Economics growth and the middle class in an economy in transitions the case of russia

Economic Studies in Inequality, Social Exclusion
and Well-Being
Series Editor: Jacques Silber

Zoya Nissanov

Economic Growth
and the Middle
Class in an Economy
in Transition
The Case of Russia

www.ebook3000.com


Economic Studies in Inequality,
Social Exclusion and Well-Being
Series editor
Jacques Silber, Ramat Gan, Israel


More information about this series at />
www.ebook3000.com


Zoya Nissanov

Economic Growth
and the Middle Class
in an Economy in Transition
The Case of Russia

123


Zoya Nissanov
Department of Economics and Business
Management
Ariel University
Ariel
Israel

ISSN 2364-107X
ISSN 2364-1088 (electronic)
Economic Studies in Inequality, Social Exclusion and Well-Being
ISBN 978-3-319-51093-4
ISBN 978-3-319-51094-1 (eBook)
DOI 10.1007/978-3-319-51094-1
Library of Congress Control Number: 2016962034
© Springer International Publishing AG 2017
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or

for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Printed on acid-free paper
This Springer imprint is published by Springer Nature
The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

www.ebook3000.com


Acknowledgements

I would like to thank Jacques Silber for his constructive comments. I also thank
Raphael Franck and Grazia Pittau for their comments on earlier versions of some
of the chapters. Finally, special thanks to the “Russia Longitudinal Monitoring
survey, RLMS-HSE”, conducted by the National Research University Higher
School of Economics and ZAO “Demoscope” together with Carolina Population
Center, University of North Carolina at Chapel Hill and the Institute of
Sociology RAS, for making these data available.

v


Contents

1 What Does the Middle Class Refer To? . . . . . . . . . .
1.1 Importance, Measurement and Characteristics
of the Middle Class . . . . . . . . . . . . . . . . . . . . . . .
1.1.1 Importance of the Middle Class . . . . . . . .
1.1.2 Measurement of the Middle Class . . . . . . .

1.1.3 Characteristics of the Middle Class . . . . . .
1.2 The Russian Middle Class . . . . . . . . . . . . . . . . . .

.............

1

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

1
1
3
5
7

2 On the Transition in Russia . . . .
2.1 Russia in Transition . . . . . . .
2.2 Data . . . . . . . . . . . . . . . . . . .
2.2.1 Data description . . . .
2.2.2 Panel Dataset . . . . . .
2.2.3 Summary Statistics . .

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.

.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.

.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.

.

.
.
.
.
.
.

9
9
11
11
13
14

3 Distributional Change and What Happened to the Middle
Class in Russia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1 Review of the Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.1 Relative Distribution . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.2 Decomposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.3 Relative Polarization . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2 Empirical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.1 Relative Polarization . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.2 Decomposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.
.
.

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.

.
.
.
.
.
.

21
22
22
23
24
25
25
29
35

4 Bipolarization and the Middle Class in Russia .
4.1 Review of the Literature . . . . . . . . . . . . . . . .
4.1.1 Bipolarization Curves . . . . . . . . . . . .
4.1.2 Measures of Bipolarization . . . . . . . .
4.1.3 Extensions of the FW Measure . . . . .

.
.
.
.
.

.

.
.
.
.

.
.
.
.
.

.
.
.
.
.

37
37
38
40
41

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.

.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.

.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.

.
.
.
.

.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.

.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.
.
.

.
.
.

.
.

vii

www.ebook3000.com


viii

Contents

4.1.4 Decomposition of the Bipolarization Measure . . . . . . .
4.1.5 Inequality and Bipolarization . . . . . . . . . . . . . . . . . . .
4.1.6 Bipolarization and Mobility. . . . . . . . . . . . . . . . . . . . .
4.2 Empirical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.1 Bipolarization Measures . . . . . . . . . . . . . . . . . . . . . . .
4.2.2 Decomposition of the Bipolarization Measure (FW)
by Income Sources . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.
.
.
.
.

42
43
44

45
45

....
....

51
55

....
....
....

57
57
57

.
.
.
.

.
.
.
.
.

.
.

.
.
.

.
.
.
.
.

5 On Polarization in Russia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.1 Review of the Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.1.1 Polarization: Definition and Properties . . . . . . . . . . . .
5.1.2 Measuring Polarization with an Arbitrary Number
of Poles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.1.3 Decomposition of Polarization Indices . . . . . . . . . . . .
5.2 Empirical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.2.1 Measuring Polarization . . . . . . . . . . . . . . . . . . . . . . . .
5.2.2 Decomposition of Polarization Measures
by Income Sources . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.
.
.
.

58
62
62

62

....
....

65
68

6 The
6.1
6.2
6.3

Socio-Economic Characteristics of the Middle Class .
Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Empirical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7 Income Mobility and the Middle Class . . . . . . .
7.1 Methodology: Income Mobility . . . . . . . . . .
7.2 Empirical Results . . . . . . . . . . . . . . . . . . . . .
7.2.1 Mobility Between Income Groups . . .
7.2.2 Income Growth . . . . . . . . . . . . . . . . .
7.2.3 Mobility Within Income Groups . . . .
7.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . .

.
.
.
.

.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.

.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

71
71
74
92

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.

.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.

.
.
.

.
.
.
.
.
.
.

93
93
94
94
98
99
101

8 Concluding Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Appendices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113


List of Figures

Figure
Figure
Figure

Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure

2.1
2.2
2.3
2.4
2.5
2.6

2.7
2.8
2.9
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
3.10
3.11
3.12
3.13
3.14
3.15
3.16
3.17

Figure 4.1
Figure 4.2

Average incomes (in June 1992 rubles) . . . . . . . . . . . . . . . .
Equivalent incomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ratio of mean to median . . . . . . . . . . . . . . . . . . . . . . . . . .
Share of zero-incomes . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Gini index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Theil entropy measure . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Between-group inequality . . . . . . . . . . . . . . . . . . . . . . . . . .
Within-group inequality . . . . . . . . . . . . . . . . . . . . . . . . . . .
The share of the middle class (equivalent incomes) . . . . . . . .
PDF overlays for 1992–1995 . . . . . . . . . . . . . . . . . . . . . . .
PDF overlays for 1995–1998 . . . . . . . . . . . . . . . . . . . . . . .
PDF overlays for 1998–2001 . . . . . . . . . . . . . . . . . . . . . . .
PDF overlays for 2001–2003 . . . . . . . . . . . . . . . . . . . . . . .
PDF overlays for 2003–2005 . . . . . . . . . . . . . . . . . . . . . . .
PDF overlays for 2005–2008 . . . . . . . . . . . . . . . . . . . . . . .
PDF overlays for 1992–2008 . . . . . . . . . . . . . . . . . . . . . . .
Decomposition of the relative distribution for 1992–1993 . . . .
Decomposition of the relative distribution for 1993–1995 . . . .
Decomposition of the relative distribution for 1995–1996 . . . .
Decomposition of the relative distribution for 1996–1998 . . . .
Decomposition of the relative distribution for 1998–2001 . . . .
Decomposition of the relative distribution for 2001–2002 . . . .
Decomposition of the relative distribution for 2002–2005 . . . .
Decomposition of the relative distribution for 2005–2008 . . . .
Decomposition of the relative distribution for 1992–2008 . . . .
Decomposition of the relative distribution for 1992–1996
and 1996–2005 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
First degree bipolarization curve . . . . . . . . . . . . . . . . . . . . . . .
Second degree bipolarization curve . . . . . . . . . . . . . . . . . . . . .

14
15
15
16
16
16

17
17
18
25
25
26
26
27
27
29
29
30
31
31
32
32
33
33
33
34
39
39

ix

www.ebook3000.com


x

List of Figures

Figure 4.3
Figure 4.4
Figure 4.5
Figure 4.6
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure

4.7
4.8
4.9
5.1
5.2
5.3
6.1

6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
6.10

Foster-Wolfson bipolarization measure based
on Lorenz curve . . . . . . . . . . . . . . . . . . . . . . . .
FW-index . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
WT-index . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Distance between mean and median incomes
(divided by median) . . . . . . . . . . . . . . . . . . . . .
PG index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FW-index for positive incomes . . . . . . . . . . . . .
Gini index for individual incomes . . . . . . . . . . .
DER-index, a = 0.5 . . . . . . . . . . . . . . . . . . . . . .
Identification component of the DER-index . . . .
DER-index for positive incomes . . . . . . . . . . . .
Estimated proportions of the groups . . . . . . . . .
Estimated normalized incomes of the groups . . .
Age of the main earner . . . . . . . . . . . . . . . . . . .
Households where the main earner is male . . . .
Households in rural area . . . . . . . . . . . . . . . . . .
Households in Moscow and St. Petersburg . . . .
Households in Northern Caucasus . . . . . . . . . . .
University graduates . . . . . . . . . . . . . . . . . . . . .

Pensioners (retired main earners) . . . . . . . . . . . .
Self-employed main earners . . . . . . . . . . . . . . . .

...........
...........
...........

41
45
47

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

48

50
52
52
63
64
66
77
77
81
81
81
82
82
83
83
83

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.


List of Tables

Table
Table
Table
Table
Table
Table
Table
Table
Table
Table

Table
Table
Table
Table

2.1
2.2
2.3
2.4
2.5
3.1
3.2
4.1
4.2
4.3
4.4
4.5
4.6
4.7

Table
Table
Table
Table

4.8
4.9
4.10
4.11

Table 4.12
Table 5.1
Table 5.2
Table 5.3
Table 5.4
Table 5.5
Table 5.6
Table 5.7

Macro-indicators for 1992–2008 . . . . . . . . . . . . . . . . . .
The details of the samples . . . . . . . . . . . . . . . . . . . . . .
Number of observations in the panel datasets . . . . . . . . .
Gini index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The share of the population belonged to the middle class.
Relative polarization for 1992–2008 . . . . . . . . . . . . . . .
Relative polarization for 1992–1996 and 1996–2005 . . . .
FW-index for individual and household incomes. . . . . . .
The FW-index for equivalent incomes . . . . . . . . . . . . . .
The main significant changes in the FW-index over time .
WT-index (θ = 1, r = 0.5) . . . . . . . . . . . . . . . . . . . . . .
PG index for equivalent incomes. . . . . . . . . . . . . . . . . .
The FW-index for positive incomes. . . . . . . . . . . . . . . .
The main changes over time in bipolarization
for individual positive incomes . . . . . . . . . . . . . . . . . . .
The share of the income sources in total income . . . . . . .
The Gini index for the various income sources . . . . . . . .
The FW-index for the various income sources . . . . . . . .
Decomposition of the FW-index by income sources
using the Shapley decomposition procedure . . . . . . . . . .
Decomposition of the changes in the FW-index . . . . . . .

