediate threat, all of which would come to be hallmarks of the abortion discourse during the 1980s and 1990s. Against the backdrop of reporting and editorializing of the era, the activists' ad stands out, ahead of its time. The language of abortion rights, particularly choice, was apparent before Roe in advocacy literature, but mostly ignored by mainstream establishment liberals.

The greater variety of arguments offered by advocates before Roe is broadly apparent in the print coverage of the era. An article from January 2, 1973 (New York Times), just a few weeks before the decision in Roe, is illustrative. The article, titled “Both Sides Gird for Renewal of Fight on Legalized Abortion,” examines the activists mobilized on both sides of the abortion debate and includes interview excerpts. Although editorials in the New York Times and other national newspapers of the era were largely devoid of absolutes, the activists speak a language infused with principles. Representatives from the Women's National Abortion Coalition and a monsignor of the Catholic Church were quoted alike rejecting proposed legislative compromises, speaking of “a principle that cannot be bent to conform to the ideals of a pluralistic society, since what is involved is no less than the murder of the unborn” and that women should have “absolute control,” including “the right to an abortion at any stage of pregnancy” (New York Times, January 2, 1973). These exceptional cases are instructive; they foreshadow our era, in which these sorts of claims dominate our collective consciousness. Although these fundamental arguments were familiar to activists, the abortion debate embraced by mainstream opinion elites was less accommodating to such innovative arguments.

Despite the successes of 1967 and 1968, abortion on the eve of Roe was still illegal in most of the United States. The debate surrounding the New York abortion law, for example, makes up the majority of pre-Roe abortion discussion in the New York Times. By the time the Court decided Roe, 18 states had partially liberalized their abortion laws. As Rosenberg points out, the liberalization of abortion attitudes during this period was driven primarily by factors exogenous to the political system; an outbreak of German measles and the widespread use of thalidomide to treat infertility both significantly increased incidence of congenital disorders in the United States. These incidents help explain the preponderance of concern over congenital disorders in pre-Roe abortion discourse. The census includes 10 such arguments prior to 1973, while in the 7 years following Roe the argument appears only four times, despite a vastly greater number of total editorials. I return to these data in more detail later in this article.

In the years preceding Roe, newspaper coverage of abortion was increasingly in favor of liberalization. In contrast to the dire warnings of the Washington Post two decades prior, in 1971 the Chicago Tribune warned against a “cruel” law leaving women with a choice to “patronize[] an abortion mill where the mortality rate is likely to be as much as 33 times that in a hospital” (Chicago Tribune, June 11, 1971). The popular syndicated columnist Ernest Fergurson wrote that year that legalization was inevitable so that women in need would not need to fear the criminal and health risks associated with prohibition (Los Angeles Times, January 20, 1971).

On the eve of Roe, Americans talked about abortion in a pragmatic framework: sterile operating rooms, preventing exposure of vulnerable women to career criminals, and the distinct emotional traumas imposed by illegal abortions. These notions are a world away from the sterile metaphysical conception of abortion that emerged after Roe, consumed with abstract rights at the comparative expense of real consequences. Among the most compelling of post-Roe advocacy are the stories of life before Roe.

A 1969 editorial article in the Chicago Tribune included two arguments. The editors wrote, “Most Americans think of Canada as a relatively staid, conventional country with a moral outlook not greatly unlike our own. This appraisal is doubtless quite correct, but parliament has now virtually approved sweeping changes in Canada's criminal code … liberaliz[ing] the present law to permit abortions …” before quoting Canadian Prime Minister Trudeau saying the reform was intended to “take the government out of the bedrooms of the nation” (Chicago Tribune, July 5, 1969). The first of these is clearly an argument that the United States should follow international example. The second, presented by Trudeau, is subtler. Without uttering the word, getting “out of the bedrooms” invokes a widely shared desire for personal privacy. The connection here between abortion and privacy is clear, but it is not structured in the formal way privacy claims would come to be in the years following Roe.

Many editorials of the era make only a few arguments but use dramatic context to amplify their impact. An editorial titled “Indifferent in Illinois,” from the Chicago Tribune in 1971, is reproduced in full below:

This editorial features two distinct claims. References to the mortality rate of abortion mills, here with a statistic attached, promote reform in the interest of public health. In the second half of the editorial is a plea for compassion, as the authors describe the dangers abortion prohibition poses to women.

In 1972, the New York Times published at least one argument about health and safety in every one of its 11 editorials endorsing liberalization. Of the 45 editorials they published in the years prior to Roe, 31 featured an argument from public health and 24 argued for modernizing abortion laws, an argument that almost always hinged on the idea that the laws were a product of another medical era.

