Abstract
The literature base pertaining to abortion decision-making and adjustment has grown substantially since legalization of abortion in the U.S. 30 years ago. However, the available research has suffered from various theoretical and methodological shortcomings and the findings do not seem to do justice to the complexity of abortion experiences among women residing in a cultural context that continues to exhibit intense conflict over the legality and morality of abortion.

The purpose of this review is to summarize previous research, offer suggestions for improving the quality of work on the topic of abortion, and to highlight specific content areas holding considerable promise for enhancing our understanding of the risks and benefits of abortion.

Introduction

...When the post-abortion literature is carefully examined, unanswered questions
abound, underscoring the need for researchers to step back and take stock of
theoretical and methodological shortcomings in order to facilitate more fruitful
work in this area. Further, with 43% of American women making the decision to
abort at least once prior to age 45 (Henshaw, 1998), approaching this area of study
in a focused and substantive manner should be a national priority. Abortion has
been legal for three decades in the U.S., rendering the time to put aside political,
social, and economic agendas and take an honest look at both the positive and
negative aspects of women’s experiences with this common medical procedure well
overdue.
Motivated by the desire to promote research with the control and depth of
exploration needed to more definitively ascertain how women with diverse backgrounds,
characteristics, and life circumstances process an abortion experience and
continue with their lives, three objectives are pursued in this article: (1) to review
the existing literature pertaining to the psychology of abortion; (2) to offer suggestions
for introducing more conceptual sophistication and methodological rigor to abortion
research; and (3) to consider a few content areas in need of further exploration as this
literature base begins to mature.

Overview of the existing literature
Most women view an unintended pregnancy as a stressful personal situation
(Adler & Dolcini, 1986; Cohen & Roth, 1984; Olson, 1980), with estimates of the
percentage of births in the U.S. resulting from pregnancies considered unwanted
(unintended, no child desired), or mistimed (unintended, child desired in the future)
at conception ranging from 49 to 60% (Forrest, 1994; Henshaw, 1998; Squires,
1995). Moreover, approximately 77% of births to women over 40 and 86%
of births to teenagers are the result of unintended pregnancies (Squires, 1995).
Many studies indicate that concerns with becoming a single parent and partner
relationship difficulties are among the most common motives for seeking an abortion
(Soderberg, Andersson, Janzon & Slosberg, 1997; Torres and Forrest, 1988). Other
frequently cited reasons include the following: (1) concerns that carrying a pregnancy
to term will interfere with the continuation of one’s current intimate relationship,
future education, career, or personal plans (Allanson and Astbury, 1995; Faria,
Barrett & Goodman, 1985; Patterson, Hill & Maloy, 1995), (2) age (Faria et al.,
1985), (3) not feeling ready for parenting (Faria et al., 1985; Kero, Hoegburg,
Jacobsson & Lalos, 2001), (4) insufficient finances (Faria et al., 1985; Glander,
Moore, Michielutte & Parsons, 1998), (5) desire to postpone childbirth (Kero et al.,
2001; Tornbom, Ingelhammar, Lilja, Moller & Svanberg, 1994), and (6)
feeling as though one does not have the time and energy for another child (Kero
et al., 2001).
Abortion tends to bring relief and a reduction in women’s perceptions of stress
(Adler, 1975). However, there is relative consensus among scholars in the field
that at least 10–20% of women who have had an abortion suffer from serious negative
psychological complications (Adler et al., 1990; Lewis, 1997; Major and Cozzarelli,
1992; Zolese and Blacker, 1992). With over 1.3 million abortions performed annually
in the U.S. (Alan Guttmacher Institute, 2000), using the more conservative 10% figure
would result in 130,000 new cases of women experiencing related psychological
problems each year. Among those who are adversely affected, many stress-related
symptoms have been identified, including anxiety (Franco, Tamburrino, Campbell,
Pentz & Jurs, 1989; Moseley, Follongstad, Harley & Heckel, 1981; Niswander,
Singer & Singer, 1972), depression (Coleman & Nelson, 1998; Cougle, Reardon &
Coleman, 2003; Gould, 1980; Moseley et al., 1981; Reardon & Cougle, 2002a,
2002b; Thorp, Hartmann & Shadigian, 2003), sleep disturbances (Barnard, 1990;
Gould, 1980), substance use/abuse (Coleman, Reardon, Rue & Cougle, 2002b;
Drower & Nash, 1978; Reardon & Ney, 2000; Yamaguchi & Kandel, 1987), and
increased risk of suicide (Gissler, Kauppila, Merilainen, Toukomaa & Hemminki,
1997; Reardon et al., 2002). A few recent studies have further identified relations
between maternal history of abortion and problematic parenting (Benedict, White &
Cornely, 1985; Coleman, Reardon & Cougle, 2002; Ney, Fung & Wickett, 1993).
As noted by Coleman and colleagues (2002a), an abortion history is essentially a
package variable comprised of numerous personal, relationship, and situational factors
precipitating an unplanned pregnancy and the decision to abort, while carrying the
potential to trigger negative psychological effects.
Women with a history of abortion and women who decide to continue an unplanned pregnancy may be distinguishable in various ways that are related to mental health. For example, in a study by Bradley (1984) of Canadian women who had recently given birth, women with a history of abortion tended to describe themselves as self-reliant, independent, rebellious, and to enjoy being unattached or unconnected to other people, places, and things.
Miller (1992) found that women who abort tend to be unmarried, independent minded,
and are likely to view abortion as both personally acceptable and as
acceptable in the eyes of family members. In a study by Skjeldestad (1994),
Norwegian women opting to abort usually favored liberal abortion legislation;
however, age and occupational status were unrelated to the decision. Research by
Russo and Denious (2001) revealed that the statistical association between abortion
and psychological problems was reduced considerably when the effect of partner
violence was statistically controlled. These results suggested that the experience of
violence, which was systematically related to the choice to abort, could be a
salient factor responsible for declines in mental health.
Numerous studies have now identified many of the demographic, individual,
relationship, and situational characteristics that place women at risk for psychological
disturbance in the aftermath of abortion. The available data specifically indicate
that women are more prone to post-abortion psychological problems when they
have any of the following characteristics: (1) low self-efficacy for coping with the
abortion (Major et al., 1990), (2) low self-esteem (Cozzarelli, Karrasch, Sumer &
Major, 1994), (3) external locus of control (Cozzarelli, 1993), (4) difficulty with
the decision (Bracken, 1978; Osofsky & Osofsky, 1972), (5) when there is emotional
investment in the pregnancy (Lyndon, Dunkel-Schetter, Cohan & Pierce, 1996;
Remennick & Segal, 2001), (6) perceptions of one’s partner, family members, or
friends as non-supportive (Major et al., 1990; Major & Cozzarelli, 1992), (7)
timing during adolescence, being unmarried, or poor (Adler, 1975; Bracken,
Hachamovitch & Grossman, 1974; Campbell, Franco & Jurs, 1988; Franz &
Reardon, 1992; Osofsky & Osofsky, 1972), (8) pre-existing emotional problems or
unresolved traumatization (Speckhard & Rue, 1992), (9) a poor or insecure attachment
relationship with one’s mother or a childhood history of separation from one’s
mother for a year or more before age 16 (Cozzarelli et al., 1998; Kitamura et al.,
1998; Payne, Kravitz, Notman & Anderson, 1976), (10) involvement in violent
relationships (Allanson & Astbury, 2001; Russo & Denious, 2001), (11) traditional
sex-role orientations (Gold, Berger & Anders, 1979), and (12) conservative views of
abortion and/or religious affiliation (Bogen, 1974; Osofsky & Osofsky, 1972;
Soderberg, Janzon & Slosberg, 1998). Further, adjustment problems have been
documented to be more common when a pregnancy is initially intended (Ashton,
1980; Friedman, Greenspan & Mittelman, 1974; Lazarus, 1985; Major et al., 1985;
Miller, 1992), abortion occurs during the second trimester (Anthanasiou, Oppel,
Michelson, Unger & Yager, 1973), and when women are involved in unstable
partner relationships (Llewellyn & Pytches, 1988; Soderberg et al., 1998). Finally,
feelings of being forced into abortion by one’s partner, others, or by life
circumstances, increase the risk for negative post-abortion outcomes (Friedman
et al., 1974).
...no researcher has offered a truly broad theoretical model to help bring together the
existing data in a cohesive manner. A framework for the purpose of integrating the
existing knowledge and for providing guidance pertaining to future research efforts is
offered in the next section.

An ecological framework for the study of abortion
Given the many diverse characteristics of individuals opting for abortion as well as the
great environmental variability in which decisions are embedded, a framework is
needed to incorporate information related to how the following variables might be
related to differential experiences: (1) individual difference factors including demographic
variables (e.g., age, ethnicity, socioeconomic status, reproductive history,
and marital history), history of stressful life experiences, personality variables,
intelligence, personal beliefs, and psychological and physical health, (2) relationship
history variables including family of origin/attachment dynamics, present family
situation, current and past intimate relationships, and friendships, etc., (3) social
support systems prior to, during, and after the decision to abort, (4) material and
social circumstances surrounding the abortion decision, and (5) cultural values and
norms pertaining to abortion. Such a model would generate numerous logical
hypotheses related to differential psychological responses to abortion decisionmaking
and adjustment based on a wide range of personal and social characteristics.
For example, very little research has explored the associations between normal and
abnormal variability in personality characteristics and abortion decision-making
and adjustment. Using clinical data alone one might hypothesize that women
with narcissistic personality disorder would be more inclined to experience rage and
less likely to experience guilt and a full range of emotions in response to the experience
when compared to others (Barnard, 1992; Siomopoulos, 1981). Adoption of a
well-developed model, however, would offer insights pertaining to how additional
individual difference, relationship, situational, social, and cultural factors may influence
this general expectation. In addition to including the factors described above,
such a model would need to incorporate sensitivity to the complex bi-directional
and multi-directional relations among the many personal characteristics and
environmental factors influencing the decision to abort and adjustment afterwards.
Brofenbrenner’s (1979, 1986, 1989, 1993) bioecological model of human development
is well-suited as a perspective for exploring the psychology of abortion decisionmaking
and adjustment in the full contextual richness that the topic deserves. Unlike a
more formalized theory, the bioecological framework does not offer the possibility
of generating specific predictions relative to the psychology of abortion or any
other human experience for that matter. Instead it offers an organized conceptual
umbrella for more specified theory development and provides insights regarding the
development of research designs likely to generate logically comprehensible data.
Although the model has been primarily applied to development during childhood,
Brofenbrenner (1998) has emphasized the utility of the model relative to understanding
developmental processes across the lifespan. In addition to acknowledging the role
of biologically based characteristics and other individual difference factors, the model
provides a framework for understanding the role of a wide array of intra-familial and
environmental systems in development. Influences stemming from the individual’s
immediate context microsystem), family members’ daily experiences that impact
the individual (mesosystems), community settings including schools, places of employment,
and churches (exosystem), and the current state of society encompassing mores, laws, and norms (macrosystem) are conceptualized as mutually interacting to
define development over time.

