A professor of economics and finance at Baruch College, Joyce has studied the impact of various state regulations, including mandatory waiting periods, the cutoff of Medicaid funds, and parental notification laws, on abortion rates. He tells FRONTLINE that his study of abortions in Mississippi after the state enacted a 24-hour waiting period found a 12 percent decrease in abortion rates; however the number of women who went out of state for abortions increased by 40 to 50 percent and the number of second trimester abortions also increased by 40 to 50 percent. Joyce discusses a study in which he tried to measure the effect of unwanted pregnancies on infant health. His team studied sisters in an attempt to factor out socioeconomic differences and found no difference in low birth weight or infant health. "It's poverty, again, poverty and always poverty, that is driving the differences we're seeing here, and not really unwanted pregnancy," he says. Joyce also talks about the effects of parental notification laws and the potential impact of the federal Child Interstate Abortion Notification Act, passed by the House, which says that if minors cross state lines to have an abortion they should be governed by the abortion laws of their home states. This is the edited transcript of an interview conducted on Aug. 31, 2005.
… Let's talk about your study that appeared in JAMA [Journal of the American Medical Association about the effects of Mississippi's mandatory waiting period before abortions can be performed]. … Tell me about what you studied and what you found.
The question we asked was, what's the effect of a mandatory delay that requires women to go to the clinic twice and wait at least 24 hours after their first visit, after receiving a counseling message, before they could have the termination?
So the question was, did that impact abortion rates? Did it lower abortion rates? Did it increase the timing of abortions, and did it cause them to go out of state? And what we found was, abortion rates fell about 12 percent after the law. And we think that's associated with the law. The percent of second-trimester terminations increased about 40 to 50 percent. And we found an increase of women going out of state of about 40 and 50 percent as well. And we're fairly confident that this occurred because of the law, because it wasn't occurring in other states -- Georgia and South Carolina -- that were similar.
Why does it matter if somebody's waiting to terminate later?
Clinicians are better to talk [to] about it, but it's fairly known that the complications increase the later you go into pregnancy to have the termination. So the vast majority, about 90 percent of all abortions, are performed in the first trimester nationally. As you go into second trimester and later, you move into more general anesthesia as opposed to local anesthesias. You move into different types of procedures. The risk of a perforated uterus increases. Other uterine bleeding can increase. Infection increases. So there's an increase in complications associated with later terminations. …
When a law forces a woman to have terminations later, it becomes more expensive to acquire the termination, because the procedures become, again, more complicated. And moving to general anesthesias as opposed to local anesthesias, that's more expensive. Recovery time is more expensive.
There's more counseling involved in second-trimester terminations. There may be more complications that need scrutiny in second-trimester terminations. They may require return visits to the clinic because of … other complications. So the economic cost of delivering a service second trimester is more expensive than delivering the same service first trimester. …
Before we even looked at the data in Mississippi, we came at it from an economic angle, which says if you increase the cost of access in the service, you should expect to see the service used less. That has nothing to do with abortion. That has to do with general medical services in general.
Let's just take an example of prenatal care. If we flip it around and said: We're going to make women come and get a speech about how important birth is and how serious it is, and how serious she should be about having this birth, and if she wants, we'll finance her abortion, because if she's not serious, we'll pay for everything. We'll do that for her, and after she gets the speech, she has to go home, wait 24 hours, think this is really an important thing to do, and then reschedule and come back, and then we'll start her prenatal care, all right. But we're not going to start it until she's had this speech and waited 24 hours.
Do you think you would see women show up later in pregnancy for their first prenatal care visit if you imposed a law like that? I think without a doubt. One, because you've made it more expensive to access that service, you'll see that service used less regularly, less effectively. I think that goes without saying for an economist.
So when we go to Mississippi and we're looking at the data for abortions, we expect abortion rates to fall for that same reason. You've made the service more expensive to access. So the question there becomes, how much does it decline -- not whether it's going to decline, but how much -- and can we measure that? And those are the difficult empirical problems we had to confront in Mississippi.
