American Advertising and Societal Values During the Twentieth Century

AMERICAN ADVERTISING AND SOCIETAL VALUES DURING THE TWENTIETH CENTURY

Richard W. Pollay, University of British Columbia [Curator, History of Advertising Archives, Vancouver, Canada]

[This paper was prepared for the Association for Consumer Research conference, "Historical Perspectives in Consumer Research," July 1985. The reported research was funded by Canada's Social Science and Humanities Research Council, and assisted by Ms. Yvonne Jordan. Access to the whole data base is readily available to interested scholars.]

ABSTRACT -

Scholars in many fields have expressed their concerns as to the potential or actual effects of aggregated advertising on the culture and, more specifically, the values of its citizens. This paper reports a synthesis of the hypothesized effects based on the intellectual history of North America. To enable testing of some of these hypotheses, a method was developed to content analyze advertisements for values manifest. The results of such an analysis of 2,000 ads from 1900-1980 in U.S. magazines is presented. These data are compared to similar data for contemporary U.S. television ads, and to self-descriptions by the population and segments based on sex, age, social class and ethnicity.

INTRODUCTION

Although they have received relatively little attention in consumer behavior research traditions, values are a basic -concept in understanding behavior and culture. They are invaluable in characterizing advertising and consumers, whether collectively as a common culture or as differentiated segments. Values are the very soul of advertising. The strategic decision of what product attributes or consumer benefits to focus on is a value decision, answering the question "what is, or should be, of importance to the consumer?" All of the art work and copy writing is an effort to enhance valuation by consumers of the product and its aspects in focus in a particular ad.

The value dimensions of advertising are also at the heart of the critique of advertising, whether judging the effectiveness of individual ads or concerned with the social and cultural ramifications of the larger system of commercial activity. The knowledge that advertising unavoidably communicates and ratifies certain values is of much concern to many social scientists. They see advertising as a highly important socialization institution in societies because it is both pervasive and persuasive. The presence of commercial messages in multiple media, not only the obvious major media, but also on street signs, clothing, vehicles, etc., make the commercial culture environmental to the individual. Thus, the relentless persuasion on behalf of consumption is not only impossible to avoid, but also hard to detect as its commonplaceness leads to its being taken for granted.

Much professional talent and money is spent to maximize potential effectiveness of ads through both consumer behavior research and advertising production values, making the commercial component of media content the most likely element to be impactful and persuasive. The pervasive and persuasive nature of advertising makes most social scientists quite concerned, especially when they consider the diminishing socialization roles played by other social institutions in affluent societies, institutions like organized religion, education, law and the courts, extended families, traditional cultures and philosophies, etc. While advertising is in the ascendancy, these others seem to fade in importance as influences on the new generations of youth and hence the emergent Culture in many societies. Advertising may be one of the most potent factors eroding traditional cultural character and leading to a transnational consumer culture.

Because of this possibility, it is worth studying, in a far more systematic manner than has been done to date, the role of values in advertising. Worthy of inquiry are the questions like what is the nature of values?, how do they change?, how are they manifest in advertising?, what are reasonable hypotheses about the value consequences of advertising?, how can manifest values be measured?, what does such a measurement suggest are the value characteristics of advertising?, and how do these characteristics relate to the value character of society and the direction of its change? Each of these questions will be explored in this paper for the North American context beginning with the issue of what hypotheses are worth researching. While the literature and data reviewed are North American, the issues and research methodology are universal.

ESTABLISHING A RESEARCH AGENDA

There has been a long standing debate in Western societies about the social and cultural role of advertising. The evolution of advertising into a national, big budget, professionally executed activity around the beginning of this century led many commentators to express concern with advertising's influence on traditional cultural values. This was true not only in North America and England, but perhaps most dramatically in Germany where the highly evolved scholarly community of the 1920s write much on commercial culture (Fullerton and Nevett 1985). This early literature, particularly in the U.S., was not very articulate in a way that would facilitate research, as it tended toward either the muckraking fault-fiading or the abstract, aesthetic and philosophic. This continued to be the case through the depression years of the 1930s, even with the contributions of critiques by experienced admen (Rorty 1934). But as the social sciences matured in the post WWII era, the precision and articulateness of the discussion advanced considerably, although to date it has not yet much influenced  the research of consumer behavior scholars.

