Back to Journal Cover Page
Volume 5, Number 20
Submitted: August 25, 2000
Resubmitted: September 23, 2000
Accepted: September 23, 2000
Publication date: September 26, 2000
Department of Human Development
Undergraduates were given three hypothetical moral dilemmas in questionnaire format. Responses were coded as being indicative of either consequential or non-consequential reasoning. Consequential moral decisions are based on the consequences of an act. Non-consequential decisions are based on factors other than the consequences, such as moral principles or rules. Significant gender differences were found. Males were much more likely than females to choose the consequential response to two of the three dilemmas. Also, all of the participants chose the non-consequential response to a much higher degree than was expected. The results suggest that moral judgment is less rational and consistent than many theorists contend, and that different people especially the opposite sexes make moral judgments in very different ways.
There are two basic ways in which people make moral decisions (Lapsley, 1997). In the first method, people decide what is morally right by gauging the consequences of each decision. So for example, if choosing Action A results in the best consequences for everyone involved, then Action A is the morally right choice. In the second method, people decide what is morally right by applying some socially defined template of moral rules for behavior. So if Action A would entail breaking a rule while Action B abides by the rule, this second philosophy would recommend Action B as the morally right choice, even if the consequences of Action B are not as good as the consequences of Action A. Traditionally, the philosophy that assesses the morality of an act by gauging its consequences has been called teleological reasoning. The philosophy that assesses the morality of an act by applying a rule, principle or ethic has been called deontological reasoning (Frankena, 1973). Ill simply refer to these two opposing philosophies, respectively, as consequentialism and non-consequentialism.
Are humans consequentialists or non-consequentialists? Obviously, the answer is both. Sometimes we look at the consequences and sometimes we apply a rule. Usually we do both. This is part of the reason why moral reasoning is so complicated. But what do we do when these two types of reasoning conflict? For example, what if Action A has the best consequences, but it requires breaking a rule? Philosophers often demonstrate this type of conflict by posing a moral problem such as the "trolley dilemma." In this scenario, a trolley is rolling down the tracks at full speed. Two people have become stranded on the tracks, and it is too late to merely stop the trolley. As the track controller, you could derail the trolley, killing the trolley conductor but saving the lives of the two people on the tracks. What do you do?
The conflict is clear. The consequences of derailing the trolley is one death, while the consequences of not derailing the trolley is two deaths. Therefore, if Action A is to derail, then the consequentialist would say that Action A is the right choice. However, derailing the trolley is an act of killing, because it is your action of derailing which will kill the conductor. Since all human societies have some form of restriction against killing innocent people, the choice to derail would be breaking a cardinal rule. Therefore, if Action B represents the choice to not derail, then the non-consequentialist would say that Action B is the right choice.
Clearly, this dilemma is not quite as simple as it seems. For instance, there are rules against doing nothing when other peoples lives are at stake. Also, some people may question whether it is possible to equate the value of human life with the number of human lives involved. Despite these problems, the trolley dilemma could still elicit responses that may answer some pertinent questions regarding moral reasoning. For example, are there gender differences between the conflicting sides of consequentialism and non-consequentialism? Is there a point where consequential concerns override non-consequential restrictions? If so, where does this point arise?
The present study was designed to investigate these questions. Undergraduate students were given the trolley dilemma in questionnaire format. Two similar dilemmas were also administered to test for consistency. The primary hypothesis was that there would be significant gender differences in the response sets. This assumption stems from the theoretical assertion that males and females use different styles of moral judgment (Gilligan, 1982, Holstein, 1976, Chodorow, 1989, Eisenberg, 1982, Ford & Lowery, 1986, Hoffman, 1977).
According to Jean Piaget (1932), moral judgment is derived from the childs understanding of conventional rules that eventually become internalized into an autonomous respect for law and order. Lawrence Kohlberg (1969) expanded upon Piagets work by adding higher stages of moral development. The highest stage of development according to Kohlbergs model is reached when the individual surpasses the confines of "conventional" rules and can make impartial moral decisions based on "post-conventional" principles of justice.