The bioecological model is based on two primary propositions (Brofenbrenner,
1998). First, optimal intellectual, emotional, social, and moral development throughout
life is dependent upon active participation in progressively more complex, reciprocal
interaction with persons, objects, and symbols in the individual’s immediate
environment. These enduring forms of interaction in the microsystem comprise the
primary driving force in development. Second, the nature and strength of proximal
processes are viewed as varying as a function of the characteristics of the developing
person and both the immediate and more removed features of his or her environment.
The individual’s age, historical periods experienced, and the nature of the particular
developmental outcome under consideration also factor into a complete understanding
of proximal processes. Bronfenbrenner (1989, 1993) has emphasized the
‘‘developmentally-instigative’’ characteristics of the individual, encompassing one’s
personality characteristics, skills, goals, beliefs, and motivations as powerful determinants
of the manner in which contextual elements are experienced by the developing
individual...

Moderators of post-abortion adjustment
A moderator is any variable that influences the magnitude or direction of an
association between a designated independent variable or predictor and a given
dependent variable or criterion (Baron & Kenny, 1986). In the general overview of
the literature provided above, several moderators of post-abortion adjustment were
identified; however due to space limitations, three were selected to highlight the
importance of including modifiers in the study of abortion, with a final paragraph
briefly covering a few additional moderators (ethnicity, socioeconomic status, and
cultural values). Age, emotional investment or attachment to the fetus, and beliefs
regarding the humanity of the fetus were specifically chosen based on preliminary
research indicating that they carry considerable potential to expand our knowledge
regarding distinct emotional trajectories in addition to enhancing our understanding
of the contextualization of abortion experience.