This may begin to change, for a new major literature review is forthcoming in the Journal of Marketing (Pollay 1986). This review summarizes the conventional wisdom of North American intellectual thought about the consequences of advertising. Much of this stems from and/or expresses concerns about the value character of commercial persuasion. The long list of possible effects is shown in Exhibit 1, but note that few of these allegations are supported by data. Most are, however, supported with convincing rhetoric by famous and influential scholars and major international organizations like UNESCO (MacBride 1980). The full list seems an impressive indictment of advertising. if even a fraction of these charges were to be validated by research, that research would be of social importance and moral urgency. Failure to explore these research directions would expose the consumer behavior intellectuals as servants to, rather than objective scholars independent of, corporate practices.

The authors contributing to this critique range quite broadly in their knowledge of advertising practices and its constraints. They also vary on the extent to which they attribute power to advertising. At the extremes some write as if advertising were an all powerful- mercenary martipulator, a corrupting influence in an otherwise innocent culture, while others write as if advertising were the innocuous innocent, merely reflecting aspects of society. Neither extreme position is easily defended, and yet both have some validity. While advertising does indeed work within a cultural context, the "mirror" of advertising is a very distorted one, for advertising operates very selectively in its portrayals of some values, some lifestyles, some motivations, etc. Thus, the feedback to the culture through commercial communication is a highly selective reinforcer of certain values. This reinforcement may well alter the audience's value hierarchy by effecting the salience of various values. Both extreme positions, with their metaphorical images of puppet mastery or passive mirroring, capture the interdependence between culture and communication and suggest the selective reinforcement process. The selectivity of advertising's employment and validation of certain values derives from the facts that not all values are as readily dramatized in ads, are as readily responded to by consumers, are as readily attached to available products, or are values cherished by the advertisers themselves.

ON THE NATURE OF VALUES; ASPECTS OF AXIOLOGY

Before we can hope to study values we must, of course, establish some definitions and discuss the importance of values in everyday life and how values change within society and within individuals. Values are a very basic concept. We often talk about the hierarchy of attitudes, opinions and beliefs, noting each successive concept as the more fundamental and durable dimension of character, the less ephemeral and changeable. Values are particular kinds of beliefs that incorporate criteria or standards of judgment. For example, Rokeach (1973, p. 5) defines values as "an enduring belief that a specific mode of conduct or end state is  personally or socially preferable   (to its converse). A value system is an organization of (such) beliefs." Note that values apply to both processes and their consequences, both means and ends, and hence the universe of values can be partitioned into "instrumental" and "terminal" values. Within the broad classification of instrumental (means) and terminal (ends) values, other distinctions are of interest. Terminal values can be segmented into those which are personal (salvation) or social (peace). Similarly the instrument values can be segmented into those that are personal competence (logical) and those that are interpersonal moral values (forgiving, obedient or helpful). This distinction is manifest in the feeling evoked when these values are violated. violation of personal values produces shame and a sense of inadequacy or unworthiness, whereas violations of moral values leads to quilt.

Note that values are a critical basis for decisions, judgments and even simple perception, the most elemental judgment. They are also, of course, evident as the foundation for religions, cultural belief systems, political ideologies and judicial processes. They act, however, in everyday life, not just in abstract cultural institutions, as all goal directed behavior implicitly expresses the   "value" of the chosen goal.

Values are therefore both prosaic and profound, commonplace in daily life and also at times apparently worth even the sacrifice of life itself. Values underlie all resource allocations, like a consumer's allocation of money, time and/or effort. Family and social reward and punishment systems are built upon a value hierarchy. Our judgments of others, and hence our presentation of self to society, is also structured around value sets. Political decisions, like personal consumer choices, reflect value preferences.

The most difficult of choices involves the competition and trade off between competing value premises. In isolation all values are, by definition, a "good," a desirable and laudable behavior or consequence. Indeed, we value so many things that most serious political decisions rest on the trade off between the relative importance of several values, such as family sanctity, justice, security, autonomy, democratic participation, peace, equity of opportunity, individual or economic progress. What distinguishes major political camps is the relative importance of these and other goals.

So too with the individual. safety, luxury, reliability, While "economy, generosity, or cleanliness," may be widely held values in a culture, what distinguishes the character of one individual from another, or one subculture from another, is often the relative importance of values, or their value hierarchy. Cultures typically also have rules governing resolution of value conflicts, shared understandings as to which values dominate under what conditions. For example, cultures have norms about how a son should decide between staying with aging parents (family loyalty) vs. leaving to seek his fortune (economic prosperity). The cultural rules governing such value conflicts are built up over generations and embodied in cultural traditions. Normally the nature of the value hierarchy implicit in such rules imply is not obvious.

