The Newspaper and Public Opinion

Author(s): George A. Lundberg

Source: Social Forces , Jun., 1926, Vol. 4, No. 4 (Jun., 1926), pp. 709-715
Published by: Oxford University Press

Stable URL: https://www.jstor.org/stable/3004449

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Oxford University Press is collaborating with JSTOR to digitize, preserve and extend access to
Social Forces

This content downloaded from

177.106.167.81 on Fri, 08 Mar 2024 00:02:20 +00:00
All use subject to https://about.jstor.org/terms
 THE NEWSPAPER AND PUBLIC OPINION 709

THE NEWSPAPER AND PUBLIC OPINION

GEORGE A. LUNDBERG

One of the conclusions very generally

I
agreed on by writers in the field is the
U P TO the present time most of the conclusion that the most powerful, if
conclusions of sociology have not the all important factor in influencing
been derived from more or less public opinion, and consequently the
cursory observations by men interested votes of a community, is the newspaper.
in social problems. Unfortunately, ob- A survey of opinions on this subject
servations of this kind and the generaliza- (they are stated not as opinions but as
tions drawn from them too frequently established facts) yields a great number of
reflect the desires or fears of the observer emphatic and unqualified statements to
rather than the facts as they are. That is, this effect. For example:
the conclusions of sociology are frequently . .. . They [the newspapers] create great men
statements of what someone thinks ought out of next to nothing and destroy the reputation of
to be, or ought not to be, rather than a men truly fit for leadership. They decide questions
of war and peace. They carry elections. They overawe
statement of what is. The difficulty with
and coerce politicians, rulers, and courts. When
much current sociological literature is
they are virtually unanimous, nothing can withstand
that many hypotheses and theories by them.2 (Italics mine.)
virtue of long standing and repetition by . . . . the great body of readers take the paper
eminent men have become accepted as seriously and its influence goes a long way in making
public opinion for them.3
statements of fact. The fact that Aris-
. . . . Since the newspapers exercise so vast a
totle made some statement and that the power they should be held to corresponding responsi-
subsequent respectable writers on the bility. Since they are the public par excellence, they
subject have accepted or repeated this should be subject to a public regulation adequate

statement is frequently regarded as con- to secure the public welfare . E.. . Every daily
paper that enjoys more than a specified circulation
clusive evidence of the validity of the
should by constitutional amendment be required
statement. In the absence of more scien- to place a certain amount of space in every issue at the
tific conclusions, these statements of disposal of each of the four parties-whenever there
opinion are, to be sure, of great sig- are four-that cast the highest vote at the last State
election.4
nificance and a new science will necessarily
The American press has more influence than it has
have to rely on a great many such formula-
tions, at least as working hypotheses. 1 Some recent writers, such as Robert E. Park,
But too frequently these hypotheses are have qualified this assumption by emphasizing the
regarded as established facts. The result tendency of the press to reflect, rather than to make,
public opinion. See also "The Waning Power of the
is that many proposed investigations are
Press," by F. E. Leupp, Atlantic Monthly, February,
discouraged as attempts to prove the
I9I0.

obvious. It is only comparatively re- 2 Yarros, V. S., "The Press and Public Opinion,"
cently that the need for extensive and A. J. S., November, I899, p. 32..
painstaking scientific inquiry into the 3 Groves, E. R., Social Problems and Education,
p. 380.
validity of the generally accepted con-
Elsewhere (p. 377). Groves admits that the press
clusions of sociology has been emphasized.
"more often echoes public opinion."
So far, the science consists chiefly of con- 4 Hayes, E. C., "The Formation of Public Opin-
clusions and very little scientific data. ion," Jr. of Applied Soc., September-October, I92.5.

This content downloaded from

177.106.167.81 on Fri, 08 Mar 2024 00:02:20 +00:00
All use subject to https://about.jstor.org/terms
 7IO SOCIAL FORCES

