Public Opinion and Public Opinion Polling[1]

Herbert Blumer
University of Chicago 

This paper presents some observations on public opinion and on public opinion polling as currently performed. It is hoped that these observations will provoke the discussion for which, I understand, this meeting has been arranged. The observations are not along the line of what seems to be the chief preoccupation of students of public opinion polling, to wit, the internal improvement of their technique. Instead, the observations are designed to invite attention to whether public opinion polling actually deals with public opinion.

The first observations which I wish to make are in the nature of a prelude. They come from a mere logical scrutiny of public opinion palling as an alleged form of scientific investigation. What I note is the inability of public opinion polling to isolate "public opinion" as an abstract or generic concept which could thereby become the focal point for the formation of a system of propositions. It would seem needless to point out that in an avowed scientific enterprise seeking to study a class of empirical items and to develop a series of generalizations about that class it is necessary to identify the class. Such identification enables discrimination between the instances which fall within the class and those which do not. In this manner, the generic character of the object of study becomes delineated. When the generic object of study is distinguishable, it becomes possible to focus study on that object and thus to learn progressively more about that object. In this way the ground is prepared for cumulative generalizations or propositions relative to the generic object of investigation.

As far as I can judge, the current study of public opinion by polling ignores the simple logical point which has just been made. This can be seen through three observations. First, there is no effort, seemingly, to try to identify or to isolate public opinion as an abject; we are not given any criteria which characterize or distinguish public opinion and thus we are not able to say that a given empirical instance falls within the class of public opinion and some other empirical instance falls outside of the class of public opinion. Second, there is an absence, as far as I can determine, of using specific studies to test a general proposition about public opinion; this suggests that the students are not studying a generic object. This suggestion is supported by the third observation—a paucity, if not a complete absence, of generalizations about public opinion despite the voluminous amount of polling studies of public opinion. It must be concluded, in my judgment, that current public opinion polling has not succeeded in isolating public opinion as a generic object of study.

It may be argued that the isolation of a generic object, especially in the realm of human behavior, is a goal rather than an initial point of departure—and that consequently the present inability to identify public opinion as a generic object is not damning to current public opinion polling. This should be admitted. However, what impresses me is the apparent absence of effort or sincere interest on the part of students of public opinion polling to move in the direction of identifying the object which they are supposedly seeking to study, to record, and to measure. I believe it is fair to say that those trying to study public opinion by polling are so wedded to their technique and so preoccupied with the improvement of their technique that they shunt aside the vital question of whether their technique is suited to the study of what they are ostensibly seeking to study. Their work is largely merely making application of their technique. They are not concerned with independent analysis of the nature of public opinion in order to judge whether the ap-

( 543) -plication of their technique fits that nature.

A few words are in order here on an approach that consciously excuses itself from any consideration of such a problem. I refer to the narrow operationalist position that public opinion consists of what public opinion polls poll. Here, curiously, the findings resulting from an operation, or use of an instrument, are regarded as constituting the object of study instead of being some contributory addition to knowledge of the object of study, The operation ceases to be a guided procedure on behalf of an object of inquiry; instead the operation determines intrinsically its own objective. I do not care to consider here the profound logical and psychological difficulties that attend the effort to develop systematic knowledge through a procedure which is not a form of directed inquiry. All that I wish to note is that the results of narrow operationalism, as above specified, merely leave or raise the question of what the results mean. Not having a conceptual point of reference the results are merely disparate findings. It is logically possible, of course, to use such findings to develop a conceptualization. I fail to see anything being done in this direction by those who subscribe to the narrow operationalist position in the use of public opinion polls. What is logically unpardonable on the part of those who take the narrow operationalist position is for them to hold either wittingly or unwittingly that their investigations are a study of public opinion as this term is conceived in our ordinary discourse. Having rejected as unnecessary the task of characterizing the object of inquiry for the purpose of seeing whether the enquiry is suited to the object of inquiry, it is gratuitous and unwarranted to presume that after all the inquiry is a study of the object which one refuses to characterize, Such a form of trying to eat one's cake and have it too needs no further comment.

