Data on late-term abortions in the United States

by Wm. Robert Johnston
last updated 14 April 2019

Abstract: Available data are sufficient to refute some claims regarding late-term abortions in the United States ("late-term" abortions here referring to abortions at >20 weeks' gestation). Here we use official data (from the Centers for Disease Control and state health agencies) to develop estimates of such abortions for 2005 to 2018. For the year 2018, best estimates (and plausible ranges in parentheses) for such abortions are: 11,500 (9,100 to 15,400) at >20 weeks' gestation; 900 (400 to 1,600) at >24 weeks' gestation; and 160 (50 to 260) at >28 weeks' gestation. There were significant changes in the top states for percentages of late abortions in 2005-2018, reflecting relocations of late-term abortion "services". Based on information reported by Arizona, Florida, and Utah, probably 30-80% of late-term abortions are "elective" (non-health related), 20-60% are in cases of fetal health issues, and only 3-10% are in cases of maternal health. These figures are generally consistent with those reported both in Guttmacher Institute surveys and in past testimony by individuals who perform such abortions. Estimated current annual revenue to the abortion industry from late-term abortions is about $20,000,000 ($12,000,000 to $40,000,000). Abortion on demand (i.e. for elective reasons) at any point in pregnancy through birth is currently permitted in 7 states, placing the United States among only four countries in the world with policies this extreme. However, since abortion practice in some countries is inconsistent with their respective laws, actual levels of late-term abortions are comparable between the United States and a number of developed countries.


In early 2019 several states enacted or proposed legislative changes to expand late-term abortions, renewing attention on this practice in the United States. Discussion of such policies has included some confusion regarding data bearing on the subject, due in part to misleading claims by the abortion lobby. What follows is a review of the best available data on such abortions, addressing nationwide incidence, variation over time and between states, and reasons for such abortions. Note that many of these topics have previously been well addressed by various publications of the Charlotte Lozier Institute, e.g. Prentice et al. (2019).

In this review, late-term abortion refers to abortions at greater than 20 weeks' gestation. Some state restrictions on late-term abortions are in terms of probable post-fertilization (PPF) age. PPF age is typically 2 weeks less than gestational age (gestational age is measured from the last menstral cycle).

Despite the significant limitations in data, there is still meaningful information on late-term abortions as reviewed in this report. Where assumptions are involved, these will be stated and, particularly where large uncertainties result, best estimates will be accompanied with plausible ranges.

Abortions at greater than 20 weeks' gestation

Abortion data is inconsistently reported in the United States, particularly with regard to late-term abortions. The U.S. Centers for Disease Control (CDC) publishes annual reports on abortion based on whatever data are volunteered by the states. During 1997-2015, 42 states (plus D.C.) reported to the CDC every year and two states never reported. During 2005-2015, the number of states from which data by gestational age were available varied from 37 to 42. Most of the CDC data correspond to data directly published by the respective state health agencies. Some states do not publish the data they report to the CDC. Limited data are reported directly from a few states but not included in the CDC reports (generally due to untimely reporting or incomplete nature of data).

The Guttmacher Institute (GI) publishes estimates of U.S. abortions at intervals of 3-4 years (e.g., Jones and Jerman, 2017a). Their estimates are based on direct surveys of abortion providers plus official data. Since 2005 about 30% of abortions estimated by GI went unreported to the CDC, and 20-25% were unreported in any official form. GI data by gestational age are very limited. It should be noted that the incompleteness of official abortion data results in part from the opposition of abortion lobbies to data collection and reporting--in fact, in some states abortion providers frequently fail to abide by regulatory reporting requirements. Due to lack of monitoring in some states, accuracy of some data (such as gestational ages) may be in question.

From 2005 to 2014, about 55% of abortions estimated by GI were reported by the CDC as having known gestational age (or age range). Averaged over this time period, the CDC reports 1.30% of abortions with known gestational age were at greater than 20 weeks. Many discussions of late-term abortions assume this percentage applies to the other 45% of U.S. abortions. This is consistent with GI statements that 1.3% of abortions were at greater than 20 weeks' gestation, although examination of GI and Planned Parenthood (PP) reports suggests that they have merely adopted the CDC figure (PP, 2015).

