The Direction of Consumer Complaint Research

Citation:

E. Laird Landon, Jr. (1980) ,"The Direction of Consumer Complaint Research", in NA - Advances in Consumer Research Volume 07, eds. Jerry C. Olson, Ann Abor, MI : Association for Consumer Research, Pages: 335-338.

Advances in Consumer Research Volume 7, 1980     Pages 335-338

THE DIRECTION OF CONSUMER COMPLAINT RESEARCH

E. Laird Landon, Jr., University of Houston

INTRODUCTION

My purpose in the first section of this paper is to present some concerns about the three papers in this session. No attempt is made to evaluate the net worth nor the relative merit of the papers, as each reader will be able to better judge with respect to the reader's frame of reference. The second section presents some observations on the state of consumer complaint research.

FORNELL AND DIDOW

Hypothesis 1

Using Hirschman's (1970) theory that dissatisfaction produces either exit (brand switching), voice (complaining), or loyalty(repurchasing), the authors reason in the following way:

"If the exit option was blocked or unavailable, complaining would be the only way the dissatisfied consumer could (overtly) react. Economic theory predicts that complaints would be more frequently registered in markets characterized by monopoly than in markets resembling the notion of perfect competition."

"H1: The number of voiced complaints relative to the number of dissatisfied consumers is negatively related to the number of competing organizations offering identical or similar goods and services in the relevant trading area."

If consumers have few or no alternative sources for a product or service it is reasonable to expect that exit behavior will be less likely than if several alternative suppliers are available. But whether voice responses are therefore more likely depends on other factors. As Hirschman (1970) hypothesizes the probability of voice depends on the value gained from a successful complaint times the probability of achieving a successful outcome. A key point regarding this first hypothesis is that monopolies may be expected to be less responsive to complaints than are firms in more competitive industries. The relative power of the buyer and seller is such that a consumer faces a take it or leave it proposition and the seller has very little reason to respond to a complaint. Consumers then may see less not more benefit from complaining. If this is true, then market concentration may reduce the probability of both voice and exit for consumers. The third option, loyalty, may be the most likely expected outcome.

If I were to categorize the 13 industries studied I would classify only telephone, natural gas, and water service as monopolies. The others are all very competitive. Clothing (92 outlets) and auto repair (78) are no more competitive than appliance repair (9) and jewelry (34). To use the number of yellow page listings as an interval scale measure of industry competitiveness is questionable. Further, the three monopolies are greatly regulated, particularly with respect to due processing of complaints. Thus, the possible inhibitor of voice responses (monopolistic indifference) is not clearly present.

Hypothesis 2

"...economic theory of rational choice postulates that the frequency of buyer-seller interaction affects the preference of voice. The more frequent the interaction ...the less likely it is that the consumer will be able to "afford" to invest the time and effort required to register a complaint..."

The hypothesis is not very compelling for at least two reasons. First, one might expect that frequent purchases could lead to a higher voice index because the cost of complaining is probably lower than for infrequent purchases. The consumer knows right where to go -- in fact, a complaint may be incorporated into the next purchase episode, and the consumer may expect better attention because she or he is a valued customer.

Second, the hypothesis will be difficult to test because of the likely confounding effect of satisfaction. Ceteris paribus, satisfied consumers will make more purchases than dissatisfied consumers -- both over time and with one particular seller. Therefore, a low voice index for frequent purchases may be a natural result of satisfaction, not the number of purchases. If this is true, a more elaborate test is required of hypothesis two.

The operationalization of the frequency of interaction for retail products are relatively straightforward, but I do not agree that the average number of interactions per year for water services, natural gas, and telephone are .16, .18, and .64, respectively. It seems to me that respondents were effectively restricted to the number of requests for repairs -- itself a voice response. How else can a 100% complaint to dissatisfaction ratio (voice index) be explained? On the other hand, many more interactions occur which may or may not lead to voice responses. Consumers pay for utility services monthly in most cases (this interaction would increase FI by 12!) I don't know why the authors excluded operator assistance. These omissions bias the FI measure in favor of the hypothesis, and do not seem reasonable.

GR+NHAUG AND ARNDT

Following an interesting and wide ranging introduction the authors state:

"...the empirical study to be reported will be directed towards propensity to complain as well as motives for not complaining in the two delivery systems." (public and private)

The authors give no clear reasons why this study was conducted nor how the study might be useful to researchers or decision makers in either delivery system. No hypotheses are presented, se it is not clear what the researchers expected or hoped to find. The data were collected in an area which "include(d) a medium size Norwegian town" without any guidance as to their usefulness in other settings. Response rate was high, 79%, but not discussed is the possible bias due to non-response.

