Download doc - RCEF2016

1
Econometric model of Russian Federation: what about the perspectives of
growth?
A final version
Sergey Mitsek
Dean of the Faculty of Business and Management
Liberal Arts University in Yekaterinburg, Russia
Doctor of Economics
Address: 620049 Russia, Yekaterinburg, Studencheskaya str. 19, ap.412, Phone number:
+79122815576, Fax number: +7(343)3740252, E-mail: [email protected]
JEL classification codes for the paper: C320, C510, C520, C530, C540, E170, E210, E220,
E230, E240, E270, E310, E470, E510, E520, E620, F170, F470, O420, O470, O520
Key words: Russian Federation; econometric model; external shocks; international sanctions;
international trade; capital flight; exchange rate; inflation; fiscal policy; monetary policy
JEL classification codes of the author specialization fields: C100, C130, C150, C200, C210,
C220, C320, C510, C520, C540, C820, D240, E000, E010, E170, E200, E210, E220, E230,
E240, E270, E310, E370, E410, E430, E470, E510, E520, E580, E620, E630, F410, F430,
F470, L000, L160, O400, O410, O420, O440, O470, O520, O570, P240, P250, P280, R110,
R150
Abstract
The paper describes a new author’s version of econometric model of the Russian
economy which is tailored to analyze current trends in Russian economy and to forecast its
dynamics for the next years. Its other task is to show how different factors and policy
2
instruments affect the main macroeconomic variables under various scenarios of the external
economic situation and of economic policy options.
In this version such new variables as monetary base, the government consumption’
deflator, the aggregate demand index, and international reserves were included in the model
(the first two were taken as exogenous). The money mass is represented as endogenous
variable (and corresponding equation is added to the model). All the equations of the model
were re-estimated based on the enlarged samples of the new time series presented by Rosstat.
The new indicators such as indexes of physical volumes of major national accounts’
aggregates were included also.
The model consists of 25 equations and 38 identities that describe the relationships
between 73 variables. They consist of 10 exogenous and 63 endogenous variables. Among the
firsts there is the capital account balance, the Bank of Russia’ key loan rate, the monetary
base, the reserve requirement, economically active population, government consumption
deflator and export and import prices. The main macroeconomic indicators such as the GDP
volume, different internal price indexes, investment in fixed assets from different sources of
financing, bank loans and deposits, employment and average wages, and the ruble-to-dollar
exchange rate were included as the endogenous variables in the model.
The parameters of the most of equations were estimated by ML – ARCH method
(some of them were estimated by least squares). The quarterly data for Q1 1999 – Q4 2015
(that is, 68 points) were used as a sample. The estimates of equations’ parameters are
presented in the Appendix 3.
The model showed that in the short run the rates of Russian economy will be negative.
If all exogenous variables will change with the average rates they had during last three years
before 2014 the average annual growth rate of Russian economy for the next four years is
3
slightly positive of about 1 % annually while inflation remains as strong as about 6 %
annually.
Under the assumption of the restoration of export prices’ growth the average annual
GDP growth increases but only slightly. Under the assumption of rapid import prices increase
the recessions strengthens sufficiently. Active monetary policy increases the GDP annual rate
not very much and at the expense of much higher inflation. The latter strengthens if the
salaries in government sector increase sufficiently. Aggressive fiscal policy has a strong
negative impact on gross capital formation.
The model demonstrates a relative weakening of external economic factors and
strengthening a role those of the internal ones.
4
1. Introduction
Over the past two years (2014 – 2015) the Russian economy demonstrates a dramatic
economic decline. The average annual rates of the GDP growth were equal to -1.05 %, much
lower than in 2000 – 2008 (6.37 %) and in 2009 – 2013 (2.06 %).1 The rate of reduction of
gross fixed capital formation was even more - -3.78 % annually in constant prices (11.72 % in
2000 – 2008 and 3.46 % in 2009 – 2013). As a result fixed capital volume in constant prices
decreased in absolute terms.
The ruble devaluation exceeded 2.2 times that caused the jump of CPI index to more
than 10 % annually. The economic decline caused the reduction of total tax payment and a
budget deficit – for the first time since the early 2000s.
At first glance it may seem that the troubles for Russian economy began just in the
middle of 2014 after the drastic drop of the world crude oil prices (the average dollar index of
export prices dropped by 38 % during 2014-2015) and by introduction of international
sanctions. But we shall prove that the backgrounds of today’s crisis were laid long before
current events. Understanding the causes of the current state of Russian economy will explain
what had happened.
The first and the most important reason of the crisis came is the stagnation of total
factor productivity (TFP) since 2008.2 It was caused by a complex set of reasons, and the
main are:

Technological gap;

Exhaustion of free economic capacity inherited from the Soviet Union;

Natural reasons;

Structural factors
5
Let’s describe all of them briefly. Technological gap of Russia in many fields is
widely recognized.3 Russia’s lagging in ICT is worth to be mentioned specially4 as various
studies support the conclusion that the leap forward in productivity in developed countries
took place due to advances in this field.5 A significant fixed capital wear and tear as a
consequence of low investment in technological modernization is another one indicator of
technological gap.6
Russian rapid economic growth in 2000 – 2008 can be explained in large degree by the
better use of production capacity inherited from the Soviet Union. Thus the degree of capacity
utilization in Russian industry increased from 36 % in 1995 to 60 % in 2007 but then stopped
to grow with the exception of mining.7 That is this source of growth has mostly been
exhausted.
Among “natural” reasons we can mention the exhausting of mature fields and
movement to hard-to-reach ones in Eastern Siberia and in Arctic shelf in oil extraction and in
mining in general.8 For rapidly developing sectors such as trade, services and finance the
“natural” reason of the slowdown a gradual saturation of demand (poorly satisfied in the
Soviet Union) can be mentioned.
The structural factors include a noticeable drop of total productivity in such rapidly
growing sectors as mining, trade and government services,9 and a dramatic drop of the growth
rate of construction industry (the latter caused by general slowdown of investment activity)
since 2008.10 Some other reasons of productivity decline can be mentioned also.11
The second reason of the crisis came is a dramatic decrease of the rates of national
labor force growth since 2008 (first time since the Second World War) due to low birth rates
and aging.12 The country was not ready to such changes and tries to compensate the shortage
by inviting of immigrants.13
6
The third reason of current crisis was a rapid increase of wages which outpaced that of
marginal revenue on labor (up to 2008): the annual rates in real terms were 6.26 % for the
former and 5.03 % for the latter in 2000 – 2008 when deflated by the GDP deflator.14 Such
trend partly caused by government policy15 and partly by structural factors.16 It led to decline
of profitability of the economy and of gross fixed capital formation as a consequence.17
The fourth precursor of future crisis there was a rapid increase of households’
consumption (by 9.76 % in real terms annually in 2000-2008)18 enhanced not only by growing
incomes but by rapid development of consumer credit and of government expenditures. A
decline of national saving and of gross fixed capital formation was a result (see data in
Appendix 4).
Finally the fifth reason which led economy to a crisis was a worsening of terms of
trade since 2012 year (see Appendix 4). Index of dollar export prices stopped to grow since
2013.19 The real dollar depreciation20 and the decrease of real import prices in 1999 – 2008 led
to a rapid growth of import.21 As a result we see a dramatic decrease of the net export as a
share of the GDP (from 12.4 % average in 2000 – 2008 to 7.0 % in 2014 – 2015).
We analyze and explain current trends in Russian economy and to forecast its
dynamics for the next 2 - 3 years using an econometric model we constructed for this purpose.
An additional important function of this model is to show how different factors affect the key
macroeconomic variables under the different variants of macroeconomic policy and scenarios
of the external economic situation.
Among the econometric models of Russian economy we mention first of all those
made by Basdevant (2000), Aivazian et al. (2006, 2013a, 2013b) and Benedictow et al.