Note that if the sample of abortions with known gestational age are not representative of all abortions, the 1.3% figure would not apply to all abortions. Examples of causes for relevant bias are:

While abortion proponents have claimed that recent state restrictions on abortion have increased the fraction of abortions that are late-term, this claim is not supported by the data. The blue line in the figure below shows CDC figures for fraction of late-term abortions for 1972-2015. For abortions reported to the CDC with known gestational age range, the fraction at greater than 20 weeks averaged 0.95% in 1970-1990, rose rapidly in 1990-1992, averaged 1.41% in 1992-2004, then declined slightly to an average of 1.30% in 2005-2015. These figures represent those states reporting to the CDC which are not consistent throughout this time period. The other lines show results each for a set of 23 states that reported consistently for limited time periods--they show generally similar trends, tending to affirm that the overall U.S. trends do not depend heavily on the changes in reporting states. Nonetheless, the actual national percentage of late-term abortions may differ somewhat from the CDC figures given that a large fraction of U.S. abortions occur in states not reporting to the CDC (particularly California). Some of the short timescale changes, such as the peaks in 1992 and 2012, are linked to changes in individual states' reporting levels as will be discussed later.

Here, we have used available state-level data from both the CDC and state health agencies to produce state-level estimates of late-term abortions, with totals for the United States as well. Since little or no data are available on abortions by gestational age for five states (California, Florida, Massachusetts, Maryland, and New Hampshire) which account for 30% of total U.S. abortions, the overall U.S. figures are uncertain. For selected years, estimates (with plausible ranges) are:

The figure below shows these estimates (with plausible ranges) for 2005-2018. The grey line shows reported numbers of late-term abortions (these numbers decline in 2016-2017 because data are currently available for only a limited number of states).

The table below gives estimates for each state for 2008 and 2015. Sources and methodology are given after the table. For most states, reported percentages of late-term abortions have been applied to the total states' abortions (which usually includes estimated as well as reported abortions). Figures indicated as approximate (~) are based on assumed percentages of late-term abortions as described in the methodology.

all abortionsabortions
at >20 wks
% of abortions
at >20 wks
all abortionsabortions
at >20 wk
% of abortions
at >20 wks
Alabama 11,268 10 0.09 5,899 40 0.76
Alaska 1,759 14 0.80 1,334 1 0.08
Arizona 19,500 150 0.77 12,655 160 1.23
Arkansas 4,890 0 0.00 4,070 40 0.93
California 214,190~4,070~1.90152,100~2,890~1.90
Colorado 15,960 140 0.85 13,020 380 2.96
Connecticut 17,030 ~50~0.30 12,310 ~40~0.30
Delaware 7,070 80 1.05 2,420 6 0.25
District of Columbia 4,450 ~0~0.00 5,990 ~0~0.00
Florida 94,360 ~940~1.00 75,260 ~750~1.00
Georgia 39,820 1,340 3.37 29,260 690 2.35
Hawaii 5,630 40 0.68 3,470 50 1.55
Idaho 1,800 4 0.20 1,272 0 0.00
Illinois 54,920 1,060 1.94 43,790 220 0.50
Indiana 10,999 0 0.00 8,020 6 0.08
Iowa 6,560 6 0.09 4,350 10 0.23
Kansas 10,642 380 3.61 6,974 40 0.65
Kentucky 4,430 60 1.32 3,420 50 1.44
Louisiana 14,860 280 1.88 9,311 17 0.18
Maine 2,800 ~1~0.04 2,020 ~4~0.22
Maryland 34,290 ~450~1.30 27,120 ~350~1.30
Massachusetts 24,900 ~350~1.40 20,670 ~290~1.40
Michigan 36,790 340 0.93 28,620 450 1.59
Minnesota 13,060 60 0.49 9,861 120 1.24
Mississippi 2,772 2 0.08 2,330 0 0.00
Missouri 7,440 34 0.46 4,830 80 1.62
Montana 2,230 11 0.47 1,611 8 0.50
Nebraska 2,840 9 0.30 2,004 0 0.00
Nevada 13,450 ~80~0.58 11,240 50 0.44
New Hampshire 3,200 ~40~1.30 2,540 ~33~1.30
New Jersey 54,160 1,580 2.92 44,170 1,240 2.80
New Mexico 6,150 34 0.55 3,350 250 7.44
New York 153,110 3,470 2.26107,430 2,510 2.33
North Carolina 33,140 3 0.01 33,640 21 0.06
North Dakota 1,400 ~1~0.07 1,166 0 0.00
Ohio 33,550 550 1.63 22,500 160 0.69
Oklahoma 7,160 10 0.14 4,690 ~1~0.02
Oregon 12,920 210 1.61 9,760 170 1.73
Pennsylvania 41,000 370 0.90 31,818 330 1.04
Rhode Island 5,000 8 0.16 3,510 13 0.38
South Carolina 7,300 8 0.11 6,870 ~5~0.07
South Dakota 848 8 0.94 444 ~2~0.45
Tennessee 19,550 9 0.04 12,800 15 0.12
Texas 84,610 570 0.68 55,287 250 0.46
Utah 4,000 4 0.10 3,176 30 0.96
Vermont 1,510 ~4~0.27 1,430 10 0.71
Virginia 28,698 29 0.10 19,490 80 0.41
Washington 24,320 420 1.72 18,570 350 1.89
West Virginia 2,280 6 0.26 1,770 ~5~0.26
Wisconsin 8,229 200 2.41 5,900 ~70~1.18
Wyoming 90 ~0~0.00 120 ~0~0.00
United States1,212,900~17,500~1.44896,000~12,300~1.37

