The incidence of abortion in the United States declined for more than a decade, but this trend may be ending, or at least leveling off. Nationwide, the number of abortions peaked in 1990, at 1.61 million, and dropped 25%, to 1.21 million, by 2005.1 Similarly, the abortion rate declined 29% over the same period, from 27.4 per 1,000 women aged 15–44 to 19.4 per 1,000. However, data from the Centers for Disease Control and Prevention (CDC), based on records from health departments in 48 reporting areas, show that the number and rate of abortions increased 3% between 2005 and 2006.2

The number and rate of abortions are in part dependent on the accessibility of abortion services, which may be affected by the number of providers, gestational limits, cost and antiabortion harassment. The number of abortion providers in the United States has been declining steadily:* It peaked in 1982, at 2,900 facilities, and had fallen to 1,800 by 2005.1 In that year, 87% of counties lacked an abortion provider, and 35% of women aged 15–44 lived in those counties;1 some of these women may lack the time or resources to travel to a provider.

In 2009, the murder of Dr. George Tiller—an abortion provider in Kansas—brought renewed attention to the issue of antiabortion harassment. Extreme acts of violence against abortion providers may lead to declines in the number of providers in the area.3 In 2000, some 82% of facilities providing 400 or more abortions per year experienced some type of harassment. Most commonly, harassment took the form of picketing and physical contact with or blocking of patients, but 15% of large providers received a bomb threat.4

This analysis provides abortion information for 2007 and 2008 to examine if the CDC data represent a reversal of the long-term decline or a “blip” in abortion incidence, and also presents updated measures of the accessibility of services. Recent information about antiabortion harassment can serve as an indicator of access to services and provide insight about the types of facilities that are most in need of legal protections against these activities.


Questionnaire Content and Fielding

The census of abortion providers described here, the 15th in a series dating back to 1973, was conducted in 2009; follow-up of nonrespondents extended into 2010. The questionnaire was modeled on the instrument used in the previous census, which collected data for 2004 and 2005.1 All respondents were asked the number of induced abortions that were performed in their facilities in 2007 and 2008, minimum and maximum gestations at which abortions were offered, and whether early medication abortion was offered. The data represent abortions according to the state in which they occurred, not the state of residence of women having the abortions. Clinics and physician providers (but not hospital providers) were also asked about the number of medication abortions performed (with separate items for mifepristone and methotrexate), charges for surgical and medication abortions, experience of harassment and the proportion of provider services accounted for by abortions. Questions about gestational limits, charges and proportion of client services for abortion referred to the time when the questionnaire was completed, so this information applies to 2009, since the majority of responses came in that year. We asked fewer questions of hospitals because the individuals answering the questionnaires in these settings typically have access to less information about clients. Information restricted to nonhospital facilities represents the experience of most women having abortions, since these providers perform the vast majority of all abortions (95% in 2005).1

All nonhospital facilities known to have performed abortions in 2005 were surveyed, and possible new providers were identified through various sources: provider listings on the Internet, newspaper ads, telephone directories, and the membership directories of the National Abortion Federation and the Abortion Care Network. Additional providers were identified and surveyed throughout the data collection process.

In May 2009, we mailed questionnaires to all potential providers, and two additional mailings were sent at three-week intervals to nonrespondents. In July and August 2009, the distributor of mifepristone (the drug used for most early medication abortions) sent the same questionnaire to a subset of approximately 1,200 providers (most of which we had likely already identified) that had purchased mifepristone. Twenty-nine providers that were not previously in our database responded to these surveys.

We also obtained information about abortion incidence from state health departments in 45 states and the District of Columbia. Many departments obtain only incomplete data from abortion providers, but we sometimes found the information useful even if it was incomplete. We used the health department figures only if the providers did not respond to any of our mailings or, in a few instances, if the number from the state was very different from the providers’ report and we had reason to believe the provider-supplied information was inaccurate (e.g., if the provider caseload differed substantially from that in the prior census and the facility was in a state with strict reporting requirements).

Intensive telephone follow-up of nonrespondents was carried out between September 2009 and June 2010, and particular effort was made to obtain the total number of abortions performed. During this phase of data collection, more than 7,100 contacts were made with approximately 1,000 providers.