Gilligan (1982) criticized Kohlbergs paradigm as being gender biased. This claim was sparked by Holsteins (1976) longitudinal study in which female respondents were typically scored at stage 3, whiles males were typically scored at stage 4. Holstein explained that these scores were a result of gender differences in moral reasoning. She argued that females were scored at stage 3 because their decisions were influenced by empathy and emotion, while males were scored at stage 4 because their decisions were less empathic and more impartial and detached. Holsteins main complaint was that womens reasoning styles were arbitrarily devalued by Kohlbergs scoring system because it was considered less developed that mens reasoning styles. "Emotional response to moral conflict which is exemplified by females more than males results in adult female reasoning categorized with children" (Holstein, 1976, p.61).
Gilligan (1977/1982) used Holsteins findings as a basis for her theory of gender differences in ego and moral development. She claimed that men are more "justice" oriented while women are more "care" oriented. A justice orientation is motivated by logic and reason, requiring the moral actor to treat others impartially and objectively and basing moral decisions on abstract principles of justice that can be universalized to every person and every situation. A care orientation allows the actor to use subjective feelings and sentiments when making moral judgments. The caring decision is motivated by empathy. It recognizes particular relationships between people and extenuating circumstances in each situation.
Both Gilligan (1977, 1982) and Holstein (1976) criticized Kohlbergs (1969) original research because he studied males only. They argued that this "andocentric" focus in his research resulted in a paradigm of development that was appropriate for male moral reasoning, but not for female moral reasoning. Furthermore, Gilligan claimed that Kohlbergs model rates individuals who function according to an ethic of care at a lower stage of reasoning, because these people make decisions based on special obligations and feelings rather than on impartial principles of abstract justice. The "caring" persons judgment is concerned with maintaining ongoing relationships between people rather than making judgments that are indicative of abstract and objective impartiality. An ethic of care seems biased and therefore less developed within the paradigm of the justice orientation. Gilligan argued that if the tables were turned, an ethic of justice would seem cold, calculating and heartless within the paradigm of a care orientation.
Kohlberg and his colleagues (1983 & 1984) response to Gilligans criticism argued that justice and care are not mutually exclusive orientations, but rather, they represent two different forces that influence moral judgments simultaneously. Gilligan and colleagues (1987) rebutted, arguing that although the two orientations are not exclusive, men "gravitate" towards a justice orientation and women "gravitate" towards a care orientation. Although Gilligan claimed support for her theory through the open-ended moral interviews that she conducted with women, most of the empirical research investigating gender differences in moral reasoning has not confirmed her ideas, (Walker, 1984, & 1987, Rest, 1983 & Ford & Lowery, 1986). Her reply to these contradictory findings was that the subtle differences between an ethic of care and an ethic of justice could not be drawn out via objective empirical testing. These differences can only be elicited within the context of an open-ended interview process. Unfortunately, this form of testing presents an obvious problem with reliability, in that the subjects responses are not uniform, and the coding of participant responses rely to a significant degree upon the interviewers subjective interpretations.
Given the absence of any uniform objective coding scheme, it is impossible to discern if womens moral development actually conforms to Gilligans (1982) developmental sequence for her "care" paradigm. Furthermore, Holsteins (1976) findings of gender differences in her studies which used Kohlbergs rigorously tested and objective scoring system were not replicated in later studies (Lapsley, 1997). In Walkers (1984) meta-analysis of 108 studies, there were only 8 clear indications of sex differences, but even in those cases, the effects were confounded by other sociological and scoring factors and the differences themselves were very small less than half a stage. Walker concluded that "rather than arguing over the extent to which sex bias is inherent in Kohlbergs theory of moral development, it might be more appropriate to ask why the myth that males are more advanced than females persists in light of so little evidence" (as cited in Lapsley, 1997, p.139). In addition, significant sex differences were not found in studies using alternate tests and measures, such as Rests "Definition of Issues" test (Rest, 1979, 1983 & Walker 1987, 1989).
Though these theories are debatable, the fact that there are differences between the moral behavior of men and women is plainly observable. Men are nine times more likely than women to commit murder, and 92.5% of all prison inmates are male (U.S. Department of Justice, 1999). There does seem to be face validity in the notion that women and men approach moral decisions in different ways. However, in order to prove the existence of gender differences in moral reasoning, a model for objective and empirical testing must be devised.