Values, like other environmental variables, are typically taken for granted as "human nature" or "the way we are" until they are threatened or violated. At this stage they are so important that cultures and individuals are willing to risk all to maintain certain values.

Values are traditionally very slow to change, but with the emergence in recent history of professional persuasion, potential value change is accelerated as persistent value reinforcing communication competes with the traditional sources of value education, families, schools, religious orders, the courts, etc. Because advertising seems to be in ascendency while the role of competing social institutions seems to be in descendency, much concern has been expressed, as seen above, about the value character of advertising and the manner in which it might work to create a new generation dramatically different from its forefather's culture.

HOW VALUES CHANGE

Value change can be dramatic, as in total religious conversions and cataclysmic political transformations, where some values are abandoned and others embraced. More commonly it is a gradual process whereby the relative importance of various values is modified slowly. The source, or stimulus, for value change can be either personal knowledge, social reinforcement, or other persuasive external agents. Personal knowledge can lead to value change by changing perceived linkages. For example, learning that some industries are polluting, and that pollution is potentially harmful, may change the valuation of that industrial activity. Personal knowledge can also change values when an individual passes through a critical incident, such as a disillusionment, or has a transcendent religious revelation. Social reinforcement of values occurs virtually all the time as friends and family express approval and disapproval of those around them, thereby exposing their norms and values. Peer groups and even reference groups which the individual may only admire from afar can function as value models. Even the relative importance of family vs. friends, the elderly vs. peers, or historical vs. the aspirational as sources of valid influence on the individual is an expression of that culture's relative valuing of these groups. Lastly, change can occur as a result of the action of an external agent, like the teacher, the preacher, the political speaker or the advertiser. Here too, change can be dramatic or gradual, but gradual change, as might more commonly be the result of seductive commercial culture, is none the less important.

There are at the least seven processes of change an individual might experience. At one extreme is (1) the total ideological conversion, be it political or religious, with wholesale acquisition and abandonment of entire sets of values. Political values can change (2) by revolutionary means, as when there is a collapse of political consensus, or (3) by evolutionary means, as social order goes through stages of environmental adaptation. Dominant values within a culture or subculture change through (4) an adoption/ diffusion process, as new ideas gradually spread. Individual change in the role of values can be effected by (5) rescaling, the setting of new expectations or aspirations; (6) redeployment, the extension or restriction of the domain over which a value is seen as relevant; and/or (7) salience shifts, revised perspectives on the topicality or emphasis various values deserve. Values are often invisible as they are exactly those things which cultures take for granted, but they become visible and salient when change occurs. Cultural actors, like parents, may spend considerable effort to communicate traditional values when they are seen as directly challenged. Lastly, some value changes produce 'cascade effects' as the change in the subset of value importances upsets the multiple equilibria of social rules and roles concerning value dominance. The changes in some core values may undermine the relative importance of other values and the whole hierarchy becomes disturbed, just as a grocer's display of oranges might collapse if items are taken from the bottom row. For a more detailed discussion see the branch of philosophy known as axiology (Rescher 1969).

It has been shown that the importance of certain values, and hence the overall structure of a value hierarchy, can be changed by experimental manipulations (Rokeach 1973, 1979a, Rokeach and Grube 1979). it follows that continuing reinforcement will work to perpetuate such changes, leading to more durable restructuring of value systems. Thus it is of importance to question what values are reinforced by advertising, not just occasionally, but on an ongoing basis. The persistent reinforcement of certain value premises by their use, validation and display can induce long term changes in value systems. Note that this is true even though individual ads may be less than totally effective. The persistent themes that transcend individual ads are the keys to understanding the macro effect of a commercial system. The effect of the aggregate totality of commercial persuasion may be consequential even when individual ads have limited power to convert behaviors, just as a flood has far more power than a raindrop.

