Explaining Differences in Consumption By Working and Non-Working Wives

Shreekant G. Joag, Oklahoma State University
James W. Gentry, Oklahoma State University
JoAnne Hopper, Oklahoma State University
ABSTRACT - The rapid increase in the number of women in the work force has affected society greatly. The growth in the number of two-income families has resulted in changing lifestyles, and many marketers have expected to find corresponding differences between the consumption patterns of working and non-working wives. However, most studies have not found those patterns to be significantly different. This study investigates the explanatory value of one of the recent refinements in the concept of "working wives"--the delineation of the high-occupational-status working wife from the low-occupational-status working wife (Schaninger and Allen 1981)--in a study of the consumption of a service. While the results indicate some support for the approach, we find it to be somewhat limited. Thus we propose that a broader perspective be used in further studies, one that incorporates Reilly's (1982) concept of role overload along with the wife's occupational status. We discuss the expected relationship between the match (mismatch) of a wife's work role with her work goal and her consumptive behavior.
[ to cite ]:
Shreekant G. Joag, James W. Gentry, and JoAnne Hopper (1985) ,"Explaining Differences in Consumption By Working and Non-Working Wives", in NA - Advances in Consumer Research Volume 12, eds. Elizabeth C. Hirschman and Moris B. Holbrook, Provo, UT : Association for Consumer Research, Pages: 582-585.

Advances in Consumer Research Volume 12, 1985      Pages 582-585


Shreekant G. Joag, Oklahoma State University

James W. Gentry, Oklahoma State University

JoAnne Hopper, Oklahoma State University


The rapid increase in the number of women in the work force has affected society greatly. The growth in the number of two-income families has resulted in changing lifestyles, and many marketers have expected to find corresponding differences between the consumption patterns of working and non-working wives. However, most studies have not found those patterns to be significantly different. This study investigates the explanatory value of one of the recent refinements in the concept of "working wives"--the delineation of the high-occupational-status working wife from the low-occupational-status working wife (Schaninger and Allen 1981)--in a study of the consumption of a service. While the results indicate some support for the approach, we find it to be somewhat limited. Thus we propose that a broader perspective be used in further studies, one that incorporates Reilly's (1982) concept of role overload along with the wife's occupational status. We discuss the expected relationship between the match (mismatch) of a wife's work role with her work goal and her consumptive behavior.


Due to increased time constraints, changes in sex role norms, role overload, and increased monetary resources, the impact of the wife's working status has become a major variable in the prediction and understanding of family behavior. Not only have women increased their participation in the workforce, but their reasons for working have also changed. Work is viewed by many women as a means to gain prestige, esteem, and independence in addition to monetarY rewards (Scanzoni 1977).

The large number of females in the workforce provided incentive for marketers to investigate differences in the consumption patterns of non-working and working women. The basic premise of most of the research was that role overload would occur among working women, which would bring about more convenience-oriented consumption behavior than with non-working females. That is, it was hypothesized that the wife's employment would take away a share of her time and energy from home production activities and thus result in increased purchase of time-saving products and services. Another explanation for the expectation of differences comes from economic theory based on the notion that the wife's income is treated as transitory in nature. Mincer (1960a, 1960b) argued that households with working wives are more likely to purchase durables as a form of saving the transitory income. Similarly, Drucker (1976) suggested that such durables would be bought as an extraordinary purchase from extra income. Strober and Weinberg (1977) argued, however, that with time, the wife's attachment to the labor force has become more permanent and the income is no longer treated as transitory.

Most of the studies that have investigated differences in working and non-working wives have found very few differences in consumption patterns. In a study by Anderson (1972), no difference was found in the consumption of convenience food items among working and nonworking women, although working women were found to be more brand loyal and make fewer grocery shopping trips than non-working wives. In an investigation of expenditures on time-saving durables, Strober and Weinberg (1977) found that the wife's labor force participation was not important in the purchase of time-saving durables, although the size of the family's income was significant in explaining expenditures on time-saving durables. In a replication of the Strobe- and Weinberg (1977) study, Weinberg and Weiner (1983) concluded that the results of their study also supported the hypothesis that the wife's employment status is not significantly related to a family's purchase of time-saving durables.