The foregoing series of logical observations has been made merely to stress the absence of consideration of a generic object by those engaged in public opinion polling. Apparently, it is by virtue of this absence of consideration that they are obtuse to the functional nature of public opinion in our societyand to questions of whether their technique is suited to this functional nature. In this paper I intend to judge the suitability of public opinion polling as a means of studying public opinion. This shall be done from the standpoint of what we know of public opinion in our society.

Admittedly, we do not know a great deal about public opinion. However, we know something. We know enough about public opinion from empirical observations to form a few reasonably reliable judgments about its nature and mode of functioning. In addition, we can make some reasonably secure inferences about the structure and functioning of our society and about collective behavior within our society. This combined body of knowledge derived partly from direct empirical observation and partly from reasonable inference can serve appropriately as means of judging and assessing current public opinion polling as a device for studying public opinion.

Indeed, the features that I wish to note about public opinion and its setting are so obvious and commonplace that I almost blush to call them to the attention of this audience. I would not do so were it not painfully clear that the students of current public opinion polling ignore them either wittingly or unwittingly in their whole research procedure, I shall indicate by number the features to be noted.

1.) Public opinion must obviously be recognized as having its setting in a society and as being a function of that society in operation. This means, patently, that public opinion gets is form from the social framework in which it moves, and from the social processes in play in that framework; also that the function and role of public opinion is determined by the part it plays in the operation of the society. If public opinion is to be studied in any realistic sense its depiction must be faithful to its empirical character. I do not wish to be redundant but I find it necessary to say that the empirical character of public opinion is represented by its composition and manner of functioning as a part of a society in operation.

2.) As every sociologist ought to know

( 544) and as every intelligent layman does know, a society has an organization. It is not a mere aggregation of disparate individuals. A human society is composed of diverse kinds of functional groups. In our American society illustrative instances of functional groups are a corporation, a trade association, a labor union, an ethnic group, a farmers' organization. To a major extent our total collective life is made up of the actions and acts of such groups. These groups are oriented in different directions because of special interests. These groups differ in terms of their strategic position in the society and in terms of opportunities to act. Accordingly, they differ in terms of prestige and power. As functional groups, that is to say as groups acting individually in some corporate or unitary sense, such groups necessarily have to have some organization—some leadership, some policy makers, some individuals who speak on behalf of the group, and some individuals who take the initiative in acting on behalf of the group.

3.) Such functional groups, when they act, have to act through the channels which are available in the society. If the fate of the proposed acts depends on the decisions of individuals or groups who are located at strategic points in the channels of action, then influence and pressure is brought to bear directly or indirectly on such individuals or groups who make the decisions. I take it that this realistic feature of the operation of our American society requires little explication. If an action embodying the interests of a functional group such as a farmers' organization depends for its realization on decisions of Congressmen or a bureau or a set of administrators, then efforts on behalf of that action will seek to influence such Congressmen, bureau, or administrators. Since in every society to some degree, and in our American society to a large degree, there are individuals, committees, hoards, legislators, administrators, and executives who have to make the decisions affecting the outcome of the actions of functional groups, such key people become the object of direct and indirect influence or pressure.

4. The key individuals referred to who have to make the crucial decisions are almost inevitably confronted with the necessity of assessing the various influences, claims, demands, urgings, and pressures that are brought to bear on them. Insofar as they are responsive and responsible they are bound to make such an assessment in the process of arriving at their decisions. Here I want to make the trite remark that in making their assessments these key individuals take into account what they judge to be worthy of being taken into account.

