Appendix

The Interviews and Interviewing in China

This appendix provides additional detail and explanation concerning the interview set and data management system used in this volume. A brief consideration of where this research fits into the broader methodological firmament of contemporary China studies is the appropriate place to begin. For a thorough tour d’horizon of the research methods currently employed in the quest to understand China in comparative perspective, I recommend Contemporary Chinese Politics: New Sources, Methods, and Field Strategies.1 In terms of understanding the specific challenges of fieldwork in the PRC, see Doing Fieldwork in China.² An indispensable tool for basic biographical work on the PRC is China Vitae at www.chinavitae.org. Finally, the Hoover Institution’s China Leadership Monitor at www.chinaleadershipmonitor.org is a unique source on a broad variety of PRC leadership-related topics.

WHERE THIS RESEARCH FITS INTO THE POLITICAL AND SOCIAL SCIENCE FIELDS

This research is in what Calvin Chen and Benjamin Read would broadly term the “ethnographic” tradition of interviewing, supplemented by rich contextual detail drawn from primary and secondary documents, statistical compilations, and contemporary scholarly and journalistic accounts. With respect to the interviews, in a few cases (e.g., my 1982 interviews in the Yangzi River Basin) I was embedded in a Chinese bureaucratic system and a particular organization within that system long enough to get to know people and get a feel for the overall circumstance in which they found themselves. In other cases, I interviewed or met with the same people periodically over a substantial span, sometimes decades, episodically cutting into their lives as their careers unfolded (e.g., Zhu Rongji as Shanghai mayor, vice-premier, and premier). My approach is ethnographic in its attempt to understand how people perceived their situation and the society and organizations in which they were involved. Over more than four decades, I have perpetually sought to understand what drove individual and organizational behavior. This is a bottom-up, inductive approach, the purpose of which is to produce generalizations and propositions about the system and specific policy issues that others can test using a broader array of methodologies. I have always viewed Philip Selznick’s TVA and the Grass Roots: A Study in the Sociology of Formal Organization as an exemplary piece of work in this regard.3 In the “ethnographic” tradition there is an objective both to understand subjects as they understand the world and simultaneously to protect them as human subjects. An ethnographic approach is most useful and appropriate, as Benjamin Read says, “when what we are studying is subtle . . . , and when what we are studying is hidden, sensitive, or otherwise kept behind barriers that require building trust.”⁴ As explained below, in my interviews I used what Lily Tsai calls “conversational or flexible interviewing,” not standardized interviewing techniques.⁵

This work clearly would be categorized as elite studies within the broad field of Chinese politics and political science more generally, though it does not rely, for the most part, on aggregate attribute data that are important sources for many scholars, including Robert Scalapino’s early work on elites, Cheng Li’s more recent contributions, and the contemporary work of Victor Shih, Wei Shan, and Mingxing Liu.⁶ This research focuses on leaders at various levels throughout the functional and territorial systems of the PRC because in an authoritarian system such individuals exert outsized influence over policy and because, generally speaking, they constitute the selectorate that perpetuates the elite and political apparatus.

THE DATA SET

To elaborate on what already has been said in the Introduction to this volume, the respondents are a diverse group in terms of functional field, level in the system, citizenship, personal characteristics, era encountered, and the length of time I have known them—some interviewees I have known for decades, while others I met only once. Seventy percent of my interviews were from the PRC (not including Hong Kong), 19 percent from Taiwan, 1 percent from Hong Kong, and 10 percent “other” (see figure 4). Each interview was systematically coded along several dimensions, as explained below.7

FIGURE 4. Nationality of interviewees.

