the influence of psycho social resources on the as so citation between self-reported empthaty and...

8/8/2019 The Influence of Psycho Social Resources on the As So Citation Between Self-reported Empthaty and Depressive Sy

1/15


2/15

Social Psychology Quarterly2001, Vol. 64, No. 4,376-389

"When Feeling Other People's Pain Hurts":The Influence of Psychosocial Resources on the Association

Between Self-Reported Empathy and Depressive Symptoms*SCOTT SCHIEMANUniversity of Maryland

HEATHER A. TURNERUniversity of New Hampshire

Is self-reported mpathy associated with depressive ymptoms? Drawing on stressprocess theory and social psychological research on empathy, we hypothesize thatempathy s more strongly associated with depressive ymptoms when personal andsocial resources are deficient. Using data from a 1985-1986 community ample ofadults n southwestern Ontario, we examine he mpact of empathy n depressive ymp-toms and the potential moderating ffects of mastery, elf-esteem, ocial support, andeducation. Results how that the relationship etweenempathy and depressive ymp-toms s more stronglypositive among ndividuals who report esseducation, owerself-esteem, nd lower mastery. We discuss mplications f these indings.

Sociologists have long been interested inemotional dynamics within interpersonalrelationships (Cooley [1902] 1964; Scheler1992). Processes related to empathy are con-

sistent with this tradition. Shott (1979)defines empathy as "the arousal in oneself ofthe emotion that one observes in another orthe emotion one would feel in another's situ-ation" (p. 1328). Accordingly, empathyinvolves what an individual might feel if he orshe were in another person's circumstances.Researchers also have identified differentaffective and cognitive aspects of empathy.Affective features of empathy, or vicariousemotional responding, involve the replica-tion of emotion between the empathizer andthe target (Eisenberg and Fabes 1990). Otherdefinitions emphasize cognitive features suchas role taking and awareness of other's prob-lems and emotions (Davis 1996). In general,empathic individuals tend to expend "theeffort to understand the internal mental and

* This study was supported by a research grant anda National Health Scientist Award from the NationalHealth Research and Development Program(NHRDP) of Health Canada to Dr. R. Jay Turner. Theauthor would like to express sincere thanks to Dr.Turner for permission to use these data. Address cor-respondence to Scott Schieman, Department ofSociology, University of Maryland, 2112 Art-Sociology Bldg., College Park, MD 20742-1315; e-mail: [email protected].

emotional events of other human beings"(Rosenberg 1990:8).

Empathy and Depressive Symptoms

Is empathy associated with a greater riskof depressive symptoms? Centuries ago,Adam Smith ([1759] 1976) recognized thathumans tend to experience a "fellow feeling"and share in the sorrows and joys of otherpeople. Although empathy may extend toothers' joy, as well as pain, we examineresponses to a self-report measure of empa-thy that focuses on the awareness of andsharing in other people's problems, difficul-ties, and sorrows. Positive outcomes relatedto empathy, such as conformity to norms,moral conduct, and altruistic behavior, arefairly well established (Davis 1996;Eisenbergand Miller 1987); we contend, however, thatempathy may be associated with negativeemotionality under some conditions.Drawing on stress process theory (Pearlin etal. 1981), we propose that in the context of anegative self-concept and weak social ties,

empathy may deplete limited resources, eav-ing individuals more vulnerable to depressivesymptoms.

Two common processes related to empa-thy are sympathy and personal distress.Sympathy is defined as "an affective

376


3/15


4/15

SOCIAL PSYCHOLOGY QUARTERLY

another or anger on another person'sbehalf).

People are likely to vary in their propen-sity for parallel versus reactive outcomes as afunction of their capacity to deal with inter-

personal adversity. Research on stressprocesses suggests a number of resourcesthat are likely to affect this capacity. Twoaspects of self-concept, self-esteem and mas-tery, have been identified as important per-sonal resources, while social supportrepresents a crucial social resource affectingstress and mental health processes (Pearlin1999; Turner and Roszell 1994). Given thesignificance of education for cognitive func-

tioning, perspective taking, and mental healthoutcomes (Mirowsky and Ross 1998), we alsosuspect that educational attainment mightaffect empathy-related processes andresponses.