We've been asked whether it was the delay forcing them to come twice to the clinic, or was it the speech and the message that she received that caused the abortion rates to fall after the law in Mississippi?
And for an economist, I'm more inclined towards the two-visit requirement, because that really is an added expense. The woman has to go, take time off from work perhaps, get baby sitters perhaps, access the clinic. If she lives far, she's traveling 50, 60, 70 miles. All that, for an economist, says cost, cost, cost, more expensive. …
The speech I'm not quite so convinced about, because that information's available. You could call up the clinic. They could tell you over the phone; they could give it to you in the mail. Many states do that. … And in those states, you don't find much impact of these mandatory counseling laws on abortion rates. Actually, we found none.
And one study that we did have some evidence on was from South Carolina.
South Carolina has a one-hour delay, so the woman can go to the clinic on one visit, receive the mandatory counseling, wait the one hour and have the procedure. … The one hour really is not an economic burden in that sense. And she's already committed to come, and she's already at the clinic. Or it's a minor burden, per se.
So we found that South Carolina's mandatory delay with a one-hour delay as opposed to 24 hours had no effect on abortion rates. So that gave us some evidence that it was the two-visit requirement forced by the 24 hours and not the message itself that was associated with the decrease in abortions.
Medicaid: What happens when [abortion] is something that the state is going to pay for versus when it's not?
Again, basic economics says that if I finance abortions for Medicaid-eligible women, I'm basically saying the out-of-pocket expenditures associated with an abortion -- about $300 -- are going to be covered by the state. Question is, will that increase abortion rates? Or if you take that funding away, will it decrease abortion rates? There's good, solid evidence that if you take away funding, abortion rates will fall among women who are Medicaid-eligible.
And I think the best study was done by colleagues at Duke University -- Philip Cook and other colleagues. And they looked at North Carolina's Medicaid funding of abortions. And it's a very interesting law, because the state sets aside a certain amount of money for Medicaid-eligible women to obtain abortions, and once that runs out in the fiscal year, there's no more funds left.
So if the fiscal year starts July 1 and you run out of funds March 1, then the women who are Medicaid-eligible and looking for abortions after March 1, there are no funds for them. They have to pay for it themselves. And what the study found is that when the funds ran out, abortion rates fell and births rose among Medicaid-eligible women. So that was very convincing evidence that women were very sensitive to the financial support of Medicaid to obtain their abortions. I think it's the best study done in the area.
Can you extrapolate into slightly larger public health context, related to cost of termination, cost of care, cost of caring for a child? Is there a way to quantify how [the] cost of a termination compares to the cost of a pregnancy across the whole community spectrum?
In the study I just mentioned about North Carolina, I had spoken to the lead author, and he was a little bit incredulous when he saw the results. He goes, "You mean to tell me that people faced with a $300 expenditure can't come up with the $300 to terminate a pregnancy they don't really want, and instead are willing to bear the expenses of raising a child from year one through year whatever with all the attendant expenses associated with child rearing?" And he goes, "For a middle-class person like myself, that's -- wow."
Three hundred dollars now or thousands and thousands of dollars to raise this child later -- it matters. It matters for poor people. I think they're more vulnerable to not having this kind of financial wherewithal to terminate a pregnancy and then just say, "OK, well, I guess I'll carry to term." …
Abortions are relatively inexpensive as medical procedures go. First trimester's about now $300, compared with a birth, which is in the thousands of dollars -- let's say $3,000 to $5,000; if there are complications, Caesarean sections, more. So the relative expense of just delivering the baby and prenatal care alone as compared to the abortion -- they differ by an order of magnitude. It's much more expensive. When you go onto raising the child, educating the child, feeding the child, the difference in expenses is just staggering.