Rank

Rank (nine categories) refers to the level of the interviewee within party, state, military, and societal hierarchies. In the PRC’s bureaucratic system, every official has a precise rank, but these internal designations were not always made known to me. Further, rank systems are not the same in the various administrative jurisdictions and localities from which my interviews are drawn—differing between Hong Kong, Taiwan, Macau, and the PRC, not to mention among non–Greater China governments and organizations (e.g., the World Bank, the International Monetary Fund, or the United Nations, where Chinese nationals now often are employed). Consequently, I have assigned my own, standardized rankings based on the individual’s title, available biographic material, and his or her own description where available. The interview set includes 25 supreme leader interviews, 10 from the PRC and 15 on Taiwan; 58 Politburo-level interviews, of which 45 were in the PRC; 113 governor and ministerial-level interviews, of which 49 were in the PRC; 91 vice-governor and vice-ministerial-level interviews, of which 89 were in the PRC; 85 Bureau-level or equivalent interviews, of which 77 were in the PRC; 65 lower-level and miscellaneous interviews, of which 60 were in the PRC;⁸ 17 ambassadorial-level interviews, 17 of which were with PRC individuals;9 32 department director–level interviews, all of which were in the PRC; and 11 CEO or chairman of the board–level interviews, 8 of which were in the PRC, though in reality there were additional interviews in this category in Taiwan because the CEOs involved (e.g., Koo Chen-foo) had other governmental titles and ranks that were employed. Much more detail on the ranking data can be found in the note.¹⁰

FIGURE 5. Rank of PRC and HK interviewees. Note: “Minister-level and higher” refers to supreme leaders, Politburo members, vicepremiers, state councilors, CMC vicechairmen, governors, ministers, municipality mayors, and ambassadors.

FIGURE 6. Rank of Taiwan interviewees. Note: “Minister-level and higher” refers to Taiwan equivalents to PRC ranks in figure 5.

Concerning the frequency of high-level, midlevel, and lower-level interviews on China’s mainland versus Taiwan, the island has over the last four decades made a greater effort to expose visitors to people who, on average, are of higher formal rank than the individuals to whom one is given entrée on the mainland—over the years, 32 percent of my interviews in the PRC were with ministerial-level and higher officials, while the corresponding figure for Taiwan was 85 percent (see figures 5 and 6).

Domain

Domain refers to the location of the interviewee in the hierarchy of central places, in this case “center” (meaning the apex of the political system in the capital), 395 total interviewees, of which 296 were in the PRC; “province,” 85 interviewees, of which 75 were in the PRC; and “local,” 16 interviewees, of which 14 were in the PRC (see figures 7 and 8). The PRC data set is top-heavy, with well over a majority of respondents at or near the pinnacle of the hierarchy of central places (75 percent of interviewees were at the “central” level, and only 4 percent at the “local” level)—there is very little direct representation of China’s rural or small- and medium-sized city leaders among those interviewed. This bias in the data is even more pronounced in the interviews with Taiwan respondents, with 95 percent of interviews at the “central” level. This fact limits what can, and what cannot, be concluded from this body of information. Though provincial-level leaders in the PRC may seem to be “local” actors by some reckonings, in fact, provincial-level officials there are responsible for geographic units with populations and land areas that would constitute a medium to large nation-state in much of the world. In short, this data set provides only indirect and circumscribed insight into small-town, much less rural, China.11 Rural and local China is an enormous part of the challenge of governing the PRC, and grassroots perspectives shape what Daniel Yankelovich memorably called “the boundaries of the permissible” in both domestic and foreign policy. As one senior academic put it in 2013, “The leaders are afraid of silent farmers.”¹² To the extent that the interviews used in this study provide insight into rural and local Chinese society, it is generally through the eyes of leaders at the apex of the system—an important perspective, albeit one with major blind spots.

FIGURE 7. Domain of PRC and HK interviewees. Note: Provincial-level interviewees include provincial-level cities and Hong Kong.

FIGURE 8. Domain of Taiwan interviewees.

