Comparison of Consumer and Store Manager Attitudes: the Case of Racial Effects in Inner City Retailing


Alan R. Andreasen, Marcus Alexis, George H. Haines Jr., and Leonard S. Simon (1971) ,"Comparison of Consumer and Store Manager Attitudes: the Case of Racial Effects in Inner City Retailing", in SV - Proceedings of the Second Annual Conference of the Association for Consumer Research, eds. David M. Gardner, College Park, MD : Association for Consumer Research, Pages: 56-75.

Proceedings of the Second Annual Conference of the Association for Consumer Research, 1971     Pages 56-75


Alan R. Andreasen, State University of New York at Buffalo

Marcus Alexis, Northwestern University

George H. Haines Jr., University of Rochester

Leonard S. Simon, Community Savings Bank and University of Rochester

[The research reported in this paper was supported by a grant from the Consumer Research Institute, Inc. and by the Systems Analysis Program, The University of Rochester, under Bureau of Navy Personnel Contract No. N00022-70-C-0076. However, the conclusions, opinions, and other statements in this paper are those of the authors and not necessarily those of the Consumer Research Institute, Inc. or the Systems Analysis Program. The authors are indebted to Charles Miersch, Ronald Homer, John Mosely, and Marilyn Dapses for assistance in data collection and to Jose da Costa for assistance in data processing.]

[Associate Professor, School of Business Administration; Professor, Department of Economics; Associate Professor, Graduate School of Management; Vice President, Marketing, and Associate Professor, Graduate School of Management respectively.]

This paper studies whether consumers must cope with inner city retail store managers whose attitudes about the effect of racial factors upon market performance differ from those of the consumers. The fundamental issue is whether racial attitudes create market imperfections which cause all participants, consumers and retailers alike, to be worse off than if differential perceptions of racial effects did not exist. It is almost impossible to provide a definitive answer to this question. However, the present study attempts to provide some insights into this issue.

A second issue is whether effects due to place of residence of consumers dominate effects due to race. A classic statement of this issue is given by Greenberg and Dervin (1970, p. 225):

"The...hypothesis...(is) to expect considerable similarity between low-income blacks and low-income whites. Considerable research suggests that the poor live in a subculture that makes them much alike no matter what their race or ethnic origin.


Andreasen undertook a study in Buffalo, New York, of the nature of businesses in the inner-city and of their operators in 1968 (Andreasen, 1968). Among other things, he asked five attitudinal questions designed to measure the views of store operators toward their customers. A subsequent study in 1969 of inner-city and central city consumers in Rochester, New York, administered the same attitudinal questions to consumers. The intent was to allow a comparison of the Buffalo study data and the Rochester study data.

The methodology of applying the same attitudinal questions in different cities is obviously strange. However, there is no evidence that attitudes on racial matters differ among metropolitan areas in upstate New York. [The original research design of the Rochester study included collection of information about racial attitudes held by retail businessmen in Rochester, Such data would have allowed exploration of whether differences in attitudes exist between metropolitan areas. This stage of the design was never carried out because funds for the necessary data collection were not available.] Further, previous research (Haines, Simon, Alexis, 1971) has shown that the structure of retail enterprise types existing in the inner city in Rochester is not significantly different from that identified in the Andreasen study of Buffalo. Finally, there are great similarities in the general characteristics of the population in the inner-city portions of both cities in terms of percent non-white, percentage, population change over the four to six year period from 1960, and in 1960 census median reported income. Therefore, the possibility of testing for racial attitude marketplace effects proved too intriguing to pass up.

The five attitudinal questions used in the Rochester study which were common to both studies are displayed in Exhibit I. The wording used in the Rochester study was slightly different from that used in the Buffalo study. The substance of each attitudinal question remained the same. [The exact wording used in the Buffalo study may be found in (Andreasen, 1970, p. 328-329).]

A sixth attitudinal question was added in Rochester: "Black consumers prefer to shop in stores that hire Black people. How do you feel about this?" This question is not directly comparable to anything in the Buffalo study, but does yield some interesting findings in and of itself.


The inner city area selected for the Buffalo study is comprised of Buffalo's Model Neighborhood Area plus an area to the northeast towards which the black community is moving. The total area is approximately twenty-five blocks wide and thirty-five blocks long.



The Rochester study collected consumer survey data in two inner city neighborhoods and two other central city neighborhoods chosen for purposes of comparison. All are within the City limits of Rochester. Brief descriptions of the areas are:


Almost completely white; many single, young persons and old, retired persons; few families with children; single and two family housing predominates; lower middle class outlook.

19th Ward

Black homeowning families penetrating one border; almost all families, many with grown children; economically solid middle class. Single and two family houses.

Third Ward

About 608 black and definitely in transition; mostly families but some singles; primarily one, two, and three story housing; economically, lower middle and upper lower class; definitely a slight step up from the worst ghettos in the cities.

West Half of Model Cities

Named for urban renewal project; heaviest concentration of blacks in city; traditional ghetto for -all types of backgrounds; deteriorated housing; economically, contains many of the very poor and those on relief; many multi-unit dwelling units and some high rises; mixture of families and single persons.