ever had in any other time in any other country. men and 350 women were interviewed
No other extra-judicial force, except religion, is half
and schedules filled out with their an-
so powerful.,
swers. For the purpose of this inquiry,
In nearly every case of such pronounce- four public questions which had been
ment, the author makes no reference to prominently before the electorate within
the data, if any, on which he bases his the last eight months preceding the
statements. There is good reason for investigation were selected and each
this. So far as the present writer has person asked his position on these ques-
been able to discover there are no scientific
tions. In a different connection the per-
data on the subject.6 The most that can son was asked what newspaper he read
be found is a considerable number of most frequently. The results of this
opinions, many of them eloquently ex- part of the inquiry are shown in Table I,
pressed. But brilliant style and great which indicates the degree of relationship,
learning will not permanently serve as a or lack of relationship, between the read-
substitute for data. ers' attitude on a public question and the
It is the purpose of the present paper to attitude of the newspaper which each
approach quantitatively, one aspect of group read most frequently.
this problem, namely, the question of the Table I shows no significant relationship
direct influence of newspapers on public between the attitude of the newspaper
opinion on certain specific public issues. and the attitude of the reader on the
The aim is to discover to what extent a questions investigated. On the question
person's attitude on public questions, on of the city manager plan, which had been
which the press is assumed to be so influ- rejected at the polls about four months
ential, correlates with the attitude on before this investigation was begun,
these questions of the newspaper to whichThge Times took a vigorous position against
he is most frequently exposed. For this the measure. Fifty-two per cent of its
purpose 940 of the residents of the city of
readers agreed with their newspaper in
Seattle were selected at random.7 590 this position. The Post-Intelligencer took
5 Irvin, W., "The American Newspaper," Collier's, a favorable stand on the question, and 50
January 2.i, I9II, p. i8. per cent of its readers agreed in this posi-
6 "Evidently the newspaper is an institution that tion. So far there is a slight indication
is not yet fully understood. What it is, or seems of correlation between the attitude of the
to be, for any one of us at any time is determined by
paper read most frequently and the atti-
our differing points of view. As a matter of fact we
do not know much about the newspaper. It has
tude of the reader. When we consider
never been studied." Robert E. Park "The Natural The Star, however, this correlation be-
History of the Newspaper," A. J. S., November, comes negative, 55 per cent of the follow-
I92X3. ers of this paper being opposed to the city
7The sample was compared with the general
adult white population of the city (according to the
census of i920) with respect to nativity, citizenship, had only I4 per cent of the total circulation in the
and occupations, and was found not to vary from thesample, and in the case of the Union Record which has
general population in these characteristics, by more iS per cent of the total daily circulation in the city
than I4 per cent in any one respect. The proportionbut only 3 per cent in the sample. While there is
of the sample reading each paper (see Table I) also reason to believe that this sample is for the present
corresponded closely to the proportion of the totalpurpose representative of the great bulk of the popu-
circulation of each paper. This was true except in lation of the city, there is no need to insist on this
the case of The Star, which has 30 per cent of the total point as the results are sufficiently significant as
daily circulation of the papers in the city, but which applied only to the class here investigated.

This content downloaded from

177.106.167.81 on Fri, 08 Mar 2024 00:02:20 +00:00
All use subject to https://about.jstor.org/terms
 THE NEWSPAPER AND PUBLIC OPINION 7II

manager plan, although the paper en- the common factor which accounts for
dorsed the plan. In the case of The both the attitude of the paper and of
Union Record the most positive relationship its readers. Among those who follow
is shown, 72. per cent of its readers agree- out of town papers we find, as might be
ing with the newspaper's attitude. Only expected, a greater degree of indifference

TABLE I

RELATIONSHIP BETWEEN THE ATTITUDE OP THE NEWSPAPER AND THIE ATTITUDE OP THE, READER ON
QUESTIONS (Seattle)

PUBLIC QUESTIONS

City manager plan Rainicr Valley Skagit Pro ject Political preference at last
PAPER READ MOST FREQUENTLY AND PER CENT Lin presidential election
OF WHOLE SAMPLE READING EACH PAPER I

u u~~.
z4 zZ >4 z z >U4 z z u .s 0 Z
p,r p,r p,r p,r p,,r pr p,r per p.r p.r per p.r p,r p'r
cmt cent cm crnt rcent cmt cent cmt cmt c:nt cant c: cmnt Mt

Against Against Against Coolidge

Times (39 per cent) ................... 4

For Friendly Friendly No stand

Post-Intelligencer(33 percent).......... 51 40
For No stand For La Follette

Star(I44percent) .................... 35
Against Against For La Follette

Union Record (3 per cent) . ............. 20 72. 9 2.1 72. 7 86 7 7 0 O 27 34 I7 I3

Out-of-town paper (14 per cent) ....... 1 41 | 43 | i6 1 2.3 | 6i | I5 6i 2.5 1 14 15 | 51 | 4 0 O 30