Self-esteem. Rosenberg (1965) definesself-esteem as "the evaluation which the indi-vidual makes and customarily maintains withregard to himself or herself: it expresses anattitude of approval or disapproval towardoneself" (p. 5). Although research documentsthat high self-esteem is related to positiveoutcomes such as physical health, psycholog-ical well-being, and happiness (Turner andRoszell 1994), little is known about the wayin which self-esteem combines with empathyto affect individuals' welfare. We apply theidea of parallel outcomes to generate expec-tations about the synergistic effects of lowself-esteem and high empathy: individualswith low self-regard may be at greater risk

for emotional "fallout" associated withempathy because they experience other peo-ple's problems in the context of their ownpoor self-image. That is, low self-esteem mayreduce the capacity to separate one's owndifficulties from those of the distressed other.Moreover, n so far as low-self esteem reflectspersonal troubles that must be managed, tak-ing on the additional difficulties of othersmay become overwhelming and stressful.Conversely, a solid sense of self may enhanceone's capacity for feeling empathy whileavoiding personal distress.

Sense of mastery. n addition, we proposethat the sense of mastery moderates the asso-ciation between empathy and symptoms ofdepression. People with high mastery reject

the notion that life's outcomes are due simplyto chance or fate (Gecas 1989). Like self-esteem, mastery is a personal resource thathelps people remain resilient in the face ofadversity. The emotional benefits of masteryare well established (Ross and Sastry 1999),but we are aware of no studies that examinethe joint effect of mastery and empathy onemotions. Although a greater awareness ofother's difficulties and sorrows can be stress-ful, masterful individuals may be better ableto avoid, manage, or control distressing out-comes. Indeed, individuals low in masterytend to respond to stress with greater psychi-atric and/or physical symptomology (Pearlin

et al. 1981). Thus, empathic responding-without also feeling personally distressed-may require the perception of control overone's own life outcomes.

We suspect that individuals with a strongsense of personal control view their ownproblems and opportunities as relativelyindependent of other people's. Conversely,more fatalistic individuals may draw strongerconnections between the misfortunes of oth-ers and their own potential for misfortune.

Social support. An abundance of evi-dence shows the emotional health benefits ofsocial support (e.g., Thoits 1995). Supportivesocial bonds foster communication and oblig-ation between individuals. Moreover, theaffective qualities of positive relationsinvolve "the sharing of personal thoughts andfeelings (i.e., self-disclosure) and other, relat-ed expressions of intimacy, appreciation, andaffection" (de Vries 1996:252). The social

exchanges that typify supportive networksoften reflect a high degree of affectiveinvolvement, reciprocity, and emotional con-cern (Lin and Peek 1999). These qualitiesappear to contribute to the direct positiveeffects of social support on mental health. Inaddition, considerable research has shownthat social support buffers the negativeeffects of stress on psychological well-being(Turner 1983; Turner and Turner 1999).Individuals who perceive themselves as beingsupported by others are less likely to experi-ence depressive symptoms as a consequenceof exposure to stress than are those who per-ceive low levels of support.

Social bonds are an important contextfor one's exposure to others' problems and

378


5/15

EMPATHY AND DEPRESSION

emotions, and for one's empathic responsesto such exposure. We suspect that the qualityof relationships and interactions with otherscan help to shape empathic outcomes.Specifically, we hypothesize that empathicindividuals with high levels of support will beless distressed by others' difficulty and sor-rows than will those with lower levels of sup-port. Given that heightened empathyincreases one's awareness of others' prob-lems, we expect that social support is aresource that can reduce that potential stress.In contrast, without the availability of sup-portive others, one may have fewer opportu-nities for empathic feelings and actions that

cultivate mutually supportive interpersonalbonds. Empathic individuals who areinvolved in hardly any supportive relation-ships may have fewer outlets for empathicconcern; therefore they may be more likelyto personalize or ruminate about the prob-lems of others. That is, supportive relation-ships may provide a channel for empathicresponding that is other-oriented rather thanself-focused.

