SEP 1 l 2001 052,} 01 ABSTRACT PRESS COVERAGE OF CONGRESSIONAL CAMPAIGNS: A CONTENT ANALYSIS OF THE 1960 CAMPAIGN IN MICHIGAN AND TENNESSEE by William C. Spragens This research project in political science was a study of political communications behavior. The specific design called for content analysis of microfilm files of a daily news- paper sample composed of six Michigan dailies and six Tennessee dailies published in the congressional campaign period from September 1 through November 8, 1960. Two newspapers were also analyzed for coverage from September through November, 1962, for limited comparisons. Tennessee newspapers analyzed were the Memphis Commercial Appeal, the Nashville Banner and Tennessean, the Chattanooga Times and the Knoxville News- Sentinel and Journal. The Chattanooga Times and Knoxville News-Sentinel were also analyzed for the 1962 campaign period for congressional campaign coverage. Michigan newspapers an- alyzed were the Traverse City Record-Eagle and Lansing State Journal and the following member newspapers of the Booth News- papers, Inc., chain: The Grand Rapids EEEEE: the Huskegon Chronicle, the Saginaw News and the Flint Journal. Two of the Tennessee newspapers analyzed, the Commercial Appeal and the News-Sentinel, are Scripps-Howard newspapers. The Knox- ville Journal was bought by Scripps-Howard interests, but still maintains an independent editorial policy. The major hypothesis was that press content (as ana- lyzed from microfilm files of the 12 newspapers) would reflect the interaction between press structure (ownership, editorial policy, news coverage and related factors) and the political structure of the congressional districts, one type of elec- toral decision-making institution. The major hypothesis was in large part validated. Specific hypotheses found valid in large part posited that bias in the Press could be measured, that urban-rural demographic factors would be reflected in campaign coverage, that VIP presence at campaign events and type of events would be reflected in difference in coverage, and that competitive- ness of the districts would be reflected in news coverage of the campaign in those districts' newspapers. A specific hypothesis not conclusively validated suggested that the incumbency variable would be an important factor in news coverage. The problem of "coattail reporting" (transfer of pulling power from major to minor candidates in press coverage) as a major factor in congressional campaign coverage was left for a later study. We also examined editorial policy, especially in the form of editorial endorsement of candidates, choice of columns used on editorial pages, and selection of letters to the editor. We found it possible to make a crude measurement (the FUN Rating) of opinion bias shown in news coverage, and this same methodology can be applied with some success to editorial statements of opinion. Treatment of news stories as another factor of news- er policy was also examined. Although we found consider- 6 effort to maintain fairness in news columns, as dis- guished from editorial columns, we found some evidence of s, even here. The design could be applied to senatorial races, sidential and gubernatorial contests and even to legis- ive and local contests, and could be applied to a longer torical period than merely one or two consecutive elec- ns, as we used it. More than two states could be analyzed this manner. We also suggested that other methods of lysis (besides content analysis) could be employed, for mple a combination of content analysis with survey meth- logy, or an application of "simulation" techniques to paign communications behavior. This is to certify that the thesis entitled PRESS COVERAGE OF CONGRESSIONAL CAMPAIGNS’ A CONTENT ANALYSIS OF THE 1960 CAMPAIGN IN MICHIGAN AND TENNESSEE presented by WILLIAM C. SPRAGENS has been accepted towards fulfillment of the requirements for i_“ D degree in Wflzfi‘z I wae /Major professor Date—Mk 0-169 PRESS COVERAGE OF C ON GRESS IONAL CAMPA IGNS : A CONTENT ANALYSIS OF THE 1960 CAMPAIGN IN MICHIGAN AND TENNESSEE BY (A William C? Spragens A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Political Science 1966 ." arr“ in... t", l '1‘ un‘ “1a.? © Cepyright by WILLIAM CLARK SPRAGENS 1967 ACKNOWLEDGEHENTS The author of this dissertation is indebted to staff members of the Michigan State University Department of Political Science for counsel and advice given at various stages of the doctoral program and thesis work. In chronological order, acknowledgement is made to Dr. Ralph M. Goldman for his guidance while I was a Legislative Intern at the 1961 session of the Michigan Legislature. Ac- knowledgement is made to the department and its chairman at that time, Dr. Joseph LaPalombara, for a Falk Fellowship during the 1960-61 academic year and a graduate assistantship during the 1961-62 academic year, which made possible com- pletion of my doctoral course work. For his guidance and counsel in developing the research design I am indebted to Dr. Joseph A. Schlesinger. Further acknowledgement is made to Dr. Herbert Garfinkel and to Dr. Charles Press for their suggestions in connection with the research design. To Dr. Charles R. Adrian, department chair- man, who made teaching assignments available during the 1964-65 academic year, I extend thanks. I am deeply indebted to the chairman of my thesis com- mittee, Dr. Robert Sciglianc, for his interest and patient encouragement of my work. I am also indebted to the other members of the thesis committee, Dr. Erwin P. Bettinghaus of the MSU Department of Communication, and Professors Schlesinger and Press whose earlier contributions I have already mentioned. Finally I express my gratitude to my wife Elaine with- whose constant encouragement and patience and typing of MS., this thesis would never have been completed. Library facilities of Michigan State University, the ,igan State Library in Lansing and the University of .essee Library in Knoxville were made available with cour- ‘ and much appreciated co-Operation. WILLIAM C. SPRAGENS East Lansing, Michigan August 8, 1966 iv —"_""—"7 TABLE OF CONTENTS Page OWLEDGEPENTS o o o o o o o o o o o o o o o o o o o 111 'OFTABLIESOOOOOOOOOOOOOOOOOOO. V11 Iter I. A PROBLEM IN POLITICAL COMMUNICATION . . . . . l The Research Problem The Relevant Literature Gaps in the Literature Major Hypothesis and Specific Hypotheses ( Time Period Chosen for the Sample Compilation of Raw Data The Operational Design Summary Statement -.-— “11 iI. CONGRESSIONAL DISTRICTS AND NEWSPAPERS: BACKGROUND FOR THE STUDY . . . . . . . . . . 42 Purpose of the Chapter Analysis of Congressional Districts Background for the Michigan Districts Background for the Tennessee Districts Comparison of Michigan and Tennessee Politics Background for the Michigan Newspapers Background for the Tennessee Newspapers Chain Ownership Factors Comparison of Michigan and Tennessee Newspapers Implications for Theory I. HOW BIASED IS NEWSPAPER COVERAGE OF POLITICAL CAMPAIGNS? . . . . . . . . . . . . 64 Introduction Analysis of the "FUN Rating" Tabulation The Specific Hypothesis Analysis of Tabulations for the Full Sample Newspaper Policy and Political Background Other Types of Media Bias Qualitative and Quantitative Analysis mfl_._.. TABLE OF CONTENTS--Cont1nued Chapter Page IV. THE URBAN-RURAL DIMENSION IN CAMPAIGN COVERAGE AND ITS RELATION TO INCUMBENCY . . 103 Statement of Specific Hypotheses Implications of the General Findings Inferential Discussion of Specific Findings Incumbents in 1960 Contests Generalizations V. WHAT KIND OF EVENTS DRAW THE GREATEST c OVERAGE ? O C O I O O O 0 C C O O O O I O O 148 The Specific Hypothesis Compilation of the Data Analysis of the Data Inferential Discussion Conclusions on News Event Differentiation VI. COVERAGE IN COMPETITIVE AND ONE-PARTY D ISTR ICTS O O O O O G I O O O D O O O O O O 182 Statement of Specific Hypothesis Analysis of Data for Full Sample Analysis of Single-Newspaper Data Some Unexplored Aspects Inferential Discussion Some Tentative Conclusions VII. SUMMARY OF CONCLUSIONS AND NEW RESEARCH IDEAS 220 Summary of the Findings Toward an Analysis of Campaign Coverage Other Suggestions for Further Study ”FWD m 0 O O O O I O 0 I I O I I O O O D 0 D O 0 I 2#2 BIBLIOGRAPHY 0 I O O O O I O I O 5 O O O 0 O D U C O 243 vi LIST OF TABLES Chapter III Table I. Biased Coverage Variable: Adjusted FUN Ratings for 12-Newspaper Sample (1960) . . II. Biased Coverage Variable: Adjusted FUN Ratings for 12-Newspaper Sample (1960) . . III. Biased Coverage Variable: Adjusted FUN Ratings for l2-Newspaper Sample (1960) . . IV. Biased Coverage Variable: Adjusted FUN Ratings for 12-Newspaper Sample (1960) . . V. Biased Coverage Variable: Category Content Analysis of References to Candidates in an Urban District Newspaper (The Grand Rapids PreSS) (1960) o o o o o o o O O I o o s a VI. Biased Coverage Variable: Category Content Analysis of References to Candidates in a Rural District Newspaper (Muskegon Chronicle) (1960) o o o e o o o o o e o o 0 VII. Biased Coverage Variable in One Contest in Three Newspapers: News Columns vs. Opinion Columns (Knoxville News-Sentinel and Journal and Flint Journal) . . . . . . VIII. Biased Coverage Variable in Comparison of Two Newspa ers (Nashville Banner and Tennessean : Senatorial Contest and Two Districts (One Contested and One Uncontested) . . . . . . . . . . . . . . . Chapter IV Table I. Urban-Rural Factors: Newspapers Published in Urban Districts . . . . . . . . . . . . II. Urban-Rural Factors: Newspapers Published in Rural Districts . . . . . . . . . . . . III. Incumbency, Non-Incumbency Factors: Newspapers Published in Districts With Incumbent Candidates . . . . . . . . . . . vii Page 84 85 87 9O 93 96 99 101 128 130 132 Chapter Table IV. V. VI. VII. VIII. Chapter Table I-A. 1-30 I-C o II-Ae II-Bo LIST OF TABLES--Cont1nued IV--Continued Incumbency, Non-Incumbency Factors: Newspapers Published in Districts Without Incumbent Candidates . . . . . . . Urban-Rural Variable: Urban-Rural Variable: Incumbency Variable: Incumbent Coverage Constituency Press 1960) o e o e e O Incumbency Variable: Total Coverage for All Candidates in an Urban District Paper (Memphis Commercial Appeal: 1960) . . Total Coverage for All Candidates in a Rural District Paper (Muskegon Chronicle: 1960) on... Incumbent vs. Non- in the Urban (The Flint Journal: Incumbent vs. Non—Incumbent Coverage in the Rural Constituency Press (Traverse City Record- Eagle: 1960) . . News Events Variable: "Oral" Events vs. "Printed" Events (Mbmphis Commercial Appeal: 1960) . . News Events Variable: "Oral" Events vs. "Printed" Events (Memphis Commercial Appeal: 1960) . . News Events Variable: I O O O O O O O O 0 “Oral" Events vs. "Printed“ Events (Memphis Commercial Appeal: 1960) . . News Events Variable: National-Angle Events vs. District-Oriented Events (Flint Journal: 1960) . News Events Variable: National-Angle Events vs. District-Oriented Events (Flint Journal: 1960) . viii Page 134 138 140 143 169 172 174 176 178 LIST OF TABLES-~Continued Chapter V--Continued Table III. News Events Variable: Televised News Events vs. Direct Personal Appearances (Knoxville Journal: 1960) . . . . . . . . . . . . . . News Events variable: VIP Presence at Events as a Factor in News Coverage (Lansing State Journal:1960).............. Chapter VI Table I. II. III'A. III‘B. III-C. Competitiveness Variable: Total Candidate Coverage in a Competitive Congressional District (Frequency and Percentage) -- (Lansing State Journal: 1960) . . . . . . Competitiveness Variable: Total Candidate Coverage in a One-Party District . (Knoxville News-Sentinel: 1960) . . . . . Competitiveness Variable: Total Candidate Coverage Compared in a Set of Competitive Districts Flint Journal (1960) Compared With ChattanOOga Times (1962): Presidential vs. Off-Year Comparison 0 o o o o o o c o o o o Competitiveness Variable: Total Candidate -Coverage Compared in a Set of Competitive Districts Flint Journel (1960) Compared With Chattanooga Timee (1962): Presidential vs. Off“Year Comparison 0 o e o o o o o o e o o Competitiveness Variable: Total Candidate Coverage Compared in a Set of Competitive Districts Flint Journal (1960) Compared With Chattanooga Times (1962): Presidential vs. Off-Year Comparison 0 o o o o o o e e o o 0 Competitive, Non-Competitive Factors Newspapers Published in Competitive Districts , , , . . . . . . . . . . ix Page 180 181 202 205 207 210 212 214 ‘4'. ,‘ur n n . e a o . s r a o LIST OF TABLES--Continued Chapter VI--Continued Page Table V. Competitive, Non-Competitive Factors Newspapers Published in Non-Competitive Distr1Ct8 O O O O O O O 0 O O O 0 O O 0 O O 216 VI. Competitive, Non-Competitive Factors Rank Order of Coverage in Individual 8 O O 21 Newspapers.............. CHAPTER ONE A Problem in Political Communication I. The Research Problem Political science and political communications liter- ature, as well as general communications theory, includes much analysis of the structure and functions of political parties and party organizations, of the influence flow and the role of Opinion leaders in the public Opinion process, and of many other facets of the political communications process. Some relevant studies will be mentioned in this chap- ter, but there has not been -- either in the political par- ties literature, or in general communications theory -- any 1 beginning toward an effort to explain a communications in-. stitution (such as the Press) in terms of the political en- vironment in which newspapers are published, and in terms of the relationship between political institutions and private institutions like the Press. There has been no systematic effort to link political and communications variables by explaining a communications institution in terms of the political environment. This study is designed to make a beginning toward that end, by examining the communications behavior of a select segment of the Press (newspapers in a two-state, twelve-newspaper sample) to gain a better understanding of how the Press cov- ers one type of campaign -- the congressional campaign. Our focal problem (major hypothesis area) will be to es- tablish the relationship between a specific facet of politi- cal party life -- congressional campaigning -- and the in- stitution of the Press. We will do this by examining content of the newspaper press. We are pep interested in ex osure, but only in content. Content refers to the news stories, photographs, edito- rials, columns, cartoons and other items printed in the Press. In this context, we are concerned with content deal- ing with the campaign. Exposure refers to that portion of the content to which the reader-voter is "exposed" by a reading of the Press. Al- though this is conceded to be an important part of the proc- ess which precedes the voting decision, it is outside the scope of this dissertation. We will consider congressional campaigning for two states, Michigan and Tennessee, with a peeie concern for the congressional elections of 1960 but with a secondary inter- est in the 1962 congressional contests, and an auxiliary con- cern with the 1960 presidential campaign. Our primary in- terest is in the political structure of congressional dis- tricts, along with the institutional structure of the Press. Political variables of interest include election proce- dures, organization of the two major parties, incumbency and non-incumbency factors, competitiveness of district factors, urban-rural factors, attention paid by the newspapers to VIP's, and candidate campaign activity such as organizational efforts, public relations, and other campaign communications aspects. The principal communications variable in our analysis will be the editorial policy of newspapers and their news- handling policies as reflected in the treatment of campaign news stories. In order to understand this properly, we will also have to examine the end product of the behavior of in- dividual communicators in the Fourth Estate -- editors, pub- lishers, reporters and other participants in the communica— tions process. This and product is of course the printed page. We recognize that we are not analyzing the role and behavior of each of these communicators, but it will be pointless to ex- amine Press content without an understanding of how it is pro- duced. Our major hypothesis, which will be stated in detail later in this chapter, is that Press content provides an im- portant clue to the dynamics of the interplay of the con— stituencies and the Press (1.9., the reflection of the po- litical structure of congressional districts as filtered through the institutional structure of the Press), and that content analysis of Press content will disclose important facts about the campaigning in the districts and about the way in which district campaigns are reported in the Press. We as- sume that press coverage -- the end product of the institu— tion of the Press -- must be the output factor of its insti— tutional character. In operational terms, we will be using content anaylsis techniques to relate political variables (factors in the cam- paigns being covered by the Press) and communications vari- ables (the institutional structure of the Press, which em- braces among other things editorial policy and newshandling policies) to each other. We will be doing this in an effort to derive some tentative generalizations that may help us in beginning to develop a workable theory of political communi- cation. As we explain some of the phenomena which occur in reporting of congressional election campaigns, we can perhaps derive some generalizations which may broaden our understanding of the campaign process. Our concern here is less with the conduct of the cam- paign by the candidate for office than with the way in which that campaign is described in the Press. We recognize that newspaper coverage gives us a certain picture of that campaign, but it is the image of the candidate and his campaign that is projected through the medium that is the focus of our interest. In this specific study, we have focused on the Press as one of the important mass media. Our thesis is that we can show a great deal about candidate image, newshandling of polit— ical stories in the Press, and other characteristics of polit- ical communication through a careful content analysis of press coverage during the campaign period. We are not probing campaign coverage's influence on the outcome of an election, and indeed this question was specif- ically ruled out when we decided to confine this study to con- tent and remain away from the question of exposure. Nonethe- less, we can operate on the assumption that campaign coverage in the media must be one of the important influences on the campaign -- as it affects the voter's understanding of the campaign and its participants, the issues and the stances of the parties for whom the candidates are slated —- or else why would the candidates, their managers and other key figures in the campaign be so concerned with the kind of coverage their campaign is afforded? Not all hypotheses about what the reader learns from the Press are subject to empirical proof. Those which cannot be tested with measurable empirical data can only be the object of speculative inference. But we will focus primarily on those hypotheses which are susceptible to empirical testing with the data derived from our content analysis and tabulated for the tables appearing in this thesis. “.2— "-~s'T 9*: More importantly, we expect to find that demographic and political characteristics of a district will to a considerable degree show up in press coverage of the campaign in that dis- trict. To illustrate, if labor is an important factor in the district, discussion of economic issues should show up prom- inently in the coverage in the district. If minority group voting is an important factor in the district, civil rights and issues in which the group is interested should get an im- portant play in news coverage. These natural district inter- ests may be filtered through the screen of editorial policy or newshandling policy and may not get the prominent attention we feel they should, but it seems highly unlikely that a news- paper published in this kind of district would be able to ig- nore issues so highly pertinent to the voting population in the district. We may also anticipate that the content of newspapers at campaign time will tell us a great deal about the true nature of the institution of the Press. Along with the factors which we defined in describing the institutional structure of the Press, we are concerned in a secondary way with the circulation figures of each newspaper, which will be given in Chapter Two.1 Intimately related to the structure of the Press and struoture of the congressional districts, to resolve the prob- lem we have raised, is our typology of congressional district patterns of competition and our typology of newspapers. These classification schemes will be dealt with in detail, later in the chapter, when we discuss the methodology of our research. II. The Relevant Literature Major publications in our problem area fall into three types -- those dealing with political parties and party or- ganization, those dealing with public opinion and political Geommunicaticn, and those in general communications theory. We will list each in turn, with appropriate comments. Writings on parties over the past century have been con- cisely summarized by Austin Ranney2 who in his discussion of gee Doctrine ei Responsible Peppy Government treats in turn the observations of Woodrow Wilson in Copgressional Govern- 92233 A. Lawrence Lowell in Essays ep Government and Public epinion epe Popular Government: Henry Jones Ford in Tee Ripe egg Growth ep American Politics; Frank J. Goodnow in Politics egg Administration: M. I. Ostrogcrski in Democracy egg pee Organization ei Political Parties: Herbert Crcly in Tee gpep- iee e; American pipe, and those of other pioneer political scientists interested in parties. We will return later to Ranney's critique of Ostrcgorski's penetrating criticism of the party system in the United States. The next important category of parties literature in- cludes the commonly used texts of today. Some of the lead- ing ones among these are V. 0. Key's Politics, Parties, epe Pressure Groups;3 Hugh A. Bone's American Politics epe pee Peppy S stem;4 William Goodman's pee Tee-Peppy System ip pee united States;5 Austin Ranney and Wilmoore Kendall's "W". spur. ' ~. Democracy and the American Party System,6 and Avery Leiserson's Parties epe Politics: pp Institutional epe Behavioral Approach.7 The late Professor Key also contri- buted an important volume on public opinion and parties, Public Opinion egg American Democracy.8 Here we remind the reader that our basic concern is with operation of parties as they relate to conduct of election campaigns, and specifically as they relate to campaign re- portage in newspapers (the other media we propose to analyze in a different study not included in the scOpe of this dis- sertation). We find the strain of “boss literature" dealing with big city politics that was quite prevalent during the Progressive Era and the New Deal period in American politics has been largely supplanted today by publications about metropolitan politics and case studies of party clubs. Two representative titles from this literature are Edward Banfield's Political Influence9 and an excellent study of New York City, Chicago and California club politics, James Q. Wilson's pee Amateur Democrats.10 In the context of our interest in parties in elections, valuable theoretical constructs have been contributed to the literature of political communication in party campaigning by Joseph A. Schlesinger who developed a theory of "Political Party Organization"11 and Donald E. Stokes who developed "Spatial Models of Party Competition".12 In the analysis of campaigning and political communication, important landmarks in the literature include a collection of articles on congressional elections, edited by McPhee and 13 Glaser. The pertinent articles in the McPhee-Glaser col- lection include a discussion of the concept Of "political 14 and a discussion of the immunization" by McPhee and Ferguson "contextual dimension in voting" by Ennis.15 McPhee and Ferguson relate press content to their theory of "political immunization" stressing the factor of overexposure of polit- ical figures. The Ennis article categorizes the environ- mental context in which voting takes place. Finally, an article by Glaser and Kadushin on "Political Behavior in Mid- "16 is particularly relevant to any compari- term Elections son of 1960 and 1962 voting and campaigning. These writers distinguish between presidential year coverage and off-year coverage in comparing turnout and minority party performance in an effort to explain variations in the pattern of party competition. 17 we find a discussion of the In Burdick and Brodbeck, mass media and voting. This treatment of party propaganda, Opinion leaders, exposure and impact, by Kurt and Gladys Engel Lang,18 is primarily a general theoretical discussion and gives us little hard data, however. In the last volume he published before his death, V. O. Keylg discussed problems in analysis of the mass media. Here he dealt with structure of the mass media, including news- papers, and presented a critique of the Lazarsfeld-Berelson2o "two-step flow hypothesis" of influence in communications 10 behavior. (The "two-step flow hypothesis" suggests that the newspapers and other mass media influence "Opinion leaders" who pay greater attention to public affairs than the general public and that these opinion leaders in turn influence their friends and associates. Thus the influence of the mass media occurs in two steps, one direct and the other indirect.) This hypothesis has been refined somewhat since the discus- sion of newspapers in the 1940 presidential campaign which appeared in Tee People's Choice (a study of presidential voting in Erie County, Ohio),21 and a full statement of the "opinion leaders" or “two-step flow" hypothesis in the 22 pub_ Katz-Lazarsfeld "Decatur study", Personal Influence, lished in 1955. Key himself indicated some preference for the idea of a "multi~step flow" of Opinion leadership and in- fluence (suggesting that there were more than two steps in the process of influence by which Opinions are shaped) -- an idea treated in the "diffusion" literature of general com- munications theory, especially by Melvin L. DeFleur in "The "23 and also by 25 Emergence and Functioning of Opinion Leadership Charles Winick24 and Herbert Menzel and Elihu Katz. Pro- fessor Key also dealt with political parties and elections and their role in opinion formation. Finally, in a discus- sion of public Opinion and democratic politics, Key dealt with his theory of interaction in Opinion formation, intro- ducing the concept of “Opinion dikes" regulating the flow of influence in the Opinion process. His construct was quite general, however. 