Of the 2,344 facilities surveyed, 1,024 responded to the mailed questionnaire, 501 responded during nonresponse follow-up and health department data were used for 451 facilities. We determined that 14 providers had closed or stopped offering abortion services during the survey period, and excluded 15 facilities that we could not confirm provided abortions. For 109 facilities, we obtained estimates of the number of abortions performed from knowledgeable sources in the area, including other providers of reproductive health services and organizations that worked on reproductive health issues. We made our own estimates for the remaining 230 facilities, usually on the basis of information obtained in prior abortion censuses. If a provider had not previously participated in the census, our estimates were based on informal data, such as information from the provider’s Web site, caseloads of other providers in the immediate area and information obtained from telephone calls to the provider (e.g., hours of operation, gestations at which abortions were provided).

In the prior census, California’s health department provided information about hospitals, but only for inpatient abortions (typically procedures performed late in the second trimester). For the current survey, we obtained hospital data on both inpatient and outpatient procedures; this allowed us to identify 65 additional hospital providers in the state in 2008. These facilities performed 470 abortions in 2008, and we expect many of them had provided small numbers of abortions in previous years as well.

Of the abortions that occurred in 2008, some 82% were reported by providers, 9% came from health department data, 6% were estimated by knowledgeable sources and 3% were projections or internal estimates. By comparison, in 2005, some 76% of abortions were reported by providers, 12% came from health departments, 9% were external estimates and 3% were estimated internally.1


We distinguished among four types of providers: abortion clinics, other clinics, hospitals and physicians’ offices. Abortion clinics are defined as nonhospital facilities in which half or more of patient visits are for abortion services. Other clinics are sites in which fewer than half of patient visits are for abortion services; these include physicians’ offices that provide 400 or more abortions per year. Physicians’ offices are facilities that perform fewer than 400 abortions per year and have names suggesting that they are physicians’ private practices.

We obtained information on other aspects of abortion care from a majority of nonhospital facilities: Sixty-six percent provided information on the number of early medication abortions, 67% on gestational limits and 66% on charges for abortion services. Because response rates varied by facility type and caseload, we constructed weights that accounted for these differences. Item-specific weights were applied to medication abortion, gestational limits, charges and harassment. Unless otherwise noted, all abortion data presented include both surgical and medication abortions.

Census Bureau data on the population of women aged 15–44 for July 1, 2007, and July 1, 2008, were used as denominators for calculating abortion rates for the entire United States and for each state and the District of Columbia.5 We estimated the national abortion ratio as the proportion of pregnancies (excluding those ending in miscarriages) that ended in abortion; to do this, we combined our abortion counts with National Center for Health Statistics data on the number of U.S. births in the one-year periods beginning on July 1 of 2007 and 2008 (to match conception times for births with those for abortions).6–8


Abortion Incidence

The incidence of abortion in the United States changed little between 2005 and 2008: The number of abortions increased by 0.5%, from 1,206,200 to 1,212,350, and the abortion rate increased 1%, from 19.4 to 19.6 per 1,000 women aged 15–44 (Table 1). The abortion ratio did not change over this period, remaining at 22 abortions per 100 pregnancies.

The lack of change in abortion incidence nationally masks variations by state. Delaware had the highest abortion rate in 2008 (40 per 1,000 women), partly because of a 37% increase in the number of abortions (Table 2, page 44). Most of this increase can be attributed to one provider that acknowledged underreporting abortions in the 2005 survey. New York and New Jersey had the second and third highest abortion rates (38 and 31 abortions per 1,000 women, respectively). The abortion rate in the District of Columbia dropped 45% between 2005 and 2008, from 54 to 30 per 1,000, which made it the fourth highest in the country. High rates were also seen in Maryland, California, Florida, Nevada and Connecticut (25–29 per 1,000).

Wyoming had the lowest abortion rate, less than 1 per 1,000 women; the five states with the next lowest rates were Mississippi, Kentucky, South Dakota, Idaho and Missouri (5–6 abortions per 1,000). Notably, rates based on abortions performed in a given state may differ from rates based on abortions obtained by a state’s residents. For example, while only 70 abortions were reported in Wyoming in 2005, an estimated 1,100 were obtained by Wyoming residents in that year, and almost all of them occurred out of state.9

Change in abortion rates varied both within and across regions between 2005 and 2008. As in prior years, the rate was highest in the Northeast (27 abortions per 1,000 women), followed by rates in the West, the South and the Midwest (22, 18 and 14 per 1,000, respectively). Both the South and the West showed slight increases in rates between 2005 and 2008 (1–2%). The largest rate increases in the South were in Delaware and Louisiana (39% and 38%, respectively); Georgia and Kentucky also had substantial increases (18% and 16%, respectively). California accounted for 18% of the nation’s abortions in 2008, and its 2% increase was responsible for most of the increase in the West. The abortion rate in the Northeast did not change between 2005 and 2008; over this period, the rate rose by 23% in Pennsylvania, while it declined by 8% in Massachusetts and 9% in New Jersey. In the Midwest, the abortion rate was unchanged; the sizable increases in North and South Dakota mainly reflect small absolute increases in the relatively small number of abortions performed in those states. More notable were changes in two states that account for almost half of abortions in the Midwest: The abortion rate in Illinois rose by 9%, while Michigan showed a 5% decrease.

