Printer Friendly

An economic approach to abortion demand.

This paper uses econometric multiple regression techniques in order to analyze the socioeconomic factors affecting the demand for abortion for the year 1985. A cross-section of the 50 states and Washington D.C. is examined and a household choice theoretical framework is utilized. The results suggest that average price of abortion, disposable personal per capita income, percentage of single women, whether abortions are state funded, unemployment rate, divorce rate, and if the state is located in the far West, are statistically significant factors in determination of the demand for abortion.

Part I: Introduction and Review

of Past Research

On Monday, January 22, 1973 a decision on the controversial case of Roe v. Wade was made by the United States Supreme Court:

The constitutional concept of "liberty" as well as the right to privacy protects a woman's right to choose to terminate a pregnancy, at least in the early weeks of pregnancy ... After the first trimester of pregnancy, the state may regulate abortion; after the second, it may forbid it (except when it is necessary to preserve the life or health of the mother), Roe v. Wade 410 US 113 (1973), paraphrased in Rubin, 1982, p. 66.7).

Thus it was ruled that the states do not have the power to "forbid" a woman from having an abortion except in the third trimester of pregnancy. However, there have been efforts to curb the ability of women to obtain legal abortions. In 1977 Congress approved the Hyde Amendment which restricts federal Medicaid funding of abortions to cases where the life of the woman in endangered, or she has been a victim of rape or incest, (public law 95-205, 1977). In addition, by 1985 only fifteen state and the District of Columbia provided (nonrestrictive) state funds for abortions for women who qualify, (the Alan Guttmacher Institute, 1988, p. 37). Before 1983, many states required that abortions done on second trimester pregnancies must be performed in a hospital. But, in 1983 the United States Supreme Court determined in Akron v. Akron Center for Reproductive Health that these regulations did not aid the health of the pregnant woman, and that by not allowing second trimester abortions in a nonhospital setting, there was a "heavy" and "unnecessary" burden put on a woman's access to a "safe" and "relatively inexpensive" abortion. The regulations were thus declared illegal, (City of Akron v. Akron Center for Reproductive Health, Inc., et al. 462 US 416 (1983); Rodman, Sarvis, Bonar, 1987, p. 120-1).

Since the historic decision of Roe v. Wade in 1973, there has been a lot of research which involves analysis of statistical percentages of the demographic characteristics of women in the United States who obtain abortions. This type of research is continually done by the Alan Guttmacher Institute which conducts a national survey of abortion providers in order to obtain this type of information.(1) They have found that during the eighties most women who obtain abortions are white, young, single, and from metropolitan areas. The nonwhite abortion rate is higher than that for whites, a large percentage of women obtaining abortion shave no previous children, and Catholics are just a likely to obtain abortions as others while Evangelical Christians are somewhat less likely. Women covered by Medicaid have a higher abortion rate, (The Alan Guttmacher Institute, 1987, p. 5-9, 1988, p. 158-68).

Another study uses individual-level data on women who obtained abortions in eight states and New York City for the year 1980 in order "... to investigate the net effects of selected sociodemographic variables on abortion," (Powell-Griner and Trent, 1987, p. 553). Using multivariate analysis the study found that women are more likely to obtain abortions if they have one or more of the following characteristics: residence in metropolitan areas, no previous children, completion of high school, black, reaching the end of their child-bearing years (35-39), unmarried.

In addition, research has been done on effects of abortion legislation. One such study examines the effect of the Roe v. Wade decision on maternal health. It uses an interrupted time-series design and ordinary least squares regression analysis with monthly data from January 1951 to January 1973 (pre-Roe) and February 1973 to December 1978 (post-Roe). The results suggest that in the short-run there was a significant reduction in abortion-related deaths after 1973, (Legge, 1985, p. 130).

A study analyzing the public costs of the Hyde Amendment found that restricting federal medicaid funds for abortion to a qualifying woman which causes her to give birth will result in a "... discounted future public cost ... for 1978 to be almost 100 times the cost of an abortion," (Sommers and Thomas, 1983, p. 340). This takes into account the public costs associated with raising the child until she is eighteen. Similar research has also been done by the Alan Guttmacher Institute. In fiscal year 1978, after the implementation of the Hyde amendment, "... if even one-fifth of pregnant indigent women go on to have unwanted births paid for by Medicaid rather than obtaining legal Medicaid abortions, the public costs of the births will far outpace the savings, if only the costs of the abortions and deliveries are considered," (The Alan Guttmacher Institute, 1980, p. 130). Similar results were estimated for c. 1985, "... for for every public dollar spent to pay for abortions for poor women, more than four dollars is saved in medical and social welfare costs over the next two years," (The Alan Guttmacher Institute, 1988, p. 45).

There has been some research undertaken which utilizes an economic approach to abortion issues in the United States.(2) Medoff (1989) uses economic analysis and multiple regression techniques to determine which states would be most likely to illegalize abortion if the Roe v. Wade decision is overturned. In order to determine this he attempts to take social, economic, and ideological factors into account. The dependent variable is a measure of whether the two senators from each state voted for or against the Hatch/Eagleton Amendment.(3) The results suggest that the higher the percentage of women in white collar jobs, the higher the percentage of nonwhites, and if the state ratified the Federal Equal Rights Amendment, the more likely the state will be pro-choice. The higher the percentage of fundamentalist Christians, the less likely the state will be pro-choice, (p. 188).