To repeat, values normally change very gradually as tradition slowly gives way to a cultural adaptation to new circumstances. The industrial revolution, with its component revolutions in communication technologies and the arts of persuasion, create a situation today where in many cultures a new commercial culture is in direct clash with a traditional, nonmaterialistic culture. Identifying the ways in which advertising has played a role in encouraging this cultural revolution requires a discussion of how advertising employs and deploys values and how the value character of advertising might be measured. This measurement methodology can then be applied to sample advertisements to identify the value profile of commercial persuasion. The pattern of this profile can then be compared to data about actual change in society.

HOW VALUES ARE MANIFEST IN ADVERTISING

Values are incorporated into advertising in every way possible, for the attachment of "value" to an inert product or intangible service is the very core of persuasion. It is the business of advertising to convert products into "goods," that is, objects with enhanced perceived values. This may result from a reminder of certain aspects of the product's performance and the benefits the consumer might derive, or the suggestion of new attributes or benefits or which the consumer was not previously aware. But whether a new idea, or the rehearsal of an old idea, advertising makes certain value dimensions salient and suggests the instrumentality of product consumption toward realization of those values. This is done in the art work, the written copy and through the interplay between them (see Pollay and Mainprize 1984). Either or both aspects of an ad can call attention to value aspects through direct display and assertion, or through association, metaphor, analogy and/or the role modeling of behavior.

However the value is displayed, the value is typically self evident as the obvious core benefit proposition. Advertising works hard to make obvious to all potential customers why a product should be perceived as a good, so the core value premise is made prominent. Where there are multiple value bases employed, as in detailed copy, the basic value dimensions of the ad can be found at the rhetorical frontier, that place where the argument stops. It is at this point, where the copy writer assumes that no further discussion is necessary, that defines the shared value between copywriter and reader/viewer, the sender and receiver of this communication. It-is at this rhetorical frontier that "goodness" goes without saying, and no further elaboration is required.

In deciding what values to utilize in an advertising strategy, the creative function may consider what values are most readily and credibly attached to the product in question -- what values are - most easily dramatized or pictorially displayed, what value premises the consumer most reliably responds to, what value premises competition is employing in its positioning, what values are important in distinguishing the target market from the larger culture, what values have the broadest base of appeals, etc. Since not even the tangible reality of the product much constrains the creative options, it is likely that even considering all of these question leaves many options available. Since the choice of what values are employed ultimately leads to the consequence of what values get displayed, modeled, and reinforced in the culture, it is potentially a very important task to delineate an Inventory of values that might be employed and to develop a methodology to reliably measure the values manifest in ads.

MEASUREMENT OF VALUE CONTENT OF ADS

While we might ultimately like to measure the effects Of communication, by being able to monitor what actually happens within the consumer value structure, we must begin by measuring aspects of the stimulus communication, the ad itself. Thus content analysis is called for and the task is to identify a universe of values that is exhaustive (comprehensive) and mutually exclusive (discriminating and unambiguous). The items in this universe then need to be defined, and rating procedures established so that trained research assistants acting as content coders can reliably identify values manifest in ads. Although the research literature on values in social sciences is still emergent and not fully developed, some work has been done measuring the content of written persuasion (White 1951) some of it specifically on advertising (Andren 1978), and another procedure well developed for self reported individual value hierarchies (Rokeach 1973). The latter is particularly valuable for it provides, in its profiling of the value character of the American population, a set of benchmarks against which advertising can be compared.

These suggestions for value dimensions from these sources were integrated into a master list, as no one source was adequate for the task. For example, the Rokeach lists of 18 terminal and 18 instrumental values were not readily applicable to analysis of advertising content because many of his scales were infrequently used and, conversely, some were used with such frequency that they needed decomposition for analytic precision. The elaborated list was amended through a series of pilot efforts which both collapsed and expanded various cells to attain a set of categories which was both rich and relevant to advertising. Other criteria employed in this method development were inclusivity (comprehensiveness), consistency with the literature about advertising social effects, analytical flexibility (the ability to apply the final form to other types of communication), mutual exclusivity so that each category was ideally a unique, orthogonal dimension. The end result was a universe of 42 concepts, organized into 21 pairs of related concepts, such as tradition vs. modernity, science vs. magic, independence vs. affiliation, etc. For the complete list, all definitions and a more complete discussion of the development of the measurement methodology see Pollay (1983).

