THE UNBALANCED DYNAMICS OF RUSSIAN REGIONS Towards a real divergence process

Journal of Economics and Business
Vol. VII – 2004, No 1 (11 – 37)


THE UNBALANCED DYNAMICS OF
RUSSIAN REGIONS
Towards a real divergence process



Frédéric Carluer1
University of Grenoble II

Elena Sharipova2
RECEP


Abstract

This article examines the uneven evolution of Russian regional per capita income
and productivities during the late 1980s and 90s. Following the studies by Baumol
and Barro and Sala-i-Martin, we examine the convergence process of the Russian
economy under two aspects. First we point out the sources of growth stressed by
the new growth theories (accumulation of physical capital, education effort, and
public expenditures). Secondly, we examine the economic-geography perspective,
by analysing the impact of the distance to Moscow and the North-East diagonal on
growth. This first application to Russian regional data confirms the regional

1 PEPSE-Espace Europe, University of Grenoble II, France: Email: Frederic.carluer@upmf-
grenoble.fr
2 RECEP; Email: esharipova@recep.ru.

© 2004 EAST-WEST University of THESSALY. ALL RIGHTS RESERVED

---

EAST-WEST Journal of ECONOMICS AND BUSINESS

divergence which results from the paradox of a beta-divergence (except in the case
of gross regional product), a fifteen years sigma-divergence and a weak
conditional convergence in the late 1990s, from both the macroeconomic and
geographical perspectives.

KEYWORDS: Russian regions, unbalanced dynamics, real divergence

JEL classification: L00; O41; R1


Introduction

This study of the convergence of Russian regions is especially interesting, in
regards to Russia’s secular feature of a center, the Moscow region, which is clearly
defined by its size, its location and above all its widely accepted leadership in
technologies and industry. Considering this, there is every incentive to analyse
polarization phenomena and, more generally, spatial-economic asymmetries, more
so when taking into account that Russia outstretches across eleven time zones.

For such a country, a regional analysis is useful from both the empirical and
theoretical points of view. Empirically, it is easier to compare data derived from
the same sources than to undertake international comparisons. Theoretically, the
assumptions made, such as those regarding the unicity of structures and
infrastructures as well as the institutional framework, preferences and available
technologies are directly relevant, since exchanges impediments do not exist
(except for natural barriers).

In the light of the difficulties encountered in ensuring harmonious development
within the European Union (Dunford, 1994 ; Armstrong and Vickerman, 1995;
Tumpel and Mooslechner, 2003; Carluer and Gaulier, 2002), despite the use of
structural funds and a decade of continuous growth (Artobolevskiy, 1997; Bachtler
and Turok, 1997; Jovanovic, 1997; European Commission, 1999), it is not
surprising that Russian regional disparities have been accentuated by the opening
of markets (Sapir, 1999; Babeski and Maurel, 2002; Kocenda and Evzen, 2001;
Carluer, Mercier and Samson, 2004). Indeed, only a small number of the Russian
regions have benefited from “cumulative growth”, on the contrary, the majority is
threatened rather, by the “poverty trap” (Azariadis and Drazen, 1990; Pritchett,
1997; Carleur, 2004).

Our approach is a three-stage one. First of all, a descriptive analysis of Russian
regional disparities is undertaken from the standard convergence perspective (beta-
12

---

Carluer, F., and Sharipova, E., The Unbalanced Dynamics of Russian Regions: towards a
real divergence process

and sigma-convergence are tested; section 1). We then use the equation of
conditional convergence to analyse the main sources of regional growth from a
macroeconomic (section II) and geographical perspective (section III). Thus we try
to measure the impact of investment, education and public expenditures on
regional growth and to shed light on the importance of geographical position.

Descriptive analysis

This study draws on data on per capita income (in nominal terms) for the 88
Russian regions (Annex 1), available to the Ministry of Economic and Finance and
obtained by the Russian European Center for Economic Policy (RECEP) in
Moscow since 1985. In order to avoid certain problems related to changes in
measurements or even in particular definitions, especially given the substantial
political and economic changes in the country during the transition, the data are
smoothed using the moving average method for three years, using weightings of
0.25, 0.5 and 0.25 respectively for the dates t-1, t and t+1. The initial (1985) and
final (1999) levels are not modified, so the procedure does not change the
regressions and calculus carried out in cross section. However, in the light of the
immense size of the country (see the map in Annex 2) and the cultural diversity of
its regions, as well as the differences in the degree of monetization of the economy
and the absence of regional price deflators, the results should be interpreted with
caution.

Relative performances

An initial analysis of Russian regional disparities reveals that the gap between the
groups of regions remains substantial. The ten richest regions have, on average, a
per capita income four times higher than the ten poorest regions in 1985 and the
trend is upward (more than six times higher in 1999). This is confirmed for the
twenty richest and poorest regions - but the gap increased less during the same
period, from 2.5 to 3.8. Nevertheless, the difference between the richest (Moscow)
and the poorest region (Ingushetia) reached an incredible level in 1999 (the
downward direction of the trend at the end of the period is due to slower growth
and the emergence of a contest for the national leadership between Yamalia, a
central northern region, and Moscow).