Although the hypothesis that the wife's employment was related to the purchase of time-saving durables and consumption of convenience products has not been well supported by most studies, researchers have begun to look for additional variables and constructs to explain the differences in consumption patterns of working and non-working wives. Two major concepts have been introduced that have contributed to the support of the hypothesis of differences in the consumption behaviors of working and non-working wives. Schaninger and Allen's (1981) approach included the development of a three-way classification scheme based upon the occupational status of the wife. Reilly (1' 82) introduced the importance of role overload and role conflict in the working wives' family and job position duties. Both Reilly (1982) and Schaninger and Allen (1981) stated the importance of the differences in levels of role overload experienced by working wives.

Schaninger and Allen's (1981) approach was based upon the Rapoports' (1971) distinction between dual income and dual career families. Allen and Schaninger (1980) viewed dual income working women as being employed in not permanent or less meaningful work. Dual career wives were seen as having a high degree of commitment to their careers. The three-way scheme developed by Schaninger and Allen (1981) included non-working wife, low-occupational-status working wife, and high-occupational-status working wife. It was hypothesized that the high-occupational-status working wife would experience the greatest role overload due to her dedication to her career. The results of the study indicated that the three-way classification scheme explained some differences in food, beverage and alcohol consumption, makeup usage, clothing purchases, shopping behavior and teal proneness, media wage, and major and minor appliance ownership. Nichols and Fox (1983) used the Schaninger and Allen (1981) three-way occupational status scheme in their investigation of time-saving and time-buying strategies. In general, they found income to be a better explanatory variable than the occupational status of the wife. The greatest differences were found in high status working wives. They found that high status working wives were more likely to use time-saving strategies of preparing fewer meals at home, having less leisure time, and reducing their time in household production. Time-buying strategies used by the high status working wives included buying more child care, eating meals away from home, and purchasing disposable diapers for children.

Reilly (1982) investigated convenience consumption by working wives. He proposed that role overload results in role conflict which occurs when demands on the working wife's family position exceed the amount of time and energy available. Thus the role overload should lead to more convenience-oriented consumption. Although role overload and the purchase of time-saving durables was found to be causally related to the wife's employment and convenience consumption, the amount of variance explained was very small. Reilly (1982) also pointed out that other ways of reducing demands of the family position include redivision of household labor assignments, eating out frequently, buying fast food for home consumption, and hiring household help as possible alternatives. This study will investigate the use of some of those alternatives; specifically, the study will investigate the ability of Schaninger and Allen's (1981) classification scheme in explaining differences in the frequency of dining out and the use of food delivery services. Kinsey (1983) found that the income earned by women working full time did not increase the marginal propensity to consume food away from home. However, she noted that she did not measure the frequency of eating out (but rather the dollars spent on meals away from home) and that, if the reason for dining out is to save time, working women may eat more frequently at faster (and relatively cheaper) eating establishments.


The study will investigate the following hypotheses (stated in alternative form):

H1: Households with high-occupational-status wives (HSW) will dine out more frequently than those with low-occupational-status wives (LSW) or nonworking wives (NWW). It is expected that wives in more traditional households will feel more obligation to perform duties such as cooking.

H2: HSW households will have foot delivered more frequently than NWW and LSW households. Since having food delivered is an alternative to cooking (as is dining out), the same rationale exists for this hypothesis as was expressed in H1

H3: HSW households will be more patient than LSW or NWW households should the foot delivery service be slow. The desire to be removed from household duties such as cooking should be greater in HSWs, since their dedication to the workplace has supplanted their total dedication to the home environment somewhat.


A survey of foot delivery service usage was conducted in a college town of approximately 40,000 residents. An area sample was designed, using a traffic zone sap and population-per-traffic-zone data. Thirty areas of the city with approximately equal population were identified and then points were chosen randomly within the areas as starting points. Thus households were selected in proportion to the city r S population distribution. The questionnaires were distributed by interviewers, but left with the respondents to complete within a short time period. This hand-delivered, self-administered questionnaire approach has been used successfully in a variety of studies. (Dunning and Calahan 1974; Lovelock et al. 1976; Stover and Stone 1974).