What about Medicaid though? What we found in the Delta is that Medicaid actually pays for women who are pregnant to give birth and then gives them money for their kid. …
We did a study published in Family Planning Perspectives, and the title of the study was "Is Medicaid Pronatalist?" In other words, does Medicaid actually provide an incentive to have a birth as opposed to not to? So how do we test that? Well, we looked before and after the states expanded their Medicaid eligibility requirements. … We found evidence that births actually increased and abortions actually fell when states expanded Medicaid eligibility into higher income brackets. So that's some complicated evidence. …
Part of the counseling that is given to women in these mandatory delay situations is that they are told that there is funding; that the state, if they are eligible, will pay for the delivery and the natal care associated with the birth, that the child will be covered by Medicaid after delivery, that the mother will be eligible for Medicaid for a year after delivery as well, and if necessary, the state will help the mother seek support from the father of the child.
In other words, what the message seems to be is: "Don't worry about financing the child if you're poor. We can get or help you get the financing for this birth." Is that an incentive for some women? I think if they were not aware of it before, it has to be an incentive. It has to kind of say, "Wow, you're going to help me support this child and pay for my medical care?" Again, as an economist, you are subsidizing behavior, you expect to see more of that behavior, yes.
Let's talk about… Mississippi. It has some of the highest poverty rates, infant mortality rates, low birth weight rates. It has very restrictive abortion policies, and those have impacted the [abortion] rates. How do those things relate to each other?
We did a study years back looking at the effect of abortion on the rate of low birth weight and infant mortality, and the question was this: In states that had easy access to abortion and supported abortion and had high abortion rates, we hypothesized that you would see less unwanted births, and if you saw less unwanted births, you would see lower rates of low birth weight and lower rates of infant mortality. In that study, we found some evidence to support that. I'm not convinced as years have gone on that that evidence is as strong as I'd like it to be. …
There are a lot of surveys that ask women after they've delivered. They say, "At the time you were pregnant, did you want to become pregnant?" And the woman answers yes or no. If she says no, they ask them, "Was this because the pregnancy was mistimed, or was the pregnancy not wanted at all?" And so the woman answers either mistimed or unwanted. So these are what we call post hoc. These are after the mother has had the birth. …
If you look at women who said the pregnancy was unwanted, they tend to have higher rates of low birth weight. They tend to smoke more during pregnancy. They tend to show up later for prenatal care. Many people in the Institute of Medicine have interpreted that evidence [as there being] a negative association between infant health and unwanted pregnancies -- more unwanted pregnancies, worse infant health. Interesting question. The association is there.
Is it causal? Does unwantedness really cause women to basically not care for their child as well or their pregnancy as well and therefore have a bad birth outcome? So we looked at it. How do you test it? What you have to sort out is, you have to sort out the women who are more likely to say their pregnancies are unwanted tend to be poor, tend to be young, tend to be unmarried. So you have to kind of filter out those socioeconomic factors when you do these studies, and that's really hard to do.
So what we did is, we looked at sisters who came from the same family. … They both grew up in the same kind of similar environment, and so we could kind of eliminate the socioeconomic differences between the two sisters. And when you do that, we found no difference in the rate of low birth weight. We found no difference in the health of the child. We found no difference in the cognitive development of the child between the sister who had an unwanted birth and the sister who had a wanted birth.
The conclusion was, it's really socioeconomic status and not unwantedness. It's poverty, again, poverty and always poverty, that is driving the differences we're seeing here, and not really unwanted pregnancy.
This is a post hoc study, right. These are women who give birth. And there's a lot of problem with using data post hoc about their feelings during pregnancy, because if the birth was successful, if she's holding this child and she's nursing the child, the likelihood that she's going to say this child was unwanted, false. I mean, there's just a natural instinct to kind of embrace this child, often. And so the data you get afterwards is not nearly as clean as the data you would get if you ask women immediately when they became pregnant, did they want this or not? …
So unwantedness is a very hard thing to measure. If you ask people, their feelings are affected by the outcome. I could have had a child who was born low birth weight or had congenital anomalies, and maybe I'd be more inclined to say that the pregnancy was unwanted. In other words, the result of the birth can often affect how women retrospectively review that pregnancy, and that can affect the association between unwantedness and health outcomes.