The most important criteria for these content categories is that they be valid, measuring meaningful dimensions of advertising content, and reliable, measuring these dimensions in a manner which can be replicated by various researchers at various times. To enhance the reliability several steps were taken. The condensation of scales to reach the final set of 42, and their structuring into 21 dualities, reduced the number of dimensions a rater needed to hold in mind while assessing ads. Coding of the dominant appeals, as manifest in illustrations and headings, was distinguished and executed separately from coding of subordinate appeals in the fine print of the body copy. Forms and procedures were designed to clarify that some scales referred to product attributes while others described benefits that might be realized by the consumer.

Pretesting through several pilot studies on ads f rom various historical eras led to revised procedures and clarifications of the definitions and the addition of illustrative examples. These were particularly helpful in providing unambiguous evidence as to the location of the boundaries between concepts using the types of phrases commonly evidenced in advertising copy. When the time came to check the reliabilities of the revised procedures, new raters were assigned to the task and their "learning curve" observed so that appropriate levels of training and experience could be judged. This identified how much orientation and warmup experience raters needed before their performance stabilized into consistent patterns. These procedures led to an average interater reliability of 70% successful replication of coding. For some data, multiple coders were assigned and any disagreements between them was adjudicated by the author. This arbitrated process had a reliability performance of successful reproducing about 80% of ratings. These, compare very favorably with -the norms sugggested by statisticians and with the levels of reliabilities achieved by other content analysis processes.

VALUES IN AMERICAN ADVERTISING, 1900-1980

Sampling Strategy

To demonstrate the feasibility of measuring values in advertising, and to generate data descriptive of North American advertising throughout the twentieth century, an extensive sample of print ads was compiled. Print was chosen because (1) it spans the entire century, (2) it is accessible, having been retained by libraries, (3) it's static quality makes it more amenable to coding, recoding and generally facilitates information storage and retrieval, (4) it is readily photo-reproducible. Electronic media, such as radio and TV possess none of these advantages. The contemplation of electronic samples is a moot point, as no major publicly accessible historical collection exists in North America from which a representative sample might be drawn.

Our sample of 2,000 ads is constituted by 250 ads for each of the first eight decades of the twentieth century, 1900-1980. These were drawn from random pages of randomly selected issues of the best selling magazines within each decade, where magazines include those circulated as Sunday supplements to newspapers. Thus, for each decade the ten best circulated magazines were identified, then an issue per year was selected for each of the years 190 to 190 to create a mid-decade sample for the five-year span. Five ads were drawn from each issue, for five years, for the 10 magazines, yielding the 250 total per decade. For exact titles represented, see Pollay (1985). These magazines are characterized as mainstream, middle class and domestic in nature, as exampled by Life, Look, Saturday Evening Post, Readers' Digest. 14 y were also aimed primarily at the female domestic head of households as suggested by the inclusion of McCall=s, Ladie's Home Journal, Woman's Day.

The nature of this sample has been described elsewhere in terms of its level of information, the changing formats, tactics and strategy of the ads (Pollay 1984b, 1985). These results need not be reiterated here, except perhaps to note the counterintuitive finding of the subsiding sizzle." Conventional wisdom suggests that advertising might display an increasing evidence of a marketing orientation by greater and greater attention to benefits, rather than product attributes, or greater use of people, rather than focusing just on products, or greater use of rhetorical styles which are based on ethos or pathos rather than logos. But by none of these measures has print advertising shown a monotonic drift. Advertising reached a pinnacle of display of consumers, attention to benefits or use of emotive rhetorical styles in the 1950s or so. Since then the trend has reversed itself and less of these aspects are present than before, although the majority of ads are still consumer Focused, not just product glamorizing.

Basic Results

The total number of value bases evoked by the ads in our sample is about 4.6 on average and this has not changed much over the twentieth century, peaking at 5.4 values in the 1930s when copy was more crammed with multiple reasons to buy. But this apparent stability masks a very real change. When those values that are dominant, in the heading and illustration, are differentiated from those that are subordinate, in the body copy, it seems that there are more values being employed in the dominate mode and less in the subordinate. Perhaps in recognition of the fact that few readers actually digest the detail of the body copy) this text has become less likely to introduce new selling points, and is more likely to reinforce the main value basis of an ad. At the same time, copywriter and art directors have gotten more clever and creative in constructing visuals and captions which simultaneously communicate multiple values at a glance, in contrast to the earlier tendency to write ads with a single value thesis in the headings. This evocation of multiple values represents greater mastery of the arts of ambiguity, using poetic language, metaphor, double entendre, synonyms and other artistic and literary devices to simultaneously convey parallel meanings.