Methodology: Where the CDC reported the percentage of abortions with known gestational age that occurred at >20 weeks, those percentages were assumed to apply to abortions of unknown gestational age (this included abortions reported in CDC reports and those otherwise estimated, either by GI or by the author). For some states and years, CDC percentages were not provided due to small numbers (<5) in which case numbers were estimated by the author (note that the CDC data still provide upper limits in these cases). In some cases missing data were available directly from state reports. For states and years without direct constraints on percentage, the average percentage from the closest three years of available data were applied. This left several states with no data on percentage. These states with median assumed percentage (and basis) are: California (1.9%, average percentage for Washington and Oregon, given the large number of late-term abortion providers in California), Florida (1.0%, based on modeling using reported gestational age data), Maryland (1.3%, U.S. average), and New Hampshire (1.3%, U.S. average). Plausible ranges are considered to range from 0.6% (median across reporting states) to 3.0% (average value for New Jersey). CDC data are from Gamble et al. (2005); Jatlaoui et al. (2016, 2018); Jones et al. (2017); Pazol et al. (2009, 2011a, 2011b, 2012, 2013, 2014, 2015). State agency data are from the respective health or statistics agencies of Alabama, Alaska, Arizona, Arkansas, Delaware, Florida, Hawaii, Idaho, Illinois, Indiana, Iowa, Kansas, Louisiana, Maine, Michigan, Minnesota, Mississippi, Missouri, Nebraska, New York, North Carolina, Ohio, Oregon, Pennsylvania, Texas, Utah, Vermont, Washington, West Virginia, and Wisconsin, as cited in Johnston (2018). Estimates of total U.S. abortions are from Johnston (2017, 2019a) which in turn are based on GI estimates for 2005-2014 from Jones and Jerman (2017a) and sources therein supplemented by CDC and state health agency data.

Trends among states in abortions at greater than 20 weeks' gestation

Despite the relatively uniform late-term abortion fraction nationwide in the last 20 years, there is signficant variation in this fraction among states. In the figure below, states have been assigned to three groups, either high, medium, or low abortion percentage based on average percentage in recent decades. States are assigned to their respective groups throughout the timeseries (i.e., states do not move between groups during the time period shown). The groups used are: high (9 states--California, Colorado, Georgia, Kansas, New Jersey, New Mexico, New York, Oregon, and Washington), medium (12 states--Hawaii, Illinois, Kentucky, Louisiana, Massachusetts, Michigan, Minnesota, Missouri, Ohio, Pennsylvania, Texas, and Wisconsin), and low (remaining 27 states plus the District of Columbia, omitting Florida and New Hampshire due to lack of data). Only reported abortions are used for the calculated group percentages (i.e., no assumed state-level data).