Education. Self-esteem, mastery, andsocial support are psychosocial resources.Educational attainment, in contrast, is oftencharacterized as a source of "human capital"(Becker 1993). Education can improve indi-viduals' ability to think, symbolize, and com-municate (Becker 1993; Mirowsky and Ross1998). Moreover, through its impact on anarray of attitudinal measures, education canbe considered an indicator of "cognitivesophistication" (Jackman and Muha 1984;

Phelan et al. 1995).Although the effect of education on

emotions is well documented (Ross and VanWilligen 1997;Schieman 2000), we could notlocate any studies that investigate the jointinfluence of education and empathy ondepressive symptoms. We suspect that empa-thy may be more distressing for individualswith fewer years of education. This hypothe-sis is based on theoretical propositions thatdistinguish between cognitive and affectivefeatures of empathic responding (Davis1996). It seems likely that the increased cog-nitive functioning, flexibility, and complexity,which are often cultivated by higher educa-tion, may encourage more cognitive featuresof empathy and fewer affective features. A

sufficient level of cognitive sophisticationmay be required if one is to engage in per-spective taking without also experiencingpersonally distressing features of empathy. Inother words, we contend that people withhigher levels of education may tend to expe-rience the cognitive aspects of empathy andavoid the negative affective componentsrelated to the self. This expectation impliesthat empathy should have a stronger positiverelationship with depression at lower levelsof education.

Hypotheses and Analytic Strategy

First, we examine the main effects ofempathy on depression, holding constanteducation, support, mastery, and self-esteem.Because previous research suggested thatthese variables themselves may be associatedwith both depression (Ross and Van Willigen1997) and empathy (Schieman and VanGundy 2000), we wish to determine theeffects of empathy on depression, indepen-dent of these other factors. In addition,because the

original sampleinvolved a

largeoversample of disabled community residents(see "Methods" section for details), we con-trol for disability status (disabled or nondis-abled) and health status. Insofar as disabilitymay influence both depression and one's ten-dency to empathize with the problems of oth-ers (and/or influence our hypothesizedmoderators), disability and health statusshould be controlled. For similar reasons, wealso adjust for age, sex, employment status,and married status. Both empathy anddepression may be distributed differentiallyacross those variables; hus it is important tocontrol statistically for that possibility in ouranalyses.

We hypothesize that empathy will berelated more positively to depressive symp-toms among people with low self-esteem,mastery, social support, and education.Multiplying empathy by each of those vari-ables creates interaction terms, which weinclude in separate ordinary least squaresregression models to assess their moderatinginfluence on the relationship between empa-thy and depression. We center all variables inour interaction terms to reduce multi-

379


6/15


collinearity between interaction coefficientsand lower-order terms (Mirowsky 1999).

METHODS

SampleThe data used in this study are derived

from second-wave interviews of 731 commu-nity-dwelling physically disabled persons and850 nondisabled respondents matched onage, sex, and area of residence (see Turnerand Wood 1985). In 1981, a two-stage clustertechnique produced a random sample of10,972 households and identified 22,680adults age 18 and over in 10 counties in

southwestern Ontario. A screening proce-dure obtained the final sample of 967 self-identified disabled respondents, based on thequestion "Do any adults in the householdhave any physical health condition or physi-cal handicap that has resulted in a change intheir daily routine or that limits the kind oramount of activity they can carry out?" Heartdisease (16.2 percent), arthritis (12.1 per-cent), osteoarthritis of the spine (7.9 per-

cent),and rheumatoid arthritis

(5.4 percent)were the most frequent conditions. Sixteen ofthe most common disorders account for 75percent of the sample.

Four years later, 731 of the original 967respondents participated n a follow-up inter-view. Of the 967, 13 percent had died, 4 per-cent were institutionalized or too ill toparticipate, 5.6 percent refused, and 1.7 werenot located. In addition, 850 comparisonrespondents who reported no disabling con-

dition were interviewed.In the present study, we use a merged

sample of 1,567;we dropped 14 cases becauseof missing values on items of interest.Disabled subjects are less likely than nondis-abled to be married (63 percent versus 73percent), and they report lower education(10.7 versus 11.8 years) and householdincome (40 percent with less than $15,000peryear versus 22 percent).