Status

Status refers to the self-defined functional hierarchy in which the interviewee places him or herself—the three most important of which are “military,” “party,” and “state.” Many leaders in the PRC are in two of these vertical, systemically central hierarchies simultaneously; an exalted few are in all three simultaneously, most notably the PRC’s “supreme” or (what used to be called) “core” (hexin) leader, but also individuals such as some senior military officers on the Party’s Central Military Commission, for instance.13 Indeed, the definition of supreme or “core” leader has traditionally been the one who is simultaneously the head of all three systemically central hierarchies—party, state, and military. The outside world, as well as the Chinese people, will know that political system transformation has occurred in the PRC when membership in a single party is not the overwhelming criterion for holding positions of authority in all systemically central hierarchies and when a constitution effectively divides at least some components of executive, legislative, and judicial power.¹⁴

FIGURE 9. Status of PRC and HK interviewees.

FIGURE 10. Status of Taiwan interviewees.

Among the PRC interviewees in this data set, 56 percent were “party and state,” 14 percent “military,” 25 percent “scholar and researcher,” and 5 percent “other” (see figure 9). Most senior PRC interviewees were very likely in the Communist Party—however, the CCP membership status of many academics and researchers was often unknown or uncertain to me. With respect to the Taiwan interviewees (figure 10), the “party and state” category is even more prominent, reflecting the higher level of attention that the authorities on the island have been willing to provide visiting academics and others. Ironically, in Taiwan one can easily get a more party- and state-dominated view of things than on the mainland, though it is importantly true that it has been much easier to gain access to Taiwan society over much of the last decades than has been the case on the mainland.

These three systemically central hierarchies (party, state, and military) dominate my interview set. Thus the people I have predominantly interviewed and interacted with over the years in Greater China are significant players in the “system,” not common citizens, much less dissidents. The number of dissidents I have spoken with over the years is small, though I have interacted with a few high-profile individuals.15 Further, though the interviews as a whole are dominated by members of the systemically central hierarchies, there are deep recesses of those hierarchies that foreigners rarely see up close—Public Security comes to mind, as does the Communist Party’s Organization Department (which handles personnel). It was not until 2011, for instance, that I had a meaningful conversation with senior officials of the Chinese Communist Party’s “International Liaison Department” (Zhonglianbu), the department historically responsible for connecting the CCP to communist parties abroad. In more recent times, that department has broadened its portfolio of those with whom it conducts exchanges (including the Republican and Democratic parties of the United States) as the number of ruling communist parties abroad has dwindled to near the vanishing point. In English it is known as the “International Department,” with a large, modern headquarters on Beijing’s west side that increasingly hosts noncommunist foreigners. Therefore, I do not claim to represent every Chinese organizational perspective in this volume; rather, I point out a number of blind spots obviously circumscribing findings.

Another important aspect of my interview set is the fact that there are other statuses (which are not mutually exclusive with the above three) that I have encountered with increasing frequency as we have entered the second decade of the new millennium: “corporate” personalities and “nongovernmental organization (NGO)” individuals are two of these statuses.¹⁶ Another sector represented in my data set is the rapidly growing nonstate sector. In short, the proportion of interviewees who may not be involved in the party, state, or military apparatuses has climbed over time (while remaining modest), suggesting that China’s leadership and society are becoming more diverse, pluralized, and transparent—a reality that is having dramatic consequences for governance in the PRC, as seen in chapter 2 and throughout this volume.¹⁷ Also, the frequency of foreign interaction with China’s academics and researchers has dramatically increased since the turn of the millennium, as has the less inhibited nature of those interactions.

Gender

Gender distribution among interviewees is interesting and revealing in both the PRC and Taiwan subsets. Among the PRC interviewees, about 5 percent were female, while the Taiwan subset was about 17 percent female (see figures 11 and 12). The percentage in the Taiwan group is even more notable given that these interviewees include women who have been political party leaders (Tsai Ing-wen), presidential candidates (Lu Hsiu-lien and Tsai Ing-wen), legislators (Bi-khim Hsiao), ministers (Lai Shin-yuan), and a vice president (Lu Hsiu-lien/Annette Lu). In the PRC, on the other hand, while overall rights and status of women have improved and women have made great strides in many walks of life, their presence at the upper reaches of the policy and political system remains very small—Cheng Li reports that in 2007 only two out of twenty-nine PRC ministers were female.18 Indeed, in the run-up to the Eighteenth Party Congress of November 2012, just one woman was known to have been considered for membership on the Standing Committee of the Politburo (Liu Yandong), but she was not selected, instead being placed on the lower-ranking twenty-five-person Politburo, along with Sun Chunlan, Tianjin’s female first party secretary. Comparatively, Taiwan has made considerably more progress in this arena of female representation at the top of the political system.