(2013). Other works use models in order to analyze the impact of different policy aspects on
Russian economy besides (see e.g. Alexeev et al. 2003; Jensen et al., 2004; Rutherford et al.,
7
2005). BOFIT’s scientists (Kerkela, 2004) used computable general equilibrium (CGE) model
to study the effects of price liberalization in Russian energy markets and vector autoregressive
(VAR) model for the analysis of the role of oil prices in Russian economy (Rautava, 2002).
Merlevede et al. (2009) estimated a macroeconomic model in order to analyze of the impact
of oil price on Russian economic performance.
We have developed a macro econometric model in order to answer the following
questions.
1.
What are the main driving forces of Russian economy?
2.
What are its perspectives in the coming years?
3.
What are the role of fiscal and monetary policy and the impact of the external
economic’ and the demographic factors on the dynamics of Russian economy?
4.
What are the consequences of different scenarios of fiscal and monetary policy
and of external shocks for it?
The actual data for our research were obtained from official Russian statistical sources
– Rosstat,22 The Bank of Russia23 and RIM.24
Our model is a system of equations and identities that describe the relationships
between different variables. For given values of exogenous variables (which represent fiscal
and monetary policy and external and demographic factors) it is solved to determine the
values of endogenous ones.
The estimated equations are interpretable in accordance with economic theory and
satisfy standard statistical tests of time series and of residual properties and of parameter
stability. The model gives a satisfactory explanation of past dynamics of the endogenous
variables.
In section 2 we present a brief description of methods we used to estimate the
parameters of the model. Section 3 gives a detailed description of model’s equations and
8
identities. In sections 4 and 5 we discuss the post forecasts of exogenous variables and
different forecasts’ scenarios for Russian economy for coming years. In section 6 we try to
explain the results we got. Section 7 concludes. Appendixes contain the total list of variables
and equations and identities, main statistical data, econometric estimations and the main
results of the model calculations.
In our approach we follow the traditions of econometric modeling established by Klein
(1950), Klein and Goldberger (1955), Fair (1984, 1994, 2005), and by the creators of the
Wharton, MPS, DRI Brookings, and authors of other well-known models.25
2. Estimation Methods
The model is fully recursive.26 Its parameters were estimated mostly by Maximum Likelihood
– Autoregressive Conditional Heteroskedasticity (ML – ARCH) method, some of them by
Ordinary Least Squares (OLS).27 Unrestricted VAR models were used as the first step to
specify the equation.
Such procedures as the Breush-Godfrey test for serial correlation,28 ARCH and
Breusch – Godfrey – Pagan tests to test for heteroscedasticity;29 the Quandt – Andrews and
Chow breakpoint tests for parameter stability,30 the Augmented Dickey-Fuller (ADF) test for
unit root as for series and for residuals in each specification to test cointegration,31 and other
tests were performed for each equation to verify the quality of estimates.32 Standard errors and
covariance are White heteroskedasticity-consistent in OLS,33 Bollerslev-Wooldridge robust
standard errors and covariance specification was used in ML – ARCH estimation.34 The
results of the most important statistical coefficient and tests are shown in Appendix 3.
The final selection of estimated equations was carried out in accordance with the
following criteria:
1.
equation has appropriate statistical properties;
9
2.
the signs of parameters are consistent with economic theory;
3.
all independent variables are statistically significant and
4.
the inclusion of the equation in the model [as a system of equations]
demonstrates the best values of Theil coefficients in a post forecast imitations carried out by
means of the entire model.
3. Model
3.1.
General description of the model
The model consists of 25 equations and 38 identities that describe the relationships between
73 variables. They consist of 10 exogenous and 63 endogenous variables. A total list of
variables in alphabetical order is given Appendix 1. The list of all equations and identities of
the model in mathematical form is given in Appendix 2. The estimation outputs of equations
you can see in Appendix 3. The symbols used in estimation outputs that make them shorter
are presented also in Appendix 1.
Among the exogenous variables there is the capital account balance, the Bank of Russia’ key
loan rate, the monetary base, the reserve requirement, economically active population,
government purchases’ quantity and prices’ index, export and import prices’ dollar indexes
and transportation tariffs’ index. All the variables can be grouped in seven units:
1. Social Unit;
2. The Investment Unit;
3. The Production Unit
4. The Prices’ Unit
5. Monetary Unit
6. Bank Unit
7. Fiscal Unit
10
8. Foreign Economy Unit
A list of variables grouped in units is presented in Appendix 1.
The supply side of the model consists first of all of two-factor production function and
the equations of labor and fixed capital supply. All these equations are part of Production
Unit. The gross fixed capital formation that determines the supply of fixed capital is a part of
Investment Unit and is determined in turn by three equations and one identity:
a) Equation of investments financed by the gross profit;
b) Equation of government investments;
c) Equation of investments financed by the bank loans;
d) Identity that sums up all these three types.
The outcomes of Production and Investment units depend in turn on both supply and
demand factors. Thus the supply of production is determined by capital and labor in Cobb –
Douglas production function. But the latter depends also on demand factors which are
presented by aggregate demand index and by variables that reflect external demand.
The investments financed by the companies’ profit depend on its volume; the
government investments depend on the amount of taxes collected; investments financed by
the bank loans to companies depend on the amount of the latter. But at the same time private
investments depend on their profitability (the net marginal revenue on fixed capital is used as
a proxy for the latter)35 and on prices on investment goods.
In turn gross and net corporate incomes are determined by well-known
macroeconomic identities. Bank credits are determined in the Bank Unit and depend on the
volume of deposits and the latter depends in turn on households and corporate incomes. Ruble
and currency credits and deposits are determined by separate equations as that they depend on
different factors in Russia. The volume of taxes collected depends on income produced by the
economy; they are calculated in the Fiscal Unit.
11
The supply-side in the number of employees’ equation is represented by the number of
economically active population while the demand one by the difference between marginal
revenue on labor and average wages per employee.
The demand side of the model is represented by the variables of internal and of
external demands that together with money mass help to determine the price level. The
aggregate demand index constructed specially in the identity 41 depends on indexes of
households’ consumption, gross fixed capital formation and government purchases (internal
demand) and on external demand, that is, on indexes of export (positively) and of import
(negatively). Variables that determine the aggregate demand index influence on some other
variables of the model separately.
The level of households demand depends on consumer prices, on incomes and on
consumer credit. The external demand (export and import) depend on internal demand and on
prices on export and import goods and on taxes on external economic activity. The
government demand is an exogenous variable in this model.
The Price Unit determines the different prices indicators such as the GDP deflator, the
consumer price index and the gross fixed capital formation deflator. These indicators depend
on monetary variables, on supply and demand factors, on variables of Foreign Economy Unit
(the dollar rate and export and import prices) and on regulated prices (transportation tariffs).
The prices on government purchases are exogenous.
The GDP deflator depends also on supply change. The latter is determined by identity
47 as the difference between aggregate supply and aggregate demand. When this difference is
positive the prices reduce and vice versa that moves the model towards equilibrium.
The price level in turn affects the amount of nominal incomes, that is, gross profit and
gross wages first of all. The incomes, on the other hand, affect the elements of aggregate
demand and supply of factors of production.
12
The Monetary Unit determines the volume of money mass and the interest rates. The
latter variables have an impact on liquidity indicators and on the volumes of bank deposits.
The liquidity in turn have an impact on the volume of production and investment and of bank
deposits.
Finally the dollar exchange rate which is critically important for Russian economy
depends on export and import prices and on the level of liquidity.
3.2.
Data
The model was estimated mostly on quarterly time series for the period Q1 1999 to Q4
2015 (as the data accuracy for the previous periods is questionable). The largest part of data is
from Rosstat (2016a). Data for money mass and for banks’ credits and deposits are from the
Bank of Russia (2016); data for corporate income tax and ruble to dollar exchange rate are
from Institute of Economic Forecasting (IEF, 2016). Based on Rosstat (2016a) and IEF raw
data we calculated cumulative indexes for prices and for GDP and its elements.
3.3.
What is new in the 2016 year version of the model?
In this year’ version of the model we had done the following innovations.
1.
Some new variables were included in the model. Among them there is
monetary base, the government consumption’ deflator and the aggregate demand index (the
first two were taken as exogenous). The indexes of physical volumes of GDP and its elements
calculated anew can also be viewed as new variables.