A second hypothesis in the present study was that there would be a certain point where the majority of responses will switch from non-consequential to consequential choices. In order to elicit this effect, the trolley dilemma was administered in six different forms. Each form had a different number of people stranded on the tracks. It was hypothesized that almost nobody would derail when only one person was on the track, but that nearly everybody would derail when there were two or more people on the tracks. This hypothesis was based on the assumption that when the ratio of consequences is 1:1, (that is, 1 stranded person to 1 conductor), the rule against killing would stop people from choosing the derail option. But when the ratio of consequences is 2:1 or higher, (more people on the tracks than in the trolley), then the utilitarian reasoning which saves the greatest number of human lives will take precedence.
Participants were 366 female and 125 male undergraduates at a large research university. The students earned extra credit in their respective classes in return for their participation. The students responses were gathered over the course of the 1999-2000 academic year. Participants were presented with dilemmas in questionnaire format and responded by circling one of two choices for each dilemma.
Participants were presented with hypothetical moral dilemmas in questionnaire format and asked to choose between one of two responses. The consequential implications for each response were pointed out explicitly i.e. "derailing the trolley would result in the death of the trolley conductor while saving the lives of the people stranded on the tracks." Three hypothetical moral dilemmas were used to elicit moral judgments:
The Trolley Dilemma presents a hypothetical situation in which the participants imagine that they are the track controllers for a trolley system. In the dilemma, people have become stranded on the tracks while a trolley is approaching. The participant must decide between the consequential choice of derailing the trolley, which would result in the death of the conductor while saving the lives of the people on the tracks, and the non-consequential choice of not derailing the trolley. Six different forms of the dilemma were administered: Form A had one person on the tracks, Form B had two people, Form C had five people, Form D had eight people, Form E had ten people and Form F had twenty people. A larger number of participants were given Form B (the dilemma in its original format with 2 people on the tracks), in order to test for a main effect of gender.
In the "bridge" dilemma, the participant is assigned the role of a general in a country which is being invaded by an enemy army. If the enemy army reaches the bridge leading to the capital city, the city and its inhabitants will be attacked and captured. However, there are civilians crossing the bridge. The consequential choice is to blow up the bridge, sacrificing the crossing civilians but saving the city. The non-consequential is to not blow up the bridge.
In the factory dilemma, a corporation's factory has been losing money for years. As the CEO, the participant must decide whether or not to shut down the factory. The consequential choice of shutting down the factory would result in the laying off of a thousand workers. The non-consequential choice of keeping the factory open will result in the continued loss of money by the corporation.
A chi-square analysis was calculated for all of the responses. Tables 1 6 display the proportion of responses by gender to the trolley dilemma. The only significant difference between male and female responses was found in the response sets to Form B of the trolley dilemma, which had two people on the tracks (see Table 2). 26.76 % more males chose to derail the trolley than females. This difference was highly significant (p < 0.0001).
Table 7 displays the overall proportion of responses to the trolley dilemma. As predicted, the proportion of participants who chose to derail the trolley tended to increase as the number of people stranded on the tracks increased. The largest shift in the proportion of participants choosing to derail was between Forms B and C. In Form B, in which two people were on the tracks, 31.64 % of the participants chose to derail. In form C, in which five people were on the tracks, 80.49 % of the participants chose to derail. This shift of 48.85 % was by far the largest shift in response proportions. It was 3.5 times larger than the shift between Forms A and B (13.86 %), and it was more than two times greater than the differences between Forms C and D, Forms D and E and Forms E and F combined (1.09 % + 6.68 % + 12.60 % = 20.37 %).
Tables 8 and 9 display the overall proportion of responses to the bridge and factory dilemmas. In the responses to the bridge dilemma, a higher proportion of women chose the non-consequential response of "not blowing up the bridge" (74.90 %) as compared to men (60.00 %). This difference was highly significant (p = .0039). The factory dilemma yielded no statistically significant gender difference in non-consequential responses (m = 77.80 %, f = 75.40 %, p = 0.7718).