Deyak and Smith (1976) did research to determine the economic benefits accrued by women seeking abortion due to the Roe v. Wade decision made in 1973. They estimate a demand equation for abortions using "... the number of women receiving abortions in Upstate New York after traveling from each of 47 states and the District of Columbia (Hawaii and Alaska are excluded) during the 1971 calendar year," divided by the number of live births in each state, in order to estimate the abortion rate, (p. 88).(4) Their suggest that travel-cost, illegal abortion price (data is for only 29 states from court cases ranging from 1948-71), and percentage of urbanization have a negative effect on abortion demand, while percentage of Catholics has a positive effect. Using consumer surplus analysis, it is determined that for women who would have traveled to New York State for abortions, savings range from $13 to $33 per woman, or totalling from $1.9 million to $4.7 million, (p. 96).

There has been only one instance or research done in economics which attempts to determine the socioeconomic factors influencing abortion demand in the United States after the 1973 Roe v. Wade decision. Medoff (1988) undertakes a cross-section analysis of abortion demand in the fifty states and District of Columbia for the year 1980. In his research he uses an economic model of fertility control and a linear regression technique.(5) The dependent variable he uses is the number of abortion per 1000 pregnancies (A). Medoff's results suggest the average income of women aged 15-44 (Y), percentage of women 15-44 unmarried (SNGL), labor force participation rate of women 15-44 (LFP), state located in the far West (W), and states with Medicaid funding for abortions (M) have a significant positive impact on the abortion demand. Average price of abortion (P) has a significant negative impact on abortion demand and percentage of Catholic population in each state (CATH) has an insignificant effect on abortion demand. The specific results of his estimated model are as follows:(6)

A = - 207.780 - .942P + .031Y

(1.42) (3.22) (3.31)

+ 4.194SNGL + 4.456LFP

(1.74) (2.57)

+ 18.287w + 1.207CATH
 (1.74) (1.50)
 + 43.775m [R.sup.2] = .77
 (2.12)


In his research, Medoff also found the percentage of nonwhite women 15-44 to have an insignificant effect on the abortion rate as well as the percentage of women 15-44 who have competed twelve years of school and the percentage of women 15-44 in each state in poverty.

Although there has been little economic research on the subject of abortion, economic theories have been developed on the related topic of fertility control and the demand for children. Michael (1973) and Becker (1981) have made important contributions in this area. Abortion is a means of fertility control - it is a decision made by a pregnant woman to not have her child.

Michael applies a "... household production function model to the household's fertility behavior ..." (p. S130). According to the theory, if the net benefit ("... excess demand for children at the prices, income levels ... and so on ...") associated with an additional child is positive, then the household may attempt to conceive a child, if it is negative, "... the couple may engage in fertility control," (p. S132).(7) Becker also develops a theory of the demand for children, in which the price of child and real income contribute in determining the number of children a couple desires to have, (p. 102). Becker uses the theory of the demand for children to examine the relationship between the quantity and quality of children. Quantity and quality are inversely related. Quality rises when households invest more time into their children and provide them with more education; thus cost of a child must increase when quality increases. In addition, as quality of children rises, households can afford to invest less in each child, thus quantity demanded falls, (p. 103). Priorities of quantity and quality of children will affect fertility decisions in different directions.

[In Part II of this paper, a model for the estimation of the demand for abortions in 1985 will be presented. part III will discuss the results of the estimation, and Part IV will discuss how this type of research on abortion could be extended and the outlook for the future of abortion.]

Part II: Model - Abortion Demand

An application of Michael and Becker's economic fertility control models to abortion demand suggests that economic factors and preferences (including moral considerations) will be elements a household will consider in choosing whether or not to terminate a pregnancy through an abortion.(8) If the household's consideration of the benefits of a legal abortion, outweigh its costs, a legal abortion will be demanded. The following is the equation of the abortion demand model to be estimated:

[ARATE.sub.i] = [beta.sub.0] + [beta.sub.1][PRICE.sub.i] + [beta.sub.2][DPINC.sub.i]
 + [beta.sub.3][MEDST.sub.i] + [beta.sub.4][SINGL.sub.i]
 + [beta.sub.5][WESTV.sub.i] + [beta.sub.6] [UNEMP.sub.i]
 + [beta.sub.7][DRATE.sub.i] + [beta.sub.8] [HIGHS.sub.i]. (1)


ARATE is the dependent variable in the abortion demand equation. It is the percentage of pregnancies of women aged 15-44 which are terminated by legal abortions in each of the fifty states and the District of Columbia for the year 1985.(9) This better reflects the abortion decision than the percentage of women aged 15-44 obtaining a legal abortion or percentage of abortions per live births.(10) However, this is an imperfect measurement of the abortion rate because under-reporting of the number of abortions may have occurred. In the survey of abortion providers implemented by the Alan Guttmacher Institute some small hospitals and physician providers may have been missed. However, in recent years (1985 included) the under-reporting of hospitals has been lessened because "Most state require hospitals to report data on abortions to the state health department," (The Alan Guttmacher Institute, 1988, p. 75). In addition, the abortion rate is an imperfect measure of the abortion decision made by women residing in each particular state due to the fact that women sometimes travel out of state to obtain abortions.(11) Thus states with fewer abortion providers (usually states with large percentages of nonmetropolitan populations) tend to have lower abortion rates because many of their residents receive abortions in neighboring states. "In 1982, the latest year for which we have data, six percent of abortions nationwide were performed outside the woman's home state ..." (The Alan Guttmacher Institute, 1988, p. 19).