Late-term abortion percentage was about 1.5% for the high group in 1982-1991, then abruptly rose to about 2.5% where it remained through 2015. Percentage for the low group has ranged from 0.2% to 0.5%, and percentage for the medium group has ranged from 0.5% to 1.4%. This illustrates the significant variation between states in late-term abortion practice. These variations are only partly due to policy differences; note that policy differences were less significant for most of the time period shown than in the last decade. Some of the variation reflects differences in abortion "services" among states (i.e., the presence or absence of abortionists that conduct late-term abortions). Variation resulting from this issue is exaggerated in the data because of signficant levels of interstate travel to obtain late-term abortions (most of the available data are by state of occurrence, not state of residence). Nonetheless, the variations are also partly due to broad differences in abortion practices by women in various states.

The figure below shows data for 1982 to 2015 on the fraction of abortions that are late-term for eight states (plus New York City) with high percentages, based on CDC and state data. In the past Kansas was a prime location for late-term abortions, having in the 1990s attained a late-term fraction of 9%. New Mexico has recently taken this role, reaching a late-term fraction of 7% in recent years.

Several one-year "spikes" of elevated late-term fraction are evident, such as for New Jersey in 1992, New Mexico in 2009, and Colorado in 2011. When reported late-term abortions and abortions with no reported gestational age are compared in successive years, several such cases are shown to result from likely underreporting of late-term abortions. This evidence points to the possibility that larger numbers of late-term abortions are going unreported in state and national data. This is also suspected in cases such as the District of Columbia, which permits abortion on demand throughout pregnancy but has no reported data from some facilities that conduct abortions.

The maps below show reported percentages of late-term abortions by state, averaged for reported years in each decade (1980s, 1990s, 2000s, 2010s). Many states have show significant changes in late-term abortion fraction. Underreporting is also an issue for some states. Several states with little or no data are likely among those with late-term fractions over 1% (California, Massachusetts, Florida). Other areas are suspected to have higher late-term fractions than reported, based on known deficiencies in reported data (Connecticut, District of Columbia, Maryland, Nebraska).

Abortions at later points in pregnancy

Recent CDC reports do not provide separate estimates of abortions for gestational ages beyond >20 weeks. However, several states report data on abortions at >24 weeks, and three states (Iowa, Michigan, and Minnesota) have reported data on abortions at >28 weeks. Although many states now restrict abortions after 20-26 weeks, nearly all still permit abortion up until birth to save the mother's life and/or for other narrow reasons; thus, late-term abortions can potentially occur in any state (particularly if those narrow reasons are abused).

For years with available data from 2005 to 2017, the average percentage of all abortions that were at >24 weeks were 0.53% in Iowa, 0.25% in Michigan, 0.05% in Alabama, and 0.02-0.03% in Florida, Idaho, Minnesota, Ohio, and Texas. Several states specifically reported zero abortions at >24 weeks for any years reported in this time period: Alaska, Indiana, and Pennsylvania.

The average percentage of all abortions that were at >28 weeks, for years with available data from 2005 to 2017, were 0.05% in Michigan and 0.003% in Iowa and Minnesota. Michigan reported an average of 12 abortions per year at >28 weeks for 2008-2017. Minnesota reported a total of 3 such abortions in 2010-2017: two at 29 weeks and one at 31 weeks. Iowa reported a total of 2 such abortions in 2005-2017. Alabama did not report data at >28 weeks but did report a total of 7 abortions at >26 weeks in 2010-2017. (Note that data used for Alabama were data for Alabama residents which include abortions performed in other states and reported to Alabama). Importantly, data are not available for states with clinics known to conduct third trimester abortions--this reflects the correlation between more permissive abortion policies and lower quality of abortion reporting (Donovan and Gonzales, 2016).