While this is of interest, and evidence of evolution of the art of advertising, our central concern is not how many but which values are utilized. Using the full set of scales developed in Pollay (1983), the frequencies of various values manifest can be seen in Exhibit 2. The most common themes dominant in ads are those of practicality, family, newness, cheap, healthy, sexy/vain, wisdom and uniqueness, in that order with all appearing in 10% or more of ads. it appears that the pattern of values used in the dominant role is different from that in the subordinate role. Some values, it would seem, work better as "hooks" to capture attention and involvement, whereas some other values might serve as a valuable reinforcing role in the text of the ad providing additional value reasons why the product Should be seen as a good. Exhibit 2 also provides data on the same dimensions for a sample of 250 television ads from the 1970s. There is a sigificant and very high rank order correlation (res=.89) between the magazine and TV data, suggesting that both media utilize values in much the same relative frequencies. Hence, our magazine data may be indicative of not just print, but perhaps all advertising.

If one examines the intercorrelations between the eight decades, one finds that all, are quite high. Even after collapsing cells into the eras identified by previous research (Pollay 1985) to ref lect periods of differential tactics and strategy, one still finds intercorrelations between eras of .85 or higher. This suggests that the hierarchy in the utilization of values is relatively stable over time and not as ephemeral as the superficial changes in style and format. While language and art styles may be trendy, reflecting fads and fancies of the day, the basic appeals endure.

Population Correlations

This also suggests that further analysis should not focus so much on the changes in the value character of the advertising, for it changes only slightly over time. Rather the more interesting analysis is how the value profile of advertising in total compares to the value profile of the population, or Various elements thereof. Most interestingly, we can compare the value profile with the directions and rate of change of values in the population, although here the available data is sparser. A correlation of the value hierarchy of the print ads and data from Rokeach (1973) for the population shows, somewhat surprisingly, no significant correlation, and what correlation there is, is negative. This seems to challenge the conventional wisdom that advertising is (merely) a reflection of societal values. it seems to be quite independent of cultural values, emphasizing a short list of values, and not apparently the short list most dearly held by the population. To investigate this latter point, and to respond to the observation that advertising can only employ certain values, as not all values are as readily commercialized or dramatized in ads, a truncated list of values was identified as those values most frequently employed in advertising. The correlation between this truncated list and the relative importance within this short list for the population was also found to be negative but not significant.

It can be argued that ads shouldn't be expected to mirror the values of the whole population, for the ads are written by people of a certain social class and are directed toward specific market segments. In our case, one might suppose that, if anything, the ads should show better correlation with the values for women, for middle class, for those in the prime purchasing age segment 18-35, or for whites because these are the predominant characteristics of the audiences for the mass magazines in this sample. Rokeach (1973) provides data for these segments in a highly detailed appendix, but no significant relationships were found. The values of advertising are apparently better correlated with those of a youthful audience, but the relationship is not significant. Advertising values are more consistent with the more affluent classes, but consistent with the more affluent classes, but again the relationship is not statistically significant. Neither sex nor race show even the suggestion of a relationship. In sum, whether studying the population as a whole or specific component segments, there seems to be no relationship between the overall value structure of ads and the population. Thus, the mirror hypothesis, that advertising is a passive reflector of societal values and plays no significant role in selectively emphasizing specific values, is not supported. Advertising seems to employ values quite independent of the importance they have in the larger cultural setting.

Social Change

Whether this selective employment of values by advertising is of consequence is a far more difficult research question. Despite the large volume of criticism, such as has been reviewed above, very little empirical work has been done to demonstrate the causal linkage between the character of advertising and the nature of social change. While some changes in society seem self evidentally linked to those behaviors and attitudes encouraged and modelled by advertising, the satisfactory demonstration of this relationship is a most difficult problem. This research issue is discussed elsewhere (Pollay 1986), pointing to the central issues of (1) the difficulty of isolating independent effects of advertising in longitudinal social changes or (2) the near impossibility of quasi-experiments because the pervasiveness of advertising leaves us with no untreated groups except those that are outliers along other major dimensions.