There is reason to believe that after 1983 there have been less abortions performed out of state. In the controversial United States Supreme Court decision on the City of Akron v. Akron Center for Reproductive Health, Inc., et al. made in 1983, it was determined that second trimester abortions could be performed in nonhospital settings. Because this initial restriction was costly, time-consuming for the woman, had limited availability, and was not implemented in all states, women may have traveled to other less restrictive states in order to get much less costly abortions, (The Alan Guttmacher Institute, 1984, p. 260). Another restriction that was struck down by the Supreme Court in this case was that of an obligatory 24-hour waiting period before the woman signed her (informed) consent for the abortion. This also added to the cost of the abortion, and applied pressure on the woman, (Rodman, Sarvis, and Bonar, 1987, p. 122). Since not all states had this requirement, this may also have been a motivation for a woman to travel to another state to get the abortion; this way she would avoid the pressure and time constraint which the restriction put on her.(12)

PRICE, an explanatory variable, is the average cost of an abortion performed in the first trimester in a nonhospital facility with local anesthesia for each state and the District of Columbia in 1985, (the Alan Guttmacher Institute, 1988, p. 21).(13) It is expected that price will have a negative impact on the abortion rate. This assumes (as previous research has suggested) that abortion is a contraceptive (service) and follows the economic law of demand (as price rises quantity demanded falls) for normal goods and services.

The reason that the restrictions still allow for a fair estimate of average abortion price is because the amount of abortions performed after the first trimester is about ten percent, only a small amount of facilities offer general anesthesia rather than local anesthesia, and "Eighty-seven percent of the abortions performed in 1985 were done in nonhospital facilities," (The Alan Guttmacher Institute, 1988, p. 15, 21).(14) Thus most abortions are the type which the average price data measures. The measures of average abortion price is imprecise because data on the actual state-by-state price is not made available. The Alan Guttmacher Institute does provide a list of states which have prices above or below the average of $213 and those price ranges are given, thus a price estimate for each state has been obtained. (The Alan Guttmacher Institute, 1988, p. 21).

There is a reason to believe that the supply of abortions is elastic. In other words, an increase in abortion demand does not affect the price level, and the supply of abortions can be expanded relatively easily so that this demand is met.(15) Hospitals provide one way in which the demand for abortion can be met. According to the Alan Guttmacher Institute,

Although the number of abortions performed in hospitals has declined, these facilities are still important in making services available. Physicians in small communities may judge that existing demand does not justify their equipping their offices to provide abortion services, and such doctors may, therefore, prefer to use hospital facilities, (1988, p. 19).

Both hospitals and physicians taking on more abortion patients provide ways in which the abortion services can be relatively easily expanded to meet demand without really affecting cost. In addition, there is evidence which suggests that price is not the driving factor in the supply of abortions. According to the Alan Guttmacher Institute, "on average, seven percent of all patients in nonhospital facilities are charged reduced fees;" this basically is done for low-income women, (1988, p. 22).(16) this demonstrates that this is not a profit maximizing situation, thus a change in demand for abortions will most likely not affect abortion price.

DPINC, disposable personal per capita income for the fifty states and the District of Columbia in 1985, is another independent variable in the model.(17) Income serves as a budget constraint for the household. Disposable income is used because it is after-tax income which is spent on goods and services. Although disposable personal income per woman of childbearing age 15-44 might be a better measure of income, it can be assumed that states with higher income will also have relatively higher income for women.

According to economic consumption theory, as income rises, more of a normal good is demanded, thus in this case it is expected that the effect of income on abortion demand will be positive. Becker (1981), in considering an economic model of fertility control behavior, states that higher earnings for a woman cause her time to become more valuable. This will cause the cost of an additional child to rise and she will then demand less children. Thus the higher the woman's income, the more likely she will engage in fertility control including abortion. In addition, as household income rises, the demand for quality in children rises; this causes cost per child to rise which results in a lower demand for children and a higher usage of fertility control methods, (p. 98). Due to these factors, the impact of income on abortion demand is expected to be relatively large.

MEDST, an independent variable, consists of states which have funding through Medicaid for abortion for poor women in 1985. Although the Hyde Amendment, implemented in 1977, restricts Federal funding of abortions through Medicaid, in 1985 there were fifteen states and the District of Columbia which provided state funded abortions. A dummy variable is used for MEDST which equals one for the states which provide unrestrictive funding of abortions (Alaska, Colorado, District of Columbia, Hawaii, Maryland, Michigan, New York, North Carolina, Oregon, Washington, West Virginia, California, Connecticut, Massachusetts, New Jersey, Vermont) and zero for the other states, (The Alan Guttmacher Institute, 1988, p. 37).(18) It is expected that MEDST will have a positive impact on abortion demand because without a cost consideration, poor women will be more able to obtain an abortion if they choose.

SINGL, an explanatory variable, is the percentage of unmarried women aged 15 and older in the fifty states and District of Columbia for 1980.(19) More recent data could not be obtained, but it can be assumed that on average the state-by-state percentages of single women remained relatively the same for 1985. The Alan Guttmacher Institute points out that abortion rates are higher among unmarried women, (1988, p. 163). This may be because a woman is generally under a larger economic constraint when she is single. In other words, she may not to able to afford to expend the time and money needed to have and care for a child, thus abortion becomes a more likely alternative.(20) In addition, applying Becker's theory to this variable, the value of a woman's time may be higher when she is single, thus she will demand less children. It is expected that as the percentage of single women rises, the abortion rate will also rise on average and the impact will be large.