Table 1. Form A (One person on the tracks).
|Derail||28.57 %||15.79 %||12.78 %||0.43|
|Do not derail||71.43 %||84.21 %|
|N||7||38||Total N = 45|
Table 2. Form B (Two people on the tracks).
|Derail||51.39 %||24.63 %||26.76 %||0.00|
|Do not derail||48.61 %||75.37 %|
|N||72||203||Total N = 275|
Table 3. Form C (Five people on the tracks).
|Derail||80.00 %||80.65 %||0.65 %||0.97|
|Do not derail||20.00 %||19.35 %|
|N||10||31||Total N = 41|
Table 4. Form D (Eight people on the tracks).
|Derail||81.82 %||81.48 %||0.34 %||0.98|
|Do not derail||18.18 %||18.52 %|
|N||11||27||Total N = 38|
Table 5. Form E (Ten people on the tracks).
|Derail||70.02 %||76.47 %||6.45 %||0.69|
|Do not derail||29.08 %||23.53 %|
|N||10||34||Total N = 44|
Table 6. Form F (Twenty people on the tracks).
|Derail||80.00 %||90.91 %||10.91 %||0.30|
|Do not derail||20.00 %||9.09 %|
|N||15||33||Total N = 48|
Table 7. Overall Responses to the Trolly Dilemma
|Response||Form A||Form B||Form C||Form D||Form E||Form F|
|Derail||17.78 %||31.64 %||80.49 %||81.58 %||74.90 %||87.50 %|
|Do not derail||82.22 %||68.36 %||19.51 %||18.42 %||25.10 %||12.50 %|
|Total N = 491|
Table 8. Overall Responses to the Bridge Dilemma
|Blow up||40.00 %||25.10 %||14.90 %||0.00|
|Do not blow up||60.00 %||74.90 %|
|N||125||366||Total N = 491|
Table 9. Overall Responses to the Factory Dilemma
|Shut down||22.20 %||24.60 %||2.40 %||0.77|
|Do not shut down||77.80 %||75.40 %|
|N||125||366||Total N = 491|
In the original form of the trolley dilemma, in which 2 people were stranded on the tracks, males were much more likely to derail the trolley than females. This male preference for the consequential choice was repeated in the responses to the bridge dilemma. There are many ways to explain these highly significant results. One explanation is that males are more aggressive than females (Lorenz, 1966, Eibesfeldt, 1972, Freud, 1961), so they are more likely to choose the aggressive actions of derailing the trolley and blowing up the bridge rather than the passive action of doing nothing.
Another possibility is that males might be more likely to associate a sense of duty with the position of being a train conductor or general, because men have traditionally been set in the role of soldiers, and because train conducting is a stereotypically male position (Maccoby & Jacklin, 1974). Therefore, men might be more inclined to perceive the conductor as a soldier or policeman-like character who understands the risk of his/her position and must accept the consequences accordingly. Women, on the other hand, might be less likely to associate the conductors position with a sense of duty. They may perceive the conductor as just another innocent bystander, who doesnt deserve to die any more than the people stranded on the tracks. Though these possibilities are merely conjecture, the significant gender differences in the responses to this dilemma are intriguing and seem to beg for explanation.
The second significant finding was in the shift of the proportion of people who chose to derail the trolley. When there were only 2 people on the tracks, less than 32% of the participants chose to derail. But when there were 5 people on the tracks, more than 80% chose to derail. This shift of nearly 49% was much larger than any of the other shifts, even the shift between Form E (10 people on the tracks) and form F (20 people on the tracks). This big shift is somewhat counter-intuitive, because it was predicted that the largest shift would come between the form in which 1 person was stranded on the tracks and the form in which 2 people were stranded. It was assumed that the vast majority of participants would automatically favor the saving of 2 people over the saving of just 1. However, the majority of participants did not make the purely utilitarian choice to derail the trolley until Form C, in which the consequences were 5 to 1 against the conductor. Apparently, non-consequential reasoning is a very strong factor in moral judgment.
Before discussing a human phenomenon, one must first confirm that the phenomenon actually exists. If only a small percentage of respondents did not choose to derail the trolley with 2 people on the tracks, then the phenomenon of non-consequential reasoning would be demonstrated. The fact that the vast majority of respondents (68.36 %) did not choose the "derail" response illustrates undeniably that not only is non-consequential reasoning a factor in moral judgment, it is often a more powerful factor than consequential reasoning.