The Times included during these years only one invocation that abortion is a form of murder, a quotation of Judge T. Emmet Clarie commenting that a decision by an appellate court in Connecticut “invite[d] unlimited feticide” (New York Times, April 23, 1972). Shortly after, the editors approvingly cited a Connecticut legislator saying that if abortions are to be performed in Connecticut, “they [should] be done by doctors in hospitals and not ‘by somebody with a coat hanger' ” (New York Times, April 23, 1972). The editorial also referred to the Connecticut statute banning abortions as “ancient,” a not-too-subtle argument for modernization. Another editorial from the New York Times that year laid the facts out in more detail:

This simple argument, that liberalizing abortion laws would reduce the risk of harm to women, was typical of pre-Roe discussions of abortion policy, even where abortion had been legalized by state legislatures. The same editorial referred to “the medieval status” of prohibition, while another referred to prohibition as “a medieval form of coercion” (New York Times, May 1, 1972). A Los Angeles Times editorial from 1966 was more straightforward still. In the piece, titled “Our Archaic Abortion Law,” the editorial board wrote, “California's law of abortion, like that of most other states, is an antiquated, cruel, and—to a great extent—an unenforceable statute.” The editorial continued, “Many of these illegal (abortion) operations are performed by persons with absolutely no medical training at all. Many others are performed, or attempted, by women on themselves” (Los Angeles Times, May 1, 1972).

The pre-Roe period often featured explicit detail about the consequences of prohibition. A 1971 Chicago Tribune editorial warned of “surgery … without benefit of qualified doctors or hospital services” (Chicago Tribune, June 12, 1971) and an op-ed in the Los Angeles Times the same year described illegal abortions “carried out in … unsanitary conditions” (Los Angeles Times, January 20, 1971). A vivid op-ed by William Farrell, published less than a year prior to Roe, was accompanied by a large photo of a man in a business suit, extending to the reader his outstretched hand holding a coat hanger. The image, accompanying an op-ed discussing the problem of subjecting women to policies made by legislatures dominated by men, anticipated a day 31 years later when President George W. Bush would sign the Partial Birth Abortion Act, flanked on stage by nine men and no women.

A reality of illegal abortions comes through these texts, a reality that was lost in the metaphysical thicket of the post-Roe discourse. Public health would continue to be referenced, albeit far less prominently, after Roe, but the nature of reference underwent a significant change. It was a change that stemmed from the Court's declaration that abortion is a civil right, and consequently from the fact that by the end of the decade, illegal, unsanitary abortions seemed to most elites to be more a memory than a reality.

The Transformation of Privacy

Privacy was not absent from public discourse about abortion prior to Roe, but it looked nothing like the conception of privacy familiar to us in contemporary abortion politics. Where reform advocates talked about privacy before Roe, they almost always talked about the privacy afforded any medical decision. In the year immediately preceding the Court's decision in Roe, the New York Times published 11 editorials dealing primarily with abortion. Across these the editors made three arguments stemming from any conception of privacy, in each case relying on the principle of medical privacy. The dominance of the medical framework is inescapable; talk of decisionmaking with doctors would not give way to the autonomy of women for several years. Prior to Roe, hypothetical women made abortion decisions in concert with their physicians. A representative editorial from 1970 referenced “the decision of a woman and her physician” (New York Times, April 9, 1970) while another from the same year argued the abortion decision “rightfully belongs with the woman and her physician” (New York Times, April 11, 1970). An editorial from 1972 argued that “the decision to have an abortion is one properly to be left with the individual woman and her physician” (New York Times, April 29, 1972). The next reference, in a follow-up editorial a few days later, consisted of an approving citation of a commission's finding that “the matter of abortion should be left to the conscience of the individual concerned, in consultation with her physician” (New York Times, May 1, 1972). Another simply stated that the abortion decision “belongs with the woman and her doctor” (New York Times, May 3, 1972).

In an op-ed in the Los Angeles Times less than a year before the decision in Roe, Ernest Conine approvingly quoted George McGovern, writing that “abortion should be a decision between doctor and patient” (Los Angeles Times, June 22, 1972), a formulation that not only denies the woman agency but reduces her to the role of the patient as well. The rhetoric of the woman as a patient, whether consulting with or at the mercy of her doctors, is a far cry from the post-Roe rhetoric of choice. Before Roe, doctors performed abortions. After Roe elites came to talk about abortion as though it were a procedure women performed on themselves.