Moderation by age. Slightly under 25% of U.S. abortions are performed on women
under age 20 (Alan Guttmacher Institute, 1996) and although the data are somewhat
inconsistent, most of the available studies suggest that younger women when compared
to older women are more inclined to experience post-abortion difficulties
(Adler, 1975; Bracken et al., 1974; Campbell et al., 1988; Franz & Reardon, 1992;
Osofsky & Osofsky, 1972). Adolescents are generally much less well-prepared, both
emotionally and financially, to assume the responsibilities associated with parenthood,
and they are logically the recipients of much greater social pressure to abort. Further,
when compared to older women, younger women are more inclined to engage in
denial and delay in decision-making, necessitating the use of procedures that are associated
with heightened physical and emotional risk (Bracken & Swigar, 1972; Cates &
Grimes, 1981; Lemkau, 1988).
The decision to abort during adolescence may have become more conflict
ridden over the last 10–15 years compared to earlier periods as adolescents of
both sexes have become more inclined to express conservative political views
and pro-life attitudes (Broggess & Bradner, 2000; Stone & Waszak, 1992). Even if
an individual approves of abortion in the abstract, a personal decision to abort
may be more difficult when the peer culture is disapproving and/or if one’s
partner is opposed to abortion. Studies pertaining to endorsement of abortion
tend to focus on females or do not differentiate between male and female
respondents; however data from the National Survey of Adolescent Males
reported by Broggess and Bradner (2000) revealed that only 24% of U.S. males
aged 15–19 years in 1995 agreed that it was right for a female to obtain an abortion
for any reason. This represented a significant decrease from a 37% endorsement
rate in 1988.
Studies designed to identify predictors of the decision to abort vs. deliver among
adolescents are limited and tend to be narrowly focused on demographic variables.
For instance, one recent large-scale effort employing data from the 1995 National
Survey of Family Growth revealed that the following variables were related to
higher rates of abortion compared to birth: discrepancy in age between a woman
and her partner (an older male), race (Blacks were most likely to opt for abortion,
followed by Whites, and then Hispanics), partner’s religious orientation (none or
non-protestant), and higher educational attainment of the partner and of the
woman’s mother (Zavodny, 2001). Another study conducted in Australia showed a
strong partner influence in adolescent decisions to abort and those who had a
mother or a sister who had aborted were more inclined to abort compared to adolescents
without such a familial history (Evans, 2001). Additional work is clearly needed
to examine personal, relational, and social predictors of the choice to abort among
adolescents.
Adolescents obviously differ from older women undergoing abortion in many ways.
For example, their life experiences are more limited, they are more inclined to possess
idealized views of the future, they are usually less focused in terms of life goals, and
they may have a less supportive social network in addition to being at lower levels
of intellectual, moral, and emotional maturity. Further research is needed to identify
the mechanisms through which environmental and maturational differences may
converge to define distinct post-abortion experiences among women who abort at
different stages of life. An added challenge, consistent with a bioecological framework,
will be to incorporate individual difference variables such as socioeconomic status and
personality characteristics, relationship, and contextual factors into the more global
patterns that emerge.
The psychology of abortion 243
Moderation by attachment to the fetus. The woman’s emotional investment in the
pregnancy or attachment to the fetus is another variable with potential power to
moderate relations between the decision to abort and psychological adjustment
relative to abortion. In discussing the fact that researchers have tended to tiptoe
around this topic, two Australian researchers note that ‘‘given the powerful influences
in our culture promoting the sanctity of pregnancy, that is personhood from conception,
and abortion as murder at a symbolic level or at the level of the pregnant
woman’s experienced reality, it is little wonder that consideration of maternal attachment
issues in the context of abortion has generally been avoided’’ (Allanson &
Astbury, 2001, pp. 146–147). One small-scale, interview-based study by Patterson
et al. (1995) revealed that women who felt more of a bond to the fetus prior to
abortion experienced more difficulty afterwards compared to women who did not
feel such a bond. In this study, bonding tended to emerge as a function of the
participant’s awareness and embracement of pregnancy-related physical changes.
There are several studies, which have addressed issues related to this topic. For
example, pregnancy intendedness, which was mentioned previously (Ashton, 1980;
Friedman et al., 1974; Lazarus, 1985; Major et al., 1985; Miller, 1992), meaning
attached to the pregnancy (Major et al., 1985; Remennick & Segal, 2001;
Zimmerman, 1977), commitment to the pregnancy (Lyndon et al., 1996), and
later-term abortions (Cohen & Roth, 1984) have all been found to be associated
with more post-abortion distress responses including significant levels of guilt, anxiety,
and depression. In each of these circumstances, the probability that women have
developed more of an attachment to the fetus is increased. In a recent study by
Kero et al. (2001), slightly over one-third of respondents who obtained first-trimester
abortions reported positive or slightly positive initial feelings toward the pregnancy
indicating that a fairly large segment may form some level of attachment and
experience vulnerability to adjustment problems afterwards. Further, in a study of
college students’ emotional responses to an abortion experience, 30% agreed or
strongly agreed with the following statement: ‘‘I sometimes experience a sense of
longing for the aborted fetus’’ (Coleman & Nelson, 1998).
There is considerable evidence indicating that many women develop strong
emotional connections to the fetus prior to birth (Leifer, 1977; Cranley, 1981;
Condon, 1986). In fact, research by Leifer (1977) revealed that attachment to the
fetus may begin shortly after conception. Attachment dynamics would be expected
to differ based on the degree of emotional investment in the pregnancy and the
investment seems likely to be relatively low among women considering an abortion.
However, this assumption should be tested in light of work by Kemp and Page
(1987), which indicated that in pregnancies considered high risk due to the possibility
of serious health complications or even loss of life for the fetus or the woman, women
reported comparable levels of attachment to those undergoing uncomplicated
pregnancies. More research is needed to address the extent to which feelings of
attachment to the fetus are common prior to an abortion and to explore the extent
to which the level of attachment is related to negative post-abortion outcomes.
Moderation by beliefs regarding the fetus. When women have not developed any form
of attachment to the fetus, their beliefs about the humanity of the fetus may still
moderate emotional reactions to the experience. In a study of over 800 women,
244 P. K. Coleman et al.
Conklin and O’Connor (1995) found that women who believed that the fetus was
human and underwent an abortion scored significantly lower than women who had
not had an abortion on measures of self-esteem and satisfaction in addition to reporting
more negative affect. However, women who had an abortion and did not endorse
beliefs pertaining to the humanity of the fetus were indistinguishable from women
without a history of abortion in terms of the three post-abortion adjustment measures.
Weak beliefs about the humanity of the fetus expressed as only slightly disagreeing
with the statement that fetuses are human were sufficient to result in relations between
abortion and compromised well-being. The authors pointed out that the results
held up even after statistical controls were instituted for various contextual variables
and they emphasized the importance of the findings in light of the inherent
difficulty in obtaining moderator effects in field research. Clearly the robustness of
this association demonstrates the merit of more concentrated attention on the role
of women’s beliefs regarding the humanity of the fetus in adjustment to abortion.
In a comprehensive analysis describing factors leading to pathological mourning
following an abortion, Array (1968) pointed out that the defenselessness of the
fetus is likely to be a salient factor underlying powerful guilt feelings associated with
the loss. Future research efforts should endeavor to dissect women’s conceptions of
the humanity of the fetus to identify more specific barriers to positive adjustment.
In addition to defenselessness, researchers might explore the extent to which
the woman’s mental image of a fetus resembles a human being physically and
psychologically (e.g., is able to experience pain). Longitudinal research is also
needed to examine the degree to which women’s beliefs in the humanity of the
fetus are susceptible to change, the conditions under which changes are inclined to
occur (e.g., education regarding fetal development, a religious conversion, etc.), and
how shifting beliefs may result in a delayed adverse response to an abortion.
Moderation by culture, ethnicity, and socioeconomic status. Many additional moderators
merit more focused research attention; however, in light of space limitations, only
three more (cultural attitudes, ethnicity, and socioeconomic status) will be described
briefly. Women’s psychological responses to abortion are embedded in widely varying
cultural contexts. In many nations throughout the world there are strong moral and/or
legal sanctions against abortion; whereas in several other countries abortion is a
passively accepted medical practice. A few studies have yielded results suggesting
that women electing abortion in an anti-abortion social context may be more inclined
to experience negative post-abortion emotions (Adler, 1975; Illsley & Hall, 1976;
Major et al., 1997; Miller, 1992). A careful analysis of women’s responses to abortion
in different cultural contexts defined by distinct levels of social acceptance should
enable researchers to begin to gauge the extent to which negative responses are socially
constructed.
Not only do vastly different attitudes regarding the morality of abortion exist across
cultures, but belief systems regarding the acceptability of abortion within the same
nation may vary considerably based on ethnic and socioeconomic group affiliations.
Although most studies designed to examine attitudes toward abortion and postabortion
adjustment have been conducted with White participants, a number of
studies have compared White women with Blacks, Hispanics, Asians, and women
of other races. The literature pertaining to Black women is the most well-developed
The psychology of abortion 245
and will be the focus here. Black women are more likely than White women to have
early pregnancies (Presser, 1971), experience an unintended pregnancy (Pratt &
Horn, 1985), delay abortion decision-making (Bracken and Swigar, 1972; Kerenyi,
Glascock & Horowitz, 1973), and they choose abortion at a rate that is three times
that of White women (Centers for Disease Control and Prevention [CDC], 1994).
Interestingly, however, many studies conducted over the past several decades have
consistently revealed that Black women are less supportive of legalized abortion
than White women (Dugger, 1998; Hall & Marx-Ferree, 1986). More research is
needed to examine predictors of abortion decision-making and post-abortion
adjustment among Blacks, particularly in light of the fact that Black women are a
very heterogeneous group divided by numerous variables including class and
religiosity among others that influence their ideological stance on abortion (Dugger,
1998). Nevertheless the general disparity between expressed attitudes and behavior
relative to abortion among Blacks suggest that Black women may be more inclined
than White women to opt for abortion when they believe it is morally wrong and/or
without social support for the decision. In this context, Black women may be more
vulnerable to post-abortion psychological problems and more research is needed
to explore this possibility.
A few studies have indicated that poor women compared to their more financially
secure counterparts are more inclined to experience adverse reactions to abortion
(Adler, 1975; Osofsky & Osofsky, 1972). However, comparisons between women
from different socioeconomic groups relative to post-abortion psychological adjustment
are complicated by the fact that many variables that may co-vary with poverty
(e.g., low self-esteem, timing during adolescence, involvement in unstable partner
relationships, being unmarried, low levels of social support, and feelings of being
forced by circumstances to abort etc., reviewed previously) are also predictive
of negative post-abortion adjustment difficulties. Studies designed to examine the
potential moderating role of socioeconomic factors should incorporate controls for
the many possible confounding variables. Future studies should also explore possible
differences in abortion attitudes and decision-making based on women’s socioeconomic
backgrounds as these topics have been neglected in the published literature.
Just as Brofenbrenner’s bioecological model lends itself to exploration of a wide
array of moderator variables relevant to abortion decision-making and adjustment,
it is likewise conducive to examination of mediational processes. This is the topic
of the next section.
Mediators of post-abortion adjustment
The study of mediators in psychological processes offers insight pertaining to how
characteristics of the individual or experiences are able to partially or fully explain
relations between specific predictor variables and outcomes (Baron & Kenny, 1986).
More precisely, mediators are defined as pathways through which an independent
variable like abortion history has an impact on a dependent variable, such as the
experience of positive or negative emotions. Research conducted to explore mediators
of relations between factors surrounding or post-dating an abortion and adjustment
reactions has tended to focus on a few variables. By far most research to date has
dealt with self-efficacy beliefs, with other mediators including the form of blame associated
with the abortion experience and reproductive events following the abortion.
246 P. K. Coleman et al.
Research pertaining to each of these potential mediators is described below, with an
emphasis on the more thoroughly studied construct of self-efficacy.
Mediation by self-efficacy beliefs. Bandura (1982) defines self-efficacy as judgments
incorporating both knowledge and confidence relative to executing the actions
necessary to successfully complete various life tasks. The power of self-efficacy beliefs
to mediate the effects of other personal and situational determinants of behavior has
been emphasized by Bandura (1989), rendering this construct potentially very useful
to a systemic or contextual analysis of women’s experiences with abortion. Consistent
with the self-efficacy theory, research by Major et al. (1990) suggested that selfefficacy
for coping fully mediated the link between perceptions of social support of
various forms and positive post-abortion adjustment. More specifically, perceptions
of strong social support from three sources: one’s partner, family members, and
friends were associated with high self-efficacy relative to coping with abortion and
enhanced self-efficacy was related to lower depression, more positive mood,
and fewer expected negative post-abortion consequences. However, no direct relation
was observed between social support and the various indicators of negative
adjustment. Other studies have highlighted the central role of self-efficacy in
abortion-related coping (Cozzarelli, 1993; Major et al., 1985; Mueller & Major,
1989). These data underscore the importance of including measures reflecting
women’s perceptions along with environmental factors in efforts to understand
responses to abortion, as Bandura (2002, p. 278) recently noted, ‘‘personal agency
and social structure operate interdependently rather than as disembodied entities.’’
Future research might explore additional socio-demographic, personal, and experiential
antecedents to feelings of efficacy relative to coping with an abortion in addition
to examining the extent to which women continue to feel efficacious over several years
after the abortion. Bandura’s (1989) description of the four primary informational
sources that relate to the development of personal efficacy provides a useful direction
for such efforts. First, personal accomplishment history (successes and failures)
represents the most direct influence on mastery expectations. Women who have
successfully coped with similar life experiences (a previous abortion experience or
any other form of perinatal loss) would be expected to develop an enhanced sense of
self-efficacy relative to coping with an abortion. Second, watching others engage in
task-relevant activities can generate vicarious estimations in observers pertaining to
their own capacity for mastery. Therefore, women who have observed their acquaintances,
friends, and/or relatives work through an abortion experience in an adaptive
manner would be expected to have higher self-efficacy relative to coping with an abortion.
However, Bandura (1989) emphasized the fact that inferences derived from social
comparison are indirect and are theoretically more susceptible to change than those
fostered through direct experience. Verbal feedback from others regarding one’s
potential for coping effectively is the third avenue through which self-efficacy beliefs
may develop. Like social comparison, appraisals from others tend to be weaker sources
of information in the formation of self-efficacy beliefs than those derived directly from
one’s own experiences. As noted previously, research does suggest that when others are
supportive of one’s decision to abort, self-efficacy tends to be higher. However, given
the volatile nature of the abortion topic and the resulting diversity of opinions,
individuals facing an abortion decision are the likely recipients of conflicting feedback
The psychology of abortion 247
regarding the probability of effective coping. The fourth mechanism described by
Bandura (1989) with relevance to the emergence of self-efficacy beliefs relates to
emotional arousal. Individuals anticipate failure when they experience high levels of
aversive physiological arousal; whereas lower levels of arousal tend to be linked
with success expectancies. Women who experience high levels of personal stress and
anxiety manifested physically before the abortion decision may therefore be expected
to show lowered levels of self-efficacy for coping during and after the procedure.
Mediation by Attributions of Blame. When faced with negative life events, individuals’
abilities to cope with the stress and make positive adjustments often relate to the
degree to which they feel the situation may have been modifiable (Abramson,
Seligman & Teasdale, 1978; Janoff-Bulman, 1979; Mueller & Major, 1989).
Viewing others as the cause of negative events and engaging in self-blame targeting
enduring traits tend to be associated with ineffective coping and problematic adjustment
(Mueller & Major, 1989). On the other hand, when self-blame for negative
events focuses on one’s own behavior, which could have theoretically been averted,
more positive outcomes are likely (Mueller & Major, 1989). Therefore, women who
blame an unintended pregnancy on their partner or an aspect of their own character
or personality, such as impulsivity or lack of responsibility, would seem to be inclined
to suffer more than women who blame the problem on some personal behavior such as
having forgotten to purchase birth control. Research by Mueller and Major (1989)
with 283 women who underwent first trimester abortions supported the operation
of attributions of blame as a mediator between the experience and adjustment.
Those who were low in other-blame and low in self-character blame demonstrated
the most positive psychological adjustment at 3-weeks post-abortion.
Mediation by subsequent reproductive events. As an effort is made to conduct more
long-term studies of post-abortion functioning, reproduction-related events including
having another abortion or other forms of perinatal loss such as a miscarriage or stillbirth,
difficulty conceiving or problems with a desired pregnancy, and giving birth
may be found to operate as mediators of adjustment several years after the abortion.
A few small-scale studies and case reports have indicated that reproductive events
often bring back thoughts and emotions associated with the procedure even among
women who report no distress at the time of the abortion (Lemkau, 1988; Congleton
& Calhoun, 1993; Stotland, 1998). However, more systematic analysis of the topic
is needed.
After discussing the bioecological framework as a potentially useful model for bringing
clarity and vision to the study of the psychology of abortion, a number of related
moderating and mediating variables were considered as potentially fruitful areas for
further research. Having outlined these conceptual issues, we now turn our attention
to various avenues for enhancing the methodological rigor of research pertaining to the
topic of abortion.
Needed methodological innovations in abortion research
Research designed to explore post-abortion emotional responses has generally been
wrought with many methodological problems. Most studies have been conducted
248 P. K. Coleman et al.
with small samples (typically under 300), limited to one geographical area (Speckhard
& Rue, 1992; Wilmoth, deAlteriis & Bussell, 1992), and initial consent to participate
rates are often as low as 60% (Adler, 1975; Cohen & Roth, 1984), with attrition rates
reported to be as high as 60% (Major et al., 1985). Additional limitations of the
existing post-abortion literature include the following: (1) insufficient attention to
the personal, interpersonal, and contextual complexity of women’s choices to abort
which carry the potential to produce both positive and negative outcomes, (2)
exclusive reliance on self-report data, (3) few prospective, longitudinal investigations,
(4) limited use of appropriate control groups, and (5) reliance on non-standardized
measures of psychological health (Zolese & Blacker, 1992).
A number of large-scale record-based studies using medical claims data in
the United States, Finland, and Canada have successfully avoided many of the
methodological limitations of other post-abortion research (Coleman et al., 2002b;
David, Rasmussen & Holst, 1981; Ostbye, Wenghofer, Woodward, Gold &
Craighead, 2001; Reardon et al., 2002, 2003). In particular, problems of concealment,
recruitment, attrition, and inadequate measurement of psychological symptoms
are averted in these studies as actual medical claims are used as the data source.
Further, all the studies except for the one by Ostbye et al., incorporated data collected
over several years in addition to utilizing women who delivered as a comparison
group. The results of these studies have consistently revealed that women with a
known history of abortion experience higher rates of mental health problems
of various forms when compared to women without a known history. However,
attempts to infer causality from these record-based investigations are restrained by
minimal controls for potentially confounding factors. Only a few demographic and
psychosocial variables have been effectively controlled in the record-based studies
due to the limited number of variables available to select from.
Although studies conducted in recent years have been designed to overcome a
number of the shortcomings, several problems remain and until they are sufficiently
addressed definitive answers to the many questions raised over the years regarding
the meaning of abortion relative to women’s psychological health will be difficult to
reach. In this section, we describe four areas wherein methodological innovations
are greatly needed to advance our efforts to understand how abortion impacts
women’s lives: (1) the need for more diversified research strategies, (2) an increased
emphasis on longitudinal designs, (3) incorporation of appropriate control groups,
and (4) instituting controls for pre-existing psychological state.
The need for more diversified research strategies
Over two decades ago, research led by Kent involving a group of Canadian women,
who had previously indicated no problems associated with an abortion, revealed
considerable differences between the initial questionnaire data and information
subsequently gathered during in-depth psychotherapy sessions (Kent, Greenwood,
Loeken & Nichols, 1978; Kent & Nichols, 1981). Moreover, a firm rational decision
for an abortion was found to frequently coexist with feelings of deep pain and
bereavement. In a major national poll by the Los Angeles Times, 56% of women
admitting to a past abortion reported a sense of guilt and 26% reported regretting
the choice to abortion, suggesting indirectly that the behavioral choice to abort may
frequently conflict with beliefs and values (Skelton, 1989). Difficulties assessing and
The psychology of abortion 249
ultimately understanding the full complexity of women’s responses to abortion may be
related to self-denial of emotional experiences at the time of the abortion in order to
‘‘get through’’ the procedure once women have made the intellectual decision to
abort. A participant in a study by Patterson et al. (1995, p. 687) conveyed this
type of response well: ‘‘I was in a state of numbness, just really going through whatever
motions were required to get this job done.’’ Accurate assessment may also be