FIGURE 11. Gender of PRC and HK interviewees.

FIGURE 12. Gender of Taiwan interviewees.

Distribution of Interviews over Time

This data set contains interviews spanning 1971–2013, with the distribution over time reflecting the explosion of access to all levels of Chinese society over the forty-plus-year period covered in this volume. The temporal distribution of interviews in the PRC is a relatively good, albeit imperfect, indicator of the expanding openness, access, and transparency of post-Mao Chinese society. About 1.4 percent of the PRC interviews were conducted in the 1971–76 period when Mao was alive. Even after the Chairman’s death, the PRC remained relatively closed to outsiders throughout the rest of the 1970s—only 2.3 percent of these interviews occurred in the period from Mao’s September 1976 death until the end of that decade. The 1980s account for about 9 percent of my interviews; the 1990s nearly 22 percent; and the new millennium (2000–2013) 65 percent. My interview rate has more than doubled every decade, reflecting both increased travel to China in general and the increased capacity of Chinese to go abroad. This distribution of interviews, of course, also reflects my own career development and wider range of contacts in China over time, but the basic point remains. In the early days of reform, even the most senior people in the China studies field were not meeting with many Chinese leaders, not to mention then junior scholars in the field, such as myself.

This temporal distribution of interviews is an imperfect but important indicator of access to, and openness of, China for many reasons. Whereas in the 1970s three to four weeks were required at a minimum to make a productive trip to China, in 2013 air transportation to China (and all forms of travel within the country) has so improved that it is easy to make more numerous, shorter, and highly productive trips to the PRC. In China, for example, it usually is about equally quick to take the high-speed train (Gao Tie) from central China’s Nanjing to Beijing in the north as it is to fly. Moreover, in the earlier years, once one finally arrived at a location in the PRC for research, things moved at a glacial rate. Simply making an appointment was a laborious undertaking, an adventure that began with trying to use the telephone system for which there was no publicly available phone directory. City maps were practically classified documents, and, oh yes, there were no taxicabs, even if one had an address—one had to rely on transportation provided by a designated unit host (jiedai danwei). And I remember sometimes asking my assigned guide on the way to an interview, “Where are we?” only to be told that they were “unclear” (wo bu qingchu). Finally, even if one cleared these hurdles, any interview or conversation usually involved a peitong, or minder, assigned by the local Foreign Affairs Office—whose presence generally exerted a chilling effect on respondent and interviewer alike. Interview sessions sometimes actually involved rather large numbers of extraneous Chinese in the room, for what purposes one could only imagine.

Often it seems that today’s foreign demands for openness, transparency, and access to Chinese people, organizations, and information are made without a full appreciation of how far China has come since 1971 or, for that matter, how far the country has come in the new millennium. Chinese leaders having news conferences, bureaucracies having spokesmen, and ministries having online chats with citizens are very recent sociopolitical “innovations,” as is the fact that the Foreign Ministry would begin to open its archives to foreigners. As the outside world loses any memory of a truly closed, indeed reclusive, China (a recollection that might help temper foreign demands), Beijing will face an increasingly impatient and insistent outside world. The more open the PRC becomes, the more it will face demands for additional openness and transparency.

A FEW IMPORTANT ASIDES

Protecting Respondents

The 558 interviews constituting this data set contain remarks by many more than 558 individuals—multiple respondents (and groups of visitors asking questions) participated in numerous conversations. This requires explanation as to the terms on which the data were obtained, the responsibilities of the interviewer, and the human subject safeguards that have governed this research and publication.