Though contemporary psychologists and sociologists tend to portray the human mind as a rational machine, we must admit that human behavior frequently belies any true logic or reason. People are essentially irrational motivated by emotions, ungrounded beliefs, absurd convictions and inconsistent principles. Irrationality and inconsistency is seen most evidently in moral reasoning and behavior. People kill abortion doctors because they believe killing is wrong. People fight for the right to own guns for their own protection, though the largest danger they face is from their own guns. We cannot pretend that humans are consistent in their moral judgments and that they could be classified into neat little categories. Theories which attempt to explain moral reasoning using purely rational models are at best reductionist and, at worst, simply wrong.
Neither can it be assumed that all humans develop their moral senses in the same way. The present study lends support to this assertion. The large gender differences in response to the trolley dilemma in its original form suggests that men may be more influenced by purely consequential factors, while women may be more inclined to take non-consequential factors into account. Though the gender differences were highly significant in the responses to the dilemma in its original form, no significant differences were found in the responses to the other forms of the dilemma. So while it may be possible to elicit gender differences in certain far-fetched hypothetical moral dilemmas, for the most part, moral judgment in men and women is extremely similar.
Lawrence Kohlbergs (1969, 1984) studies of moral reasoning also elicited some gender difference in moral judgment. Carol Gilligan (1982) and her colleagues used these differences as a starting point to propose completely different patterns of moral development for men and women. Though it is certain that there are some broad gender differences in moral behavior, the possibility of eliciting these differences through moral dilemmas is still undetermined. The dilemmas may only be eliciting responses to very specific hypothetical scenarios, rather than moral judgments relating to real-life situations. If this is true, then the variance found in the present study is not indicative of gender differences in consequential and non-consequential reasoning. The variance may simply point to a discrepancy in the way males and females respond to this specific type of hypothetical moral dilemma.
But even though these gender differences may not relate to everyday morality, they may tell us something about moral judgment in extreme situations, such as times of war or tragedy. Throughout history, men have seen themselves cast in the role of the soldier. It is reasonable to infer that the cold and brutal decisions demanded of soldiers in war may in some way affect the moral judgments of men. Battles are all about sacrifices and numbers. The method of war requires the death of some in order to ensure the safety of many. Since soldiers have traditionally been males, it is possible that men may be more inclined than women to make unemotional calculations regarding the life and death of other people.
The perspective of consequential and impartial moral judgment, which Carol Gilligan labeled an "ethic of justice," may indeed be more indicative of male moral reasoning than female moral reasoning. However, even if this is true, it is only true in very rare situations, so that the differences may be more hypothetical than real. Still, it is possible that gender differences in consequential and non-consequential reasoning may have some more subtle effects on social issues. For example, laws regarding hotly debated issues such as abortion, gun control and capital punishment are directed by a gender-biased Congress. Presently, there are more than 10 times as many men in the Senate as there are women (Womens Studies Database, 2000). Whether non-consequential reasoning is a factor or not, if there is a gender difference in moral judgment, it is quite possible that it is directly affecting our governments policies and legislation.
Chodorow, Nancy (1989). Feminism and psychoanalytic thinking. New Haven: Yale University Press.
Eibl-Eibesfeldt, Irenaus (1971). Love and Hate: The Natural History of Behavior Patterns. New York: Holt, Rinehart and Winston.
Eisenberg, Nancy (1982). "The development of reasoning regarding prosocial behavior." In Nancy Eisenberg (Ed.), The Development of Prosocial Behavior (pp. 219-246). New York: Academic Press, Inc.
Ford, M. R., and C. R. Lowery (1986). "Gender differences in moral reasoning: A comparison of the use of justice and care orientations." Journal of Personality and Social Psychology, 50:777-783.
Frankena, William K. (1973). Ethics. Englewood Cliffs, NJ: Prentice Hall, Inc.
Freud, Sigmund (1961). Civilization and Its Discontents. (James Strachey, Trans. & Ed.), New York: W.W. Norton & Co., Inc.