hindered if negative experiences are expressed less directly in the form of maladaptive
behaviors or psychosomatic complaints or if women with underlying ambivalent
feelings regarding abortion are reluctant to openly express problems encountered.
Unfortunately, most of the existing post-abortion data are based on the exclusive
use of narrowly focused questionnaire-based self-reports. There are many logical
ways to expand and diversify the methods used to study the psychology of abortion.
Qualitative studies probing women’s thoughts and feelings pertaining to personal,
relationship, and contextual factors that entered into their decisions to abort as well
as postabortion emotions, thoughts, and experiences (personal and professional) are
needed to do justice to the inherent complexity of this area of study. The use of
open-ended questions posed by empathetic interviewers, who convey the wide
range of emotions women may experience in response to an abortion is likely to
result in rich data that is less vulnerable than other methodologies to social desirability
biases. For example, in a study of Israeli and Russian immigrants by Remennick
and Segal (2001) using an interview methodology, widely ranging experiences were
reported with comments bordering on exhilaration afterwards ‘‘when it was over I
felt alive and a boss to myself again’’ (p. 50) to reactions suggesting profound
trauma ‘‘I couldn’t stop thinking about this, counting what week in pregnancy I’d
be by now, and how the baby would have looked, and all that . . . When I saw mothers
with babies in the street I winced. In my dreams, I saw the hospital, the nurses, and
myself in the stirrups . . .’’ (p. 50). In a large Swedish study of 854 women one year
after an abortion, which incorporated a semi-structured interview methodology
requiring 45–75 min to administer, rates of negative experiences were considerably
higher than in previously published studies relying on more superficial assessments
(Soderberg et al., 1998). Specifically, 50–60% of the women experienced emotional
distress of some form (e.g., mild depression, remorse or guilt feelings, a tendency to
cry without cause, discomfort upon meeting children), 16.1% experienced serious
emotional distress (needing help from a psychiatrist or psychologist or being unable
to work because of depression), and 76.1% said that they would not consider abortion
again (suggesting indirectly that it was not a very positive experience).
Given the political, social, and moral issues surrounding abortion, disclosure of
sensitive, substantive data is likely dependent on the extent to which researchers
are able to provide a truly accepting interpersonal context. A study conducted in
Tanzania, where pregnancy interruption is prohibited unless continuation is life
threatening, demonstrated the salience of the setting for improving data quality
(Rasch et al., 2000). When women were admitted to hospitals for incomplete
abortions and assured of confidentiality within the context of in-depth personalized
dialogues with interviewers, they were much more likely to reveal an induced abortion
than when information was gathered in a less empathetic manner. Similarly, Patterson
et al. (1995) found that assurances of anonymity, researcher political neutrality, and
that researchers would not pass judgment of any kind as they were simply interested
in understanding the decision and adjustment processes, gave the participants the
250 P. K. Coleman et al.
necessary confidence to describe their experiences in very candid detail. One logical
method for creating a comfortable environment conducive to generating women’s
true thoughts and emotions associated with an abortion experience would be to
use women who have had abortions themselves and thus are inclined to serve as
compassionate interviewers. An alternative is to collect data in a group discussion
forum conducted over several sessions, which would enable ample opportunity to generate
genuine and substantive discussions among women with a history of abortion. A
recent study conducted in Thailand by Whittaker (2002) incorporated a combination
of data collection strategies including a survey on reproductive health, in-depth
interviews, and vignettes in focus group discussions and revealed that the latter two
methods were the most effective means of gathering sensitive, abortion-related data.
Focus groups typically include 6–10 participants with a knowledgeable moderator
guiding the discussion (Whitaker, 2002). Further, in a study using the newest cycle
of the National Survey of Family Growth, a computerized private recording system
was employed in addition to the standard interview and the combined methodology
produced an abortion reporting response rate which was 59% of the expected rate
based on prevalence data (Fu et al., 1998). This represented a considerable increase
from the 45% figure previously reported using the interview methodology alone.
In addition to the need for qualitative studies, more research incorporating information
from other sources is needed. Data gathered from significant individuals in
women’s lives (e.g., partners and family members) and/or behavioral assessments
(possibly from counselors and other abortion provider personnel or conducted by
researchers) should enhance efforts to assess the complexity of women’s positive
and negative experiences before, during, and after the decision to abort. For example,
if the researcher is interested in the effects of abortion on partner relationships or marital
quality, information could be derived from the partner, friends or family members
who know the couple well, and the researcher might conduct a laboratory assessment
of relationship factors such as communication, supportiveness, trust, and/or anger.
As indicated in the literature overview section above, there has been a recent trend
in post-abortion research toward conducting large record-based studies with this
methodology offering considerable promise relative to avoiding numerous pitfalls
associated with post-abortion research. However, the utility of such large-scale efforts
relative to enhancing our understanding of the psychology of abortion is necessarily
dependent upon the extent to which the records contain demographic and contextual
data. Although assessments of pregnancy intendedness and other relevant factors
of the abortion experience may not be readily obtained with this methodology,
researchers can work with the data in creative ways to construct variables that approximate
the constructs of interest. For example, an exclusive focus on women taking
birth control pills prior to their births or abortions would result in a sample of
women likely to fall into the ‘‘unintended pregnancy’’ category. Unfortunately,
accessing complete medical records on large populations of people is nearly impossible
in the United States. However, this research technique is promising in countries with
socialized medicine and centralized records.
Longitudinal research
Most of the existing abortion studies have been conducted within a framework
suggesting that an abortion experience, even if experienced as traumatic, will be of
The psychology of abortion 251
short duration. Data on post-abortion reactions have typically been collected
within hours or weeks of the event, with assessments extending beyond six months
uncommon. Recent research, however, indicates that women undergoing an abortion
may experience long-term negative effects. For example, in a study of women involved
in clinical trials of the abortifacient, RU-486, regret increased from 2 weeks to 6–8
months post-abortion and Miller and his colleagues concluded that ‘‘the low point
following the abortion may not occur for days, weeks, or even months’’ (Miller,
Pasta & Dean, 1998, p. 262). Miller (1992) had previously found evidence of delayed
reactions in a study covering three years. More recently, Major and colleagues (2000)
analyzed the psychological outcomes of women one hour pre-abortion, and 1 h,
1 month, and 2 years post-abortion. They reported an increase in negative emotions
and a decrease in relief and positive emotions between the assessments at 1 and 2 years
following the abortion. The results also revealed an increase in depression and a
decrease in satisfaction with the abortion decision over time.
Evidence from professionals who work with women who have had abortions
and studies incorporating a case study methodology suggest that while abortion
may be an effective short-term coping strategy, it may also function as an insidious
long-term stressor (Butlet, 1996; De Veber, Ajzenstat & Chisholm, 1991; Joy,
1985; Speckhard & Rue, 1992). Longitudinal research incorporating opportunities
for women to express the process whereby the sense of relief might fade and feelings
of dissatisfaction with the decision may begin to add stress to their lives is needed.
Studies should be of a prospective nature as retrospective feelings and impressions
surrounding the events preceding the decision to abort and at the time of the
abortion are undoubtedly distorted by life events as well as one’s actual emotional
and intellectual adjustment to the decision. In many cases, researchers have measured
psychiatric variables prior to the abortion (e.g., Major et al., 2000), but rarely is
there prospective data available prior to the pregnancy. The use of pre-abortion/
post-conception psychological assessments offer poor baseline measures as women
who are about to have an abortion are not likely to be in their ‘‘normal’’ psychological
state, given the stress associated with the unintended pregnancy and possible concerns
regarding the procedure (Adler & Dolcini, 1986; Cohen & Roth, 1984; Olson, 1980).
Assessment of pre-conception measures of psychological health would necessitate a
large-scale prospective study in order to identify a sufficient number of women
opting to abort. There has not yet been a national study designed to prospectively
examine psychological adjustment relative to reproductive events.
Finally, research is needed to help differentiate between women who are more or
less likely to experience long-term negative effects of abortion. One small study
indicated that long-term adverse reactions were more common when women felt
they received poor treatment during the abortion, experienced conflict over the
meaning of abortion, felt ambivalent about the pregnancy, or experienced a bond to
the fetus prior to the abortion (Patterson et al., 1995).
Adequate control/comparison groups
One of the most significant methodological problems facing the post-abortion
literature is the fact that very few studies incorporate an appropriate control group.
The ideal control group has been suggested to consist of women who wanted an
abortion and did not obtain one for personal reasons (e.g., guilt, anxiety, fear, etc.)
252 P. K. Coleman et al.
or due to external pressures (e.g., from others such as a partner or parents) (Keshen,
2003). Following this line of reasoning, the ideal ‘‘treatment’’ group would consist
of women who really wanted an abortion and were not behaving against their
primary desire or personal belief system. As research reported throughout this
review suggests, abortion decisions are often not easily made and are likely to
represent the culmination of an array of mixed emotions and external circumstances
that are not well understood. Therefore, abortion ‘‘wantedness’’ is perhaps best
conceptualized on a continuum rather than as a discrete variable, with most women
falling somewhere in between the two extremes. Assuming that researchers are able
to identify ‘‘clean’’ groups by instituting the above inclusion criteria, sample sizes
are likely to be reduced considerably and the information obtained will be limited
in terms of generalizability. From a practical and conceptual standpoint, women
who simply carry an unintended pregnancy to term would therefore seem to represent
a more logical comparison group, with studies using this strategy offering greater
potential to provide a representative assessment of relative risk than the previously
described methodology.
Most of the published work pertaining to post-abortion psychological effects is
based on studies that do not address relative-risk and the use of an appropriate control
group is often overlooked. However there is an emerging literature using women who
have delivered as the comparison group without assessment of wantedness.
Small-scale studies comparing psychological reactions within a short period following
childbirth or abortion have either reported no significant difference in psychological
outcome (Anthanasiou et al., 1973; Zabin, Hirsch & Emerson, 1989) or have
revealed a heightened risk of emotional difficulties such as anxiety and depression
during pregnancy and the postpartum period among women who abort (Colman &
Colman, 1971; Bradley, 1984; Kumar & Robson, 1978, 1984; Linares, Leadbeater,
Jaffe, Kato & Diaz, 1992). Large scale investigative efforts using women who delivered
as a comparison group (described briefly above) have only been conducted recently.
These studies have consistently indicated that abortion is associated with significantly
more mental health problems (Coleman et al., 2002b; Cougle et al., 2003; Reardon
et al., 2003), higher rates of substance use (Coleman et al., 2002a), and a significantly
higher risk of suicide (Gissler et al., 1997; Morgan, Evans, Peter & Currie, 1997;
Reardon et al., 2002). For example, the results of the largest U.S. post-abortion
study to date comparing over 54,000 low-income women on state medical assistance,
indicated that women who had an abortion in 1989 with possible subsequent
pregnancies had significantly higher rates of outpatient psychiatric diagnoses than
women with only birth experience in the target year and no history of subsequent
abortions after eliminating all cases with psychiatric claims 12–18 months prior to
the initial pregnancy (Coleman et al., 2002b). This difference was revealed when
data for the full time period were examined (17% higher) and when only data from
women with claims filed on their behalf within 90 days (63% higher), 180 days
(42% higher), 1 year (30% higher), and 2 years (16% higher) of the pregnancy
event were considered. Data using the same sample and focusing on inpatient claims
revealed similar findings (Reardon et al., 2003). These different rates were observed
after controlling for age, months of eligibility for services, and the number of
pregnancies. Although pregnancy intendedness was not directly assessed in this
study, women living under compromised economic conditions are logically less
likely to plan a pregnancy than the average woman.
The psychology of abortion 253
There is not a sizable body of literature designed to specifically compare women
who have had abortions to women who carry unintended pregnancies to term.
While there are many studies examining these women as separate groups, few direct
comparisons are available. One recent analysis of the National Longitudinal Study
of Youth (Reardon & Cougle, 2002a,b) revealed that at an average of 8 years following
their first pregnancies, women who aborted a first pregnancy were significantly more
likely to be at risk for clinical depression compared to similar women who carried a
first unintended pregnancy to term. The risk of depression was most elevated
among women who were married and those who had no history of divorce. These
results were obtained even after controlling for age, income level, race, and a psychological
measure taken prior to the women’s first pregnancies. An observed difference
such as this between women who deliver and abort becomes more meaningful when
viewed relative to the stressfulness associated with carrying an unintended pregnancy
to term. New mothers frequently feel vulnerable, inadequate, and depressed after
giving birth (Fleming et al., 1990) and an unintended pregnancy seems likely to
accentuate some of these feelings. In a recent meta-analysis, postpartum depression
was estimated to afflict approximately 13% of women regardless of intendedness
status and unplanned pregnancy was found to have a small yet significant association
with postpartum depression (Beck, 2001). Work by Leathers and Kelley (2000)
similarly revealed that unintended pregnancy was associated with maternal depression.
Any differences detected between women who abort and deliver an unintended
pregnancy demonstrating more negative outcomes for the women with a history of
abortion underscore the potential for abortion to initiate adjustment problems. As
potentially useful as the comparison between women with unintended pregnancies
resolved though abortion versus delivery is in many ways, it remains possible that
pre-existing psychological factors leading to the choice to abort operate as critical
factors in determining post-abortion mental health. Moreover, the classification of
pregnancies as unintended, untimely, or unwanted introduces a whole complex area
of nuance and uncertainty, since these reactions are likely to vary across the pregnancy
and are inclined to be influenced by third party responses to a woman’s pregnancy.
While non-pregnant women have seldom been used as a control group, this
comparison may be useful with the logic for this comparison strengthened by the
notion that abortion, in theory should ‘‘return’’ a woman to her pre-pregnancy state.
A record linkage study conducted in Finland was one of the few studies to employ this
type of control (Gissler et al., 1997). The relative risk from death among women
who had an abortion the previous year when compared to women who had not
been pregnant was equal to 3.7 for suicide, 2.2 for accidents, and 4.3 for homicide.
The use of non-pregnant women as a control group is likely to produce the most
meaningful results if the sample is restricted to a matched group of never pregnant
women and women who abort their first pregnancy. A few comparisons to the general
population have also been conducted. For example, a record-based study in Canada
compared admission rates for psychiatric hospitalization over a period of 5 years
following an abortion to a matched sample of women who had not had abortions,
irrespective of other reproductive events (Badgley, Caron & Powell, 1977). The
results indicated that 13% of women who had abortions were hospitalized compared
to 4% of the control group. Using a much larger sample of Danish women, David et al.
(1981) found that the psychiatric admission rate in the first 90 days after an
abortion was 2.5 times higher than that of the general population of women.
254 P. K. Coleman et al.
In other cases, researchers who have not directly utilized control groups have
sought to place their findings into the context of the broader literature pertaining to
the prevalence of psychological problems in the general population. Since this practice
involves comparisons across study designs, researchers need to exercise caution
in making comparisons. For example, in a follow-up study of 442 women who had
abortions two years earlier, conducted by Cozzarelli et al. (2000), the results revealed
that 24.5% of the sample had scores above the cutoff for clinical depression on the
Brief Symptom Inventory (BSI). The researchers concluded that the depression
rate detected in their study was only slightly over that of American women in
general by reference to a study of national prevalence conducted by Blazer et al.
(1994), which indicated a 20% lifetime prevalence rate of major depression
among women 15–35 years of age. The problem with this comparison is that
Cozzarelli and her colleagues were comparing symptoms of depression measured in
the most recent month to lifetime prevalence rates. Fortunately, Blazer and colleagues
(1994) also report the prevalence of current (30 day) major depression for
females aged 15–24 and 25–34, years as 8.2% and 4.3% respectively. This
suggests that the depression rates two years after abortion are 3–5 times higher
among women who have had an abortion compared to the general population of
women.
An additional logical set of comparisons might involve a detailed sociodemographic,
psychological, and lifestyle analysis of four groups of women: (1) those who report
predominantly positive feelings surrounding their decision to abort and the procedure,
and satisfaction with their decision over time, (2) those who experience a preponderance
of negative emotions before, during, and in the years following an abortion, (3)
those who experience considerable distress before and during the procedure, but heal
rapidly and do not report any long-term suffering, and (4) those who are not very distressed
prior to and during the abortion, but experience negative reactions afterwards.
Previous studies have tended to either target the average woman seeking an abortion
(most of the studies cited herein) or have analyzed clinical samples (Reardon, 1997;
Speckhard, 1987). However systematic examination of women reporting distinct
emotional trajectories within the same report are rare in the published literature.
One study by Congleton and Calhoun (1993) compared the experiences of women
who reported emotional distress in conjunction with an abortion with women
who reported relieving/neutral responses. The groups were similar in terms of
many sociodemographic and abortion-related characteristics. An interesting pattern
of similarities and differences emerged in the results. Forty-eight percent of the
distressed group reported recalling feelings of loss immediately after abortion compared
to none in the non-distressed group. Other responses that were considerably
more common among the distressed group included the desire to replace the fetus,
sadness/grief, behavioral changes such as increased drug use, and depression around
the anniversary date of the abortion. However, 20% of the non-distressed group
reported depression and 44% of the non-depressed group expressed sadness/grief in
conjunction with the experience over time. The majority of women in the distressed
group (88%) and in the non-distressed group (72%) reported long-term post-abortion
‘‘catalytic’’ events including childbirth and learning about early fetal development
that aroused thoughts or emotions regarding the abortion and both groups (36% of
each) reported fantasizing about the fetus prior to the abortion. This study was conducted
with a very small sample (n¼50) and relied on a retrospective methodology;
The psychology of abortion 255
however, it does offer a strong impetus for larger scale prospective work adopting
similar comparison groups.
A final way to examine the impact of abortion in a manner that enables sensitive
exploration of personal and situational determinants of abortion-related adjustment
problems while controlling for individual difference factors is to study women
who have had more than one abortion and report more emotional difficulties with
one of them. The less difficult of the two situations would function as the ‘‘control’’
condition. By conducting such a within-subjects design incorporating an extensive
analysis of the relationship dynamics and life circumstances surrounding the decision
to abort that comprise the context of abortion at two points in time, many individual
difference factors are effectively controlled. Obviously attention should also be given
to life events occurring before and after each abortion. Although nearly 43% of
women who have one abortion will abort again (Henshaw & Silverman, 1988),
securing a sample of sufficient size for this strategy could prove problematic, as
women who have suffered from one abortion seem inclined to suffer again.
Similarly, women who do not have problems the first time around seem unlikely to
have difficulties the second time unless they experience emotionally significant and
related events before or after the second abortion. Support for the notion of emotional
continuity from one abortion to another was provided by Kero et al. (2001), who
found that 94% of women with repeat abortion experiences used the same words to
describe feelings associated with the two occasions. However sufficient research has
not been conducted on the emotional continuity idea and it seems equally probable
that responses are cumulative with levels of negative affect differing considerably
from one experience to the next, particularly among women who have some level of
difficulty handling the first abortion.
Unfortunately, results generated from this within-subjects comparison strategy
may lack generalizability to the general population of women undergoing an
abortion in light of research suggesting considerable lifestyle and psychological
differences between women with a history of one versus two or more abortions.
Specifically, women who repeatedly choose abortion when compared to women
with one abortion tend to be more sexually active (Berger et al., 1984; Howe,
Kaplan & English, 1979), are more inclined to be involved in less satisfying and/or
shallow partner relationships (Berger et al., 1984; Fisher, 1986; Szabady &
Klinger, 1972), are less likely to live with their partners (Tietze, 1978), express
negative feelings more frequently (Leach, 1977), are more often dissatisfied with
themselves (Leach, 1977), report less concern about moral or social issues (Bracken
& Kasi, 1975), and are less likely to report being religious (Leach, 1977). Women
with repeat abortions also experience more sleep problems (Tietze, 1978; Freeman,
1980; Berger et al., 1984), tend to be more immature (Fisher, 1986), are more
prone to being emotionally detached (Fisher, 1986), report a lack of nurturing in
their families of origin more often (Fisher, 1986; Kitamura, Toda, Shima &
Sugawara, 1998), and show significantly higher distress scores on interpersonal sensitivity,
paranoid ideation, phobic anxiety, somatization, hostility, and psychoticism
(Freeman, 1980).
The above discussion suggests that the selection of an appropriate comparison
group in post-abortion research is not an easy, readily discernable process.
Moreover, it appears necessary to employ a variety of reasonable control groups,
recognizing that while no single comparison is perfect, all can be informative
and each may have advantages in teasing out a clearer picture of abortion’s risks
and benefits for particular groups of women facing an abortion under various
circumstances.