A total sample of 696 households was obtained. The respondents were grouped as to their marital status: 32 non-response; 391 single; 38 divorced, separated or widowed, and 235 married. The high percentage of single households was due to the fact that the survey was conducted in a college town. [The fact that the survey was taken in a college town also was responsible for the failure to include a measure of income in the study. Pretesting indicated that income" was interpreted by some students to mean their own while others interpreted it to mean their parents' income. Consequently, the sponsor of the study decided that no measure of income was necessary.] Those respondents who were not married were deleted from the sample. The married households were separated into three groups based on the occupational status of the wife. There were 132 households in which the wife was not working (NWW), 52 LSW households in which the wife held a job in one of the lower categories of the Hollingshead Index of Social Position (Hollingshead and Redlich 1958), and 51 RSW in which the wife held a position in one of the top three white-collar categories of the Hollingshead Index. This approach is the same as the one taken by Schaninger and Allen (1981).

The family classification scheme discussed above was related to a variety of variables dealing with the frequency of dining out and the use of fast-food delivery services. Given the limited number of households in the LSW and HSU categories and given the ordinal (and occasionally nominal) nature of the scales used to measure the dependent variables, chi-square analyses were performed. When the dependent variable had several possible responses (for example, when a five-point scale was used), the data were collapsed into two or three possible response categories so as to insure that the minimum expected in frequencies were not close to zero. [For example, the question dealing with frequency of dining out had six possible responses (less than once a month, 1-3 times a month, once a week, 2-3 times a week, 4-6 times a week, 7 or more times a week). The following crosstabulation resulted: TABLE

Due to the sparseness of the cells associated with dining out more frequently, the tables were collapsed in order to compare those who dine out less than once a week (columns 1 and 2) to those who dine out more frequently (columns 3-6).]


Table 1 presents the results of the chi-square analyses. Hypothesis 1, that HSW households will dine out more frequently, is supported as 73: of them dine out at least once a week compared to 54% of the LSW households and 42% of the NWW households. The higher frequency of dining out for RSW households is significantly greater than that for LSW households (X2 (ldf) 3.9, p < .05). However, the type of occupation is not important when one considers the frequency of dining out for pizza. More than 50% of the households with working wives (regardless of occupation) tine out for pizza more than once a month, compared to 33% of the nonworking wives. One implication of these findings may be that RSW dine out more frequently at a higher class of restaurant than do LSW households.



Hypothesis 2, that HSW households will order foot to be delivered more frequently, is not supported. Households with working wives were more likely to have ordered a foot delivery, but the higher percentage (75%) of HSW households doing so is not significantly different from the percentage (68%) of LSW households doing so (p > .1). Further, there were no differences in the frequency of ordering foot deliveries among households regardless of family classification, as only 18% of the HSW households ordered more than once a month compared to 17% of the LSW households and 11% of the NWW households.

Hypothesis 3 deals with the actual process of ordering the food. It is hypothesized that HSW households will show more patience should problems arise during this process, since the alternative of cooking the meal themselves (once they have decided not to cook) is less appealing to them than it is to more traditionally-oriented households. The results provide modest support for this hypothesis. Households with working wives express greater willingness to wait longer for a food delivery than do households with nonworking wives. The difference between HSW households (61% willing to wait more than 30 minutes) and LSW households (48%) is not significant (p > .1) however. The relationship between the family classifications and their willingness to wait (rather than to cancel the order) if the estimated delivery time exceeds their expectations is not significant when all these classifications are examined. However, HSW households are more willing to wait (rather than cancel) than are LSW households (X2 (ldf) - 3.2, p - .08). No differences were observed in terms of how many times a household would let the phone ring before calling another establishment.


Schaninger and Allen (1981) found that HSW households differ from LSW and NWW households in their consumption of a wide variety of goods. The results of this study indicate that the type of occupation helps to explain differences in the consumption of services as well. HSW households dine out more frequently than LSW and NWW households and they are somewhat more patient when it comes to waiting for a food delivery. Delineating between LSW and HSW households was not found to be important for the frequency of eating pizza at a restaurant or having used a food delivery service; for these variables, a delineation of working vs. nonworking wives explained the differences.