The DER-index (α = 0.5). . . . . . . . . . . . . . . . . . . . . . .
The main significant changes in the DER-index (α = 0.5)
over time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The components of the DER-index (α = 0.5) . . . . . . . . .
The DER-index for positive incomes (α = 0.5) . . . . . . . .
The main changes over time in polarization
for individual positive incomes . . . . . . . . . . . . . . . . . . .
The DER-index (α = 0.5) for the various income sources.
Decomposition of the DER-index (α = 0.5) by income
sources using the Shapley decomposition technique. . . . .

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

10
12
13
17
18
28
34

46
46
47
48
49
51

.
.
.
.

.
.
.
.

.
.
.
.

52
53
53
54

...
...
...

54
54
63

...
...
...

64
64
65

...
...

66
66

...

67
xi

www.ebook3000.com


xii

List of Tables

Table 5.8
Table 5.9
Table 5.10
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table

6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
6.10
6.11
6.12

7.1

Table 7.2
Table 7.3
Table 7.4
Table 7.5
Table B.1

Decomposition of the DER-index (α = 0.5) by income
sources using the Araar (2008) method . . . . . . . . . . . .
Shapley decomposition of changes in the DER-index
(α = 0.5) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Decomposition of changes in the DER-index (α = 0.5)
using Araar’s (2008) method . . . . . . . . . . . . . . . . . . .
The BIC values of the mixtures up to 5 components . . .
Parameters of the components . . . . . . . . . . . . . . . . . .
Socio-economic characteristics of the poorest group . . .
Socio-economic characteristics of the middle class . . . .
Socio-economic characteristics of the richest group . . . .
Ordered probit estimation. . . . . . . . . . . . . . . . . . . . . .
Marginal effects after ordered probit estimation for 1992
Marginal effects after ordered probit estimation for 1995
Marginal effects after ordered probit estimation for 1998
Marginal effects after ordered probit estimation for 2001
Marginal effects after ordered probit estimation for 2004
Marginal effects after ordered probit estimation for 2007
Mobility between the components
(percentage of the households in the base year) . . . . . .
Ordered probit estimation for the change in component
(decrease, no change, increase) . . . . . . . . . . . . . . . . . .

Average annual equivalent income growth rates
per groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Mobility within the groups (in percentage of the group
in 1995). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Estimated proportions of the components for three types
of datasets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The components of the FW-index . . . . . . . . . . . . . . . .

....

67

....

68

.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.

68
74
76
78
79
80
84
86
87
88
89
90
91

....

95

....

96

....

98

....

100

....
....

100
110

.
.
.
.
.
.
.
.
.
.

.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.


Abstract

This book attempts to study the evolution of the middle class in Russia during the
last decade of the twentieth century and the first one of the twenty-first century.
This is a critical and even fascinating period because it follows the dissolution
of the USSR in 1991 and covers what is known as the transition period, during
which many fundamental reforms were implemented. This study focuses on income
and is based on the RLMS dataset (Russian Longitudinal Monitoring Survey).
This book includes eight chapters, most of them taking a somehow different
view of the middle class and implementing a distinctive measurement technique.
Chapter 1 presents a survey of the literature on the middle class. Chapter 2

intends to provide a description of the economic situation in Russia during the
transition period. Although the analysis in all the other chapters is based on the
distribution of incomes, each chapter takes a specific view of the middle class and
adopts generally a different methodology to apprehend this concept. Thus, while the
focus of both Chaps. 4 and 5 is on polarization, the analysis in Chap. 4 is based on
the idea that there should be a clear link between the middle class and the notion of
bipolarization while that in Chap. 5 is centered on the idea that the potential for
social conflict is related to the degree of polarization. But the measures used in the
empirical sections of these two chapters are completely different. Similarly, while
Chap. 3 tries, on the basis of the “relative distribution” approach, to discover the
main features, in terms of both location and shape, of the change that took place
over time in Russia in the distribution of income, Chapter 6, using the so-called
“mixture model”, aims at finding out how many different groups could be derived
from the income distribution in Russia and what the main socio-economic and
demographic characteristics of these groups are. The goal of Chap. 7 is to check
whether, during the period examined, many individuals moved in and out of the
middle class, and what the determinants of such mobility are. Chapter 8 finally
summarizes the main findings of this study.
Focusing on equivalent incomes, during the whole 1992–2008 period, it seems
that while inequality did not change and average income increased, the rich group
moved closer to the median (in relative terms), so that the gap between the poor and
rich and the ratio of the mean to the median decreased; thus, bipolarization, when
xiii

