previous chapter
next section

Appendix 1—
Using Government and Legal Documents

Unlike many other methods, archival historical research does not have a how-to manual to guide the researcher. Those of us who use this method often find ourselves groping in the dark until we develop a sense of the material. I frequently wished there was a tried and proven set of steps I could follow to quicken the pace of my research. In that spirit, I am sharing my trials and tribulations with others who may wish to join in this tedious, time-consuming, but satisfying search. I warn the reader that this appendix is not a handbook for using documents but merely a summary of the approach I took in researching the case studies included in this book.

I began these case studies as part of an attempt to understand the meaning of the statistical evidence on interlocking corporate directorates generated by the MACNET research group at the State University of New York at Stony Brook. I could see that the patterns of interlocking directorates were strong, but what did they mean? Many of the researchers insisted that they indicated power, but I was still unable to follow the empirical analyses to the substantial conclusion. Finally, Michael Schwartz wisely challenged me to look at some cases and see if I could trace the meaning of the interlocks. Thus began the search that resulted in this book. I started with the theoretical framework of the debates concerning interlocking corporate directorates. Later, many of those questions


would change as the theoretical framework broadened to include issues of the state and questions of power.

My selection of corporate cases began as an attempt to test Beth Mintz and Michael Schwartz's notion that banks become more central when corporations experience a crisis and then recede when the crisis ends. I also realized that many corporate relations and processes are hidden until a firm runs into a crisis and such information becomes a matter of public record. To untangle this material, I began with a simple typology of the possible sources and outcomes of crisis. This led me to select cases that are generally well known but had not been analyzed in this particular framework. The Chrysler Corporation bailout had received extensive coverage in the business and popular press, as had W. T. Grant Company's bankruptcy and Leasco Corporation's struggle with Chemical Bank. I decided to begin with these materials to get a rough idea of the sequence of events, the major actors in each case, and the location of documents and other materials. Once I was satisfied that I had a reasonable understanding of the information available from these sources, I interviewed several business analysts and lawyers involved in these cases for further insight into the events, actors, and source materials, particularly those not mentioned in the popular and business press. These interviews also alerted me to important elements to watch for in my examination of the available documents. Occasionally I was able to locate documents myself, give them a preliminary examination, and then perform interviews. Later I returned to the documents for a more careful inspection.

The interviews conformed to the professional code concerning research with human subjects. I told all interviewees the purpose of the research, promised anonymity if they wished, and provided the telephone number of the university's office of research administration to call to report any mistreatment they felt they encountered. I sent interviewees a transcript of their taped interview to allow them to identify any remarks they wished to have off the record, to correct any inaccuracies they found, or to add any new information they wished to include.

Armed with a basic outline of the cases, I then visited the government documents section of the library (all government documents used in this research are included in the bibliography). Both the


Leasco and Chrysler cases had gone before congressional committees that investigated and held hearings about these firms' troubles. Congressional hearings are gold mines of data. When the hearings end, the Government Printing Office publishes the transcripts of the entire proceedings, including any documents submitted to the committees during the hearings. Since the Leasco hearings had occurred almost a decade before my research began, I had to spend long hours in the library copying notes from the transcripts. I also spent quite a bit of money copying the pertinent supportive documents included in the volume of testimony. I was luckier in the case of Chrysler Corporation's bailout hearings. Since those were current and not yet published, I was able to request a copy from the committees holding the hearings. If you can keep abreast of congressional hearings relevant to your research, you can call the committee in Washington, D.C., and request a copy of the proceedings. To date, these publications are free and provide an incredible amount of research material.

I was far less fortunate in my attempts to access the materials concerning W. T. Grant's bankruptcy. First, the case was mired in a long, drawn-out series of litigations that were still going on when I began my research. Second, the materials were not neatly bound and available in the library as government documents, but were lodged in the basement of the United States Southern District Courthouse in New York City. No one had filed and categorized them so that I could quickly access the ones I wanted. There were hundreds of thousands of pages of testimony transcripts, depositions, documents, and legal briefs related to the case, all strewn about in corrugated boxes in a dusty room.

Sifting through these documents was not quick, but it was dirty. There appeared to be no rationale behind the storage of documents other than a haphazard tossing of paperwork into the first available box. I simply had to roll up my sleeves (literally) and dig in. The first difficulty was getting comfortable with the legal language of many of the documents. Interviews with the lawyers involved in the case helped clarify the meaning of the professional language. Breaking the code of legalese helped me realize that many of the documents were simply legal applications for continuances and extensions of the various hearings, real estate leases, and so on. These were not relevant to the case study and could


therefore be omitted. I still had to untangle the mountain of remaining documents, particularly the transcripts, depositions, and supportive documents. This task took many months of intensive digging, categorizing, copying, and analyzing. I periodically telephoned Grant's lawyers to ask for translations, clarifications, and updates on the status of the case as it worked its way through the courts. They were most generous with their time in telephone conversations and straightforward in their explanations. Their guidance was instrumental in sorting out the complicated pieces of the case. I also tried to keep up with the various hearing dates and attended court sessions for the case whenever possible. The hearings helped me become accustomed to the legal proceedings and the patterns of questions and answers.