Justice Blackmun's opinion for the Court was a direct product of the medical conception of privacy, its influence apparent in his handling of the doctor–patient relationship and understandable in biographical terms considering his experience as legal counsel to the Mayo Clinic. As the Los Angeles Times wrote in a supportive editorial the day after the Court released its ruling in Roe, “the decision will not satisfy those who had argued the mother should make the decision. It insists that the decision be made with the personal physician” (Los Angeles Times, January 23, 1973). Justice Blackmun's decision ultimately placed the physician's counsel in the medical service of a woman's private decision, but the medical framework in which he situated the right would soon disappear from the social conception of abortion. Although Roe constitutionalized the discourse, it did so in a somewhat crude way. Ultimately, it is the activists, legislators, and other elites who perpetually recreate the constitutional discourse inaugurated by the Court.

Opinion elites responded in predictable ways to the Court's decision in Roe; those responses collectively transformed the relationship of the concepts of privacy and abortion. In its editorial applauding the Court's decision in Roe, for example, the Los Angeles Times wrote “a woman and her physician now have unrestricted discretion as to whether a pregnancy should be aborted in the first three months” (Los Angeles Times, January 23, 1973). Particularly significant in the editorial, however, were the board's repeated references in the piece to a “right to privacy” that “surely must include protection from unreasonable intrusions by government in private matters” (Los Angeles Times, January 23, 1973). That the editorial board of the Los Angeles Times thought in 1973 that the right to privacy so surely included an abortion right is surprising. If this argument were so obvious to them, however, why did the editorial board not advance it in any of the 11 editorials in favor of liberalization in the decades prior to Roe? Although the Los Angeles Times argued on January 24, 1973 that abortion was a private medical decision for a woman and her doctor to make, that editorial endorsing Justice Blackmun's opinion for the Court was the first time the paper ever suggested that a right to privacy entailed a right to abortion. The strange but purported obviousness of this argument is a consequence of the power of Supreme Court justices to set the terms of public debate.

Most journalists presented the Court's argument on its own terms, adopting and directly quoting the language of Justice Blackmun's opinion and paying some lesser but nontrivial attention to Justice White's dissent. The Los Angeles Times' front page article about the decision, titled “Abortions and the Right of Privacy,” described and then uncritically invoked the existence of a right to privacy entailing a right to an abortion (Los Angeles Times, January 23, 1973: A1). The Washington Post did the same. The New York Times did not author an editorial about the decision, but the newspaper's lead story, titled “High Court Rules Abortions Legal in First Three Months,” covered the decision broadly, recapitulating arguments on both sides and providing political context (New York Times, January 23, 1973: A1). The New York Times also printed excerpts of the opinion and dissent in a 3:1 ratio, roughly equivalent to the 7-to-2 vote, providing a good summary of each (New York Times, January 23, 1973: A20).

The New York Times did not run an editorial focused on abortion policy during the 2 years following Roe. It returned to the issue following a Boston jury's 1975 conviction of Dr. Kenneth Edelin, a prominent abortion provider. The heart of the editorial was an acknowledgment of the complexity of the abortion issue, suggesting that absent good reason to say life begins at one point or another, the law should favor the welfare of women (New York Times, February 19, 1975). In 1976 Times ran its first editorial to invoke a “right to privacy” entailing an abortion right. The editors wrote of “the Court's reaffirmation of this most intimate of privacy rights” (New York Times, June 3, 1976). The same editorial described the abortion decision itself not in terms of doctor and patient, but instead as “a decision for the woman to make for herself” (New York Times, June 3, 1976). The legislative change in the abortion law in New York did not presage a change in the way the New York Times talked about abortion. The Court's ruling in Roe did.

Discussion of public health did not disappear immediately; the decline of public health took place over the course of the decade following the decision, whereas the rise of a language of rights was stark and sudden. Another editorial that year made reference to a National Academy of Sciences report finding that “legal liberalization was followed by a sharp decline that would undoubtedly be even steeper had the full horrors of the past not been hidden by the secrecy that prevailed prior to legalization” (New York Times, February 5, 1976). The Times explicitly endorsed the core holding of Roe in a 1977 editorial in which the board wrote, “the state has no business intruding into the individual woman's abortion decision.” The editorial concludes with a powerful call for privacy as a form of compassion, quoting an abortion provider saying, “abortions reside in the realm of individual struggle, personal defeat, private hell” (New York Times, January 31, 1977).