Provider Numbers

The national trend in the number of abortion providers paralleled that of abortion rates, showing very little change: 1,793 in 2008, compared with 1,787 in 2005 (Table 3). Twenty-seven states and the District of Columbia experienced a decrease in providers, while nine had overall increases and 14 experienced no change. The number of providers declined in the South (10%), the Northeast (8%) and the Midwest (5%). In contrast, it grew 15% in the West, largely because of a 23% increase in California. Without the newly identified facilities in California, the number of providers there would have increased by only 8%.

The potential impact of the loss of providers in a given state varies depending on the total number of providers in that state. For example, the largest decreases in the absolute number of providers occurred in New York and Florida, each of which had 12 fewer providers in 2008 than in 2005. However, this numerical loss of providers represented a 12% decline in providers in Florida but only a 5% decline in New York, because the latter had about 250 providers in both years. The loss of 10 providers in New Jersey represented a decline of 12% of that state’s providers.

In 2008, the overwhelming majority of U.S. counties (87%) lacked an abortion provider, and 35% of women of reproductive age lived in these counties. The proportions were lower in the Northeast (53% and 18%) and the West (74% and 13%). The former is the most densely populated region, which partially accounts for the relatively good coverage. In addition, the fact that states in both regions have larger (and fewer) counties than states in the Midwest and the South helps explain the below-average proportions of counties without a provider.

Abortion services are concentrated in cities. However, 69% of counties in metropolitan areas lacked a provider (not shown), and 25% of metropolitan women aged 15–44 lived in those counties. Almost all nonmetropolitan counties—97%—lacked an abortion provider, and 92% of women of reproductive age in these areas resided in those counties. These figures were virtually unchanged from those for 2005.1

Types of Providers and Abortion Caseloads

•Clinics. The 378 specialized abortion clinics accounted for 21% of all abortion providers, but performed 70% of all abortions in 2008 (Table 4). Most of these facilities reported 1,000 or more abortions during the year. A total of 473 nonspecialized clinics accounted for 24% of all abortions; some were similar to abortion clinics in having caseloads of 1,000 or more abortions per year. Overall, the number of very large providers (those performing 5,000 or more procedures) increased by more than 50% between surveys: Twenty facilities of this size accounted for 12% of all abortions in 2005,1 whereas 31 such facilities provided 17% of abortions in 2008.

•Hospitals. Thirty-four percent of abortion providers were hospitals in 2008, but these facilities accounted for only 4% of all abortions. Many hospitals provide abortions only in cases of fetal anomaly or serious risk to the woman’s health, and a majority (65%) performed fewer than 30 abortions in 2008. Twenty-two hospitals reported 400–999 abortions during the year, and only nine reported 1,000 or more.

•Physicians’ offices. Some 19% of providers were physicians’ offices, but these facilities accounted for only 1% of all abortions. A majority of these offices (57%) reported fewer than 30 abortions; our survey may have missed a number of small providers in this category.

Early Medication Abortion

Fifty-nine percent of facilities provided one or more early medication abortions in 2008, a slightly higher proportion than in 2005; a 4% increase in the number of such providers occurred over this period (Table 5). The number of nonspecialized clinics that provided early medication abortion services increased by 23%, but the numbers of hospitals and physicians’ offices doing so decreased (by 13% and 9%, respectively). Eighty-three percent of abortion clinics and 88% of other clinics performed at least one early medication abortion in 2008, whereas 25% of hospitals and 55% of physicians’ offices did so. The likelihood of providing early medication abortion services increased with caseload—from 30% among the smallest providers to 94% among the largest.

A substantial number of clinics and physicians’ offices—164 facilities, or 9% of all providers—offered early medication abortions, but not surgical abortions (not shown). Eleven percent of physicians’ offices were in this group, as were 27% of nonspecialized clinics. (Information on number of early medication abortions was not available for 34% of nonhospital facilities and 49% of physicians’ offices, and some of these facilities may have provided only this service; hence, our estimate is a conservative one.)