www.ebook3000.com


xiv

Abstract

measured by indices such as those proposed by Foster and Wolfson (FW) or Wang
and Tsui (WT), decreased, indicating an increase in the share of the middle class.
A mixture model estimation confirms this trend since it shows an increase in the
total percentage of the two middle groups from 56% in 1992 to 68% in 2008.
Assuming an arbitrary number of poles, between-group inequality which is a proxy
for alienation and within-group inequality which represents a loss of identification
did not change; as a result, polarization over the whole period did not change.
Despite the stability of inequality and polarization and a decrease in some bipolarization measures, relative polarization estimation shows that the income distribution diverged between 1992 and 2008 and the mass of the middle class in the
median-adjusted distribution decreased, moving mostly to the lower quartile. Thus,
one can conclude that the share of the middle class increased only if the impact of
changes in the median income is not isolated.
It seems therefore that one should be very careful in drawing conclusions based
on traditional bipolarization and polarization indices because they are not able to
make a distinction between the impact on the relative importance of the middle
class of an increase (or decrease) in the average (and median) income and that of a
variation in the shape of the income distribution (changes in higher moments of the
distribution). The relative distribution is on the contrary able to make such a very
useful distinction.


Introduction

Birdsall et al. (2000) claim that policymakers should take into consideration the
effects of economic policies on the middle class not only because of the requirement
of a healthy market economy for the active involvement of middle groups, but also
because reforms, which are critical to market-based economies in a developing and
transitional world, cannot be sustained over long periods if the middle income strata
does not grow and if those in the middle perceive themselves as losing as a result of
reforms. Thus, the fluctuations in the size of the middle class may be especially

important in economies that are in transition from one economic system to another,
such as post-Soviet Russia. A growing middle class during the period of
socio-economic transformations in those countries could be seen as an important
indicator of the effectiveness of reform. The role to be played by a strong middle
class should therefore be a serious policy concern in transitional economies.
To determine who belongs to the middle class, however, is a complex task,
among other reasons because of the scarcity of data sources that would encompass
all the potentially relevant variables. In addition, as will be clear from the review
of the literature, there is no agreement, among those who have attempted to define
the middle class, about the most important features of the middle class, although the
amount of income available is almost always mentioned. But even when the main
focus is on the income level, there is no consensus concerning the critical thresholds, those that distinguish the middle class from the poor and from the rich.
There is hence a need to be very careful when attempting to assess the size of the
middle class in a given country, to find out whether its importance grew over time,
to detect its main characteristics or to determine whether the identity of those
belonging to the middle class does not change or varies a lot over time.
Notwithstanding these caveats, the present book attempts to study the evolution
of the middle class in Russia during the last decade of the twentieth century and the
first one of the twenty-first century. This is a critical and even fascinating period
because it follows the dissolution of the USSR in 1991 and covers what is known as
the transition period, during which many fundamental reforms were implemented.
This study will focus on income and will be based on the RLMS-HSE dataset
(Russian Longitudinal Monitoring Survey—Higher School of Economics).
xv

www.ebook3000.com


xvi

Introduction

RLMS-HSE1 is a series of nationally representative surveys conducted by the
National Research University Higher School of Economics and ZAO “Demoscope”
together with the Carolina Population Center, University of North Carolina at
Chapel Hill and the Institute of Sociology RAS. This dataset should be useful in the
context of a research on the transition, as it covers the relevant period and monitors
the effects of reforms on the health and economic welfare of households and
individuals in the Russian Federation. A repeated cross-section as well as a longitudinal analysis will be undertaken.
This book includes eight chapters, most of them taking a somehow different
view of the middle class and implementing a distinctive measurement technique.
Chapter 1 is entitled “What Does the Middle Class Refer to?”. This introductory
chapter presents a survey of the economic and sociological literature on the middle
class and reviews the various definitions which have been proposed to characterize
the middle class. It also discusses the potential link between a strong middle class
and sustained economic growth.
Chapter 2 (“On the Transition in Russia”) provides a description of the economic
situation in Russia during the transition period, looking at the main
macro-indicators. The chapter describes also the database on which this study of the
middle class in Russia is based.
The title of Chap. 3 is “Distributional Change and What Happened to the Middle
Class in Russia”. It intends to analyze changes over time in the relative importance
of the middle class, using data on the distribution of income in Russia between
1992 and 2008. Another method used here to determine who belongs to the middle
class is based on the concept “relative distribution” (see, Morris2 et al. 1994;
Bernhardt et al. 1995; Handcock and Morris 1998, 1999). This nonparametric
method allows a comparison of two distributions in terms of their ratio. The
advantage of the non-parametric estimation is that, unlike the parametric approach,
which assumes that the data are drawn from a known distribution, this procedure
does not require that the data fit a parametric distribution. In the second part of the

chapter, a decomposition technique is implemented that enables one to make a
distinction between the impact of changes in the median income and in the shape
of the distribution of income.
The purpose of Chap. 4 (“Bipolarization and the Middle Class in Russia”) is to
analyze the determinants of changes in income bipolarization in Russia between
1992 and 2008. A survey of the literature on bipolarization measurement is first
presented and the link between this concept and that of the middle class is
explained. The empirical part gives first the value of various measures of the degree
of bipolarization in Russia during the 1992–2008 period. Then, one of these
measures is decomposed, using the so-called Shapley decomposition procedure

1

Available at />Actually, the aspects of the relative PDF and CDF as a basis for the comparison analysis were
examined earlier by, for example, Parzen (1977 and 1992), Eubank et al. (1987), Cwik and
Mielniczuk (1989 and 1993).