5.) The above points give a crude but essentially realistic picture of certain important ways in which our society operates. The fifth feature I wish to note is that public opinion is formed and expressed in large measure through these ways of societal operation. This point requires a little elaboration. The formation of public opinion occurs as a function of a society in operation. I state the matter in that way to stress that the formation of public opinion does not occur through an interaction of disparate individuals who share equally in the process. Instead the formation of public opinion reflects the functional composition and organization of society. The formation of public opinion occurs in large measure through the interaction of groups. I mean nothing esoteric by this last remark. I merely refer to the common occurrence of the leaders or officials of a functional group taking a stand on behalf of the group with reference to an issue and voicing explicitly or implicitly this stand on behalf of the group. Much of the interaction through which public opinion is formed is through the clash of these group views and positions. In no sense does such a group view imply that it is held in equal manner and in equal degree by all of the members of the group. Many of the members of the group may subscribe to the view without understanding it, many may be indifferent about it, many may share the view only in part, and many may actually not share the view but still not rebel against the representatives of the group who express the view. Nevertheless the view, as indicated, may be introduced into the forum of discussion as the view of the group and may be reacted to as such. To bring out this point in

( 545) another way, one need merely note that in the more outstanding. expressions of view on an issue, the individuals almost always speak either explicitly or implicitly as representatives of groups. I would repeat that in any realistic sense the diversified interaction which gives rise to public opinion is in large measure between functional groups and not merely between disparate individuals.

I think that it is also very clear that in the process of forming public opinion, individuals are not alike in influence nor are groups that are equal numerically in membership alike in influence. This is so evident as not to require elaboration. It is enough merely to point out that differences in prestige, position; and influence that characterize groups and individuals in the functional organizations of a society are brought into play in the formation of public opinion.

The picture of a series of groups and individuals of significantly different influence interacting in the formation of public opinion holds true equally well with reference to the expression of public opinion. By expression of public opinion I mean bringing the public opinion to bear on those who have to act in response to public opinion, This expression is not in the form of a parade or array of the views of disparate individuals, in an open forum. Where the views are voiced in open forum they are likely, as has been indicated, to be in one way or another the expression of group views. But in addition to the voicing of views in the open forum, the expression of public opinion is in the form of direct influence on those who are to act in response to public opinion. Through such means as letters, telegrams, petitions, resolutions, lobbies, delegations, and personal meetings interested groups and individuals bring their views and positions to bear on the key persons who have to make the decisions. I am not concerned with whether such forms of expressing public opinion should occur; I merely wish to emphasize that in any realistic consideration of public opinion it must be recognized that such means of expressing public opinion do occur. A society which has to act will use the channels of action that it has in its structure.

6.) The last feature of public opinion that I wish to note is that in any realistic sense public opinion consists of the pattern of the diverse views and positions on the issue that come to the individuals who have to act in response to the public opinion. Public opinion which was a mere display, or which was terminal in its very expression, or which never came to the attention of those who have to act on public opinion would be impotent and meaningless as far as affecting the action or operation of society is concerned. Insofar as public opinion is effective on societal action it becomes so only by entering into the purview of whoever, like legislators, executives, administrators, and policy makers, have to act on public opinion. To me this proposition is self-evident. If it be granted, the character of public opinion in terms of meaningful operation must be sought in the array of views and positions which enter into the consideration of those who have to take action on public opinion.

It is important to note that the individual who has to act on public opinion has to assess the public opinion as it comes to his attention, because of the very fact that this public opinion comes to him in the form of diverse views and usually opposed views. Insofar as he is responsive to public opinion he has to weigh the respective views. How this assessment is made is an obscure matter. But one generalization even though trite, can be made safely, to wit, that the individual takes into account different views only to the extent to which such views count. And views count pretty much on the basis of how the individual judges the "backing" of the views and the implication of the backing. It is in this sense, again, that the organization of the society with its differentiation of prestige and power, enters into the character of public opinion. As was explained above, the key person who has to act on public opinion is usually subject to a variety of presentations, importunities, demands, criticisms, and suggestions that come to him through the various channels in the communicative structure of society. Unless one wishes to conjure in his imagination a very fanciful society he must admit that the servant of public opinion is forced to make an assessment of the ex-

( 546) -pressions of public opinion that come to his attention and that in this assessment consideration is given to expressions only to the extent to which they are judged to "count."