Measures

Table 1 shows a factor analysis of depres-sive symptoms and empathy items.Depressive symptoms are assessed with theCenter for Epidemiological Studies (CES-D)

items, in which respondents are asked howoften they experienced each of twenty symp-toms in the past seven days. The validity andreliability of the CES-D are well documented(Radloff 1977). We summed the items such

that higher scores reflect more depressivesymptomology. Reliability is .87. The empa-thy index contains items that are similar oridentical to items from several larger indices(see Davis 1996; Mehrabian and Epstein1972).We summed those items; higher scoresreflect greater empathy. Reliability is .70.

Hypothesized moderating variables. Inself-esteem items, respondents are asked toindicate how strongly they agree or disagree

with these statements: "I feel that I have anumber of good qualities," "I feel that I'm aperson of worth at least equal to others," "Iam able to do things as well as most otherpeople," "I take a positive attitude towardmyself," "On the whole I am satisfied withmyself," "All in all, I'm inclined to feel thatI'm a failure" (reverse coded)(Rosenberg1979). Choices range from 1 (strongly dis-agree) to 5 (strongly agree). The index sumsthe responses; higher scores reflect higherself-esteem. Reliability is .78.

Mastery items assess the extent to whichrespondents agree or disagree with thesestatements: "I have little control over thethings that happen to me," "There s really noway I can solve some of the problems I have,""There s little I can do to change many of theimportant things in my life," "I often feelhelpless in dealing with the problems of life,""Sometimes I feel that I am being pushed

around in life," "What happens to me in thefuture mostly depends on me" (reversecoded), "I can do just about anything I reallyset my mind to" (reverse coded) (Pearlin andSchooler 1978). Choices range from 1(strongly agree) to 5 (strongly disagree). Weaveraged the items; higher scores reflecthigher mastery. Reliability is .71.

Social support is assessed with theRevised Kaplan Scale (see Turner, Frankel,and Levin 1983). In a set of nine vignettes,respondents are asked to identify with one offive descriptions that best describes theirlevel of received social support. This scaleassesses the degree to which respondents feelcared for and loved, esteemed and valued,and part of a network of communication and

380


7/15


Table 1. Factor Loadings or Empathy and Depression tems (N = 1,567)

Empathy DepressionEmpathy tems

1. I tend to get emotionally nvolved with friends' problems.a

2. I don't get upset because a friend s troubled.3. When a friend starts o talk about his or her problems, try to steer theconversation o something lse.

4. Sometimes don't feel very sorry or other people when they are havingproblems.

5. Other people's sorrows do not usually disturb me a great deal.6. I am usually aware of the feelings of other people.a7. I feel that other people ought to take care of their own problems hem-

selves.8. Many times I have felt so close to someone else's difficulties hat they

seemed as if they were my own.a

Depression Items

1. I was bothered by things hat usually don't bother me.2. I did not feel like eating; my appetite was poor.3. I felt that I could not shake off the blues even with help from my family

or friends.4. I felt that I was just as good as other people.a5. I had trouble keeping my mind on what I was doing.6. I felt depressed.7. I felt that everything did was an effort.8. I felt hopeful about the future.a9. I thought my life had been a failure.

10. I felt fearful.11. My sleep was restless.12. I was happy.a13. I talked ess than usual.14. I felt lonely.15. People were unfriendly.16. I enjoyed ife.a17. I had crying pells.18. I felt sad.19. I felt that people disliked me.20. I could not "get going."

.52

.50

.59

.18

.04-.04

.66

.73

.41

.65

.37

.00

.04-.04-.01

.15

.03-.02

.01

-.08.00.00

-.04-.12-.04

.01

.01-.10-.06

-.02-.03-.06-.07

.09-.03

.00

.53

.47

.75

.33

.56

.80

.63

.46

.61

.62

.54

.65

.52

.66.38

.61

.56

.74

.50

.61

Notes: Principal-components actor oadings with retention of two factors and promax otation. Response cate-gories for empathy tems: 1,very much ike me; 2, much ike me; 3, somewhat ike me; 4, not very much ike me;5,not at all like me. Response categories or depression tems:0, rarely or none of the time; 1,some or a little ofthe time; 2, occasionally r a moderate amount of the time; 3,most or all of the time.a

Reverse coded.mutual obligation in which they can count onother people. The support index sums thenine scores; higher scores reflect greatersocial support. Reliability is .80.