Prior psychological health
With research indicating that pre-existing psychological problems represent a risk
factor for post-abortion psychological problems (Anthanasiou et al., 1973; Osofsky
et al., 1973; Lask, 1975; Miller, 1992; Major et al., 2000), more research with
controls for mental health prior to the abortion is necessary to ascertain the extent
to which particular post-abortion psychological problems can be attributed to the
experience. The few recent studies that have included controls for prior psychological
difficulties or psychological status suggest that abortion is associated with a heightened
risk for in-patient and out-patient treatment of various psychological problems,
depression, and suicide (Coleman et al., 2002b; Cougle et al., 2003; Reardon et al.,
2002, 2003). The contention that only psychologically vulnerable women are inclined
to exhibit mental health problems in the aftermath of an abortion can no longer be
sustained by the evidence. However, further research should be devoted to a careful
analysis of how abortion might exacerbate pre-existing problems. Furthermore,
additional research with sociodemographically diverse samples using more extensive
controls for pre-existing psychological problems of varying forms and severity that
extend back several years prior to the abortion is in order.
There have been a few post-abortion studies that have exclusively focused on
psychologically vulnerable women. For example, in one study of women with a
prior history of psychiatric problems, none of those who carried to term subsequently
committed suicide over an 8–13 year follow-up, whereas 5% of those who aborted did
take their lives (Jansson, 1965). Additional research has indicated that pregnancy
and childbirth reduce the risk of suicide (Appleby, 1991; Appleby & Turnbull,
1995; Drower & Nash, 1978; Hoyer & Lund, 1993; Jansson, 1965). Further, prior
suicidal behavior is apparently not predictive of abortion, nor does it explain the
increased risk for suicide attempts after abortion (Morgan et al., 1997). With these
findings suggesting that childbirth may reduce the risk of subsequent suicide attempts
whereas abortion may aggravate that risk, a greater sense of family obligations and
a fear of hurting one’s children may account for fewer suicide attempts and suicidal
thoughts among those who deliver (Linehan, Goodstein, Nielsen & Chiles, 1983).
The same connectedness to family may also help protect women from exacerbation
of other mental health problems. More research is needed to directly address this
possibility.