A recurring question in the analysis of attitude data has been that of what measurement properties should be attributed to the data (Coleman, 1964). The present study assumes only that responses can be classified into one of six groups: the five groups shown in Exhibit I, and a sixth group, into which all respondents who refused to answer the attitudinal question are classified. The model used is a multinomial distribution with five parameters. [An elementary discussion of the multinomial distribution and its properties is given in 11.] The analytical method employed in processing the data to yield statistics which bear on the questions outlined above is a test for a significant difference between or among multinomial distributions. Each of the response categories is considered as one outcome on a multinomial distribution, and comparisons are then made across neighborhoods, store types, race, and consumers and store operators. The test procedure gives an empirical x2 value, along with the degrees of freedom associated with the empirical x2. This statistic is then used to test the null hypothesis of 6 no significant differences between the observed multinomial distributions. [A full description of the test derivation and use may be found in (Pothoff and Whittinghill, 1966). A BASIC program to perform the computations is available from the authors on request.]


Each of the attitudinal questions from Exhibit I will be referred to by a single adjective (in order to simplify the discussion) as follows: (1) "Interest", (2) "Quality", (3) "Over-Charged", (4) Shoplifting", and (5) "Manners".

The first set of tests performed was for homogeneity of operators attitudes by store type on the Andreasen data. The null hypothesis of homogeneity could not be rejected except on the Quality and Overcharged attitudinal questions at an 0.05 level of Type I error. The Andreasen store types were then further split into those that were primarily retail units (categories 5 to 18) and those that were primarily service units (categories 19 to 26). The null hypothesis of no difference then was not rejected on any of the attitudinal questions. It may therefore be speculated that something in the character of a service store operation leads to differential operator response on those attitudes which focus on quality and over-charging. These results are Presented in Table 1.

The second set of tests performed was upon the attitudes of consumers in the Rochester study across the four areas of the central and inner city in which respondents had been questioned. Table 2 presents these results. The null hypothesis of homogeneity was not rejected on any of the attitudes except that pertaining to Quality. Figure 1 presents the empirical data for this attitude question. It is perfectly clear from this Figure that Maplewood responses are different. Therefore, the Maplewood group was separated out and a test for homogeneity performed on attitudes toward quality using only the remaining three areas. The null hypothesis of no difference could not be rejected. [A possible alternative explanation would be that the difference arises from differences in the non-response and don't know categories. Therefore, the homogeneity test on Quality was re-run using only responses in categories 1-4. Once again the null hypothesis of homogeneity is rejected at * =.05, which indicates the data do not support this alternative explanation.] Maplewood is an almost exclusively white lower middle-class neighborhood; in fact, the sample of Maplewood respondents is entirely white. The difference on the Quality attitude on the part of the Maplewood residents is particularly interesting since the Quality attitude was one of the two which had to be examined differentially by type of store.

The responses of all the different owners and operators of all types of stores were then pooled based on the fact that the null hypothesis of homogeneity of attitudes could not generally be rejected. Similarly, the responses of the respondents in the Rochester study in all four areas were also pooled again taking advantage of the fact that there was generally homogeneity of responses. When the two groups are compared, the null hypothesis of no difference is rejected on all the attitudinal questions, if as in the case of the Quality attitude, the Maplewood consumers are excluded on the findings mentioned above. If the Maplewood area only is used to represent consumer views on Quality, then the null hypothesis on that question is not rejected. These results are presented in Table 3. Again, a differential result due to the effects contributed by the Maplewood residents is identified. There are apparently different attitudes held by consumers and store operators. This disparity in viewpoint may be a substantial contributing factor to the general hostility that is sometimes reported occurring between consumers and store operators in the inner-city.







The data for operators of stores in the inner-city in Buffalo then was divided to control for race and ownership versus management status. When black operators were compared, no significant differences in attitudes among female owners, male owners, and managers could be found. Similarly, the null hypothesis cannot be rejected when making the same test on the white operator group. All the black respondents store operators were then pooled and contrasted to all the white operators pooled. Again a test for homogeneity was conducted. In this particular instance, the null hypothesis of no difference was rejected in every case except on the attitudinal question having to do with Manners. Again, apparently there is a substantive difference in viewpoints between the black store operators and white store operators with the exception of attitudes having to do with Manners. These results are reported in Table 4. [Once again it can be seen that there is no change in the qualitative nature of the results if categories 5 and 6 (don't know and non-response) are excluded (see Table 4, part IV).]

The Rochester data were then divided to control for race. A test for similarity of attitudes was then made across the respondents in the four different areas of the city in which samples had been taken. The sixth attitudinal question, which appeared only in the Rochester data and which is referred to as "Blacks Hired", is included in this analysis. Table 5 presents the results. The null hypothesis of no difference could not be rejected on any of the six attitudinal scales for data from all the black consumers only and for data from all white consumers only. The black consumers were then pooled and compared to the pooled white consumers. The null hypothesis of no difference is rejected in every case but Manners [The result is essentially the same if the analysis is performed excluding "don't know" and no response". The only change is in shoplifting where in this analysis the null hypothesis of homogeneity cannot be rejected.]. The similarity of this finding to that of store operators in the Buffalo study when race is controlled is striking.

