. 7
( 55 .)


able.) This approach is not particularly useful in addressing questions
about process, dialectical relations, and sequence of events and may
even lead to teleological and over determined accounts. However, it is
an expedient means to engage in case study comparison for it holds
many variables, such as cultural factors or position in the world econ-
omy, reasonably constant. This case design lends itself to macrolevel
research where the dependent variable is analyzed in the context of
transformations in policies, institutions, and relations, resulting in a
description that measures the overall variation and shift in parameters.
The diachronic method analyzes the interaction between variables
through a de¬ned time frame (from t1 to t2). Here the processes and
mechanisms that engendered the transformation in the dependent vari-
able are the units of analysis. Rather than structural forces, the diachronic
approach identi¬es sequences of events, intermediary variables mediating
structural shifts, and bargaining between groups as critical factors. By
stressing hermeneutics, this approach may introduce a large number of
factors and contingences alongside a causal argument. Yet, diachronic
perspectives are critical for comparing and differentiating causal chains
that lead to the same outcome and often demonstrate the multiplicity of
paths to a single outcome. Also, tracing these processes will remind us
that our worlds consist of continuities as well as discontinuities.
By consciously applying these two approaches to study change,
one can shift back and forth between the trees and the forest. The
synchronic approach is privileged in Chapters 3 and 4, in which I
compare the form of network governance of the Tehran Bazaar and the
transformation of the state in the two general eras (1963“79 and 1979“
2000). In Chapters 5 and 6 the three selected sectors within the Tehran
Bazaar and the variations in the Bazaar™s capacity to mobilize, are
analyzed by taking a more diachronic approach to evaluate alternative
frameworks by investigating the sequence of events.
Finally, I have chosen a time frame (1963“2000) in order to hold
several key variables constant. First, the historiography is in agreement
The puzzle of the Tehran Bazaar under two regimes 29

that Iran™s modern rentier state was established in 1963 when the price
and production of crude oil steadily began to increase and constitute the
bulk of the state™s budget.61 The petro-dollar economy, with all its
virtues and shortcomings, began to develop in 1963 and has persisted
uninterrupted until today. In terms of political rule, 1963 is also a
useful starting point. Owing to foreign interference, democratic and
separatist movements, and the inexperience of the young Mohammad
Reza Shah, the period from 1941 to 1963 was marked with political
instability and uncertainty. It was only after 1963 that the central state
reestablished its supremacy and the monarchy™s autocratic rule was
fully imposed through strict control over the press, parliament, and
political parties.

Data collection
As already mentioned, secondary sources on the Tehran Bazaar, and
Iranian bazaars in general, are sparse, not comprehensive, and not
directly related to my research questions. Thus, in order to make valid
descriptive and causal inferences, I needed reliable and accurate indi-
cators of key variables. No such data are readily available, and as a
result, it was necessary to generate my own data and take into account
interpretive matters. The methodological goal during my ¬eld research
was to tap into a wide variety of primary sources, especially ones that
could capture the workings of the Bazaar. I used various forms of
interviews, participant observation, and primary and secondary docu-
ments, including surveying newspapers and dissertations and theses
written in Iranian and U.S. universities. Appendix 1 provides a fuller
treatment of various aspects of the ¬eld research.
This book follows a thematic rather than a chronological format. The
next chapter presents a brief historical background, a typology of the
literature on bazaars and markets more generally, and outlines my
approach to bazaars that views them as embedded networks. Chapter 3
details the variation in the form of governance between the Pahlavi
monarchy and the Islamic Republic by describing the commercial,
¬nancial, religious, and social bases of these networks and the mechanisms
that reproduce them. It pays particular attention to how the Bazaar™s

Homa Katouzian describes the post-1963 era the ˜˜Petrolic Despotism.™™ Homa
Katouzian, The Political Economy of Modern Iran 1926“1979 (New York: New York
University Press, 1981). For similar periodization see Ervand Abrahamian, Iran:
Between Two Revolutions (Princeton: Princeton University Press, 1982), p. 426.
30 Bazaar and State in Iran