The Chicago Tribune argued for abortion rights in seven editorials from 1973 to 1979, describing a choice involving a physician only once. Although the board would write in 1976 of “physicians' rights to make professional decisions in the best interests of their patients” (Chicago Tribune, November 26, 1975), every reference to the abortion decision itself in the remainder of Roe's decade vested the choice entirely with the woman. The paper wrote of “the rights of individual women,” (Chicago Tribune, September 7, 1975) that “society does not have any right to deny a woman an abortion” (Chicago Tribune, November 10, 1976), of “a woman's right to have an abortion” (Chicago Tribune, December 11, 1976), “a woman's right,” (Chicago Tribune, January 11, 1979), a “woman's untrammeled freedom of choice” (Chicago Tribune, July 6, 1980), and so forth.

As the conception of an abortion right founded on privacy rose to rhetorical dominance during the 1970s, advocates found themselves writing of an abortion decision no longer made in consultation with a physician but instead by a lone woman, in the privacy not of a medical office but instead the solitude of her conscience. This is a profound change in the social conception of abortion, one apparent only in the slow transformation of American political language following Roe. It is a transformation that occurred amidst a broader social change, as women achieved an increasing share of social power and respect. It was not conjured by the Court from nothing. The Court catalyzed it. It is a transformation in which the unique authority of the Court, and the pen of Justice Blackmun, can be clearly felt. Before Roe, the decision to terminate a pregnancy was a medical decision made by a patient. Like all other medical decisions, it was made in consultation with a doctor. After Roe the patient receded, replaced by a moral agent. And in the years following the Supreme Court's decision, the doctor disappeared.

Advocacy for legalizing abortion in the years following Roe came to be situated in three rhetorical frameworks: privacy, choice, and autonomy. Each of these is closely related to the other two. Choice is an inherently individual act. Autonomy is the necessary condition for choice as well as a quality of the realized freedom to choose. The transformation of traditional medical privacy into novel legal conceptions of privacy—personal privacy, personal choice, and personal autonomy—began prior to and partially alongside the broader trend toward the de-medicalization of abortion. In place of health care emerged a discourse of constitutional and moral imperatives. For male news reporters and editors metaphysical matters may have been more comfortable subjects than abortion procedures themselves, dangerous and safe alike. But conversations about the metaphysics of abortion elide a critical aspect of abortion services: that abortion services are health care.

Discussions of abortion in the context of public health in the post-Roe period recall the past rather than comparing distinct presents, and the change in the lived experiences of American women seeking abortions was reflected in the emerging social conception of abortion. A 1982 editorial opposing a proposed constitutional amendment to devolve jurisdiction over abortion cases from the Supreme Court to the federal Congress or the states (explicitly favoring the more restrictive law in cases of conflict) was also a product of this change. The editors wrote, “the Hatch Amendment would, at best, take the country back to the time before [Roe], when legal and safe abortions were available only to women who could travel to states that permitted them” (New York Times, March 10, 1982). An editorial from 1983 approvingly quoted Justice Powell, writing that additional abortion regulations would “drive the performance of many abortions back underground, free of effective regulation and often without the attendance of a physician” (New York Times, June 17, 1983). But these are unusually stark examples. Far more often, the public health argument for abortion was reduced in the post-Roe years to little more than a shibboleth uttered without any elaboration. Typical editorials referred to a legal right to “safe abortion,” as opposed to the unmentionable alternative (New York Times, December 18, 1985). One article described the alternative to safe and legal abortion access as “terrible risks” (New York Times, 1995). For the generation who came of age after Roe, the reality of those risks lost its once prominent place in the public consciousness.

It is possible that advocates for abortion access felt the argument to be so well known and understood by the 1990s that they no longer had to actually say it aloud. Instead, a few symbols came to stand in for the once dominant discussion about the risks illegal abortions pose to women's health and safety. Typical references frequently mention “back-alley abortions” (New York Times, October 5, 1991). A piece from 1977 reminded the public of the “common experience that where safe abortions are forbidden by law or by parental fiat, back-alley abortions flourish” (New York Times, June 3, 1977).

These subtle but real changes in the language of public health mirrored the transformation of the social conception of privacy, declining instead of growing. Although it is important not to lose sight of qualitative changes within arguments, the heart of the story of the transformation of the social conception of abortion is of conflict over the right more than the good, and arising not from a commitment to the best possible medical practice but instead from a controversial vision of a constitutional right to a particular conception of privacy itself.

The Transformation of Abortion

Most people, Rosenberg argues, believe Roe radically changed abortion laws. Rosenberg thinks that view is mostly mistaken, that the case did little to accelerate access to abortion services, and that this failure substantiates a skeptical account of judicial power to effect social change (Reference RosenbergRosenberg 2008). I argue, however, that by changing the legal basis of abortion access, Roe profoundly changed the social conception of abortion and the framework in which abortion policy continues to be debated. The influence seems likely to last as long as the Court's recognition of the abortion right itself. In the decades that followed Roe, abortion advocates developed a language of rights, held by individual women in opposition to the state, and responsibilities of the state to ensure the promise of those rights was fulfilled.