2


Introduction

xvii

(see Shorrocks 1999), in order to estimate the contribution of different income
sources to the change that took place over time in bipolarization.
In Chap. 5 (“On Polarization in Russia”) the concept of polarization, rather than
that of bipolarization, is described. This concept is different from that of bipolarization, first because it is assumed to capture the formation of any arbitrary number
of local poles in the income distribution (see, Duclos, Esteban and Ray 2004). But
this alternative approach is also relevant because it is grounded on different theoretical premises, namely the notions of “identification” and “alienation”

(see, Esteban and Ray 1994). The idea is that polarization is positively related to the
“alienation” that individuals or groups feel one for another, but also to the sense of
within-group “identification”. Since “alienation” and “identification” are assumed
to be critical determinants of the potential for social conflict, this approach
emphasizes the possible social implications of a “polarized” income distribution. In
addition it has the ability to estimate the degree of polarization for the total population without having to divide the sample into an a priori defined number of
groups. The chapter presents estimates of the DER (Duclos, Esteban and Ray)
polarization index in Russia during the 1992–2008 period. This index is then
broken down on the basis of both the so-called Shapley decomposition procedure
and the method proposed by Araar (2008). The goal of this decomposition is again
to determine the contribution of the various income sources to the change in
polarization over time.
The purpose of Chap. 6 (“The Socio-Economic Characteristics of the Middle
Class”) is to find out whether the Russian society is stratified in groups which can
be defined on the basis of household incomes. More precisely, the chapter aims at
finding out how many groups may be identified and what their main
socio-economic and demographic characteristics are. The so-called mixture model
method is used to identify the groups. This is a semi-parametric method which
enables one to model unknown distributional shapes. The advantage of this
approach is that it allows representing sub-populations (income groups in our case)
and parameters of their densities without having to define in advance the number
and characteristics of these groups.
The RLMS data may be used to conduct a cross-section as well as a longitudinal
analysis. Thus, when income groups are identified and all the individuals or
households are allocated to these groups, the panel dataset allows one to examine
mobility between and within groups over time. This is precisely the goal of Chap. 7
(“Income Mobility and the Middle Class”), whose focus is on the relationship
between the middle class, individual income growth and mobility. Its main purpose
is to examine the impact of some household characteristics on the probability to
move from one component to another. More specifically, an attempt is made to

determine the factors (like age, gender, locality, education and work status) which
affect the probability that a household that belonged to the middle class in the first
post-Soviet years would remain in the middle class. The chapter examines also the
factors which affect the probability that a poor household manages to improve its
economic situation and enter the middle class or that a rich one moves down to the
middle class. The chapter tries also to shed some light on the relative income

www.ebook3000.com


xviii

Introduction

growth of the middle class, as compared with that of people belonging to the high
and low classes, and to test which group benefits the most from income growth,
using both an anonymous and a non-anonymous approach.
Concluding comments (Chap. 8), based on the results presented in the various
chapters, summarize finally the main findings of this study. This study should
contribute to a better assessment of the main factors affecting the formation of a
sizable middle class in Russia and of the determinants of the changes in its relative
size that followed the implementation of economic reforms. This book may also
help better understanding the effect of various socio-demographic variables on
disparities in individual income growth during the transition period and the impact
these characteristics have on the probability to move along the income distribution.


Chapter 1

What Does the Middle Class Refer To?

This chapter presents a survey of the economic and sociological literature on the
middle class and reviews the various definitions which have been proposed to
characterize the middle class. It also discusses the potential link between a strong
middle class and sustained economic growth. The chapter ends with a brief review
of the literature on the middle class in Russia during the transition period.

1.1
1.1.1

Importance, Measurement and Characteristics
of the Middle Class
Importance of the Middle Class

There is some evidence that the share of the middle class in a population is positively correlated with political stability, economic growth and the quality of state
institutions. Thus, a sizable middle class is an important factor in economic
development (Thurow 1984; Foster and Wolfson 1992 and 2010; Landes 1998;
Easterly 2001; Birdsall et al. 2000; Pressman 2007; Birsdall 2007b and 2010;
Boushey and Hersh 2012, among others).
Easterly (2001) found that an increase in the share of the middle class is associated with a rise in per capita income and that an increase in the share of the middle
income increases the growth rate. In addition, a greater income share of the middle
class leads to better health outcomes, higher levels of publicly provided health
services and higher levels of political rights and civil liberties, raises financial depth
and reduces consumer price inflation.