Product selection was rather unfortunate. Different criteria were used for selecting private goods than public goods. The resulting private goods are all products, whereas the public goods are all services. The four selected goods in each area are neither representative of their sector, nor are they matched in a way which might make comparison meaningful. For example, health services are available in both sectors (in the U.S.), and a comparison of complaint behavior in both sectors might be useful. It would also be possible to pick products which differ on divisibility and/or number of suppliers (see Figure 1) within one sector to see if these factors explain between sector differences. Finally, the use of students as interviewers when many crucial questions are open-ended , is questionable.

BARNES AND KELLOWAY

Barnes and Kelloway compare consumer activists (members of the Consumers' Association of Canada -- CAC) to a provincial sample of Canadian voters. The comparison is intended to give businessmen a better understanding of the differences between these two groups.

The Canadian sample came from members of the CAC (55.5% response rate) and a random sample of a "...recent provincial voters' list" (27.3% response rate). Non-response may be a particular problem in the voter list because of self selection related to interest in the survey. Non-respondents and non voters may be more different from consumerists than are respondents.

The six hypotheses are "...based on earlier studies..." little of the earlier research is presented. Further, the material work of Sentry(1977) and Thorelli, Becker and Engledow (1974) are not cited. However, the most frustrating part of the paper is that the six hypotheses--four of which have 2 parts, are not directly and clearly analyzed.

1.  "Consumerists are more likely to perceive problems and to complain then are consumers selected at random."

Respondents were asked if they perceived problems with grocery items, appliances, and clothing if they had complained about a grocery product (in the last month) or about appliances or clothing (in the last year). The percentages were presented only for each category. I totaled the responses for CAC members and the voter samples across the three categories. The two groups do differ on the perception problems (29.6% to 23.7%; t=1.90; p<.05; one tail) but not on complaints (53.9% to 50.1%; t=1.06; p>.05; one tail). Of course, problem perception and complaint behavior is very likely to differ across product categories as can be seen from Table 1 in the Barnes and Kelloway paper.

2.  "Consumerists will express greater concern regarding social and consumer issues."

Hypothesis 2 appears to have two parts: social issues and consumer issues. Table 2 in the paper is titled "Consumer Concern with Social Issues" while Table 3 is titled "Attitudes Towards Consumer Issues". However, the issues listed in Table 2 clearly include a number of consumer issues, for example, inflation, fair deals for consumers, the costs of health care, etc. Furthermore, consumer issues per se are never discussed in the paper. The reference to Table 2 suggests that consumer and social issues are intermingled and the analysis pertains to the total.

Nevertheless, Table 2 presents equivocal results. Five of the 15 issues receive significantly different responses from the two samples in the direction indicating that CAC members hold more concern, 5 of the issues show significant differences with the voter sample showing more concern, and 5 of the issues do not show significant differences between the two groups. The authors merely report these significant differences without discussing their meaning nor accepting or rejecting the hypothesis. As a footnote, it is reported in the table that a 4.6% difference on the energy issue is non-significant, whereas a smaller difference of 3.7% on the pension issue is significant.

3.  "Consumerists will express more negative attitudes towards business as the cause of consumer problems and more positive attitudes regarding greater government in seeking solutions."

Table 3 presents 15 Likert statements and the percentages from each group agreeing with each. Seven statements which did not produce significant differences between the groups are not presented. Of the 15 Likert statements presented no analysis is done which relate specifically to negative attitudes towards business or to the involvement of government in consumer issues. Five statements do not appear to relate to either negative attitudes towards business nor government involvement in consumer issues. Seven appear to me to pertain to negative attitudes towards business, and three pertain to the role of government in consumer issues. Of the seven on business attitudes, four show significantly more negative attitudes by CAC members and three show significantly more negative attitudes by the voter sample members. In my opinion, this hypothesis is not supported by these data. All three government statements indicate significantly more support for government intervention by members of the voting sample, in direct opposition to the hypothesis.

4.  "Consumerists will be more likely to perceive as effective certain consumer actions taken to solve consumer problems."