We can, however, begin this analysis by looking at the correlations between the value hierarchy of advertising and the changes in values that we can observe. There is not much hard data on value change, despite the popularity of the topic in verbal discussions, and after much data library network searching, only two sources were found, both measuring change from 1968 to the early or mid 1970s. Rokeach (1979b) provides data on terminal values for 1971 which is directly comparable to his earlier 1968 data. Calculating the change in average scores for the various values and comparing the ranking of this change data to the ranking of advertising value utilization yields a negative but insignificant correlation of -.387. This is very nearly significant at the 10% level and suggests at least that the value change for this period was opposite to that modelled by advertising. Data from Barnes (1979) exposes change from 1968-1974. Whether calculated from the full list of values or the truncated list of those commonly employed by advertising, a significant negative correlation is found for the value profile of advertising in general.

DISCUSSION

This result of observed social change for 1968-1970s occurring in a direction opposite to that modelled by advertising is at first glance very surprising and counter-intuitive. But to an historian this result is totally expected, for the period of the late 1960s and early 1970s was a period identified as a countercultural reversion.

Led by a hippie movement in the mid 1960s, but generalizing into revised attitudes toward governments, wars, ecological concerns, feminism and family roles, etc. this was a period of great social ferment in North America where many of the received cultural norms were being challenged and re-evaluated. Even in the midst of this era, it was obvious that consumer values, as are obviously vividly expressed in advertising, were central to the culture that was being reacted against. Thus, the conventional wisdom of that era is that we should expect to see cultural backlash and a movement away from the cultural norms -- not just those expressed in advertising but certainly including those norms.

While this data does not show dramatic statistically significant patterns, this non-result is itself of interest. Many have argued that advertising must be a cultural mirror reflecting social values. To do anything else, it is argued, is to create a communication that runs a high risk of being misunderstood, or culturally alien, or offensive. Why, then, do we not find some positive relationship, even on the short list of value frequently employed in advertising? The two possible answers to this are that the data is suspect (invalid), or that the relationship does not indeed exist (that the theory is wrong). We know that the measurement is at least a reliable one and that it seems to have validity when applied to individuals and decision contexts. There is no reason to suspect that an application of the concepts to the domain of advertising has any inherent reason to invalidate the concepts and their observation, although the judgments involved in content analysis might make the data noisy and the true relationships harder to detect.

Rokeach's data must be thought suspect for socially desirable self-reporting. His procedure and data might be biased such that socially directed values could be systematically ranked higher than personally directed values. This was tested by Kelly, Silverman and Cochrane (Rokeach 1973) who had their subjects complete the value survey twice, the second time with the request that the values be ranked in a .1 socially desirable" way. A high correlation between the first and second rankings would indicate that the subjects had provided socially desirable rankings in both instances, but no such relationship was observed.

On the other hand, how valid is the theoretical expectation that advertising will echo the value hierarchy of the culture? It would appear, not very. The argument is not very closely developed, even though it has been repeated with great frequency by those wishing to excuse advertising from any moral responsibilities resulting from social impact. The argument typically overlooks a number of key points, especially the following: (1) a culture is a mosaic of multiple and interrelated values and is best characterized by the relative importance of these; (2) not all values are as readily commercialized as some or more easily dramatized, visualized, linked to products in current production, and reliably responded to by the population in question. Thus, while it is no doubt true that advertising employs culturally derived and accepted values, it does so on a selective basis which has at least the potential for restructuring the value hierarchy. Because of the potential for cascade effects, this restructuring can produce dramatic shifts in cultural behaviors and can be far reaching in importance.

Obviously we are in need of better data on the changes in values for society. For North America we need data which covers a larger time span, and/or a time span not known to be an atypical countercultural reversion. Fortunately there will soon be available data for the 1980s so that change from 1968-1980s can be noted and related to the advertising data. It would be desirable, of course, to have data going back further in time. Since we have data for advertising from the turn of the century, it would be ideal to have parallel data on actual social value norms. Such data is however impossible, at least with the techniques now used. The value data is now derived by self reporting techniques and hence is always contemporary at the time of its collection. Perhaps an adaptation of the scales to a content analysis procedure might permit the identification of values from previous eras, but even if such a method were to be developed, like our advertising content analysis method, there would still be a very substantial problem of what would be sampled. It is not clear what documents should be analyzed as reflective of actual social values. Clearly, newspaper editorials or popular culture sources both have major issues of validity as indicators of the values for the average person, for they, just like advertising, may well reflect just a portion of the range of values.