Education is measured in years ofschooling.

Control variables. Age is measured inyears. Average age is 58 years, with a rangefrom 22 to 92. Gender is coded 1 for female.Fifty-five percent are women. Married iscoded 1, all others 0. Employed is coded 1, allothers 0. The health index combines twoitems. One item asks, "In the past month, howoften has any health condition caused youpain?" Response choices range from 1

(never) to 5 (every day). The other item asksabout general health in the past year; choicesare 1 (very poor), 2 (poor), 3 (fair), 4 (good),and 5 (excellent). We standardized andsummed the two items to form the index.Appendix Table Al shows a correlationmatrix and descriptive statistics.

RESULTS

Is empathy related to symptoms ofdepression, even with adjustment for otherpsychosocial variables and sociodemograph-ic characteristics? Eq. (1) of Table 2 showsthat empathy is associated positively with

381


8/15


9/15


LowSelf-Esteem

High Self-Esteem

I I I15 20 25

Empathy ndex30I0 35 40

Note: Lines correspond o Eq. 2 of Table 2 and reflect predicted depression cores or nondisabled, married,and employed women; other variables re held constant at their means. ntercepts willvary with other combi-nations of gender, marital, work, and disability tatuses.

Figure 1.The Association Between Self-Reported Empathy and Depressive Symptoms t Low (TenthPercentile) and High (Ninetieth Percentile) Levels of Self-Esteem

Itappears

that anotherpsychosocialresource, the sense of mastery, has an influ-

ence similar to that of self-esteem.Specifically, the negative interaction coeffi-cient (b = -.10) shown in Eq. (3) indicatesthat empathy has a positive effect on depres-sion only at lower levels of mastery-anotherfinding in support of our prediction that mas-tery is a buffer against the distressing effectsof empathy. Figure 2 illustrates these pat-terns: among individuals with high mastery,predicted depressive symptoms are low anddo not change across the entire range ofempathy scores. Among people with lowmastery, however, predicted CES-D scoresincrease from about 10 to 18 across the rangeof values on the empathy index.

In contrast to evidence about "resource-deficient" individuals (i.e., people with lowerself-esteem and sense of mastery), our find-ings indicate that social support does not mod-erate empathy's relationship with depressivesymptoms. Eq. (4) shows that the support-by-empathy coefficient (b = -.01) is not statisti-cally significant: as self-reported empathyscores increase, the size of the increase inCES-D scores remains relatively stable across

levels of socialsupport.

Thesepatterns

fail tosupport our prediction that social support is aresource protecting individuals from theharmful effects of empathy-related distress.

Finally, we examine the role of educationas a moderator. In keeping with our hypothe-sis,we find that education moderates the pos-itive impact of empathy on depression: asempathy increases, scores on the CES-Dindex also increase, but only at lower levels ofeducation (see Eq. (5)). Figure 3 illustratesthe significant interaction between empathyand education, showing their joint influenceon symptoms of depression.

DISCUSSION

Guided by a stress process framework, weexamined the impact of empathy on depres-sion, and identified several factors that condi-tion this association. Findings confirmed ourhypothesis that empathy can increase individ-uals' risk for symptoms of depression. Otherpeople's problems, difficulties, and sorrowsoften present distinctive emotional challenges.One might suspect that the more deeplyinvolved individuals become in such adversity,the greater the potential for emotional trans-

18 '

16 '

14 "

12 '

10 '

aE

0.EnU._

a,0.

4

10M=r=

383


10/15


18 -

16 -

u 14E0

12 '

0

c 8 =

10 15 20 25Empathy ndex

30 35 40

Note: Lines correspond o Eq. 3 of Table 2 and reflect predicted depression cores or nondisabled, married,and employed women; other variables re held constant at their means. ntercepts willvary with other combi-nations of gender, marital, work, and disability tatuses.