Notes on Table I

I. The city manager plan had been rejected at the polls four months before this investigation was made,
55 per cent of those voting being opposed to the plan as propbsed. The Rainier Valley Line and the Skagit
Project were definitely proposed municipal ownership and development projects before the city council at the
time of the investigation.
2. A statement of the attitude of each newspaper on each of the public questions involved was secured
from the editor of each paper except The Times. The attitude of the latter was determined from the paper
itsclf as it attacked vigorously all three of the proposed city measures.
3. The vote for president in Seattle at the last election was as follows: Coolidge, 55 per cent; La Follette,
38 per cent; Davis, 7 per cent.
4. The table was worked out separately for men and women but no significant differences in the two groups
were cvident.

3 per cent of the whole sample were to the question, but an approximately
readers of the Union Record, however, even division on the question among
and as it is the organ of organized labor, those who expressed themselves.
there is reason to believe that this third On the question of the purchase of the
factor, namely the union organization, is Rainier Valley Street Railway line by the

This content downloaded from

177.106.167.81 on Fri, 08 Mar 2024 00:02:20 +00:00
All use subject to https://about.jstor.org/terms
 712. SOCIAL FORCES

city, the results The are Timesmore

supported Coolidge,
difficultand 66 to
analyze due to the per lack of
cent of its a indicated
readers positive the same stand
on the questions by two of the papers. attitude. Fifty-three per cent of the
But the results reflect the same situation readers of the Post-Intelligencer, which
as found on the previous question, namely,took no stand on candidates, also indicated
that the correlation between the attitude a preference for the Republican ticket.
of the newspaper and that of the readers is At the same time the proportion of
too small to be regarded as significant. readers of the latter paper who indicated
We do find, it is true, a slightly greater their preference for Davis and La Follette
percentage opposed among the readers were practically the same as in the case of
of the two papers which were opposed the Times. The difference in the percent-
to the measure. As compared with the age of readers supporting Coolidge in the
group reading out of town papers, the case of these two papers can be accounted
difference between this group and the for almost entirely on the basis of the
larger percentage of the readers of the
readers of the Times is I 5 per cent, with the
complicating factor that a much larger Post-Intelligencer who failed to indicate
proportion of the former failed to answer their preference. Again, the proportion
the question. A truer mneasure of the of La Follette supporters among the
difference perhaps is the difference betweenreaders of the Star, which supported La
the Times (positively opposed) and the Follette, was no greater than the propor-
readers of the Star (no stand), a differencetion supporting this candidate among the
of 7 per cent. Obviously, the small readers of the other two papers, one of
difference makes causal connection ex- which was indifferent, and the other
tremely doubtful or at least extremely vigorously opposed to this candidate.
small. In the case of the Union Record we find the
In the case of the next question, namely greatest scatteration between different
the question of the Skagit Project (a candidates, but in view of the fact that
proposal to enlarge and develop the city'sthe city as a whole is regarded as heavily
hydroelectric power plant) we have the Republican, the 34 per cent support of
most striking example of the lack of cor- La Follette among the readers of the
relation between the attitude of news- organ of organized labor, indicates again
papers and the attitude of its readers. that in a homogeneous group of this kind,
The Times vigorously opposed this proj- there tends to be a closer correspondence
ect and yet 75 per cent of its readers
between the attitude of the newspaper
favored the project, which was the read by such a group and the attitude of
the group. Its special position as an
attitude of practically the same percentage
organ of special group, however, intro-
of readers of the papers friendly or favor-
duces an element which invalidates to
able to the project. The Union RIecord
some extent the comparison of it with
again shows the most positive relation-
ordinary newspapers.
ship, undoubtedly due to its being the
organ of a special group. The results, as far as this study is con-
Finally, the relation between the atti-cerned, indicates unmistakably (i) that
tude of the newspapers and the attitudea modern commercial newspaper has little
of the voters in the last presidential elec-direct influence on the opinions of its
tion appears to be entirely negligible. readers on public questions. It probably