Needed Cross Cultural Research

Cross cultural studies may prove to be the best approach, for the comparisons of different cultures may expose those ways in which they converge or diverge over time. Studies in other cultures also have a key advantage over studies in North America. The U.S. has had reasonably wide spread affluence for so long that the major social change of values may well have occurred some time ago in history and be difficult to isolate in contemporary studies. Other cultures, where affluence is more recent and still spreading, may prove to be far better laboratories for observing the processes and directions of social change.

The method outlined here can serve to identify the value character of the advertising in any cultural context and the value changes within those cultures can be more directly observed happening as it has, between generations still alive. Hopefully this methodological development will facilitate such research in other cultures so that we can all better understand the directions of change and hence the probable futures of consumer cultures.

EXHIBIT 1

HYPOTHESES ON THE VALUE CONSEQUENCES OF ADVERTISING

EXHIBIT 2

TOTAL ARGUMENT RELATIVE FREQUENCIES

EXHIBIT 3

ROKEACH'S (1973) VALUES

REFERENCES

Andren, Gunnar, et al. (1978) Rhetoric and Ideology in Advertising. Stockholm: Liberforlag.

Barnes, S. (1979) Political Action. Beverly Hills: Sage.

Darden, D.K. (1982) "Stratification and Mass Society: A Question of Values," Sociological Spectrum, 2, pp. 273-289.

Fullerton, Ronald A. and T.R. Nevett. (1985) "Advertising and Society: A Comparative Analysis of the Roots of Distrust in Germany and Great Britain," AMA Historical Workshop.

MacBride, Sean, et al. (1980) Many Voices, One World: Communications and Society Today and Tomorrow. New York' United Nations (UNESCO) Commissions Reports.

Pollay, R.W. (1983) "Measuring the Cultural Values Manifest in Advertising," Current Issues and Research in Advertising, pp. 71-92.

Pollay, R.W. (1984a) "The Identification and Distribution of Values Manifest in Print Advertising, 1900-1980," in A.G. Woodside and R. Pitts, eds., Personal Values and Consumer Behavior, Lexington Press, pp. 111-135.

Pollay, R.W. (1984b) "Twentieth Century Magazine Advertising: Determinants of Informativeness," Written Communication, Vol. I No. 1, pp. 56-77.

Pollay, R. W. (1985) "The Subsidizing Sizzle: Shifting Strategies in Twentieth Century Advertising," Journal of Marketing.

Pollay, R.W. (1986) "The Distorted Mirror: Reflections of the Unintended Consequences of Adveftising", Journal of Marketing, forthcoming.

Pollay, R.W. and S. Mainprize (1984). "Headlining of Visuals in Print Advertising" A Typology of Tactical Techniques," Proceedings, American Academy of Advertising, pp. 24-28.

Rescher, Nicholas. (1969) Introduction to Value Theory. Englewood Cliffs: Prentice-Hall.

Rokeach, Milton. (1973) The Nature of Human Values. New York: Free Press, p.438.

Rokeach, Milton. (1979a) "Long-term value change initiated by computer feed-back," in Understanding Human Values, M. Rokeach, ed. New York: Free Press, pp. 210-225.

Rokeach, Milton. (1979b) "Change and Stability in Value Systems, 1968-1971," in Understanding Human Values, M. Rokeach, ed. New York: Free Press, pp. 241-156.

Rokeach, M. and J.W. Grube. (1979) "Can Values Be Manipulated Arbitrarily?". in Understanding Human Values, M. Rokeach, ed. New York: Free Press, pp. 241-156.

Rorty, James. (1934) Our Master's Voice: Advertising. New York: John Day.

Sanders, K.R. and L.E. Atwood. (1979) "Value Change Initiated by the Mass Media, in Understanding Human Values, M. Rokeach, ed. New York: Free Press, pp. 226-240.

White, Ralph K. (1951) Value Analysis: Nature and Use of the Method, Ann Arbor: Society for the Psychological Study of Social Issues.

Williams, R.M. (1979) "Change and Stability in Values and Value Systems: A Sociological Perspective," in Understanding Human Values, M. Rokeach, ed. New York: Free Press.

White, Ralph K. (1951) Value Analysis: Nature and Use of the Method, Ann Arbor: Society for the Psychological Study of Social Issues.

Williams, R.M. (1979) "Change and Stability in Values and Value Systems: A Sociological Perspective," in Understanding Human Values, M. Rokeach, ed. New York: Free Press.