11 26 public opinion text author, pro- Clarence Schettler, vides a general discussion of the newspaper's role in public opinion formation. His treatment, written in 1960, though fairly descriptive, provides good factual background on newspaper ownership, wire services’ role in newspaper cov- erage, and related factors. In their Personal Influence?7 commonly referred to as the "Decatur Study", Elihu Katz and Paul F. Lazarsfeld dis- cuss the role of opinion leaders in formation of political and other types of opinions, deal with public affairs leaders and the "two-step flow" hypothesis. Their report on a major research project done in Decatur, Illinois, in the early 1950's by survey techniques, deals with the relationship be- tween personal influence on decision-making and how it sup- plements or supplants media influence. Despite its valuable data, this study is primarily focused on exposure and opin- ion leaders. We are not concerned in our design with ex- posure, but we are interested in the editorial output Of such opinion leaders in the Press as editorial writers, syndicated columnists, and political writers. We are interested in these Opinion leaders in their relationship to the institu- tional features Of the Press and to party and demographic characteristics of congressional districts. 28 on public opinion and propaganda deals The Katz reader with mass media including such aspects as the effect of a newspaper strike on readership patterns, an optimism-pessi- mism analysis of newspaper headlines, and statistics on 12 effects on freedom of the Press from diminishing daily news- paper competition in most metropolitan areas of the United States. 29 discusses the theoretical importance Karl W. Deutsch Of the measurement of information and the fidelity of chan- nels in dealing with communication models and decision- systems. (If a newspaper is to have fidelity as a channel for a candidate, it must convey his message as intended.) Deutsch also discusses information as a "quantitative con- cept“ at the middle level of communications and command. This middle-level concept can be adapted and applied to political decision-systems, such as parties, and a study of the electoral process itself would be of great pertinence in applying Deutsch's construct. Joseph T. Klapper30 ties together many strands of com- munications research into a single theoretical statement. One might argue that it is the most meaningful analysis of general communications research done up to the date of its publication in 1960. Klapper deals with broad general hy- potheses, deriving the "nexus hypothesis" which states that the media are but one of a whole series of variables affect- ing public attitudes. In our case, dealing with voter atti- tudes and newspaper attitudes (represented respectively by ag- gregate voting data and editorial content and coverage) in a congressional campaign, we are relating political and com- munications variables. Our findings should strengthen Klapper's early formulation by giving evidence of this nexus 13 in a very specialized area of communication. Relevant reference works on this topic of campaigning and political communications include Ayer's Directory,31 the 32 Editor and Publisher Yearbook, Wilbur Schramm's collection 34 of articles,33 the Congpessional District Data Book, the Congressional Quarterly Census Analysis,35 Ithiel de Sola Pool's Trends ip Content Analysis?6 Berelson's article on "Content Analysis",37 and Robert C. North, Holsti, Zaninovich and Zinnes, Content Analysis: e Handbook With Applications 38 for the Study ei International Crisis. In the handbook, North and his co-authors state: Content analysis research usually involves the following stages. First, the research question, theory, and hypotheses are formulated. The sample is then se- lected, and the categories are defined. Next, the doc- uments are read and coded, and the relevant content is condensed onto special data sheets. After coding, items placed in each category may be scaled, whereupon counts in frequency or intensity are made. Finally, interpretations of the findings are made in the light of the appropriate theory. Besides the four major studies of presidential voting,39 and the growing literature on "coattail voting”,4° we find several volumes and articles on congressional elections, or applicable to their study. These include the Schlesinger formulation for classifying states according to degree of inter-party competition,41 the Cutright-Rossi discussion of candidate organization at the grass-roots level,42 the Miller- Stokes treatment of constituency influences on congressmen,45 Charles L. Clapp's treatment of congressional campaigns from 42, the legislator's point of view, Cortes A. M. Ewing's study of Southern primaries written in 1953.45 Lewis A. Froman Jr.'s ”-4 ‘w -“ ‘V . '7- a: ‘ “an e 6 . . .._' ‘ ' a' .1'“' S" ’ ~ " - _ .~ 5‘ .Qf .’ ~ __ t e — 14 volume on regional and party differences in voting turnout in congressman-constituency relations,46 the important Ferguson- Smuckler study which made a comparative content analysis for senatorial elections in two states, Wisconsin and Connecticut, in 1952,47 and of course Key's discussion of congressional nominations in his standard parties text.48 III. Gaps in the Literature Considerable as the literature is within our problem area, most of it is either too broad, emphasizes different as- pects of the problem, or uses different approaches from our design. In the above summary, we have found much concern with the "coattail problem" as it relates to voting and electoral data, discussion of mid-term voting patterns, analysis of regional and other influences on turnout, and a growing interest in the effect of constituency relations Of the incumbent on voting patterns both in elections and in roll-call votes in Congress. What aspects of our topic has the existing body of writings either failed to deal with, or approached from an- other attitude of inquiry? Schlesinger focuses on party or- ganization rather than the Press. Except for the Ferguson- Smuckler study, focusing on senatorial elections, we have no detailed literature stressing relationships between the po- litical context of campaigning and the institution of the Press. Most articles and writings have either stressed the behavioral or the institutional approach. Very few analysts have used a synthesis of the two to examine patterns Of Nb 15 behavior within the institutional context of political and communications institutions. Key's analysis, based on survey research data, deals with the public opinion process in the whole sweep of Amer- ican politics. Since the discussion in ipe People's Choice of the role of the mass media in presidential campaigns, we have little detailed reference to the political communica- tion process in the campaign. In pinpointing the area of our analysis, we were aided in developing a profile of the po- litical complexion of the congressional districts by Congges- sional Quarterly special reports giving demographic data on the composition of various districts in the two states chosen for our analysis.49 Basically this research project aims to get at some Of the unresolved issues we have discussed above, especially in an effort to develop an approach to a coherent theory of the end product resulting from the relation between communications media -- in this specific study, the Press -- and the polit- ical units of electoral decision-making, such as the congres- sional district, that are involved in the biennial election contests. The end product of this relationship is, of course, the content of the Press during the campaign period. IV. Major Hypothesis and Specific Hypotheses A theoretical statement at whatever level should state a major hypothesis, some specific hypotheses derived from it, and a list of inquiries which when answered will illuminate and support the individual specific hypotheses as well as the m 16 major hypothesis. Expected findings should be summarized when actual findings based on the data are stated after the data have been compiled into tabular form. We have followed this procedure in the remainder of this work. The following is the major hypothesis: The way ie which e congressional cappaign ie reported ip the Press ie e result ei two major variables -- (e) the political context, 1. e., district structure, electoral competitiveness and incumbeney, and (p) the structure ep the Press, 1. e., newspaper policy whose basic components are endorsements ei candidates and handling e; political stories. This can pe demonstrated through content analysis e; newspapers published during the campaign time period. For clarity's sake, we include the following glossary of terms used in the major hypothesis: District structure -- The geographical division of the districts and the various socio-economic components affecting district politics. Electoral competitiveness -- Division of the major party vote in the specific elections examined, during the period from 1952 through 1962. (This time segment was chosen 17 because the apportionment was the same in both states for the period Of this case study, except that an at-large district, the Nineteenth, was added in Michigan for the 1962 election only. Using figures for the 1952-62 period, we could deter- mine average party strengths, making comparisons across states, and thus we could analyze the 1960 returns in the light of the districts' voting behavior for the previous decade.) I Incumbeney -— The factor of whether the congressional candidate is holding congressional office or is an outsider seeking the office. Strpgture.e; the Press -- Factors which go to make up the structure of the Press as an institution include its ownership, its policy preferences, and its newshandling pol- icies (we shall define the latter two, which are our prin- cipal concern in the major hypothesis). We can find data on its ownership which is presented elsewhere in the thesis; through editorial endorsements we can find data on its policy preferences; through content analysis of newspages we can discover the general trend of its newshandling policies. Editorial policy -— The political attitude of a specific newspaper or chain of newspapers (e. g., the Scripps-Howard chain or the Booth Newspapers, Inc., chain in Michigan) to- ward parties, issues or candidates for public office. This is one aspect of newspaper policy, a key variable in the hy- pothesis.) Newspaper policy is a basic component of the in- stitutional structure of the Press. 18 Newshandling policy —- The basic guidelines according to which a newspaper handles news about candidates, parties, is— sues and other aspects of the campaign. These guidelines will be reflected in the treatment given these political phenome- na in coverage of the campaign in the newspages of the daily newspapers in the sample. (A separate glossary of other definitions relating to the specific hypotheses will be presented after these have been stated later in this chapter.) In dealing with the minor variables of our specific hy- potheses, to be stated later in the chapter, we find two major variables running through the design. These are the political context of the congressional districts (including district structure, electoral competitiveness and the incum- bency factor), and the institutional structure of the Press (principal aspects of which in our concern are editorial policy, especially candidate endorsements, and newshandling policy). Also we find a linking variable (to borrow V. O. Key's terminology), press content and its measurements and categories. In the political context of the districts as outlined above, the most important factor as a lesser variable is prob- ably the competitiveness factor as measured by the total major-party vote polled by each major party during the period being analyzed (1952-62) and for the specific election we are interested in, 1960, in using our content data. Factional strength would be of major importance if we were analyzing a 19 primary campaign, but this is of only secondary concern to us in this analysis as the content data cover the period of the general election campaign. Next to party control in importance is the factor of in- cubency and non-incumbency. Another significant lesser var- iable is the urban-rural dimension. While such things as campaign organizational efforts, candidate personality factors and choice Of issues for the campaign are Obviously a part of the content measured and analyzed in our research, we have not treated these as separate sub-variables. In communications variables, important factors are amount and percentage of potential exposure for each party's candi- date, circulation of the various newspapers in the sample, their ownership and key personnel, and their editorial and newshandling policies. Any careful content analysis should develOp these variables. We have chosen in this research to limit the scepe of analysis to content and its immediately re- lated factors, but this does not mean this exhausts the pos- sibilities. The linking of the communications variables with the political ones will have to be done through painstaking con- tent analysis, using both quantitative egg qualitative tech- niques, for statistical factors mean little without careful verbal analysis. The methodology will be explained in Sec- tion VI of this chapter. Before stating our specific hypotheses, we should define basic terms used in describing congressional districts. 2O Basically, a demographic classification of congressional dis- tricts according to pOpulation characteristics would suggest three types of districts -- rural, suburban, and urban. How- ever, because urban-suburban newspaper circulation and rural newspaper circulation seem to be two basic types into which the Press can be categorized, we have chosen to use two cat- egories for the CD's: (1) Rural districts, and (2) urban districts (the latter including both urban and suburban voters living in a metrOpolitan area such as Memphis-Shelby County or Nashville-Davidson County in Tennessee). If a more de- tailed analysis were desired, the suburban areas could be separated out. This might be apprOpriate for the study of some densely-pOpulated Eastern states where some districts are suburban in their entirety. However, the more common pattern in Michigan and Tennessee (especially since we are excluding Detroit newspapers from this study) is the exist- ence of rural districts alongside urban-suburban (or metro- politan) districts. Thus we believe the two basic categor- ies of rural districts and urban districts would be adequate for the analysis of these two states. It is also necessary to define the terms we use in de— scribing the Press in our two states. We can refer to small city dailies, in journalistic terminolOgy, and to metrOpol- itan or big-city dailies. In political science terminology, the small city daily is likely to be one published in a rural congressional district, while the metrOpolitan daily is likely to be one published in an urban district. When the 7 __ 77‘_—1___,,_, —. _r_. _ _, .— ....... _-_. 1...”... --I.-—-:' -—‘ - 21 political science term is used, the journalistic term that corresponds to it is automatically implied. Examples of rural district dailies would be the Muskegon Chronicle and Traverse City Record-Eagle in Michigan. An example of an ur- ban district daily in Tennessee would be the Memphis gee- mercial Appeal. POpulation densities are somewhat differently distri- buted in Michigan and Tennessee and newspaper circulation patterns are correspondingly somewhat different. Whereas ’ Michigan has one extremely large metrOpolitan center, the Detroit-Wayne County urban area and its suburban ring in Wayne, Oakland and Macomb Counties, and several smaller cities on the order of Grand Rapids, Lansing, Benton Harbor and Ann Arbor (all in the Lower Peninsula), the political controversies such as reapportionment involving rural—urban factors have tended to pit the Detroit-Wayne County area against outstate Michigan, with the sparsely pOpulated Upper Peninsula sometimes being a separate factor from the out- state portion of Lower Michigan. Traditionally, outstate areas in the Lower Peninsula (such as Grand Rapids, for ex- ample) have been Republican strongholds, while Detroit-Wayne County has been a Democratic stronghold and there has been some Democratic strength in recent years in the Upper Pen- insula and in scattered outstate areas such as Flint. In Tennessee, however, while Memphis is clearly the largest city, there are three other major cities all of which have importance in state politics. When reapportionment 22 became an issue in the famous Tennessee reapportionment case decided by the United States Supreme Court (eepep v. 9e32, 1962), the four major cities of the state, Memphis, Nashville, Knoxville and Chattanooga, co-Operated in the interests of the urban majority in the state. But the most deeply rooted political divisions in the state (some times more deeply rooted than partisan differences) are sectional intra-state distinctions dating back to the Civil War and Reconstruction. Middle and West Tennessee (whose principal cities are Nash- ville and Memphis) were pro-Confederate areas at the time of the Civil War secession referenda, and the current views of their citizens on issues like civil rights still reflect this historical conditioning. While Upper vs. Lower Peninsula or Detroit vs. outstate are still of great importance as factors of division in Michigan politics, it is doubtful that these divisions are quite as deep-seated as the century-old intra- state sectional differences in Tennessee. These sectional differences within the State of Tennessee are so important 'that the Tennessee Constitution recognizes them by formally dividing the state into three Grand Divisions -- East Tennessee, Middle Tennessee and West Tennessee -- and seeing that the interests of each Grand Division are substantially represented in political and electoral arrangements for the state. Another important factor in comparative analysis of Michigan and Tennessee politics is the issue dimension, as well as the historical factor. The issues which strike 23 closest to the hearts of Tennesseans are probably still racial issues, for all that Tennessee is considered by some to be a "border state" and institutions like the University of Tennessee and the school systems of the major cities have been integrated with much less ado than the similar insti- tutions of Alabama and Mississippi (dynamitings at Clinton and bombings at Nashville accompanied the early stages of in- tegration in Tennessee in the late 1950’s but there has been less of this sort of activity and more of the "sit-in" activ- ity typical of states like the Carolinas since 1960.) Given the continuing importance of racial issues in Tennessee, it is not surprising that some of the outlines of a "political relief map" of Tennessee still follow lines that were first established during the time of the Civil War and Reconstruc- tion. Tennessee, too, reflects the imprint of the politics of the "Crump Era" when Boss Ed Crump of Memphis regularly had an alliance lining up Shelby County with East Tennessee interests in legislative politics to the detriment of Middle Tennessee, Nashville-centered interests. The existence of such an alliance shows that partisan affiliation in Tennessee may be of less importance as a political factor than in Michigan, where party is an important factor due to the dif- fering nature of key issues in the state's politics. One- party Democratic strength in Middle and West Tennessee and one-party Republican strength in East Tennessee have probably promoted a sort of casualness about party affiliation that has created a climate in which "deals" for bipartisan 24 coalitions such as the Crump-East Tennessee "understanding" could flourish. To cite another example of such "dealing" in two districts we studied, we could point to reports of covert support given by Knox County Democrats in the Second District to a moderate Republican, John Duncan, in his race for the "non-partisan" mayoralty post in Knoxville. Pre- sumably Duncan kept some of this Democratic support in his successful race for Second District congressman in 1964. In the Third (ChattanOOga) District, a similar factor might be noted in the press speculation in 1962 that the defeated in- cumbent conservative Democrat, Rep. J. B. Frazier, and Gov. Frank G. Clement (reportedly aligned with conservative ele- ments in Hamilton County) gave aid and comfort to Rep. William E. Brock III, the successful Republican congressional candidate from the ChattanOOga district. This would indicate the develOpment of a kind of "three-party politics" of con- servative Democratic factions, liberal Democratic factions, and Republicans in the Chattanooga case, and in the Knoxville case Of Democrats and moderate Republicans lined up against Old Guard Republicans. This is much the same kind of tri- factional politics that is found in Kentucky, which like Tennessee has a traditional enclave of Republican strength in its eastern mountains which goes back to the era of the 1860's. Michigan, on the other hand, probably underwent the shaping of its present political divisions in the 1930's when economic issues were dominant and the “sit-down strikes" 25 in the Detroit area auto plants during the term Of Gov. Frank Murphy split liberals and conservatives along economic lines and started a class politics pattern which may still be ob- served today. Although the growth of Negro pOpulation in Michigan brought on by the expansion of the auto industry and the hiring of in-migrating Negroes from the South made civil rights an issue of some importance in the state, an Observer can find many points to support the thesis that economic is- sues are still more important in Michigan politics than ra- cial issues. Thus the politics of the two states can be said to Operate in a markedly different context, despite the in- creasing nationalization of our political system which made it possible for Senator Estes Kefauver of Tennessee, victo- rious in his August primary contest for re-election in 1960, to campaign for Michigan congressional candidates such as Genesee County Prosecutor Jerome F. O'Rourke of Flint. One further word should be said in any comparative anal- ysis Of Michigan and Tennessee politics, and that is that while civil rights is a dominant issue in Tennessee politics, economic issues are not unimportant there. We might say that they occupy the same kind of secondary role in Tennessee that racial issues probably do in Michigan. We hOpe that in later research it may be possible to analyze some of these factors in a grouping of four states, such as Kentucky-Tonnessee-Michigan-Illinois, to see whether these differences are due simply to characteristics of states or to a differing nature of regional politics within the 26 United States. That question is beyond the scOpe of this thesis, however. The four specific hypotheses are as follows: 1. Biased coverage (1. e., "slanted" coverage, or cover- age tending to favor one candidate or party at the expense of the other) of congressional campaigns is more likely to be found in rural-district papers than in urban-district papers. 2. Total coverage of candidates will be greater in newspapers published in rural districts than in those pub- lished in urban districts (e. g., Memphis in the Ninth Tennes- see CD, or Grand Rapids in the Fifth Michigan on). This will be so in two specific cases, stated in the following sub- hypotheses: (a) Both major-party candidates get more total coverage (in terms of measurement of actual newspaper space and not potential exposure) in rural than in urban districts. (b) Non-incumbents get more coverage in newspapers pub- lished in rural congressional districts than in newspapers published in urban districts. As a related prOposition, news- paper coverage of non-incumbents will tend to be found in "round-up" or "wrap-up" stories dealing generally with the campaign, rather than in separate individual stories in papers published in urban districts. Non-incumbents will, however, get somewhat more coverage where they mount an active or in- tensive campaign, regardless of whether the newspaper giving coverage is published in an urban or rural district. For our purposes, increased campaigning is assumed to be 27 reflected in greater coverage of the non-incumbents and we do not go beyond the evidence found in the content analysis data. In the case of non-incumbents nominated by the major- ity party, we expect to find them getting more mention than minority party non-incumbents but less than incumbents of either party. 3. Noticeable differences will occur in coverage -- in terms of space allotted, news play according to position on the page and page number, and emphasis according to headline size -- for different types of campaign events, according to: (a) whether a party "VIP" was present at the event; (b) wheth- er events were nationally oriented or district oriented: (c) whether events were televised or involved direct personal ap- pearances of candidates and other campaign figures, and (d) whether events were "printed" or "oral" in nature. The three types of events categorized as "oral" included rallies and speeches (including debates), news conferences, and organi- zational activities (caucuses, parties, headquarters Openings, and photo sessions). The four-type categorization of "printed" events included press releases and reprints (in- cluding "handouts"): letters to the editor: editorials, col- umns and cartoons, and background features. As a corollary subhypothesis, we anticipated that organizational activities would generate more coverage than routine party rallies. 