Overall, the results support the use of the wife's occupational status in investigating differences in dining out and in ordering food deliveries. As Allen, Deberec and Chan (1983) point out, the family classification scheme used in this study is a relatively crude one. The separation of working wives into HSW and LSW households based upon occupation is quite simple and, as our results indicate, useful. However, it is necessary to infer that commitment to the wife's job in the HSW household is greater (as was done in the rationale of our hypothesis) since no direct measure of labor force attachment was obtained. Thus, any relationship between Reilly's (1982) concept of role overload and occupational status is based on the very tenuous assumption that role overload increases with the status of one's occupation. Sieber (1974) observed that the positive outcomes of outside-employment may outweigh the role conflict and role overload and, as such, the working wife may not perceive role overload. Thus, the use of a demographic variable (occupational status) is insufficient, as it is the perception of the role overload by the individual herself that would affect her behavior and not the overload by itself.

What constitutes "role overload varies, as the overall capacity of an individual to meet various time and energy demands is not a fixed quantity. This capacity is flexible, and it may depend upon the individual's motivation for performing these roles. Several researchers have attempted to capture the "work motivation" construct by using various classification schemes. Bartos (1977) discusses a Yankelovich survey that asked working women: "Do you consider the work you do 'just a job' or a career?" In her own research, Bartos (1977, 1978) defined four groups of women--stay at home, plan to work, just a job, and career-oriented. Langer (1982) suggested the use of four categories of working women, as based on their attitude toward their work: high energy achiever, involved worker, conscientious worker, toiler. Hafstrom and Dunsing (1978) used monetarily-motivated and non-monetarily motivated classifications, and suggested that those who work for monetary reasons may view their purchases as ends in themselves while those who work for non-monetary purposes may be more likely to regard their purchases of consumptive items as means to ends at the time of purchase. Similarly, Rosen and Granbois (1983) found that the wives' motivation for working (for monetary reasons or not) affects the husband's and wife's roles in the family's financial decision making.

Previous research efforts that have examined consumption differences between working and non-working wives have focused on the wife's occupation, and recently, role overload as primary explanatory variables. The results of these studies have provided some insight into the consumption differences in married women, but have failed to explain comprehensively convenience consumption differences in working and non-working wives. We propose to integrate the wife's work role (occupational status) with the wife's work goal (her motivation for working) in a broader model that will provide a new basis for the examination of convenience and time-saving consumption behaviors of working and nonworking wives.


Most conceptualizations of a working wife's purchase behavior take a snapshot view and are essentially static in nature. It is important, however, to recognize the dynamic nature of the phenomenon. Our approach will be to present first our representation of how the interaction between roles and goals affects consumer behavior at any one point in time, and then to discuss how the nature of this interaction between roles and goals will change over time.

Our view is that role overload is not just a function of the demands of a total position set as Reilly (1982) suggests, but rather a function of the match (or mismatch) of the wife's roles and goals. Her perception of role conflict and role overload would not result merely from conflicting and competing demands and pressures caused by different roles in her position set. Rather the match (or mismatch) between her goals (which define her desired state) and her roles (which define her actual state) would determine her motivation to carry out the various roles ant, in turn, have an impact on her purchase behavior. A working wife with a career as her goal may perceive role overload in housework, while one whose goal is that of being a good homemaker may perceive role overload in her job. Therefore the actions taken by the two in reducing their perceived overload may be totally different, with one trying to minimize her commitments at home and the other office commitments.

Our model will incorporate a very simplified view of one's work motivation, using the labels "Home/Inward Oriented-- and 'Work/Outward Oriented." In some respects these are somewhat similar to the concepts of "modern" and "traditional' used by Reynolds, Crask, and Wells (1977) or the "feminist," "moderate," and "traditional" classification used by Venkatesh (1980). However, we acknowledge that what we are dealing with is continuum, as many wives are both home-oriented and work-oriented. Similarly, there are wives who to not work for a wage, but who are not home-oriented as they spend much of their time and energy outside the home in civic, social, or self-developmental activities. Hence we use the broader label "work/outward oriented to portray the career-oriented extreme. Our model thus should be interpreted as indicating that those wives who tent to be more work-oriented than home-oriented, for example, will be more likely to make the types of purchases indicated in Figure 1.