Table 4 presents nine consumer actions with the ratings of the two samples as to the percentage who perceive each action as effective. Four of these nine are seen as being effective significantly more often by CAC members, two of nine are seen as effective significantly more often by members of the voters' sample and three of nine are not viewed differently by the two samples. However, the Table is poorly prepared. The text indicates that five of these differences are significant, whereas the Table shows that 6 are significant. Table 4 shows differences of 0.3 percent and 1.1% to be significant, whereas a difference on another action of 5.2% is reported to be nonsignificant. Further, the 5.2% difference which is non significant at the .05 level does not gibe with another difference of 5.4% which is reported to be significant at the .01 level.

5.  "Consumerists are more likely to make use of non-commercial sources to obtain information on consumer rights and major purchases."

Table 5 presents the use of information sources for the two groups for more than a dozen information sources. However, no attempt is made to separate these sources into commercial and non-commercial categories and determine whether the result supports the hypothesis. Only two categories are commercial in nature: newspaper -TV - radio and banks. Two of the 505 CAC respondents indicate newspaper -TV - radio and none indicate banks. There is no significant difference between these total mentions for commercial sources, therefore one may conclude that for information on consumer rights, these two samples did not differ with respect to preference for commercial or non-commercial sources. For the CAC sample 27.5% of all information sources used would be commercial in nature, whereas 34% for the voter sample would be commercial in nature. The difference is significant indicating support for the hypothesis. These percentages were adjusted both for the number of respondents who could not recall their past use of information and for the total number of information sources used.

6.  "Consumerists will be more likely to perceive little progress in recent years in the resolution of consumer problems."

Table 6 presents the results of the two sample respondents to 10 issues. Respondents were asked to indicate whether things are better or worse today than they were 3 years ago in each of these areas. Unfortunately, the authors do not present the number of respondents who indicated they were not sure, even though this category appears to have been included in the calculation of Chi-square values and probably had a considerable influence. Nonetheless, it appears as if the two groups differ on their perceptions of a change in these consumer areas over the past 3 years. In some cases, CAC members believe there has been more positive change and in other cases the voter sample feels there has been more positive change.

ISSUES IN CURRENT COMPLAINT RESEARCH

What is a Complaint?

Complaining behavior has not been very thoroughly described or defined. Because most studies have been empirical in nature, operationalization has been done, but these operational definitions are specific rather then global, and are often limited by the nature of available data. For example, studies often examine the complaints received by companies which are recorded, often written complaints. Oral complaints, because they are not recorded are less often studied. Yesterday a sales clerk totaled my lunch order incorrectly. When I complained the correct total was calculated. Complaints of this type are certainly not recorded and have not to my knowledge been studied.

The overwhelming diversity of complaint behavior insures that any one study deals with only a small part of it. Therefore, researchers must clearly indicate what type is being studied. Also, work would be quite helpful on a global definition and taxonomy of complaints.

The American Heritage Dictionary defines a complaint as "an expression of pain, dissatisfaction or resentment.'' To get us started, I offer the following definition of a consumer complaint:

An expression of dissatisfaction on a consumer's behalf to a responsible party.

Any expression of dissatisfaction would classify communication as a complaint, even if the communication includes expression of satisfaction. Firms receive letters from consumers which are mixtures of praise and scorn. For the purpose of this definition, a letter which indicates dissatisfaction in any way would be counted as a complaint.

Communications from consumers may or may not give reasons for the stated dissatisfaction. If reasons are given, they may vary considerably from complaint to complaint. Complaints may be categorized according to the reasons given for the dissatisfaction. Also, the medium of communication of the dissatisfaction may be studied.

The expression of dissatisfaction must be given on the consumer's behalf. The consumer must have a purpose for expressing the dissatisfaction. This purpose may be to receive restitution, to protect other consumers, or to assist the firm in correcting a problem. Regardless of the specific reason, the expression of dissatisfaction must have some consumer motive in order for it to qualify as a consumer complaint. Without a consumer motive, the expression of dissatisfaction is difficult to analyze.

To qualify as a complaint, the expression must be made to a responsible party. Casual expression of dissatisfaction to one's friends may be important for the marketer to study, but does not qualify as a consumer complaint. The responsible party may be within the channel of distribution or may be a third party outside the channel. Examples of these extra channel third parties would be the Better Business Bureau, attorneys, and the federal government. These third parties may advocate on the consumer's behalf to a person or firm who can take action.

As the number of studies of consumer complaint behavior continues to increase, it will be useful for research to define complaints and categorize them according to these dimensions.