This content downloaded from

177.106.167.81 on Fri, 08 Mar 2024 00:02:20 +00:00
All use subject to https://about.jstor.org/terms
 THE NEWSPAPER AND PUBLIC OPINION 7I3

seeks to discover and reflect that opinion sports relatively low as compared with the
rather than to make it. (z) The stand rating by the men. 8
of a newspaper on public questions is a A ranking of the types of news on the
negligible factor in the reader's estimation basis of interest value gave the following
in selecting his newspaper. (3) Only order for the group as a whole: (i)
when a paper is the organ of some homo- Politics, (z) Disasters, (3) Graft, (4)
geneous group does a slight correspond- Financial, (S) Crime, (6) Theater, (7)
ence between the attitudes of the news- Marriage, (8) Society. The order was
paper and its readers appear. (4) News- the same for the male as for the female
paper opinions are perhaps themselves group, but the women ranked graft,
the products of the various forces which financial news, and crime relatively low,
make opinion in a commnunity. Of these, and theater, marriage, and society news
the newspaper is undoubtedly one, but its correspondingly high.
influence as such has perhaps been grossly Finally, in response to the question,
overestimated or at least the nature of "Do you think the papers reliably report
this influence has been misunderstood. the news of the world," 44 per cent
answered "Yes," 44 per cent, "No,"
II
9 per cent "Fairly," and 3 per cent failed
Further light on the attitude of people to answer. The estimated time spent
towards newspapers, in so far as the ques- reading newspapers daily was as follows:
tionnaire method is reliable in revealing less than IS minutes, ii per cent; i6 to 30
such attitudes, was secured from a ranking minutes, 37 per cent; 31 to 45 minutes,
by each person of the different departments I3 per cent; 46 to 6o minutes, 2.4 per cent;
of the newspaper on the basis of its rela- 6i minutes and up, 8 per cent; no answer,
tive interest to the reader. This part 4 per cent.
of the inquiry is obviously open to the III
objection that few persons are capable or
willing to give the actual facts in such The interest in the question of news-
cases, but tend to respond with the con- paper influence on public opinion has
ventional attitude. The difficulty of ade- centered largely around the relationship
quately classifying the contents of a between election success of candidates
newspaper is also a complicating factor. for office and their newspaper support
The results, however, may have some in the campaign. This relationship and
degree of validity and hence are briefly influence has been variously estimated
submitted here. and ranges from the view which holds
For the group as a whole, the interest that the two are directly, positively, and
value of different departments was rankedcausally related, to the view which
as follows: (i) General News, (Y) Edi-
8 The method employed was the same as was
torials, (3) Cartoons, (4) Advertising, employed by Hotchkiss and Franken of New York
(S) Sports, (6) Communications from University in a simnilar study in i9Zi. "Newspaper
readers, (7) Radio, (8) Puzzles. The Reading Habits of Business Executives and Profes-

order was the same for both men and sional Men in New York." Copyrighted i92z by
New York Tribune. The findings in this study, and
women and corresponded closely, except
others cited in this report, agree in general with
that the women's rating of advertising those of the present study, except that the Seattle
was relatively high and their rating of group shows a much higher appreciation of cartoons.

This content downloaded from

177.106.167.81 on Fri, 08 Mar 2024 00:02:20 +00:00
All use subject to https://about.jstor.org/terms
 714 SOCIAL FORCES

holds that there is little or no relationship. returned the heaviest tnajorities for the
The former view is undoubtedly more Democratic candidate for governor in the
generally held. But much evidence may last election had combined Democratic
be adduced in support of either position. circulation of 85I as against 3oo Repub-
There is, on the one hand the unquestioned lican, and 3334 Independent. Likewise,
fact that candidates for mayor and other the three counties returning the heaviest
offices in many cities have been elected Republican majorities were found to
in the face of the united opposition of have no Democratic newspapers.
all or nearly all the newspapers.' There A similar comparison of the usual
is on the other hand apparently clear political classification of different states
cases where newspapers have been very shows a more positive tendency for states
influential in causing public reforms. of usually heavy Republican or Demo-
These latter cases, however, have fre- cratic majorities to have the bulk of its
quently been cases where the great role newspaper circulation listed as of the
of the newspaper has been that of exposing
same respective party affiliation. In this
conditions not generally known and caus-connection the most significant finding
ing agitation over them. But the ques- was the fact that the so-called doubtful
tion as to the nature and degree of the states, e.g., Indiana, have an unusually
relationship existing in either case may large percentage of their newspaper circu-
still be said to be open. lation listed in the independent column.
As a more general and indirect approach The fact that there seems to be, as was
to this problem, an attempt was made to be expected, a positive relationship
to ascertain the general relationship on the whole between the avowed polit-
between the political alignment of various ical affiliation of newspapers in a given
counties and states as evidenced by their area with the usual political complexion
votes, and the avowed political affiliation of that area, does not, of course, provc
of the newspapers of these areas, stated in anything as to the influence of the news-
terms of the total circulation of all the papers in causing the political attitude
papers of each party affiliation. For the of the people among whom it circulates.
state of Washington a study was made of It is quite as reasonable to assume that
the circulation of newspapers located in the paper locates and takes its stand in
the different counties and the political accordance with the prevalent public
alignment of these counties in the last view, thus being a reflector rather than a
gubernatorial election. This was obvi- creator of public opinion. In view of the
ously an imperfect method due to the fact that the modern newspaper is prima-
fact that many newspapers circulate out-
rily a business enterprise, the latter is a
side of the county of their location. The much more reasonable conclusion. This
problem is further complicated by the correspondence, therefore, on which many
fact that a great many papers prefer to assumptions as to the direct power of the
list themselves as Independent. This is press in politics is based, furnishes very
especially true of papers in small town little legitimate ground for such assump-
and rural areas, where a paper must ap- tions. In the first place, it is at most an
peal to all factions for financial support. indirect and general approach to the
Nevertheless, the three counties which question of the relationship between
'See Bernays, Crystallizing Public Opinion, p. 7I newspaper support and the success of a
and
following for a summary of such cases. given candidate in a given election. In