For selected years, estimates (with plausible ranges) of total United States abortions at >24 and >28 weeks' gestation are:

Abortions at >24 weeks' gestation:

Abortions at >28 weeks' gestation:

Methodology: State-level estimates of abortions at >24 and >28 weeks' gestation were constructed based on the above data/estimates for abortions at >20 weeks, plus any state-level data for later abortions, plus modeling of the fraction of later abortions. For the modeling, the limited available data suggest an exponential drop in the fraction of late-term abortions with increasing gestational age, and that the rate of this drop is more consistent between states than is the late-term abortion fraction. This exponential relationship was applied to estimate abortions at >24 and >28 weeks for states with no data on such abortions, with the range in reported rates of drop vs. gestational age used to produce plausible upper and lower bounds, and with some consideration of state-level abortion policies (e.g., higher levels of extremely late abortions in states with no limits on abortion through full term pregnancy). No abortions at >28 weeks' gestation were estimated for 2005-2018 for 27 states (plus D.C.).

The figures below show these estimates (with ranges) for 2005-2018.

Anecdotal evidence is also consistent with the preceding data regarding extremely late abortions. An estimated 17 to 42 women from Quebec obtain third trimester abortions in the United States annually (Fidelman, 2016). Some facilities that specialize in late-term abortions are well known to conduct abortions past 28 weeks' gestation. An Albuquerque, New Mexico, clinic openly offers abortions up to 28 weeks' gestation and beyond (Heild, 2016). This clinic was reportedly willing to schedule an abortion at 37 weeks (Texas Right to Life, 2017). Morana (2016) cites cases in multiple states of abortions at 30, 33, and 35 weeks' gestation. Sullenger (2019) lists one clinic each in Colorado, New Mexico, Maryland, California, and Ohio known to conduct abortions at any point in pregnancy (in contrast to any stated limits in their literature, or to respective state regulations).

The figure below places these late-term abortion figures in the context of U.S. abortions in general. It shows, for increasing gestational age, estimates of the number of abortions occurring later than that point in pregnancy. Few pregnancies are aborted prior to 4 weeks' gestation because the pregnancy is not yet identified. Nearly all abortions occur between 5 and 18 weeks' gestation, with declining fractions towards later in pregnancy. Abortions drop off faster after 20 weeks' gestation, due to regulations in some states as well as trends in abortion practice--late-term abortions are significantly more dangerous for women, and very few abortionists are willing to perform them. However, the numbers of late-term abortions are much more uncertain due to lack of reporting.

Reasons for obtaining late-term abortions

Abortion proponents tend to either imply or outright claim that few if any late-term abortions are elective. At the same time they do not support reporting of data that could verify these claims, and as a result there are extremely little data to bear on this question. But some data is available. Three states report some data on reasons given for obtaining abortions in cases of late-term abortions: Arizona, Florida, and Utah. (To be clear, the term "elective" is used just as it is used in the cited sources; in reality, virtually all of these abortions are by choice as will be seen when "health-related" abortions are examined closely.)

Arizona reported 714 abortions in 2012-2013 and 2015-2017 at >20 weeks' gestation. Of these, maternal medical conditions were cited for only 19 (3%) and fetal medical conditions for only 129 (18%). Thus 79% of these abortions were not health-related.

Utah reported 134 abortions in 2008-2017 at >20 weeks' gestation. Numbers (and percentage) by reason given were: maternal life, 8 (6%); abnormal fetus, 84 (63%); therapeutic, 35 (26%); and elective, 3 (2%). It should be noted that many or most of those for "therapeutic" reasons are not, in fact, for reasons of physical health, given that "therapeutic" reasons were claimed for 84% of abortions at any gestational period in this period in Utah, in stark contrast to other states (Johnston, 2016).

Florida reported 230 abortions in 2000-2007 at >24 weeks' gestation. Numbers (and percentage) by reason given were: personal choice, 89 (39%); maternal physical condition, 6 (3%); maternal mental condition, 15 (7%); abnormal fetus, 90 (39%); and other, 29 (13%).

A study by Foster and Kimport (2013) published in a GI journal surveyed 272 women in the U.S. who obtained abortions at greater than 19 weeks' gestation. This study was specifically limited to women not seeking abortions for reasons of "fetal anomaly or life endangerment". Relative to women who obtained earlier abortions, the women obtaining later abortions more frequently cited issues with finding and reaching an abortion facility and paying for the abortion, and to a lesser extent factors such as not recognizing they were pregnant, being undecided about getting an abortion, and disagreeing with the baby's father about the abortion. These factors are generally consistent with findings by Jones and Jerman (2017b) in a larger study (but only separating abortions at >12 weeks' gestation from earlier ones). The authors do acknowledge that "data suggest that most women seeking later terminations are not doing so for reasons of fetal anomaly or life endangerment," citing Foster et al. (2012) in this regard.