Post-abortion content areas in need of attention
In addition to the need for a broad theoretical framework and the many avenues
for enhancing the methodological integrity of the post-abortion research, there are
several content areas in great need of focused research attention. In keeping
with our emphasis on the bioecological framework, we discuss three areas of pressing
concern: (1) investigation of the positive effects or benefits of abortion to women’s
health; (2) the dynamic association between abortion decision-making and adjustment
to the experience; and finally (3) the relevance of domestic violence to understanding
abortion decision-making and adjustment.

Psychological responses to abortion apparently involve a complex combination of
positive and negative emotions and cognitions. A recent study of 211 Swedish
women seeking an abortion revealed that two-thirds of the respondents expressed
both positive and negative feelings about the abortion, with the remaining one-third
reporting only negative feelings (Kero et al., 2001). Anxiety, relief, grief, anguish,
and emptiness were the commonly reported emotions. At the close of their report,
these authors noted ‘‘The relief to be saved from unwanted parenthood did not
exclude painful feelings that may reflect experiences of ethical conflicts and feelings
of loss. This complexity is seldom recognized in abortion studies’’ (p. 1489). Other
studies have likewise revealed how abortion may serve as a coping strategy ushering
in an affective response characterized by a sense of relief while also provoking
simultaneous or subsequent negative emotions (Barnard, 1990; Selby, 1990;
Vaughan, 1991).
Legalization of abortion 30 years ago was based on the idea that abortion benefits
women, yet amazingly, well-designed research specifically documenting how the
procedure enhances women’s quality of life is generally absent from the professional
literature. A few medical researchers have voiced the reminder that the onus of
proof lies with those who perform or support any medical intervention to demonstrate
beyond a reasonable doubt that the procedure is therapeutic (e.g., Ney, 1993). At the
cultural level, there are widely held assumptions that when women are able to
avoid undesired childbearing, are free to pursue more highly valued paths, focus on
the children they already have, or postpone childbirth until they are physically and
psychologically ready to assume the responsibilities and enjoy child rearing, then
they are far better off materially and psychologically...
Given the accumulating data pertaining to the risks associated with abortion,
documentation of the presumed benefits is needed to assist women in making
well-informed decisions. Micro and macrolevel analyses designed to explore the
questions of how women as individuals may benefit from abortion and how females
in general may have prospered economically and socially from access to abortion
are needed. As suggested by Reardon (1997), an appropriate way to examine the
personal benefits of abortion might be to collect prospective data pertaining to
why women seek abortions and then follow them over several years to investigate
the extent to which the abortion has, in fact, led to fulfilled expectations. For
example, are abortion decisions that are based on relationship dynamics, educational,
or occupational plans linked with the anticipated benefits? These data could be
subsequently pooled in order to offer a more extensive assessment of how women
have benefited generally.
. The trend toward ‘‘evidence-based medicine’’ is a reflection of the need for medical advice to be more solidly
grounded in well substantiated benefits as opposed to assumed benefits (Grimes,
Bachicha & Learman, 1998).

Abortion decision dynamics and post-abortion adjustment
Studies suggest that decisions regarding how to resolve an unplanned pregnancy
are difficult for many women, even when they express an unwavering decision to
terminate (Brett & Brett, 1992; Gilchrist, Hannaford, Frank & Kay, 1995; Handy,
1982; Mueller & Major, 1989). Research by Husfeldt, Hansen, Lyngberg, Noddebo
and Pettersson (1995) indicated that 44% of the women surveyed had doubts
about their decision when the pregnancy was confirmed and 30% continued to
express doubts when the abortion date arrived. When Kero et al. (2001)
interviewed 221 Swedish women seeking an abortion, 46% revealed that their
thoughts regarding termination evoked a conflict of conscience. There is also evidence
indicating that many women who initially request an abortion will subsequently opt
not to go through with the procedure (Gilchrist et al., 1995; Handy, 1982).
Further, the results of a study noted earlier indicated that 76.1% of women who
had an abortion would never consider repeating the experience again (Soderberg
et al., 1998). Studies also suggest that many women who have an abortion become
pregnant again within one year and elect to carry the subsequent pregnancy to term
(Tietze, Rowland-Hogue & Cates, 1982). In this situation, the second pregnancy
may be a result of women feeling as though the previous abortion was a mistake.
The decision-making process has been identified as one of the primary variables
differentiating between women who have post-abortion psychological adjustment
problems and those who do not (Adler, 1975; Shusterman, 1979), with decision
difficulty found to be specifically associated with post-abortion guilt (Osofsky &
Osofsky, 1972), anxiety (Bracken, 1978), and negative emotions such as regret,
depression, and anger (Adler, 1975). In particular, when ambivalence regarding the
decision to abort is rooted in some pregnancy intendedness or desire to have the
child (Ashton, 1980; Friedman et al., 1974; Lazarus, 1985; Lyndon et al., 1996;
Major et al., 1985; Miller, 1992; Remennick & Segal, 2001), and/or feelings
of pressure or coercion by one’s partner (Lemkau, 1991; Miller, 1992), women
are more prone to regret their decisions and experience postabortion emotional
difficulties.
There is considerable evidence indicating that the choice to abort is
often instigated by partners and men frequently play a primary role in women’s final
decisions (Lieh-Mak, Tam & Ng, 1979; Walter, 1970; Zimmerman, 1977). Delay of
an abortion decision beyond the first trimester is likely to be a marker for ambivalence
and as noted previously, women who have an abortion during the second trimester
have been found to exhibit more post-abortion adjustment problems. Osofsky et al.
(1973) found that 51% of women who had a second trimester abortion reported
decision difficulty compared to only 12% of women who had a first trimester
abortion. As recently pointed out by Kero et al. (2001), very few studies have offered
an in-depth analysis of ambivalent abortion decisions. Further, given the centrality of
this predictor, more attention should be devoted to examining the quality of
decisions and post-abortion reactions among women with diverse backgrounds, characteristics,
and abortion-related circumstances. An association between decision
ambivalence and ethnicity indicating that Black women tend to be more ambivalent
than White women has been reported (Faria et al., 1985).
...As Mathews-Green (1994, p. 34)
suggests: ‘‘. . . some hold to their right to regulate reproduction so strongly that the
sudden intrusion of motherhood is often perceived as a complete loss of control
over their present and future selves, and this can paralyze their ability to think more
rationally and realistically.’’ Consistent with this idea, Allanson and Astbury (1995)
found that the most common argument offered for an abortion was that continuing
the pregnancy would jeopardize one’s future.
Based on the conceptualization of abortion as a period of personal crisis for
many women, Landy (1986) observed that decision-making abilities may indeed be
temporarily compromised. Specifically, she described four types of faulty thinking
frequently observed in abortion clinics: (1) the ‘‘spontaneous approach’’ in which
the decision is made rapidly without sufficient time given to explore the options
and examine possible conflicting feelings; (2) the ‘‘rational-analytic approach’’
which emphasizes practical reasons for pregnancy termination (finances, single
parenthood, etc.) and excludes emotional considerations such as attachment to the
pregnancy; (3) the ‘‘denying-procrastinating approach’’ which involves avoidance of
decision-making due to internal conflict pertaining to continuing versus terminating
the pregnancy with the likelihood of the conflicts remaining as time pressure
necessitates a decision; and (4) the ‘‘no-decision making approach’’ characterized
by the woman deferring to others to make the decision (partner, parents, a health
care professional, etc.). Any of these patterns may result in lower levels of satisfaction
postabortion and may precipitate problematic adjustment. In a study of coping
strategies, Cohen and Roth (1984), found that women who used denial or avoidance
as a means for coping with an abortion reported higher levels of post-abortion anxiety
and depression than women who did not adopt such a strategy. Further, those who
engaged in approach strategies characterized by behaviors such as contemplating
the procedure and discussing the decision with others experienced greater decreases in
anxiety from before to after the abortion when compared to women who did not use
such direct means of coping. The results of this study suggest the importance of
encouraging women who are considering an abortion to thoughtfully work through
the decision and to reach out to others as they explore the pros and cons...
There is evidence to indicate that women confronted with an abortion decision are
likely to engage in distorted thinking possibly due to conflicts between their desire to
go through with the abortion and personal beliefs tied to moral issues. For example, in
a qualitative study by Simonds, Ellertson, Springer and Winikoff (1998) designed to
examine how medical abortion methods affect private experiences of abortion, it was
common for women to view a medical abortion in a way that distorts the reality of the
procedure. Many women described the medically induced abortion as ‘‘more natural’’,
‘‘like menstruation’’, ‘‘more humane’’, and ‘‘less bad’’ than surgical abortion.
The authors noted that the women’s references to abortion as similar to severe menstrual
cramps suggested a distorted or wishful conceptualization of the process as
‘‘not-really-abortion’’ but as a late period that finally arrives. Further, a study by
Foster and Sprinthall (1992) revealed that adolescent and young women’s level of
reasoning associated with abortion decision-making was significantly lower than their
general reasoning abilities. When a decision involves a violation of one’s conscience or
belief system, which appears to be rather common in the case of abortion as evidenced
by high levels of guilt reported in the literature (reviewed above), particularly among
adolescent women who abort (Martin, 1973; Perez-Reyes & Falk, 1973), regression in
cognitive functioning may represent a way of coping with the decision difficulty.
Sadly, however, after the stressfulness of the decision and the abortion are over,
women’s abilities to distort their experience or rationalize their behavior may
become decidedly more challenging as cognitive abilities return to a normal level...
Unfortunately, many women who make the decision to abort do so without a
thorough understanding of the procedure and research suggests that feelings of
having been misinformed or denied relevant information are related to post-abortion
difficulties (Congleton & Calhoun, 1993; Vaughan, 1991; Franz & Reardon, 1992).
Making accurate information pertaining to fetal development available to women,
particularly those who request it, should help to ensure that women feel that they
had adequate knowledge to arrive at a decision that is consistent with their beliefs
and value systems. Avoiding discussion of fetal development or using terms like
‘‘tissue’’ or ‘‘a clump of cells’’ to refer to a fetus that is 6-weeks-old or older when
counseling women seeking information is undoubtedly viewed by many health care
providers as helpful, because it keeps the decision simple and focused on what the
woman desires for her life. However, others may see this practice as denying
women the respect they deserve and as somewhat deceitful because it obscures
women’s right to make a fully informed decision...