So there have been these studies in … Czechoslovakia. When [a] woman wanted to have an abortion, she had to go in front of a committee and petition the committee to have the abortion. … So eventually, in the last leg of this study, what they did was, they went to the woman who had this child because she was denied an abortion, and they compared the outcomes -- the employment, the mental health, the criminal propensities of this child relative to a brother or sister born to that same mother who wanted that child -- and they found relatively little differences. So the evidence against unwanted birth and pregnancy and its income on the well-being of the child is not nearly as strong as people want to think it is. …
Has [there] been anything that studied wantedness and delay in the first prenatal visit?
Yes. We studied it. Again, we found an association between pregnancies that women described as "unwanted" when they conceived versus pregnancies that were described as "wanted" when they conceived. We found that there was about a 5 percentage-point difference in the proportion of women who showed up for first trimester.
In other words, in women whose pregnancies are wanted, 80 percent show up first trimester; for those whose pregnancies are unwanted, 75 percent showed up first trimester. There's about a 5 percentage-point difference. It wasn't really what we call robust. The association is there, but when you start to adjust for many other factors, the association often went away. So we were not convinced that it was a significant or very important relationship, but certainly there's some evidence for it. …
The analysis of parental notification laws is very, very challenging. … The point is this: It's very hard to measure abortions. Kids go out of state for abortions, and so the classic study is this: You go to a state. You measure abortions by where they occur. You pass a parental notification law. You see the abortions go down. Now, is that because the kids had less abortions, or the kids went to another state and had the abortion? Unless you can follow the kids to the other state, you don't know whether that observed decrease in abortions in that state is a real decrease or just an artifact, in the fact that you couldn't measure where the kids really went.
The early evidence was this: In Massachusetts is the first very important study [that] found that -- I think abortion rates among minors fell 40 percent when Massachusetts imposed a parental notification statute in 1986. They found the kids went to Rhode Island, New Hampshire, Maine. And when you took into account where the kids went, there was no change in the abortion rates to minors who were residents of Massachusetts. In other words, the law had no effect.
A colleague, Stanley Henshaw, did the same study in Mississippi. Again, if you looked at where the abortions occurred, you'd see a decrease in abortions to minors that happened in Mississippi after the enforcement of a parental notification law. …
What's changing is, the number of states that are imposing these laws is now up to 34. In other words, the borders are closing. It's becoming much harder for minors to go out to a nearby state to have the termination. As an economist, we think they have to travel so much further that the law will start to have more bite.
So who's done the most recent study? We're working on a study now in a very big state that is surrounded primarily by states that impose similar laws, so the kids don't have access to really out-of-state providers. And we're finding that the abortion rates of minors in these states is falling about 10 to 15 percent with the implementation of a parental notification statute in this state, so the point being the older studies were done in the '80s and '90s, when these laws were relatively rare. As these laws become much more common and the borders are closed, the laws are going to have more bite.
So if we can go further, and the federal legislation that's pending, that's been passed by the House, the Child Interstate [Abortion] Notification Act, says that if a minor is from a state that has a parental notification law and goes to another state to have that termination, in essence the minor will take her state law with her. In other words, the provider in the state, even if they didn't have a law, must verify that this kid is coming from a state that has a law and then honor the law in that state. In other words, the kid will carry her state law on her back to the other state. We think that will have a fairly important impact on the ability of kids to go to other states and avoid their own state's law. And therefore you'll see abortions fall for real. …
The big test case would be California, because if California votes this November to enforce a parental notification statute -- Nevada doesn't have one; Oregon doesn't have one -- it's very easy to get across the border. Anyone going to Las Vegas knows to go to Reno from California, and minors, we would predict, would go across the border to avoid compliance within California.