Some 199,000 early medication abortions were performed in nonhospital facilities in 2008, representing a 24% increase from 2005. Mifepristone was used for 94% of these procedures (187,000), and methotrexate for the remainder (not shown). Slightly more than half of early medication abortions were administered by abortion clinics, and most of the rest by nonspecialized clinics. Physicians’ offices averaged about two medication abortions per month and accounted for only 2% of all such procedures. In 2008, some 17% of all abortions performed in nonhospital facilities were early medication abortions; nonspecialized clinics had the highest proportion of such abortions (30%). Early medication abortions accounted for a larger share of procedures at facilities with smaller caseloads: 37–49% at facilities in the two smallest caseload categories, but only 9% at those with the largest caseloads. We did not collect data on the gestational age at which abortions were performed, but using gestation data from the CDC,2 we estimate that in 2008, slightly more than one-quarter of eligible abortions, or those before nine weeks’ gestation, were performed using medication.

Accessibility of Abortion

In addition to number and type of providers, gestational limits, cost and antiabortion harassment can affect the accessibility of abortion services.

•Gestational limits. Most providers have limits on the earliest and latest gestations at which they will perform abortions, and women who are very early in their pregnancy or in the second trimester may have a difficult time locating appropriate services. Some 42% of providers offered abortions at four or fewer weeks since a woman’s last menstrual period (not shown). The greatest proportion of providers offered abortions at eight weeks’ gestation (95%), and 64% offered at least some second-trimester abortion services (13 weeks or later). Twenty-three percent offered abortions after 20 weeks’ gestation, and 11% did so at 24 weeks. Access to very early and later abortions changed little since 2005, when 40% of providers offered abortions at four weeks and 8% did so at 24 weeks.1

Gestational limits varied by provider type. The proportion of nonspecialized clinics performing abortions dropped markedly after nine weeks—98% offered abortions at nine weeks’ gestation, while 63% did so at 10 weeks—probably because some provided only early medication abortion services. By comparison, 98% of abortion clinics offered abortions through the first trimester. Hospitals were more likely than other types of providers to offer abortions at later gestations: Fifty-eight percent reported that they performed abortions at 20 weeks’ gestation, whereas 36% of abortion clinics did so.

•Cost. A majority of women who access abortion services are poor or have a low income, and most women pay for the procedure out of pocket.10 The cost of obtaining an abortion, which varies by provider type and gestational age, may prevent some women from accessing this service. In 2009, the median charge for a surgical abortion at 10 weeks’ gestation was $470 (Table 6, page 48). Abortion clinics charged the least ($425), and physicians’ offices the most ($535). Surgical abortions at 10 weeks’ gestation were most expensive at facilities that performed fewer than 30 abortions per year ($629) and least expensive at facilities with the highest caseloads ($400).

We weighted the cost data by number of abortions to account for the fact that more women obtain abortions at facilities with lower charges; the resulting measure represents abortion patients’ mean out-of-pocket expenditures. Women obtaining a surgical abortion at 10 weeks’ gestation paid $451 in 2009, on average. The comparable figure in 2006 was $413, which is equivalent to $440 in inflation-adjusted 2009 dollars.11 Thus, the average amount that women paid for a first-trimester surgical abortion increased by only $11 between 2006 and 2009.

Abortions after the first trimester cost more because of the extra time, skill and resources required. The median charge for an abortion at 20 weeks’ gestation was $1,500, and charges across provider types and facility caseloads ranged from $1,100 to $1,650.

The median charge for early medication abortions was $490. Patterns in these charges by type of facility and caseload were similar to those for surgical abortions at 10 weeks, though the price difference between the lowest and highest cost facilities was smaller ($50). Notably, the median charge for this procedure was higher than that for surgical procedures at 10 weeks, perhaps because early medication abortion is a newer technology, and providers consider the cost of the drug an add-on to the cost of their services. This general difference in the median cost held across provider types and caseloads, with two exceptions: At physicians’ offices and at facilities with the smallest caseloads, a medication abortion cost less than a surgical procedure. These providers may specialize in medication abortion and, in turn, charge more for surgical abortion because it requires more training and specialized equipment.