Domestic violence and abortion
Many studies indicate that partner relationship problems are among the most
common motives for seeking an abortion (Torres & Forrest, 1988; Russo, Horn &
Schwartz, 1992; Soderberg et al., 1997), with the experiences of partner sexual assault
frequently found to factor into the choice to abort (Allanson & Astbury, 2001; Borins
& Forsythe, 1985; Russo & Pope, 1993). A woman who is a victim of domestic
violence may choose to abort for various reasons related to the abuse: (1) because the
current or past pregnancies precipitated increased violence, (2) due to fear that the
fetus will be harmed by violence, (3) due to coercion from an abuser, (4) because
the pregnancy was the result of rape, or (5) based on a lack of personal interest in
and/or fears regarding the prospect of having a child with an abuser (Coleman &
Maxey, 2004). Although extensive exploration of abortion as a risk factor for domestic
violence has not occurred, Hedin and Janson (2000) did report an association between
abortion and violence during a subsequent pregnancy.
The psychology of abortion 263
In addition to operating as a predictor of the choice to abort and as a possible
negative outcome of abortion, relationship violence has been identified as a risk
factor for negative post-abortion adjustment (Allanson & Astbury, 2001; Llewellyn
& Pytches, 1988; Soderberg et al., 1998; Russo & Denious, 2001). Adding to the
complexity of relations between partner violence and abortion decision-making and
adjustment is the overlap in possible negative outcomes including anxiety, depression,
and substance use/abuse among victims of partner violence (Burnam et al., 1988;
Goodman et al., 1993a,b; Koss, Koss & Woodruff, 1991) and among women
who have had an abortion (Coleman & Nelson, 1998; Coleman et al., 2002a;
Cougle et al., 2003; Drower & Nash, 1978; Franco et al., 1989; Gould, 1980;
Reardon & Ney, 2000; Reardon & Cougle, 2002a,b; Thorp et al., 2003;
Yamaguchi & Kandel, 1987). Finally, the literature on partner violence and abortion
is complicated by the fact that both partner violence and unwanted pregnancy are
more common among women with particular sociodemographic characteristics
including poverty, low levels of formal education, and being unmarried (Adams,
1985; Amaro et al., 1990; Miller, 1992; Russo, 1992; Williams & Pratt, 1990).
More research is clearly needed to examine the rather convoluted associations
among partner violence, abortion, and mental health, with sensitivity to the sociodemographic
context within which abortion decisions are made.

Conclusion
For various political, social, and ideological reasons, the psychology of abortion has
probably not received the amount of concentrated scholarly attention that a topic,
which touches the lives of so many contemporary women, deserves. Moreover,
the body of work that has accumulated throughout the world has proceeded in a
predominantly atheoretical manner and has been plagued by numerous methodological
shortcomings as well as content gaps...As theoretically driven, methodologically sound assessments are
conducted in the years to come, the information gathered should lead to meaningful
insights pertaining to abortion decision-making and adjustment among women with
widely varying characteristics contemplating abortion under diverse circumstances.
The need for a large nationally representative, longitudinal study of women faced
with an unintended pregnancy has been voiced repeatedly by researchers (e.g.,
Cougle et al., 2003; Speckard & Rue, 1992; Thorp et al., 2003) and is further conveyed
by the work reviewed herein...Strong emotions
infiltrating the academic study of this topic render the conduct of research that is free
from moral, political, and philosophical biases a difficult, perhaps unattainable goal.
However, in the interest of the millions of women who undergo one of the most
common surgical procedures currently available in the United States and elsewhere
throughout the world, it is evident that more probing and substantive research
should be conducted. Such research will continue to be the target of political attacks.
However, the comments by the editors of the Canadian Medical Association Journal,
(2003) in response to readers’ criticisms of their decision to publish a study linking
abortion and psychiatric hospitalization offers an appropriate reminder that scientists
must continually investigate the risks and benefits of one of the most politically
charged medical procedures: ‘‘This debate is conducted publicly in religious,
ideological and political terms: forms of discourse in which detachment is rare. But
we do seem to have the idea in medicine that science offers us a more dispassionate
means of analysis. To consider abortion as a health issue, indeed as a medical
‘‘procedure,’’ is to remove it from metaphysical and moral argument and to place it
in a pragmatic realm where one deals in terms such as safety, equity of access,
outcomes and risk–benefit ratios, and where the prevailing ethical discourse, when
it is evoked, uses secular words like autonomy and patient choice’’ (p. 93).