The foregoing six features are, I believe, trite but faithful points about public opinion as it functions in our society. They may serve as a background for the examination of public opinion polling. I may state here that in this discussion I am not concerning myself with the problem of whether the individual opinions one gets through the polling interview are reasonably valid. My discussion, instead, is concerned with the question of the value of poll findings even if one makes the dubious assumption that the individual opinions that are secured are valid.

In my judgment the inherent deficiency of public opinion polling, certainly as currently done, is contained in its sampling procedure. Its current sampling procedure forces a treatment of society as if society were only an aggregation of disparate individuals. Public opinion, in turn, is regarded as being a quantitative distribution of individual opinions. This way of treating society and this way of viewing public opinion must be regarded as markedly unrealistic. The best way I can bring this out is by making continuous reference to the common sense empirical observations of public opinion that were noted previously. We do not know at all whether individuals in the sample represent that portion of structured society that is participating in the formation of public opinion on a given issue. That the sample will catch a number of them, or even a larger number of them, is very likely. But, as far as I am able to determine, there is no way in current public opinion polling to know much about this. Certainly the mere fact that the interviewee either gives or does not give an opinion does not tell you whether he is participating in the formation of public opinion as it is being built up functionally in the society. More important, assuming that the sample catches the individuals who are participating in the formation of the given public opinion, no information is given of their part in this process. One cannot identify from the sample or from the replies of those constituting the sample the social nitch of the individual in that portion of the social structure in which the public opinion is being formed. Such information is not given in the conventional items of age, sex, occupation, economic status, educational attainment or class status. These are rarely the marks of significant functional position in the formation of public opinion on a given issue. We do not know from the conventional kind of sample or from the responses of the interviewee what influence, if any, he has in the formation or expression of public opinion. We do not know whether he has a following or whether he doesn't. We do not know whether or not he is speaking on behalf of a group or groups or whether he even belongs to functional groups interested in the issue. If he does, perchance, express the views of some such functional group, we don't know whether or not that group is busily at work in the channels of society to give vigorous expression to their point of view. We do not even know whether he, as an individual, is translating his opinion into what I have termed previously "effective public opinion."

In short, we know essentially nothing of the individual in the sample with reference to the significance of him or of his opinion in the public opinion that is being built up or which is expressing itself functionally in the operation of society. We do not know whether the individual has the position of an archbishop or an itinerant laborer; whether he belongs to a powerful group taking a vigorous stand on the issue or whether he is a detached recluse with no membership in a functional group; whether he is bringing his opinion to bear in some fashion at strategic points in the operation of society or whether it is isolated and socially impotent. We do not know what role, if any, any individual in the sample plays in the formation of the public opinion on which he is questioned, and we do not know what part, if any, his opinion as given has in the functional public opinion which exists with reference to the issue.

What has just been said with reference to the individual component of the public opinion poll applies collectively to the total findings. The collective findings have no assur-

( 547) -ance of depicting public opinion on a given issue because these findings ignore the framework and the functional operation of the public opinion. If this is not clear from what has already been said, I would like to point out the enormous difficulty that occurs when one seeks to assess the findings of a public opinion poll in terms of the organization of society with which an administrator, legislator, executive, or similarly placed person has to contend. As I have stated earlier such an individual who is presumably responsive to public opinion has to assess public opinion as it comes to his attention in terms of the functional organization of society to which he is responsive. He has to view that society in terms of groups of divergent influence; in terms of organizations with different degrees of power; in terms of individuals with followings; in terms of indifferent people—all, in other words, in terms of what and who counts in his part of the social world. This type of assessment which is called for in the instance of an organized society in operation is well-nigh impossible to make in the case of the findings of public opinion polls. We are unable to answer such questions as the following: how much power and influence is possessed by those who have the favorable opinion or the unfavorable opinion; who are these people who have the opinion; whom do they represent; how well organized are they; what groups do they belong to that are stirring around on the scene and that are likely to continue to do so; are those people who have the given opinion very much concerned about their opinion; are they going to get busy and do something about it; are they going to get vociferous, militant, and troublesome; are they in the position to influence powerful groups and individuals who are known; does the opinion represent a studied policy of significant organizations which will persist and who are likely to remember; is the opinion an ephemeral or momentary view which people will quickly forget? These sample questions show how markedly difficult it is to assess the results of public opinion polling from the standpoint of the things that have to be taken into account in working in an organized society. This difficulty, in turn, signifies that current public opinion polling gives an inaccurate and unrealistic picture of public opinion because of the failure to catch opinions as they are organized and as they operate in a functioning society.