WESTV is an independent variable which consists of states located in the far West. A dummy variable is used for WESTV which equals one when the state is located in the far West (Arizona, California, Hawaii, Nevada, Oregon, Washington) and zero if the state is not. Drawing from the fertility control model, both economic factors and tastes interact to determine if a woman will obtain an abortion. Medoff (1988) points out that tastes in states in the far West may differ from other states due to social norms, ties, etc.(21) In addition, the Alan Guttmacher Institute states that "In general, [abortion] rates are higher on the ... West Coast ...", (1988, p. 18). Thus the expected impact of WESTV is positive, states located in the far West will have higher abortion rates on average.

UNEMP, an independent variable, is a measure of the unemployment rate of the civilian labor force aged 16 and over in the fifty states and Washington D.C. for the year 1985.(22) The female unemployment rate was not used because the labor force unemployment rate captures the state of the economy better, which may be an important factor in the abortion decision.

It is not clear in which direction the unemployment rate will affect the abortion decision. Economic theory suggests that as the unemployment rate rises, households will be able to purchase less goods and services, thus abortion rate will fall. Drawing from Becker's theory of fertility control and the demand for children, it is possible that as unemployment rate rises, households will be less able to afford to raise and care for a child, thus abortion rate will rise. Conversely, as unemployment rate rises, the value of a woman's time falls so that she will be able to spend more time with her child at less cost, thus abortion rate will fall.(23) One other likely possibility is that in times of high unemployment and economic uncertainty, people will be more careful about their contraceptive use because they realize the cost of getting pregnant, thus not only will the birth rate fall, but the abortion rate will also fall.

DRATE is another explanatory variable in the abortion demand equation. DRATE is the divorce and marriage annulment rate in the fifty states and District of Columbia for the year 1985.(24) Ideally, the data would be a measure of the divorce and marriage annulment rate involving women aged 15-44 in 1985. However, it can be assumed that this measurement will reflects this because it does include women aged 15-44, and a large percentages of divorces and marriage annulments do occur among women who are still in their child-bearing years. (U.S. Bureau of the Census, 1988, p. 83). It is expected that the divorce and marriage annulment rate will have a positive impact on the abortion rate. This is because divorce or marriage annulment usually causes a period of turmoil and change, and the woman will probably be less likely to want to have the child of the man she is separating from. In addition, Becker's theory of fertility control can be applied to the effect of divorce rate on the demand for children. When a woman gets divorced she will most likely depend on her salary; this will cause the value of her time to rise which will make her desire less children and engage in fertility control, thus abortion rate will rise. Due to these factors, the impact of DRATE on the demand for abortion is expected to be relatively large.

The final independent variable included in the abortion demand model is HIGHS. HIGHS is the percentage of women aged 15 and over who have completed 12 years of education or more in the fifty states and District of Columbia for 1980.(25) A better measure would be the percentage of women aged 15-44 who have completed 12 years of education or more for 1985. However, it can be assumed that the variable captures the relative education levels of women who are of child-bearing age in each state for 1985. The direction in which the education level affects the abortion rate is not clear. Michael (1973) states that "... more-educated couples have greater access to fertility-control information, and are therefore more successful in preventing unwanted pregnancies." (p. S140).(26) Following from this, education will have a tendency to lower the abortion rate. Becker (1981) notes that education causes the value of the woman's time to rise, thus the cost of having a child also rises and she demands less children, (p. 98). Therefore, she will use more fertility control measures, and since abortion is a method of fertility control, the abortion rate will rise.

Part III: Empirical Results - Abortion

Demand Equation

The abortion demand equation (1) was estimated using ordinary least squares and the empirical results are as follows:(27)

ARATE = - 11.654 - .041PRICE
 (.993) (1.714)
 + .002DPINC + 3.837MEDST
 (4.027) (2.722)
 + .852SINGL + 6.140WESTV
 (5.833) (3.285)
 - .993UNEMP + 7.429DRATE
 (3.312) (2.200)
 - .165HIGHS,
 (1.517)


[R.sup.2] = 87.4%, [R.sup.2] = 85.0%

A Goldfeld-Quandt procedure was used in order to test if the variance was lower for states with smaller populations of women aged 15-44. The results indicate that the null hypotheses of homoscedasticity cannot be rejected at both the 5 and 1 percent significance levels. In addition, a plot of the fitted value of ARATE against the residuals squared shows that the size of the residual is also not related to the size of the fitted value of the abortion rate.

The signs of the variables in the estimated abortion demand equation coincide with expectations. PRICE has a negative impact on abortion demand. It can be concluded at the 10 percent significance level that PRICE has a significant effect on abortion demand.(28) As PRICE rises by $40 (about one-fifth of its mean value), the abortion rate falls by 1.64 percent on average. The low significance level and degree of impact support the hypothesis that price will not have a large effect on abortion demand due to the fact that many facilities offer reduced costs for abortion services for women with low incomes.(29) The calculated price elasticity of demand computed at the sample mean is - .348. This suggests that abortion demand is relatively inelastic with respect to price and coincides with the fact that the demand for medical services has been found to be inelastic, (in Nicholson, 1985, p. 185-7).