Foster et al. (2012) surveyed 5,109 women at a single clinic, including cases for 1,122 abortions sought at >19 weeks' gestation in 2008. Of these 1,122 abortions, only 7 (0.6%) were reported as pregnancies resulting from rape, and only 7 (0.6%) were cases where a fetal anomaly was reported as present. While there is likely a bias towards late abortions for elective reasons in this sample because the clinic attracts women seeking such abortions, the cited authors have nonetheless concluded that most abortions sought late in pregnancy are not for reasons of maternal or fetal health. Foster is also quoted in Barry (2018) stating that abortions in cases of fetal anomaly "make up a small minority of later abortion".

These results are consistent with testimony in the 1990s by individuals who performed partial-birth abortions (PBAs) (approximately corresponding to those abortions termed intact dilation and extraction), summarized by Johnston (2007). In an interview with American Medical News, M. Haskell stated that about 80% of the PBAs he performed were "purely elective", with the remaining 20% performed for genetic reasons (Gianelli, 1997, cited by Canady, 1996; Sprang and Neerhof, 1998). In testimony to Congress, J. McMahon reported that for about 2,000-2,100 PBAs he had performed, 1,183 (56%) were for fetal "flaws" or "indicators", 175 (9%) were for maternal "indicators", and the remainder (about 700, or 35%) were elective (National Right to Life, 1996; Canady, 1996; Sprang and Neerhof, 1998). McMahon further indicated that elective abortions comprised 20% of those he performed after 21 weeks gestation, and none of those he performed after 26 weeks (Canady, 1996).

McMahon's 1995 testimony to the House Judiciary Committee was analyzed by physicians P. Smith and K. Dowling. Among maternal indicators, the single most frequent was maternal depression (39, or 1.9% of total for all reasons), with 28 attributed to maternal health conditions "consistent with the birth of a normal child (e.g. sickle cell trait, prolapsed uterus, small pelvis)" (1.3% of total) and the remainder (5% of total) for other maternal factors ranging from maternal health risk to "spousal drug exposure" and "substance abuse". Those performed for fetal indicators included cases ranging from conditions involving risk to the baby's life to lesser conditions such as 9 (0.4% of total) for cleft lip-palate, 24 (1.1% of total) for cystic hydroma, and others for conditions either surgically correctable or involving lesser degrees of neurological/mental impairment (National Right to Life, 1996; Canady, 1996). This affirms multiple lines of evidence that even when maternal or fetal health issues are cited for an abortion (including in cases to circumvent gestational limits), the issues are often far from being life-threatening or dehabilitating.

The table below summarizes the preceding data on reasons for late-term abortions. Collectively, the data show that large fractions (likely a majority) of late-term abortions are performed for elective reasons rather than reasons of fetal or maternal health. This conclusion is consistent whether from official data, from studies published by pro-abortion groups, or from testimony by abortionists. Even in cases where fetal or maternal health is cited, these fractions may be misleading because available data show that such cases are subjectively defined to include situations of limited impact to the mother and/or the baby. Most non-elective late-term abortions are in cases of fetal health, but this includes cases where fetal health issues are possible but unconfirmed, and others where the health issues are of lesser consequence and/or correctable. This does not diminish the fact that many or most abortions for reason of fetal health involve conditions that threaten the life of the child or involve potentially life-long conditions. Rather, the point is that fetal heath abortions are not limited to such extreme cases, even apart from the ethical questions involved in such abortions.

of data
number of
percentage of late-term abortions
for reasons of:
maternal healthfetal healthelective
Arizona 2012-17 714 3 18 79
Florida 2000-07 230 6-13 39 39-55
Utah 2008-17 134 6-26 63 2-28
Foster et al. (2012)2008 5,109  0.6 
Haskell testimony ?-1993 >700 0 20 80
McMahon testimony ?-1995 2,000 7-9 5535-37

Again, abortion proponents defend late-term abortions by appealing to cases of maternal health issues or severe fetal health issues. Available data--including data from abortion proponents--refutes any assertive claim that such cases represent most such abortions, apart from any discussion of the medical or ethical issues these abortions present. Importantly, the reason for the lack of better data to bear on these questions is specifically that the abortion lobby successfully opposes gathering of such data. But the evidence presented here suggests that currently about 6,300 late-term abortions are performed annually for reasons other than fetal or maternal health (with a possible range of 3,000 to 12,000 annually).