What I have said will appear to many as distinctly invalid on the ground that public opinion polling has demonstrated that it can and does detect public opinion faithfully, by virtue of its marked success in predicting election returns. This contention needs to be investigated carefully, particularly since in most circles polling, wherever applied, is regarded as intrinsically valid because of its rather spectacular success in predicting elections. What I think needs to be noted is that the casting of ballots is distinctly an action of separate individuals wherein a ballot cast by one individual has exactly the same weight as a ballot cast by another individual. In this proper sense, and in the sense of real action, voters constitute a population of disparate individuals, each of whom has equal weight to the others. Consequently, the sampling procedure which is based on a population of disparate individuals is eminently suited to securing a picture of what the voting is likely to be. However, to regard the successful use of palling in this area as proof of its automatic validity when applied to an area where people do not act as equally weighted disparate individuals begs the very question under consideration. I would repeat that the formation and expression of public opinion giving rise to effective public opinion is not an action of a population of disparate individuals having equal weight but is a function of a structured society, differentiated into a network of different kinds of groups and individuals having differential weight and influence and occupying different strategic positions. Accordingly, to my mind, the success attending polling in the prediction of elections gives no validity to the method as a means of studying, recording or measuring public opinion as it forms and functions in our society.

There is a very important contention in this connection which has to be considered. The contention can be stated as follows:

An election by public ballot is in itself an

( 548) expression of public opinion—and, furthermore, it is effective and decisive expression of public opinion. It is, in fact, the ultimate expression of public opinion and thus it represents the proper norm of the expression of public opinion. In the election by ballot each voter, in accordance with the basic principles of democracy, has his say as a citizen and has equal worth. to every other citizen in casting his ballot. If election by ballot be recognized as the genuine referendum in which true public opinion comes to expression, then the preeminence of current public opinion polling as the device for recording and measuring public opinion is established. For, public opinion polling with its current form of sampling has demonstrated that it can predict reliably and effectively the results of the election. Accordingly, public opinion polling, in itself, can be used as a type of referendum to record and measure the true opinion of the public on issues in the instances of which the public does not go to the election polls. Thus, public opinion poling yields a more reliable and accurate picture of public opinion than is represented by the confused, indefinite, slanted, and favor—ridden expressions of opinion that come ordinarily to the legislator, administrator, or executive who has to act on public opinion. The public opinion poll tells us where people stand. It gives us the vox populi.

My remarks with reference to this contention will be brief. It should be evident on analysis that the contention is actually a normative plea and not a defense of polling as a method of study of public opinion as such public opinion functions in our society. The contention proposes that public opinion be construed in a particular way, to wit, that public opinion ought to be an aggregation of the opinions of a cross section of the population rather than what it is in the actual functioning of society. To my mind it is highly questionable whether in the day by day operation of our society public opinion ought to be of the nature posited by the public opinion poll. Many appropriate questions could be raised about how and to what extent public opinion is expressed at the election polls, and, more important, whether it would be possible or even advisable for public opinion, in the form of an aggregation of equally weighted individual opinions, to function meaningfully in a society with a diversified organization. However, such questions need not be raised here. It is sufficient to note that if one seeks to justify polling as a method of studying public opinion on the ground that the composition of public opinion ought to be different than what it is, he is not establishing the validity of the method for the study of the empirical world as it is. Instead, he is hanging on the coat-tails of a dubious proposal for social reform.[2]