Also we need to clarify our labels for the work roles: "non-working," "working-stagnant career," and "working-progressive career." To a large extent, this categorization is very similar to the Schaninger and Allen (1981) categorization (non-working, low-occupational status, high-occupational status). However, the use of the terms "stagnant" and "progressive" reflects our desire to include the nature of responsibility faced by the wife. If she can move up the organizational latter as a result of her performance in decision-making tasks, then we would classify her role as "working-progressive career." On the other hand, if her reward structure is more-or-less independent of her performance but dependant on the decision-making of others, then we would classify her role as "working-stagnant career."

We combine the roles and goals in Figure 1 to provide some testable propositions as to how the catch/mismatch of those roles and goals will affect their consumption patterns. At the extremes of this table we find situations in which there is harmony between the wife's role and her goals. In box A, for example, we find nonworking, home/inward oriented wives. To the extent that role overload is experienced, it will be in outward roles. These wives are not as likely to purchase home-oriented convenience items, as their self-concept is dependent upon providing home production and services with a personal touch. Thus, other things being equal, this segment would be more likely to purchase items related to the personal touch in home-making: cookbooks, sewing materials, more traditional household appliances, etc.



At the other extreme (box F) we find harmony between the wife working in a progressive job who is career/ outward oriented. Her perception of role overload will probably be related to her home role. As such, she will be more likely to purchase home-oriented timesaving goods and services (microwave oven, housecleaning service, food delivery service, dining out more, etc.) Her dedication to her career may also be reflected in such purchases as a work wardrobe, a briefcase, and technological aids (such as a home computer, a telephone-answering service, or a call waiting/call forwarding system).

In the remaining boxes we find disharmony between the wife's roles and goals, and we propose that this disharmony may result in purchase behavior that would not be consistent with the wife's role or goals by themselves. For example, boxes B and C both relate to working wives who are home/inward oriented. The perceived overload would be directed at the career role, and we would expect to find these wives more likely to purchase job-related convenience products. To the extent that there are income differences between those wives in a progressive career (box C) as opposed to those in a stagnant career (box B), we would expect to see more convenience-oriented durables (for example, a home-computer with word processing capabilities) purchased by the wives in box C.

The disharmony noted in boxes D and E will result in a different perception of role overload (the home role is perceived to be overloaded) and in a different type of purchasing behavior. These wives will be more likely to purchase home convenience items. Again, income differences due to the work status of those in box E compared to those in box D may result in the purchase of more time-saving consumer durables.

Our model indicates that a wife who is working in a stagnant career is going to be in disharmony with her goals. Given our use of the extremes of the home versus career continuum, this is true. However, wives who have a combination of home and career goals may well be content in a work role that offers little opportunity for self-advancement.

The propositions generated in Figure 1 tend to indicate that goals are better predictors than roles. For example, we hypothesize that wives in the same work roles but with different goals (within the same row in Figure 1) will tend to purchase goods and services aimed at reducing different types of role overload: home/inward oriented wives would reduce the job-oriented overload while career/outward oriented wives would reduce the home-oriented overload. On the other hand, the primary type of role overload reduction is consistent within different goal structures (within the same column in Figure 1): again, home-oriented wives tend to reduce job/outward role overload while career-oriented wives would tend to reduce home-oriented overload. However, as discussed above, we to expect differences in the types of purchasing strategies used by wives in different work roles even though they have a common goal.

The proposed broader model is not all inclusive. The model's limitations include the deletion of other variables that may also influence wives' consumption behaviors. The impact of children, the wife's commitment to the family, and the wife's educational level are among the variables which may be included in future revisions of the model. A more fundamental limitation of the model is its implicit assumption that the wife is the family member responsible for most household purchases. As Burns, Foxman, and Myers (1984) note, both husbands and wives can face role overload in our society and an emphasis solely on the wife's roles and goals may well result in an incomplete representation of the household's purchase of time-saving and time-buying items.

A test of the model presented in Figure 1 would require only marginal additions to the standard research instrument used in working/non-working wife research. In addition to the wife's occupation question (and the subsequent use of the Hollingshead Index as was used in this study), one should also ask a question relating to her ability to determine her own career path in her current position. Further, it would be advisable to determine if the position is a part-time or a full-time one. The "goal" construct may be captured by some of the instruments used to measure the modern/traditional delineation, through an instrument designed to measure labor force attachment such as the Maret-Havens index (Maret-Havens 1977), or through a simpler process such as asking whether the work is just a job or a career (Bartos 1977).