Several states legalized abortion prior to the Court's decision; New York did so in 1970. In anticipation of the governor's pre-announced signing of the bill, the New York Times wrote a supportive editorial, reproduced in full below:

This editorial features an implicit reference to medical privacy, but it is a narrow one that follows the tendencies of pre-Roe discourse and otherwise references only modernization (“an antiquated restriction”) and invokes a kind of pluralism (“[t]he new law should offend the conscience of no man [sic] … it leaves to the conscience of each”).

The language of rights is not constrained to affirmations of a woman's abortion right; it permeates thinking about abortion and the state. In 1976, the editors of the Chicago Tribune wrote, “society does not have the right to deny a woman an abortion” (Chicago Tribune, November 10, 1976) and elaborated a little more than 2 years later, discussing late-term abortions, that “there is a point … where priority must be given to society's right to protect itself from having to condone … evils” (Chicago Tribune, January 11, 1979). In an op-ed a few years later, Stephen Chapman quoted Jesse Jackson, writing, “life ‘is really a gift from God. Therefore, one does not have the right to take away (through abortion) that which he does not have the ability to give' ” (Chicago Tribune, February 19, 1984). Here, opposition to abortion is expressed as a powerful statement of rights derived from a kind of natural law.

The language of rights, established in the general case, necessarily informs discussions of the narrower abortion controversies that came to define abortion politics after Roe. Advocates of restrictions focused on issues like parental and spousal notification, public funding for abortion services for the poor, and support for international programs providing abortion services. The editorial board of the Los Angeles Times argued in 1976, for example, against parental notification laws, writing that “parents do not have the right to veto their daughter's decision” (Los Angeles Times, June 5, 1976). Even when discussing the potential personhood of a fetus, the New York Times would, by the 1980s, refer to “the rights of a nonviable fetus” (New York Times, May 27, 1985).

As the scope of rights talk grew larger over time, its exercise grew more rote. In a 1989 op-ed, Mary Steichen Calderone made an extended argument for abortion access as a means of reducing the incidence of congenital disorders. The argument had been common prior to Roe but fell out of favor after. Calderone, however, argued from a “fetus' right not to be born,” a formulation unimaginable in mainstream discourse prior to Roe (New York Times, September 16, 1989). Abortion rights are often employed illustratively, as in an op-ed in which Richard Berke quoted L. Douglas Wilder, the Democratic candidate for Governor of Virginia, characterizing his opponent as “someone who would ‘take away your right to choose and give it to the politicians’ ” (New York Times, October 15, 1989).

The qualitative account I have developed can be subjected to quantitative tests intended not to replicate the insights of the reading but instead to substantiate its core content. Specifically, we can make theoretical predictions about changes in the composition of the discourse, predications that can be falsified and, absent falsification, provide insight into the relative size and scope of these changes. By moving from a qualitative to a quantitative analysis, I aim to step back from examining the leaves in order to count the number of trees in the forest. The quantitative reading I offer in the next section is focused on the language of mainstream establishment opinion elites. Although I discuss examples from reporting, hypothesis tests focus on editorials and op-eds (not letters to the editor), in which the pretense of objectivity is traded for argumentative clarity.

Quantitative Measures

Thus far I have developed a critical qualitative reading of the abortion policy discourse, one emphasizing composition and change. In the final section of this article I present the results of a traditional quantitative content analysis. It is not easy to empirically test an interpretation. But the interpretation I have offered earlier entails testable consequences. Because the nature of the interpretation limits the Court's power to a contingency—it demands an obvious candidate constitutional justification, nascent but organized social movements, and high public salience—quantitative evidence can demonstrate the existence of the consequent, that public discourse did change. Although the quantitative data cannot, by definition, provide causal evidence, they can substantiate the element of the argument hinging on the direction and magnitude of change in the public discourse about abortion. Evidence that the Court was the driver of a significant part of that change rests on the qualitative interpretation above. Below, I report the results of three hypothesis tests intended to observe judicial influence in democratic discourse.

