Appendix: Survey Methodology

Full text

Page 1

Appendix: Survey Methodology
BORIS KUZNETSOV1& SAUL ESTRIN2
1Bureau of Economic Analysis, Moscow, Russia
2Centre for New and Emerging Markets, London Business School, Regent’s Park, London
NW1 4SA, UK
Comparative Economic Studies (2003) 45, 204–212. doi:10.1057 /palgrave.ces.8100013
The survey was designed to enable the analysis of the relationships between
performance, ownership, corporate governance, restructuring and finance
among privatised Russian enterprises. The questionnaire was developed
between mid-1998 and end-1999, and tested in two pilot runs in 1998 and at
the beginning of 2000. Significant corrections were made after the first pilot
and some minor amendments after the second. The full survey of 437 firms
was undertaken in the spring of 2000. In this appendix, we first outline how
the population frame was stratified to meet the sampling criteria with respect
to ownership, size, sector and region. We go on to compare the major
indicators of our sampled firms with those of the Russian industrial
population as whole.
The survey data were collected by direct face-to-face interviews with one
of the top managers of the enterprise: in most of the cases the general
manager (director/general director) or economic/financial director.1In fact,
43% of the respondents were Company Directors and 50% were Deputy
Directors. Hence, we were successful in obtaining high-level respondents; an
important fact given that a number of our key variables are qualitative and
perceptual.2
1In Russia, the top manager position may have different names. In the ‘Director’ category we
include: Director General, Executive Director, Acting Director, Director (if he is the only one with a
title). In three cases, respondents were presidents of the company and in one the Chairman of the
Board. The category ‘Deputy Director’ includes Deputies of the Top Manager and in one case Chief
Engineer. In the ‘other’ category there are mostly Heads of Departments (Planning, Economic, etc).
2For some questions where pilot surveys showed a low response rate, several options were
given to the respondent. For example, if it was impossible to get information on separate shares for
workers and managers, the revised instrument included an option to report the cumulative share of
insiders. This layering of data is used in the papers, see for example, Estrin and Angelucci in this
volume.
Comparative Economic Studies, 2003, 45, (204–212)
r 2003 ACES. All rights reserved. 0888-7233/03 $25.00
www.palgrave-journals.com/ces

Page 2

The questionnaire contained six major blocks of questions:
*
indicators of economic activity and factors of production (output,
capacity and labour utilisation, costs, financial in- and outflows,
structure of assets, investment activity, etc);
information on restructuring activities of the enterprise (such as
shedding of labour, introduction of new technologies, new products, etc);
market structure data (competition, price elasticity, enterprise activity on
different geographical markets);
ownership and corporate governance data (ownership structure, owner-
ship concentration, board composition, some information on top
management of the enterprise);
data on financial constraints (availability of external financing, state
assistance, etc);
a block of control variables such as region, industrial code, legal type of
enterprise, date and method of privatisation and others.
*
*
*
*
*
Where feasible, the data were generally collected for the years 1997–
1999; to cover pre- and post-crisis years. In composing the questionnaire, we
tried to duplicate information regarded as most important; hence, for
example, more than one question (quantitative, rank, qualitative) would be
included to permit construction of an indicator for main characteristics of the
firm.
SAMPLING STRATEGY
The survey was undertaken to enable evaluation of the adjustment of Russian
privatised industrial enterprises to new conditions in the transition economy
of the late 1990s. This rationale, together with the restricted number of
enterprises surveyed (around 400 enterprises were to be interviewed),
determined the character of the sample. It is clearly impossible to seek
representativeness of ‘Russian industry’ in a country so regionally diverse as
Russia with such a small sample. We therefore stratified by region, size,
industry, ownership form and age, and then undertook random sampling
within each sampling frame. The initial list of enterprises from which the
sample was drawn was based on the Goskomstat Enterprise Registry
data included in ALBA-Y database. The registry includes information for
about 30,000 medium and large Russian industrial enterprises, accounting for
65%–85% of output and employment in the selected industries. Utilising this
database enabled us to use historical time series data in the analysis and at
B Kuznetsov & S Estrin
Survey Methodology
205
Comparative Economic Studies