In this paper I have presented criticisms of "public opinion polling" as a method for the recording and measurement of public opinion, These criticisms have centered around the distortion that stems from the use of a sample in the form of an aggregation of disparate individuals having equal weight. These criticisms should not be misinterpreted to mean that such a sampling procedure is invalid wherever applied or that wherever polling makes use of such a sampling procedure such polling is intrinsically invalid. Clearly, the criticism applies when such a sampling procedure is used to study a matter whose composition is an organization of interacting parts instead of being merely an aggregation of individuals. Where the matter which one is studying is an aggregation of individual units then the application of the sampling procedure spoken of is clearly in order. I make this banal statement only to call attention to the fact that there are obviously many matters about human beings and their conduct that have just this character of being an aggregation of individuals or a congeries of individual actions. Many demographic matters are of this nature. Also, many actions of human beings in a society are of this nature —such as casting ballots, purchasing tooth

( 549) paste, going to motion picture shows, and reading newspapers. Such actions, which I like to think of as mass actions of individuals in contrast to organized actions of groups, lend themselves readily to the type of sampling that we have in current public opinion polling. In fact, it is the existence of such mass actions of individuals which explains, in my judgment, the successful use in consumer research of sampling such as is employed in public opinion polling. What I find questionable, and what this paper criticizes, is the use of such sampling with its implicit imagery and logic in the study of a matter which, like the process of public opinion, functions as a moving organization of interconnected parts.

The last item I wish to consider briefly refers to the interesting and seemingly baffling question of how one should or can sample an object matter which is a complicated system of interacting parts, having differential influence in the total operation. Perhaps the question in itself is absurd. At various times I have asked different experts in sampling how one would sample an organic structure. With a single exception these individuals looked at me askance as if the question were idiotic. But the problem, I think, remains even though I find it difficult to state. In human society, particularly in modern society, we are confronted with intricate complexes of moving relations which are roughly recognizable as systems, even though loose systems. Such a loose system is too complicated, too encumbered in detail and too fast moving to be described in any one of its given "cycles" of operation adequately and faithfully. Yet unless we merely want to speculate about it we have to dip into it in some manner in order to understand what is happening in the given cycle of operation in which we are interested. Thus, using the public opinion process in our society as an illustration we are able to make a rough characterization as to how it functions in the case, let us say, of a national issue. However, if we want to know how it functions in the case of a given national issue, we are at a loss to make an adequate description because of the complexity and quick movement of the cycle of its operation. So, to know what is going on, particularly to know what is likely to goon in the latter stages, we have to dip in here and there. The problems of where to dip in, how to dip in, and how far to dip in are what I have in mind in speaking of sampling an organic structure.

I suppose, as one of my friends has pointed out, that the answer to the problem requires the formulation of a model. We have no such model in the instance of public opinion as it operates in our society. My own hunch is that such a model should be constructed, if it can be at all, by working backwards instead of by working forward. That is, we ought to begin with those who have to act on public opinion and move backwards along the lines of the various expressions of public opinion that come to their attention, tracing these expressions backward through their own various channels and in doing so, noting the chief channels, the key points of importance, and the way in which any given expression has come to develop and pick up an organized backing out of what initially must have been a relatively amorphous condition. Perhaps, such a model, if it could be worked out, would allow the development of a realistic method of sampling in place of what seems to me to be the highly artificial method of sampling used in current public opinion polling.


  1.  Paper read before the annual meeting of the American Sociological Society held in New York City, December 28-30, 1947.
  2. 2 I refer to such a program as dubious because I believe the much needed improvement of public opinion in our society should be in the process by which public opinion organically functions, i.e., by arousing, organizing, and effectively directing the opinion of people who appreciate that they have an interest in a given issue. A reliance, instead, on a mere "referendum" by an undifferentiated mass, having great segments of indifference and non-participation, is unlikely to offer a desirable public opinion. At the best, in my judgment, such a "referendum" could operate as a corrective supplement and not as a substitute. The important question concerning the directions in which public opinion might secure its much needed improvement is, of course, outside of the scope of this paper.

Valid HTML 4.01 Strict Valid CSS2