4. In establishing categories of congressional districts according to degree of party competition, we expect to find greater campaign coverage (in terms of total column inches) in 28 newspapers published in highly competitive districts than in non-competitive (one-party or one-party predominant) dis- tricts. Thus a competitive district with a normally close margin and a high rate of turnover in party control could be expected to have more intensive campaign coverage (as a usual pattern) than a one-party district where readership interest and voter interest is comparatively light. Each of the above specific hypotheses will be examined individually in more detail in Chapters Three through Six. Before concluding our discussion of the major hypothesis and the four specific hypotheses derived from it, we wish to add this glossary defining terms used in the minor or spe- cific hypotheses: Election procedures -- The legal, customary and actual practices involved in a contest for office, in which the voters act as decision-makers. Organization ei the two mejor parties -- The structure of the basic groupings for political action and nomination of candidates, the Republican and Democratic parties, which in- cludes the positions of the formal organization and the candi- dates it is running for Office. Incumbency and non-incumbency factors -- The factor of whether a candidate is holding office or is out of office, as applied to each of the contests covered in the newspapers in our sample. Incumbency relates to the party strength in the districts on the basic premise that an incumbent, by virtue. of his holding office, adds to the party's vote-pulling power. 29 Candidate campaign activity -- That behavior in which contestants for office engage in the formal and informal ef- fort to win nomination or election. As applied in this study, their goal is election in a contest between nominees of the parties. Organizational efforts -- Activities of candidates de- signed to put together committees and other groups of cam- paign workers in an effort to mount an effective campaign. Public relations -- Efforts made on behalf of candidates and their parties to reach the voting citizens, primarily through the mass media but also through direct contact meth- Ods. For our study, we are concerned primarily with the use of the media for this purpose. Competitive district -- This is a constituency in which there is a high degree of competition between the parties. For our study, a district is competitive if the average for all elections during the period 1952-62, when districting re- mained the same in both states, reflected the winning party's candidate carrying the district with less than 57 per cent of the vote. Non-competitive district -- A constituency which may be described as either one-party or one-party predominant (see definitions below). One-party district -- This is a constituency in which the same party holds the seat for the entire 1952-62 period, regularly winning without serious Opposition from the minor- ity party. In Michigan and Tennessee during this period, we 30 expect the average percentage of the winning party's vote to exceed 65 per cent for it to fit into this category. While it is specifically excluded from our definition, we expect the historical pattern Of voting in the district can be traced back to the last critical election, and in the case of some Tennessee districts, back as far as the Reconstruction. One-party predominant district -- This is a district with an average percentage of at least 57 per cent and not more than 65 per cent for the winning party during the 1952-62 period in our two states. Reportage -- Coverage of campaign events in the regular news columns of the daily newspapers of the district being covered. A detailed analysis of the congressional districts and of the newspapers included in the sample will be found in Chapter Two. V. Time Period Chosen for the Sample The time period covered in our analysis is that period from September 1, 1960, through the day Of the presidential- congressional general election, November 8, 1960. The pri- mary coverage was not included in the sample in the case of Tennessee, although this would have been useful had it been feasible to do so. But the primary coverage was beyond the scOpe of our project. The Detroit News and Free Press, originally considered for the Michigan portion of the sample, were deliberately ex- cluded because they were considered qualitatively different 31 from the smaller Michigan newspapers, a distinction that was not so apparent in comparing the Memphis Commercial Appeal with the smaller Tennessee papers. We felt that our selection of the Michigan newspapers would give us, outside the Detroit metropolitan area, areas of non-competitive Republicanism (one-party in the Fifth Dis- trict, and one-party predominant in the Ninth District) and a close competitive situation (in the Sixth District, especial- ly in terms Of voting trends in the 1960's). In the case of the Tennessee newspapers, the Memphis area (Shelby County comprises the Ninth District) is an ob- vious area of interest. This is because of the interesting voting pattern in the Ninth Tennessee congressional district balloting.50 For the period from 1952 through 1960, Rep. Clifford Davis, the Democratic incumbent during this period, won without Republican Opposition in 1958 and 1960. Davis polled 71.7 per cent to his Republican Opponent's 28.3 per cent in 1956, received 83.5 per cent to his Republican Oppo- nent's 16.5 per cent in 1954, and in 1952 polled 85.7 per cent to 14.3 per cent for his Independent Opponent. Thus the district has all the earmarks of a safe "one-party" district. Yet in 1962,51 he had 50.6 per cent to 49.4 per cent for his Republican Opponent, and in 196452 when Davis was defeated in the primary, the successful Democratic nominee, George W. Grider, polled 52.5 per cent of the total vote to 47.2 per cent for his Republican Opponent and 0.3 for an Independent candidate. Thus we have the interesting question of whether 32 a "safe" Democratic district all through the 1950's was trans- formed into a competitive district as a result of the reac- tion to the Ole Miss crisis in 1962 or whether this was an aberration. The 1964 results in the district merely show the district fell within the boundaries of competitiveness that year, but the possibility Of an important shift makes the Ninth Tennessee CD one of particular interest to us. The Fifth Tennessee CD (which includes metrOpolitan Nashville, identical with Davidson County) is of interest as a traditional one-party Democratic area, and also as an area where Negro and labor votes are of greater importance than in rural Tennessee. This is also true to some extent of Memphis as well, where according to one SNCC (Student Non-Violent Co-ordinating Committee, a civil rights group) source in Knoxville, there is considerable cohesion in Negro voting and pressure-group activity. By examining Knoxville and Chattanooga papers, we should be able to gain more information about erosion of Republican strength in areas of traditional one-party Republicanism (such as the Second District) and Republican inroads into Democratic strength in traditional areas of Democratic domi- nance (such as the Third Tennessee CD, where Chattanooga candy manufacturer William E. Brock III was elected to Con- gress in 1962 and remained in office in the 1964 election on a wave Of pro-Goldwater sentiment). Although we will discuss the congressional districts in detail in Chapter Two, we will mention here that in terms of 33 the percentage definitions of competitiveness stated on Pages 29 and 30, the First Tennessee CD (using the 1952-62 averages) is one-party predominant Republican, the Second Tennessee CD is one-party Republican, and the Third, Fourth, Fifth, Sixth, Seventh, Eighth and Ninth CD's are all one- party Democratic. In demographic terms, to restate the Tennessee categor- ies, our collapsing of the 93 classification (described in detail in Chapter Two) into a two-fold system gives us the First, Second, Fourth, Sixth, Seventh and Eighth CD's as rural, and the Third, Fifth and Ninth as urban. VI. Compilation of Raw Data TO compile the data for content analysis of our 12-news- paper sample, we used a data sheet made up by two different methods -- one set up for quantitative analysis and the other set up for qualitative analysis. The quantitative data sheet included: (1) Name of newspaper and its circula- tion: (2) date: (3) page and section (where applicable, in the case Of section) for each story; (4) position on page (above or below fold) and column position: (5) identification of story (by headline); (6) head size (in column width, not points of typd; (7) column inches devoted to story; (8) type of campaign event described (press release, rally, etc.), and (9) issues and types Of issues mentioned. The qualitative data sheet included: (1) Name Of newspaper; (2) date; (3) political figure mentioned: (4) wording Of the mention, and (5) whether the reference was considered favorable, 34 unfavorable or neutral. Amount of raw data recorded con- sisted of approximately 500 sheets of quantitative and 200 sheets of qualitative data. These data, in the refined form tabulated into summary tables, we used to test the four spe- cific hypotheses for validation. In selecting the newspaper sample for content analysis, we reduced the work by taking our data from each fourth day's newspaper, plus Sunday editions where these were published by the newspaper sampled. Since two newspapers in each state had no Sunday edition, this kept the sample balanced, and the Sunday data gave us a sampling of the type of "in-depth" cov- erage not found in weekday editions. The only alternative to this methodology would have been to restrict the number of newspapers analyzed, or else choose a less representative sample. VII. The Operational Design Once our data were in hand in the raw form, we began testing each specific hypothesis with the data apprOpriate to it. We did this by using an Operational design for each of the four specific hypotheses. The design appears below. Specific Hypothesis 1 states: "Biased coverage (1. e., "slanted" coverage, or coverage tending to favor one candi- date Or party at the expense of the other) of congressional campaigns is more likely to be found in rural-district papers than in urban-district papers." Data employed to validate this include data from cate- gorization Of content into favorable, unfavorable and neutral U ‘V 35 for newspapers published in rural CD's and the same cate- gorized data for newspapers published in urban CD's. For close-up analysis Of segments of the sample, tables compar- ing one newspaper in each category with another in the Op- posite category have been constructed. In constructing our sample, we ruled out metrOpolitan Detroit newspapers and have no cities larger than Memphis rep- resented. With this in mind, in analyzing different news- papers published in the same city for structural factors such as newspaper policy, these factors will cause one of the two newspapers to be more heavily biased than the other. For example, the Nashville Banner may be more so than the Tennessean, or vice versa, or the Knoxville Journal more than the News-Sentinel, or vice versa. (In the Knoxville case, chain control of editorial policy may be a distinguishing factor, and in the Nashville case, long-standing partisan commitments may be a factor.) Each newspaper will be guilty of slanting in some instances, but the “FUN Rating" numer- ical representation of instances of bias, explained fully in Chapter Three) will give us a glue to degree Of bias. Specific Hypothesis 2 calls for separate treatment of parts (a) and (b), which suggest two specific cases of the specific hypothesis that “total coverage of candidates will be greater in newspapers published in rural districts than in those published in urban districts”. Specific Hypothesis 2(a) states: "Both major-party candidates get more total coverage (in terms Of measurement 'm . («TS-war £9. ,,*" " fr 31.3- m. .~—.—— -* firm-ream" a, g “I, “VII-T34.“ ’(V 36 of actual newspaper space and not potential eXposure) in rural than in urban districts." To test this, we compiled data on total coverage for both candidates in newspapers published in urban CD's, with the expectation that data on one newspaper as a segment of each sample would pinpoint the analysis. We also gathered parallel data on total coverage for both candidates in rural CD's, with detailed attention to one newspaper. These data were expressed in terms of a frequency count of items, col- umn inches, and percentage of newspage content. (The news- page content for each newspaper was checked out to be sure the newspapers were suitable for cross-comparison across news- papers, and not overloaded with advertising in the case of any one of the papers.) Specific Hypothesis 2(b) states: "Non-incumbents get more coverage in newspapers published in rural congressional districts than in newspapers published in urban districts." To test this, we compiled data in the form of a breakdown for the same newspapers used to test Specific Hypothesis 2(a), testing this time for the incumbency, non-incumbency factor. A complete list of incumbents running in 1960 and 1962 is given in Chapter Two. Specific Hypothesis 3, dealing with type of events cov- ered, states: “Noticeable differences will occur in cover- age -- in terms of space allotted, news play according to position on the page and page number, and emphasis according to headline size -- for different types of campaign events, 37 according to (a) whether a party "VIP“ was present at the event; (b) whether events were nationally oriented or dis- trict oriented; (0) whether events were televised or involved direct personal appearances of candidates and other campaign figures, and (d) whether events were 'printed' or 'Oral' in nature." Data were taken for column inches, position, page and head size for each type of campaign event, but only the col- umn inches were used when we grouped the data according to these differentiations: (a) VIP-attended events vs. non-VIP- attended events; (b) nationally-oriented vs. district-orientem (c) televised vs. direct personal appearances, and (d) 'printed' vs. 'oral' events. Specific Hypothesis 4 states: “In establishing cate- gories Of congressional districts according to degree of party competition, we expect to find greater campaign coverage (in terms of total column inches) in newspapers published in high- ly competitive districts than in non-competitive (one-party or one-party predominant) districts." Variables Of party com- petition were compiled in a set Of tables to examine this prob- lem. Data needed to validate Specific Hypothesis 4 include total column inches Of coverage for all competitive CD news- papers, with a sub-total for one newspaper for close analysis, and total column inches for all non-competitive CD newspapers, with a sub-total for one newspaper for close analysis. This will be expressed as well in terms Of percentage of political 38 coverage for the grouping and for single newspapers chosen from the grouping. VIII. Summary Statement Once again, we wish to restate the purpose of our re- search project. Our aim is to deveIOp more systematic infor- mation about political communication by exploring the rela- tionship between communications variables on the one hand and political variables on the other. The political variables we are concerned with have to do with electoral structure and electoral behavior in con- gressional districts. The communications variables have to do with Press structure, principally editorial policy and news--I handling policy. Exploration Of the specific hypotheses in Chapters Three through Six, when done against the background of Chapter Two in which the districts and the newspapers are discussed, will lead to some useful conclusions which are presented in Chapter Seven, the final chapter of the thesis. )7 fi., _-. 39 Footnotes for Chapter One 1. These are found in N. W. A er & Son's Director , Newspapers ppg Periodicals, 1964 and 1962 editions. The Ed tor ppg Publisher Yearbook, published annually, is an- other source Of this information. 2. Austin Ranney, gpp Doctrine of Res onsible Part Government, Urbana, Ill., university 3?‘ no 3 Press llini Books paperback), 1962 (revision of 1954 edition). 3. V. 0. Key Jr., Partipp, Politics ppp Pressure Gropps (Fifth Edition), New York, Thomas Y. Crowell Co., 1964. 4. Hugh A. Bone, American Politics ppp the Party System (Third Edition), New York, McGraw-Hill, Inc.,_fi65. 5. William Goodman, Tpp TppfiParty S stem 3p ppp united, States, Princeton, N. J., Van Nostrand, . . Austin Ranney and Wilmoore Kendall, emocrac¥ and ppp American Part S stem, New York, Harcourt—Brace. 9337 7. Avery eiserson, Parties app Politics: Ag Insti- tutional Egg Behavioral A roach, New York, Knopf, 1953. 8. V. 0. Key Jr., Public Opinion pp; American Democ- racy, New York, Knopf, 19 1. 9. Edward C. Banfield, Political Influence, New York, Free Press Of Glencoe, 1961. 10. James Q. Wilson, Tpp Amateur Democrats, Chicago, university of Chicago Press, 19 2. _ 11. Joseph A. Schlesinger, "Political Party Organiza- tion", in James Gardner March, ed., Handbook p; Or anizations, Chicago, Rand-McNally, 1965. 12. Donald E. Stokes, "Spatial Models Of Party Com- petition", American Political Science Review, Vol. LVII, NO. 2, June 1963' . pp." 36W?”— 13. William N. McPhee and William A. Glaser, eds., Public Opinion ppg CO ressional Elections, New York, Free Press of Glencoe, 1962. 14. W. N. McPhee and Jack Ferguson, "Political Immu- nization", in McPhee and Glaser, pp. pip., pp. 155-179. 15. Phillip H. Ennis, "The Contextual Dimension in Voting", in McPhee and Glaser, pp. p$§.. pp. 180-211. 16. William A. Glaser and Charles Kadushin, “Political Behavior in Mid-Term Elections", in McPhee and Glaser, pp. p13., pp. 251-258. 17. Eugene Burdick and Robert Brodbeck, eds., American Votin Behavior, Glencoe, Ill., Free Press, 1959. 8. Kurt and Gladys Engel Lang, "The Mass Media and Voting", in Burdick and Brodbeck, pp. pip., pp. 217-235. 19. V. 0. Key Jr., Public Opinion ppg American Democ- racy, New York, Knopf, 19 1. 20. See Footnote 22 and Bernard Berelson, et al., Votipg, Chicago, university Of Chicago Press, 1952. 21. Paul F. Lazarsfeld, et al., 1gp PeO le's Choice: How the Voter Makes gp gig Mind (Second Edition), New York, CEIuEETa University Press, 1948. 1. (I (I) "MI: I!“ NJ‘O\.,.HU. r I I . . e I O C C I e O I I I O I o O O l O o O I O n . w e e . C a i O o e O o I l O . a e a a o a I o n a s u . U o e o . e O u g . . g s o p C a C e 0 e 0 e I v e e I e e e p a o v o v e a e 0 e u e I I 0 e s e O n . e v 0 e e O I r n e . a s o e a o e e O 4o 22. Elihu Katz and Paul F. Lazarsfeld, Personal Ip- fluence, Glencoe, Ill., Free Press, 1955. 23. Melvin L. DeFleur, "The Emergence and Functioning of Opinion Leadership“, in Decisions, Values and Groups, New York, Pergamon Press, 1962. 24. Charles Winick, “The Diffusion of an Innovation Among Physicians in a Large City“, Sociometry, Vol. 24, No. 4, December 1961, pp. 384-396. 25. Herbert Menzel and Elihu Katz, “Comment on Charles Winick, 'The Diffusion of an Innovation Among Physicians in a Large City'“, Sociometpy, Vol. 26, NO. 2, June 1963. PP. 125-127 e 26. Clarence Schettler, Public Opinion and American Societ , New York, Harper & Brothers, 1960. 27. Katz and Lazarsfeld, pp. pip. 28. Daniel Katz et al., eds., Public Opinion and PrOpaganda, New York, Dryden Press, 1956. 9. Karl W. Deutsch, The Nerves pf Government, New York, Free Press Of Glencoe, 1963. p 30. Joseph T. Klapper, The Effects p: Mass Communica- tion, Glencoe, 111., Free Press, 1960. 31. N. W. Ayer p Son's Directory, Newspapers and Periodicals, Philadelphia. (Published annually) 32. Editor and Publisher Yearbook, New York. (pub- 1ished annually by American Society of Newspaper Editors) 33. Wilbur Schramm, ed., The Process and Effects of Mass Communication, Urbana, University of Illinois Press, 1958. 34. Congressional District Data Book (Districts of the 88th Congress), Washington, Government Printing Office, 1963. (Edward D. Goldfield, chief, Statistical Reports Division, U. S. Bureau of the Census, Department of Commerce, ed.) 35. Congressional Quarterly Census Analysis, Washing- ton, Congressional Quarterly Service, 1964. 36. Ithiel de Sola Pool, Trends ip Content Analysis (reviewed in Spring 1961 Public Opinion Quarterly). 37. Bernard Berelson, "Content Analysis", article in Vol. I, Handbook p; Social Psychology, Gardner Lindsey, ed., Addison-Wesley, 1954. 38. Robert C. North, Holsti, Zaninovich and Zinnee, Content Analysis: p Handbook With Applications for the Stpdy pf International Crisis, Evanston, Ill., Northwestern University Press, 1963. 39. Paul F. Lazarsfeld, et al., Tpp Peo le's Choice (1944); Bernard Berelson, et al., Voting (195%): Campbell, Converse, Miller and Gurin, The Voter Decides (1954), and Campbell, Stokes, et al., The American Voter (1960; abridged edition, 1964). 41 40. Warren Miller, "Presidential Coattails: A Study in Political Myth and Methodology“, gpp, Winter 1955-56, Vol. XIX, pp. 353-368. Malcolm Moos, Politics, Presidents and Coattails, Baltimore, Johns HOpkins Press, 1952. Louis H. Bean, Hp! pp Predict Elections, New York, KnOpf, 1948, Chapters 2 and 4, pp. 12-21, pp. 31-36. Charles Press, “Presidential Coattails", Midwest Journal pg Political Science, Vol. VII, No. 4, November 1963, pp. 320-335. 1. Joseph A. Schlesinger, "A Two-Dimensional Scheme for Classifying States According to Degree of Inter-Party Compe- tition", APSR, Vol. XLIX, December 1955. PP. 1120-1128. 42. Phillips Cutright and Peter H. Rossi, "Party Organi- zation in Primary Elections", American Journal pg Sociology, Vol. XLIX, November 1958, pp. 262-269. 43. Warren Miller and Donald E. Stokes, "Constituency Inflzencg in Congress", APSR, Vol. LVII, NO. 1, March 1963, PPO 5'5 0 44. Charles L. Clapp, Tpp Congressman: His Work pp Hp Sees 2p, Washington, Brookings Institution, 1963: See especially Chapter 8, "The Problem of Being Returned", pp. 330‘392 e 45. Cortez A. M. Ewing, Primapy Elections ip the South, Norman, Okla., University Of Oklahoma Press, 1953. (Quota- tion from pp. 105, 106 cited in Prospectus) 46. Lewis A. Froman Jr., Congressmen and Their Consti- tuencies, Chicago, Rand-McNally, 1963. 7. LeRoy C. Ferguson and Ralph H. Smuckler, Politics in the Press, East Lansing, Michigan State University Press, I954. 48. V. 0. Key Jr., Parties, Politics and Pressure Grou 3 (Fifth Edition), New York, Crowell, 1964. See espe- cially Chapter 16, pp. 475-497. and Chapter 20, pp. 591-621. 49. Congressional Quarterly Special Reports: Report on Voting in 1962 Congressional Elections and Report on Re- apportionment, Washington, Congressional Quarterly Service, 19 3. 50. Figures based on congressional voting statistics found in America Votes for 1960 and in Congressional Quarter- ly Special Report: Complete Returns Of the 1962 Election by Congressional District. ‘51. Congressional Quarterly Census Analysis, pp. pip., see Footnote 35. he ._. CHAPTER TWO Congressional Districts and Newspapers: Background for the Study I. Purpose of the Chapter Our starting point in this analysis is the description of Michigan and Tennessee congressional districts, since the purpose of the content analysis will be to learn more about the Press in these districts, and the relation between the Press and the politics Of the districts. The chapter will take up first the background for the Michigan districts, then the background for the Tennessee districts, before making a comparision of Michigan and Tennes- see politics in terms of the degree Of competitiveness or one- partyism found in them. 42 43 After a discussion of the political context for con- gressional campaigning in the two states, we will turn to the background for the Kichigan newspapers, followed by a discus- sion of the background for the Tennessee newspapers. For each newspaper, its general policy, pattern of endorsements, news- handling policies and ownership will be discussed. Before leaving the tOpic of newspapers, we will compare kichigan papers (representing a basically competitive state) with Tennessee papers (representing a basically one-party state). We will also discuss the implications of chain ownership of newspapers (e.g., Scripps-Howard as Opposed to independently owned dailies) for the dynamics of the relationship between congressional districts and performances of the Press in those districts. II. Analysis of Congressional Districts In setting Up the district analysis, we shall focus on 18 Michigan districts and nine Tennessee districts. We shall examine 18 of the 19 Michigan districts that were in existence prior to the redistricting of 1964. These 18 districts had remained unchanged since 1914, when the lines were drawn which were in effect during the 1960 campaign. Our analysis in no way deals with the at-large district known as the 19th, which existed during the 88th Congress in 1963 and 1964 and to which Representative Neil Staebler of Ann Arbor, a Democrat, was elected. This at-large district was abolished in 1964 when present lines were drawn. 44 In Tennessee, we shall focus on all nine districts, which were established in 1952 after Tennessee's share of congressional seats was reduced from 10 to nine. In most Tennessee districts in the past, the real decision in the congressional contest has occurred in the primary, but this was not the case in the 1962 Brock-Thrasher contest in the Third Tennessee CD, when a factional division among Democrats was given credit for contributing to the election of the Republican, Representative Brock. This was obviously an instance in which the real decision occurred in the November election, even though defection growing out of the Democratic primary figured in the outcome. It is these 27 districts in the two states that we are examining for the period embracing biennial congressional elections dating from 1952 through 1962. We have gathered aggregate voting data and precentages for the two-party vote in each of the 27 districts for this period, essentially a decade of electoral contests during which district lines were not changed, making comparison possible.1 Now we are examining the voting data along with such descriptive demographic data as the urban-rural dimension (with some attention to surburban factors) and other socio- economic characteristics. We are also interested in the incumbency factor in the districts and will include a list of incumbents who ran for office in the two states in 1960 and 1962. 