© Springer International Publishing AG 2017
Z. Nissanov, Economic Growth and the Middle Class in an Economy
in Transition, Economic Studies in Inequality, Social Exclusion
and Well-Being, DOI 10.1007/978-3-319-51094-1_1

www.ebook3000.com

1


2

1

What Does the Middle Class Refer To?

Birdsall et al. (2000) point out that the political support and economic participation of the middle class are critical to market-driven economic growth and, thus,
to poverty reduction in the long run. Furthermore, Birdsall (2007a) links a small and
weak middle-income group to weak state institutions which are the main problem of
non-sustainable growth.
Pressman (2007) claims that a large and vibrant middle class contributes to
economic growth, as well as to social and political stability, since it helps
democracy flourish, mitigates class warfare and may be necessary for good
macroeconomic performance.
Amoranto et al. (2010) considered that the middle class is “important in its
demand for better goods and services and keeping governments accountable”.
Loayza et al. (2012), using a cross-country panel dataset containing information
on 128 countries, found that the increase in the size of the middle class was
accompanied by more activity in social policy on health and education and by
improvement in the quality of governance regarding democratic participation and
official corruption.
Bussolo et al. (2014) point out that the expansion in the size of the middle class
in developing countries can lead to a shift in demand towards more globalization
supportive policies and is likely to promote policy goals such as improved transparency, intensified anticorruption efforts, and demand for a more open society and
cleaner environment.

Boushey and Hersh (2012) claim that a strong middle class is important for
economic growth in the following ways: (i) it promotes the development of human
capital and a well-educated population; (ii) it creates a stable source of demand for
goods and services; (iii) it incubates the next generation of entrepreneurs, and
finally, (iv) it supports inclusive political and economic institutions, and, thus,
underpin economic growth.
These empirical regularities evidently raise questions regarding the conditions
which favor the emergence of the middle class.
Berkowitz and Jackson (2005) point out that an equitable income distribution is
conductive to the formation a powerful middle class because it enables the existence of institutions that are critical for market economies.
Birdsall (2007b and 2010) defined inclusive growth as growth which builds a
middle class and implies an increase in the share of the middle class (implying that
some people exit poverty) and the proportion of total income they command
(implying gains at the “expense” of either the initially poor or the initially rich).
According to Pressman (2007), macroeconomic conditions might affect the size
of the middle class. More precisely, while economic expansion, which creates new
jobs and generate higher wages, enables more households to earn enough money to
enter the middle class, recession can throw people out of work, decrease wages and,
thus, reduce the size of the middle class. Also important is governments’ spending
on transfer payments directed toward low- and middle-income families, including
negative taxes to workers. This leads to greater overall income equality and,
therefore, raises the share of the middle class. In addition, Pressman (2007) found
that liberal welfare regimes tend to have the smallest middle classes and the largest


1.1 Importance, Measurement and Characteristics of the Middle Class

3

declines in size of the middle class over time. Conversely, social democratic states

tend to have the largest middle classes; moreover, these countries managed to
maintain the size of their middle classes despite adverse economic circumstances.

1.1.2

Measurement of the Middle Class

As pointed out by Pressman (2007), there are three meanings of the middle class:
sociological, economic and self-identification. These approaches were described by
Levy and Michel (1983) as follows. According to the first one (sociological), the
middle class is measured by a set of behavioral characteristics and socio-economic
criteria. Thus, the middle class include individuals “who have achieved a certain
educational level, whose jobs have a certain level of social status, and who have a
particular set of values and attitudes”. The economic meaning implies having an
income level that is somewhere in the middle of the income distribution. The last
one (self-identification) is the most problematic since it allows people to identify
themselves as the middle class even if they are not.
Foster and Wolfson (1992) suggested the following steps in measuring the
middle class: (i) choosing a “space” (individual/family/household income, salary,
expenditure etc. in “income-space” or “people-space”); (ii) defining the middle
(median or mean income, the individual at the 50th percentile); (iii) fixing the range
around the middle (identifying the middle class by fixing a percentage interval
above and below median or mean); (iv) aggregating the data.
There is however no agreement on how the middle class should be defined. Even
when the main focus is on the income level, there is no consensus concerning the
critical thresholds, those that distinguish the middle class from the poor and from
the rich. Various approaches and definitions have been used to measure the size of
the middle-income group. Needless to say, different definitions lead to different
results.
Some researchers proposed relative measures based on the “income space”. In

this case, the thresholds of the middle class are multiples of the median income (or
consumption) or based on certain income quintiles. The advantage of these relative
measures is that they are sensitive to changes in the income distribution across and
within countries and over time.
While Thurow (1984) defined the middle class as those with incomes between
75 and 125% of the median income, Lawrence (1984) set the middle-class brackets
at approximately 66% and 132% of men’s median weekly earnings. Blackburn and
Bloom (1985) broadened the middle income range to 60–225% of the median and
Davis and Huston (1992) defined the middle class as including those families
whose incomes are between 0.5 and 1.5 times the current-year median incomes.
Some studies used as a specific share of the income distribution to define the
middle class. For instance, Barro (1999) and Easterly (2001) computed the share of
the middle class as the three middle quintiles.

www.ebook3000.com


4

1

What Does the Middle Class Refer To?