If this federal law is passed, when they go to Nevada, the California law would go with them. They go to Oregon, the California law would go with them. Given the size of California, this becomes a huge, huge change that would, we think, affect the abortion rates of minors and affect the birth rates of minors. We'd expect to see abortion rates to fall and birth rates to rise. …
Have you found that parental notification or parental involvement regulations impacted the timing of terminations?
There hasn't been much analysis of whether parental involvement laws are associated with minors having abortions later in pregnancy. We have some evidence from the state we're doing work in now that this state's notification statute increased the proportion of second-trimester abortions among a very specific group of minors who are 17 years of age.
So yes, there's some evidence that it causes them to delay. And it makes sense, because the law requires a 48-hour delay between the time the parents are notified and the termination is done. So mechanically you're talking two days. And if the child goes through a [judicial] bypass procedure, all right, that could further delay the child from having an abortion, which again would push them later into pregnancy. …
What can you tell me about how state regulations intersect with race and class?
Well, in Mississippi, in our study that [was] published in the Journal of the American Medical Association, we found that the mandatory delay of 24 hours had a larger impact on the abortion rates of whites than it did for blacks. … And it was greater on white teens than it was on black teens. And that's somewhat consistent with our knowledge of how teens across race communicate with their parents about sex, pregnancy and abortion.
… [What was the effect of the Casey Supreme Court decision?]
The [Planned Parenthood of Southeastern Pennsylvania v.] Casey decision again reaffirmed Roe v. Wade. But at the same time, it allowed states to regulate abortion so long as the regulations did not impose quote "an undue burden" unquote on the woman's right to an abortion. What it did is it opened up to interpretation the term "undue burden" and actually made evidence from social science much more relevant in court cases, because everyone's trying to define or struggle with the definition of an undue burden. And so different states, different judges and different courts have interpreted undue burden differently. …
I really don't know what courts are looking for exactly when a regulation imposes an undue burden on a woman. Is an increase in second-trimester abortions associated with the law? Is that an undue burden? Is forcing minors or pregnant women to go out of state for an abortion an undue burden? Is a decrease in the abortion rate an undue burden?
It may be to me or to you or to another person. But again, it may not be to others. Why? Well, forcing women to have later terminations is not necessarily an undue burden. All she has to do is basically stay on top of her pregnancy and go earlier, and she can offset that.
You don't have to go out of state. You can just comply with the law within your state and have the termination in your state. So, in other words, the law may have these effects -- and I think these effects we've documented are real. … But it's not fundamentally denying anyone the right to an abortion per se, and I think that's how they interpret it. …
But is [it] impacting how accessible abortion is?
Yes. Again, as an economist, you're raising the cost of getting an abortion. You're making it more expensive and more difficult. For me, it's easy to conclude that you're going to see fewer abortions, all right?
[For] judges, I think cost is not an aspect. Now, you can go back [to] the Supreme Court ruling in the Casey [case]. But I don't think cost is an issue that they want to entertain as being an undue burden, all right? And so again, I think it just throws open the entire ruling on these restrictions or on these regulations to the interpretation of the particular judge. And undue burden is a pretty vague standard. …
[Talk about what could happen if the Child Interstate Abortion Notification Act is passed and how it would intersect with state laws.]
The Child Interstate [Abortion] Notification Act -- CIANA as we call it -- could have a very profound effect as it interacts with parental notification laws. And California becomes the very interesting test case.
For example, if CIANA is passed by the Senate, and the Senate and House have a joint bill that's signed by the president -- and the Senate majority leader, Sen. [Bill] Frist, has said it was on his top 10 list of legislative agenda. Let's say there's a strong likelihood that CIANA is passed in some form, and minors who leave their state that have a law carry their state law with them. Let's say that goes into effect.
At the same time, let's assume that California voters vote in favor of the parental notification statute in California come November. The intersection of CIANA, that law by the feds, and California's implementation of their parental notification statute could have a profound effect on the abortion rates of minors in California because if the two intersect, … minors cannot go to Nevada; they cannot go to Oregon to have their terminations. They can't go anywhere.