Surveys Have Problems But Can Be Useful

Among other problems, surveys of consumer complainers are plagued by self-selection and self report problems. Even though most consumers have had dissatisfactory experiences in the market place, consumers who are relatively pleased and not currently dissatisfied, are less likely to respond and participate in surveys on complaining behavior than are consumers who are currently dissatisfied or are presently dealing with a complaint.

Surveys necessarily depend on respondents self report of complaint behavior. Problems are very likely with complaint studies in this area. Consumers memories of dissatisfactory experiences and complaint behavior are likely to be flawed. Selective retention and recall seems particularly possible. Second, strong social norms may influence respondents to report complaint actions when actions did not occur. After lengthy questioning on a survey about dissatisfactory experiences, respondents are likely, when asked, "What did you do about your dissatisfaction?" to respond that, "I complained," when in fact the consumer may have done nothing about the problem. Finally, self interest on the consumer's part may influence the self report of complaint behavior. Some organizations audit their complaint handling units by sending questionnaires to consumer complainers after the complaint has been resolved. Consumers may believe that if they are still dissatisfied the firm may give a better response. These consumers misreport their satisfaction.

Survey researchers of complaining behavior have not done very thorough or sophisticated analysis of survey data. For example, several researchers have reported that complainers, when compared with non-complainers tend to be younger. However, very little has been done to discover why this result is obtained. Are complainers younger because they have more education than older consumers? Is it because younger people tend to have more liberal views than older people? Is it because the young tend to feel a sense of social outrage that vents itself in a consumer complaint? Is it because younger consumers have more at stake when the products go wrong because they have relatively less income than older consumers? Or is it because younger consumers have not yet learned that complaining behavior doesn't get anywhere? Ralph Day has suggested that the relationship between age and complaining behavior may be product specific. Because most products are purchased by younger people in the full-nest family life-cycle stage, it makes sense that they would complain more often. However, when products which are used by older people are considered, the relationship with age may not obtain.

Surveys can help answer these questions. In fact, many of the types of data necessary to answer these questions are available in the various surveys which report the age-complaint relationship. However, analysis of data bases has not occurred. I am sure that as particular issues become more clearly focused in the growing area of complaint research, more sophisticated research techniques will be used to resolve them.

Surveys can be very useful in selecting data to address these problems. However, more attention should be paid to the generation of specific testable hypotheses, the development of operational measures which are valid and reliable, and an increased use of multivariate analysis on survey data.

Complaining is a Process

It is possible to study consumer complaints by examining the complaint and understanding the reasons for it. But it is also extremely useful to study the complaining process. This process involves the consumer's evaluation of dissatisfaction and the choice to make the dissatisfaction known. The complaint is then made to the responsible party and some type of response is given. The consumer then evaluates the response and may decide on further action. The interaction between the consumer and the responsible party represents a process of confrontation and negotiation. Presently the largest amount of research on this process is being conducted in large commercial organizations in the United States. These organizations are attempting to understand the causes for consumer complaints so that responses can lead to consumer satisfaction and continuing purchase of the company's product. It seems to me, that we will all benefit from the process perspective in seeking to understand the causes, types, motives, and style of consumer complaining.

REFERENCES

Barnes, James G. and Kelloway, Karen R. (1980), "Consumerists: Complaining Behavior and Attitudes Toward Social and Consumer Issues," in Jerry C. Olson (ed.), Advances in Consumer Research, Vol. VII.

Fornell, Claes and Didow, Nicholas M. (1980), "Economic Constraints on Consumer Complaining Behavior," in Jerry C. Olson (ed.), Advances in Consumer Research, Vol. VII.

Gr°nhaug, Kjell, and Arndt, Johan (1980) "Consumer Dissatisfaction and Complaining Behavior as Feedback: A Comparative Analysis of Public and Private Delivery Systems," in Jerry C. Olson (ed.), Advances in Consumer Research, Vol. VII.

Hirschman, Albert O. (1970), Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States, Cambridge: Harvard University Press.

Landon, E. Laird, Jr., "A Model of Consumer Complaint Behavior," Ralph L. Day (ed.), Consumer Satisfaction, Dissatisfaction and Complaining Behavior, Indiana University, 1977.

Sentry Insurance, "Consumerism at the Crossroads," 1977

Thorelli, Hans, Becker, Helmut, and Engledow, Jack, The Information Seekers (Bloomington, Indiana: Indiana University Press, 1974).

----------------------------------------