It can be concluded at the 1 percent significance level that the variable DPINC, disposable personal per capita income, does have an affect on abortion demand. As predicted income has a large positive effect on the demand for abortions. As income rises by $2000 (about one-fifth of its mean), abortion demand rises by 4 percent on average. This large impact is probably due to the combined forces of an income effect, the fact that as income rises the woman's time is more valuable, and that if a household has a higher income, its demand for quantity of children falls. The income elasticity of demand calculated at the sample mean is .810 which supports the assumption that abortion is a normal good.(30)

States with Medicaid funded abortions will have a higher abortion rate than those without. It can be concluded at the 1 percent significance level that MEDST has an impact on abortion demand. If a state has an unrestrictive policy on abortion funding, it will have an abortion rate that is 3.837 percent higher than other states on average. This supports the hypothesis that if price is not a constraint for a low-income woman, she will be more likely to be able to obtain an abortion if she chooses. It can also be concluded at the 1 percent significance level that percentage of unmarried women in the state will have an impact on abortion demand. As expected, the impact of SINGL is large and positive. As SINGL rises by 8 percent (about one-fifth of its mean), the abortion rate rises by 6.816 percent on average. The results support the ideas that single women are under a larger economic constraint and thus can less afford to have another child and that a single woman's time has a relatively higher value than a married woman's (since she must support herself) and thus she will demand less children.(31)

States located in the far West have a higher abortion rate than those that are not. It can be concluded at the 1 percent significance level that WESTV has an effect on abortion demand. If states are located in the far West, they will, on average, have an abortion rate that is 6.140 percent higher. This supports the theory that states in the far West have different social norms, ties, etc. At the 1 percent significance level it can also be concluded that the unemployment rate, UNEMP, has an effect on abortion demand. As the unemployment rate rises by 1.4 percent (about one-fifth of its mean value), the abortion rate falls by 1.390 percent on average. The results suggest that a combination of the following hypotheses may have occurred: An income effect may occur so that as unemployment rate rises households are able to purchase less goods and services. As unemployment rate rises, the value of a woman's time falls thus the cost of a child falls, and abortion rate falls.(32) And, another theory which seems very plausible is that at times of high unemployment and economic uncertainty, people are more careful about their contraceptive use, thus abortion rate falls as does the birth rate.(33)

DRATE, the divorce and marriage annulment rate, has an expected positive impact on the abortion rate. It can be concluded at the 5 percent significance level that DRATE affects abortion demand. AS DRATE rises by .1 percent (about one-fifth of its mean), the abortion rate rises by .743 percent on average. The overall effect on abortion demand is not as large as expected. This may be because the yearly divorce rate affects a small percentage of the women in the entire population. however, the fact that it has a significant positive effect supports the idea that in periods of turmoil caused by divorce and marriage annulment, the woman may be more likely to choose to terminate he pregnancy. And, that a woman is more likely to depend on her income once divorced, thus the value of her time rise which causes the cost of an additional child to rise, which in turn causes demand for children to fall. HIGHS, the percentage of women aged 15 and over with twelve years of education or more, has an insignificant impact on abortion demand.(34) It may be that of the effects which Michael and Becker suggest, neither dominate. However, the fact that the sign is negative suggests that perhaps an effect Michael hypothesizes is stronger.(35) More educated women have more knowledge about contraceptive methods, thus causing abortion demand to fall.

The value of the [R.sup.2] suggest that 87.4 percent of the variation in the abortion rate can be explained by the estimated abortion demand equation. Taking into account the degrees of freedom, the adjusted-[R.sup.2] is equal to 85.0 percent. The effect of price coincides with that found by Medoff (1988) and Deyak and Smith (1976) as does the insignificant impact of education on abortion demand. In addition, the direction of the impact of income, unmarried women, far Western states, and medicaid states also coincides with the results of Medoff's abortion demand equation. Neither Medoff nor Deyak and Smith consider the unemployment and divorce rate variables in their papers.

Other relevant variables were considered before the final abortion demand equation was decided upon. The equation was estimated with the percentage of nonwhites in each state and the District of Columbia in 1984.(36) The Alan Guttmacher Institute points out that the nonwhite abortion rate is higher than that for whites. However, after the other variables (such as income) are controlled for, the percentage of nonwhites has an insignificant impact on abortion demand. Becker (1981) suggests that aid to mothers with dependent children reduces the costs of a child, (p. 97). This may reduce desire in poor women to obtain abortions. The abortion demand equation was estimated with the percent of AFDC recipients in each state and the District of Columbia in 1985.(37) Ideal data would be the percentage of women aged 15-44 receiving AFDC in 1985, but it is most likely that women in this age group are a high percentage of the recipients, (Becker, 1981, p. 97). The results suggest that the percentage of AFDC recipients does not have an impact on the abortion rate.(38) this suggests that AFDC does not act, n general, as an incentive for women to have children. In both cases, the value of the estimated coefficients of the variables remained virtually the same as in equation (1). Thus the empirical results remain robust with respect to the addition of percentage of nonwhite population and percentage of AFDC recipients.

Becker (1981) also suggests that as the female labor force participation rate rises (and the earning power of women), the demand for children will fall, (p. 98). The abortion demand equation was estimated using the labor force participation rate of women in the fifty states and the District of Columbia in 1985.(39) The results suggest that the impact is negative, although it could not be concluded that the effect of the variable was significantly different from zero. This might have resulted because the effect of the variable is captured by the unemployment rate and the education level, both of which it is highly correlated with. However, the Alan Guttmacher Institute states that in the 80's the percentage of women in the labor force with newborns is consistently rising, (1989, p. 37). This suggests that more and more women in the work-force are having children; thus indicating that though women in the labor-force may demand less children because of the value of their time, this may be somewhat offset by the increasing trend of women both having children and staying in the labor force.(40) The values of the estimated coefficients remained virtually identical to the ones in equation. (1).