Potential for survival of babies aborted in late-term abortions

The attention on late-term abortions stems in part from the fact that babies born alive at these stages of pregnancy have significant chances of survival. Rysavy et al. (2015) reported (for hospitals in their survey) overall infant survival rates of 5.1% at 22 weeks' gestation, 23.6% at 23 weeks, 54.9% at 24 weeks, 72.0% at 25 weeks, and 81.4% at 26 weeks.

Assume that babies aborted in late-term abortions were instead delivered at the gestational ages of their abortions, with survival probabilities as reported by Rysavy et al. (2015), except with 10-50% having fetal health issues making survival unlikely. The resulting estimates of babies that could have survived for selected years are:

Due to the current prevailing abortion-favoring interpretation of U.S. law, a child is not afforded constitutional protection of the right to life until delivered from the womb. However, even this is not secure given that abortion proponents actively oppose legislation that would protect babies born alive after abortions. There are significant numbers of cases where babies have survived abortions. Melissa Ohden is one such survivor (Ohden, 2017) and has founded the Abortion Survivors Network which reports knowing of 260 survivors (Abortion Survivors Network, 2019). The efforts to oppose life-saving care in such cases demonstrate the degree to which a pro-abortion mentality has warped both the legal and medical professions.

Abortion industry revenue associated with late-term abortions

Revenue to the abortion industry from late-term abortions is significant in part because such abortions are significantly more expensive than those at earlier stages of pregnancy. Representative costs of such abortions are reported in a number of studies (here converted to 2018 $ in parentheses):

Abortions later in pregnancy are even more expensive. One abortion clinic states "Third trimester abortions are quoted individually but are reported to be in excess of $5,000" (Rockville Women's Center, 2019). Other anecdotal reports affirm that extremely late abortions climb significantly in cost (e.g., Texas Right to Life, 2017).

Applying these ranges of cost estimates to the ranges of late-term abortions previously estimated, the following estimates of revenue from late-term abortions result (figures in 2018 $):

Given that abortions in general (dominated by early abortions) cost about $500 each on average, the above estimates imply that late-term abortions account for about 5% of the abortion industry's revenue (which overall runs near half a billion dollars annually).

U.S. late-term abortion policies vs. incidence

As of early 2019, seven U.S. states permit abortion for any reason (i.e., on demand) at any point in the pregnancy up to natural birth: Alaska, Colorado, New Hampshire, New Jersey, New Mexico, Oregon, Vermont, plus the District of Columbia. Eighteen other states permit abortions for any reason through viability, which is generally left to the determination of those performing the abortions. Based on late-term abortion practice in some of those states, "viability" is not interpreted in an objective manner. Nearly all states permit abortions throughout pregnancy for various health reasons.

The table below groups states by gestational limit for abortion on demand as of early 2019. Also shown are late-term abortion fractions by group. The overall late-term abortion fraction is for total abortions for the given group. The weighted median late-term abortion fraction is the median among each group's states weighted for the total estimated abortions (of any gestational age) for each group; bounds for the middle 95% of the weighted range are also given. Unsurprisingly, later (or no) gestational limits are associated with higher late-term abortion fractions.