Conversely, other studies rely on an absolute measurement, which defines the
incomes corresponding to the upper and lower bound of the middle class, usually in
terms of purchasing power parity (PPP). Milanovic and Yitzhaki (2002), for
example, defined the world middle class as all those with an annual income higher
than 3470$ and lower than 8000$ in the PPP terms.
Other studies in this line of research include Banerjee and Duflo (2008), where
the middle class is defined (in the context of developing countries) as the households whose daily per capita expenditures valued at purchasing power parity varied

either between 2$ (the international poverty line) and 4$ or between 6$ and 10$.
Ravallion (2009) proposed a definition close to that of Banerjee and Duflo (2008),
identifying the developing world’s middle class as those living between 2$ and 13$
PPP a day. Similarly, Kharas (2010) considered as global middle class all households with daily per capita incomes between 10$ and 100$ in PPP and in the report
of the Asian Development Bank (2010) the middle class is defined as those with
daily per capita consumption expenditures of 2$–20$ PPP.
Torche and Lopez-Calva (2012, p. 11) also identified the middle class as the
households which are “above the threshold of absolute economic deprivation, as
defined by the poverty line, but below the threshold that ensures virtually no risk of
falling into poverty”.
Birdsall (2007b) combined relative and absolute measures and identified the
middle class as people at or above the equivalent of $10 day, and at or below the
90th (95th in the developing world according to Birdsall 2010) percentile of the
income distribution in their own country.
Another possibility to identify the middle group in an income distribution is to
apply the semi-parametric “mixture model” approach (Pittau and Zelli 2006; Pittau
et al. 2010; García-Fernández et al. 2011).
Another approach to defining the middle class is based on the “people-space” as
exemplified by Levy (1987). In this case, the size of the middle class is fixed and
can neither shrink nor expand. The focus is then on the evolution of the income
share, in contrast to the “income-space” where the population size is the main
concern (Atkinson and Brandolini 2013). Levy (1987) defined the middle as the
50th percentile and the range from the 20th to the 80th percentile is identified as the
middle class.
When examining the middle class, Atkinson and Brandolini (2013) set limits in
both the people and income space. They argued that although it is normal among
economists to think of classes as income groupings, the concept of “class” requires
the examination of some other dimensions beyond income, like property and the
position in the labor market.
Some other researchers suggested that the definition of the middle class should

include characteristics other than income. For instance, Wheary et al. (2007),
focusing on American families, devised a set of formulas for measuring five key
benchmarks of economic security in terms of assets, education, ability to meet
essential expenses, health care access and housing costs. If three or more of the
factors in a family’s profile supported financial security, they considered that family
as part of the middle class.


1.1 Importance, Measurement and Characteristics of the Middle Class

5

The Pew Research Center (2008, 2016), explored the middle class in America
and, in addition to the households included in the middle-income group, looked also
at those who self-identify as the middle class. It claimed that a class should be
defined not only in terms of incomes but also reflect “a state of mind”, so that “it
could be a matter of self-identification”.

1.1.3

Characteristics of the Middle Class

Kenny (2011) writes that it is hard to find a set of characteristics that can consistently and uniquely identify the middle class across countries and time. However,
based on different definitions, he concluded that the middle class members are those
who are “confident of not slipping into deprivation”, “secure in their assets” and
“can spend on things other than basic necessities”.
Similarly, the Australian Government report (2007, p. 4) defines the global
middle class as “those who have scope for discretionary expenditure over and above
the basic necessities of life such as food, clothing and shelter; but who, at the same
time, face some constraint on that expenditure. Members of this group may, for

example have access to items such as consumer electronics, cars, the possibility of
owning their own housing, leisure travel – whether domestically or internationally –
and, perhaps most importantly, the ability to invest in higher levels of education for
the next generation. At the same time, their access to these things is, in general,
restricted by some significant budgetary constraint – unlike the ‘rich’”.
Following Birdsall et al. (2000), information on occupation and education can
provide a basis for defining the middle class “that has been broadly understood and
reasonably comparable across countries (even given large differences in average
incomes among countries)”. They however note that a high education level that
provided opportunities in the past is “no longer sufficient for upward occupational
mobility”. More precisely, while in the 1960–1970s a secondary education was
sufficient to attain the middle-class living standard, by the 1990s “it neither guaranteed a well-paying job nor protection from falling into poverty”.
In developed countries, it is usually considered (see Hayashi 2007) that the main
characteristics of the middle class are “economic independence, strong professional
orientation and a civic sense of values”. In addition, the middle class consists of
small entrepreneurs, lawyers, medical doctors, engineers, scholars and urban
white-collar workers.
Furthermore, Amoranto et al. (2010) argued that the degree of political activism
driven by the middle class is greater than that of the upper and lower classes. They
found that the middle class is related to higher education, more skilled and stable
jobs, and the ability to save.

www.ebook3000.com


6

1

What Does the Middle Class Refer To?