•Harassment. Exposure to antiabortion harassment was common among nonhospital abortion providers in 2008: Fifty-seven percent experienced at least one of six types of harassment (Table 7). Picketing was the most common form of harassment (reported by 55%), followed by picketing combined with blocking patient access to facilities (21%). Internet harassment was assessed for the first time in this survey, and 3% of providers reported that protesters had posted pictures of patients on the Internet.

The overwhelming majority of abortion clinics—88%—experienced at least one form of harassment in 2008. Eighty-seven percent reported picketing; 42% reported picketing with patient blocking, and 21% cited incidents of vandalism. Nearly two-thirds of other clinics reported any type of harassment, but only 10% of physicians’ offices did so.

Harassment was also commonly reported by facilities that performed 400 or more abortions per year (89%). In 2000, when harassment was last assessed, 82% of providers with this size caseload reported at least one of five forms of harassment,4 which suggests a slight increase over this period. (Internet harassment was not measured in 2000, but even when this item was excluded from the 2008 data, 89% of these providers experienced at least one type of harassment.) Additionally, between 2000 and 2008, the proportion of such providers reporting picketing increased from 80% to 88%, and the proportion reporting picketing with contact or blocking access increased from 28% to 37%; the proportion that received bomb threats declined from 15% to 5%. Almost all providers that performed 1,000 or more abortions had been picketed in 2008, and 63% of facilities that performed 5,000 or more abortions reported picketing that involved blocking or physical contact. Nearly one in five of the largest facilities reported a bomb threat.

The incidence of harassment varied by region; 85% of providers in the Midwest and 75% in the South experienced any form of harassment, compared with 48% and 44% in the Northeast and the West, respectively. All types of harassment were more common among facilities in the Midwest and the South than elsewhere. Levels of harassment did not vary by gestational age at which abortions were offered among providers that performed 400 or more per year (not shown).


The long-term national decline in abortion incidence has stalled and may have ended. Both the number of abortions and the abortion rate increased slightly between 2005 and 2008. Notably, the small change in abortion incidence at the national level masks substantial changes in some states.

Delaware’s abortion rate is twice the national average and reflects, in part, that residents of other states obtain abortions in Delaware. Out-of-state residents accounted for an estimated 25% of abortions performed in the state in 2004.9 However, the apparent dramatic increase in the state’s abortion incidence is probably spurious, because abortions in 2005 were underreported at one facility.

In other states, shifts in abortion incidence may be partially explained by interstate dynamics. For example, the decrease in the number of abortions in New Jersey was paralleled by an increase in the number in neighboring Pennsylvania. In 2004, some 12% of abortions among Pennsylvania residents were obtained out of state,9 but this proportion may have dropped in 2008. Meanwhile, although the District of Columbia still has one of the highest abortion rates in the country, it had the greatest decrease in abortion rate between 2005 and 2008, while neighboring Virginia had an increase. Over this three-year period, the District of Columbia experienced a substantial decline in providers because several clinics closed. In 2004, almost one in five abortions in the District were obtained by nonresidents,9 and in more recent years women likely chose to obtain an abortion in their state of residence, or had no choice but to do so. Indeed, the increase in abortion incidence in Virginia may reflect that fewer of this state’s women traveled to the District to terminate their pregnancies, as well as that more women traveled from the District to Virginia for this purpose.

Changes in abortion incidence may also be due to developments within a state. For example, the abortion rate in Georgia increased 18% between 2005 and 2008. The state gained three large clinic providers (including an abortion clinic), but it lost five small providers (hospitals and physicians’ offices). As a result, the proportion of women of reproductive age who lived in counties without a provider declined by five percentage points from the 62% level found in 2005,1 and the shift in provider types may have increased access to services and, in turn, the abortion rate.

This is the first census since 1982 showing no decline in the number of providers,9 but we believe that the very small increase in the number of providers between 2005 and 2008 is due to the collection of more accurate data. In particular, new information from the California health department resulted in the inclusion of 65 more hospital providers, which performed a small number of abortions in 2008. If these facilities had not been included, the national number of abortion providers would have declined by 3% between 2005 and 2008.

Early medication abortion has become an integral part of abortion care. Although the proportion of providers offering this service increased only slightly between 2005 and 2008, both the number of early medication abortions and the proportion of all abortions accounted for by this method grew substantially. Mifepristone use has grown steadily since its introduction in the United States in 2000, and substantially in recent years. We found a large increase in the number of mifepristone-induced abortions, from 158,000 in 200712 to 187,000 in 2008. This increase over one year corresponds with recent usage estimates from the manufacturer and may suggest an increased reliance on this procedure.13 Early medication abortion appears to be particularly important for nonspecialized clinics; it accounted for 30% of all abortions at these facilities, and a minimum of 27% of nonspecialized clinics offered only early medication abortion services.