gestational limit on abortionsnumber
of states
states with recent
late-term abortion data
states without recent
late-term abortion data
2018 abortion rate,
(per 1,000 women
ages 15-44)
2018 abortions,
late-term abortion percentage,
late-term abortion percentage,
weighted median (and 95% range)
20-22 weeks*19Alabama, Arkansas, Georgia, Indiana, Iowa, Kansas, Kentucky, Louisiana, Mississippi*, Nebraska, North Carolina*, North Dakota, Ohio, Oklahoma, South Carolina, South Dakota, Texas, West Virginia, Wisconsin(none)8.4189,0000.660.19 (0.03-2.37)
24-25 weeks**6Massachusetts, Nevada, Pennsylvania, Rhode Island, Virginia**Florida15.3152,0001.021.25 (0.25-1.40)
viability18Arizona, Delaware, Hawaii, Idaho, Illinois, Maine, Michigan, Minnesota, Missouri, Montana, New York, Tennessee, Utah, WashingtonCalifornia, Connecticut, Maryland, Wyoming15.4411,0001.541.94 (0.13-2.29)
no limit8Alaska, Colorado, New Jersey, New Mexico, Oregon, VermontDistrict of Columbia, New Hampshire17.080,0002.963.02 (0.81-7.19)
U.S. total5144713.1832,0001.391.25 (0.06-3.02)

* Limit in Mississippi and North Carolina is 20 weeks' gestation, remaining states in this group have limits of 22 weeks.
** Limit in Virginia is 25 weeks' gestation, remaining states in this group have limits of 24 weeks.

International context of U.S. late-term abortion policy and practice

The seven U.S. states (plus the District of Columbia), listed above, that permit abortion for any reason (i.e., on demand) through birth are extreme outliers globally. Very few countries in the world are as permissive as these U.S. states regarding late-term abortions. As reviewed by the Charlotte Lozier Institute in 2014 (Baglini, 2014), only six other countries plus some regions in an eighth country (Australia) permit abortion for any reason after 20 weeks' gestation:

Australia, like the United States, permits states or territories some degree of control over abortion policies and has regionally varying policies as a result. Victoria is the only state with unlimited abortion up to 24 weeks (Marie Stopes Australia, 2019), although several other states/territories permit late-term abortions with a doctor's approval though with little or no restriction on reasons.

However, there are some qualifications for some countries listed above. P.R. China since 2005 has banned abortions for reason of sex selection. Thus, those U.S. states permitting abortion on demand throughout pregnancy match the policies of only three countries in the world: Canada, North Korea, and Vietnam. Note also that P.R. China, North Korea, and Vietnam are countries known to conduct forced abortion (either currently or in recent years) against the will of the mother (Jacobson and Johnston, 2018).

Just as in the United States, however, abortion practice may differ significantly from what the law permits. Sex-selection abortion is understood to still be common in P.R. China (just as forced abortions continue in spite of national policies supposedly ending the practice). More importantly, many developed countries that limit late-term abortions to certain health cases appear to have much more permissive policies in practice. The figure below shows trends in late-term abortion fractions for countries with available data (from Johnston 2019b, 2019c).

Late-term abortion fractions are comparable to the United States average in several countries: Australia, Canada, Czech Republic, Denmark, Germany, Iceland, Netherlands, Russia, Spain, and the United Kingdom. Many of these countries show an increase in late-term abortion fraction in recent years. It is likely that this increase results from expanded late-term abortions of babies known or suspected to have health issues. Several European countries have virtually eliminated births of children with Down syndrome using abortion (Forte, 2018). All countries known to have recently increasing fractions of late-term abortions have universal healthcare, where abortions for fetal health conditions are cheaper than providing care for a child with even minor conditions: a late-term abortion is cheaper than surgery for cleft palate, for example.

This is not to say that late-term abortions in these countries are only conducted for health issues. Limited data from Australia suggest about half of late-term abortions there are for reasons other than fetal health or maternal physical health.

The increasing numbers of late-term abortions are accompanied in several countries (including Australia, Canada, Italy, Norway, the United Kingdom, and the United States) by increasing cases where the baby is delivered alive after the abortion and then either left to die or actively killed. This is why late-term abortions are increasingly identified as a numan rights issue on behalf of the unborn child (ECLJ, 2014).


In the United States, the subject of late-term abortions has occasionally become prominent in debate about abortion, in the past in relation to partial birth abortions and most currently in relation to aggressive efforts by abortion lobbies to expand legalization of extremely late abortions. The fraction of late-term abortions in the United States has, however, been relatively constant for decades.

(under construction)


© 2019 by Wm. Robert Johnston.
Last modified 14 April 2019.
Return to Home. Return to Other Policy Issues. Return to Abortion Statistics.