Figure 1 represents the relationship between the role/ goal match (mismatch) and consumer behavior at any one point in time. However, a wife's goals and roles may change as she passes through the life cycle and experiences success and failure. In the long run, it is not just the goal versus role dyad that plays a major role in the wife's perception of role overload; the perceived chances of changing roles and of achieving the goals are also of vital importance.

Figure 2 presents a model depicting the dynamics of the wife's goals and roles. The model categorizes wives by the goals held to be most important at a point in time. Career/outward oriented wives consider a career as their goal, while home/inward oriented wives consider being a good homemaker as their primary goal. Satisfaction with the status quo exists when the role and the goal are in harmony. However, when there is disharmony due to a mismatch of role and goal, the wife may experience stress. Further, she may try to relieve this stress by either trying alternative job roles or by changing her goal. Thus, over time, the roles and the goals interact and change. Further the model proposes that the direction of change will be toward the two extremes (boxes A and F) shown in Figure 1. Evidence supporting the existence of role/goal stress is provided by Albrecht (1978), who found that one-third of the 147 promotable' women in her study would reject a possible promotion. The most common reason cited was related to the fear of added responsibility, while family considerations were also cited frequently. In order to reduce the potential role/ goal stress, Albrecht (1978) suggested that training emphasize the woman's social orientation toward the job (in other words, emphasize changing the woman's goals) rather than training that emphasizes technical skills.

The model proposes that most of the changes in the roles or goals are due to the desire to remove disharmony between them. One other type of incentive for change is included; we hypothesize that there may be an encouragement loop as well. The situation in which the encouragement loop would be most likely to occur is the case of a home-oriented wife working in a progressive career. In such a situation, we hypothesize that the wife will be encouraged to change her goal to conform with her role. An example of this encouragement loop was discussed by McCall (1977), when she projected that future workwives may identify more closely with the consumer behavior patterns associated with the social class of their employers than with that of their husbands.



The most important implication of this model is that, a any point in time, work status alone may not correctly reflect the perceptions of the wife. However, if the work role has been the same over time, then we should be able to assume that the wife's work role and her goals are in harmony. Thus, for those individuals who have maintained the same work role for a number of years, the measurement of occupational status by itself may serve as a proxy for both role and goal. On the other hand, measurement of both work role and goal, as discussed in the previous section, will be necessary for those wives who have not been able as yet to match theirs.


The original purpose of this paper was to investigate further Schaninger and Allen's (1981) use of occupational status to explain differences in the consumption of convenience products. We investigated the use of time-saving alternatives such as dining out and the use of a home-delivery foot service, and found that the delineation of high-occupational-status working wives from low-occupational-status working wives was able to explain differences in behavior for some, but certainly not all, of the dependent variables.

Our results suggested to us that one could develop a richer independent variable by incorporating Schaninger and Allen's (1981) approach with the role overload concept proposed by Reilly (1982). As such, we developed a model (shown in Figure 1) that discusses the interaction of the wife's work role with her goals and their impact on her perception of role overload. Further, we discuss the expected nature of the consumer behavior resulting from the desire to reduce the role overload. Finally, we acknowledge that this model is static in nature. Then we discuss the expected interaction between the wife's roles and goals over time, and the implication which the long run interactions have on the choice of the independent variables selected for use in investigating differences in the purchase behavior of working and non-working wives.


Albrecht, Maryann (1978), "Women, Resistance to Promotion and Self-Directed Growth," Human Resource Management, 17 (Spring, Number 1), 12-17.

Allen, Chris T., Kathleen Deberec, and Kenny R. Chan (1983), "Nontraditional Family Forms, Time Observation, and Household Behaviors: Perspectives and Research Directions for the Consumer Researcher," Proceedings, AMA Educational Workshop in the Changing Household--Its Nature and Consequences. Cambridge, MA.

Allen, Chris T. and Charles M. Schaninger (1980), "Dual Career, Dual Income, and Non-Work Wife Families: Perspectives and Research Directions," in the Proceedings, American Marketing Association Educators' Conference, R. P. Bagozzi, K. L. Barnhardt, P. S. Busch, D. W. Cravens, J. F. Hair, Jr., and C. A. Scott, eds., Chicago: American Marketing Association, 93-96.