2.
The indirect taxes were split in three variables such as export and import duties
and other indirect taxes.
3.
The money mass is represented as endogenous variable now (and
corresponding equation is added to the model).
13
4.
All the equations of the model were re-estimated based on the enlarged
samples of the new time series presented by Rosstat. The specifications of the majority of
equations were changed consequently.
3.4.
Detailed description of the model: equations and identities36
Equation 1 shows that fixed capital stock depends on gross capital formation.37 The
variables in the equation are deflated by the gross fixed capital formation deflator. Appendix 4
demonstrates statistical tables of annual growth rates of variables included in the model in
different periods, main macroeconomic ratios and elasticity calculated by means of estimated
equations. Together with equation 1 they show that rapid increase of investment in 2000 –
2008 determined a significant increase of fixed capital in 2009 – 2013 (its quickest increase
took place in 2005 – 2011) but then its dramatic drop after. As a matter of fact in 2014 – 2015
Russian economy demonstrates a loss of fixed capital.
Equation 2 shows that number of employees depends on the economically active
population and on difference between net marginal revenue on labor and average wages but
negatively on the share of gross wages in the GDP. The number of economically active
population represents the supply factors of labor in this equation; two other variables
represent the factors that affect the demand on labor.38 The data in Appendix 4 show that
supply factors dominate and this is an explanation of relatively low level of unemployment in
Russia (usually less than 6 %).39
Equation 3 is the Cobb-Douglas, constant returns to scale production function.40
Production per employee here depends not only on capital per employee but also on lagged
GDP values per employee (that reflects seasonal fluctuations in Russian economy), and on
some other variables that reflect on output. Among the latter we see the indicators of external
and internal demand, the tax burden, and the indicators of liquidity.
14
The crux of the matter lies in the fact that traditional economic theory defines the
production function as maximum output for a given amount of production factors’ input.41 But
in reality we estimate production function using actual but not maximum output.42 Taking
into account this fact we need to explain what factors will affect the gap between maximum
and actual output.43 We can suppose that the elements of demand will affect on this gap first
of all.44
The tax burden also affects the gap as it affects the optimism or pessimism of
producers (and their desire not to expand but to send earnings abroad as a consequence). The
liquidity level has also a crucial effect on Russian economy as we’ll show further.
Equation 3 and the figures of Appendix 4 can explain us that dramatic reduction of
growth rates of Russian economy that we can see successively first in 2009 – 2013 and then in
2014 – 2015 are explained by:
a) Reduction of rates of fixed capital formation;
b) Slowdown of total factor productivity;
c) Slowdown of aggregate demand;
d) Slowdown of external demand separately;
e) Dramatic shortage of liquidity.
Equation 4 shows that the GDP deflator depends on money supply, on export and
import prices and on dollar exchange rate, and on transportation tariffs that are under state
control. It depends negatively on volume of real GDP and on inventory change. An important
role of prices on government purchases we mention specially: the increase of their ratio to the
GDP deflator can serve as additional evidence.
Figures in Appendix 4 show a steady reduction of overall inflation in Russia. By
means of equation 4 it can be explained by:
a) Tightening of monetary policy;
15
b) Weakening of the impact foreign economic factors on inflation;
c) State control on tariffs’ growth and on prices on government purchases.
Identity 5 determines the volume of nominal GDP; identities 6 and 7 define elasticity
of output on capital and labor, respectively. A long-term elasticity of output on capital and
labor calculated by means of this equation are equal to 0.31 and 0.69 respectively.45 Identities
8 and 9 define the net marginal revenue on fixed capital and on labor, respectively.
Data in Appendix 4 show a dramatic reduction of growth of both indicators as a result
of a slowdown of total factor productivity. This is one more explanation of an overall
slowdown of Russian economy.
Equation 10 shows that average gross wage per worker depends on the net marginal
revenue on labor and on government purchases.46 The role of the first one can be explained by
microeconomic theory of a firm. Average real wages improved on 4 times in Russia since
1999 mostly due to the improvement of the marginal revenue of labor. The influence of
government purchases is obvious as
a) large share of employed operate under state order;
b) salaries paid in government institutions had a significance impact on salaries in
private sector.47
Further description of the model will give additional proof of important role of
government sector in Russian economy. Equation 10 explains that significant reduction of
growth rates of real wages in Russian economy is due to the same trend of both indicators – of
marginal revenue on labor and of government purchases (see Appendix 4).
Identity 11 defines the total wages in the economy.
Equation 12 shows that the households’ consumption depends on the volume total
income, on government purchases, on the level of consumer prices and on bank loans to
households and on import volume.48 Households’ consumption in Russia increased 2.8 times
16
since 1995 mostly due to remarkable growth of the GDP (on about 1.8 times) and of real
average wages and of real social transfers (both on about 4 times). Nevertheless we see a
dramatic decrease of annual growth of households’ consumption and equation 12 explains it
by:
a) significant reduction of incomes’ growth;
b) high inflation in consumer sector;
c) reduction of growth of government purchases and of consumer credit.
The equation of households’ consumption illustrates once more a significant role of
government in Russian economy. It also shows a strong dependence of Russian consumer
sector on import.
Equation 13 determines the level of consumer prices which depend on dollar exchange
rate, on export and import prices, on the volumes of export and import, on monetary
indicators, on consumer demand and on fiscal deficit. It is worth mentioning that the impact
of monetary variables on CPI is not very strong (the total value of elasticity is about 0.12)
while the intercept in the equation is rather large.
We can explain it by strong (though often indirect) government regulation of consumer
prices. A steady reduction of real consumer price index (when they are deflated by the GDP
deflator – see Appendix 4) until 2012 year is an additional support to such conclusion. At the
same time the impact of foreign economic variables on consumer prices is noticeable. That’s
why the annual rates of the latter don’t correspond exactly with the rates of monetary
variables but do it with the exchange rate. The increase of annual rates of consumer prices in
2015 can be partly explained by panic on markets which is illustrated by dummy in this
equation.
Identity 14 calculates the volume of households’ consumption in current prices.
17
Equations and identities 15 – 24 represent an Investment Unit of the model. Identity 15
calculates the level of national savings. Appendix 4 shows that annual rates of this indicator
dropped dramatically since 2008 due to a noticeable increase of households’ consumption and
government purchases in the GDP. That’s obvious that such trend serve as an obstacle to
investment.
Identity 16 shows that gross fixed capital formation volume in the model is equal to
the sum of three summands which reflect the financial origins of investment process: own
resources of business, state budget and bank loans.49
Identities 17 and 18 define gross and net corporate income respectively.50 Appendix 4
shows that the share of corporate incomes in Russian economy increased steadily due to
following reasons:
a)
reduction of the share of indirect taxes in Russian economy since 2008 due to
changes in rates and to structural changes;51
b)
a steady reduction of corporate income tax as the share of gross corporate
income due to privileges and tax evasion;
c)
a slowdown of wages’ increase since 2014.
These trends give some support to investment increase and economic growth.
Equation 19 shows that Moscow Interbank Credit Rate (MIACR) depends on
monetary indicators, on marginal revenue on capital and on dollar exchange rate. Its
significant dependence on the key rate of the Bank of Russia is remarkable and supports the
conclusion of importance of monetary policy we do later. Its strong dependence on the
marginal revenue on capital is evidence in favor of microeconomic theory and of
strengthening of market mechanisms in Russian financial market.
Equation 20 shows that investment in fixed assets at companies’ own expense (that is,
financed by retained earnings and accrued depreciation) depends on marginal revenue on
18
fixed capital (as a proxy for investment profitability), on prices on investment goods and on
net corporate income first of all. It depends also on government investment, on total national
savings,52 on government purchases, on liquidity indicators and on import and on
transportation tariffs.
Here and in other equations of the Investment Unit we see again an important role of
liquidity indicators. Dramatic drop of the rates of private investment after 2008 are well
explained by severe liquidity constraints. Other explanations of such trend include a severe
reduction of annual rates of:
a)
net marginal revenue on fixed capital;
b)
net corporate incomes (in 2014-2015);
c)
national savings;
d)
government investment;
e)
negative investment climate since 2013 as shown by dummies.
