Compared to most of the other countries in the ICPP Project, Australia presented no problem in locating and obtaining material dealing specifically with political parties and party politics. In fact, the literature was found so easily that we overran our target figure of 2,500 pages, set as an arbitrary limit on the size of the file for any given country. This is not to say that the literature on Australian party politics is perfectly developed, but it certainly is extensive.

Unfortunately, the researchers who read and indexed the Australian literature assumed other responsibilities on the project and never wrote an essay recording their experiences and impressions. Not having read the vast bulk of this literature, I will limit my comments to salient features of the material in the file.

The 3,039 pages in the file were tagged with an average of 2.59 substantive index codes per page, meaning that nearly 7,900 codes were distributed by topic as shown in the first of the two tables which accompany this essay on Australia. The frequency of codes used (Table 1) can be read directly as the number of pages in the microfilm file that are tagged with a given code. Thus the most frequently used of the specific indexing codes is 530, which indexes the discussions of the parties' "issue orientations."

A total of 391 pages in the Australian file are tagged for their discussions of party positions on issues. The next most frequent code is 490, used for "centralization of power," which we found treated on 346 pages. An examination of the table of codes and frequency of usage discloses the topics that are well-covered in the literature and those that draw little or no attention. The resulting profile of the literature on party politics in Australia reflects the major features of Australian political parties and political life.

The heavy usage of the "issue orientation" code (530), for example, tends to be typical of countries with highly institutionalized parties competing openly for electoral support with good prospects for heading a future government. The strong reliance on the "centralization of power" code (490), however, reflects the more particular aspects of Australian politics, for it deals with the power of state parties vis a-vis the party organization at the federal level. This issue is important in all three Australian parties, whose organizations parallel the federal structure of the government. Note that the third most frequent code is 420, used to reference discussions of "regional party organizations."

Turning to codes that are used rarely or not at all, we find, for example, only two pages indexed with 270, "causes demonstrations, riots," and only one assigned 760, "activities of students." In the former case, we can dismiss the absence of codes as a reflection of politics in Australia, where parties tend not to play violently disruptive roles. The latter case was thought to be .a matter of neglect in scholarship--assuming that student participation in campaigns warranted more discussion than can conventionally be contained on one page--but our consultant states that students did not indeed play any part in party politics during our time period.

A broader perspective on the coverage of the literature can be purchased by moving from the specific codes to the major coding categories. The major category that subsumes the most frequently used codes is 4--, "party organization," which accounts for nearly one-quarter of all the indexing codes assigned. Code 4--has come to be an indicator of the quality of the literature for coding, parties on the variables in the ICPP conceptual framework. Heavy reliance on party organization codes signifies that the literature discusses parties per se. On the other hand, heavy usage of the 6--codes (political environment) and 7-- (social conditions) exposes a literature that merely mentions parties in passing. Judged by these standards, the Australian literature is very good indeed.

The quality of the literature on parties in any country depends on the way parties are approached as objects of study. Codes 0-- are used to index discussions concerning the "study of parties." In all of our countries, these codes virtually always stand at the bottom for frequency of usage, as they do for Australia. But their incidence of use in indexing the Australian literature is much higher than usual, attesting to the analytical quality of the information. Some 258 pages, 8 percent of the total in the file, were addressed to matters of definition, typologies, functions, theory, approaches, or methods in the study of parties. While not the highest proportion of pages tagged with 0--codes in our study, it is one of the highest and fits with our previous observations on the literature.

A second approach to evaluating the quality of the parties' literature in a given country is to analyze the characteristics of the items in the bibliography, document by document. The second table accompanying this essay summarizes our attempt to assess document characteristics by assigning codes to seventeen different "data quality" variables for each of the 154 documents in the file. Only the four most frequent codes for each variable are given in this abbreviated table, (the full set of codes is given in Chapter 14) but those four account for, the most salient features of any file.

We see under "document type," for example, that most of the documents, 60, are journal articles, sections in books supply another 51 documents, -books in whole or part provide 21 citations, and only 9 documents are news features--usually from newspapers. In stark contrast to the Australian situation, the information files for some other countries show news features and even news items as the most common type of document. The scholarly character of the Australian file is confirmed by the fourth data quality variable, "position of author," where academics outnumber journalists by 112 to 16.

The other data quality variables relate much the same story in comparison with the literature for most other countries. Almost half the documents engage in "quantitative analysis," even if it only means reporting some data in tables or text. About one-fifth of the documents provide some "theoretical treatment" of sorts, and almost half are rated above zero on "traditional scholarship," as they attribute their sources of information through footnotes.

In addition to the first fourteen "objective" data quality variables, the indexers coded each document on three "subjective" variables, which are reported in the last three rows in the table. The indexers' overall judgment of quality for any document was never "low," although indexers were forced to accept some low quality documents for files of other countries. Their evaluation of the "author's ideology" was overwhelmingly "centrist," indicating that ideological biases did not complicate the literature. Their judgment of the "author's objectivity," however, found about 8 percent of the documents to be tainted by values of some other sort. Nevertheless, the Australian file, in sum, stands out as one of the best in the ICPP Project.