Wheary et al. (2007), focusing on American families, found that 65% of the
respondents belonging to the middle class were married; 76% were “white”
(non-Hispanic); and 85% lived in urban areas. The average age was 44 years old
and the average family size was 2.8.
In developing countries, the characteristics of the middle class are somewhat
different. Birdsall (2010) found that the middle class in most developing countries
has close to or above 10 years of education. This average is below that of the OECD
countries where a high school education is a minimum for a middle class status. The
situation is even worse if the middle class is defined as those between $2 and 10$,
which is the definition adopted by Banerjee and Duflo (2008). In that case, the
educational levels of the household heads are lower than or equal to the average
educational level of their countries. Concerning the employee status, Birdsall (2010)
reported a small distinction between the poor and the middle class, at least in the
cases of Turkey and the Dominican Republic, and noted that the households in an
income range between $2 and 10$ “enjoy regular if low-wage participation in the
formal sector”.
Banerjee and Duflo (2008) claimed that “nothing seems more middle class than
the fact of having a steady well-paying job” and pointed out that members of the
middle-class prefer having “an income coming in every month and not just the
income itself”. The authors found many entrepreneurs among the middle class, they
however argued that it is mostly because middle-class members are still relatively
poor and “every little bit helps” and “if they could only find the right salaried job,
they might be quite content to shut their business down”. Banerjee and Duflo (2008)
also found that the households belonging to the middle class have smaller families
and fewer children and they spend much more than the poor on education and health.
According to the report by the Asian Development Bank (2010, p. 27), the
middle class in Asia, compared to the poor, is “less connected to agriculture, less
likely to own land and less likely to be wage laborers. It is much more likely to hold
salaried jobs, has a greater propensity for migration, a higher propensity to seek

more expensive medical care when ill, and has fewer children and invests more in
health, nutrition, and schooling. The middle class is also better educated, and more
geographically concentrated (in urban areas or along coasts)”.
Torche and Lopez-Calva (2012) investigated the Latin American middle class,
taking Chile and Mexico as case studies, and found that a higher social class is
associated with a higher education level and a higher occupational status. They
pointed out that, unlike the US where there is stratification in marriage rates and
female-headed households, the proportion of female-headed and married-headed
households in Chile and Mexico hardly varies across classes; and, unlike some
developing countries, such as South Africa, the demographic factors do not matter.
The report by the Asian Development Bank (2010) concluded that the two main
factors in the creation and sustenance of a middle class are higher education and
stable and secure well-paid jobs.


1.2 The Russian Middle Class

1.2

7

The Russian Middle Class

Remington (2010) points out that the Russian leadership (as well as the American
one) recognizes that “the development of a larger middle class would be beneficial:
by reducing the polarization between rich and poor, it would improve the provision
of growth-enhancing public goods, reduce social tension and increase stability, and
would make government more effective and accountable”. Thus, an increase of the
middle class is a serious policy concern in modern Russia.
The Russian middle class estimations vary from 2% (Shkaratan et al. 2003) to

80% (according to self-identification criteria) of the country’s population. The
reason for such a discrepancy is that researchers investigating this issue use different definitions of the middle class. Bobkov (2014) writes that the lack of a
uniform acceptable definition of the middle class does not allow to evaluate the real
size of its group in Russia, and, thus, makes it difficult to formulate a policy aimed
at increasing the share of the middle class. Moiseev (2012) points out that due to
differences in approaches, Tichonova and Mareeva (2009) divided the researchers
into different groups; at one end there are those who deny the existence of the
middle class in Russia, at the other those who insist on the presence of the sizable
middle class.
Based on the material position (income/consumption and housing criteria), the
size of the Russian middle class in 2008 was in the range of 15% (Rosgosstrach and
Levada Center) to 19% (Bobkov 2014).
Most of the studies conducted by Russian researchers however relied on a
sociological approach where income is not considered as the main indicator
(Trusova 2001; Maleva 2002 and 2008; Carnegie Moscow Center 2003; Avramova
2008; Tichonova and Gorunova 2008; Maleva and Ovcharova 2009; Tichonova
2009; Beliaeva 2011). They argue that the middle class cannot be defined by one
criterion and mostly use Max Weber’s approach that assumes a multidimensional
definition of the middle class, including all the three “meanings” described by Levy
and Michel (1983) mentioned above.
The Russian Academy of Sciences’ (RAS), Independent Institute for Social
Policy which is one of the leading academic institutions in Russia in the field of
middle class analysis, uses the following criteria: the income level (equal to or
higher than the median for a given region), the educational level (at least
secondary-special), the work characteristics (not physical) and self-identification.
According to their report (2014), the Russian middle class increased between 2003
and 2008 from 29% to 34%. Then, due to the economic crisis, it decreased but since
2010 it increased again and in 2014 reached 42% of the total population.
There are a few studies that used relative measures based on the “income space”.
For example, Birdsall et al. (2000) suggested as a measure of the middle class the

share of households with per capita income in the range of 75%–125% of the
median household per capita income. They thus found that the Russian middle class

www.ebook3000.com