Government investment has a strong positive impact on private investment as the latter
very often seeks for support from the former in Russia. At the same time the government
purchases (its expenditures on current needs) have a strong negative impact due to “crowding
out” effect.53 The volume import has predominately negative effect on private investment due
to import substitution.
Equation 21 shows that investment in fixed assets at the expense of a consolidated
state budget depends on the tax payment and of government expenditures first of all. The sum
of indirect taxes and corporate income tax serves as a proxy for government revenues in this
equation (they constitute together 66 % of total revenue and 80 % of tax revenue of
consolidated budget of Russian Federation). Similar to private one the government investment
depends on liquidity and on import.
19
That’s why a drastic drop of government investment in fixed capital since 2008 can be
explained by:
a)
reduction of tax revenues;
b)
reduction of government expenditures;
c)
liquidity shortage;
d)
general tendency to minimize government investment since 2012 that is
illustrated by constant and by dummies.
Equation 22 determines the investment in fixed capital by means of bank loans. They
depend on total volume bank loans to companies and on net marginal revenue on fixed capital
first of all. They depend also on investment from internal sources and on government
investment as the majority of large Russian banks are state-owned and they tend to finance the
fulfillment of government orders.54 The “crowding out” effect of current government
purchases is present here too; and like the other types of investment in fixed capital the
transportation tariffs have a significant negative effect on them.
The investments in fixed capital through bank loans also demonstrate a severe drop of
annual rates since 2008. The equation 22 and Appendix 4 explain that this trend is due to the
drop of:
a) net marginal revenue on fixed capital;
b) private investment from own companies’ resources;
c) government investment;
d) annual rates of banking system in general
The increase of prices of investment goods and general tendency to reduce the volume
of investment-type loans (which is illustrated by dummies in this equation) serve as additional
explanation. The ratio of this type of investment to total volume of bank loans to companies
20
reduced from 13 % in 2000 – 2008 to 8 % in 2014 – 2015 that illustrates worsening of
investment climate and unwillingness of banks to give investment-type loans.
Equation 23 calculates the gross fixed capital formation deflator that depends on the
GDP deflator and on import prices first of all. The latter is obvious as Russia depends
significantly on the import of equipment. A strong impact of prices on government purchases
in this equation is an evidence of an important role of state intervention in Russian economy.
Identity 24 calculates the gross fixed capital formation index in constant prices.
Equations and identities 25-36 represent a Bank Unit of the model.55
Equation 25 determines that households’ denominated bank deposits depend on money
mass and on the level of GDP, and negatively on the share of households’ consumption in the
latter.56 The interest rates have positive impact and the exchange rate – a negative one on the
volume of these deposits.
Appendix 4 shows a slowdown of annual rates of this type of deposits as a result of a
slowdown of economic growth and tightening of monetary policy.
Equation 26 explains that corporate denominated bank deposits depend also on money
mass and on the level of GDP, and negatively on the exchange rate. The explanation of the
slowdown of their rates is the same as for households’ bank deposits.
Equation 27 shows that of households’ currency deposits depend positively on average
gross wages and on transfers paid and on dollar exchange rate as they try to protect their
savings when ruble weakens. They depend negatively on the money mass (that is obvious)
and on the GDP: that is households preserve their savings nominated in foreign currencies in
bad times but reduce them in the time of prosperity. A noticeable increase of annual rates of
this type of deposits in 2014-2015 supports such conclusion.
Equation 28 shows that corporate currency deposits grow together with their
denominated deposits and with the volume of the GDP and that their value increases when
21
ruble weakens. In both types of currency deposits we see a strong impact of government
purchases.
Identities 29 - 31 calculate the total volume of denominated and currency deposits, and
overall bank deposits, respectively.
Equation 32 shows that the amount of denominated bank loans to companies depends
positively on the amount of deposits and on the GDP. They depend negatively on dollar rate
as when ruble weakens companies need more loans nominated in foreign currency. At the
same time bank loans to business are supported by the credits of the Bank of Russia and by
government purchases.57
Bank loans to companies nominated in foreign currency (equation 33) also depend on
the volume of deposits and on dollar rate. Ruble weakening strengthens the demand for such
credits. In this equation we see the positive effect of the Bank of Russia support and of
interest rates increase also. Both types of bank loans to business are sufficiently affected by
the government purchases.
Identity 34 summarizes the value of denominated and currency bank loans to
companies. Appendix 4 shows a noticeable slowdown of annual rates of bank loans to
business. It can be explained not so much by the slowdown of the offer of bank deposits (the
ratio of loans to deposits declined) as the reduction of demand to bank loans due to overall
slowdown of economic growth and of government purchases in particular.
Equation 35 shows that the volume of bank loans to households depends positively on
the volume of deposits and on real average wages. The influence of dollar rate on them is
negative as credit balances depreciate in the situation of ruble weakening. The slowdown of
the rates of consumer loans can be explained by the critical level of households’ indebtedness.
Identity 36 determines the level of Bank of Russia support of commercial banks as a
ratio to the volume of bank loans to companies and to households.
22
Equations and identities 37 – 47 describe the Foreign Economy Unit of the model.
Equation 37 shows the negative dependence of the dollar exchange rate on the export
prices increase and its positive dependence on import prices. We see also a noticeable positive
impact of monetary variables and of the GDP volume on the exchange rate.
It is worth mentioning the significant and positive parameters in front of dummies for
H2 2014 and H2 2015 that reflects significant role of macroeconomic instability and panic on
markets in ruble weakening in these periods. That is a severe reduction of export prices is not
a single reason of ruble devaluation since H2 2014.
Identities 38 and 39 determine the ruble indexes of export and import prices,
respectively.
Equation 40 shows that the volume of Russian exports depends on export prices and
on the volume of the GDP and on the government purchases (as the latter are partly exported
also). But at the same time negatively depends on internal investment demand.
It is worth mentioning that the volume of export depends negatively on index of
internal real prices for energy products as the increase of the latter makes internal sales of
them more profitable.
Due to equation 40 the slowdown of the rates of export can be explained by general
slowdown of economic growth.
Identity 41 calculates the volume of export in current prices.
Equation 42 shows that the amount of import depends on GDP but negatively on
import prices. The increase of the value of transportation tariffs reduces the volume of import.
Equation 42 explains a severe reduction of Russian import after 2008 by the slowdown
of economic growth and for 2014-15 in particular by the impact of international sanctions (the
presence of dummies for in this equation reflects it).
23
Identity 43 calculates the volume of import in current prices and identity 44
determines the value of the net export. Due to the rapid growth of import prior to 2014 the
ratio of net export to the GDP reduced from 12.4 % in 2000 - 2008 to 7.6 % in 2009 – 2013;
but since 2014 such tendency almost stopped due to a dramatic reduction of import.
Identities 45 and 46 calculate the volume of export and import duties as a share of the
value of export and import, respectively. It is worth mentioning that the ratio of the value of
export duties to the value of export exceeded 21 % in 2009 – 2013 but then reduced to 17.5 %
in 2014 – 2015 due to decline both the export prices and the rates of duties. In 2013 – 2014
export duties provided 20 % of the state revenues in Russia (and the lion’s share of them there
were oil and gas export duties). But in 2015 their share reduced to only 12 % of state
revenues.
The role of import duties was much more modest. They constituted 5-6 % of the value
of import (4 % in 2014 – 2015) and only 3 % of state revenues (2.5 % in 2015).
Equation 47 shows that the change of international reserves depends on the exchange
rate, on the volume of the net export and of capital account.
Equations and identities 48 – 54 belong to the Fiscal Unit of the model.
Identity 48 calculates the volume of indirect taxes other than export and import duties
as a share of the GDP. This share was rather stable all the period 2000 – 2015 on the level 1112 % as indirect taxes play a very important role in Russian government finance. Their total
value is determined by identity 49 and share in the GDP (together with export and import
duties) was remarkably stable during all this period on the level 17-18 %, in 2013 – 2014 the
provided about 64 % of all revenues of Russian state (59 % in 2015).