cooperative hierarchies, unlike coercive hierarchies, generated a reputation
system able to appraise reputations, sanction behavior, and inculcate
cooperative norms and a sense of solidarity. Chapter 4 explains why this
shift occurred, arguing that the particular transformative programs of the
two regimes generated different forms of governance in the Bazaar by
altering the institutional and physical setting of networks. Chapter 5 traces
the dynamics in the structural change in the hand-woven carpet sector, tea
sector, and the china and glassware sector. These narratives show that the
sectors, although all moving from cooperative to coercive hierarchies,
followed different trajectories because of contingencies related to the
particular nature of the commodities traded and the institutional patterns
negotiated by the state and bazaaris, rather than mere presence or absence
of the state in the economy. Chapter 6 considers how the Bazaar™s structure
relates to its capacity to mobilize against the state. It presents a summary of
the literature that focuses on Islam and the clergy as the explanatory factor
for the Bazaar™s vigilance against state encroachment, and through a re-
reading of the historical record shows that this prevailing view cannot
account for various dimensions of bazaari mobilization. I return to my
discussion about cooperative and coercive hierarchies to posit that the
transformation in the form of governance accounts for both high levels of
mobilizing capacity during the Pahlavi era and lower levels under the current
regime. I conclude by drawing out the broader implications of the analysis
for the study of state“society relations, and Iranian politics in general.

Appendix: Methods of Data Collection and Evaluation
This appendix makes the research process more transparent for the
reader and forces me to self-consciously re¬‚ect on how information for
this project was collected and evaluated. Speci¬cally, data were gleaned
from several different types of interviews, participant observation, and
various primary and secondary texts. The bulk of this research took
place during ¬fteen months of ¬eld research including stays in Tehran in
the summer of 1999 and from August 2000 to August 2001, when I also
made short visits to Dubai, UAE, and Hamburg, Germany, to conduct
interviews with Iranian merchants.

My main method was in-depth interviewing of bazaaris using closed and
open-ended questions. The interviewing process was not a simple affair.
Public opinion polling and surveying are not well-established practices
in Iran, where authoritarian governments from the shahs to the mullahs
The puzzle of the Tehran Bazaar under two regimes 31

have tended to abuse such information to the detriment of the people “
or at least that is how it is perceived. Gathering information about
bazaaris is especially dif¬cult since, as part of today™s limited private
sector, they are often avoiding taxes and involved in strictly illegal
activities. Strolling into a shop and declaring myself a researcher was not
the best method of soliciting information, let alone engendering trust
and cooperation, from the cautious bazaaris.
I gained access to bazaaris through what is known as the ˜˜snowball
sampling method.™™ In contexts where members of a community know
each other and entry into the group is dif¬cult (e.g. local elites, insular
minorities, drug addicts in a neighborhood), access is gained by iden-
tifying a few members to refer you to further members, who in turn
direct you to other members, and so forth. This entails using a few initial
contacts to generate further interviews via referrals and thus create entry
points into the community. Through previous contacts I established six
independent entrees into the Bazaar. In turn, as I met these bazaaris,
explained my project and earned a level of trust, I asked them to
introduce me to other members of the Bazaar. The snowballing system
was quite effective since references and interpersonal relations are a
critical means of gaining access to information and earning trust. In this
manner, I was directed through networks of relatives and commercial
partners. A simple mention that I was so-and-so™s friend would usually
solicit cooperation. On some occasions my interviewees would call on
my behalf to arrange for an appointment, write a letter of introduction,
or personally take me to their colleagues. Over time and by following the
leads established by the initial contacts, I was brought into contact with
a large number and relatively disparate group of bazaaris.
To reassure interviewees that evidence was not going to be used
against them, all names of interviewees are strictly con¬dential.62
Moreover, I typically did not tape record interviews. In cases when I felt
that it would not alter the content of the interview, I requested per-
mission to tape the interview. Most interviews were conducted with
me simply jotting down important phrases, dates, and names during the

Although all interviews were entirely con¬dential, bazaaris were generally wary of
discussing their personal ¬nances and speci¬cs of their business practices. Thus, I did
not solicit information about their income, assets, and speci¬c investments. On several
occasions bazaaris made references to land and real estate they owned, but I never
sought to systematically gather ¬nancial information. Also, sensitive political topics
(allegiances to particular parties or political personalities, views on the legitimacy of the
Islamic Republic, or speci¬cs of their political activity during the Revolution) were
pursued only after a rapport was established or if the information was volunteered.
32 Bazaar and State in Iran