45 We list below in tabular form percentages of the major- party vote in each district over the 1952—62 decade, and individual year percentages for 1960 and 1962, the two years we are mainly concerned with (primarily 1960) in dealing with the newspage content data. In content analysis, we decided to choose a small group Of districts for intensely focused study because of the party situation in the districts and the availability,of microfilm files of daily newspapers published in the districts, and we are concentrating our analysis on these selected districts. III. Background for the Michigan Districts Average major-party percentage for the 18 Michigan districts for the 1952-62 period, as computed from the per- centages for these six elections individually, appears below in tabular form. Listed in sequence by districts, the per- » centages are listed to the tenth of one per cent, with Demo- cratic percentages in the first column and Republican per- centages in the second column. The analysis of Michigan districts for 1952-62 follows: District First Michigan Second Michigan Third Michigan Fourth Michigan Fifth Nichigan Sixth Michigan Seventh Nichigan Eighth Michigan Ninth Michigan Tenth Hichigan Eleventh Michigan Twelfth Michigan Thirteenth Michigan Fourteenth Michigan Fifteenth Michigan Sixteenth Michigan Seventeenth Michigan Eighteenth Michigan The Nineteenth Michigan District, an at-large district, was excluded because it was abolished in the redistricting of 1964. If we now focus our attention on the two-party percent- ages for the final two elections of this period, we will find 46 Per 88.0 39.3 38.8 37.1 34.2 47.4 48.9 36.8 42.1 36.8 44.3 41.4 69.4 59.5 75.8 66.6 55.0 44.1 Cent Democratic Per Cent Republican 12.0 60.7 61.2 62.9 65.8 52.6 51.1 63e2 57.9 63.2 55.7 58.6 that some of the out-state Michigan districts appear more closely competitive in 1960 and 1962 than in earlier years, as reflected in the averages above for the 1952-62 period. Here are percentage figures for the two-party vote in Michigan in 1960 and in 1962: (4" District 1960 1960 1962 1962 Pet. Pct. Pct. Pct. Dem. Rep. Dem. Rep. First Michigan 88.7 11.3 89.2 10.8 Second Michigan 40.3 59.7 41.6 58.4 Third Richigan 39.3 60.7 40.5 59.5 Fourth Michigan 37.6 62.4 36.2 63.8 Fifth kichigan 33.1 66.9 33.0 67.0 Sixth Lichlgan 43.3 56.7 45.5 54.5 Seventh Michigan 53.4 46.6 56.3 43.7 Eighth Richigan 37.7 52.3 39.5 50.5 Ninth Michigan 40.3 59.7 40.6 59.4 Tenth Hichigan 37.8 62.2 38.5 61.5 Eleventh Michigan 45.1 54.9 43.3 56.7 Twelfth hichigan 39.1 60.9 36.7 63.3 Thirteenth Michigan 71.7 28.3 71.2 28.8 Fourteenth Richigan 62.8 37.2 61.8 38.2 Fifteenth Michigan 79.6 20.4 83.0 17.0 Sixteenth hichigan 66.2 33.8 67.9 32.1 Seventeenth Michigan 57.7 42.3 59.3 40.7 Eighteenth kichigan 44.1 55.9 40.4 59.6 These sets of percentages for the entire 1952-62 decade and for the two election years at the conclusion Of this period (1960 and 1962) give us a broad idea of the political context of the Michigan districts we are examining. Before turning to the Tennessee districts, we wish to note the rural—urban-suburban characteristics of each Michigan district. We will also explain how we categorized rural and urban constituencies for purposes Of our analysis. 2 Congressional Quarterlyis Census Data Book refers to these Michigan districts as urban: The First, Thirteenth, Fourteenth, Fifteenth and Seventeenth. Rural Kichigan dis- tricts are the Second, Third, Fourth, Eighth, Ninth, Tenth, Eleventh and Twelfth. Suburban districts are the Seventh, Sixteenth and Eighteenth (the latter in Oakland and Macomb Counties). Two districts, the Fifth (Grand Rapids and Kent N‘. n... C .16 Q‘s N\VN.U \I‘ A . 48 County) and the Sixth (Ingham, Livingston and Genesee Counties) are described as mixed urban-rural-suburban. In order to set up a two-fold classification, for pur- poses of our constituency analysis, we will describe as rural districts the Second, Third, Fourth, Eighth, Ninth, Tenth, Eleventh and Twelfth. Constituencies of mixed nature, which are urban at least in part, we will describe as urban districts. Any district described as suburban we are including in the urban category. Thus, along with the First, Thirteenth, Fourteenth, Fifteenth and Seventeenth, we will classify as urban the Fifth, Sixth, Seventh, Sixteenth and Eighteenth CD's. Since we have already made a general comparison of Michi- gan and Tennessee politics, and we have outlined the political context of the districts for Hichigan, we will reserve any further comments on the specific districts for later chapters and will move to an analysis of the political context of the Tennessee districts in the next section. IV. Background for the Tennessee Districts Average major-party percentages for the nine Tennessee districts for the 1952-62 period, as computed from percentages for these six elections, will be found below: District Per Cent Per Cent Democratic Republican First Tennessee 35.3 64.7 Second Tennessee 22.4 77.6 Third Tennessee 72.0 28.0 Fourth Tennessee 100.0 0,0 Fifth Tennessee . 86.6 13.4 Sixth Tennessee 100.0 0.0 Seventh Tennessee 100.0 0.0 Eighth Tennessee 100.0 0.0 Ninth Tennessee 84.} 15,7 49 Some interesting recent tendencies show up when we examine the figures for 1960 and 1962, especially some trends toward a more competitive politics in Tennessee, which like most of the South has had one-party Democratic bastions for many years (with pockets of one-party Republicanism as well). Three Tennessee districts are of Special interest for indications of tendencies toward two-partyism. It would be an overstatement to say that a long-term trend toward two- partyism is indicated by the results of only a few elections, but there is some evidence of tighter interparty competition, especially in 1962 voting. At any rate, the 1962 congressional voting pattern placed the First Tennessee District within the range of what we have defined as competitive, with the Democratic percentage reaching 44.9 there. In the Third Tennessee CD, traditionally Democratic, a Republican candidate won and was re-elected in 1964. In the Ninth Tennessee CD, also traditionally one-party Democratic, the 1962 Republican candidate, Robert James, polled 49.4 per cent of the two-party vote. James, the GOP candidate in Memphis again in 1964, did less well that year because of the presence of an Independent aspirant in the race and because a younger, more moderate Democratic nominee with more advanced civil rights views pulled out the Crump-registered Negro vote in Nenphis much better than did Davis in 1962. Tennessee district party division percentages for 1960 and 1962: District 1960 1960 1962 1962 Pct. Pct. Pct. Pct. Dem. Rep. Dem. Rep. First Tennessee 24.6 75.4 44.9 55.1 Second Tennessee 0.0 100.0 29.4 70.6 Third Tennessee 100.0 0.0 48.9 51.1 Fourth Tennessee 100.0 0.0 100.0 0.0 Fifth Tennessee 100.0 0.0 * * Sixth Tennessee 100.0 0.0 100.0 0.0 Seventh Tennessee 100.0 0.0 100.0 0.0 Eighth Tennessee 100.0 0.0 100.0 0.0 Ninth Tennessee 100.0 0.0 50.6 49.4 *Because of contested Democratic primary, both Fifth District candidates ran as Independents in November, so there were no party percentages. According to the gg Census Data Book} only one district, the Ninth (Memphis-Shelby County) was urban. The First, Second, Fourth, Sixth, Seventh and Eighth are categorized as rural. The Third (Chattanooga) District is described as mixed urban-rural, and the Fifth (Nashville-Davidson County) is termed mixed suburban-urban. In setting Up a two-fold classification, we will describe as rural districts the First, Second, Fourth, Sixth, Seventh and Eighth CD's. Those falling in the urban category are th Third, Fifth and Ninth CD's. V. Comparison of Richigan and Tennessee Politics In a general comparison of the politics of hichigan and Tennessee, considered on a state-wide basis, we would have to consider Tennessee a one-party state in which some areas of two-party competition are beginning to develOp, and Hichigan a two-party state where interparty competition occurs in varying degrees of intensity. 51 Both traditionally one-party Republican districts in Tennessee, the First and Second, are moving in the direction of competitiveness as the 1962 figures indicate, but the Second is still the most solidly Republican. The Third, traditionally one-party Democratic, must now be considered competitive in the light of recent elections, as must the Ninth. Apparently, how- ever, the Fourth, Fifth, Sixth, Seventh and Eighth have pri- marily one-party characteristics even yet. Despite our description of Michigan as a two-party state, it does contain a number of congressional districts with one- party characteristics. If we consider voting percentages through the 1962 election, the First and Fifteenth remain most heavily one-party Democratic, while the Democratic party has dominant but lessening strength in the Thirteenth, Fourteenth, Sixteenth and Seventeenth Districts. Most strongly one-party Republican districts are the Fourth, Fifth, Eighth, Tenth and Twelfth Districts. The Sixth, Seventh and Eleventh may now be con- sidered competitive districts. One-party Republicanism appears to be losing strength in the Second, Third, Ninth and Eighteenth Districts. From the above it can be seen that while Michigan taken as a state-wide whole appears to be an area of two-party com- petition, when broken down geographically into congressional districts, only about three of the eighteen districts analyzed are truly competitive, while four districts each are one-party predominant Democratic and one-party predominant Republican. Of the remaining seven districts, five are one-party Republi- 52 can and two are one-party Democratic, according to our defini- tion of degrees of competition. Not surprisingly the one-party Democratic districts are to be found in Wayne County and the one-party Republican districts are to be found outstate, mainly but not entirely in more rural areas of the state. host inter— esting perhaps are the "swing" districts, the Sixth, Seventh and Eleventh, which are becoming more competitive and in the case of the first two contain a large suburban vote and are in areas where urbanization and industrialization are increasing rapidly. VI. Background for the Michigan Newspapers In this section we shall discuss the policies, candidate endorsements, newshandling and ownership of the six Kichigan newspapers chosen for our sample. Newspapers which we chose include the Grand Rapids Press, Lansing-East Lansing State Journal, Flint Journal, Saginaw Laws, Traverse City Record—Eagle, and Muskegon Chronicle. A bit further on in this section of the chapter, we will present a table with data about each of these papers and discuss each individually. The daily edition of a newspaper is the end product of a publication process in which the characteristics of the news- paper, its staff, owners and the raw material of daily events all interact to produce a certain kind of newspaper. If we assume that the coverage of political events does not get into print willy-nilly but is the end product of a process of this kind, it becomes important for us to know something about the characteristics of that newspaper. These are the input factors ‘Vv 1“ 's HA 53 that result in the output which is newspaper content. Thus in analyzing structure of the Press as an institu— tion, we must know the individual paper's circulation, how long the newspaper has been in existence, its editor and publisher and, as a starting point for analysis, its policy stance (i.e., editorial policy) as well as newshandling policy. It would be interesting to amass all these data for each newspaper of all descriptions (weeklies, ethnic papers, suburban papers, etc.) in both states. But we are limited by the practi- cal consideration that microfilm has come into general use only recently, and only a limited number of newspapers (mainly dailies) can be analyzed by examining their microfilm files. This led to two basic decisions in compiling data on owner- ship, policy and circulation. The first was not to use weekly newspapers. Use of the daily newspaper sample made it possible, too, to observe chain newspaper behavior in editorial policy- making and newshandling practices (Booth Newspapers, Inc., in Michigan, Scripps-Howard in Tennessee in the cases of kemphis and Knoxville). Also time and funds were not available to undertake a weekly newspaper analysis, even where files are conveniently kept. The same was true of the suburban press. It was decided to gather these data for newspapers of general circulation which are dailies published in urban and rural constituencies as we have described them in Section IV. Focus of our attention was further limited by the number of these available on microfilm. A second decision was that made to concentrate detailed ‘l g"... _., fl 54 analysis on a small group of newspapers, six in each state. An original plan to "Spot check" other newspapers besides these 12 was discarded when the l2-newspaper sample proved to be a fairly representative sample in terns of the daily Press of each state. Now we present data in this section for the Michigan news- papers (and in the next section for the Tennessee papers) that we used in our l2—newspaper microfilm sample}+ This appears below in tabular form. Structural descriptive data for the (six hichigan papers5 follows: 1964 1962 Newspaper Circulation Circulation Established Grand Rapids Press D, 125,915 D, 126,547 1890 S, 101,805 8, 91,407 Lansing-East Lansing D, 70,182 D, 68,447 1855 State Journal S, 69.552 S, 67,709 Flint Journal D, 101,023 D, 96,701 1875 S, 100,920 S, 96,510 Saginaw News D, 53,521 D, 52,499 1859 3, 53,355 3, 58,798 Traverse City D, 13,688 D, 13 733 1858 Record-Eagle (No Sunday Edition) Euskegon Chronicle D, 46,391 D, 45 293 1857 (Ho Sunday Edition) Editor and Nominal Newspaper Publisher Policy Grand Rapids Press Lee E. Woodruff Indebendent Booth Newspapers ‘ Lansing-East Lansing Louis A. Well, Jr, Independent State Journal Republican Flint {QEEEEL Ralph E. CUFPY Independent Booth Newspapers Saginaw EEEE Glen A. Boissoneault Independent Booth Newspapers Traverse City A. T. Batdorf Independent Record-Eagle Record-Eagle Co. ‘ huskegon Chronicle Robert C. Herrick Independent Booth Newspapers 55 The noticeable tendency of newspapers to call themselves Independent, despite the true persuasion stated in their edito- rials, cannot be overlooked. For this reason, in order to prOperly examine the policy of the six papers we must look at their true policy preferences through editorial endorsements. The only truly Independent or non-partisan paper in the Richi- gan sample proved to be the Record-Eagle in Traverse City, which did not endorse either party's candidates and made a strenuous effort to remain truly non-partisan. The Grand Rapids Press, Lansing State Journal, Flint Journal, Saginaw News and Ruskegon Chronicle all followed a general policy line of Republicanism, endorsing the Republican ticket and for the most part taking the Republican line on major campaign issues. Each of these five newspapers endorsed Vice President Richard E. Nixon and the Republican congressional nominee in their home district in the 1960 campaign. The best indicator of degree of partisanship is probably the newshandling policies of the papers, as evidenced in con- tent which we examined (just as the best indicator of party preference is the endorsements of candidates). Judging from our impression of this factor (which will be discussed in more detail in Chapter Three when we deal with bias in the Press), the most strongly partisan Republican of the five papers were the Saginaw Eggs (notably in handling of Letters to the Editor) and the Grand Rapids Press. Judging from newshandling, the Lansing State Journal and Huskegon Chronicle appeared to be somewhat less partisan in their Republicanism, i.e., they gave 56 a more balanced coverage to Republican and Democratic candi- dates. The least partisan reporting of the congressional cam- paign was probably that found in the Flint Journal. Whether this was because the Journal appeared slightly more moderate than the other four Republican papers in Michigan, or whether it was because the Sixth Hichigan CD congressional nominee, Prosecutor Jerome F. O'Rourke of Flint, was a local man, we could not determine, but perhaps both factors entered in. Our general impression was that five of the six Richigan papers in the sample were actually Republican, that some were more strongly partisan than others but that even the most strongly partisan gave reasonably fair coverage to Opposition candidates on the newspages. VII. Background for the Tennessee Newspapers Turning to the six Tennessee newspapers in the sample, we found the following basic descriptive data:6 1964 ° 1962 Newspaper Circulation Circulation Established Chattanooga Times D, 52,864 D, 52,408 1869 3. 84.583 S, 81.797 Knoxville Journal D, 65,278 D, 66,956 1859 (Sold (No Sunday Edition) to Scrinps-Howard Knoxville News- D, 105,517 ., 104.997 interests) Sentinel S, 145,717 3, 141,507 1885 Iemphis Commercial D, 216,760 D, 221,591 1940 Appeal 3, 261,545 S, 259,472 Nashville D, 131,797 D, 131,339 1812 Tennessean S, 212,037 3, 207,101 Hashville Banner D, 95,965 D, 104,457 1976 (No Sunday Edition) 57 Editor and Noninal Newspaper Publisher Policy Chattanooga Times Iortin Ochs Independent Zen Fale Golden Democratic Knoxville Journal Guy L. Smith, JCP ex—state Independent 0': -ir rren Republican Jourml 1 Co. Knoxville News- Loye fl. Liller Independent Sentinel hers—Sentinel Fuhlishin; Co. (Scripps -Ho iqrd subsidiary) Eerphis Connercia 1 F. R. ahlgren Independent Appeal Lenphis Publishin3 Co. (3crinos-»o.nrd subsidiary) Nashville John 3ie3entha1er Democratic Tennessean Anon Carter EVSns Nashville Banner A. C. Dunkleberger Independent Aron Carter Evans For the Tennessee newspapers, it is also interesting to compare their professed policy with their true persuasion as revealed in editorial endorserents. Lenphis papers call them- selves Independent, though the Co: ercial Appeal supported Kixon in 1960. But it is only newspapers of deep-dyed persua- sion, lilte the two capita 1— —city ne i up pers, the Tennessean and the gtgtg Journal in Lansing, th: t h vs the nerve to call them- selves respectively Derocratic (unabashedly) and Independent Republican. The Ten.es seen, of course, was founded shortly before the erergence of Andrew ac‘rson as a political leader and has a long tradition of support for Denocratic candidates and causes. The ultraconservative Knoxville Journal also uses a party designation, terning itself Independent Rep‘blican, although it is actually Old Guard Republican in editorial policy, jud3in3 by its enjDrSEhGfitS of candidates. The Journal was so adara ntly reactionary on the civil rights issue during the early 1950's that it branded a Protestant chaplain, who had 58 made Space available for interracial meetings, as "having Red ties" because he once invited a "suspect" fo k-singer to perform at a student center serving University of Tennes- see students. The £23333; also publicized some of the more community-accepted right-wing groups, but buried on a back page an "advance" on a visit by General Edwin Walker of Ole Miss fame, apparently because of embarrassment over his repu- tation. General Walker immediately accused the {garnal of being too liberal when he made his appearance before a right- wing rally at the Knoxville Civic Auditorium in 1963. On the liberal side, the Tennessean was almost equally biased against Governor Frank G. Clement, the defender of integration at Clinton High School (reluctantly, it is true), although over another issue. The Tennessean attacked the governor because of his action against the TVA interests when he allowed the state sales tax to be applied to electric utili- ties power generated by the TVA. The Tennessean, liberal enough in the context of Tennessee politics, was a long—time advocate not only of TVA but also of its champion, Senator Estes Kefauver of Chattanooga, and since Kefauver's death in 1963 SUpported the Chattanoogan's successor in the Senate, former Rep. Ross Bass of Pulaski, against Governor Clement in the 1964 Senate primary in Tennessee, as well as in the 1964 general election. Compared with the virulent partisanship of the Knoxville Journal and the militant pro-Republican partisanship of the Nashville Banner (founded we note in 1975, the year of the — 33.37 V 59 contested Hayes-Tilden presidential contest, and still ada- mantly Republican), the Lansing State Journal -- which some of its Michigan readers describe as biased —— seems to verge on being an Eisenhower Republican publication of the respect- able type, employing care in newshandling although not free of distortion, as no newspaper can be because of the nature of headline writing which guarantees a certain amount of this. Since the Tennessee partisan identification of newspapers varies more than that found in Kichigan, we worked up the chart below to give an idea of party preference, endorsements and newshandling policies of the Tennessee papers in the sample. Our findings are as follows: Usual Party Newshandling Newspaper Preference Policy Chattanooga Times Democratic Balanced Knoxville Journal Republican Favors Republican candidates Knoxville Republican (endorses Balanced News-Sentinel some Democrats) Kemphis Democratic (prefers Favors conserva- Commercial Appeal conservatives) tives slightly Nashville Democratic Favors Democratic Tennessean _ candidates Nashville Republican Favors Republican Banner candidates . Endorsed for Endorsed for Newspaper President Congress Chattanooga llififi Kennedy Democrat Knoxville Journal Nixon Republican Knoxville Nixon Republican News-Sentinel ‘ hemphis Kixon Degocrat Commercial Appeal Nashville Kennedy Democrat lggggssean Nashville Nixon Republican (where bennga running) 60 The above should give a good indication of the real party preferences of the Tennessee papers in the sample, and taken together with similar information about the Michigan papers in the sample, should give us a very good idea of the charac- ter of the newspapers we are analyzing in both states, before we proceed tea more careful study of the content analysis data on specific hypotheses, beginning with Chapter Three. VIII. Chain Ownership Factors One aspect of the characteristics of the newspapers in our sample chosen from the Michigan and Tennessee daily Press that we found of interest, although it was not dealt with in our hypotheses, was the factor of chain ownership of news- papers. In Michigan, although we did not plan it that way, it deveIOped that four out of the six newspapers in the sample were a part of the Booth Newspapers, Inc., chain. In Tennes- see, we found thst two newspapers out of the six were part of the Scripps-Howard chain and that a third had been bought out by Scripps-Howard, even though it was permitted to retain its editorial autonomy. While there were variations in newspapers even within the same chain, we found some evidence of more responsible cover- age and reporting by newspapers that were part of a chain than by locally owned newspapers. This might be in part at least due to the greater reportorial resources of the chain newspapers, but it was something we had not eXpected to find. 61 This tepic will be discussed in greater detail in Chapter Three, which deals with bias in the Press. IX. Comparison of hichigan and Tennessee Newspapers A brief comment in comparing Michigan and Tennessee news- papers may be in order here, since we have just examined each set in the sample separately. We had wondered whether we mieht find some differences in degree of partisanship of the Press in one-party and two-party political contexts such as those in Tennessee and Michigan, but in general we found that degree of partisanship varied from weak to strong in both states regard- less of the situation in party competition. In other words, we found greater differences across newspapers in the save state than across states. K. Implications for Theory Since this chapter immediately precedes the chapters in which we deal with our specific hypotheses, we might briefly point out some implications for theoretical problems in politi- cal communications that seem to grow out of our examination of the political context in the two states and the newspapers in the two-state sample. One point which should be made is that differences be- tween different congressional districts in the same state and different newspapers in the same state max often be as marked, or in some cases perhaps more so, than differences between districts or newspapers in different states. By the same token, some of the Michigan districts may be more like some Tennessee 62 districts than they are like other Kichigan districts, and the same may be true of newspapers as well. In addition to the party dimension of newspaper policy and in congressional district voting, it may also be useful to pay some attention to liberal-conservative preferences in news- paper policy as well as in voting behavior. This is another factor we should bear in mind as we begin to examine the de- tailed data derived from our content analysis. Keeping these things in mind, we proceed to an exanina- tion of the individual specific hypotheses, which will begin with a report of our findings on bias in the Press, discussed in Chapter Three. MAFM'HW; _-‘ I .- 63 Footnotes for Chapter Two l. Percentages based on congressional voting statis- tics found in America Votes for 1960 and in Congressional Quarterly Special Report: Complete Returns of the 1952 Elections by Congressional District. 2. Congressional Quarterly Census Analysis, Washing- ton, Congressional ;uarterly service, 1954. 3. Congressional Quarterly Census Analysis, 92. cit. 4. Ricrofilm files were all examined at the microfilm section of the Richigan State University Library. The six Hichigan newspapers were made available on microfilm files through the courtesy of the Kichigan State Library in Lansing. Files for the Nemphis Connercial appeal and the Nashville Banner were provided by .he Kid-West Inter—Library Center in Chicago through interlibrary loan. The University of Tennes- see Library in Knoxville furnished microfilm files for the Nashville Tennessean, the Chattanooga Times, and the Knoxville Journal and flaws-Sentinel. 5. Data on hichigan newspapers from Ayer's Directory, 1964 and 1962. 6. Data on Tennessee newspapers from Ayer's Directory, 1964 and 1962. CHAPTER THREE How Biased Is Newspaper Coverage of Political Campaigns? I. Introduction The problem of bias in the newspapers is an important problem in the analysis of public opinion. This is because of the considerable likelihood that biased coverage will introduce perceptual distortion into the average reader-voter's "picture" of political campaigns, even though alternative sources of news coverage may be available to him. A dictionary definition of bias in the sense in which we use it would be "a prejudice; bent." But this definition must be refined for our purposes. 64 65 In our context, bias is the use of descriptive words or other means of affecting the reader's attitude toward a candidate so as to cause him to consider the candidate either more favorably or more unfavorably. The definition of bias as used here does not include any assumption of inten- tional use of terms which might describe the candidate either more favorably or more unfavorably, largely because there is simply no way to ascertain intent. However, we can illustrate what we mean by a favorable reference, a neutral reference and an unfavorable reference, which are the bases of the numerical rating system for bias which we propose to use. If a candidate is referred to as "experienced," " "vigorous," or "energetic," this is interpreted as "capable, a favorable reference. If he is referred to as "inept," "irresponsible," "slow-witted," or "unaware of his constitu- ents' interests," or if we say he has been charged with cor- ruption, this is considered an unfavorable reference. De- scriptive material which relates primarily to the candi— date's background, such as his pre-political experience, his home, and other factual matters of this nature can be inter- preted as neutral. To explain biased coverage another way, it is "slanted" coverage (albeit unintentional) tending to favor one candi- date or party at the expense of the other. We suggest that a partial and crude measurement of this kind of bias can be made by a technique for numerical 66 representation of bias which we refer to as the "F-U—N Rating" because it is a measurement of favorable, unfavor- able and neutral references to candidates in coverage of the campaign. II. Analysis of the "FUN Rating" Tabulation First, here is how the FUN Rating is determined in its raw form, giving an idea of the intensity of bias by record- ing the number of references in each of the three categories in the context of all references to candidates in a particular news story. Thus measuring the kinds of references listed above as examples and allowing one unit for each reference, Rep. Alvin M. Bentley, the Republican senatorial nominee in Michigan in 1960, received 12 favorable mentions, 28 neutral mentions and 16 unfavorable mentions, making his raw-score "FUN Rating" 12-28-16. In order to compare ratings of each candidate, however, a refined FUN Rating was also deve10ped, giving all neutral mentions a "0“ value, all favorable ratings a plus one value (for each mention) and all unfavorable ratings a minus one value (for each mention). Thus in this example, we could translate the unrefined FUN Rating of 12-28-16 into a score of -4 for Bentley. The unrefined score shows the ratio of the three kinds of references can be quickly determined, while the refined plus or minus FUN Rating gives the rela- tionship of a candidate's unfavorable references to his 67 favorable ones. Thus refined scores can be compared for cp- posing candidates for a clue as to whether either got favored treatment. III. The Specific Hypothesis In this chapter, we are discussing the problem of bias in the Press in terms of such factors as editorial policies and newshandling policies of the newspapers. We also have a secondary interest in ownership, which was dealt with in Chapter Two. Further we will examine the relationship be- tween the political makeup of the congressional district and the coverage of biennial campaigns conducted in the district. The first specific hypothesis derived from the major hypothesis is stated as follows on Page 26: l. Biased coverage (1. e., "slanted“ coverage, or cov- erage tending to favor one candidate or party at the expense of the other) of congressional campaigns is more likely to be found in rural-district papers than in urban-district papers. With the data we derived for our tables we might also suggest a special case of biased coverage, namely, that biased coverage is more likely to be found in a one-party state like Tennessee than in a two-party state like Michigan. We are interested in the newspaper policies in terms of the endorsement of candidates editorially and in terms of newshandling policies reflected in content analysis of cov- erage of the campaign itself. Our interest in ownership and other factors of Press structure is secondary to our interest "-‘~W~<.1M_' - .3. ;- -‘ - ixruiwl .. . __ v7. . WT -‘ - " I . O . O . F‘ O 7 _ . . . o C . . O 68 in editorial and newshandling policies. Specific factors reflecting policy include editorials embodying policy state- ments, like candidate endorsements, and the treatment of letters to the editor and editorial cartoons. We are considering the problem of biased coverage of congressional campaigns in Michigan and Tennessee, using a twelve-newspaper sample. The Tennessee newspapers in the sample consist of the Memphis Commercial A eal, the Nash- ville Banner, the Nashville Tennessean, the Knoxville N215- Sentinel, the Knoxville Journal, and the Chattanooga Times. The Commercial Appeal and the News-Sentinel are Scripps- Howard Newspapers. The Michigan newspapers in the sample consist of the Flint Journal, the Saginaw News, the Lansing State Journal, the Grand Rapids Press, the Traverse City Record-Ea lo, and the Muskegon Chronicle. Of this sample of six, the State Journal and the Record-Eagle are locally owned, while the re- maining four -- the Press, the Chronicle, the News and the Journal -- are part of the Booth Newspapers, Inc., chain (known as Boothpapers), a subsidiary of the Scripps-Howard interests. These are the twelve newspapers in the two-state sample for the period covered in our analysis, September 1 through Election Day, November 8, 1960. In examining bias in the Press, we took qualitative data of content of news stories. We also examined quali- tative data dealing with editorial content and content of 69 letters to the editor. We examined cartoons in a minor way. All these data were dealt with in the quantitative collec- tion of data as well. IV. Analysis of Tabulations for the Full Sample One finding we can make from the Michigan-Tennessee sets of tables is that prominence of the candidate tended to get him more favorable mentions. While we could point to examples in congressional coverage, this can be most strikingly seen in the case of Sen. Estes Kefauver, whose total score for the six Tennessee papers was plus 70 (/70) while his Republican opponent, A. Bradley Frazier, scored only plus 2 ({2). In examining the scores of the U. 5. Senate candidates in Michigan, we find Bentley's plus 316 (/316) compares quite favorably with Senator McNamara's plus 172 (/172), indicating a considerable bias favorable to the Republican nominee. This may be partly explained by the pro-Republican editorial policy of the Michigan newspapers sampled, which may have been reflected in a tendency to give somewhat more generous coverage to the Republican nominee. (In Tennessee, most newspapers endorsed Kefauver, but the preponderance of cov- erage given him may have been less due to pro-Democratic pol- icy than to the fact that he was better known in Tennessee than his Opponent, a "token" candidate. In Michigan, how- ever, both Bentley and McNamara were about equally well known.) However, this alone is not sufficient explanation for Bentley's advantage, as part of the difference was doubt- less the result of the fact that Bentley made a more vigorous 7O campaign than did Senator McNamara, a fact commented on in several stories. We could conclude then that there was some evidence of bias, but that it was not very strong evidence because of the presence of other factors. Looking at the congressional contests and their cover- age, we would expect to find some partisan bias for the Democratic candidates in Tennessee and for Republican candi- dates in Michigan. There issome slight evidence of this; for example, Rep. Clifford Davis (D), the conservative Democrat who represented the Ninth Tennessee Congressional District, got a score of plus 21 {/21) in the Memphis Commercial Appeal, his home town paper, which was the most of any of the Tennessee candidates covered in that paper and was noticeably greater than the plus 9 (/9) scored by Senator Kefauver. The Nashville Banner, one of the two Openly Republican papers in Tennessee, gave its heaviest coverage to Rep. B. Carroll Reece of the First Tennessee Congressional District with a score of plus 22 {/22), but this was doubtless partly because Congressman Reece also served as state Republican chairman and got publicity in that capacity, so the score may be misleading unless this is taken into account. Further, althOUgh the Banner has a considerable pro-Republican bias by reputation, it accorded good coverage to Rep. Joe L. Evins, Fourth District Democrat running for re-election without 0p- position, largely because of a newsworthy appearance he made in a Christian Church pulpit in a Nashville suburb to answer the minister's attack on Senator Kennedy. 30 newsworthiness 71 as well as partisanship can affect the kind of coverage afforded a candidate. Otherwise, Evins and Kefauver scored well in liberal Democratic papers, the Nashville Tennessean and Chattanooga Timgg, which is not too surprising. Another candidate getting especially favorable mention was Rep. Howard Baker, the incumbent Republican Second District con- gressman in Tennessee, who was the recipient of plus 22 (/22) and plus 9 ({9) scores, highest for the two Knoxville papers, and who had a plus 37 (/37) score, second only to Reece among the congressmen seeking re-election, for the full Tennessee sample of six newspapers. Noteworthy congressional coverage in the eighteen Michigan contests included that given to Rep. George Header of Ann Arbor, the Second District Republican incumbent (plus 36); to Rep. Gerald R. Ford of Grand Rapids, the Fifth Dis- trict Republican incumbent (plus 56); to Rep. Charles E. Chamberlain of East Lansing, the Sixth District Republican incumbent (plus 30) and his Democratic opponent, Genesee County Prosecutor Jerome F. O'Rourke of Flint (plus 43): to Rep. Robert P. Griffin of Traverse City, the Ninth District Republican incumbent (plus 48), and to James Kellie, Demo- cratic nominee in the 18th District who contested with Rep. William E. Broomfield, incumbent Republican, for that seat. Kellie had a score of plus 33 and Broomfield plus 13. In the case of Chamberlain and O'Rourke, the latter got heavy cov— erage in his home town of Flint, accounting in part for his good showing. In the case of Kellie (another exception to ‘ 3W9. m; "3:22;. 3‘ '63"? ‘ - a 3 ' ' —- . o a o I . F a . o s a I . O o e e e D I e v o O D e I O I . e 72 our expectation that incumbents would fare better than non- incumbents), perhaps his newsworthiness (as a combat veteran of World War II with an independent record) may have helped to win him greater coverage. Allowing for these exceptions and a preponderance of Republicans among those given the most favorable coverage in the Michigan congressional contests, a more plausible ex- planation is probably found in the incumbency factor. In- cumbents who were particularly active had a natural advantage and it was probably this combination of incumbency and active campaigning that made the difference. As this factor of amount of campaign activity appeared to occur in the Senate contest where Bentley, the Republican non-incumbent who cam- paigned actively, outscored McNamara, the Democratic incum- bent who did not work intensively until late in the campaign, this subsidiary hypothesis of activity-related coverage seems to get some support. In summary, we found considerable evidence that some candidates get more favorable treatment in the Press than others but little evidence that this was primarily or even in large part due to partisan factors. Also, incumbency was not always the determining variable. Based on the findings summarized in the tables accom- panying this chapter in the thesis, we can reach the following general conclusion about our hypothesis on "bias" in the Press: Biased coverage ("slanted" coverage tending to favor one 73 party or candidate at the expense of the other) of con- gressional campaigns is more likely to be found in rural-dis- trict papers than in urban-district papers. Our exploration of biased coverage was mainly concerned with evidence of favoritism toward one party's candidate and favorable treat- ment of the incumbent. There was more support for the view that incumbency is a factor in favorable coverage (though even here there were exceptions) than for the view that party affiliation is a factor in such coverage. Although we did have some figures comparing urban district with rural dis- trict coverage, there was no clear evidence of urban or rural district papers being the most biased. We did, however, find somewhat greater tendencies toward bias in the rural district papers, as we expected. V. Newspaper Policy and Political Background Bias in newspaper coverage of campaign events may be re- flected in partisan bias (favorable attitude toward one party over the other), liberal-conservative bias on issues, or bias in favor of one candidate over the other. This can be de- termined in terms of endorsements. Most of the hichigan newspapers endorsed Nixon and the Republican congressional candidate in the district, although the Traverse City Record- gaglg maintained neutrality. In Tennessee, the Remphis Commercial Appeal, the two Knoxville papers and the Nashville Banner endorsed Nixon while the Chattanooga Times and the Nashville Tennessean endorsed Kennedy. The Commercial appeal, 74 the Times and the Tennessean endorsed Democratic nominees for Congress, but the Banner and the two Knoxville papers en- dorsed Republican congressional nominees. We were interested in determining whether the endorsed candidates got more favorable treatment on the news pages as well as in editorial endorsements. This was most true of the Knoxville Journal and Nashville Banner, only slightly less true of the Tennessean, and least so for the Knoxville Ngwg—Sentinel, the Chattanooga Tipgg and the Memphis Commercial Appeal. The intensity of Republicanism in Tennessee may be due to the fact that, although East Tennessee is a "one-party island“ in a predominantly Democratic state, Tennessee is a modified one-party state. If this terminology is confusing, it must be explained that traditional Republicanism is breaking down in the urbanized portions of East Tennessee, while traditional Democratic views in West Tennessee are being weakened by the civil rights issue. Michigan, on the other hand, is basically a two-party competitive state like Indiana, Illinois or Ohio. This would account for some differences in intensity of partisanship between the Tennessee and Michigan newspapers. To deal in specifics, we found that the Nashville Banner was the most vitriolic (in terms of editorial comment, cartoons, etc.) and the most bitterly partisan of any of the papers, with the possible exception of the Knoxville Journal. The Tennessean, while partisan and biased in favor of liberal Democrats and the TVA, did not pursue its partisanship so 75 viciously and continously. The editor of the Knoxville Journal, Guy Lincoln Smith, is a former state chairman of the Tennessee Republican Party. The Nashville Banner, founded in 1876 (the year of the Hayes- Tilden disputed presidential election at the end of the Reconstruction Era), is an Old Guard Republican newspaper like the Journal. Both, however, supported Richard Milhous Nixon in 1960. On the other hand the Memphis Commercial Appeal, in which we found evidences of moderate bias, strad- dled party lines. We might call it a conservative Democratic newspaper, on both civil rights and economic issues. But in 1960, partly for lack of candidates more in keeping with its own viewpoint and the embarrassment that might have ensued if it had made no endorsements, the Commercial Appeal backed the mixed slate of Vice President Nixon for president, Senator Estes Kefauver (D-Tenn) for re-election to the U. S. Senate, and Representative Clifford Davis (D-Tenn), the conservative Democratic congressman representing the Ninth Tennessee (Shelby County) District. Thus one can see a mixture of party choices, and a mixture of issue preferences or ideolog- ical choices in candidate preferences, probably brought about by the practical considerations mentioned above. The Nashville Banner, on the other hand, went straight- line Republican, while the Nashville Tennessean, a strongly liberal Democratic newspaper which is firmly linked with TVA interests in Tennessee, maintains its liberal policies in Nashville, Tennessee's capital city. It endorsed 76 Representative Richard Fulton, incumbent Democratic congress— man from the Fifth District, in 1962, and endorsed Senator Ross Bass (D-Tenn), successor to the late Estes Kefauver, in his 1964 primary campaign against Governor Frank G. Clement. VI. Other Types of Media Bias Bias found in the newspaper press is not the only type of bias found in the mass media. Varying shades of bias may be found in radio or television coverage, but in neither of these two media do we find the headline factor to be as im- portant as in the newSpaper format. News display (a minor factor in the television format) is also more important in newspapers because their typOgraphy permits a wider range of headline sizes. This is also partly because of the instantaneous recep- tion of television and radio compared to the printed page medium. Therefore, senationalism through newspaper head- lines, which can be read and reread, and bias in the head— lines, as well as news "play" (position on page) above and below the fold of the printed page, may be important factors in the impression the reader gets from the newspaper about campaign events, about Specific rallies and meetings, the filings of candidates, and the various other aspects in the different stages of the electoral contest. ' This time factor is one essential difference between newspapers and the electronic communications media such as radio and television. Besides analysis of the general problem of bias in the 77 Press, we also have an interest in the "Press bias" variable in connection with the urban-rural dimension in voting, and in relation to more general demographic or socio-economic factors. First, we conjecture that all newspapers are biased be- cause total objectivity of news coverage is in practice un- attainable, whether by newspaper reporters, radio-TV report- ers, or their editors and publishers, station managers, news script supervisors, or other policy-makers and regular employ- ees in the "working press". While this test theoretically permits a score of total "neutrality" in coverage in any one news story, the problem is to recognize when bias in the form of "slanting" is no- ticeable, i. e., when it becomes measurable with a preponder- ance of unfavorable references (item references at the sentence level) in the analysis of one news item. Granting that total objectivity in either reporting or editing is unattainable, we must look at forms of "doctoring" the news. This may occur in the editing of press association dispatches (the measure of this would occur through compari— son of printed items in specific newspapers that subscribe to it with the Associated Press or United Press International AAA Wire (trunk) for the day in question). Another factor which may be biasing in the news items on the page may be the news "play" or display factor, not just in one instance (easily interpreted as an exception) but con- sistent favoring of one party or candidate over another. It ‘74 78 may also be found that partisan bias is great in certain newspapers, while it is slight in other newspapers, partic- ularly so in the case of rural district papers where the editor may know many readers personally. In this latter case, there may be an almost militant anti-party bias of the Ostrogcrski type.l An example of this would be the Traverse City Record-Eagle, whose editor refused to publish letters to the editor because he felt READERS were too biased! We can probably make only a subjective measurement of this type of phenomenon, using qualitative analysis data, even employing the FUN Rating. The best study we have of this kind, to date, is not a scholarly study. It is Ben Bagdikian's study of the nationally circulated weekly news- magazines -- Time, Newsweek, and U. 3. News and World Re port 0 2 Bagdikian found evidence of bias and distortion in all three of the weekly newsmagazines in a series of articles published first in the Providence, R. I., Journal-Bulletin and reprinted in the New Republic in 1959. The articles dis- cussed the accuracy of gipg (or lack thereof), pro-business bias in U. 8. News and World Report, and the prediction methods of Newsweek's cepyrighted special section, "Perisc0pe". Bagdikian's articles, to summarize, were concerned with his search for inaccuracies and distortions and militant partisan bias. 79 VII. Qualitative and Quantitative Analysis What kind of basic data are called for in our analysis? We have two types, the latter of which can be divided into two sub-categories. The first kind is derived from quali- tative content analysis and will be presented in tabular form and discussed briefly in this chapter. Tabulations were made for a rural congressional district daily, the Muskegon Chronicle, and for an urban district daily, the Grand Rapids ngpp (see Tables I and II with this chapter). For the Grand Rapids Press, there was a total of 66 favorable references to congressional candidates, 16 to Democrats and 50 to Repub- licans; 16 unfavorable references to congressional candi- dates, 8 to Democrats and 8 to Republicans, and 13 neutral references, 6 to Democrats and 7 to Republicans. Heaviest favorable references were to two Republican candidates, Ford (42) and Griffin (14), neither of whom received any unfavor- able references. This indicates some pro-Republican bias in an urban newspaper. For the Muskegon Chronicle, there was a total of 106 favorable references to congressional candidates, 46 to Democrats and 60 to Republicans; 27 unfavorable refer- ences, 14 to Democrats and 13 to Republicans, and 16 neutral references, 6 to Democrats and 10 to Republicans. From this limited sample we would have to conclude that there is a greater degree of pro-Republican bias in the one urban dis- trict newspaper examined than in the one rural district news- paper examined. In the Chronicle data, the greatest number 80 of favorable references found were for Griffin with 22, Meader with 14, Kellie with 15, O'Hara and McIntosh with 12 each, and Jennings with 10. Of these men, three were Repub- licans and three were Democrats. Although favorable Repub- lican candidate mentions totaled 48 and favorable Democratic candidate mentions totaled 37, this is much nearer an even balance than was found in the case of the ggggg. A second kind of analysis consists of comparative study of same-district newspapers taking the general data mentioned above, but limiting this to make comparisons such as the following two examples: (1) The Knoxville Journal (locally edited) and the Knoxville Npgp-Sentinel (guided somewhat by Scrippe-Howard chain policy). The Journal has the reputation in the com- munity of being more highly biased than the Npgp—Sentinel. Our data should show whether subjective use of qualitative content analysis data would bear this out. We can point to instances from editorial statements alone which would seem to, without even examining letters to the editor. (See Table III with this chapter, which compares these two newspapere' coverage of a two-party contest in one district with that of the Flint Journal; the greater number of mentions for both candidates in the latter reflects greater attention to the contest as well as less indication of bias -- the favorable mentions are more evenly distributed.) (2) The Nashville Banner (strongly partisan and locally 81 edited) and the Nashville Tennessean (strongly partisan and locally edited). The Banner has the reputation Of being biased conservatively, almost to the point of being extrem- ist. The Tennessean has the reputation of being liberally biased, especially on anything affecting Tennessee Valley Authority-oriented interests. Public utilities such as the TVA are strongly favored by the Tennessean, so much so that in the case in which Governor Frank G. Clement of Tennessee secured passage in the General Assembly at its 1963 session Of a bill to tax services, extending the sales tax to utili- ties services and thus adversely affecting TVA, the Tennes- ppgp violently Opposed him and carried cartoons lambasting him as being corrupt and wearing a suit marked with dollar signs. These cartoons were drawn by the regular Tennessean cartoonist, Tom Little. Does subjective use of qualitative content analysis bear out this pro—TVA, pro-liberal Demo- cratic bias? We think it does. In Table IV with this chap- ter, it is noteworthy that the Tennessean gave considerably less favorable coverage to Representative B. Carroll Reece, First District congressman and state GOP chairman, than did its Republican Opposition, the Banner. Tables V through VIII contain additional FUN Rating data. Another investigative angle, including a different type of data (this is not a directly content analysis approach, but rather deals with policy orientation of the newspapers studied), asks what are the newspapers' basic policies on political questions and issues (a dimension different from the partisan one). 82 We can find statements (1. e., a statement in each news- paper) of formally published or institutional policy orien- tation for the newspapers in the l2-newspaper sample. This can be done by inspecting the editorials published in the newspapers included in our sample for the September-November period of 1960. Typical of such editorials, either stating formal en- dorsements or making the papers' policy stands clear, are the following editorials: 1. Grand Rapids ngpp -- Editorial on November 4, 1960, "The Choice for President", three columns, 33 inches, endors- ing Nixon and Ford. Editorial on October 31, 1960, "We're Voting for Congress Too", three columns, 21 3/4 inches, en- dorsing Ford, and calling on Republican workers to work hard for the congressional slate. 2. Flint Journal -- Editorial on November 6, 1960, "Nixon Is Best Qualified to Lead Nation, Free World", three columns, 39 3/4 inches, including an endorsement of Nixon and the Republican slate. 3. Traverse City Record-Eagle —- No endorsement of either candidate. A typical non-partisan appeal, published October 7, 1960, measuring one column (wide measure) by 4 inches, consisted of an editorial calling for participation and headed "Political Apathy". 4. Lansing State Journal -— Actual endorsement occurred in a day's newspaper which was not in the sample, but a typical editorial appeared on October 23, 1960, headed :"rrezmsrt'"... -1 .5- -. ., ....~ A- at '_' await ‘ I 0 ~—. I I . . . ~ , e e . ' I o e o v . U o I D O . ‘ . l . . . 0 e o ‘ O . U V ‘ . ‘ e O . . .1 O a 7 e e a « A . ‘ . . l O 83 "Unfair Tactics in Campaign”, three columns, 12 inches. This was a complaint about tapes of the presidential debates being used for campaign purposes by Kennedy supporters, a practice the editorial described as unfair tactics. 5. ChattanOOga Tippp -- Relisting of endorsements, "In Today's Election", one column, 4 inches, on November 8, 1960, a reiteration of the paper’s policy stands for Kennedy and the Democratic slate. 6. hemphis Commercial Appeal --"Voter’s Guide", one column, 17% inches, on November 8, 1960, a repetition of endorsements of Nixon, Kefauver and Davis. In Chapter Two, a complete summary of newspaper endorse- ments was included. Next we will deal with urban-rural differences in the congressional campaign coverage, along with a brief look at the incumbency factor. '. r v 'l'" \r L 84 no X mm\ o moaHB «woosdppmso condense Auomomnonbv H \ me any menn “sum AbomoquSDv H x N x any upehopm ”new Anemonmocbv o m x any houses ”can Anomomaosbv n x n x any when ”new Apomoanosbv N x H \ Ana homes "new AbomOQQOCDV HHx N x AQV msHpm as»: Anemonmossv o H \ AQV AOHNmnm "ppm Apomoamossv m x H \ Amy “exam "paw o H . Ana psmHum NNX o “my momom "and OHX m \ ADV no>5mhom H x 0 Amy noHumnm Ammmmfi nmssmm HsHonmaaoo oHHdoshoz oHnesos mopeoHeedo Aowmav oHQamm nonmamxozlmH non mmszmm 23m “OHQmHum> owmuo>oo pommHm "AHHH nopnmsov H mqm<9 umpOanmHQ omsom “summon Apomommonav m x o o . N \ any means: ”spa Abomoanonov NNx H\ n x m x “as deem ”new Apomommonpv m x o o H \ gov swoon “new Anemonaonpv .9 me o o me any merm as»: 8 Apomomoonbv OH\ 0 o o gov elooum seem . Anomommospv an\ ax . NNx 0 Amy women ”oeN N u o o H u “av oannm Nex mx OHx N x Amy oooom «peH .eeloeoHn omoom oax m\ N x me any hoeoeuom N x Na 0 H . Amy eoHuonm «meadow Amnonmnmsoz NHmv amassed mosHpnomumzoz snowmocsow oHasem oHHaewoem oHHHewoea oHHeenoez eoeoeHoneo commandos Hana non humaaom Aomev oHoamm acamamzoZINH you mwsHpmm 29m euphonoa .oHneHne> omoeopoo eoeon .AHHH noedosov HH mamas 86 me o o m x o o Amuoamamxoz NHwV Hacksaw HosHpnomnmzoz onamm OHHprosM OHHprosu commandos Hank you humaaom Apomoamosbv H \ any aspen ”hem “pomommospv H x Ana oneness «new looosHenoov unpoanmHQ omsom samenesswm oHHHpnmmz measpHpsso ApossHpsoov AoomHv oHasmm noanmBoZuNH non mwsHpmm 25m copesaea .erloe> omeaoeoo ooeon «AHHH hopoeeov HH mamas 87 00 00 NH HF! 00 W: r-IH NH \l XV. \\ 3‘ ‘4»! 0m N \ mwx Hacksaw madam wnHmnoq o me o me o m x o m x o H x o a x o o o on o o o o o o o N x o m x o me o o o o N \ me o N x me eNx Mdmmemmmmmm exoz muHo oesopdsa :anwmm Aowmav oHasmm soammmzoZINH sou mwsHpmm 29m commence .eHmoHses omesopoo eeeon .AHHH messages HHH mamas “av seesaw mmH: “my seesaw “my neopeHoz “my osem.oe omhsom.o Amy eHoHNopsesoe ADV soamom whom: AQV masom Amv soausomo AQV MHGHU “My «H0 ”Sancho... 3v ashes Amv soonest Amy ononnmno Any uonosnomza any mumBOZoSe ApmHGOHpHpHnosmv oonmsopom QHHHom Amy soHedom nopOchnmo ”new «Sun ”new «Sun as»: «can «paw «pea «mpOHspmHn mmoom ”madden 'fi'vn’. . , "I'If; '- o o o “my odHnmom o o o and HHoonme «memH o o me Amv sHoz .es: 0 o a x Amy phoneme .mpeH go o o o Hmv HHosmoon 8 o o o 2: 33m: 3an o o o “my assesses o o 0 Amy speedome ”meNH H I o H \ ADV nsosm H x o H \ “my eons. ”seHH m x o H \ Amy seem o o m x Hmv whopsoooo. ”neoH 0 HH\ 0 Ame eweHseos a x HHx e x Hmv sHsNHuoe ”sum AoooeHeeoov «epoHneon ouoom m$m¢M|psooom mzoz mmmmmmw spasm wsfiussq tho masseuse :susam nonmcacnoo Aboaannoov Aoome onamm MonsantoZINH sou mwsHumm 23m meanness «OHDGHAN> m mumboo comaam NAHHH Houndsov HHH NAM48 _, —_._ "a 2’ NF" “=3" x!’; H ‘1’ -.. 89 pamnasosH monosont o o o heoHsehHm HHm ‘ :H hopeeHeceo memes osHms H\ o H\ Amy mHHHom ma 0 Ha Amy eHonsoon “mpmH o o Ame HHohhoz o or H\ Amv hmpHmsHmo .mszw "mpsH o o o Amv msoHo oom AeooeHesoov o o 0 Amy HmesHmome “nemH “hpoHspmHm venom onmmupsooom mzoz AmQHSOh mumpm wGHmsmq tho omsobmsa BmQmem mopmoHmeo ApmpsHosooV AoomHv oHasmm scamamzoZINH son mmsHpmm 25m pmpmsnp< .onloes omosoeoo eohon ”AHHH homeomov HHH mamas NH \ H x o H . Ana cooeem oon mm x H x o N x Amv sopsom ”mum NN x HHx a x m \ Amy shoeeHom nH \ s x m x s x Amy oeem.o. sees me x o me o Amv omsoom.o on x H \ me m \ .mv eHeHeomaesoe «new H x o H I N \ “av noamom on x o n x m: \ Amv enema «men N x o H x o “my wehem m x o a x o Amv eosmsom. «no: 0 9 H \ O N X . o :3 MHGHU a x H x N x o “my someones. ”ohm oH x o N x o “no eases on \ MH\ m x m \ Amv soonest «pnm o o o 0 Amy Hmooesoho a x H x m x m \ Amy Nonohmoose ”emH «meoHNeon omaom Nst mm\ mm\ mH x Ame ehosozoze m x o H x o aeoHeoHerHnosmv eosmsobom eHHHom anx wmx mmx 00H\ “my hOHpnom «opmnom Amnommmmsoz hHHmn 0v oHoHsosso Hangman mmosm oHeaom domHson HHsm somomoo: esHHm eeHmom edema mopeeHoeoo pooesoo Aomev oHaamm HommmmzoZINH sou mwsHpmm 29m pennanps umeoHnm> emmsopoo pomon NAHHH sopnmsov >H mam ommsopoo oomaHm NAHHH sopadnuv >H mqmma _‘_ i » “ ) I . ' f ’ - . e l 1 O i | r i , I v ) l O . . . - . 92 psmnanonH movement .m- HmemH .noHpHm -Hmosmv oHsm oHNmpso eHHHsmm Am- finesH .eoo mesa moose oHeom an- “nemH .eoHeHmHmosmv sHmseH soon an- AmmoH .ooHeHsHsosmv shoaompeom ooseHHz Am: Amps .somomuumHHoHoomv hHHHm she: «AhHso mmmnm mpHmmm psmso QH .mopmpHpsmo HNGOHmmouwsoo mpsmm panav mm\ :H\ H\ . me any mHHHmM me m x o a \ Hmv oHoHNsoosm "NNNH m x N x o n \ Ame HHouhom uHx N x o on ADV mspHssHso .mszs ”npmH ApossHpSOUV "mpOHspmHQ omsom “macadamxoz hHHmn wv oHoHsosso Housman mmosm oHnsmm smenOHz HHnm newsman: psHHm anQem cacao mopmpHosmo penance “popsHosooV AoomHV oHasmm nonmamxozuNH sou mwanmm 25m oopmswpm «oHpaHsmb owssosoo oomsHm «AHHH nomadsov >H mummy o o N Amy ededcoe H 0 3H Amy sHsNHhce «eoHeemHm mom 0 H 0 Ana cooeem omHz o o H Amv soeeom neoHNeon mom a m a Ame mooeeHos m m a “ma deem.oe ”moHsemHm mas o o 0 gas omuaom.o H a a Amv eHeHeomaemoe «ploeon new 0 o N Amy :oaoom mu o o N: 2: enema .3233 new 0 o o Amy mehom o o o “my cosmmome ”elopmHm see c o 0 Amy mseHo o o 0 Amy commence: ”poHseon can 0 o o 2: 3st o o m Amy houses: «poHNeon onN o o 0 Amy onossmno o H N Amv Nonoeeoem* «eoHseon eeH Hmsuzmz OHnmsO>mHQD oHnmsO>mm msHeom 23m opoeHoeeo ploeeHm «HHZO meonBmHQ AfiZOHmmmmwzoo AommHv Huuuuu oeHmem edema omsv nonmamzoz chnumHn sans: an :H mopmpHpqmu Op mopsououom Mo mHthmsd psounoo hsowopmo «oHpmHsm> owmsopoo pommHm «AHHH sounmnov > mqmmusD meHeom 29m 00 H0 00 00 OH 00 OO oHnmsopmm Amy MHMHU ADV HMmSHmmqt Amv meHmmom “ma HHomsHme Amy HHmz .mnz ADV pfimnmmt Amy HHozmoeHm Any wwwHQt “ma sepsoooz Amv peossome ADV czosm Amv momma Ana comm “my muonsopooa opooHesoo «eoHseeHm semH .eoHnemHm mpmH .eoHsehHm mpeH ”poHspon spmH «pOHRpmHQ spNH «DOHHpmHQ anH ”pOHHpmHQ SpOH POHHDmHQ ApossHpnoov «ano mBOHmamHQ AdZOHmmmmUZOU ApessHpsouv HoomHv Ammonm mpHamm compo onev nonmamsoz pOHspmHo sons: so SH mopmpHpseo Op.moososouom so mHmmHms< psopnoo hsowopmo ”OHansm>.owmsm>oo pommHm «AHHH seemeeov s mamas . 1: -.,',_~W_~_._ a. -Tw- - poopsaosH.mmposoQ* an NOH meN mmasos mammc w o o omesopoo spasm osHss “mpomzmmmmomHm Illmwln (mm -=mmH_ mHomoemsm 5 o, mm HH “my soHpcom .m sHpHm 5 mm am Hmv onesNZos .> moHspemw ”smmsZOU msmzmm .m .m MH mH mm hHepoemom o o H Amv mHHHom o o 0 “ma eHoHNsoohme upoHsemHm memH o o o “ma HHosnoz o o 0 “ma mmeHsmHeu .mmmo "poHseon mpNH Hospsoz oHpmsosmunD oHDmsopmm msHeem 2pm opeeHoeoo mloemHm ApopoHoaoov «HQZO mBUHmBmHQ Q420Hmmmmuzoo ApNUOHosoov Aomev Anmmsm mpHamm comma 036V sonstSoz uoHspmHn sensb_sm SH mopmchsmo on moosososom so mHthms< pnopnoo hsowopso «oHQmHsm> owmseboo ommmHm .NAHHH sopamsov b mqmme o N 0H Amv hmcHdeoe N a NN Amv eHmNHeue ”poHNemHm com o o H Amy soohem hmHm o o o , Hmv sophem «poHNpmHm sum N m NH Amy mmoeeHom H m NH “ma osom.oe "eoHneeHm sea 0 o o “my omseom.o o o 0 Amy eHoHsomaesoe «eoHNemHm new 0 o 0 any meadow m" o o o Hmv enema “poHsthm new 0 o o ADV manom o o o “my dessmome .eoHNeeHm a»: o o 0 Ana msoHo o o H .mv commences ”moHspmHm can 0 o o .Anv onhmm H H aH Amv meshes. «eoHsthm oeN H o 0 Amy onossmmo H o H Amy honoseoeze “eloemHm pmH Humanoz QHnmso>ounp mHnmsobmm msHeem 2mm oeoeHosoo poHsemHm ”Nuzo mBUHmBmHQ A<20Hmmmmozoo HoomHV Amwamsomso :Owoxmszv smnmmmzmz pOHspmHQ Hmuom m :H mopmpHosmo op moosmnomom no mHthmsm psousoo hsowopmo .oneHhos omesoeoo ooeeHm .AHHH seemenov H> mamas 97 00 00 00 HH OO r-lr-l 00 00 OO 00 00 OO 00 OO 00 HO HN HO Hmupsmz mapmmopwmnp wcham 2pm manmnopmm Amv mannom Hov HHmmnHo* Amy nHaz .mp2 ADV pawnmmt Amy HmeHoaHm Any mmmHn* Hay usunmooz Amv pamszwmt nzonm Amy N0GM$ any comm Amv mumpuocoow opmoaunwo "pOHNNNHm spmH “poHuomHn nde upoHupmHn gumH “poHupmHn spNH upoHNpan spHH ”poHupmHn sNOH peanumdn Acossznoov «NAZO mBDHmsmHQ AdZOHmmmmozou Acmanpcoov Aowmav AmHOHconso nomemszv nmmmmmzoz uoHHpmHn Hmnsm a nH mopmchcmo op moonmumuom no mHthmc< pnmuzoo knowmpmo «mHannm> mwmnmboo cmmem .AHHH umpnanov H> mHmNB pcopasocH mmpocma* mN HN mON mHH< : mH mH Hay mnmam202 .> HOHupam* 8 o, flammazoo me mwmnmbou commHm «AHHH nonmasov Hb mqm¢a unopasocH mmuo:m0* .gonwwwmn 03» no cmemc HmcHano as» Sony omvfiaoxm hammomnsg was .hHmonu mpcmasoo soan .mmwna pHonpoQ on» onm .manwp mHsp GH omczaocH non mum pcmsaoo HmHnopHcm pmoHpmms wthuumo mummmm mHsmamz and oHHanmmz .mnasaoo snopHcm m8» op mnmupqu :H no Amoapaaom Hobvalamooa wGHproo mumpuoaou no munmcnommmunoo HonpHHom mm sosmv mumpan mumpm hp mcaaaoo cmcme :H hammoa cmmmoumxm ohm one: nonmasnmp mmmmo on» :H mnoHnHmo o, .mmpmcHuzmo ANGOmeonwcoo no mwanopoo pstH hampHp o, Imaon cam nozonnmmHm cam “condmm mo mwmnmpoo m>an voponozH mmaHB mwoozmppmso cam Hmchcomumez 02» Scam manaam Noma ms» cmpm .pmopzoo Hmenmchmum no oonmpnoo hapmoa .HmHHOpmnmm no panH .moomn HanonmmanOO no mmumnm mum mpcmaaoo HmHnopHcm «wmaoz mumuom ououm mumumN Hay omnsom.o oaonmu Hum-MN NuHuHH N-:-NH Hmv :HmHnmpawso mmHNmno .nmm: ”Hmnnson unaam 0.0.0 0.0.0 0.0.0 pstnm uanNH :uoum ouono anoum Amy mommm HHoNpmo .m .amm* ”Hanuson oHHHPNon ououo ououo ououo Hov pannm nsnpn< HnOIoa onono Huouoa Amy mommm Hmonnmu .m .nmmt «Hmchnmmumxmz oHHwaon Aampoav Amnasaoo aoHanov Amcaaaou msmzv wchdm zsh manmm 29m wanwm zam nonmgmxoz AHwGMSOh anHm cam HannaOH cam Hmanammlmsz mHHHpuoch mzaaaoo QOHSHQO .m> mnasaoo mxoz mnoawmmzoz mouse :H ummpnoo «no :a mapmHnm> mwmuopoo wommHm NAHHH Ampmmsov HH> mqmnman 30H23 onwon HmHHOpHoo cm hp :oHnmmm HmemHHoo :H oonHanmpmn hapcmnvmum mH mnHmso nmmmmmzmn no and mummwmwzm: :mpHHomoupma mo hoHHoa HmHHOpHcm .mamHOHpHno :mmean and Hacomumm :H thmnoan wconpm no monopnmbo 29H; mmmmung nopcwam cum mnmwoam quHan hHHwams and mme we empamnmsw woow hupmnm a mum .mnmamamzm: pmoa mo moHHom swampuqu ms» hp oocfiaowm .mumppmH maoamnon¢ .msoHpsmo mH pamaaoo HmHnopva when; coprspHm m :H men on whox mum massaoo :HOpHom map on mumppmqg QH noproHHnsm you mnmppma mo coHpomHmm cum Amman HmHHOpHom m2» mm HHm: mm mmmmamzm: no wnHHmmgmmv massaoo cmanm no madamzmz Homanpnoov Aomachnoov «WMBOZ AHw2H50h »nHHm cam dmmmmmm cam HmchCmmnmsmz mHHprocMV msaaaoo GOHnHQo .m> measaoo mzoz mnogwamzmz manna :H pmmpsoo 0:0 SH mapmHnm> owmnmboo cmmmHm «AHHH nopmmsov HH> mqmmmMm mqm<9 102 Footnotes for Chapter Three 1. Austin Ranney says in The Doctrine of Responsible Party Government (see Footnote 2, Chapter OnET that Ostrogorski devoted more than 20 years to the study of the English and American party systems. In his principal work, Democracy and the Organization of Political Parties, Vol. II (the United States), Garden City, Anchor, 1964} p. 356, Ostrogorski commented: . . . Is not the solution demanded by the problem of parties an obvious one? Does it not consist in discarding the use of permanent parties with power as their end, and in restoring and reserving to party its essential character of a combination of citizens formed specially for a particular political issue? . . . Party as a general contractor for the numerous and varied problems present and to come, awaiting solution, would give place to special organizations, limited to particular objects. . . Ranney takes issue with Ostrogorski, in the following passage (0 £0 cit.. p. 133): Despite its broad scepe and wealth of learning . . . Ostrogorski's indictment of American parties and his attack on the party-government point of view cannot withstand close analysis. He assigned to 'democracy' an arbitrary defini- tion under which no political parties mankind has ever seen could have any place, and sought to prove by means of argu- ments derived from that definition, that parties as he observed them could not possibly be democratic institutions. He therefore won his battle against parties even before the commencement of hostilities. But it is quite as accurate to say that, from any point of view save his own, he lost the battle." 2. Ben H. Bagdikian, three articles in the New Republic, as follows: “The Weekly Newsmagazines -- “I. U. 8. News and World Report", February 2, 1959; "Those Uncanny Predictions in Newsweek", February 16, 1959; "TIME- Study", February 23. 1959. \O CHAPTER FCUI The Urban-Rural Dimension in Campaign Coverage And Its Relation to Incumbency I. Statement of Specific Hypotheses In applying the major hypothesis of this press content study to political problems, we formulated a specific hypoth- esis consisting of two parts. It may be stated as follows: Specific Hypothesis 2: In comparing rural congressional districts with urban districts, total coverage of candidates Will be greater in newspapers published in rural districts than in those published in urban districts. This will be true in two specific cases illustrated by Specific Hypotheses 2(a) and 2(b). Looking more closely at these specific cases, Specific 103 104 Hypothesis 2(a) states: Both major-party candidates get more total coverage (in terms of measurement of actual newspaper space and not potential exposure) in rural than in urban dis- tricts. Specific Hypothesis 2(b) states: Non-incumbents get more coverage in newspapers published in rural congressional districts than in newspapers published in urban districts. In relation to 2(b), we expect to find newspaper cover- age of non-incumbents will tend to he found in "round-up" or "wrap—up" stories dealing generally with the campaign, rather than in separate individual stories in papers published in urban districts. Non-incumbents will, however, get somewhat more coverage where they mount an intensive campaign, regard- less of whether the newspaper giving coverage is published in an urban or rural district. (As noted in Chapter One, in- creased campaigning is assumed to be reflected in greater cov- erage of the non-incumbents as shown in the content analysis data.) Our interest in rural-urban constituency differences in candidate coverage is related to the current interest in re- apportionment of state legislatures and the effect of this reapportionment on the redistricting of congressional seats. This essentially urban-rural issue has been the focus of a congressional struggle in the 88th and 89th Congresses over a constitutional amendment sponsored by Sen. Everett K. Dirksen (RrIllinois), the Senate minority leader, which would make it legal for states to apportion one legislative chamber on a basis other than pOpulation. -’ ‘ v—rr-uv v- _ '? g . ,‘,1 HIM-u," ' 1 ‘ :'. 105 Political factors arising from congressional redis- tricting are closely related to legislative redistricting. For many years, state legislatures had a free hand in redis- tricting their own seats, but in a Tennessee case, gaggg v. gagg, decided in 1962, the cities of Knoxville, Memphis, Chattanooga and Nashville succeeded in bringing about the in- tervention of the federal courts in the legislative apportion- ment battle. The fight is still going on in Tennessee, which held a state constitutional convention in 1965 as a last ditch effort by rural interests to restrict the impact of reappor- tionment. In hichigan, the case of Scholle v. fiagg launched the judicial aspect of the fight, but it was not until June 15, 1964, when the U. S. Supreme Court ruled that the "one man, one vote" principle applied to both houses of state leg— islatures, that the provision in hichigan's 1963 Constitution, promoted by Governor George Romney and narrowly adepted over the Opposition of Democrats partly because of its reappor- tionment provisions, was superseded and the Eichigan Legis- lature was reapportioned in compliance with the U. S. Supreme Court order. Thus while Michigan succeeded in modernizing its repre— sentation in the legislative houses, Tennessee was still struggling with the problem in late 1965. Both legislative reapportionments will eventually affect congressional dis- tricting but the most important immediate impact was felt in Michigan which in 1964 approved a new state-wide congression- al redistricting. However, Michigan may ultimately prove more I ~ o':-."“-"~P .. w“: a!!! 106 stable than Tennessee, as any change in Tennessee will prob- ably give more representation to Memphis and Shelby County and this is one of the areas of Tennessee where the Negro vote is most strongly felt. In compiling data for this chapter, we prepared tables to compare urban and rural district coverage in all 12 news- papers included in the tao-state sample. In analyzing data on the (a) and (b) special cases for this specific hypothesis, we tabulated urban-rural variables in the following forms: (1) A table showing total coverage for all candidates (congressional. presidential, vice presidential and senatori— al) in national races that were covered by all newspapers in the sample published in urban districts in 1960. Data for each candidate or political figure are presented in this table (Table I, appended to this chapter) in the form of fre- quency of items (including both news stories and photo cover- age) and total column inches for each category. This gave us a measure of the kind of coverage given in all daily newspa- pers from the sample published in urban congressional dis- tricts. From these totals, we derived two sets of figures for the "typical" urban-district paper, mean frequency and mean column inches. We also computed the total frequency of measured stories and total column inches for the stories. (2) A table showing total coverage for all candidates (congressional, presidential, vice presidential and senatori- al) in national races that were covered in all newspapers in 107 the sample published in rural districts in 1960. Data are presented for each candidate or political figure in this table (Table II, appended to this chapter) in the form of frequency of items and total column inches for each category. This gave us a measure of all daily newspaper coverage found in sample newspapers from rural congressional districts. From these totals, we derived two sets of figures for the "typical" rural-district paper, mean freouency and mean column inches. We computed the total frequency as well for measured stories, and the total column inches. (3) A table showing total coverage for all candidates in these four categories that were covered by one urban district paper, the Memphis Commgrcial Appeal, in 1960. Categoriza- tion data included frequency and total column inches figures, giving us a measure of the kind of coverage given in a daily newspaper published in one of the urban congressional dis- tricts in the sample. (Table V) (4) A table giving similar data for a newspaper pub- lished in a rural congressional district. While we have used the terms "urban" and "rural" in a loose sense, we took care to see that the paper was one published in a basically rural congressional district. For this purpose, since regional differences did not constitute a variable in examining this hypothesis and the state of publication did not matter, we chose the Nuskegon Chronicle, a Booth Newspaper, to examine the rural congressional district coverage. (Table VI) The above were the four tables pertinent to part (a) of 108 the specific hypothesis. In dealing with part (b) of the specific hypothesis, we used two approaches, and compiled two sets of tables. The first two tables pertinent to part (b), Tables III and IV (both appended to this chapter), were compiled from a tabulation of figures on all types of candidates in 1960, with frequency of items and column inches in all items for each category totaled. In similar fashion to the computations for Tables I and II, figures were computed for a "typical" dis— trict for mean frequency and mean column inches. This was done in Table III for all newspapers published in districts with incumbent candidates, while in Table IV figures were com- puted for all newspapers published in districts where all can- didates were non-incumbents. The other two tables pertinent to part (b), Tables VII and VIII, were limited to senatorial and congressional cam- paign data, and dealt with the incumbency variable as it re- lated to coverage in two Eichigan newspapers, one published in an urban constituency, the Flint igprgal, and the other published in a rural district, the Traverse City Record—Eagle (districts are the Sixth and Ninth Michigan CD's, respective- ly). (This made it possible to concentrate on the coverage given incumbents -- starred in the tables -- and non—incum— bents in the same district.) Under the senatorial and congressional categories in the coverage in each of the newspapers for Tables VII and VIII, we computed frequency of items and total column inches data. 109 We did not differentiate, in any of these single—newspaper tables (V, VI, VII, and VIII) between editorial Opinion or news coverage, using total coverage data in an effort to keep the tables statistically significant. Data were compiled in Tables VII and VIII for each congressional district race in hichigan, and for the U. 8. Senate contest in Kichigen in 1960. These four tables (like the full-sample tables, I, II, III and IV) were all limited to 1960 data. The use of 1960 data makes it possible to compare the coverage in that year with the coverage in other years when district lines were the same, such as 1958. However, the 1958 election coverage was beyond the scepe of this thesis. Besides the frequency count of news items, and the com- putations of total column inches for each candidate and each category, we took another computation to show that the news- papers were comparable. This computation was made to find the percentage of newspage content within the total content of each of the 12 newspapers represented in the entire Sample Nothing in these computations indicated a wide variation in the newspage content, although slight variations were found which could be explained by the different nature of the news- paper circulation areas of their correSponding political con- stituencies, the districts in which the papers circulated. In all the tabulations, our focus was on coverage of the congressional contests, and the presidential, vice presi- dential and senatorial data were compiled purely for purposes of comparison, so that we could see the congressional dis- 110 trict coverage in context of the entire coverage of all con- tests for national office. If an investigator wished to in- troduce a regional variable, in order to compare the South and Kidwest, we believe the newspapers we examined in hichigan and Tennessee are typical enough for him to do so. II. Inferential Discussion of the General Findings Our first interest is in seeing what findings we can make from the full-sample tables compiled for this chapter and dealing with the urban—rural and incumbency, non-incumbency variables. The "mean" figures for the typical newspaper published in an urban district showed Democratic nominees and Republican nominees each getting 10.6 stories (items), but the Republican nominees had the preponderance of coverage, with 146.75 column inches to 80.96 for the Democratic nominees. In home district coverage, the "typical" Democratic district nominee got 4.4 items to 3.3 items for the "typical" Republican, but the Re- publican stories were longer, as Republican nominees won a total of 41.96 column inches to 36.96 column inches for the "typical" Democratic nominee. This advantage in column inches is slighter than we might expect given the incumbency lineup of Democrats and Republicans in 1960. Nor would we have ex- pected the "typical" Republican to have had fewer items in the frequency count. If we look at newspa_ers published in rural districts, we find a pattern more like we would have eXpected to see in both urban and rural districts. In these newSpapers, Demo- 111 cratic nominees for U. S. representative (all in the state) had 5.6 items for 70.90 column inches, opposed to 12.2 items for 153.85 for Republican nominees. Focusing on same-district Republican and De ocratic noninees, the typical situ,tion found the Deuocretic nominee getting 2.4 itc:s and 21.65 col- umn inches of COV€P&5G to 8.0 itexs and 59.70 column inches for his Republican counterpart. This partly may be accounted for by greater interest in Republican candidotes in the rural districts in Tennessee, where the incurbent Democrat had no serious oneosition in 1960. As for tentative conclusions on this score, we sudgest that the urban-rural dimension is a more definite indicator ‘ of differences in coverage, following party guidelines, than the incunbency, non-incumbency factor which we will discuss below. This is not surprising when we consider the fact that there was only one incumbent not seeking re-election in the yecr we exeninei, 1960. In procession incumbency data for the full sauple, we deveIOped an index for the newsp;pers published in districts with incumbent candidates and an index for the nevepapers published in districts without incumbent candidates. For each grouoing (Tables III and IV), we computed the frequency of items about each condidcte or slate, the total colunn inches devoted to each candidate or slate, rezn frequency and mean column inches. The total frequency of neasured sto- ries in districts with incurbznt caniidates Was 5,727, and total column inches in reasurei stories in the "incunbent 112 running" districts amounted to 7l,636%. In the "no incumbent running" district newspapers, there were 480 stories totaling 8,5255 column inches. The significant figures to conpsre are probably the mean column inches" "mean frequency" and columns of the ta- " ner- bles, which give us the data for the mythical "average paper published in the "average" district. In newspapers published in districts with incuwbcnt candidates, Democratic nominees for U. S. representative re- ceived 8.1 stories nessuring 95.45 column inches to 10.6 sto- ries (or items) neisuring 123.61. In exnlcnation of thee figures, it shouli be said that in the Lichigcn newsneners, the "incumbent" district papers numbered five, and in all five districts, the incumbent was Republican. With the Tennessee papers, ell incumbents were seeking re-election in 1960, and four of the six were published in districts with a Democratic incumbent in Congress, while tvo of the six (published in the ssre district) had a Republican incutbent seeking re-election. So the frequency and column inches figures shouli be corn;rcd on the basis of seven newspapers published in districts with Republican incumbents, chi four n Lers nublishei in districts with Democratic incumbents. Thus tbe preference in frequency for Republicans is nerhcps a little less then this 7:4 retio might lead us to nraiict, but the rough 4:3 ratio in colurn inches appears to be more in line With what the party lineup would indicate. Thus the incumbent advantage seems to be re- flected in column inches of covers;e more than in frequency 113 of items, although even the total column inches appear less than expected. In the remaining newspaper, published in a district where the Republican incumbent congressman was not running in 1960, Democratic nominees for U. S. representative got 13 sto— ries measuring 217.50 column inches and Republican nominees got 25 stories measuring 417.75 column inches. The deter- mining factor here apparently consisted of (l) the expectation that the Republican nominee would win in a normally Republican district, and (2) the expectation that Republicans would win in many districts in the state-at—large. The newsworthiness of candidates, we would assume, is based to some extent on their estimated chances of winning their contest. In this dis— trict, the Democratic nominee in the home district had five items measuring 67 inches, and the Republican nominee received 91.25 inches of coverage in the form of 10 stories. This pre- ponderance again we would take to reflect the dominant party newsworthiness, and the expec+ation of victory for this party's candidate. Usually of course a candidate given the best chance of winning is "better c0py" in a district with one-party tend- encies. The other coverage, in the form of item frequencies and column inches, shown in Tables III and IV, gives a context for the rest of the campaign coverage, so that congressional cam- paign coverage can be judged in terms of the overall distri- bution of items and column inches. Any conclusions we could draw from these data would have ~vf‘" w; . o. _ “its 14.....-‘55 114 to be very tentative, but we believe the incumbent advantage is less than we expected to find. In tie districts where no incunbent was running, affiliation with the dominant party seehed to be a factor in news coverage, and this factor would be closely allied with expectations of victory for the candi- date who is the object of coverage. III. Inferential Discussion of Specific Findings This section of the chapter, which deals with more in— ferences —- these based on the detailed data in Tables V, VI, VII and VIII -- is intended to lay the groundwork for the con- clusions drawn in the final section of the chapter. We first introduce the question: (1) Is the urban-rural readership dirension critical to an understanding of the dif— ferences betwe3n urban and rural congressional districts? On the basis of our data, we believe that it is. Not only do the data indicate different ennhsses in the rural and urban press, but also the press content does reflect important constitu- ency differences. To cite a specific instance, home district coverage in the Kernhis QBZTerElii Aooeal, on campaign activities of in- cumbent Rep. Clifford Davis (D-Tenn), Einth District congress- man, mentioned Davis in only six news itens during the cam- paign, with a total of 76} colunn inches. As an unoppoaed incuhbent, (he was nearly defeated by his Republic;n opuonent in 196? and was defeatei in the 1964 Democratic prinery), his covers;e was limited, while the coverage for inou bent Sena- tor Estes Kefauver (D-Tenn), also unopposed but the winner of "I 115 a resounding victory in the August Democratic primary, over- shadowed that for Davis. Senator Kefauver had 11 news items for a total of 123 3/4 column inches. This would indicate that stature of the candidate and the office he holds may be more important then place of residence (Kefauver was from Chattanooga) or the type of newspaper (urban or rural dis- trict) in effecting the coverage. Turning to Michigan, a two-party state, the same prin- ciple seems to hold true if we examine congressional coverag against the backdrop of the more significant senatorial con- test and its coverage. The Huskegon Chronicle devoted lO items in 1960 to Senator Patrick V. HcNamara (Democratic in- cumbent senior senator) and 15 items to his Republican Oppo- nent, Rep. Alvin M. Bentley, compared to seven items for the Republican incumbent, Ninth District Representative Robert P. Griffin, and three for Griffin's Democratic Opponent, Don Jennings. Inch totals were 121% for hcflamara and 172 3/4 for Bentley, who waged a more active campaign than hcfiamara. Column inch totals for Representative Griffin, co-author of the Landrum-Griffin Labor Act of 1959, and for Jennings, the Democratic nominee for his seat, were 49 3/4 and 28% respec- tively. Comparative figures with presidential candidates —- to give a perspective on the congressional coverage -— are pre- sented in Tables V and VI. It is not surprising that the presidential nominees got the lion's siare of the coverage in 1960 because of the absence of an incumbency factor, because of the presence of volatile issues like the "religious issue" in the cernsign, and probably beesuse a close outcome Wes ex- pected. Th se factors enhanced the normal expectation of greater coverage arising from the status of the office sought by the presidential nominee . A more detailed analysis of opinion factors in coverere for each cetegory was previously made in cemputing "FY? Rating“ tables for Chepter Three. A point which stenis out in Table VII, for the Flint {22:35l, covering a contest in a competitive district where Prosecutor O'Rourke, the Democratic non-incuybent, made were then a token challenge to the incunbent, Representative Chem- berlain, is that while congr ssionel coverage is still over- shsiowed by senatorial, when it is in a two-party state with active campaigning going on in both contests, the place of publication does becene en inportsnt factor. When one candi- date lives in the sane city where the caper is published, he and his opeonent receive greeter coverage in proportion to the Senate contest coverage. To demonstrate this we Lerely cite the fact that Chamberlain eni O'Rourke were given 12 items each in the frequency count, compared to the figures of 31 for Eentley, 24 for Kcfiergrs, and one for Rollin I. Severance of Saginaw, the Prohibitienist nominee for the U. 5. Senate. The same thing is reflected in the figures for total column inches appearing in Table VII. The general campaign cover23e in the Traverse City Recori-nsgle, probably a typical locally owned Slall daily published in a rural congressional district, was quite light. 117 But even here, coverage of home district candidutes for Con- gress was overshadowed by coverage of the Senate contest. In the Ninth District, Griffin (a Traverse City resident) was given eight items and Jennin;s (also a local men) five by the non-partisan 322231‘22512: while Bentley had 10 items and Kchrsre six. Griffin and Bentley conducted more active cam- paigns then their opponents, so here campaign financing might have been a factor in the amount of news events and subsequent coverage generated by their carpeign activities. The ssm kind of analysis can be pursued with other newspepers in the lE-newspoper serple, thOU$h we did not do so in Table VIII. Another point about Table VIII, for the Record-Egale, is that, in addition to the coversge of the Senate carpsign, which Was a Spirited contest, attention was given in the Record—E3533 to contests in two outside congressional dis- tricts. The Record-ggile reported one item on the Kellie- Broomfield rece in the 19th Congressional District in Oeklend County, where hr. Kellie, a retired erhed forces officer, made a challenge to the established Republican incumbent, Rep. William E. Broomfield. This contest gained statewide cover- age through Sunday articles in the "Boothpspers", reporting campaign activities in this Republican district in which the Democrats mounted a serious effort. Kelli; received 14 3/4 coluhn inches of coverage in the one item recorded in the sennle from the Traverse City daily, significsnt in the light of this newspaper's hevin: the smallest circulation of any in the sample. 118 The other outside district mention in the Record-Eagle consisted of an item, the only other congressional one in his sample, about Rep. Victor A. Knox in the llth District in the Kackinac Straits area, combining counties in both Lower and Upper Peninsulas. Here Knox was given 10% inches of coverage in a single news item. Knox, in an unusual event for a Repub- lican, received labor SUpport from the Brotherhood of Railroad Trainmen, and this was the subject of the story. One additional story in the Traverse City sample con- cerned national and statewide congressional campaign activity. A total of 22% column inches devoted to this story indicates a careful editing of the newspaper, which usually had a rela- tively small newspage content, as compared with the larger deilies examined. Tvo stories were devoted to third parties, a total of 15% inches. Further details of each specific news- paper examined in this chapter will be found in Tables V through VIII accompanying the chapter. On the incumbency, non-incumbency problem, relevant data provide a comparison between the incumbent congressman and his challenger. These data consist of (l) a frequency count of mentions for incumbents and for non-incumbents, end (2) a column inches tabulation for coverage of each candidate, in- cumbent and non-incumbent. These figures may be found in Ta- bles VII and VIII with this chapter. Tables III and IV also deal with the incumbency, non-incumbency problem from a more general perspective and were dealt with in Section II of this chapter. 119 In inferential discussion of our modest statistical findings, we pronoee to analyze briefly eight questions about urben-rur:l constituency differences or incumbency factors. (1) Is the urban—rural readership dimension critical in terms of political communication? Reedership studies rode for the Press would indiczte that the neflsnipers circuleted pri- marily in a rural circulCtion zone differ nsrkedly from those with chiefly urban ClPCUlLthH. Afternoon pepers are probably more widely read in cities and metropolitan areas, while morn- ing boners get more statewide, ares-wide or region—wide circu- lation, depending on their nationsl nreotiie. So we conclude thit this dimension is critical in journalistic terms. In political terms, the type of coverage afforded would determine how important it is -- e.g., whet races are covered. (2) What is the interaction pattern between Press and district in the urban—rural context? Certainly, a newsoeoer which is circulated in e e.ecific congressional district may have nunerous resiers who are non-voters -- teen-age readers in Lichigen or Tennessee, for exzmnle, cennot vote ea can their 19-yeer-old counterparts in Kentucky, Georgia or Hawaii. There also will be voters in the district who do not read the Dress, denendin; inetec] on relio, television or word-of-mouth information frow friends. Cr nerhops they may be the apathet- ic voters who seldom vote excent in e cruciol presidential election, and have little interest in political information anyway. Despite these non-coinciding tyoes of reader—voters, a 120 number of informed voters, "Opinion Leaders and Followers", to use the Katz—Lazarsfeld terminology from the Decatur study,1 or "Opinion Givers and Ashers”, to use the terminology em- ployed by Troldahl and Van Dam in their 1964 interview survey of the Detroit area,2 measured on a "Perceived Opinion Leader- ship Scale",3 or the 1952 study by Troldahl, Van Dam and Ro- beck on communicstion networks in the Detroit Office of Civil Defense during the Cuban missile crisis,4 or an earlier study by Troldahl5 carried out in the suburban Boston area in 1952 -- ‘ do take an active interest in politics and current issues of national or statewide debate. These are the important readers out of the entire circulation, and they are the most apt to be registered voters, unless they are disfranchised by residence requirements or other types of voting restrictions which af- fect highly informed and intelligent voters. We conclude, then, that a significant segment of the readership in any congressional district is important to the study of the flow of political commuiication —- expecielly at politically crucial times, such as a time of national crisis as in October, 1962, or at the time of a major nationwide election, such as an "off—year" congressional contest or a presidential election. (3) When we compare congressional campaign coverage with presidential campaign coverage in our data, do we find more straight news reporting or more feature-tyne coverage? Feature coverage tends to be heavier in the more important contests. At the congressional level, it is lighter except 121 in the week before the election, when biographical sketches of the nominees and position statements submitted by them to the newspapers often appear in Sunday editions. Taere does appear, however, to be more background material on candidates and issues in the newspapers than in the pre-television era. While we did not document this with detailed statistical data, it is our impression that perhaps about 5 per cent of the overall coverage of at least the important congressional can- didates, i.e., incumbents or nominees who have gained state and local prominence in other offices, consisted of this type of material. (4) Is there a difference in issues emphasized in the urban and rural press? Economic issues seem more important in the rural press, especially those affecting farm—business interests as Opposed to labor interests. gowever, the metro- politan press (as in Kemphis) gives these good coverage too. It seems logical, and our examination of the data bears it out as far as we have pursued it, that the policy of any news- paper will shape its cover;5e of candidate statements, what- an economic issue, civil (I) ever the issue, whether it b rights, or a foreign policy issue. (5) We doubt that we can comment with any confidence on whether the incumbent, non-incumbent relationship is critical in coverage, because it is difficult to hold other variables constant enough to conclude anything significant with cer- tainty. The demographic variable is obviously easier to ge at, and it probably tells us more than the incumbency vari- 122 able, especially since U. 8. Census Bureau reports have been compiled for political analysis purposes by Conzressionai Quarterly in a Special Report issued in 1964. (6) However, there is an important distinction between second-race incumben s, still in the process of establishing them elves, and third-race-plus incumbents, who are usually (I) considered well established. (This was not dealt with in the hypothesis, but a concern with this problem grows out of a study of the data compiled for the hypothesis.) A soothpa- pers" correspondent reporting in the Saginaw £232 and other Hichigan papers in the sample noted this phenomenon, in con— nection with the NcIntosh-O'Hara contest in the Seventh Richi- gan District in 1960. Eore attention centers on the incumbent nominees seekin their first re—election, than on the senior .5 incumbents, who are ordinarily considered to he e establishe therselves in the House. (7) The coverage does not appear to distort differences already appearing between non-incumbent nominees of the major parties in terms of background (especially the recruitment of leadership factor), in their previous experience in lower po- litical office or in offices similar in prestige and status to those of congressmen, or in visibility of position t0 the DUb‘ 110, if they hold a lesser or equal office. Ve say this be- cause these are primarily political factors, and not communi- cation factors. The "image" is important, to borrow the lan- uase of Radison Avenue, but the language of suns anc3 and s.) U} power is more significant here. 123 (8) We do not believe our data can answer the following question: Where is the advantage of incumbency negated to large degree, and where is it most advantaieous? Is there anything to the probability that an incumbent is most disad- vantaged or vulnerable in his second race (this seems at least probable according to Point 6), holds a distinct advantage in his third through tenth race (while established), and then is relatively disadvantaged if relatively inactive and over 60 in age? We do suggest this as a possible hypothesis for fur- a ther exploration, but we doubt we can provide solid proof for it from the data we have collected for this thesis. IV. Incumbents in 1960 Contests In order to clarify our consideration of the incumbency factor in our analysis of Specific Hypothesis 2, we are pre- senting a list of incumbents who were candidates for re- election in Michigan and Tennessee in the 1960 election, al- though their names are marked with an asterisk in some of the accompanying tables. Incumbents running in Michigan in 1960 were:6 First District -- T. M. hachrowicz (D); Second District -- George Meader (R); Third District -- August E. Johansen (R); Fourth District —— Clare E. Hoffman (R); Fifth District -- Gerald R. Ford (R); Sixth District —- C. E. Chamberlain (R); Seventh District -- James G. O'Hara (D); Eighth District -- NO in- cumbent running; Ninth District -- Robert P. Griffin (a); Tenth District -- Elford A. Cederberg (R); Eleventh District 124 Victor A. Knox (R); Twelfth District -- John 3. Bennett (3); Thirteenth District -- Charles C. D1553 (D); Fourteenth Dis- trict -- Louis C. Rabaut (D); Fifteenth District -- John D. Dingell, Jr. (D); Sixteenth District -- John Lesinski, Jr. (D); Seventeenth District -- Eartha N. Griffiths (D); Eight- eenth District -- William Broomfield (R). Incumbents running in Tennessee in 1960 were:7 First District -- B. Carroll Reece (R); Second District -- Howard E. Baker (R); Third District -- James B. Frazier (D); Fourth Dis- trict -- Joe L. Evins (D); Fifth District -- J. Carlton Loser (D); Sixth District -- Ross Bass (D); Seventh District -— Tom Hurray (D); Eighth District -- Robert A. Everett (D); Ninth District -- Clifford Davis (D). V. Generalizations In part (a) of Specific Hypothesis 2, we stated our ex- pectations as follows: "Both maior party Candidates get sig- nificantly more total coverage (in terms of measurement of actual newspaper space) in rural than in urban districts." However, the data compiled in Tables V and VI for this chapter indicate that coverage is heavier in one of the urban district papers than in one of the rural district papers. Nearly twice as many items made Up the total of campaign cov- . A ‘* ,1 -. erage in the Memphis Commercial gggeal as in the nusxeuon . 1‘ n ‘t’ 1" '.-- : ._ ‘ y _ ghronicle. While both candidates in the ninth “ioniocn Dis trict not a total of 10 items and 79; 001111m 1UCFGS compared to 6 items and 76? column inches for the sole candidate in the Ninth Tennessee District, the Eemfihis coveraée was actu- 125 ally heavier when we consider that there was no contest there while the competitiveness factor was present in the Einth Kichigen District. So this much of the evidence is contrary to the first case of Specific Hypothesis 2. Looking at the data compiled in Tables I and II for this chapter, we find similarly that the total frequency of measured stories or items, 4,159, and column inches in meas- ured stories, 52,603; inches, both are greater for the urban papers' portion of the sample. Corresponding figures for the rural papers' portion are 2,048 items and 27,192 column inches. For home-district congressional candidates, the rural papers afforded a total of 10.4 items (on averaee) for both candi- dates amounting to 81.35 column inches (average). This may be compared with 7.7 items (average) for the urban papers' portion of the sample and their figure of 78.92 column inches .A (avereve). Of the four sets of figures, the latter is the \“' only one 1“at tends to confirm the hypothesis and the first three tend to disprove it, so the weight of the evidence is contrary to the hypothesis, when we consider both the full sample and single newspapers drawn from the sample. In part (b) of Specific Hypothesis 2, we hypothesized as follows: "Non-incumbents get less coverage in newspapers published in urban districts than in newspapers published in rural congressional districts." We can measure this in abso- lute terms and also in terms of comparison with incumbent . s: ~ i her rurol coverage. Non-incumbents hot little COVerwue in e t a . n, a e ma e VII and or urban district newspapers, but accordingJ to isbl s 126 and VIII they seemed to fare somewhat better in the urban press, once again contrary to our expectations. With only two noted exceptions (see Tables VII and VIII), non-incumbents got less coverage than incumbents in either rural or urban district pa- pers. O'Rourke in the Sixth Michigan and Kellie in the 18th Hichigan Districts were the exceptions, and even here O'Rourke had fewer column inches than Chamberlain, his Republican in- cumbent Opponent, so he was ahead in frequency but not in total inch measurement. As a related point, we expected to find campaign coverage heavier in districts without incumbent candidates than in dis- tricts where incumbents were running for re—election. In Ta- bles III and IV, we found the average for both candidates in districts where incumbents were running averaged 8.5 items to- taling 72.82 column inches, while the average for both candi- dates where no incumbent was running was 15 items totaling 158.25 column inches. This does therefore indicate stronger interest or at least greater activity and greater resulting coverage where neither candidate is an incumbent. Some related conclusions which are not generalizations based on the hypothesis, but are worth stating, are as follows: (1) The urban-rural dimension is critically important, else reapportionment would not be such a major issue in most of the 50 states, even at the congressional district level. Urban-rural dimensions in the press coverage of carpaigns are limited by such factors as nationwide press association cov- ‘ie —_-- _ 127 erage, the existence of chain newspapers, and the modifying factor of network television, but they are there and they are important. (2) Longer—term incumbents (third-term and beyond) with seniority appear to heve more political power and to earn greater newspaper coverage in their congressional districts. The tables seem to SUpport this finding. (3) To synthesize these points, we find that the urban- rurel dimension outweighs the incumbency, non-in umbency di- mension in importance, whether considered in historical terms, political science terms, or communications terms (such as the sociology of mass communications). Finally, for further research, we think a close study of Senate races and of presidential contests where party compe- tition is close (1950 would provide a better case than 1964) would be rewarding, using perhaps a different kind of content sample (perhaps from radio station—TV station logs or from party records). But we will develop this suggestion in the final chapter. 128 w:.mom m.om «wwam mam manonu ”monosz mm.m 0. mm o Ampmpm no uzov mw>apmpcmmonnom nwoaapsamm mm.mm m.m e\m mom mm Ampapm no paov mmpapmpnmmmunmm cauwuooawa mH.oe H.e «Hem mm Aepmpm so psov mHOpmnom caudansaom mm.una H.5H «seas ems Amumpm no pace mAOmemm oapdnooama mm.¢m H.m 3mm mm Sesame unocdmmumnnm om.mma o.mH mwma moa nmzonzmmam pnmcamoum om.H: m.m :\m mam mm mmmuwnoo you monaaoz nmoaansmmm uoaupmaa mm.om 3.: :\m mmm an mmmnwmoo you mmzaaoz oapmnooaon podnpmaa ms.eea e.oa «mmoa em mpdpepammmnamm .m .D you womCaBoz cwoaansnvm om.om w.oa :\m own :5 mpaumpnmmonmmm .m .D you moccaaoz aduwuooawo Hm.neom n.wea «mused Heoa mpeam oweoqunoxdz ow.~HHN u.aea oases «mad «seam senescensemsnos mmzozH mesonH nasaoo honmsumnm casaoo Hence news new: dance honoswonh mwanoboo ho pomnno mpoaupmao menu: ca nonmaansm muonwamxmz muopoam amusmucmnuo nfi>H noummnov H mqmH umpgegov H memes mnopomm Handmlcmnub om.mmH m.ma Ham mm masono pmmuman 0H.m m. mnm e Amemem no p900 wmpapwpsmwonnmm Canaansamm mo.ma w.a «mm m Ampwpm no psov mmpnuwpcomonamm oauwuooamn 00.0: 0.: 00m 0m Ampepm no p500 mnopmnmm cmoaapsaom nu.e0H o.m e\m mmm me Ampmpm no p500 mpoprom oapmpooaon mu; om.mm mi won an usage pcmcdmopmnxm 1. om.mwa o.NH Ham ow umzonsmmdm pcocamonm on.mn o.m mwmm o: mmmnwcoo you omzaaoz neodansaem nonhuman mw.am :.N «moa NH moonwnoo you mondaoz oapmnooamn poanpman mm.mma N.NH «men He opdpmpsommnaom .m .D you moonaaoz :moaansaom om.ou w.m wzmm mm mmpaumpsommummm .m .D you momzaaoz capmnooamn 0n.eoaa 0.05 mess mam upmam eweoquaowaz om.mwma m.ooa Hmwm mom mpmam nomnsohuhcocnmm monocH mmsosH caoaoo hoconvmum casaoo Hence new: new: Hmpos hocoswoum mmmumboo no pomnno mpodnpmam Hmysm ca vonmdansm unmawmmzoz unapomm Hmuomnzenuo uA>H nmummsov HH mqmea m:.om: 0H.0H mm.mm 1.mm.mem 1.00.N0H ms.mn 5 ma.mm 0a.:w 0H.mea ow.wua n0.HH mmsosH CBDHOU 9602 mma.um II moauopm conzmmmz Ca monosH caoaoo meow In moHHOpm nonemamz Mo mocmovmum Hence :.a: :\m :mmm m.H wmm mo: AQQH N.na :\n mmma 0.0H dam N.u e\m mum «.0 e\m mm: 0.0 Hm: 0.HH mass o.:H mwm N.a :\m mm monocH hosmsomnm cesaoo 2mm: proe mom em mm mm on o: m: mm mm kuoe mommodoum Amwanmboo munmm chase .mmanopm genome .mmHSpmom copwaomnuwawmawov nmsuo wnonuo>ou cmoaapsaom mnosumboo oaumnoosmn who>uzm one madam mpdpdpoe Hanoapmnanmwuo cwoaansamm muapapoe chodpmnacmwno oapmnooaon mpwsmm .m .D you mmmnaaoz cmoaansmmm opwnmm .m .D you momzdsoz oapmnooaoa mnmommq Hmzoapwz cwoaansmmm mnoommq Hanoapmz odpmuooaon muovomam vowooaans mwmum>oo mo pomnno Aomznapzoov mpofinaman Hmnsm a“ cogmdapsm mnmamomzmz ufi>H nepaesov HH memes mpopomm amusmucmnua 132 Hy.0mm 00.0 00.0N 00.50 0:.NOH :0.yn y:.0uy 00.0: 00.00 Ho.mma 0:.00 00.0000 00.0:0y mononH Gaoaou yam: :.0m #0000 00m 0. M00 0 0.0 0000 00 0.: «my: 5: 0.:H «s0:y :0H 0.5 :\0 :0“ :0 0.0a mammy 000 0.: H00 00 0.0 000 00 0.0a :\0 0:0y 0AA y.0 0:0 00 0.0Ny onemy 0y:y 0.50y «0:00N many mononH hoyosdmym £85000 Hmpoe can: Hmpoe hoymadoym mmsoyu pmmymvyH Ampmpm mo 950V mmpypmpcmmoymmm amoaansmmm Ampmpm yo 950v mmpypmpymmoymmm oypmyOOSmQ Aopmpm yo 950v myouwyom ywoaapsmom Ampmpm yo psov myopmymm oypmyooamn mmssya anonymoymnxm ymzosyomym anonymoym mmmywyoo you ooyyaoz cmoaanzamm oyypmyn mmoywyoo you moyyaoz oypmyooaon poyypmyn mpypmuymmmymmm .m .D you moonaaoz amoyansmom opypwuymmoymmm .m .D you mooyyaoz oypmyooamn mpmam mwooquyoxyz madam yomysohlhooyymm owmymboo yo pomnno mmpmoyncmo pamnasocH spy; mpoyypmyn yd cosmyansm myoamamzmz «A>H umpeeeov HHH memes myouomm mommnasoyHICoz .moyonesoyH .0... .32.... ‘ q '- 133 No.mwo 00.00 00.00 00.000 :0.000 00.:00 00.00 00.000 000.00 00.00 mmsoyH $854.0 0 Cfim E 0.0: 0.0 0.0 n.0m 0.00 0.00 u.m 0.00 0.00 n.00 0.0 0000000ym :00: A0000000000 0000000000100000 ‘ myopomm 00009800001002 .0000980000 «0:00 00: :\m :00 «0000 «0000 «00:0 00:00 «0000 00000 :\m 0000 000 m0non0 :85000 00909 00: mm 000 0mm 000 3:0 00 NNO 050 000 3N 0muoB mocmaw0ym mwm0.0m in m00yopm 00y5000z C0 00noy0 285000 0N0n II m00yopm 00y50002 no 00:0000ym 00909 . m0w0y0>oo hpymm 0y0se 000 00m ammoas .00y3p00m 009000mlyw0waamov ymspo myoyy0>ou :0000noa0m my02y0>ow 00pmyooa0m mh0>ysm 0:0 000om 000>apo< 00C009000y0wyo c0000nsm0m mudpapoe 00no00000s0wyo 00pmy0080n 000:0m .m .D you m00y0aoz £0000nom0m 000y0m .m .D you 00020802 oapwyooaon 0y0000q 00y00002 c0000psa0m my00000 00yoypwz 0000yooa0a my00000m 00w000ynb 0w0y0>oo 00 900090 0000 0003 mpodypw0o 00 002000900 0y0mwmmz0z ua>y 00000000 HHH 00000 ’mm '~ " " —_._,._ '52.)» " . fl oo.m00 oo.NN om.m0 00.00 00.000 0., 00.000 1 00.0mm 00.00 00.00 00.00: 00.000 00.0000 00.0000 002020 280000 2002 00 00 00 00 0M0 0N0 ho200dmym 2002 000 m 00 0 000 0 00 0 :\0 000 0 :\0 000 0 000 00 000 00 00 0 :\0 00: 00 0000 00 0000 000 00000 000 002020 280000 00208 00209 ho20000ym 0200y0 000y0020 000020 20 p000 00>0p0p2000y202 2000020202 000000 00 p000 00>0u0p2000y20m 0000y00802 000020 20 p0ov 0yop020m 20000D0Q0m 002020 20 2000 0yop020m 00p0y0080o 2080ya p200000ymlxm y0zo22000m 2200000ym 000yw2oo you 0020802 0000000000 00000000 000yw2oo you 0020802 00p0y00802 000y0002 0>0u0p2000y20m .0 .2 yo“ 00020802 2000020002 0>0u0p2000y202 .m .D you 00020802 00u0y00802 02000 0wcoqu2oN02 0200a 200220030002202 0w0y0>00 20 000020 0020000200 020220020 0002003 0uo0y0002 20 002000200 0y02020z02 0209002 002028002HI202 .0020280020 «000 00000000 >0 00000 mwmmm In 0002o0m 00220002 20 mo2o2H 282Hoo om: II 000200m 00220002 20 202020022 H0009 om.mm0 a: mmmm a: 000020>oo 00202 02029 .000200m 00202: .00220002 0000Hmm|2m00280ov 20200 mm.mm H 0\m mm H m2o220>ow 2000H22202 om.mH H mNH H m2o220>oo 0000200800 oo.mmH m 00H m 020>22m 020 mHHom 20 1) .1 00.000 0 000 0 200>000< H0200000020w2o 2000022202 00.000 0 000 0 00000000 H0200000020w2o 0000200809 00.000 00 0\m 000 00 000000 .0 .0 202 00020802 2000H22002 00.000 00 0000 00 000000 .0 .0 202 00020802 0000200802 mm.NmH m MNmH 0 0200000 H0200002 2000H22202 00.000 0 mmmm 0 0200000 H0200002 0000200802 00.0 o o o 020000Hm cmmcmHm2b 002020 002020 282Hoo 002020022 282H00 H0009 2002 2002 H0009 202020022 00020>oo no 000020 A0022002oov 0000000200 020280020 0002003 000020002 20 00200H222 0202020302 0200000 0000000000u0oz .0000000000 ”000 00000000 00 00000 .Tfi w r «nun-t“ g—L-Ivf’i‘.‘ “‘5‘. 3» 136 mm o m:a o :\n mma «an m wz mmoa mwamm mmmm mmsozH naaaou Hmpoa mmm Amopogm and mmaHOpmv meopH no .02 Aummoaaozpv “av mnd>m .q «on .nom nu poanpman nausea Aav unmanm unspu4 “my mommm Haounmo .m .Qmm II poduuman pmnam Amy nodumnm moaowum .4 “av umpsmmmm mmpmm .nmm Amv nmzosnomam .m uswdza pgooammum Amwooqncoxaz no pnoamsm CHV Amy mmnog ponmo mucmm noomwm¢pa¢ Any nomcnoh .m nooshq .nmm Ampswam mmumpm Hmcodpazv mansmm .3 prno .poo Amv noxdz .2 oumsodm pcmcammnm moa> ADV homssox .m snow .cmm muswdm aweduaaom no madcaocmo Aowma «Hmomm< Hmaonmsaoo mannamzv uQ uQ¢HBzmmHmwmm h n o w m u m o “mamm poaupmaa manna an :« mopmoaucmo HH< you owwnopoo proe «mflpadnm> Hmuam-cmpnp ufl>H umpamnov > mqmoo stcapaoca ham saupm no: smammmnwnoo comoaaonp Anamoaaozpv wen m Amy mdpmm ogouhdao .amm nu peanpmam sugaz Awomommosmv Wm H AQV mmmm mmom .mmm Aomscapcoov In poaupmdn sumam "A«onmmmmuzoo masozH Amoponm nasaoo cum modnoumv onsmdm Haoapaaom Hapoa maopH mo .02 no mpwvawcmo m A o w m p a o Acmsnapcoov Aowaa “Hammm4 fladoMoaaoo magmamzv “mamm poaupmdm easy: nu ma mopmodugmo HH< you owmnwpoo proa «magmanm> Hansmncmnup ua>H Ampmwsov > mam¢a . _ fimN .N :\m 0: mm mm 0 «HA 0 «om :\m «ma mama «an moa Mam a: mONH 138 «oaza mososH nasaoo Hmpoe “coma "maoaconno c m MO 303 G303 l\ OW NOHO HH b-m 0H mNH EmpH ooz uSmQEfiocH mm902mnt Amy mmnaccoh can “my magmano .mwm*uupo«npmda sum Amy gmochoz .gmmuwm Anv mumm.o .amm*uupo«npmao sun “av xnmao .H Hmsamm Amv cmmcmson .mom*u-»odnpmdn can AQV mamam masons Amv “momma .ammtnupoanpmaa cam Amv mmapnmm .z ad>a4 .mmm AQV anaemZOS .> Moanpmm .nmmt Amv nmzoncmmam .n unwazm pcmoammnm Amwooqunoxaz no pHOQQSm :Hv Amy mwooq ponmo hunmm nocmmmwnad AQV comnnon .m SOQShH .nmm mmmnaaoc human chase Amy Coxaz .2 camcoam pnmcammnm moa> AQV hnmnnmm .m snow .Smm mnsmam Hmodpaaom no mpmoaocwo uA¢ZOHmmmmozoo «Q¢Hmoe “AdHazmQHmmmm h n o w o p m o ommxmazv ummwm poanpmaa Hmnsm a ma mmpmoacsao HH¢ you mmmnm>oo Hmpoe ”mapmdgm> amusmncmpnp ufl>H “mpgmnov H» mam«a " .flf— pnmnasosH mmponmat moamn aom mg¢aoa :\m um : “mpomz Hansmuzapnp uan nmpgmnov H> mqm49 ‘4‘ WT.'V?W . a... w. T- '— _ — _ “ s_~n-~--n"fi—(“Z—'i~v’r‘h" — O O o o . o o ; :\m n H mm NH «mHH NH mm H mm: m mm H O H. mm H mm H mm H mm H «mm N o o “H H mmmo mm H H um: Hm mHom 3N mmsonH nasaoo mampH Hmpoa mo .02 «maannmb hononaaocH HOV nocnmm mmHz Amy mmpumm mcHnssu pamnasonH 02 Amy SmouGHoz ADV mumm.0* an omnsom.o Hmv :HmHnmnamno. HOV noaaom Amy whom: “av mausm “my Smemmomt ADV NHMHO Amv GmwsmSOht any ochmm Amy Roommzt Amy onmcnmuo “my uononnowz: upoanmHa «poHuumHn “HOHupmHa "poanmHn ”HOHuamHa upoanmHm «poanmHa “HOanmHo sum can sue sun so: cum cam pmH nmmpaoo mummom you Hmpoupsm ApmHQOHuHDHSOHmV monmnmbmm .2 ndHHom Hmv moHpnmm .2 nH>H< .aom ADV wumawZOZ .> onnumm .cmm* ouamHm HmoHpHHom Ho mvmchzwo Homma «HflflhfiOh undam 029v HA¢ZOHmmmmUZOO «A4Hm0942mm h u o m m p m o mmoum hocmzpapmnou camp: as» :H mwmnm>oo pcopasocHlnoz .m> ucmnasocH uH>H umpamnov HH> mHm.honmnasocH ~H>H umpgmsov HH> mHmH< .omm dmoa o HQV onwaozoz .> onnpom .nomt «HQHmoe¢zmm mononH casaoo maopH onstm HmoHpHHom Hopoa no .02 no opmoHocmo h n o w o p m o HoomH «admmflrnuqqmm tho monoponav moonm hocosqumcoo Hmnzm 02» SH owmno>oo pconasonHunoz .m> anonasocH ”memHnm> noomnasoon uH>H nmuomsov HHH> mom unopasocH umflflmdnm> hoEQDBDOSH «A>H Hmpmmsov HHH> Mdm<8 145 psmpasozH monocont mum on mHoaoa %ma N “Hononowv moHpnmm oane "mmmeo maa mH mpmoucoo omsom non Hopopnsm «mm H HoQonmonmcoo oonouwpmnHmcoHpmz :\m :H H Hov mHHHmH o 0 Amy oHoHnaoonmt upoanmHQ npmH o o Amv Haonnoz Hooczaocoov o 0 any manmano .mnz* upoHnumHQ Spud HA¢onmmmm0200 mononH casaoo maopH onsmHm HmoHpHHom Houoe no .02 no opononoo h n o w o p o o Acousaocoov Hooma «mamomuonooom mpHo omnobmnav moonm moomspHpmcoo Hmnzm on» :H owmno>oo pompasoaHuooz .mp uconasocH “oHnaHnm> nonopssocH uH>H umpgmsov HHH> mHm mfizm>m cq mpzmpm mzmz maapoppsm mpnmm chase pcmsm>oz mnopomam cmwumamcb my pmxoda m venisoxaz any umxoas Comanchlhomncmm «amapnmcammnm mmpmcacnmo hp mmauommpmo «amommd Hmaonmaaoo manaamzv mpnmpm :ompcaum: .mp mpzm>m gauge: uh> nmuamnov 41H mqm4e Lung? 1" -—~;: " 171 pnmpanonH mononont .cH .Hoo a\n Haaa .nH .H00 and .nH .Hoo :\m Hoaa maopH mm maopH on maopH mad mgapo4 moonmnomnoo monooogm mopuvaccwo Hm:o«pmN«:mwno msoz a mmaaamm an woauowmpmo m o a n o w o p a o p n m > m mazm>m =Am sumpsanm: .m> mpnopm :HmHOx "mapmdnm> mummpm mzwz «A> nopmmsov Hpo< ‘ mmoCmnmmcoo mmnommmm mmnsmHm HmoHuHHom Hmcodpaanmwno mzmz a moHHHmm up mmHnowwpmo m m H u o w m p a o p n m > m mmoeo< MB¢DHQZ¢UIZOZ mom mBzm>m sq ”agape mzmz ,a> prawgov m.H mQHHB -55 {Two {sfi .cH .Hoo :\m umH mampH mH HHmpOH :H omosHocH pony mHHom asHHmo mo :oHpcmz «mpoz .nH .Hoo mmmH .nH .Hoo mm .cH .Hoo :Hm mHHgoa mamHH m maopH NH mampH mm .:H .Hoo :\m oH .cH .Hoo Hm .cH .Hoo «3m ampH H mampH m mampH m :oon pmemmm= .cH .H00 0 apr H :omnwpmpm HmH©4 z, _ .nH .Hoo «OH 7. mampH N mewHHHHx penumboo l I, .:H .Hoo a\m m nwanpHm mgamHHHz ampH H .sago mow munsoo HpHmnm .cH .Hoo mm .cH .Hoo wee ampH H mamuH w muoo nonmzmm .cH .Hoo mm .:H .Hoo NH .sH .Hoo wwm ampH H mampH N wampH w nopmcHHHm nocnm>oo .sH .Hoo MH ampH H mHHm noanm ngoomHam mwcHazonu ccw mmHquHpo< moosmnmmcoo monomoam umnstm HmoHpHHom HanovaNHnmwno mzmz w mmHHHmm hp mOHnommpmo m m H u o m o p w o p s o > m mm080¢ MB¢QHQZ¢OIZOZ Mom mezm>m :H mpCmpm szho: "memem> mpsmpm mzmz uH> pmpnmsov muH mHm omoHooonav mnowcmHHmomHz Hpumm opHsa unmam>oz muonomHm omwomHmnb my pmHOHs o coHnnotz Anv amone Comanchnhcwunmx «Hmenmonmnm umpmochao hp mmHnowmpwu Aome “Hammm< HmHonmaaoo mHsmamzv mpnmpm zomannm: .m> mpampm :Hmno: «mHannm> mpcmpm mxoz «A> Hmpamnov UIH mqmda __ w w—Wfim*y' fl~._‘_ - : o - ' . . . c o 3%MH a: mampva ampH H mHapopnsm 2H0 mampH N mampH N AQV nm>smmmm .cmm ::\m mm :5 _ mampH N amuH H Haumv meaa .gom . _ :: ampH H HmIHV moomm .mmm ”Hanoumsmm Q Hmconmmuwcoo :HnHo ::\m NNHH gmaom :HNNN a. mampH mN mampH Nw mampH NMH mampH m: mHmpopnsm mu Hmuouoon mmmmmnzma swam .mum»o> omoHooosov mampH N l maowstHoomHz :wNQ :d\m Hm :NH 3mm maopH N mampH N mampH m sopH H Nahum uana :HNN :HmN =:\m mH :a\m H psmaopoz mampH N mampH N mampH m ampH H mnopomHm nowomHmcp gaxm NON gmmom ngmm gmNmH NV pmHOHe mampH m mampH N: mampH mm maopH mN o coHInotz ::\m mmN :zwn sand: a:\m NNH any pmona mampH 2H mamuH mm mampH mm mampH mH Gownsohlmcmcamx _ «Hmenmonmnm magnummm mnoopumo Ampsocnmm .HonHv unsouw é mnazHoo nopHcm as» manummm a mmpwchnau nmowm .mHmHnopHum op mnmupmq mommmHmm mmmnm hp mmHAowmpmo m o H n o w m p m u p n o > m mazmkm stBZHmm: HoomH "HNNQQN HmHoumeaoo mHnmamzv musopm :omannm: .m> mpnmpm :Hmno: umHannm> mpnmpm mxoz uA> umuamnov ouH mqm4a 422.: ‘¥.w~_wwwm“fiv—w ...._ . .r__ .H_ H. m:‘. .- I‘!‘ -. 1‘ ._,.'~ . u. "7‘" ‘ I 175 m 0 HomSSHpGoov HoomH ::\m noaH wmmm mampH mm mampm ooH ewonH =:\m Hm maopH m mampH mH g:\m N ampH H ,gmN ampH H awn EmpH H ::\m QH two: EmpH H mampH HH :3\m m5 :WH mampH m meopH m :mH :MQH ampH H waopH m :0 ampH H accounmo a mussHou HopHom as» .mHaHnoaHom on mumupmq o w w p a o p AmmnocH CESHoo 2H mum mpcmamusmmmz suwnmq HHNV .mmn mampH Ho cmHOH mampH mH a. saw mampH m gwwm mampH m gmHH mampH N :«w mampH m :MHH mampH N Ampsoocmm .Hosz mannmwm w :d\m MN mfimpH N mmmmmHmm mmmnm Q 0 P mazm>m :QMBZHmm: umHannm> wuam>m mzmz m madaoa Qzm sumannm: .mp wpcw>m :Hmno: ”H>-nmpamnov onH mHm manhooaoav :wwH ampH H muzmpmmponm HanonHH g m sopH H mpqameHH pusoo gm aopH H won .H .2 .>om g a: mamuH n .oOHm pmenmmg gmH ampH H mumpmaaoe :H\m m aopH H Hasouo gam meopH m oHouHm¢ obHum>nomnouv U :Am aopH H GOHpo< oHpmnooaoQ u HN acpH H anHouoxoom .>oo now mcmoHuoa< ,o =N aopH H sopzanom .gmm .:\m MHN mampH mH nmsosnmmHm mu :mm: maopH w comxoah .nom : owN mamuH 3N msooanHoomHz can mamuH H Hpuae .ov omom : mm maopH m nopsaNmm .cmm zwmm mampH : maaHHHHz .>oo .H\m mm mampH m nomcmpmpm HaHoN :mH mampH N nounoz .Som :o mampH o mnouoon camcoHanb .mH ampH H mmoo gmHn maopH N unsung .m huuam "mgaono a mnoxwonm opmuchaouqoz .HonN maopH :oN Hmv mmcoHuconHz .HmmmN maopH mHN Hay nomnsohnhoonnom :omno mampH on ”Hmmpunon ump20>m HprnouHmonm Hmpoa wnHUSHoGHV HaHpcoonwnm monsmHm stanamo a mopmuHocao hp mmHuowouao ZOHB mpzopm mHmz mpcm>m mzoz HH> nonmasov «uHH mqmH< .nmm "upsmpm HmHnopwsom Hapoa :mHN mampH mN any mumBMZOZ .> MOanam .nomt "mummbm HmHnoumem ZOHBNBzmHmO H mpCmpm Och mummpm mzcz uH> “mummsov <-HH NHm mpzmpm mzmz gm gmH gmN =:\m H gmH ”Wm =:\m 00H :mw gmN: smN =:\mvum :mHm gNH mempH ampH ampH sopH ampH mampH mampH mampH mampH wampH ampH mampH mampH ampH ADV HHmwcHQ N Amv nsz .mp2 Amy pposnmm Amv mwchnmh any Cmunwm mmHz Hmv smochoz Hay mumm.o Amv cHpronamno H Hmv oxusom.o Hmv ouch Hmv cwaumom Amy cmmGGSOH Amv noowmz H N N H N d‘O\ NN H H H H AQV MOHsonsowz usumH unpaH ”ngH ”spa «sum «Sun «2pm ”can «n»: «can «ch «pmH mmmsHaoz poHupmHQ Hanonmmnwcoo pnm>m no Homcoam hp mmHuowoumo «HNCADOh chHmV mucm>m uoHCmHHOIHOanmHQ .m> mpcm>m meQNIHmcoHpmz uh> umpmmcov muHH NHNNH “.mmnonH nasHoo zH mum mpcmamuzmmmz npwan HHm mpcsoonHmooq you Hmpoe :mNN mampH m Hm cam av poonB Hmecoonmnm mpcmbm hpcsoo cam Hmooq ZOH8¢BZMHmO BUHmamHQ Homanpcoov Home "Hacksaw chHmv mpcm>m oopcoHHOIHOanmHQ .m> muzwbm mez mpsm>m msmz uH> umpmmsov mnHH mHmNB :\m an MH Hapoa HmHnopmnomuHNSOHumonwnoo Hmnmnoo mm H Amy noHuaum HmHuanm .4 m N Any nwbsmumm mopmm .smm moonHaoz opmnmm .m .D «a: oH mopauHunmu Hmaonmonwgoo HH< no“ Hmpopnsm o o mouacHocmo HHN NuN .n-n upoanmHn Anuanham SHHB wanmonnav «n N Hav mmum mmom .mom pOHupmHn sprm Hammocapcm: mwunHoch O *wN o Hmv nmxwm amazom .mwm poHupmHn vacuum 8 1. HHH N Hmv oomom HHoanmo .m .gmm HOanmHo pmnHm MOSOGH SSH—HOD mawvH ODNUHCSNO mmoz¢m¢mmm¢ Am ammH>mHme woocmnwmQQH >9 and mmHHHmm no mmmnm>oo o» couHaHH .mHno mwomm HNGOHmmwnwcoo now mama «0902 Home « Manama mHHHPHocmv mmonaumogm< Hma0mnom pomnHQ .m> mpcm>m wzoz cmmHPmea "oHanua> muco>m mzwz uH> nopmmsov HHH mqm¢a 181 mmnozH sasHoo :\m moon mamuH HNN mmnonH QBSHoo :\n H:: maopH mm mmsocH CBSHoo d\m NH: mamuH mm mmnonH CSSHOU MN mampH N mmzosH cazHoo owNm mampH omN mmnonH sasHoo «nu: maouH an mmsonH sasHoo :\m NNNN mampH NmH ”mwmum>oo ”assoc umwmnoboo «pusoo “mwmumpoo «pnsoo "owmnmpoo “unsoo “wwwumpoo upcsoo «mwmnopoo «pnzoo «ammumpoo «assoc Home no pcsos¢ hocmsvmum mo pcsoa< hocosvmum Ho unsoa4 honmzvmum mo pcsoa< hocmsvmum mo unsoa< honmsvmnm no pQSoS¢ houmSUmnm Ho pasoa¢ nonmaaoum mqmz¢m Hoo mmustm hammoq Hmpoa H.090 .mamHHHHz .poo .ouom .anpmmpmv mumcwmq hunmm mumpm vcm mmmcHaoz Hmconwmnwnoo .cmammmnwnoo mnocmmq Hmooq mmmDuHm mmmmmq OszDB¢mm MBzm>m wwwnmpoo mH> Hmpoa H.0um .comxown .cOpuoz cmeHmso .wHOHNGom .m .D .NANENZOS .HmeCOmV uncommq human Hanoprz 02m mmponozmo HwHHOHNSmm Amnmnamz mpmHm mpnmm cam HmzoscmmHmv pzmonmum no mpmoHonwo Hmenoonmnm moH> wad HaHpcmchmnm azammmm mH> mmmmx mezm>m «Hacksaw mumpm mchcqu owmum>oo mzmz CH monomm m mm mpcmbm pa mocmmmnm mH> umemHnm> manopm mamz «A> Honnwsov >H mqmoo mpmoHccmo Hmpoa «HpmHum> mmmgopHpHpmgaoo «AH> nmpaagov H MHmea 203 pzmnasocH movement :\n mam mm Hapoe «mm m “as mHHHmm s\m s H Hmv eHmHeaoonm* “poHapmHn numH o o Hmv HHmnsoz cam an mspHeeru .muzs "poHupmHa numH o o Ame xano new.HmV HewsHmmq* «pOngmHa spoH «o o Hmv maanom a: H Has HHmmgHo* “poanmHn npmH mmconH :Esaou maopH Avonannoov «Amaze mmomm chonmonmcoov mwmum>oo ovacdcnao no mommavmum AomssHpcOov Aowma «chm:Oh mumpm wnHmcmqvaImemucmoumm cam hocmSUmnmv poanmHm H620HmmmnwsOo mbdepmmaoo a :H o mnwpoo cpmoHozmv Haves memHnme mmmzmpHpHpogaoo “AH> “mpgmnov H mHmea " ' "' ' " far-f» ~Y$bmwzep w. " "" , , 204 Am woo. m :mm.am Axoosu Honpnoo homnsooev u mmo.HN mHHHmm R mHm.H oHonsoonme R nan.a HawwcHat m Hum.n cmmm R mmm.m mwannmh R mmm.m zHeeHuos m sHm.H smpumm R won-O dhmm.0* m mmm.um mxnsom.o a www.mH :HaHnmnsasot R Nam.m mansm u Acmvsaonoov A poHnumHQ H unmnasocH mmponont ”nonmasoamo onwm cmoHnm no Hapoa Hmspo¢ scum :oHpmHPva "poHupmHn npmH ”poHupmHa spmH “pOHupmHn spOH upoanmHn :pm «poHupmHn sum ”poanmHa nus upOHHpmHn nae «poHupmHa a»: mwmnmpoo H¢QH§05 macaw wanHmomm mmpmoHcawo you cmusgaoo mowmpnmonmm owma ”Hann50h mpwpm wnHmnmAvanamwwpcmonwm was hocmsvmnmv aconwonwcoo mprHpmmaoo m SH owwnmpoo mpmchsmo Hmpoa mHnaHnm> mmmcmprHpmmaoo «AH> umpmmsov H mqm¢9 —‘ '~ -".-~L- 4.. ., 0" --":¢:r-.:' - a” 1. ' a: $5.25, "'~ a- .- . , . Il- -- III-null.— Jul-.- mm 205 :\m NH 0 o 0 «mu 0 w mesosH GBSHoo “coma pumnasonH mmponmnt :H wwwumpou kuoe Anemonaozsv 0 Has meaa cnoeeHHo* "90HupmHm zpsz Acmmoamocov 0 Has ppmuopm Ampmmv .< phenom: “uoHupmHa npsmHm Anemoaaocbv o ADV means: 809* «poprmHQ summpom Acmmoaaocbv m Hay mmmm mmoms «poHupmHn sprm Hommoamozbv o ADV nmmoq :opanmo .nt ”poanmHo apmHm Aommoaaocpv o an mchm .H mos: ”poHnumHa spnaom Anemoamonav o Hav umHNmum .m .e: upOanmHm cpHse Acmmomaonpv 0H Amy umxmm vnmzomt «poHnumHn uncomm 0 Has pannm nsnpu< H Amy mommm HHonnmo .m: ”poHupmHn nmuHm mfimvH mopmodcflwo mo .02 H w n o H m m m n w n o o mo¢mm>oo mBNoGMV uOHApmHQ muummuono a :H mwmnmpoo muwchQmo Haves «mHanua> mmmnmpHpHuoaaoo ”AH> umpgmgov HH mqmee ‘ - , V -“Cpmv w '..>.‘... '1'”. S.“ 206 uzonasozH mononomt Ammoo. «nonwasoamo and: nmoHum no QoHpmH>mnv R moo.ooa Axomno Honpzoov swapsoonom Hmpoa Ammonwzou Spam .wpdzom .m .D :H Sozv u mmmqwmn Hay mmam mmom .gom* ”poHuumHn npuHm R omw.mm Amv Hoxwm unazom .QoM* "DOHHpmHQ uncoom u mmn.o Hmv moose HHonnmo .m .amm: “pOHupmHo pmnHm owmnwpoo Hmpoa no owwpnoonmm mwwnm>oo nmbHo mmpwuHccao mBoo mo moNonmv InpOanmHn hpnamlwno a sH mwmnoboo oumuchmo Hauoe "QHpmHna> amazopHppraaoo «AH» umpassov HH mHmee ; a _ _ _ h. Hacksaw psHHm Hoe Hapoppsm nHml mpmmpcoo :meson HH< :Nm am d. H mmammmumnoo Hmnman mcHz msH Hay mHHme «ammHson npwH 208 0:02 «:mmHson spuH mcoz ”cmenon :pmH Ame HHmman «cmmHson numH mH 0 MN 0 Amy “Hmz mHoH .mnz ”amesOHz npsH mcoz ”zmenon :pmH Ame ppmszmm «cmeson numH Hacksaw anHm mcoz ”smeson spHH mpmoHonmo mmumoHucmo HmCOHmmmnwcoo mesocH QSSHou new mampH nonmamzmz mwoo ma Hmecmchmnm ”HmomHv mmsHe a oonmppaso ssz emsmaaoo Homev Hacksaw szHm muoHupmHa mprHpmmaoo no new a :H omummaoo mwmnmboo mpmochmo proe «memHnm> mmmsmpHpHpmmsoo “HH> “manages «-HHH mHmee .nH .Hoo om:.mm:m msopH NmH “mHgsmm :H proa ome.NHmH omH mmmww mmoocmppmno nee Hmpopnsm «n H Amy mesmh «mmmmmccme npSHz :\m N H “my nonmm :\m N H Any wchm "mmmmossme nunsom «a H Axosm mawHIQv anNmnm MHm.mom c: any umSmmnsa mmm.now m: Amy mooum "mmmmeSme cpHss o, o 0 Amy nopmwcH>HH m” mnH n “my memm “mommoscme ocoomw «o N Ame mumpHHm 4m N Amy amHHHsa «oommmzsma vmuHm oom NN AmpmoHccaonnocv hooznmm unmonmnm MNmH HH Amponnsmouconv nmzonammHm unmonmumnxm mN H mmomm Hanonmonwuoo HH¢ m H AmHHGCOHpmzv mpmnooama HH¢ :\m Nm m mmeHB mpmnoosma mmmmmzcme HH< wwoocmppmso :\m WN N COHpaNHsmwno anHHnsmmm mmwmmSCwB mpmoHocmo mmpmoHonmo Hanonmmuwnoo mesonH nasHoo pmm msmpH nmgmgmzmz moemm>oo meenHaz HmeCmnmenm «AmomHv mmaHB smoocwppmno anz cmpmmaoo Homev Hacksaw ucHHm mpoanmHQ mpHpHpmgsoo mo new a :H cmnmmeoo mwmnmpoo mpmoHesmo proe «memHum> mmmsm>HpHumasoo ”AH> nmpamsov ¢|HHH wqm¢e 211 AGoHpmemQ Nam. W Hproa Honpnoo mom.mm mm Hmcn50h unHHm you Hmpopnsm :No.: H mpmmpnou saston HHe Hdm.H H 608mmmnwnoo HmnmnHH mnHz mmm.m H Ame mHHHmH ”nmenon spmH 000.0 0 mcoz unmeson SpNH 000.0 0 enoz «cmmHSoH: npwH mmH.H N “av HHmwcHn «:menon npmH mm:.m H “my “Hmz mHoH .mp2 ”naston epsH 000.0 0 mnoz "cmenoHs spmH 503.0 H Amy ppmscmm ”nmmHsOHz npNH Hacksaw 000.0 0 psHHm mcoz «mwanon apHH mpmchnwo mom mpmcHocwo 95090 no mnstm mcwmpcmonom nmm mampH nmmmgmzmz HmOHaHHom wonmznmsmz mmmoo mo mo¢92m Aumanpnoov somHnmmaoo nmmwnumo .m> Hprnovacnm «HmomHv moaHe swoonmppmno 39H: oonmaaoo HoomHv Hacksaw ngHm mpOHupmHa mprHpomaoo mo pom m GH cmnmmaoo mwwnmboo mpmcanmo Hmpoa ”oHanHwb mmmnm>HpHpmmaoo «AH> umpamnov mnHHH mam<9 ommm :mN.0 . 000.0 000.0 000.0 212 000.0 maH.o 03H.0 omN.o mmo.om mmm.mm 000.0 Ham.o mmm.o Nom.o mpwchnmo pom mmwapzoouom OOOH 0 Amv mmawh ADV memQ H00 ppmuopm A00 means: any mmmm Hummpnoo mpumm oz “pmmpcoo mmocsm :H ummoq one CouHsmv NNMOMOHHH 41': mpaeHosmo Mom wfiopH Amv nonwm HOV mchm Axosn mama I my anNmHm A00 umnmmuna Amy Moonm ADV nonmwchHH Amv “exam HOV muobHHm Amy uoHHHsa moaHB mwoocmppmno noawamsoz "mommeCma nchz «commences npstm "mommmscme nunmpmm «commences npNHm "mommmcnme spuHm «commences spasom “mowmoSCme ouHse «mommmcnma ocoomm ”ommmmSCoe pmnHm nacho no mustm HmoHpHHom monazumsoz mmm<2m3mz m0 medeQz¢U mam mooo mo mUNBzmommm GomHnmano ummHIMMo .m> Hmenmonmnm mpOHuumHQ opHpHpmmaoo mo pom m CH nonmaaoo mmmnm>oo mpmoHonmo Hmuos «mHannm> mmmnm>HpHpmmsoo “HmomHV wmaHe «woocmppmno :pHa cmumaaoo Homev Hazusoe chHm «AH> umpamgov ouHHH mHmee “coHpmemn 500. ”.0308 Houpsoov II mmm.mm omH moaHa wwoocwppmso you mHmqupsm 213 0mm.H H AHHHmQOprzV mmomm Hmsonmonwnoo HH< mm:.o H HHHHmonpmzv mpmuoosoo HH< mum.H m mumnooamm mommofiCme HH¢ HmH.N HH umsossmmHm psmonmneusm NmN.0H NN momccmm meUHmmnm mmaHB nmm.H N «moonwppmso AmmmmmCCmev coHpmNHnswno SQOHszmmm mpmnHosmo pom opmcHozmo macho no onstm mommpnmonmm new mampH Hoammmzoz HmoHpHHom wonmznmsmz mmxfizmzmz Mo MB¢QHQZ¢O mmm mo¢mm>oo &o moHqumaaoo no new a :H cmnmafiou mwmumboo mpmcHosmo Hmpoe «mHannm> mmmnmpHpHpmgsoo «AH» nopmmsov UIHHH mHm4B .U‘rll . _ mH.NmN m.mN was“ Hm masono pmouoch 1.. H _ mN.HH m.H MNN m nonopm ho psov . mopHuopcomonmom nooHHnsaom mm.mm 0.m :\m mHH 0H Ampmpm no 9500 mmpHumpComonaom oHpouooaoQ mn.NN m.m mm: m Hopmpm no 9300 mHOpmcom amoHHnsmom on.HsH “.mH Hem mm Hopmpm mo 9:00 whopocom Opruooaon “H mm.o: m.: :\m 00 m smegma pcoonoumuxm a.mm.-~ m.mH s\m as: on Hosancmem pcmonmum mw.ww 0.0 mmmH NH moonwaoo you oosHaoz smOHHpsomm poanmHo 0m.mn m.m mHH mH moonwnoo you monHaoz oHpouooSoQ poanmHQ mN.mom m.HN Mmmm m: opHpousomonamm .m .D you moocHBoz conHpsmom mm.mmH “.mH :\m can Hm opHpmpnmmmpoom .m .o you moonHaoz OHponooaoQ oo.w:om m.:oH moo: mNm mpmHm mweoHuoost mN.man 0.0:H moamm NmN opon nomCSOHIhooQCoM monocH monocH nESHoo mocozumnm nasHoo Hmpoa nmoz smoz Hmpoa monoovonm mwonoboo no poomno mpoHHpmHQ opHpHpomBoo SH omananm muoaoamxoz mnemoom opHpHpomaooucoz .m>Hquomaoo «AH> nmpamsov >H mamaa 215 MH.:mm mm.mm mu.mm mo.m:: m~.::H mu.o:H om.:0m oo.omH MEN oo.o mmsocH csoHoo 9mm: 0.0: m.m 0.5 m.oN m.NH 0.MH m.oN m.mH 0 mo 0 o 0 $“ HN honosvonm zoos Mano.mH In maopH venomoo: SH monocH GBSHoo owNH nu mampH ooHSmooz no monoavonm proa «momH :\m 05H «mmH «Ham Hmmm mmmm mmm com mm: «so: 0 monoaH :ESHoo Hapoa om HH 0H mm mm mm mm mm on on Hmpoa honosooum Howmnoboo hpnom oane .moHHOpm xenon: .monzpmom oopoHomuanomamov nonuo muonuo>ou soOHHnamom muonnopoo OHpmnooaom whobnsm one mHHom spHpHpoe HonoHumNHnmwno GooHanmom mpHpHpoe HmcoHpmNHnowno oprnooBon opmnmm .m .D you moonHBoz cmOHHQBQom opmnmm .w .0 now moonHaoz oprnooamQ whoomoH HocoHpmz anHHpsmmm whooooH HonoHpmz OHQmHooBoQ macpoon cowooHQQD mwouopoo ho aoonpo Hooachnoov mpoHupmHQ opHpHuomaoo CH omanHnam mummommzoz muopoom opHpHpmmaoousoz .o>HpHpoQBoo ”HH> neugmgov >H mHmHuoqumouaom oHuonooaoQ mo.m: N.: «on: N: Ampapm oo p200 unapmcom cmoHanmom 00.0:H 0.NH :\m :0:H wNH Houopm no 9500 mnopwnom oHpmHooEoQ Ac mu mmumw N.w :\m 0mm N0 masons unoonoumuxm mm mmH 0.NH :\m mmnH wNH nozosnomHm psoonoum mm.H: H.n «mH: Hm mmouwnoo mom ooSHaoz sonHnsaom poanmHn 00.:N m.N mom 0N mmmnwnoo you monHBoz OHpoHoOEoo poanmHQ 00.NOH N.m mN0H N0 opprpcmmopmom .m .D you momcHaoz cooHHpsmom 00.0n H.& :\m com Hm opHumpnmmmoomm .m .0 you moocHaoz oHpmuooaom 00.0HmH «.mmH oOHmH mNNH mpmHm mmeoH-notz om.ummH n.3mH mummH msmH mpaHm nomcnonumcmzsmm monocH monocH mazHoo hocosooum n85H00 Hopes mam: zoos Hmuoa honosooum omonoboo No poofino mpoHupmHm o>HquoQBoon:oz :H oosmHHnsm mnoamamzmz mnopomm mpHqumoaoonsoz .mpHpHpmosoo "HH> umpomsov > mHmee fl: “M.... . . .. i r :\m mm:.do II mmHROpm oopfimdmz CH mososH CEDHoo Ham: II mmHuopm UmHSmmoz Mo mozosvmum Hopoe mN.Hmo m.ma WNHmo mm: Hmwmpmpoo spasm ouHsa . .moHHOpw :0000: .mopfipmom . oopmHomIanomaoov nonpo mm.mN N.N :\m mmN NN mnonuobou noOHHnsaom mH.:m u.m :\m How 00 muocuopoo oHpmnooaoa 7. om.HmN «.mH “Ham NmH msm>nsm com mHHom l 9. mm.HHH N.HH :\m mHHH NHH spHpHuoe HmBOHpmNHnowuo QoOHHnsgom wo.mNH o.mH :\m omNH oNH spHpHpoe szoHpoNHmmwuo OHuwnooaoQ 00.Nm 0.0 mNm 00 oposom .m .D How momcHaoz cmOHHnanmm mo.mm m.oH :\m 0N0 HOH opacmm .m .0 you moocHaoz oHponooBmQ m:.mmH N.NH sxm :mmH mNH mumemoq choHpmz omOHHpsomm 0H.mmH N.0H HmmH NwH mnoomoq HoQOHpoz oHpoHooaon om.mm :.N mmm em mnopomHm ommomHooo monozH mononH nazHoo mononuonm :89H00 Houoe Goo: Goo: Hmpoa honosvonm owono>oo no pomnno Acoanpnoov muOHupmHa o>HpHpomaooIcoz SH nosmHHpam muogmamzmz mQOpomm opHpHpogaooIzoz .o>HpHpomaoo «AH> popamsov > mqm«9 218 NH 0H MEG oHoU 0 NH 0H -& H mnom .0lo 800H NJNIQN 0-0 :\m nHHIHH :IH «sNHIJH «oNIm 0-0 0-0 s\m mHI: «mum «NHNImN «NBIHH 0-0 «mmINH s\m msum omqu :\m nmum 0.0 “ms-“ HIH wommumm wumuc mmHIm :\m ooHua msoun MmmHnmm s\m Ho-“ s\m NNI: «mmum :\m mm.: :\m mmmumn HHmIOH mo-n :\m NHsumN mmHmumH Hmomm< HmHonoaaoo mHsmaoz possum oHHH>£moz HosHpcomImzoz mHHbeocm HmonSOh oHHanoaM oHoHnonso cowoMmsz owwmmIonooom muHo monopoue ozoz sonmem :\m eweuom moHHIHH «n IH szumH HOHIn mmmnm moHoam ensue mummmm pOanmHn m>HpHpomeooInoz momoIm: Mmmuo Momum ONNIQH wommumH Hmouaon psHHm «Hmmumn :\m mmuo wwo-n Mmosuum HooHIHH Hasnson spasm mnHmcaH mmmnopoo monHaoz mocHaoz mooGHBoz moonHBoz .wcoo .nammm .Bom .wcoo .wuoo mummmm poanmHQ Hopes pngpmHo poHupmHo .omm .smo mpHpHpmoaoo muommamzmz Hm50H>H0GH :H owmuopoo mo “mono Hamm mnouome mpHpHpmoaoounoz .mpHpHpmoaoo .HH> umpamnov H> mHmHe .ompuHao one; mGoprszmoo Hump spano one cnoomm on» .noono Mama honoswonm BopH :H wstan npnm>om one pmnHu onp non moHp who: mums» oosz .mnomopoo pas» you wmsozH naoHoo no pnoamuamooa on» mH onooom 02» .p9500 monosvmnu BopH on» wH hopes: pmnHm map .wonstm mo pom some SH "opoz m m :mHIHN 0.0 :\m ONIN «oHI: naHImH cammmmccoa aHHHpgmaz HH HH wsmIHH ouo N-N :\m mum mmmuo mmsHe «woonaupmno Mamm owmnoboo oozHBoz mocHaoz mmmnHaoz moonHaoz Hmom .doum .waoo .nomom .aoa .wcoo .MCOU mummom pOanan .H.o ampH Hmpoa poanmHo noanmHo .oom .smo opHpHpmosoo Acoschcoov muommmmsoz HmsoHpHosH 2H owouoboo no noono Mama muonowm opHpHpoQBooIcoz .opHqummaoo «AH> umpmmnov H> mqm