Identity 50 determines the volume of the corporate income tax collected. Though the
official corporate income tax rate in Russia is equal to 20 % the internal revenue service
collects now only 9.5 % of corporate’ income (the reason is tax evasion and privileges). This
24
figure was equal to 15 % in 2000 – 2008 and 13 % in 2009 - 2013. In the forecasts we used
the figure equal to 2014 – 2015 level.
Identity 51 sums the volumes of corporate income tax and indirect taxes. The ratio of
this value to the value of the GDP was stable in 2000 – 2013 on the level 22 % (in reduced to
20 % in 2014 – 2015). It provided about 74 % of all state revenues (70 % in 2015).
Identity 52 calculates the value of government purchases in current prices (two factors
in the right side of the identity are exogenous variables in the model). The share of this value
in the value of the GDP is rather stable during all the period 2000 – 2015 on the level 18 %
(19 % since 2009). But this stability was provided mostly by the rapid increase of prices on
government purchases (their physical volume increased rather slowly).
Identity 53 determines the volume of government social transfers. This ratio of this
variable to the GDP demonstrates a noticeable growth from 8.5 % in 2000 – 2008 to 11 % in
2009 – 2013 and 12 % in 2014 – 2015. In 2000 – 2013 the annual rates of growth of
government social transfers in real terms (the CPI deflator) was 11 %; since Q1 1999 their
real value increased by 5 times. But since 2014 their real growth stopped due to slowdown of
economic growth and budget shortage.
Identity 54 determines a proxy of consolidated budget surplus (shortage).58 In 2000 –
2006 the ratio of this value to the GDP was equal to 7 – 10 %, but since then it decreased
steadily to almost zero in 2015.
Equation 55 belongs to Monetary Unit of the model; it shows that the volume of
money mass59 depends on the volume of monetary base and reserve requirement (both
exogenous variables). Capital account, lagged values of interest rate and fiscal variables also
have an impact on the money mass. A noticeable decline of its annual rates can be well
explained by slowdown of monetary base, that is, by tightening of monetary policy.
25
Equation 56 shows that index of energy prices in real terms (deflated by the GDP)
depends on export and on import prices and on transportation tariffs. The equation show that
the reduction of the annual rates of energy prices prior to 2014 can be explained by a
slowdown of overall inflation in Russia but since 2014 by government intervention also what
is illustrated by dummy.
Identity 57 calculates a specially constructed index of aggregate demand. Its nominator
is a product of indexes of households’ consumption, gross fixed capital formation and
government purchases in constant prices weighted by their shares in the GDP. The
denominator is index of import weighted by its share of the GDP.
The slowdown of aggregate demand index’ annual rates in 2009 – 2013 in comparison
with the period of 2000 – 2008 can be well explained by the general slowdown of Russian
economy which is illustrated by the slowdown of the rates of all four indexes in its nominator.
But in 2014 – 2015 the rates on aggregate demand didn’t reduce due to a severe reduction of
import (and by a modest increase of annual rates of export too).
The weights that are necessary to calculate the aggregate demand index are calculated
by identities 58 – 62.
Identity 63 determines the volume of inventory change that plays a certain role in the
model as affects the GDP deflator in equation 4 and thus helps to move aggregate supply and
aggregate demand towards equilibrium.
4. The Ex-post Forecast Results
We carried out the ex-post “historical” simulation for the period Q1 2009 – Q4 2015 first of
all and then ex-post-forecast for the period Q1 2012 – Q4 2015 on the equations re-estimated
on the diminished sample Q1 1999 – Q4 2011. Theil coefficients for all of the endogenous
variables in both procedures do not exceed 0.3; for 70 % of them they are less than 0.2.60 The
26
model predicts correctly all the turning points of endogenous variables. Therefore, we can
assume that the model shows good properties and can be used for the analysis and forecasting.
5. The Forecasts
The estimated model allows us to make analysis and forecasts for the Russian economy. At
first we calculated impulse multipliers of the exogenous variables and did it in following
manner. At first we calculated what we called “zero forecast”; that is, forecast variant for the
period 2016 – 2018 where all exogenous variables are constant. Than we changed one
exogenous variable by 1 % and the model calculated the percent of change of all endogenous
variables.61 We included the changes in taxes, duties and tariffs in this table which are
formally endogenous variables in the model as at the same time they are instruments of state
policy in reality.62 The results are presented in Appendix 5 (only the reactions of the most
important endogenous variables are included in this model).
Figures in table 5.1 help to make following conclusions.
1.
Russian economy still depends significantly on the demographic factors. This
once again underlines the extreme urgency of the change of economic course.
2.
Export prices have rather modest impact on Russian economy. Only the
investments in fixed capital react more or less noticeably on their increase. That means that
hopes for future economic growth due to the next increase of export prices become more and
more illusory. Nevertheless they have some effect on inflation and on dollar exchange rate.
3.
Import prices have a significant negative impact on Russian economy mostly
on profits and investments. Their increase leads to a noticeable ruble weakening.
4.
Corporate income tax has rather weak impact on Russian economy though that
of indirect taxes (other than export and import duties; the latter have influence mostly on
27
profits) is very strong. That means that some opportunity for economic recovery by means of
fiscal policy lies in the field of loosening of indirect taxation.
5.
Government purchases have rather strong influence on Russian economy
mostly on investment in fixed capital (negative) and on export and import and on inflation
(positive). This can be explained by: a) “crowding out” effect; b) important role of
government in export – import operations; c) large role of state in economy in general.
6.
Monetary policy has rather strong positive effect on investment in fixed capital
but it impact on inflation is weak. That means that the Bank of Russia should refuse from its
restrictive policy.
7.
Russian inflation depends mostly on government purchases and not on
monetary policy. At the same time consumer prices reflect rather weakly on the change of the
majority of variables (with the exception of import prices) due to significant government
intervention.
8.
Russian economy demonstrates a significant dependence on import.
9.
The impact of social transfers and of capital account on the economy is weak
(that’s why we didn’t include them in the table).
Then the model was used to run several forecast variants. The basic forecast variant
was based on the following assumptions.
I.
The dynamics of export and import prices in 2016 – 2018 will be the same as
in 2009 – 2013. That is the abnormal circumstances of 2014 – 2015 for Russian export and
import will not continue.
II.
The dynamics of the number of economically active population, of monetary
base, of the volume and prices for government purchases and of transportation tariffs in 2016
– 2018 will the same as it was in 2014 – 2015. That is that demographic situation and the
monetary and fiscal policy will not change.
28
III.
The level of taxation will be the same as in 2014 – 2015.
IV.
The ratio of social transfers to the GDP will be on the level of 2014 – 2015.
V.
The instruments of monetary policy (key rate and reserves’ norm) will be
constant on Q4 2015 level.
The abnormal conditions of 2014 – 2015 will be abolished. In practice that
VI.
means that we exclude dummies for 2014 – 2015 from the forecasts for 2016 – 2018.
The results of the basic variant are presented in Appendix 5, Table 5.2. The series of
variables in this table are smoothed by Hodrick – Prescott filter.
The basic variant demonstrates following results.
1.
The economic recovery in Russia starts only in 2018.
2.
Inflation declines gradually and ruble strengthens.
3.
The physical volume of export increases during the entire forecast horizon but
the volume of import increases rapidly already since 2017.
4.
The investments in fixed capital reduce significantly during entire horizon.
That is, the basis for the recovery that starts in 2018 is only the growth in aggregate demand
and not the investment in fixed capital. That means that such recovery may be short-lived in
the environment of falling investments and growing import. On the opposite side the increase
of net marginal revenue on fixed capital may restore the investments after 2018.
Note that results of the basic variant are obtained under assumptions of: a) moderately
favorable international conditions; b) restrictive monetary policy; c) moderate tax burden.
Other forecast variants differ from the basic one by change in dynamics of some
exogenous variable or variable that represents monetary or fiscal policy. Their results are
represented in Appendix 5, Table 5.3 and are characterized by following features:
I.