Most of our measures of accessibility of abortion services showed little change between 2005 and 2008. As in the earlier year, 35% of women of reproductive age lived in the 87% of counties that lacked a provider. The proportions of providers offering abortions at four weeks after a woman’s last menstrual period and at 24 weeks also remained stable. Furthermore, after adjustment for inflation, the average amount that women paid for a first-trimester surgical abortion increased by only $11 between 2006 and 2009. This small change was particularly notable given that medical price inflation has increased at a faster pace than inflation in other sectors.14 The moderate rise in abortion cost may be the result of overt efforts by providers to keep the procedure affordable, or given the increase in poverty among abortion patients,10 it may reflect that more women obtain abortions at facilities that charge the least.

Most nonhospital abortion providers experienced at least one form of harassment, and the proportion of large providers reporting any form of harassment increased slightly between 2000 and 2008. For the first time, we examined variations in harassment by provider type, caseload and region; virtually all abortion clinics and large facilities had experienced at least one form of harassment, as had at least three-quarters of providers in the Midwest and the South. Extreme forms of harassment lead to fewer abortion providers,3 and the relatively high levels of harassment in the South and Midwest may have contributed to both the decline in numbers of abortion providers and their relatively small numbers in these regions.

Media reports suggest that the economic recession that began in late 2007 has led to increased demands for abortion services.15,16 This study was unable to assess these claims, as the small increase in abortion incidence began before the recession.


We are aware of several limitations in our data. Undoubtedly, some abortion providers were not counted because we were unable to identify them. An earlier survey specifically designed to assess the extent to which providers are missed found that our census overlooks a number of small providers and suggested that the actual number of abortions in 1992 was 3–4% greater than the census indicated.17 Undercounting has likely become more pronounced over the last decade because of the integration of mifepristone for early medication abortion at facilities that do not offer surgical abortions.1,12 In particular, facilities that performed few abortions may have been reluctant to identify themselves as abortion providers and may not have responded to the survey mailed out by the distributor of mifepristone.

While these dynamics might influence statistics related to the total number of providers, facilities with larger caseloads—which account for the overwhelming majority of abortions—are more easily identified because they are typically known by other providers in their communities and advertise on the Internet and in the yellow pages. In addition, although we made intense efforts to obtain data from all known abortion providers, we had to make informed estimates for some facilities. Similarly, our weighting procedure assumed that nonhospital facilities that did not respond to specific items resembled those of the same type and in the same region that did respond; if this is not the case, information about these aspects of abortion services may be somewhat inaccurate. Finally, data inaccuracies may have been introduced when facilities (particularly those that do not use electronic records) provided us with abortion estimates.


While nationally it would appear that little has changed regarding abortion incidence, abortion is only part of the larger picture of unintended pregnancy, and information on unintended births is also needed. An increase in the rate of unintended births along with the abortion rate would indicate that unintended pregnancy is on the rise. Alternately, if the rate of unintended births decreased, then the slight increase in the abortion rate might indicate that abortion had become more accessible. Between 1994 and 2000, abortion rates increased among poor and low-income women, while they decreased among those with higher incomes;18 the fact that the representation of poor women among abortion patients increased between 2000 and 2008,10 while the abortion rate declined only slightly during this period, suggests that barriers to abortion services were reduced for this population. In the context of the economic recession that was occurring in 2008, their growing representation could also signify that increasing financial instability left low-income women less able to prevent unintended pregnancy or less well equipped to carry an unintended pregnancy to term.

Patterns in abortion incidence and number of providers have several public policy implications. Abortions are usually the result of unintended pregnancies;19 affordable family planning services need to be widely available to women and their partners to reduce the number of unintended pregnancies and, in turn, abortions. In addition, it is important to remove barriers to abortion services, especially for low-income women. Only 17 states use their own funds to cover all or most medically necessary abortions for women with Medicaid coverage.20 If more states did so, or if federal restrictions on Medicaid coverage for abortions were lifted, poor women could more easily access services when confronted with an unintended pregnancy. Harassment of abortion providers continues to be a problem, particularly in the Midwest and the South. More states need to enact and enforce laws that prohibit the most overt and damaging forms of harassment and allow access to this legal, needed and basic health care service.