I have not released material in which a respondent could be identified if we had agreed to any permutation of “off the record.” Further, in meetings with very senior leaders (in Beijing, Taipei, and Washington, D.C.) multiples of people often took notes (on both the foreign and Chinese sides), and only in a few instances did leaders explicitly say that their remarks were off the record—though the unspoken expectation was (and is) in meeting with Chinese leaders that those having the opportunity to interact with these figures will exercise prudence. “Prudence” means that if you hope to have access in the future you should be mindful of what you attribute to leaders by name. Using vaguer identifiers such as “a very senior leader said,” is what is expected.

In terms of concerns about human subjects, most individuals in this data set would be categorized as “public personalities” or “public officials,” and therefore we were operating in the realm of the historical record. While academic and think tank respondents are in a grayer zone, when interviewing them, I either asked them how they wished to be identified if I thought I might wish to use their name, or I cited them in a nonidentifiable way. Throughout the forty-plus years covered by these interviews, there were only a handful of times when the respondent asked for absolute protection and confidentiality. Consequently, I will not place in a publicly accessible archive the entire set of original interviews and related documents until a number of years after my death—and presumed passing on of the respondents as well. As a practical matter, however, in many meetings with officials, there were note takers from the Chinese government or party (and often the American government as well) recording what was said.

Interviewing

Interviewing in China or in Chinese communities elsewhere is not sui generis, meaning that the commonsense and ethical approaches that work elsewhere are generally applicable.19 While the PRC is no longer the far side of the moon, one must exercise delicacy in posing politically sensitive questions and disseminating any replies for which the respondent might conceivably suffer negative consequences in the present or the unpredictable political future. Building and maintaining trust with respondents is the coin of the realm. To my knowledge, no interviewee has ever been harmed as a result of speaking with me. For one-time interviewees, of course, there is no preexistent record of personal trust (beyond your general reputation), but when you repeatedly return to the same locality, organization, or individual, this record of trust is slowly built and becomes indispensable. Without such a record, repeated access is not likely to be gained. An unobtrusive indicator of trust, therefore, is whether the researcher repeatedly gains access to localities, organizations, and individuals. In China, being trustworthy is not only an ethical imperative but a research necessity.

Related to trust, I believe, is another issue—I have never used audio recording devices, even when the Chinese side has a stenographer there presumably recording every word said. Moreover, it is likely that many conversations were recorded by devices that remained out of sight, feeding information directly into the security, foreign policy, and intelligence bureaucracies. Nonetheless, I have felt it is essential to the “atmosphere” that the interviewee always be able to say to his or her compatriots, as a last resort, “The foreigner misunderstood what I said,” without fear of definitive refutation.

The hardest work of an interview occurs before you walk into the meeting place. The interviewer should learn as much as possible beforehand about the subject(s) he or she is exploring for several reasons, one of which is that time is too valuable to waste learning what already is known, or knowable, from the publicly available record. An essential dimension of preinterview research and preparation is to familiarize oneself with the vocabulary that is specific to the area of concern. Using the correct, technical vocabulary indicates to the interviewee that you are knowledgeable, and technical vocabulary is more precise than a layperson’s phraseology, gently pushing the interviewee to see you as a member of his or her guild rather than as an “outsider.”

When I engaged in extensive interviews in the Ministry of Water Conservancy and Electric Power “system” (xitong) in 1982, for instance, I used the bureaucratic and engineering vocabulary employed in that functional domain because many interlocutors were engineers. Using the appropriate technical vocabulary was more precise and efficient, and it helped move the discussion from the unstable terrain of politics toward the far safer territory of a “technical discussion.” I recall asking one division-level Ministry of Water Conservancy and Electric Power official, for instance, about the shiyefei (operational or program funds) in his ministry, a subject about which I had read in a ministry publication. From the prior reading it was clear that these kinds of funds were problematic, but I did not understand exactly what they were and where the problems arose. The mere use of the correct term, shiyefei, elicited a long, detailed, and illuminating discussion by the interviewee that explained how discretionary operational funds had originated, how they had evolved into large reservoirs of unaccounted money, how corruption and inability to control investment by the center had resulted, and how all this also had to do with an absence of law. This illuminating conversation occurred in 1982, a period of generally tight political discipline. A section of my notes from that meeting recounts:

I then asked what the shiyefei was and why it was a problem. XX said it was originally to be “operations” money (like phone, etc.), but it has grown over the years and is used in “illegal ways” (buhefa), but the legal system was so chaotic that these funds had been diverted to all sorts of capital construction and miscellaneous uses such as furniture, buildings, land acquisition, etc. He said it was a problem of great importance and that while he didn’t know the precise size of the shiye funds in the 1950s, the size of the shiyefei had grown greatly in recent times. I asked what the qiekuai banfa (the method of cutting up the lump, literally) is, and he said it is slicing up the shiyefei among provinces (basically on fixed percentages) and then not having any accounting or control over how these monies are subsequently spent and they don’t have to be returned to the center if unspent. In short, the shiyefei had become a large pool of discretionary money for provinces. The shiyefei ceng ceng (or the shiyefei level by level) is the division of the shiyefei in the MWCEP [Ministry of Water Conservancy and Electric Power] “system,” level by level on a basically fixed percentage with no returns and no accountability or control on the funds’ use. All this, XX said, led to abuses and illegal behavior, except it was difficult to say it was illegal because [of] the disrepair of the legal system.20

A related point is that even if the interlocutor’s English language is better than the interviewer’s Chinese, I have found it preferable to conduct the interview in Chinese. This builds trust, shows respect, and drives the researcher to use the terms in which the Chinese system operates and your interviewee thinks. Moreover (to be frank), your interviewee will generally try to make his points exceedingly clear and provide the essentials out of consideration of the linguistic inequality. Nonetheless, it is essential to learn the technical vocabulary dominant in the field in which one is interviewing.

Part of “getting to know your subject” does not simply mean the functional area one is exploring, it also means the interviewee as well. Establishing a human connection helps establish a professional connection—personal connection and professionalism are not necessarily mutually opposed. The easiest way to do this is to find a point or points in common, which may be mutual acquaintances, school or geographic ties, or experiences. Starting a conversation with someone from Hubei Province (particularly Wuhan), for instance, by talking about local dishes (doupi or tangbao) can provide an expedited path to a comfortable and productive interview. But the starting point is to know something about your subject, and he/she knowing something about you. An admittedly unusual example would be an interview I had with then-retired vice-premier Gu Mu, a man about whom I had written a biographic study some years before we met. Before I met with him, I conveyed to his staff the book I had written in which his life was one chapter. I walked into Building 15 at the Diaoyutai State Guest House with a colleague, and my notes record the following as we sat down, neither having met the other before: “I walked into a large, beautifully decorated reception room with Jan Berris and met a smiling Gu Mu, with sparse teeth beaming with much gold. He is a bit frail but appeared in good health for a man of about eighty. I told him in Chinese that ‘though this is the first time we have actually met, I feel that we are old friends from my research about your life.’”²¹ This meeting proved to be one of the most comfortable and enlightening that I have had.

In terms of interview technique, I have found it most productive to use an open-ended style—“conversational or flexible interviewing.” The interviewer approaches the respondent from the vantage point of wishing to know many things, realizing that there will be insufficient time to cover everything and pragmatically selecting a subset of subjects on the basis of where the interviewee demonstrates the greatest interest and knowledge and the least apprehension. I have a long list of possible questions in mind, but I let the actual questions posed flow from the conversation’s logic to the fullest extent possible.

A practice that has served me well is to stay away from loaded words such as politics—I focus instead on how things work, letting the explanation of the process lead the discussion into the underlying views and conflicts. For instance, a common word used to describe bureaucratic processes in China is xietiao (“to coordinate,” to resolve divergences among people, organizations, and territories). Exploring who needs to be “coordinated,” why they need to be coordinated, and how they are coordinated gets you to the heart of politics without ever using the word itself.