Increase of annual rates of export prices (PEXPD);
II.
Increase of annual rates of import prices (PIMD);
29
III.
Decrease of corporate income tax as a share of corporate income (PTAX);
IV.
Decrease of indirect taxes (other than export and import duties) as a share of
the GDP (OINTAX);
V.
VI.
Decrease of export duties as a share of export (EXPODUT);
Decrease of import duties as a share of import (IMPDUT);
VII.
Increase of annual rates of government purchases index (IND_G);
VIII.
Increase of annual prices of government purchases (PG);
IX.
Increase of government transfers as a share of the GDP (TRANSFER);
X.
XI.
Reduction of the key rate of the Bank of Russia;
Reduction of reserve requirement;
XII.
Increase of annual rates of monetary base (MB);
XIII.
Increase of annual rates of transportation tariffs (TARIF).
Results that are presented in Table 5.3 help to make following conclusions.
1.
A rapid increase of export prices has rather weak impact on Russian economy
(with some exception to the investments in fixed capital). That is, illusions that return to
happy days of 2000 – 2008 with their rapid economic growth due to high export prices are
groundless. The export prices can only strengthen the ruble.
2.
The dependence of Russian economy very strong and import substitution may
have a positive effect on economic growth.
3.
Investment in fixed capital is much more sensitive to changes in economic
policy or external variables (export and import prices and demography) than other variables.
Such conclusion is consistent with generally accepted economic theory.
4.
The opportunities of fiscal policy are also limited though the loosening of the
tax burden will increase corporate incomes and give some hope for future growth.
30
5.
The increase of government purchases has a dramatically negative impact on
investment in fixed capital due to a “crowding out” effect. On the positive side they help to
increase export. Table 5.3 demonstrates a generally strong influence of government activity
on different aspects of Russian economy – on prices, incomes, investments, export and
income.
6.
Loosening of monetary policy has a significant positive influence on
investments in fixed capital with a relatively moderate increase of inflation. That means that
restrictive policy should be abolished steadily.
7.
Inflation depends mostly on government purchases and not on monetary policy.
6. Conclusions
The model and forecasts show that recovery in Russian economy will start not earlier than in
2018. The main reasons for such result are low total factor productivity and low investment
rate. The ability of Russian banking system to support the economic growth is limited due to
its low share in the economy and mostly short-term structure.
Russian economy is sufficiently dependent on the world economy; its dependence on
the ability of labor force is still very high.
The termination of restrictive monetary policy is more and more necessary to enhance
the investment and support economic growth. The losses due to higher inflation will be not so
strong. A reduction of the tax burdens especially what about the indirect taxes other than
export and import duties will help to improve profits and give some hope to economic
recovery too.
Workforce reduction which began a few years ago puts the new challenges in the
fields of improvement of education, healthcare, scientific and technological advances. The
main purpose of all these necessary changes is an overall increase of economic efficiency.
31
Failure to meet these challenges will leave Russia dependent on the world commodity
markets; will retain an inherent weakness of her economy and postpone a significant
improvement of living standards of her citizens.
References
Aigner Dennis J. and C. A. Knox Lovell, and Peter Schmidt, “Formulation and estimation of
stochastic frontier production function models”, Journal of Econometrics 6 (1977): 21-37
Aivazian, Sergey and Boris Brodsky, “Macroeconometric Modeling: Approaches, Problems,
an Example of an Econometric Model of the Russian Economy,” Applied Econometrics 2
(2006):85-111 (in Russian)
Aivazian, Sergey, Boris Brodksy, Edward Sandoyan, Mariam Voskanyan and David
Manukyan, “Macroeconometric Modeling of Russian and Armenian Economies. I.
Peculiarities of Macroeconomic Situation and Theoretical Description of Dynamic Models,”
Applied Econometrics 30 (2013a):3-25 (in Russian)
Aivazian, Sergey, Boris Brodksy, Edward Sandoyan, Mariam Voskanyan and David
Manukyan, “Macroeconometric Modeling of Russian and Armenian Economies. II.
Aggregated Macroeconometric Models of the National Economies of Russia and Armenia,”
Applied Econometrics 31 (2013b):7-31 (in Russian)
Alcala, Francisco and Antonio Ciccone, “Trade and Productivity,” Quarterly Journal of
Economics,”
May
(2004):
613-646
http://ciccone.vwl.uni-
mannheim.de/fileadmin/user_upload/ciccone/images/tradeproductivity.pdf
Alesina, Alberto and Romain Wacziarg, “Openness, country size and government,” Journal of
Public Economics 69 (1998): 305-321.
32
Alesina, Alberto et al., “Trade, growth and the size of the countries” in: Philippe Aghion and
Steven Durlauf (eds.), Handbook of Economic Growth, Vol. 2, Amsterdam: North-Holland
2005, ch.23.
Alexeev, Alexander, Natalia Tourdyeva and Ksenia Yudaeva, “Estimation of the Russia’s
trade policy options with the help of the Computable General Equilibrium Model,” CEFIR
Working Papers 42 (2003)
Ando, Albert, Franco Modigliani and Robert H. Rasche, “Equations and Definitions of
Variables for FRB – MIT – Penn Econometric Model, November, 1969,” in: Bert G. Hickman
(ed.), Econometric Models of Cyclical Behavior, National Bureau of Economic Research,
New York: Columbia University Press, 1972, pp. 543-98.
Arias, Omar S. et al., “Back to Work: Growing with Jobs in Europe and Central Asia,”
Washington, DC: World Bank (2014).
Bank of Russia, “Official Internet Site,” http://www.cbr.ru, 2016.
Barro Robert, “Economic growth in a cross section of countries,” Quarterly Journal of
Economics 106 (2), (1991): pp.407-443.
Barro Robert and J.-W. Lee, “Sources of economic growth (with commentary),” CarnegieRochester Conference Series on Public Policy 40, (1994), pp. 1-57.
Barro Robert, “Democracy and growth,” Journal of Economic Growth 1 (1), (1996): pp.1-27.
Barro, Robert, Determinants of economic growth, MIT Press, Cambridge, 1997.
Basdevant, Olivier, “An Econometric Model of the Russian Federation,” Economic Modelling
17 (2000):305-36.
33
Beck, Thorsten and Ross Levine, “Stock Market, Banks and Growth: Panel Evidence,”
Journal of Banking and Finance 28 (2004):423-42.
Benedictow Andreas, Daniel Fjaertoft and Ole Lofsnaes, “Oil dependency of the Russian
economy: an econometric analysis,” Economic Modelling 32 (2013): 400-428.
Benhabib J. and Spiegel M. M., “Human capital and technology diffusion” in: Philippe
Aghion and Steven Durlauf (eds.), Handbook of Economic Growth, Vol. 1, Amsterdam:
North-Holland 2005, ch.13.
Blanchard, Olivier, The economics of post-communist transition, Clarendon Press, Oxford,
1997.
Data Resources, Inc., The Data Resources National Economic Information System,
Amsterdam: North Holland, 1976.
De Leeuw, Frank and Edward Gramlich, “The Federal Reserve – MIT Econometric Model,”
The Federal Reserve Bulletin 54 (1968):11-40.
Duesenberry, James Stemble, Gary Fromm, Laurence R. Klein and Edwin Kuh, eds The
Brookings Quarterly Econometric Model of the United States, Chicago: Rand McNally &
Company, 1965.
Duesenberry, James Stemble, Gary Fromm, Laurence R. Klein and Edwin Kuh, The
Brookings Model: Some Further Results, Chicago: Rand McNally & Company, 1969.
Ekstein, Otto, The DRI Model of the US Economy, New York: McGraw Hill Book Company,
1983.
Evans, Michael K. and Lawrence R. Klein, The Wharton Econometric Forecasting Model,
Philadelphia: Economics Research Unit Wharton School, University of Pennsylvania, 1967.
34
Evans, Michael K. and Lawrence R. Klein, The Wharton Econometric Forecasting Model, 2nd
enlarged ed., Philadelphia: Economics Research Unit Wharton School, University of
Pennsylvania, 1968.
Fair, Ray C., Specification, estimation and analysis of macroeconomic models, Harvard
University Press, 1984
Fair, Ray C., Testing macroeconometric models, Harvard University Press Cambridge,
Massachusetts, London, England, 1994.
Fair, Ray C., “Evaluating the Predictive Accuracy of Models,” in: Zvi Griliches and Michael
D. Intriligator (eds.), Handbook of Econometrics, Vol. 3, Amsterdam: North Holland, 2005,
pp. 1979-1996.
Findlay, Ronald, “Growth and Development in Trade Models,” in: Ronald W. Jones and Peter
B. Kene (eds), Handbook of International Economics, Vol. 1, Amsterdam: North Holland,
2007, pp.185-236.
Frankel
Jeffrey
A.
and
David
Romer,
“Does
trade
cause
growth?”
(1999)
http://eml.berkeley.edu/~dromer/papers/AER_June99.pdf
Fromm, Gary, Tax Incentive and Capital Spending, Amsterdam: North Holland, 1971.
Fromm, Gary and Lawrence R. Klein, The Brookings Model: Perspective and Recent
Developments, Amsterdam: North Holland, 1975.
Green, William H, Econometric Analysis, 6th ed., Upper Saddle River, NJ: Prentice Hall,
2008.
35
Gregory Paul R. and V. Lazarev, “Structural change in Russian transition,” Economic Growth
Center,
Yale
University,
Center
Discussion
Paper
N.896,
October
(2004)
http://www.econ.yale.edu/growth_pdf/cdp896.pdf
Institute of Economic Forecasting, “Laboratory Medium-Term Forecasting Processes of
Reproduction; Laboratory Prediction of Productive Capacity and inter-Industry Interactions;
Laboratory Analysis and Forecasting of the Fiscal System,” http://www.macroforecast.ru,
2015.
Intriligator, Michael D., Ronald G. Bodkin and Cheng Hsiao, Econometric Models,
Techniques, and Applications, Upper Saddle River, NJ: Prentice Hall, 1996.
Intriligator, Michael D., “Economic and Econometric Models,” in: Zvi Griliches and Michael
D. Intriligator (eds.), Handbook of Econometrics, Vol. 1, Amsterdam: North Holland, 2007,
pp. 181-222.
Jensen, Jesper, Thomas Rutherford and David Tarr, “The Impact of Liberalizing Barriers to
Foreign Direct Investment in Services: The Case of Russian Accession to the World Trade
Organization,” World Bank Working Paper WPS3391 (2004)
Johnston, Jack, and John DiNardo, Econometric Methods, 4th ed., New York: McGraw Hill,
2007.
Kerkela, Leena, “Distortion costs and effects of price liberalization in Russian energy
markets: A CGE analysis,” BOFIT Discussion Papers, 2 (2004).
Klein, Lawrence R., Economic Fluctuations in the United States, 1921-1941, Cowles
Commission Monograph 11, New York: John Wiley & Sons, Inc., 1950.
Klein, Lawrence R., Lectures in Econometrics, Amsterdam: North Holland, 1983.
36
Klein, Lawrence R. and Arthur Stanley Goldberger, An Econometric Model of the United
States, 1929-1952, Amsterdam: North-Holland, 1955.
Klenow Peter J. and A. Rodriguez-Clare, “Externalities and Growth,” in: Philippe Aghion and
Steven Durlauf (eds.), Handbook of Economic Growth, Vol. 1, Amsterdam: North-Holland
2005, ch.11.
La Porta, Rafael, Florencio Lopez-de-Silanes and Andrei Shleifer, “Government Ownership
of Banks,” National Bureau of Economic Research Working Paper 7620 (March 2000)
http://www.nber.org/papers/w7620.pdf
Lerner Abba P., “The economics of control. Principles of welfare economics”, NY, McMillan,
1944
Levine Ross e. a., “Financial intermediation and growth: causality and causes,” Journal of
Monetary Economics 46 (1) (2000): pp.31-77
Levine, Ross, “Finance and Growth: Theory and Evidence,” in: Philippe Aghion and Steven
Durlauf (eds.), Handbook of Economic Growth, Vol. 1, Amsterdam: North-Holland, 2005, pp.
865-934.
Love, Inessa, “Financial development and financing constraints - international evidence from
the structural investment model,” Review of Financial Studies, 16 (2003): 765-791.
Merlevede, Bruno, Bas van Aarle and Koen Schoors, “Russia from Bust to Boom and Back:
Oil Price, Dutch Disease and Stabilization Fund,” Comparative Economic Studies 51 (2009),
p.213-241.
37
Mitsek Sergey and Elena Mitsek, “Econometric and statistical estimates of industrial and
regional structural changes in Russian Federation,” Finance and Credit (Finansy iCredit) 1
(577), January, 2014, pp.2-9 (in Russian)
Orphanides, Athanasios and Robert M. Solow, “Money, Inflation and Growth,” in: Benjamin
M. Friedman and F. H. Hahn (eds.), Handbook of Monetary Economics, Vol. 1, Amsterdam:
North Holland, 1990, pp. 223-62.
Pindyck Robert S. and Daniel L. Rubinfeld, “Econometric Models and Economic Forecasts”,
4th ed., McGraw-Hill International Editions, 1998
Pindyck Robert S. and Daniel L. Rubinfeld, “Microeconomics”, 7th ed., Pearson International
Edition, 2009
Rasche, Robert H. and Harold T. Shapiro, “The FRB – MIT Econometric Model: Its Special
Features and Implications for Stabilization Policies,” American Economic Review 58
(1968):123-49.
Rautava, Jouko “The role of oil prices and the real exchange rate in Russia’s economy,”
BOFIT Discussion Papers 3 (2002)
Romer, David, “Advanced Macroeconomics,” 3d ed., McGraw Hill, 2006.
Rosstat, “Federal Service of Statistical Information Internet Site,” http:// www.gks.ru, 2016a.
Rosstat, “Federal Service of Statistical Information Demographic Forecast of Russian
Federation until 2030,” http://www.gks.ru/free_doc/new_site/population/demo/progn3.htm,
2016b.
Rutherford Thomas, David Tarr and Oleksandr Shepotylo, “The Impact on Russia of WTO
Accession and The Doha Agenda: the importance of liberalization of barriers against foreign
38
direct investment in services for growth and poverty reduction,” World Bank, Working Paper
WPS3725 (2005).
Sachs Jeffrey and A. Warner (1995), “Economic reform and the process of global integration
(with discussion),” Brookings Papers on Economic Activity 1 (1995): pp.1-118.
Solow, Robert M., “Reflections on Growth Theory,” in: Philippe Aghion and Steven Durlauf
(eds.), Handbook of Economic Growth, Vol. 1, Amsterdam: North-Holland, 2005, pp. 3-66.
Wacziarg Romain and Karen Horn Welch, “Trade liberalization and growth: new evidence”
(2001) http://164.67.163.139/documents/areas/fac/gem/wacziarg_trade.pdf
World Bank, Financing growth: selection of methods in a changing word, Moscow, “Ves’
Mir” (2002) (in Russian)
World Bank, World development Indicators http://wdi.worldbank.org (2012)
World Bank, “Russian Federation - Export Diversification through Competition and
Innovation:
A
Policy
(2013)
Agenda”
http://documents.worldbank.org/curated/en/2013/04/18023843/russian-federation-exportdiversification-through-competition-innovation-policy-agenda, pdf
World
Bank,
“Russian
Economic
Report
N.
30,”
http://www-
wds.worldbank.org/external/default/WDSContentServer/WDSP/IB/2013/10/08/000356161_2
0131008122642/Rendered/PDF/816880NWP0Russ0Box0379842B00PUBLIC0.pdf, 2014a
World
Bank,
Russian
Economic
Report
N.
http://www.worldbank.org/ru/country/russia/publication/russian-economic-report-32,
2014b
32,”
pdf,
39
Notes
1
See Appendix 1 “Statistical Tables”, Table 1.