The Roman Catholic Church greatly opposes abortion and therefore might influence abortion demand, (Rodman, Sarvis, and Bonar, 1987, p. 138). The equation was estimated using the percentage of Catholics in each of the fifty states and the District of Columbia for the year 1980.(41) It can be assumed that the state-by-state percentage of Catholics remained relatively the same in 1985. The variable was found to be just under the 10 percent significance level and the impact was very small, but positive. This coincides with Medoff's (1988) results and suggests that Catholics do not uniformly support the stance of their church. The results also confirm the findings of the 1983 Gallup Poll in which 22 percent of Catholics state that they support legal abortion under any circumstances and 57 percent state that they support legal abortion under certain circumstances, (p. 142). The results indicate that the values of the estimated coefficients remain virtually the same as in equation (1).

Medoff (1989) argues that evangelical Christians (fundamentalists) are unified in their opposition to abortion, (p. 187). The abortion demand equation was estimated using the percentage of fundamentalists in each state and the District of Columbia in 1980. Heatwole's (1978) definition of "Bible belt denominations" which are mostly fundamentalist was used in order to calculate the percentage of fundamentalists. (p. 52). The results suggest that the variable is not statistically different from zero below the 10 percent significance level, although it does have the expected negative sign. There is reason to believe that the percentage of fundamentalists in each state will not strongly affect abortion demand: the Alan Guttmacher Institute states that "One out of every six women who have an abortion describes herself as an evangelical Christian," (in Salholz and McDaniel, 1989, p. 31).

Based on past research, economic theory, and the total empirical findings, equation (1) appears to be the |best' model for explaining abortion demand.

Part IV: Possible Future Research

Considerations and Conclusion

Because there has been very little economic research on abortion demand, there are many possibilities for further research. One reason that the abortion rate measurement is imprecise for the analysis done in this paper is because it does not reflect the actual abortion rate for state residents. This type of data is available from the Alan Guttmacher Institute for certain years (most recently 1982). An interesting comparison of the magnitude of the effects of the variables explaining abortion demand for a particular year could be made by estimating one abortion demand equation using the abortion rate per state and one using the resident abortion rate. A similar type of research could be done using a fertility control theoretical framework in which an equation for abortion demand is estimated, and the same factors are used to estimate a contraceptive demand equation.(42) The magnitudes of the factors could be compared and would help to indicate if contraceptive and abortion behavior are closely linked.

A comprehensive analysis of state-by-state abortion demand ranging from 1973 when abortion was made legal in the Roe v. Wade decision until the most current year in which data on the abortion rate is made available would be interesting and useful. The analysis could trace how the magnitude of factors affecting abortion demand has altered since 1973. The study could also investigate if the 1983 United States Supreme Court decision made on the City of Akron v. Akron Center for Reproductive Health, Inc., et al. had an effect on abortion demand in the states affected. In addition, the analysis could attempt to trace the impact of harassment of abortion providers on abortion demand since 1973.

The Roe v. Wade decision has currently been challenged and the Supreme Court will most likely decide its fate during the summer of 1989. If Roe v. Wade is overturned, the states will be in control of the status of abortion. What this suggests is that unrestricted abortions will only be legal in certain states. This implies that women who can afford the time and travel costs will be able to travel to neighboring states and obtain abortions. Deyak and Smith (1976) state that this was the situation before the 1973 decision. However, for those who cannot afford the costs, the alternative to a legal abortion may be either an illegal abortion or carrying to term an unwanted pregnancy. Thus no matter what decision the Supreme Court makes concerning the status of Roe v. Wade, abortion will continue to be demanded as a form of fertility control.