session. Then, after the interview was completed I would write down the
discussion in detail.63
It should be noted that although I used interviews with individual
bazaaris, the unit of my analysis is not the individual. Rather it is the
Bazaar (or sectors) at different points in time. Thus, unlike most survey
research or projects based on interviews, the data collected from these
interviews were not used to understand and explain variation at the
individual level, but rather the sets of relationships, practices, and
institutions that constitute the Bazaar. Nonetheless, it is still important
to pause here and identify the drawbacks of this type of snowball sam-
pling. Ideally the snowball method will lead to interviewing all members
of a group. In the context of the Tehran Bazaar, a large differentiated
society, this is not feasible and interviews were only a sample. Second,
since my initial contacts (or sample) were not random, and the sub-
sequent references were most de¬nitely not random, this introduces the
problem of selection bias at the level of sources,64 wherein members of
the Bazaar are not interviewed randomly (if I were to conduct regression
analysis I would have to be concerned with the potential for creating
autocorrelation between observations and error terms between the
observations that are greater than zero). Thus, in order to guard against
selection bias in sources, I used multiple and independent initial entrees.
More important though, the overall bias involved in the interview pool
was mitigated by the use of different interviewing techniques and other
methods to check facts. As a general rule, I have followed the journalist™s
rule of securing an absolute minimum of two independent and credible
(preferably primary) sources for each fact. At face value this seems like a
modest bar for fact checking, but it proved to be a high enough bar to
weed out much information.
Three types of interviews were conducted within the Bazaar:
Structured interviews were arranged in advance and typically took place
in the Bazaar. The interviews consisted of preplanned questions that
solicited basic information about the interviewees™ business and work
history as well as a series of more open-ended questions about their
views of the state™s role in the economy, the changes in the Bazaar
structure, and policy prescriptions. Interviews ran from thirty minutes to

I attempted to write up my notes immediately after the interviews were ¬nished;
however, this was not always possible and often the interviews and discussions were
recorded only later. Thus, quotes in the manuscript are not always exact and should not
be treated as verbatim.
Ian S. Lustick, ˜˜History, Historiography, and Political Science: Multiple Historical
Records and the Problem of Selection Bias,™™ American Political Science Review 90
(1996), 605“18.
The puzzle of the Tehran Bazaar under two regimes 33

an hour and a half, and at the end of each interview I sought referrals for
additional interviews.
Follow-up semi-structured interviews were conducted in almost all cases.
These interviews, which were often not prearranged, were by far the most
fruitful encounters and provided the bulk of the qualitative material. They
consisted of (1) thanking the interviewee for their initial cooperation, (2)
verifying information gathered in the initial meeting, (3) fact checking
information from other sources, (4) gathering oral histories from older
bazaaris, and (5) soliciting more information in areas that the interviewee
felt were important and where I believed they had special knowledge.
These informal interviews occasionally took place in front of other
bazaaris who would also participate. The presence of other people
(apprentices, customers, colleagues, relatives, or my friends) added a new
dynamic, introducing new topics or different presentations of issues. This
allowed me to evaluate information and identify sensitive topics. The
length of these follow-up interviews ranged from ¬fteen minutes to several
hours and ideally entailed several meetings. Fortunately for me (but
unfortunately for my interviewees), business was hardly brisk during my
¬eldwork stay, and merchants had time to discuss matters.65
Random interviews were conducted in order to offset the potential for
sample bias involved in the snowballing method. These consisted of
spontaneous discussions with members of the Bazaar. I would brie¬‚y
explain who I was and my research and then sought to ask questions
from the structured interviews. The quality of response was highly
variable and interviews ran from a couple of minutes to half an hour. If
the interviewee was busy, but seemed genuinely interested in partici-
pating, I would arrange for a meeting at a future date.
At times, I will refer to statements made by bazaaris in a preliminary
survey. This is because initially I designed a pilot survey and planned to
distribute it during interviews. The questions were derived and tested
during a research trip in the summer of 1999. Thirty-two surveys were
completed. However, this method was highly unsuccessful since (1)
most bazaaris were wary of the written method and (2) they viewed this
as too time consuming. On a number of occasions I was told that they
would answer all my questions and more, if I put away ˜˜that paper.™™
Consequently, the survey was abandoned and the questions were sub-
sumed into the various interviews.
Interviews with non-bazaaris involved shopkeepers and wholesalers
outside the Bazaar area to gain a wider perspective on economic issues,

Over time I was able to identify the best days (Saturdays) and times in the day (early
afternoon) to conduct interviews.
34 Bazaar and State in Iran

uncover how the Bazaar was viewed by other commercial actors, and
check information from interviews in the Bazaar. Also, included in this
category are the interviews I conduct during a one-week trip to Dubai in
May 2001. There I met a number of Iranian businessmen who operated
import“export businesses between Iran and the UAE, some of whom
had been based in the Tehran Bazaar. In December 2000, I spent one
week in southern Iran and visited two of the three free trade zones
(Qeshm and Kish) and several ports (Bandar Abbas, Bandar Lengeh,


. 7
( 55 .)