Figure 2. The Association Between Self-Reported Empathy and Depressive Symptoms t Low (TenthPercentile) and High (Ninetieth Percentile) Levels of Mastery

ference. In this respect, empathy-as a socioe-motional process-is a potential stressor, andtherefore also can threaten the empathizer'semotional well-being.

Yet although empathy clearly can placeindividuals at greater risk for negative emo-tional outcomes, such risk is not uniform. Wefound that empathy and symptoms of depres-sion are associated more strongly and posi-tively when self-esteem, mastery, and

education are low. Insofar as high empathyincreases one's exposure to stress by height-ening awareness or sensitivity regarding theproblems of others, self-esteem, mastery, andeducation apparently are buffering resourcesthat reduce distressing outcomes. Takentogether, our results seem to confirm a basictenet of stress process theory: the emotionalimpact of interpersonal and socioemotionalstress is often more severe when psychosocialresources are few.

Because empathic ndividuals are attunedto other people's problems, the patternsreported here are somewhat parallel to ideascontained in the "cost of caring" thesis.Although this thesis has been discussed arge-ly in the context of understanding gender dif-

ferences in distress, he basic premise involvesgroup variations n exposure and/or vulnera-bility to the hardships of others in one's socialnetwork (Belle 1982; Kessler and McLeod1984;Turner and Avison 1989).As a result oftheir greater involvement with, or sensitivityto, others' problems, women (or more empath-ic individuals) are at greater risk for depres-sion. The "cost of caring" also may reflectsensitivity to network stressors or sorrows in

the context of deficient resources. This inter-pretation is consistent with some research onwomen living in poverty, from which Belle(1990) concludes that helping other peoplewhen one's own resources are few may erodethe rewards associated with assistance andmay become a source of emotional distress.

Two components of self-concept, self-esteem and mastery, emerged as importantresources that buffer against the potentiallydistressing effects of empathy. That is, empa-thy exerts a more powerful positive effect ondepressive symptoms when self-esteem andmastery are low. The mechanisms by whichthese factors influence empathetic responsesare likely to involve differences in cognitiveand affective processes. Empathetic process-

LowMastery

HighMastery

4

384


11/15


LowEducation

HighEducation

15 20 25Empathy ndex

30 35 40

Note: Lines correspond o Eq. 5 of Table 2 and reflect predicted depression cores or nondisabled, married,and employed women; other variables are held constant at their means. ntercepts willvary with other combi-nations of gender, marital, work, and disability tatuses.

Figure 3.The Association between Self-Reported Empathy and Depressive Symptoms t Low (9 years) andHigh (16 years) Levels of Education

es that are self-focused are more likely toevoke affective matching with the target ofempathy and to result in personal distress. Incontrast, other-directed processes-focusingon another person's needs and experience-produce empathic concern for others(Eisenberg et al. 1991).

Self-conceptions may influence the ten-dency towards other- versus self-focusedempathy. Low self-esteem, for example, may

erode one's capacity to separate personaltroubles and feelings from those of distressedothers. Similarly, individuals with low mas-tery may have less capacity to manage others'emotional needs and may draw stronger con-nections between others' misfortunes andtheir own personal potential for misfortune.Moreover, given their greater effectiveness,masterful individuals may have more timeand energy to deal with other peoples' diffi-culties and concerns without being over-whelmed. Friends' problems are moredepressing if one also feels helpless in dealingwith one's own problems.

In our original hypotheses, we proposedthat individuals with high empathy and adeficit in interpersonal associations charac-

terized by intimacy, appreciation, disclosure,and affection might be most depressed.According to the stress process model, sup-portive relationships should protect againstthe distressing elements of interpersonalstrain and stress. We also contended that thelack of supportive ties might increase theself-oriented or distress responses to empa-thy. Our results, however, failed to supportthat hypothesis. We documented that the

relationship between empathy and depres-sive symptoms does not vary across levels ofsocial support. It is possible that social sup-port, unlike self-esteem and mastery, ends tocreate greater exposure to other people'sproblems and hardships. Social support typi-cally involves close contact with others andreciprocal exchanges. Whereas a positive self-conception may allow one to separate per-sonal distress from the distress of others,social support (independent of mastery andself-esteem) may simply increase the oppor-tunity to be distressed by the problems of sig-nificant others. Thus, in the context ofempathy, any stress-buffering effects of sup-port may be offset by increased exposure tonetwork stress.