Garmon, L. C., K. S. Basinger, V. R. Gregg and J. C. Gibbs (1996). "Gender differences in stage and expression of moral judgment." Merrill-Palmer Quarterly, 42(3):418-437.
Gilligan, C. (1977) "In a different voice: Womens conception of the self and of morality." Harvard Educational Review, 47:481-517.
Gilligan, Carol (1982) In a different voice: Psychological theory and womens development. Cambridge, MA: Harvard University Press.
Hoffman, Martin L. (1977). "Sex differences in empathy and related behaviors." Psychological Bulletin, 84(4):712-722.
Holstein, C. (1976). "Development of Moral Judgment: A Longitudinal Study of Males and Females." Child Development 47:51-61.
Kohlberg, Lawrence (1969). "Stage and sequence: The cognitive-developmental approach to socialization." In Goslin, D.A. (Ed.), Handbook of socialization theory and research. (pp. 347-480). Chicago: Rand McNally.
Kohlberg, Lawrence (1984). The Psychology of moral development: The Nature and validity of moral stages. San Francisco: Harper & Row.
Kohlberg, L., C. Levine and A. Hewer (1983). "Moral stages: A current formulation and a response to critics." In. J.A. Meacham (Ed.), Contributions to Moral Development. (10). Basel: Karger.
Lapsley, Daniel K. (1997). Moral psychology. Boulder, CO: Westview Press.
Lorenz, K. Z. (1966). On Aggression. London: Methuen.
Maccoby, E. and C. Jacklin (1974). The Psychology of Sex Differences. Stanford: Stanford University Press.
Piaget, Jean. (1932/1965). The Moral judgment of the child. New York: The Free Press.
Rest, J. (1979). Development in judging moral issues. Minneapolis: University of Minnesota Press.
Rest, J. (1983). "Morality." In P. Mussen (Ed.), Handbook of child psychology: Vol. 3. Cognitive development (4th ed.; J. Flavell & E. Markman, Vol.Eds.,pp.556-628). New York: Wiley.
U.S. Department of Justice. (1999). Bureau of Justice statistics. Washington, DC: Bureau of Justice Internet Website.
Walker, L. J. (1984). "Sex differences in the development of moral reasoning. A Critical review." Child Development, 55:677-691.
Walker, L. J. (1989). "A longitudinal study of moral reasoning." Child Development, 60:157-166.
Walker, L. J., B. de Vries and S. D. Trevethan (1987). "Moral stages and moral orientations in real life and hypothetical dilemmas." Child Development, 58:842-858.
Womens Studies Database. (2000). University of Maryland, College Park, MD.
William Indick received his BA in psychology and his MA in music therapy from New York University. He will earn his Ph.D. in developmental psychology from Cornell University in August 2001. For his doctoral dissertation project, he is investigating cultural and gender differences in the expression of moral norms and issues in Lawrence Kohlberg's "Longitudinal Study of Moral Development" (1955-1977) and in Snarey, Reimer and Kohlberg's "Israeli Moral Development Study" (1969-1978). Mr. Indick intends to pursue a career in teaching and research in the fields of psychology and human development. E-mail: Wi16@Cornell.edu.
John Sun Kim is completing his senior year as an undergraduate at Cornell University. He will earn his BA in Human Development in May 2001. Currently he is studying in South Africa as a student in the Cornell Semester Abroad program. E-mail: Jsk44@Cornell.edu
Beth Oelberger is completing her senior year as an undergraduate at Cornell University. She will earn her BA in Human Development in May 2001. Currently, she is active in several research projects at Cornell University, including a study of gender differences in participant responses to Lawrence Kohlberg's "Moral Judgment Interview." Ms. Oelberger intends to pursue a career in clinical psychology. E-mail: Bfo1@Cornell.edu
Lauren Semino is completing her senior year as an undergraduate at Cornell University. She will earn her BA in Human Development in May 2001. Ms. Semino is interested in moral education, moral development and the implementation of character education programs. Currently, she is active in several research projects at Cornell University and intends to pursue a career in school psychology. E-mail: Ljs25@Cornell.edu
Back to Journal Cover Page