Page 3

the same time did not significantly narrow the population of firms from
which to select.
Selection of industries
We based our work on the population of firms in the following manufacturing
industries according to Russian Industrial Classification (OKONH).3We
judged that the economic issues arising from privatisation and restructuring
were illustrated most clearly in the manufacturing sector. The following two-
digit industries were selected:
13 – Chemical and oil-chemical industry;
14 – Machinery building and metal-working industry;
15 – Wood and paper industry;
16 – Stone and clay (production of building materials) industry;
17 – Light industry;
18 – Food industry.
Our sampling was meant to ensure that the firms were evenly distributed
across the chosen two-digit industries. However, in practice, it proved difficult
to meet these quotas in some industries, given the need to sample only
privatised enterprises and to adhere to the size and regional stratification
criteria as well. The actual distribution slightly over-represents the machinery
sector relative to the industrial population (see Table A1) but there are
sufficient observations in each industry to control for industry-specific factors
in our analysis.
Size categories
We restricted the size of our sampled firms to be between 100 and 5,000
employees. Small enterprises with below 100 employees were excluded
because: (a) they work under specific tax and accounting rules that often
make them incomparable with others; (b) although extremely important for
market institutions in the long run, during this period they accounted for less
than four percent of industrial output in Russia, and (c) the SME sector
necessitates larger, more specific samples.
3The list does not include the ferrous and non-ferrous metal industries, because the
Russian Industrial Classification does not separate mining and metal production at the two-digit
level so the decision to exclude extractive industries precluded the inclusion of metal production.
Moreover, concentration in the ferrous metal industry is extremely high (12 Russian metal
plants produce about 90% of ferrous metal; the non-ferrous metal sector concentration is lower
but still very high), and metal industry enterprises tend to be large, and hence well above our
upper size limit.
B Kuznetsov & S Estrin
Survey Methodology
206
Comparative Economic Studies

Page 4

The upper size limit of 5,000 employees was chosen because in Russia,
although a small number of very big firms operate in most industries, these
have unique features that make them more suitable for case studies.4For
example, very large firms are often located in ‘mono-towns’ (see Commander
et al, 1996), or develop a special kind of relationship with the local or national
authorities. Table A1 shows that the sample was distributed more or less
evenly across the three size categories. The average firm employed 891
workers, rising from 265 in the smallest category to 1820 in the group of firms
employing more than 1,000 workers.
Selection of regions
The choice of regions was based on two considerations. The relatively small
size sample meant that we could not achieve representative sub-samples for
all or even the majority of 89 regions (oblast, kray, republics) of Russia. On
the other hand, given the regional diversity of the business and economic
environment in Russia, the regional dimension has to be explicit in any
analysis (see Brown and Earle, 2001). We decided, therefore, to select a
limited number of regions belonging to four macro-zones: European Russia,
the Volga, the Urals and Siberia. European Russia is represented by two
Russian capitals – Moscow and St Petersburg – together with their respective
oblasts (Moscow oblast and St Petersburg oblast); three regions belong to
Volga macro-zone – Nizhny Novgorod, Samara, Volgograd; the Ural macro-
zone is also represented by three regions – Chelyabinsk, Perm, Ekaterinburg
(Sverdlovskaya oblast), while the Siberia macro-zone includes enterprises
from Novosibirsk, Krasnoyarsk and Omsk. We recognise that our approach
does not permit reliable analysis of all regional specifics; regional policies can
differ significantly within one macro-zone (Tatarstan, Uliyanovsk and Nizhny
Novgorod regions in Volga zone provide well-known examples). Never-
theless, in many cases the geographical position itself and the distance from
Moscow are likely to be factors contributing to the macro-zone economic
environment and enterprise behaviour.
Table A1: Distribution of firms by size in 1999
Total 100–500%501–1,000% >1,000%
Selected industries43714733.6% 13931.8% 15134.6%
Source: Authors’ calculations
4See Kuznetsov and Muravyev (2000) for an example of a study of very big Russian firms.
B Kuznetsov & S Estrin
Survey Methodology
207
Comparative Economic Studies

Page 5

We did not apply strict regional quotas because for the purposes of this
study, we regarded size and industrial stratifications as the most important
criteria. Indeed, we intended initially to survey in only eight regions, with one
region in each macro-zone ‘reserved’ in case the size and industry quotas
could not be otherwise met. This resulted in some ‘reserved’ regions having a
smaller number of observations. Table A2 reports the regional distribution of
the sample.
Establishment year and form of ownership
Our research explores questions concern post-privatisation behaviour in
former Soviet industrial enterprises, so only firms existing before 1992 (the
beginning of transitional reforms in Russia) were eligible for selection. In
practice, the vast majority of de novo firms were anyway excluded as a result
of the lower size restriction. We also excluded fully state-owned enterprises
though our population contained some ‘mixed’ state–private joint-stock
companies including a few where the controlling stake belongs to Federal or
Regional authorities.
Table A2: Regional distribution of firms
RegionNo. of firms % of total
Central macro-zone subtotal
Moscow
Moscow oblast
St Petersburg
St Petersburg oblast
122
36
41
30
15
27.9
8.2
9.4
6.9
3.4
Volga macro-zone subtotal
Nizhny Novgorod oblast
Samara oblast
Volgograd oblast
115
64
39
12
26.3
14.6
8.9
2.7
Ural macro-zone subtotal
Ekaterinburg oblast
Perm oblast
Chelyabinsk oblast
111
50
43
18
25.4
11.4
9.8
4.1
Siberia macro-zone subtotal
Krasnoyarsk kray
Novosibirsk oblast
Omsk oblast
89
37
40
12
20.4
8.5
9.2
2.7
Total437 100.0
Source: Authors’ calculations
B Kuznetsov & S Estrin
Survey Methodology
208
Comparative Economic Studies