And so the law, I think, based on the evidence that we have from work we're doing now, would have substantial bite, so to speak. It would have an impact on abortion rates. It would most likely increase second-trimester abortions. It would have an impact on births, causing kids to carry pregnancies to term that they would have terminated had there not been this law and had not there been CIANA, all right? So I think it's a very important legislation. And combined with the magnitude and the size of California, it could be the perfect storm. …
How would you characterize the landscape post-Casey?
The landscape post-Casey is we've seen just a large increase in state regulation of abortion from regulation of clinic requirements, from parental notification statutes, parental consent statutes, mandatory delay statutes. The states are trying to intervene in as many ways often as they can if the sentiment in the state is clearly anti-abortion.
And so Casey has allowed, fostered, fueled these kinds of regulations. Some folks consider this a very good thing, clearly. Other folks consider it an infringement on their personal rights, an increase in the cost of accessing services. That's not necessarily state intervention in a private decision, but Casey has allowed it because the requirement that a regulation must impose an undue burden is really just very vague and hard to interpret and varies from state to state. So let's go further down the road.
Let's say the Supreme Court changes, and Roe v. Wade is overturned. What's the impact then? I think Casey kind of previews what happens. States that have been supportive of abortion will have laws that make abortion legal, will go back to the pre-Roe years in which New York, California, Washington, Hawaii, Ohio, I predict, let's say, Washington, D.C., maybe Maryland, your blue states … would have legal abortion.
What happens? Kids in the red states, minors in the red states would have to travel to the blue states as long as they can get around CIANA. Women in the red states would have to travel to the blue states, all right? So abortion wouldn't go off the map. It would just be concentrated in certain states, most likely blue states.
And so Casey's been a little bit of a preview of that, because you're not going to eliminate abortion. Casey's not going to eliminate abortion. It doesn't come close to it, but it chips away. …
A health care provider in the Delta said to us … out there it really is like Roe doesn't really exist. Abortion has become so restricted and so difficult to access, it's like there's no Roe. Are there places in the country for where it's like Roe doesn't exist?
I think it's a strong statement. Roe does exist. You may have to travel further to have an abortion or to access abortion. But, you have the right to an abortion and you can get it.
I think an interesting aspect of Casey legislation is that for a while in Texas, there was not a provider I think in 2004 that would do a second-trimester termination. And so, implicitly, Casey had chipped away at Roe v. Wade by basically making second-trimester abortions unattainable in the state of Texas. …
That is how Casey can start to chip away in a very serious way into the trimester system established by Roe. I think South Carolina also. There is no provider in South Carolina that performs a second-trimester termination. If you're in South Carolina and you want one, you most likely have to go to North Carolina or to Georgia, or go to Atlanta for a second-trimester termination.
So we may be seeing the regionalization of second-trimester terminations in this country even though Roe's in place. And some states may, because of their restrictions and requirements that clinics have -- in essence, the kind of surgical capability similar to a hospital -- to perform second-trimester terminations, that they become so onerous and so expensive that clinics just get out of the business of second-trimester terminations, and the state doesn't have a provider that does one. And so that is a very real impact of post-Casey laws on the statutes of Roe. …
Do you see this happening more and more across the country? …
Since '92 you've seen just more and more regulation, and I think as the groups that are against abortion realize their clout and realize that they can get these kinds of restriction passed.
I think folks who are anti-abortion are also very good economists. They know if you increase the cost of these services, less of these services are going to be accessed. I think they had that down cold. And so they're pushing through legislation that increases the cost of accessing these services.
Now, their goal is clearly to decrease abortions, and I think they're achieving their goal in many ways by trying to make these services much more expensive and much more difficult to perform. …
SUPPORT PROVIDED BY
NEXT ON FRONTLINEOutbreakMay 5th