Notes

(1.) The results of the analyses are published in a magazine entitled Family Planning Perspectives as well as in their publication Abortion Services in the United States, Each State and Metropolitan Area. (2.) Coelen and McIntyre (1978) researched pronatalist and abortion policies in Hungary using a time-series regression technique and an economic theoretical framework. (3.) The purpose of this 1983 amendment was to overturn the Roe v. Wade decision. (4.) At this time New York, Alaska, and Hawaii were the only states which had legalized abortion. (5.) The economic model of fertility control he uses is developed by Michael (1973). (6.) The absolute value of the t-statistics are in parenthesis, (p. 356). Medoff's empirical results have been included in this paper because (1) his work has influenced the abortion demand model developed in this paper and (2) the results can thus be conveniently compared to those of the abortion demand model presented in this paper. "The mean and standard deviation (in parentheses) for the dependent and explanatory variables ... are A: 250.898 (87.847); P: 213.64 (43.15); Y: 6407.2 (936.338); SNGL: 34.158 (3.956); LFP: 61.99 (4.602); CATH: 20.022 (13.806); w: .12 (.3249); M: .28 (.0448)," (p. 358). (7.) Michael also presents ideas on the effects of education on fertility control which will be brought up later in the paper. It should be noted that Michael draws on past research by Becker in order to develop his model of fertility control. (8.) In this situation, a household decision can be interpreted as being made by either a couple or a single pregnant woman. (9.) Pregnancies - Live Births + Abortions. If the number of stillbirths and miscarriages could be included, a better measure of the number of pregnancies could be obtained. The abortion rate data is from The Statistical Abstract of the United States, 1988. In 1985, "... about 30 percent of pregnancies (excluding those ending in stillbirths and miscarriages) were terminated by abortion," (The Alan Guttmacher Institute, 1988, p. 15). (10.) Deyak and Smith (1976) consider the rate of abortions per live births, and Medoff (1988) considers all three of the measurements of the abortion rate, although the model he depicts in his paper utilizes abortion rate per 1000 pregnancies. (11.) The model utilizes characteristics of women in each state in order to determine factors affecting abortion demand. (12.) Another decision the Supreme court made in this case was that overly detailed information (concerning the fetus and abortion procedure) which was often used to put pressure on the woman seeking an abortion was declared unconstitutional, (Rodman, Sarvis, and Bonar, 1987, p. 123). (13.) The average price is as of mid-1986, but is determined by the 1985 caseload. Since the average abortion price changes relatively slow and price ranges are used, it can be assumed to be nearly the same for 1985. (The Alan Guttmacher Institute, 1988, p. 21). (14.) The most recent available data on the percentage of abortions performed after the first trimester is from 1983, (The Alan Guttmacher Institute, 1988, p. 20). (15.) This assumption is also made by Deyak and Smith (1976). Medoff (1988) does estimate a supply equation by having price determined by average hospital cost per day, number of hospitals, number of abortion clinics, number of physicians per 100,000 population in each state, as well as the abortion rate, states, in the far West, and states with medicaid funded abortions. (16.) This also suggest that the variable PRICE is imprecise because the listed price does not always reflect the price charged. (17.) Data is obtained from the Statistical Abstract of the United States, 1988. (18.) As of June 4, 1985, Colorado adopted a restrictive policy on abortion funding. (The Alan Guttmacher Institute, 1988, p. 37). However, because abortions are state funded for qualifying low-income women for about half of 1985, Colorado is given a value of 1. (19.) Data is from the State and Metropolitan Area Data Book. Ideal data would be the percentage of single women in each state and D.C. in 1985 aged 15-44. (20.) For a married woman, the cost of having a child is less because she can share the actual cost of having a child with her husband, and both can share in the time cost of raising the child. (21.) Medoff cites Hamermesh and Soss (1974) in making this point. (22.) Data is from the Statistical Abstract of the United States, 1987. Sixteen is the legal employment age. (23.) This is assuming that a couple are making an abortion decision, and that the woman's income was one of two in the household. (24.) Data is obtained from the Statistical Abstract of the United States, 1988. The rates are preliminary, and for states where the data was not yet available (Indiana, Louisiana) a previous year's data was used. (25.) Data is obtained from State Reports, Detailed Characteristics, 1980. (26.) Birth control does fail, and it could be argued that an increase in education increases a woman's knowledge about the availability of abortion, and thus increases the abortion rate. (27.) The absolute value of the t-statistics are in parentheses. The mean and standard deviation (in parentheses) for the dependent and explanatory variables are as follows: ARATE: 25.690 (9.160), PRICE: 218.137 (24.264), DPINC: 11387.255 (1778.701), MEDST: .314 (.469), SINGL: 43.963 (5.138), WESTV: .118 (.325), UNEMP: 7.110 (1.936), DRATE: .518 (.181), HIGHS: 64.716(6.414). (28.) Equation (1) was also estimated using dummy variables for PRICE, the results were not statistically better. (29.) The relevant price paid is not always observed on the list of fees. (30.) Both the income and price elasticities have the same signs as those estimated by Medoff (1988). In addition, the results from the abortion demand equation are relatively consistent with the income elasticity of demand for medical services in Nicholson (1985, p. 185). (31.) In 1985 there were 1785 children ever born per 1000 women ever married, and 285 children ever born per 1000 single women, (Statistical Abstract of the United States, 1987). (32.) When the equation was estimated using the female unemployment rate, the results were qualitatively the same, and the female unemployment rate had a negative significant effect on abortion demand. (33.) During the Great Depression of the 1930's, a time of massive unemployment, the birth rate greatly declined, (Historical Statistics of the United States Colonial Times to 1970). (34.) The percentages of women aged 15 and over with four years of college or more was also found to be insignificant. (35.) As HIGHS rises by 13 percent (about one-fifth of its mean value) the abortion rate falls by 2.145 percent on average. (36.) Data is from the County and City Data Book 1988. It can be assumed that the state-by-state percentage of nonwhites is relatively the same for 1985. (37.) Data is from the Statistical Abstract of the United States, 1987. (38.) The results of the nonwhite and AFDC variables are consistent with Medoff's (1988). Percentage of families with female headed households was also found to be insignificant. (39.) The data is from the Statistical Abstract of the United States, 1987. (40.) The results disagree with Medoff (1988) who found labor force participation rate to be significant. However, his data is for 1980 which may have been before the impact of this trend became so large. (41.) Data is from Churches and Church Membership in the United States 1980 as is data for the percentage of fundamentalists in each state and the District of Columbia. (42.) Excluding abortion.