13

---

EAST-WEST Journal of ECONOMICS AND BUSINESS

Figure 1. Gaps between richest and poorest regions: per capita income

20
18
16
14
12
Max/Min
t
i
os 10
10Max/10Min
Ra
8
20Max/20Min
6
4
2
0
85
87
89
1
3
5
7
99
19
19
19
199
199
199
199
19
Years

14

---

Carluer, F., and Sharipova, E., The Unbalanced Dynamics of Russian Regions: towards a
real divergence process

From a geographical perspective, the ten worst performing regions, in terms of per
capita income, are mainly located in the Caucasus (South-West) and near the
Mongolian frontier (South). Moreover, their position and performance remained
stable from 1985 to 1999, with six of the ten poorly performing regions always in
the bottom ten (Figure 2). Only one region, Khakasiya, really soared up the table,
leaping nearly 50 places. Except for this one spectacular case of leapfrogging, no
process of convergence is revealed.

The performance of the top ten regions is also characterized by considerable
stability: eight of the ten richest regions maintained their superiority and five
increased their lead. It was mainly the Eastern and the Northern regions that out-
performed the rest, and particularly regions in the east of the Urals “frontier” (i.e.
Siberia ; Show, 1987 or Portnov, 1994): Tyumenskaya, Khanty-Mansiiky and
Yamalo-Nenetsky. Thus we notice that the richest regions are located close to the
poorest ones, such as Komi-Permyatsky. Lastly, the spectacular progress of the
Moscow region should be noted, since it occupies no fewer than ten of the top
eleven places (and therefore does not feature in Figure 3). Clearly, there is now a
genuine capital effect in Russia.

15

---

EAST-WEST Journal of ECONOMICS AND BUSINESS


Figure 2. Per capita income deviations for the ten laggard regionsla
y
vo-
k
r
m
e
n
e
t
s
a
i
-
P
cha
her
C
i
r
y
m
insky
ygeya
r
a
a
g
d
agesta
a
hakasiya
abardino-B
Bu
Ko
A
A
D
Ingushskaya
K
Alt
K
K
0
-10
-20
nces -30
IpC85
-40
IpC99
-50
-60
Average differe
-70
-80
Regions





16

---

Carluer, F., and Sharipova, E., The Unbalanced Dynamics of Russian Regions: towards a real divergence process

Figure 3. Per capita income deviations for the ten richest regions
350
300
s
e 250
c
n
200
Regions
IpC85
i
f
f
ere
d
IpC99
e 150
g
era
v 100
A
50
0
a
a
y
y
a
y
y
y
-
k
a
y
ya
ya
k
a
lo-
k
t
i
ya
s
k
t
y
iis
a
t
sky
ska
ska
ksk
ku
a
ot
e
t
ska
ns
m
n
ans
n
han
e
l
i
n
a
r
y
a
ans
a
a
Y
o
h
huk
K
Ya
Ne
rm
m
K
C
M
u
kh
mc
gad
M
a
a
Tyu
S
a
K
M



17

---

EAST-WEST Journal of ECONOMICS AND BUSINESS

Beta-convergence

In order to evaluate more precisely the convergence of per capita income we could
apply the two concepts of beta and sigma-convergence. The former refers to the
existence of a negative relationship between the initial level of income and the
growth following growth. It is a necessary but not sufficient condition for the
sigma-convergence to be verified. This second concept merely shows the variance
reduction of productivities in cross section between two dates. It should be noted
that, in the case of strong asymmetrical regional shocks, the per capita income
dispersion does not diminish even in the presence of the beta-convergence. The
results of a standard empirical analysis of regional disparities using the hypothesis
of sigma-convergence and beta-convergence corroborate this trend.

The estimate will be performed in cross section (88 regions) as it is customary to
do so (these results are confirmed by a time series estimation). Given that the
sigma-convergence is the strongest and the most intuitive3 definition of
convergence, we will refer to this concept in order to decide on the presence or
absence of convergence.

Clearly, there is no absolute convergence for the regional incomes per capita in the
1985-99 period (Table 1). A divergence process is nearly highlighted. This is
obvious for the industrial output for the last years of the century. On the contrary,
on a similar period, a strong beta-convergence could be noted for the gross
regional product.


3 However a sole statistics cannot sum up the evolutions of the regional productivities spread. An
extended study, stressing on the emergence of convergence clubs for example, will require the use of
tools developed by Quah [1996 a,b] or Durlauf and Quah [1999].
18

---

Carluer, F., and Sharipova, E., The Unbalanced Dynamics of Russian Regions: towards a real divergence process




Table 1. Beta-convergence of income, gross regional product and industrial output per capita



Period
Beta
Constant
DW
R²


+0.78 %
+0.66


Income
1985-99
(1.41)
(59.64)
2.083
0.151
[0.161]
[0.000]


-13.6 %
0.52


Gross regional product
1994-99
(-7.94)
(23.85)
1.793
0.651
[0.000]
[0.000]


+0.92 %
-0.013


Industrial output
1995-2000
(2.64)
(-2.21)
1.882
0.274
[0.010]
[0.030]
(.) = Student; [.] = p-value.


19

---

EAST-WEST Journal of ECONOMICS AND BUSINESS


Sigma-convergence

The further study of sigma-convergence confirms this preliminary diagnosis: the
variance ratio increases by nearly 40% during the nineties (Figure 4), and there is
no reversal of the trend. It shoud be noted that, on the contrary, when we consider
prices series, a significant convergence clearly appears at the end of the 1990s
(Babeski and Maurel, 2002). Otherwise, the intensity of the uneven process is
reinforced by the fact that the ten leading regions (in particular the last ones) and
the ten lagging regions (especially the middle ones) swap positions (as shown by
Figures 2 and 3) inside these two specific groups.
20

---