First, because the Court described the abortion right as a consequence of the right to privacy, we might expect to observe an increase in opinion elites' articulation of claims hinging on privacy. This is the privacy hypothesis. Second, because Roe's focus on privacy reduced the influence of public health advocates, we should observe a decrease in opinion elites' discussion of abortion in terms of health and safety. This is the health hypothesis. Third, and most importantly, the Court's declaration of a constitutional right to abortion reduced the efficacy and prominence of nonconstitutional arguments. Constitutional claims are widely treated as though they enjoy lexical priority over nonconstitutional claims. This should be observable in the form of an increase in the incidence of constitutional claims and a decrease in nonconstitutional claims. This is the constitutionalization hypothesis. Evidence for each is discussed in the Results section.

Content Analysis

The quantitative analysis derives from a census of 339 editorials and op-eds about abortion appearing in four leading newspapers of the mainstream opinion establishment. Of the 339 total texts, 251 came from the New York Times (1950–2000) and 88 from the Chicago Tribune, Los Angeles Times, and Washington Post (1950–1980). A total of 991 claims were coded on 45 different categories of three kinds: rights claims, nonrights claims assuming widely shared public goals (primarily about health and safety), and nonrights claims assuming divisive goals. The taxonomy is detailed in Table 1.

Table 1. Arguments for Liberal and Prohibitory Abortion Policies

The taxonomy indicates difference on two levels of analysis: differences between specific claims and between kinds of claims. The first of these is fairly straightforward, the second more theoretical. What are the defining qualities of the relevant conceptual frameworks? In this case, I want to ask how the rhetorical structure of the abortion debate changed following the judicial proclamation of a right to abortion. Because the core argument I make is that the Court has a special power to force discussion in constitutional terms, the most important distinction I make is between constitutional claims, mostly about rights, and other kinds of claims.

All data collected from the Washington Post, Los Angeles Times, and Chicago Tribune as well as a representative subsample of the data from the New York Times were coded by a second coder. Each coder carefully read the texts and noted instances of claims specified in the coding protocol, including references to “abortion rights,” “the right to choose,” and equivalent formulations postulating rights.

Data were grouped by newspaper and inter-coder reliability calculated for each. Reliability rates are reported as free-marginal kappas (Reference CohenCohen 1960). Kappas ranged from .71 for the Los Angeles Times to .88 for the Chicago Tribune. Although included data were coded by two coders and meet conventional thresholds for reliability (K > .7), there may be good reasons to think the taxonomy itself either misrepresents some existing features in the discourse about which we might care or that it imposes an answer upon the question. The principle constraint on data integrity is the representativeness of the purposive sample, that is, how broadly can we generalize from these four newspaper editorial boards (Reference Riffe, Fico and LacyRiffe, Fico, & Lacy 2005). The four newspapers were chosen because, although different in many ways, they occupied in 1973 positions of prestige in the mainstream liberal opinion establishment.

Results

The results of the content analysis strongly suggest that opinion elites writing in the editorial pages of the liberal establishment responded to Roe by adopting a rhetoric that emphasized constitutional claims at the expense of all other kinds of claims. These data are consistent with predictions entailed by the theoretical account above, in which the Court utilized its institutional powers to elevate constitutional arguments from relative obscurity to commanding stature. Following Roe, opinion elites in four leading newspapers of the liberal establishment came to defend abortion on rhetorical terms dictated not by a vibrant public debate but rather by nine unelected judges.

From 1950 through the end of 1972, the New York Times, Washington Post, Los Angeles Times, and Chicago Tribune printed 74 editorials directly arguing for various abortion policies. A content analysis of these texts demonstrates the prevalence of important arguments in the pre-Roe abortion discourse. When editorial boards wrote about abortion in the two decades prior to the Supreme Court's decision in Roe, this is, in an important sense, how they thought about abortion. The discourse prior to Roe was composed of a diverse range of claims. Arguments from privacy of the sort that would eventually carry the day at the Supreme Court were comparatively obscure. Arguments from public health, by contrast, cast a long shadow over the debate, appearing more than any others.

To test the three hypotheses outlined above, post-Roe ratios are compared with pre-Roe ratios. The resulting data are the ratios of post-Roe to pre-Roe incidences of claims. Table 2 shows these post-pre ratios, measures of the incidence of claims per editorial during the second period as a proportion of the first period. The Los Angeles Times, for example, has a post-Roe to pre-Roe constitutional claims ratio of 1.21, meaning that for every claim per editorial they printed in the 23 years preceding Roe, they averaged 1.21 times as many in the 27 years that followed. As detailed in Table 2, the post-pre ratio for privacy claims in the New York Times is 1.75, indicating that the average incidence of privacy claims increased 75 percent from the first time period (1950–Roe) to the second (Roe–2000).

Table 2. Evolution of Abortion Politics in Four Major National Newspapers, 1950–1972 Compared with 1973–1980 and 1973–2000, Presented as a Ratio of Post-Roe Period over Pre-Roe Period

The results of the analysis presented in Table 2 strongly support health and constitutionalization hypotheses while providing mixed evidence for the privacy hypothesis. To falsify the hypothesis, we would need to observe no substantive positive change in the ratio of privacy claims following Roe. A substantive change would constitute a failure to reject the hypothesis.

The privacy hypothesis predicts that the values for privacy should all exceed 1.00. As can be seen in Table 3, the post-pre ratios for privacy range from 0.67 to 1.75. In two of the newspapers, the Chicago Tribune and the Los Angeles Times, no substantial change was observed, whereas in the Washington Post the post-pre ratio was under 1.00, inconsistent with the theory. In each of these cases, there is sufficient evidence to reject the hypothesis. This was not the case, however, for the New York Times.

Table 3. Abortion Politics in the New York Times during Three Time Periods

Claims per editorial during 1965–Roe, Roe–Casey, and Casey–2000. K = .78.

The data are strongly consistent with the other two hypotheses. Considerations relating to public health formed the most significant pre-Roe arguments for liberalizing abortion policy. Public health concerns were cited as justifications for liberalizing abortion policy an average of 1.1 times per editorial in the pre-Roe period, the most of any argument. The health hypothesis states that we should observe a substantive decrease in the most significant frameworks unendorsed by the Court. In the case of Roe, this is the line of argument from public health that influenced Justice Blackmun's thinking but which did not form the basis of the ultimate abortion right, which hinged on the right to privacy articulated in Griswold v. Connecticut (1965). This should be observable in the post-pre ratio for the aggregate category of public health and modernization. To falsify the hypothesis, we would need to observe no substantive decrease in the ratio of public health claims to editorials following Roe.

Substantial evidence for the hypothesis was observed in all four newspapers, and none came close to falsification. The likelihood of a given editorial utilizing claims about health decreased considerably in each of the newspapers following Roe. In the extended series, the expected number of such claims decreased from 1.20 to 0.22, an 81 percent reduction. In the 13 years prior to Roe, the opinion page of the New York Times discussed abortion in a medical framework so often that a reader could expect to find in any editorial on abortion policy one or more distinct arguments for liberalization in the interest of health and safety. After Roe, that reader could expect on average to read an argument from public health in roughly one out of every five editorials.

The numbers in other newspapers are similar. The Los Angeles Times' ratio is 0.28, reflecting almost as dramatic a decrease, whereas the ratios for the Washington Post and Chicago Tribune were 0.17 and 0.21, respectively, virtually identical to the New York Times. These findings comprise strong evidence for the health hypothesis. The effects are reliably observable, theoretically predicted, and substantial in magnitude.

The constitutionalization hypothesis is the most important of the three. It states that following the Court's articulation of a novel constitutional right, opinion elites should respond by “constitutionalizing” their discourse. To observe no change or an increase in the comparative incidence of nonconstitutional claims would falsify the constitutionalization hypothesis. Evidence for the constitutionalization hypothesis is the strongest of the three and is found across all four newspapers in the study. Post-pre ratios for nonrights claims from shared goals were under 1.00 in all four newspapers, ranging from 0.26 to 0.46. Theory predicts that post-pre ratios for nonrights claims from divisive goals should be less depressed, and in one case, the Chicago Tribune, it slightly increased, but in all cases these changes were dwarfed by the magnitude of the increases in constitutional claims. These ranged from a comparatively modest 1.21 for the Los Angeles Times to 3.26 for the New York Times and 5.83 for the Chicago Tribune. Not only the arguments but the kind of arguments made changed following Roe.

The testable questions the qualitative interpretation poses are fundamentally about how the way we talk about abortion—how we do abortion politics—changed following the Court's recognition of an abortion right on January 23, 1973. Figure 1 visualizes these trends with kernel density estimations, showing trends associated with discourses characterized by constitutional rights and health and safety from 1965 until 2000 in the New York Times. The figure reveals a discourse centered on health and safety transformed in the wake of Roe. The Court's decision radically accelerated a nascent but marginal way of talking about abortion in constitutional terms. While talk about abortion in a public health framework did not disappear entirely or immediately, it went from the dominant frame to a peripheral one. These data also reveal the robustness of the Court's influence, as the discourse remains constitutionalized through the end of the century.

Figure 1. Incidence of Constitutional and Ordinary (Nonconstitutional) Claims in Discourse about Abortion Politics in the New York Times, 1965–2000. K = .78; N (texts) = 251; N (claims) = 664. The Relative Heights of the Curves Represent the Distribution of Claims.

The quantitative data are consistent with the qualitative analysis. They suggest an account of the Court as a kind of pageant judge, examining contestant constitutional arguments and declaring a winner. The Court did not invent the notion that abortion is a civil rights issue. That idea came from public intellectuals and activists, and these data are consistent with that fact. But in declaring that abortion is a civil right, the Court encouraged opinion elites to contest abortion policy on constitutional grounds. Having declared it to be a civil right, the justices guaranteed abortion would be debated as a civil right. The justices confronted a controversy marked by a variety of ambiguous and contradictory definitions and they chose one. And because they are the Supreme Court, that choice was accepted by the opinion elites who drive democratic politics.

Conclusions

I have argued, by way of a combination of qualitative and quantitative analysis, for a novel account of judicial power and a particular interpretation of the effect of that power in abortion politics. By recognizing a constitutional right to abortion predicated on privacy, the justices played a significant and unique role in transforming the social conception of abortion. These changes, evident in elite discourse about abortion, have potentially profound implications for democratic theory. Finally, in addition to providing a novel perspective on the evolution of abortion politics in the United States, this study significantly complicates Reference RosenbergRosenberg's (2008) account of a judiciary severely limited in its power to change society. The justices' decision in Roe changed the language and criteria with which abortion policy was and continues to be understood and debated.

The language of rights adopted by the Court presaged an era of bitter political entrenchment. In the wake of Roe, opinion elites, politicians, and ordinary citizens continued as they had to clash over abortion politics. They clashed over parental and spousal consent, waiting periods, and the breadth of permissible restrictions, but they did so in a conceptual framework that may have been particularly ill-suited to mutual understanding, let alone compromise. For proponents of abortion rights, initial enthusiasm over a judicial embrace of a right to abortion was followed by the reality of a constitutionalized policy domain. Public deliberation about constitutional commitments is unusually challenging; it is supposed to be. To imagine alternative policies is much harder when that imagination contests who we are. Future research into judicial influence over society should distinguish between immediate policy influence and its potential alternatives. The next question posed by this study is how that transformation may have changed ultimate policy outcomes in state legislatures. Comparative research could shed further light on the effects of constitutional frames on legislative behavior.

However specific changes in policy outcomes may have changed because of the constitutionalization of abortion politics, the way we do politics—the language we use and the criteria of judgment entailed by that language—matters immensely in itself. Public deliberation is a necessary and defining component of democracy. The terms of deliberation have important consequences. Extensive research on framing effects has demonstrated the importance of language and criteria for judging in shaping public opinion and policy outcomes, particularly in civil rights cases (Reference ChongChong 1993; Reference Chong and DruckmanChong & Druckman 2007; Nelson, Clawson, & Reference Nelson, Clawson and OxleyOxley 1997), but little research has explored the unique institutional role of the Supreme Court in shaping and entrenching rights frames.

Constitutional arguments are taken by democratic theorists to be different from other kinds of arguments because constitutional commitments are necessary for liberal democracy (Reference DworkinDworkin 1978; Reference ElyEly 1980). These kinds of commitments are not unique to abortion politics. This study suggests that constitutional options open to the justices in a variety of other policy arenas—including privacy, firearm ownership, sexual liberty, and many others—empower the Court in ways political scientists have not adequately explored. Finally, there are good reasons to think the Court may possess influence over the way we think about the scope of appropriate government power. One promising candidate is the study of the Court's effect on discourse about federal power during the Great Depression and New Deal. Another potential vantage point for judicial power is the study of nonjudicial paths to constitutionalization, such as the passage of the Civil Rights and Voting Rights Acts and the ratification process. Further study of the origins of ideas in debates about the Civil War Amendments, Women's Suffrage, and the Equal Rights Amendment offers tremendous potential to identify potential popular counters to the Court's power.

The justices of the Supreme Court influence politics in ways beyond their ability to execute policy. I have discussed one such avenue of influence, the power to set the terms of public debate in civil rights cases. That power is limited and contingent, but also important. The justices cannot enshrine constitutional commitments in our law without also enshrining them in our language. The justices may or may not be able to effectively execute social policy, but they can and have shaped the way we think about social problems. For democratic theorists who celebrate the Court's role in promoting constitutional discourse, this should come as welcome news. For advocates and public intellectuals, depending on their disposition, it can feel either limiting or empowering. And for citizens who favor a politics of negotiation and compromise, a discourse of inviolable rights, whether to choice, privacy, life, or anything else, can make the possibility of compromise seem all the more elusive. We cannot make sense of any of these important changes in our politics without first making sense of the Court's unique role in bringing them about.