This content downloaded from

177.106.167.81 on Fri, 08 Mar 2024 00:02:20 +00:00
All use subject to https://about.jstor.org/terms
 ANTHROPOLOGY, A NATURAL SCIENCE 7I5

the second place, the conclusions drawn true of other cities and especially of small
from part one of this study, as well as towns and rural areas is a question which
numerous cases where candidates win in only further inquiry can reveal. There
spite of unanimous newspaper opposition, is reason to believe, in fact, that the influ-
throw very serious doubt on the assumed ence of the press on the people of large
causal relationship. Certainly newspaper cities on questions such as those here
support cannot be regarded as in any way investigated is far less than in small towns
essential to political success, and its power and rural areas, due to the number and
in a negative way, namely, in refusing all frequency of other contacts in the city
publicity to a candidate, is perhaps more through reading several newspapers and
important than its positive stand for or through clubs, unions, and personal and
against a candidate or an issue. group influences. This paper does not, of
The conclusions here set forth, es- course, pretend to answer in any complete
pecially in part one of this paper, seemn or conclusive way the question of the
justified on the basis of the present study newspaper's influence. Obviously this
as well as from facts of general observa- study has made no attempt to evaluate
tion.)0 Whether these results will hold the more subtle and far-reaching influence
of the newspaper through suggestion and
10 For such facts see Bernays, Crystallizing Public
repetition and possible suppression of
Opinion, p. 7iff. Also McKenzie, "Community
news. The more general validity of the
Forces: A Study of the Non-Partisan Municipal
Elections in Seattle." Jr. of Soc. Forces, Jan., I92.4, conclusions remain to be tested by further
PP. 272X-3. study.

ANTHROPOLOGY, A NATURAL SCIENCE?

ROBERT REDFIELD

SCIENCE, broadly speaking, is the The term natural science, on the other
systematic investigation of ob- hand, is often reserved for scientific
served phenomena. It is recog- investigation which seeks to classify data
nized that this investigation may be and to reduce a wide range of observed
directed towards one of two distinlct phenomena to a brief statement or for-
and opposable ends. It may be the aim mula. This formula is termed a natural
of such investigation to discover and set law. It is, of course, not a law at all:
out specific sequences, temporal or spatial, it compels nothing. It is merely a short-
of objects or events. History and ge- hand description of phenomena observed
ography are scientific disciplines of this to recur.' It is the processual counter-
sort. They are sometimes called descrip- part of the generic concept. It is tested
tive sciences. In fact all sciences are pragmatically, not by any standard of
descriptive, but the events or objects
1 Karl Pearson: The Grammar of Science, 77; A. D.
of the historical-geographical sciences
Ritchie: The Scientefic Method, 55; Hobson, The
are described as they are encountered in
Domain of Natural Science, z6; L. L. Bernard, "Scien-
time or space, and each datum is unique tific Method and Social Progress," Amer. Jour. Soc.,
and not subject to verification. July, I92.5.

This content downloaded from

177.106.167.81 on Fri, 08 Mar 2024 00:02:20 +00:00
All use subject to https://about.jstor.org/terms