These results clearly suggest one final set of tests which compare consumer's attitudes against those of store operators controlling for race. The following separate analyses are made: 1. black consumers compared to black merchants, 2. black consumers against white merchants, 3. white consumers versus black merchants, and 4. white consumers contrasted to white operators. The results of this analysis are presented in Table 6. The null hypothesis of no difference in attitudes is rejected on all attitudes except Shoplifting in the first two cases. [The analysis based only on response categories 1 to 4 includes some additional cases where the null hypothesis cannot be rejected but the impression of substantial differences in attitude remains.]

It should be recalled that the pooled black consumers, pooled black store operators, or pooled white store operators had no significant differences in attitude on this particular question.

The nature of the difference in attitudes on the other questions is exemplified in Figure II,which shows the overwhelmingly less favorable response of white store owners compared to black consumers on the quality attitude question. For the third group, white consumers compared to black store operators, the null hypothesis of no difference was rejected in every case. However, only in the case of manners was the null hypothesis rejected when the don't know and no response categories were removed from the analysis; in this case, apparently much if not all of the observed difference apparently from responses to these two categories. The same pattern is seen in the fourth group of tests which compare white consumers to white retailers; in fact the only difference in results between this group and the preceding group is the similarity of responses on the Quality question; the null hypothesis of homogeneity cannot be rejected for this question in either of the question set tests. This again may be the effect of the Maplewood consumers.












1. Attitudes about the effects of race which might affect retail market operation are largely similar in inner city store operators across diverse retail store types.

2. This similarity also is found within race: consumers of similar race have similar attitudes, regardless of residence place and retailers of similar race have similar attitudes, regardless of sex and whether they are an owner or a manager of the retailer business.

3. Attitudes about the effects of race are different when retailers and consumers are compared.

4. Attitudes about the effects of race are different when black and white retailers and consumers are compared.

5. Attitudes are different when retailers and consumers are compared controlling for race.


All consumers seem to see retailers as "the enemy". There is no mitigating effect of race. Very few consumers seem to adopt the attitude of store operators, and vice versa. Being a consumer puts one on a different side of the marketplace, sufficiently so that even though a store operator may be of the same race, the consumers attitudes are not likely to be like those of the store operator.

An often stated hypothesis is that one reason living and working in inner city neighborhoods is intolerable is that the people living and working in these neighborhoods have such different perceptions of each other that they cannot communicate to one another. This lack of communication prohibits any effective group action to better environmental conditions for everyone. The results of this study could not be held to reject such a view, although it is still possible that the results reflect differences between the urban areas where the data were collected.

The results are quite clear, on the other hand, that on the consumer data effects due to race dominate effects due to place of residence. This is the exact opposite of the hypothesis stated by Greenburg and Dervin. There can be no comfort in these results for those who would claim blacks and whites have similar attitudes, and that observed differences in attitudes can be explained away by place of residence.


Andreasen, A.R. Ghetto Business: A Study of Business and Businessmen in the Core Area of Buffalo, New York, School of Management, State University of New York at Buffalo: June 30, 1970.

Coleman, J.S. Introduction to Mathematical Sociology, New York, N.Y.: The Free Press, 1964, Ch. 2.

Freeman, H. Introduction to Statistical Inference, Reading, Mass.: Addison-Wesley, 1963, Chapter 14.

Greenberg, B. and Dervin, B. "Mass Communication Among the Urban Poor", Public Opinion Quarterly, (Summer, 1970: pp. 224-235).

Haines, G.H. Jr., Simon L. S. and Alexis M. "The Dynamics of Commercial Structure in Central City Areas", Journal of Marketing, Vol. 35, No. 2 (April, 1971; pp. 10-18).

Potthoff, R. F., and Whittinghill, M. "Testing for Homogeneity: I The Binomial and Multinomial Distributions: Biometrika, Vol. 53, No. 1 and 2 (1966; pp. 167-182).



Alan R. Andreasen, State University of New York at Buffalo
Marcus Alexis, Northwestern University
George H. Haines Jr., University of Rochester
Leonard S. Simon, Community Savings Bank and University of Rochester


SV - Proceedings of the Second Annual Conference of the Association for Consumer Research | 1971

Share Proceeding

Featured papers

See More


Is Congruity Desirable for Brand Extensions? A Conceptual and Meta-Analytic Review

Qian (Claire) Deng, University of Prince Edward Island
Paul Richard Messinger, University of Alberta, Canada

Read More


Unexpected-Framing Effect: Impact of Framing a Product Benefit as Unexpected on Product Desire

Monica Wadhwa, INSEAD, Singapore
Christine Kim, Hong Kong University of Science and Technology
Amitava Chattopadhyay, INSEAD, Singapore
Wenbo Wang, Hong Kong University of Science and Technology

Read More


B8. Factors Influencing Collaborative Consumption Usage in the US market: An Exploratory Study

Pia Annette Albinsson, Appalachian State University
B. Yasanthi Perera, Brock University, Canada
Bidisha Burman, University of Mary Washington
Lubna Nafees, Appalachian State University

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.