Anderson, Beverlee B. (1972), "Working Women Versus Non-Working Women: A Comparison of Shopping Behavior,'in 1972 Combined Proceedings, eds. Boris W. Becker and Helmut Becker, Chicago: American Marketing Association, 355-359.

Bartos, Rena (1977), "The Moving Target: The Impact of Women's Employment on Consumer Behavior," Journal of Marketing, 41, 31-37.

Bartos, Rena (1978), "What Every Marketer Should Know About Women," Harvard Business Review, 56, 73-85.

Burns, Alvin C., Ellen Foxman, and Martha Myers (1984), "Role Load and Its Consequences on Individual and Household Consumer Behavior," Working Paper, Louisiana State University.

Drucker, Peter (1976), "Why Consumers Aren't Behaving," Wall Street Journal, December 1, 18.

Dunning, Bruce and Don Calahan (1974), "By Mail vs. Field Self-Administered Questionnaires: An Armed Forces Survey," Public Opinion Quarterly, 37 (Winter), 618-624.

Hafstrom, Jeanne L. and Marilyn M. Dunsing (1978), "Socioeconomic and Social Psychological Influences on Reasons Wives Work," Journal of Consumer Research, 5, 169-175.

Hollingshead, August B. and Frederick C. Redlich (1958), Social Class and Mental Illness, New York: John Wiles and Sons.

Kinsey, Jean (1983), "Working Wives and the Propensity to Consume Foot Away from Home," American Journal of Agricultural Economics, 65 (February, No. 1), 10-19.

Langer, Judith (1982), "A Typology of Working Women, Management Review, 71 (October), 19-21.

Lovelock, Christopher H., Ronald Stiff, David Cullwick, and Ira M. Kaufman (1976), "An Evaluation of the Effectiveness of Drop-Off Questionnaire Delivery," Journal of Marketing Research, 13 (November), 358-364.

Maret-Havens, Elizabeth (1977), "Developing an Index to Measure Female Labor Force Attachment," Monthly Labor Review, 100 (May), 35-38.

McCall, Suzanne H., (1977), "Meet the 'Workwife,'" Journal of Marketing, 41 (July), 55-65.

Mincer, Jacob (1960a), "Employment and Consumption, ' Review of Economics and Statistics, 42, 20-26.

Mincer, Jacob (1960b), "Labor Supply, Family Income and Consumption," American Economic Review, 50, 574-583.

Nickols, Sharon Y. and Karen 0. Fox (1983), 'Buying Time and Saving Time: Strategies for Managing Household Production," Journal of Consumer Research, 10 (September), 197-208.

Rapoport, Rhona and Robert N. Rapoport (1971), "Further Considerations on the Dual Career Family," Human Relations, 24, 519-533.

Reilly, Michael D. (1982), Working Wives and Convenience Consumption," Journal of Consumer Research, 8 (March), 407-418.

Reynolds, Fret D., Melvin R. Crask, and William D. Wells (1977), "The Modern Feminine Life Style, Journal of Marketing, 41 (No. 3, July), 38-45.

Rosen, Dennis L. and Donald H. Granbois (1983), 'Determinants of Role Structure in Family Financial Management," Journal of Consumer Research, 10 (September), 253-258.

Schaninger, Charles M. and Chris T. Allen (1981), "Wife's Occupational Status as a Consumer Behavior Construct," 8 (September), 189-196.

Sieber, Sam D. (1974), "Towards a Theory of Role Accumulation," American Sociological Review, 39, 567-578.

Stover, Robert V. and Walter J. Stone (1974), "Hand Delivery of Self-Administered Questionnaires," Public Opinion Quarterly, 37 (Summer), 284-287.

Strober, Myra H. and Charles 8. Weinberg (1977), Working Wives and Major Family Expenditures," Journal of Consumer Research, 4, 141-147.

Venkatesh, Alladi (1980), "Changing Roles of Women--A Life-Style Analysis," Journal of Consumer Research, 7 (September), 189-197.

Weinberg, Charles B. and Russell S. Winer (1983), "Working Wives and Major Family Expenditures: Replication and Extension," Journal of Consumer Research, 10 (September), 259-263.