Once I had collected the materials for the three cases, I had to work through the problem of coding and analyzing materials that defied quantification. Using government and legal documents does not make a strict coding scheme easy. For one thing, like other qualitative methods, the categorization of quotes, data, and evidence frequently changes as the research evolves. Sometimes the researcher gains new insights, alters her perspective, or modifies her analysis as new evidence confronts previous analyses and interpretations. In addition, since my task was to trace the processes and relations of specific cases, I was concerned not so much with quantifying instances or variables but with uncovering details and telling a story. Hence I cannot provide a codebook by which the reader could replicate my findings. Instead, I will describe the process I used to handle the materials and will provide a sample interview questionnaire and outline of court documents for the W. T. Grant case study (as an example of similar questionnaires used for the other cases). The reader may review these and the publicly available documents to determine if the findings are indeed accurate.

I made copies of all the materials I collected, keeping the original copies for a master file. I kept a file of the bibliographic information for each of the sources on four-by-six-inch file cards. I marked the secondary copies with colored markers representing various issues and concepts: green for managerial decisions, pink for bank hegemony, orange for loans, purple for stockholding, yellow for bank participation in decision making, and so on. (Of course, some markings changed for each case, as the important de-


tails in each differed. Only the broad categories were common, such as bank hegemony, loans, and stockholding.) I cut and sorted these marked document pieces into piles of topics and issues, noting their original source in the master file. Later I reshuffled these document pieces into new piles as new ideas and analytical frameworks occurred to me. This sorting and resorting continued until the end of the analysis.

I began each case study by writing a chronology of events and then sorting through and eliminating those events that were irrelevant to the research. Next, using the analytical framework, I broke out of the chronology to categorize events and relations by issue or concept: here was evidence of early bank knowledge of impending problems, there was evidence of bank hegemony, and so on. I revised each case study no fewer than a dozen times to smooth out the analysis.

Later, in reading through the business press, I happened to notice that many of the processes and relations involved in Mexico's foreign debt crisis bore a striking similarity to those I had found in the corporate cases. I decided to keep an eye on the case, and particularly on the congressional hearings. As in the Chrysler case, I was quick enough to catch the hearings before the Government Printing Office published the data and was able to get copies of most of the relevant documents; the rest I got from the government documents section of the library. This time categorizing the materials was easier than it was for the corporate cases. For one thing, I had developed an approach to data management that worked for me. For another, I had developed an analysis that I could now use to categorize the data in this case. I looked for similarities and differences relative to the corporate cases using the master file and colored-marker method.

The process for the Cleveland case study was quicker and smoother than the other cases for several reasons. First, I had grown comfortable with my method of data location and management. Second, there were fewer government documents related to this case than there were for the Mexico case. But the age of the case meant that I had to spend more time in the government documents section of the library than I did for the Mexico case (although surely less than I spent in the Southern District Courthouse in New York). My conceptual framework was clearer, my categories firmer,


and my approach more refined than when I first began the project. Both the Cleveland and Mexico cases underwent many revisions before the analyses became tight and clear. Like the corporate cases, both began with the popular and business press accounts for the chronological development, identification of important actors, and location of materials. I continued to look for similarities and differences between the government and the corporate cases. And I continued to re-sort and recode the document pieces until I was satisfied.

I now realize that there must be a way to make a computer do the coding and re-sorting that I did by hand with my colored markers. Yet the process of repeatedly going through the materials to recode them made me see new things each time. Often materials that I had neglected to color code the first time would suddenly become more significant after the second or third reading. I am not sure that would have happened had I left the data management to a computer.

I would have liked to interview the key actors involved in each of the cases. Indeed, I repeatedly wrote to and called several of the corporate actors in an attempt to interview them, either in person or by telephone. Each attempt met the same response: they were not interested. Although many researchers have found managers and executives more than willing to grant interviews, the people involved with these cases were not—and for good reason. These cases are all politically and economically sensitive. An executive with any intention of remaining in the business world, or a government official who wants to remain an effective state leader, must be wary of antagonizing the banking community. Hence my research depended on government and archival documents, business press information, and interviews with business analysts and lawyers.

Although the method I used is time-consuming and tedious, it is also most rewarding. It can produce clear details of relations and processes that remain hidden in statistical analyses. The very tangibility of the material allows us to trace these relations and give life and form to the quantitative data on which speculations of meaning often depend. I found the lack of a road map for the method frustrating but exhilarating—and well worth every minute.


previous chapter
next section