18 '

16 '

14 '

12 '

10 s

E.2

E0

O.aoQ3

4

10* * * am a

385


12/15


Our hypothesis concerning the moderat-ing effects of education was confirmed:empathy exerts a stronger positive effect ondepressive symptoms when the level of edu-cation is lower. Like self-esteem and mastery,education apparently is a resource on whichindividuals can draw in socioemotionalexchanges. The beneficial effect of educationcan reflect either economic or cognitiveresources. We suspect, however, that the lat-ter exerts a stronger influence on the rela-tionship between empathy and depression.

In general, the patterns found in thisresearch reinforce ideas about the cognitivesophistication aspects of education in relation

to empathy. More highly educated people maybe able to engage in empathy-related perspec-tive taking without also suffering he emotion-al costs of empathy. The cognitive flexibilityand complexity fostered through higher edu-cation allow one to disentangle one's ownexperience and emotion more successfullyfrom the experiences and emotions of others.It seems plausible that some features of edu-cation equip individuals o competently detectand interpret other people's emotional statesand to suppress egocentric viewpoints andimpulses. In contrast, ow levels of educationmay reduce the capacity to engage in other-focused empathy. For individuals with lesscognitive sophistication, mpathy-related per-spective taking may entail the drawing ofstronger connections between other people'ssorrows and one's own. Given the limitationsof our data, however, we can only speculateabout processes related to cognitive skill and

perspective-taking bility.Several potential limitations of thisresearch must be mentioned. First, we cannotmake strong claims about causal orderamong our variables. For example, t is entire-ly possible that individuals who experiencedepressive symptoms may tend to affiliatewith people who have troubles, sorrows, anddifficulties. In addition, although it was not afocus of this research, it is plausible that ourmoderating variables, themselves, influencethe level of empathy. Self-concept variablesand education not only may condition empa-thetic responses, but also may affect levels ofempathy directly. In fact, all four of ourhypothesized moderating resources are cor-related positively with self-reported empathy

(see Appendix Table Al). In addition, thereare likely to be complex reciprocal relation-ships among these variables. Alternativemodels would be tested most appropriatelywith panel data that examine the strengthand direction of an array of plausible linksbetween empathy, psychosocial resources,and emotions over time.

In the future, researchers should consid-er the joint effects of sex- and age-relatedprocesses on the relationships betweenempathy, depression, and other emotionssuch as guilt, shame, and anger. Some evi-dence suggests that the gender gap in depres-sion "rises in adulthood as women and men

enter and live out their unequal adult status-es" (Mirowsky 1996:376). Other findings sug-gest that women often experience a greater"burden of caring" (Kessler and McLeod1984;Thoits 1999), which may create differ-ent meanings, experiences, and outcomes ofempathy. Previous studies showed thatwomen (Eisenberg and Lennon 1983) andyounger people (Schieman and Van Gundy2000) tend to report higher scores on self-reported empathy measures.' Empathy maycause stress by increasing exposure to otherpeople's problems and sorrows; also, the pos-session of fewer psychosocial resources mayincrease one's vulnerability to such distress.Knowledge about age-linked gender varia-tion in different emotions could be informedby the study of interpersonal dynamics with-in social networks and their link withempathic thoughts, feelings, and behavior.

Finally, ittle is known about gender dif-

ferences in the links between anger andempathy, and how those processes are pat-terned over the life course. As individualsdevelop and embark on different role trajec-tories, they often encounter an array of emo-tional experiences in such role relationships.A growing area of research examines theprocesses that link emotions, psychosocialresources, and violent expressions in particu-

1 We tested whether thepositive relationshipbetween empathy and depressive symptoms is

stronger among women. Additional analyses (notshown) failed to support that contention: the gender-by-empathy interaction coefficient was insignificant(b = .08, p = .25). Thus, at least in our sample, the"cost" of empathy seems to be similar for women andfor men.

386


13/15


**

00 NC 00-N 00

* *

Cf

the influence of psycho social resources on the as so citation between self-reported empthaty and...

Documents