References

Abortion Services in the united States, Each State & Metropolitan Area, 1984-1985.ed. S. Henshaw and J. Van Vort. New York: The Alan Guttmacher Institute, 1988. Becker, Gary S. "The Demand for Children," A Treatise on the Family. Massachusetts: Harvard University Press, 1981. 93-112. City of Akron v. Akron Center for Reproductive Health, Inc., et al. 462 US 416 (1983). Coelen, Stephen P. and Robert J. McIntyre. "An Econometric Model of Pronatalist and Abortion Policies." Journal of political Economy. December 1978. vol. 86. no. 6. 1077-1101. Deyak, Timothy A. and V. Kerry Smith. "The Economic Value of Statute Reform: The Case of Liberalized Abortion." Journal of Political Economy. 1976. vol. 84. no. 1. 83-99. The Gallup Poll Public Opinion 1983. Delaware: Scholarly Resources Inc., 1984. 142. Grimes, David A. "Second-Trimester Abortions in the United States." Family Planning Perspectives. November/December 1984. vol. 16. no. 6. 260-5. "Half of all U.S. Women with Newborns are in the Labor Force." Family Planning Perspectives. January/February 1989, vol. 21. no. 1. 37-8. Hamermesh, Daniel S. and Neal M. Soss. "An Economic Theory of Suicide." Journal of Political Economy. January/February 1974. vol. 822. 83-98. Heatwole, Charles A. "The Bible Belt: A Problem in Regional Definition." Journal of Geography. February 1978, vol. 77. 50-5. Henshaw, Stanley. "Characteristics of U.S. Women Having Abortions, 1982-1983." Family Planning Perspectives. January/February 1987. vol. 19. no. 1. 5-9. Henshaw, Stanley and Jane Silverman. "The Characteristics and Prior Contraceptive Use of U.S. Abortion Patients." Family Planning Perspectives. July/August 1988. vol. 20. no. 4. 158-68. Legge, Jerome S. Abortion Policy-An Evaluation of the Consequences for Maternal and Infant Health. Albany: State University of New York Press, 1985. Medoff, Marshall H. "Constituencies, Ideology, and the Demand for Abortion Legislation." Public Choice. 1989. vol. 60. 185-91. Medoff, Marshall H. "An Economic Analysis of the Demand for Abortions." Economic Inquiry. April 1988. vol. 26. 353-9. Michael, Robert T. "Education and the Derived Demand for Children." Journal of Political Economy. March/April 1973.vol. 81. S128-64. Nicholson, Walter. Microeconomic Theory. third ed. New York: Holt, Rinehart and Winston, Inc., 1985. 185-7. Powell-Griner, Eva and Katherine Trent. "Sociodemographic Determinants of Abortion in the United States." Demography. November 1987. vol. 24. no. 4. 553-61. Public Law 95-205 (1977) in United States Statutes At Large, 95th Congress, 1st Session, 1977. vol. 91. Washington D.C.: U.S. Government Printing Office, 1980. Quinn, B., H. Anderson, et al. Churches and Church Membership in the United States 1980. Georgia: Glenmary Research Center, 1982. Rodman, Hyman, Betty Sarvis, and Joy Walker Bonar. The Abortion Question. New York: Columbia University Press, 1987. Roe v. Wade. 410 US 113 (1973). Rubin, Eva. Abortion, Politics and the courts Roe v. Wade and Its Aftermath. Contributions in American Studies, no. 57. Connecticut: Greenwood Press, 1982. Salholz, Eloise, Ann McDaniel, et al. "The Battle Over Abortion." Newsweek. May 1, 1989. 28-32. Sollom, Terry and Patricia Donovan. "State Laws and the Provision of Family Planning and Abortion Services in 1985." Family Planning Perspectives. November/December 1985. vol. 17. no. 6. Sommers, Paul M. and Laura S. Thomas. "Restricting Federal Funds for abortion: Another Look." Social Science Quarterly. June 1983. vol. 64. no. 2. 340-6. Trussel, James, Jane Menken, et al. "The Impact of Restricting Medicaid financing for Abortion." Family Planning Perspectives. May/June 1980. vol. 12, no. 3. 120-130. U.S. Bureau of the Census. County and City Data Book 1988. U.S. Department of Commerce. Washington, D.C.: U.S. Government Printing Office, 1988. U.S. Bureau of the Census. Historical Statistics of the United States Colonial Times to 1970. U.S. Department of Commerce. Washington, D.C.: U.S. Government Printing Office, 1975. U.S. Bureau of the Census. State Reports, Detailed Characteristics: 1980. Washington, D.C.: U.S. Government Printing Office, 1983. U.S. Bureau of the Census. Statistical Abstract of the United States. Washington, D.C.: U.S. Government Printing Office.
COPYRIGHT 1992 Omicron Delta Epsilon
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Rothstein, Donna S.
Publication:American Economist
Date:Mar 22, 1992
Words:8367
Previous Article:The efficient tariff: systematically balancing security and welfare concerns.
Next Article:The effects of monopolization on newspaper advertising rates.
Topics:


Related Articles
A note on "An Economic Approach to Abortion Demand." (response to Donna S. Rothstein, The American Economist, vol. 36, no. 1, p. 53)
Responding to Economic Arguments for Abortion What Do 40 Million Lost Lives Mean?
PRO-LIFE GRASSROOTS WORK MOVES ABORTION POLLS.
The EU and the death ethic. (Insider Report).
Using the right to life to confront unsafe abortion in Africa.
Convincing new providers to offer medical abortion: what will it take? (Viewpoints).
The economics of sexual health.
Making abortion rare.
Complex problems, comprehensive interventions.
Abortion in Chile: contributions to the debate from women's perspective.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters