This survey of political parties in fifty-three countries relied heavily on traditional methods of library research joined with modern methods of information processing. The method was traditional in that virtually all the information was collected by consulting printed sources located through extensive bibliographic searches of library materials. It was modernized through the use of microfilm and computer techniques of information storage and retrieval (see Chapter 1 of Part I). Even with these aids to information management, the success of our research was ultimately dependent on the number and quality of items in the bibliographies.

Before outlining the sources consulted and steps followed in constructing these bibliographies, we must note certain constraints impinging upon the bibliographic search which affected the final results. First, it must be understood that this study was designed as a broad survey of political parties in fifty different countries. Second, it must be recognized that this survey was undertaken by graduate and undergraduate students who initially had little or no familiarity with the politics of the countries they were assigned to research. Third, one should expect that skills in foreign languages among this group constituted a scarce resource to be spent carefully on the countries that needed it the most. Fourth, one needs to realize that the bibliographic searches constituted the very first step in a research sequence that has spanned nearly a decade since the project was launched with financial support from the National Science Foundation (see the Acknowledgments for a discussion of the phases of the project). Finally and unfortunately, no funds could be raised after 1970 to support the project and extend the bibliographies beyond their original compilations.

These constraints operated to impose some definite limitations on the bibliographies presented below. Due to the broad scope of the project no attempt was made to be exhaustive of the literature on countries where parties were well-studied, e.g., France or the U.S.S.R. Instead, the literature search was terminated when the information needs of the project were thought to be met. Due to the qualifications of the researchers and the scarcity of language resources, foreign-language materials ordinarily were not included when the English-language materials were judged to be adequate for the purposes of the project. Thus, the file on France contains virtually only English material, as students proficient in French were assigned instead to countries in former French West Africa, where the English literature was quite sparse. In countries with more exotic languages, fortunately, English-language translations of relevant material were often available--as in the case of translations of Russian publications on the Communist Party of the Soviet Union. Because most users of these bibliographies can be expected to know only English, this heavy reliance on English materials wherever possible may be more a blessing than a limitation.

Unhappily, no such blessing in disguise can be claimed for the most serious limitation of these bibliographies--the absence of material published since the late 1960s on party politics in our countries. In a few instances, citations to publications in the 1970s were added as new material had to be consulted to resolve problems detected during review of the researchers' work. But no claim can be made for any systematic coverage of material that appeared after 1967, 1968, or 1969--depending on which year the bibliographic search was completed for the country in question. Insofar as this survey of party politics centers on the time period 1950 to 1962 and as more recent publications tend to focus on more recent developments, these bibliographies can be considered as comprehensive for the period under study. They should certainly serve as a good starting point for those who wish to study party politics in our countries since 1950.

Over 3,500 documents, selected specifically for their relevance to political parties and party politics during our time period, were included in the bibliographies. While books and journal articles account for most of the documents, the bibliographies also contain citations to unpublished doctoral dissertations, papers presented at professional meetings, government publications, party publications, and newspaper and magazine clippings. These items were culled from a much larger body of references turned up through a systematic search of both standard and specialized sources.

The standard bibliographic sources searched for references to parties and politics in each of the original fifty countries included the:

International Bibliography of Political Science,

Bulletin Analytique de Documentation Politique, Economique, et Sociale Contemporaine,

External Research Reports of the U.S. State Department,

Universal Reference System,

ABS Guide to Recent Publications in the Social and Behavioral Sciences,

and Subject and Author Catalogs of the Library of Congress.

If the above sources did not generate an adequate set of citations to pursue the search through internal footnotes, the researchers were instructed to consult other sources:

Social Sciences and Humanities Index

Bibliographic Index

Public Affairs Information Service

London Bibliography of the Social Sciences

International Political Science Abstracts

Dissertation Abstracts

U.S. Bureau of Census, Bibliography of Social Science Periodicals and Monograph Series

American University Field Staff, Reports Service

In addition to these general sources, specialized area bibliographies were also identified for researchers. Moreover, the project bibliographer provided each researcher with the results of customized searches through these facilities and publications:

U.S. Government Joint Publications Research Service

the DATRIX dissertation service of university Microfilms

the Defense Documentation Center of the U.S. Government

American University's Cultural Affairs Information Analysis Center

Smithsonian Institution's Scientific Information Exchange

U.S. Government Research and Development Reports

U.S. Government Publications: Monthly Catalog

Finally, the researchers had access to extensive pamphlet files which also contained copies of the clipping files of the Christian Science Monitor and the New York Council on Foreign Relations. To all these sources, the researchers invariably added fugitive material collected through their own devices&endash;most usually with assistance of diplomatic offices in or from the country concerned and occasionally foreign desk officers at the U.S. State Department.

Once material was selected for inclusion in the information base, each page was indexed or "tagged" with one or more three-digit numbers (see Chapter 1, Part I) according to the coding categories reported in Part I, Table 1.2. The material and binary representations of the accompanying codes were then photographed on 16mm rolls of microfilm and spliced onto film cassettes organized by countries. With the use of Eastman Kodak's MIRACODE system these cassettes could be searched electronically for the presence of codes or certain combinations of codes. Pages which bore codes that satisfied the search were displayed instantaneously on a screen for study.

In all, more than 70,000 pages of material from more than 3,500 documents were indexed, microfilmed, and thereby incorporated into the information base. The students who researched our original fifty countries employed this information base and the MIRACODE technology to code their parties. Clearly, both the quality of their coding and the number of variables they were able to code for little-studied parties depended largely on the quality and quantity of the information located on party politics in their country and processed for inclusion in our microfilm files. Table 1.1 in Part I shows the number of pages contained in the file for each country. The number of pages by country, however, offers only a crude guide to the adequacy of the information base available to the researcher. A better indicator of the extensiveness of the available information is the number of pages that pertained to specific parties within each country. Because each page was tagged with the relevant identification code for each party discussed on the page, this indicator was readily obtained through an analysis of the indexing process. The results are given in Table III.1.

Only 145 of our 158 parties are represented in the table of pages by party due to departures from standard practice in several countries. As mentioned in Chapter 1 of Part I, three countries (the U.S., U.K., and Canada) did not appear in the original sample and were researched after the initial study of fifty countries was completed.

The research for these three countries did not involve the same systematic construction of information bases underlying the original sample. Eight of the thirteen parties missing from the tabulation of pages by party come from these three countries. The other five are from countries whose history of party politics was so confusing at the beginning of our study that the parties were not properly identified or assigned distinguishing codes until after the indexing of the literature was well underway or even completed. Thus the five parties lack an estimate of the number of pages on which they are discussed, but the ambiguity surrounding their existence suggests that they were unlikely to be discussed at length.

For the other 145 parties, we see from Table 111.1 that the files contain an average of nearly 392 pages of discussion per party. On the surface, this would seem to be an impressive amount of information to support a cross-national survey. But this mean figure is greatly inflated by a small number of parties. For example, the lone entry in the 1656-2755 page category in the table stands for the Communist Party of the Soviet Union, for which we indexed 2,699 pages. Another example is offered by the Indian Congress Party, discussed on 1,978 pages.

A more accurate indication of the quantity of information available for our researchers can be obtained from the modal categories in Table 111.1. Note that 69 of the 145 parties are discussed on a much smaller number of pages, from 56 to 255. Indeed, 49 of our parties are treated on fewer than 156 pages in our file. Allowing for the incidence of brief mentions on these pages and considerable redundancy in information provided, one recognizes that in sheer quantitative terms, the information available to our researchers was often wanting.

Of course, the number of pages discussing any party bears at best an imperfect relationship to the quality of the information on that party. Concerned with the quality of the available information in the course of our research, we devised the "adequacy-confidence" codes (see Chapter 2, especially Figure 2.2) for expressing the adequacy of the information available for any coding judgment and our confidence in that judgment. But those codes apply only to individual variables for individual parties. On a broader level, one can evaluate the state of the literature for an entire country by assessing the nature of its coverage in terms of the usage of indexing codes (Part I, Table 1.2) and in terms of the characteristics of the documents included in the file. Such assessments are made on a country-by-country basis along with the bibliographies for those countries in the chapters which follow in Part III. Each chapter contains two sections:

Bibliographic Essay: The bibliographic essays discuss the characteristics of the party politics literature at the time the bibliographies were completed (ranging from 1967 to 1969, depending on the country) and the coverage and quality of the material included in our files. The essays for the original sample of fifty countries conclude with a tabulation of indexing codes as applied to individual pages in that country's microfilm file and a tabulation of data quality characteristics displayed by the individual documents in the file.

Bibliographic Listing: Each chapter concludes with an author-alphabetized listing of every item indexed for the file plus some additional items that were consulted but not indexed. Most of these additional items were published after the bibliographic search and indexing phases of the project were completed. A reference code is associated with each item in the bibliographic listing, and the items in the bibliography are occasionally cited by these codes in the discussions of the party support data for the attraction, concentration, and reflection variables.