References
[1] Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. (1978). Learned helplessness in humans: Critique and
reformulation. Journal of Abnormal Psychology, 87, 49–74.
[2] Adams, H. P. J. (1985). Physical abuse during pregnancy. Obstetrics and Gynecology, 66, 185–190.
[3] Adler, N. E. (1975). Emotional responses of women following therapeutic abortion: how great a problem? Journal of Applied Social Psychology, 6, 240–259.
[4] Adler, N. E., David, H. P., Major, B. N., Roth, S. H., Russo, N. F., & Wyatt, G. E. (1990). Psychological responses after abortion. Science, 248, 41–44.
[5] Adler, N. F., & Dolcini, P. (1986). Psychological issues in abortion for adolescents. In: Melton, G. (Ed.), Adolescent abortion: Psychological and legal issues (pp. 74–95). Lincoln: University of Nebraska Press.
[6] Alan Guttmacher Institute (1996). Facts at a glance: Induced abortion. Author, New York.
Alan Guttmacher Institute (2000). The Status of Major Abortion Laws in the States (on-line). Available at http://www.agi-usa.org.
[7] Allanson, S., & Astbury, J. (1995). The abortion decision: Reasons and ambivalence. Journal of Psychosomatic Obstetrics and Gynecology, 16, 123–136.
[8] Allanson, S., & Astbury, J. (2001). Attachment style and broken attachments: Violence, pregnancy, and abortion. Australian Journal of Psychology, 53, 146–151.
[9] Amaro, H., Fried, L. W., Cabral, H., & Zuckerman, B. (1990). Violence during pregnancy and substance use. American Journal of Public Health, 80, 575–579.
[10] Anthanasiou, R., Oppel, W.,Michelson, L., Unger, T.,&Yager, M. (1973). Psychiatric sequelae to term birth and induced early and late abortion: A longitudinal study. Family Planning Perspectives, 5, 227–231.
[11] Appleby, L. (1991). Suicide during pregnancy and in the first postnatal year. British Medical Journal, 302, 137–140.
[12] Appleby, L., & Turnbull, G. (1995). Parasuicide in the first postnatal year. Psychological Medicine, 25, 1087–1090.
[13] Armsworth, M. (1991). Psychological response to abortion. Journal of Counseling and Development, 69, 377–379.
[14] Array, J. (1968). Aborto: estudios psicoanalitico. Paidos, Buenos Aires, Ediciones Horme.
[15] Ashton, J. (1980). The psychosocial outcome of induced abortion. British Journal of Obstetrics and Gynecology, 87, 1115–1122.
[16] Baetsen, K. L., Rankin, R. E., Fuller, G. B and Stack, J. M. (1985). A comparative MMPI study of abortion-seeking women and those who intend to carry their pregnancies to term. Family Practice Research Journal, 4, 199–207.
[17] Badgley, R. F., Caron, D. F., & Powell, M. G. (1977). Report of the committee on abortion law (pp. 313–321). Ottawa: Supply and Services.
[18] Bandura, A. (1982). Self-efficacy in human agency. American Psychologist, 37, 122–147.
[19] Bandura, A. (1989). Regulation of cognitive processes through perceived self-efficacy. Developmental Psychology, 25, 729–735.
[20] Bandura, A. (2002). Social cognitive theory in cultural context. Applied Psychology: An International Review, 51, 269–290.
[21] Barnard, C. (1990). The long-term psychosocial effects of abortion. Institute for Pregnancy Loss, Jacksonville, Florida.
[22] Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182.
[23] Beck, C. T. (2001). Predictors of postpartum depression: an update. Nursing Research, 50, 275–285.
[24] Benedict, M. I., White, R. B., & Cornely, D. A. (1985). Maternal perinatal risk factors and child abuse. Child Abuse and Neglect, 9, 217–224.
[25] Berger, C., Gold, D., Andres, D., Gillett, P., & Kinch, R. (1984). Repeat abortion: Is it a problem? Family Planning Perspectives, 16, 70–75.
[26] Blazer, D. G., Kessler, R. C., McGonagle, K. A., & Swartz, M. S. (1994). The prevalence and distribution of major depression in a national community sample: The National Comorbidity Survey. American Journal of Psychiatry, 151, 979–986.
[27] Bogen, I. (1974). Attitudes of women who have had abortions. Journal of Sex Research, 10, 97–109.
[28] Borins, E. F. M., & Forsythe, P. J. (1985). Past Trauma and present functioning of patients attending a women’s psychiatric clinic. American Journal of Psychiatry, 142, 460.
[29] Bracken, M. B. (1978). A causal model of psychosomatic reactions to vacuum aspiration abortion. Social Psychiatry, 13, 135–145.
[30] Bracken, M. B., Hachamovitch, M., & Grossman, G. (1974). The decision to abort and psychological sequelae. Journal of Nervous and Mental Disease, 158, 155–161.
[31] Bracken, M. B., & Kasi, S. (1975). First and repeat abortions: A study of decision-making and delay. Journal of Biosocial Science, 7, 473–491.
[32] Bracken, M. B., Klerman, L. V., & Bracken, M. (1978). Coping with pregnancy resolution among never-married women. Journal of Orthopsychiatry, 48, 320–334.
[33] Bracken, M. B., & Swigar, M. E. (1972). Factors associated with delay in seeking induced abortions: A review and theoretical analysis. American Journal of Obstetrics and Gynecology, 121, 1008–1019.
[34] Bradley, C. F. (1984) Abortion and subsequent pregnancy, Canadian Journal of Psychiatry, 29, 494.
[35] Broggess, S., & Bradner, C. (2000). Trends in adolescent males’ abortion attitudes, 1988–1995: Differences by race and ethnicity. Family Planning Perspectives, 32, 118–123.
[36] Brett, W., & Brett, A. (1992). Outcome and management of crisis pregnancy counseling. New Zealand Medical Journal, 105, 7–9.
[37] Bronfenbrenner, U. (1979). The ecology of human development. Cambridge Mass: Harvard University Press.
[38] Bronfenbrenner, U. (1986). The ecology of the family as a context for human development: research perspectives. Developmental Psychology, 22, 723–742.
[39] Bronfenbrenner, U. (1989). Ecological systems theory. In: Vasta, R. (Ed.), Annals of Child Development, Vol. 6 (pp. 187–249). Greenwich CT: JAI Press.
[40] Bronfenbrenner, U. (1993). The ecology of cognitive development: Research models and fugitive findings. In: Wozniak, R., & Fischer, K. (Eds.), Development in context: Acting and thinking in specific environments (pp. 3–44). Hillsdale, NJ: Erlbaum.
[41] Brofenbrenner, U. (1998). Growing chaos in the lives of children, youth and families: How can we turn it around? Parenthood in America. Conference Proceedings. University of Wisconsin-Madison General Library System, Madison, WI.
[42] Burnam, M. A., Stein, J. A., Golding, J. M., Siegel, J. M., Sorenson, S. B., Forsythe, A. B., & Telles, C. A. (1988). Sexual assault and mental disorders in a community population. Journal of Consulting and Clinical Psychology, 56, 843–850.
[43] Burnell, G. M., & Norfleet, M. A. (1987). Women’s self-reported responses to abortion. The Journal of Psychology, 121, 71–76.
[44] Butlet, C. (1996). Late psychological sequelae of abortion: Questions from a primary care physician. Journal of Family Practice, 43, 396–401.
[45] Campbell, N., Franco, K., & Jurs, S. (1988). Abortion in adolescence. Adolescence, 23, 813–823.
[46] Cates, W., & Grimes, D. A. (1981). Deaths from second trimester abortion by dilatation and evacuation: causes, prevention, facilities. Obstetrics and Gynecology, 58, 401–408.
[47] Centers for Disease Control and Prevention (1994). Abortion surveillance: preliminary data – United States, 1992. Morbidity and Mortality Weekly Report, 43, 930–939.
[48] Cohen, T., & Roth, S. (1984). Coping with abortion. Journal of Human Stress, 10, 140–145.
[49] Coleman, P. K., & Maxey, C. D. (2004). The choice to abort among mothers living under ecologically deprived conditions: predictors and consequences. Manuscript submitted for possible publication in Family Relations.
[50] Coleman, P. K., & Nelson, E. S. (1998). The quality of abortion decisions and college students’ reports of post-abortion emotional sequelae and abortion attitudes. Journal of Social and Clinical Psychology, 17, 425–442.
[51] Coleman, P. K., Reardon, D. C., & Cougle, J. (2002). The quality of the caregiving environment and child developmental outcomes associated with maternal history of abortion using the NLSY data. Journal of Child Psychology and Psychiatry and Allied Disciplines, 43, 743–758.
[52] Coleman, P. K., Reardon, D. C., Rue, V., & Cougle, J. (2002a). Prior history of induced abortion and substance use during pregnancy. American Journal of Obstetrics and Gynecology, 187, 1673–1678.
[53] Coleman, P. K., Reardon, D. C., Rue, V., & Cougle, J. (2002b). State–funded abortions vs. deliveries: A comparison of outpatient mental health claims over four years. American Journal of Orthopsychiatry, 72, 141–152.
[54] Colman, A., & Colman, L. (1971). Pregnancy: The psychological experience.
[55] Condon, J. T. (1986). The spectrum of fetal abuse in pregnant women. Journal of Nervous and Mental Disease, 174, 509–516.
[56] Congleton, G., & Calhoun, L. (1993). Post-abortion perceptions: A comparison of self-identified distressed and non-distressed populations. International Journal of Social Psychiatry, 39, 255–265.
[57] Conklin, M., & O’Connor, B. (1995). Beliefs about the fetus as a moderator of postabortion psychological well-being. Journal of Social Psychiatry, 39, 76–81.
[58] Cougle, J., Reardon, D. C., & Coleman, P. K. (2003). Depression associated with abortion and childbirth: A long-term analysis of the NLSY cohort. Medical Science Monitor, 9, CR105–112.
[59] Cozzarelli, C. (1993). Personality and self-efficacy as predictors of coping with abortion. Journal of Personality and Social Psychology, 65, 1224–1236.
[60] Cozzarelli, C., Karrasch, A., Sumer, N., & Major, B. (1994). The meaning and impact of partner’s accompaniment on women’s adjustment to abortion. Journal of Applied Social Psychology, 24, 2028–2056.
[61] Cozzarelli, C., Major, B., Karrasch, A., & Fuegen, K. (2000). Women’s experiences of and reactions to antiabortion picketing. Basic and Applied Social Psychology, 22, 265–275.
[62] Cozzarelli, C., Sumer, N., & Major, B. (1998). Mental models of attachment and coping with abortion. Journal of Personality and Social Psychology, 74, 453–467.
[63] Cranley, M. S. (1981). Development of a tool for measurement of maternal attachment during pregnancy. Nursing Research, 30, 281–284.
[64] Canadian Medical Association Journal (2003). Unwanted results: The ethics of controversial research, Canadian Medical Association Journal, 169, 93.
[65] Dagg, P. (1991). The psychological sequelae of therapeutic abortion. American Journal of Psychiatry, 5, 578–585.
[66] David, H., Rasmussen, N., & Holst, E. (1981). Post-abortion and postpartum psychotic reactions. Family Planning Perspectives, 13, 88–91.
[67] De Veber, L., Ajzenstat, J., & Chisholm, D. (1991). Postabortion grief: Psychological sequelae of induced abortion. Humane Medicine, 7, 203–209.
[68] Drower, S. A., & Nash, E. S. (1978). Therapeutic abortion on psychiatric grounds. South Africa Medical Journal, 54, 604–608.
[69] Dugger, K. (1998). Black women and the question of abortion. In: Beckman, L. J., & Harvey, S. M. (Eds.), The new civil war: The psychology, culture, and politics of abortion (pp. 107–131). Washington, DC: American Psychological Association.
[70] Evans, A. (2001). The influence of significant others on Australian teenagers’ decisions about pregnancy resolution. Family Planning Perspectives, 33, 224–230.
[71] Faria, G., Barrett, E., & Goodman, L. M. (1985). Women and abortion: Attitudes, social networks, decision-making. Social Work in Health Care, 11, 85–99.
[72] Fisher, S. (1986). Reflections on repeated abortions: The meanings and motivations. Journal of Social Work Practice, 2, 70–87.
[73] Fleming, A. S., Ruble, D. N., Flett, G. L., & Van Wagner, V. (1990). Adjustment in first-time mothers: Changes in mood and mood content during the early postpartum months. Developmental Psychology, 26, 137–143.
[74] Forrest, J. D. (1994). Preventing unwanted pregnancy: The role of hormonal contraceptives. American Journal of Obstetrics and Gynecology, 170, 485–1489.
[75] Foster, V., & Sprinthall, N. A. (1992). Developmental profiles of adolescents and young adults choosing abortion: Stage sequence, decalage, and implications for policy. Adolescence, 27, 655–673.
[76] Franco, K. N, Tamburrino, M., Campbell, N., Pentz, J., & Jurs, S. (1989). Psychological profile of dysphoric women post abortion. Journal of the American Medical Women’s Association, 44, 113.
[77] Franz, W., & Reardon, D. (1992). Differential impact of abortion on adolescents and adults. Adolescence, 27, 161–172.
[78] Freeman, E. (1980). Emotional distress patterns among women having first or repeat abortions. Obstetrics and Gynecology, 55, 630–636.
[79] Friedman, C., Greenspan, R., & Mittleman, F. (1974). The decision-making process and the outcome of therapeutic abortion. American Journal of Psychiatry, 131, 1332–1337.
[80] Fu, H., Darroch, J. E., Henshaw, S. K., & Kolb, E. (1998). Measuring the extent of abortion underreporting in the 1995 National Survey of Family Growth. Family Planning Perspectives, 30, 133–138.
[81] Gilchrist, A. C., Hannaford, P. C., Frank, P., & Kay, C. R. (1995). Termination of pregnancy and psychiatric morbidity. British Journal of Psychology, 167, 243–248.
[82] Gilligan, C. (1982). In a different voice. Cambridge, MA: Harvard University Press.
[83] Gissler, M., Kauppila, R., Merilainen, J., Toukomaa, H., & Hemminki, E. (1997). Pregnancy-associated deaths in Finland 1987–1994 – definition problems and benefits of record linkage. Acta Obstetricia et Gynecologica Scandinavica, 76, 651–657.
[84] Glander, S. S., Moore, M. L., Michielutte, R., & Parsons, L. H. (1998). The prevalence of domestic violence among women seeking abortion. Obstetrics and Gynecology, 91, 1002–1006.
[85] Gold, D., Berger, C., & Andres, D. (1979). The abortion choice: Psychological determinants and consequences. Concordia University, Department of Psychology, Montreal.
[86] Goodman, L. A., Koss, M. P., & Russo, N. F. (1993a). Violence against women: Physical and mental health effects Part I: Research findings. Applied and Preventive Psychology: Current Scientific Perspectives, 2, 79–89.
[87] Goodman, L. A., Koss, M. P., & Russo, N. F. (1993b). Violence against women: Physical and mental health effects, Part II: Conceptualizing post-traumatic stress. Applied and Preventive Psychology: Current Scientific Perspectives, 2, 123–130.
[88] Gould, N. B. (1980). Postabortion depressive reactions in college women. Journal of American College Health Association, 28, 316–320.
[89] Grimes, D. A., Bachicha, J. A., & Learman, L. A. (1998). Teaching critical appraisal to medical students in obstetrics and gynecology. Obstetrics and Gynecology, 92, 877–881.
[90] Hall, E. J., & Marx-Ferree, M. (1986). Race differences in abortion attitudes. Public Opinion Quarterly, 50, 193–207.
[91] Handy, J. A. (1982). Psychological and social aspects of induced abortion. British Journal of Clinical Psychology, 21, 29–41.
[92] Hedin, L. W., & Janson, P. O. (2000). Domestic violence during pregnancy: The preva