2
Crafts e. a. (2013, p.418-419) had shown that the same was evident for the USSR in 1970-80s.
3
See for example McKinsey (2009) and IEA (2014) for energy sector specially.
4
See World Bank (2014, p.47), OECD (2013a, p.268-270), OECD (2013b, p.4, 27)
5
See details in Jorgenson e. a. (2005, 2008), Oliner e. a. (2007), van Ark e. a. (2008, 2013), Bloom e. a., (2010),
Syverson (2011), Bernanke (2005); in our work (Mitsek and Mitsek, 2014) gross regional product in Russian
regions depends significantly on the number of personal computers per 100 employees.
6
Its degree increased from 43 % in 2004 to 49 % in 2014; fixed capital is worn out most of all in such sensitive
for Russian economy spheres as mining and transportation and in such important regions as Moscow City and
oil-rich Khanty-Mansi and Yamalo-Nenets districts.
7
Rosstat
8
IEA (2014)
9
See Mitsek and Mitsek (2015); in Mitsek and Mitsek (2014) we proved e negative influence of increasing share
of government services on economic growth; see also Barro (1991, 1996, 1997) and Sachs and Warner (1995),
Alesina et al. (1998, 2005) with the same result.
10
Regional aspect also should be mentioned: we keep in mind the TFP decline in such important regions as
Moscow City and oil-rich Khanty-Mansi and Yamalo-Nenets districts (see Mitsek and Mitsek, 2015).
11
World Bank (2014, p.42-52) mention other reasons of low TFP in Russia: a) inadequate investment in human
capital; b) poor infrastructure; c) unfriendly business environment; d) distortion of equal competition conditions;
e) low quality of government institutions. More detailed analysis of all these aspects one can find in OECD
(2014, chapters 1-2). Many works are devoted to the first item – the role of human capital (e.g. Sala-I-Martin e.
a. (2004), Ruses e. a. (2013) and science (OECD, 2011) in economic growth.
12
The number of people older than working age increased from 21 % of total population in Q4 1999 to 24 % in
Q4 2015; the share of retired increased from 26 % of total population to 29 % during the same period. Other data
are contradictory. The ratio of economically inactive population in working age to those who are active
decreased from 28 % to 24 % during this period but only due to the reduction of the ratio of the number of
economically inactive who want to work from 10 % to 5 % of those who are active. The ratio of inactive to
active in the age 15 – 72 years reduced from 57 % in Q4 1999 to 44 % in Q4 2015.
About aging as a factor of economic growth one can see Dedry A. e.a. (2014); estimates of its role in Russia in
World Bank (2015, p.61-67). The recent increase of labor force (on about 0.6 % annually in 2014-2015) can be
attributed to inclusion of Crimea labor.
13
Official estimate of foreign workers who have the permission to work is 1 million, that is, about 1.5 % of total
employment. But if we pay attention on other sources this figure is probably 3 million or 4.5 % of total (see
Labor and employment in Russia, tables 5.11 and 5.13). In August 2015 about 11 million of foreign people lived
in Russia that is about 8 % of total population (World Bank 2015, p.39).
When deflated by CPI these rates were 11.39 % annually in 2000 – 2008; see details in Appendix 1. The
increase of real wages often serves as an indicator of coming crisis, see Zarnovitz (1992).
14
15
The real wages in Russia increased by 3.53 times in real terms since 2000 up to 2013 (when deflated by CPI).
At the same time it increased by 3.93 times in government services and by 4.98 times in education and by 4.83
times in health care and by 4.21 in social services. The last three sectors are subject to strict government control;
see also World Bank (2015, p.38) and World Bank (2014, p.53).
40
16
Rapid growth of construction and finance sectors and the increase of relative share of such regions as Moscow
City, Moscow Oblast, St. Petersburg City, Sakhalin and Yamal with their high wages made contribution to
average wages increase.
17
It is important to mention that since the year 2005 the profit margin in such export-oriented industries as oil
extraction, petroleum production and metallurgy decreased drastically.
The households’ consumption expenditures increased in large degree due to dollar exchange rate real
depreciation and real import prices decline as a consequence prior to 2009, see picture. The real import prices
decrease mostly for meat and milk products and for sugar and footwear. Another one important reason was a
rapid development of consumer credit from near zero in the beginning of 2000s. The third reason is low personal
income tax rate (13 % for the majority of citizens) and abundance of price subsidies.
18
19
The world prices dropped not only for crude oil but for natural gas and for nickel and aluminum which are
Russian export goods too.
The real appreciation/depreciation of the currency is often measured as ratio of non-trade goods’ to traded
goods’ price’ indexes. If we use such rough measure of it as relation of the GDP deflator to export price index
this value increased by 65 % since Q1 1999 to Q1 2010; but then it decreased by 6 % up to Q4 2015.
20
The real increase of import was about by 6.5 times in 1999 – 2014. Blachard (2006) on the case of Portugal
showed that real appreciation of national currency leads to the loss of competitiveness and reduction of output as
result. The same case in Argentina, see Hausmann and Velasco (2002).
21
22
http://www.gks.ru/
23
http://www.cbr.ru/
24
http://www.macroforecast.ru/
25
See Evans and Klein (1967, 1968); Rasche and Shapiro (1968), De Leeuw and Gramlich (1968), Ando,
Modigliani and Rasche (1972); Data Resources, Inc. (1976), Ekstein (1983); Duesenberry et al. (1965, 1969),
Fromm (1971), Fromm and Klein (1975); review and references in Intriligator et al. (1996, pp. 430-56),
Intriligator (2007, pp. 204-7).
26
OLS gives consistent estimates for such type of models, see e.g. Green (2008, p. 372), Intriligator et al. (1996,
pp. 336-9, 388-9), Pindyck and Rubinfeld (1998, p.346-48) on the use of fully recursive models.
27
See Johnston and DiNardo, (2007, pp. 316-7) on the widely used practice of the OLS estimation of structural
equations.
When ML – ARCH estimates were used correlogram of residuals with Q-statistics (Box – Pierce – Ljung
statistics) was used to test autocorrelation.
28
29
ARCH test was used for ML-ARCH estimates (1 lag of square residual and regressor in test equation) and
Breusch – Godfrey – Pagan test for OLS estimates (regressores of equation in test equation).
30
The dummies were used for the points where breaks were detected.
31
We used automatic lag length based on Schwartz info criterion; and 5 % MacKinnon p-values.
32
Among the other tests there are: Chow forecast tests for 4, 8 and 12 forecast points; Ramsey 1, 2 and 3
specification tests. Recursive tests were performed to test for parameters’ stability. The residuals were tested for
normality by Jarque – Bera test besides.
33
Variances of residuals were used here as weights for such tests.
34
Presample variance: backcast with parameter 0.7; starting coefficient values OLS / TSLS; Marquardt
optimization algorithm.
35
We used such proxy as official statistics of profitability is not credible and the corresponding variables were
not significant probably due to this reason.
41
36
As the model is fully recursive the equations and identities are numbered in the order in which the model is
solved as a whole.
37
The lags in this equation reflect long construction lags in Russian economy and the specifics of Russian
accounting practice. Here we mean a lag between a fixation of investment and statement of fixed assets in
accounting books. In this connection it is worth mentioning that investment in residential and non-residential
construction account for about 60 % of total investment in fixed assets in Russia.
38
In accordance with classical economic theory at the point of equilibrium of the company the marginal revenue
on production factor should be equal to its price. But Russian economy is not in equilibrium; the labor market is
not an exception as is it is subjected to strong government regulation. That’s why the difference between the net
marginal revenue on labor and average wages may create an additional demand for labor. Abba Lerner gave a
similar argument about capital when he wrote that the main force that increases or reduces the fixed capital
amount is the difference between the marginal revenue on capital and the interest rate (see Lerner, 1944, p.335).
39
One more explanation of relatively low unemployment is certain increase of the ratio of marginal revenue on
labor to average wages since 2010 to 2013 from 1.44 to 1.50 (since 2013 it had been constant).
40
A gross fixed capital formation deflator is included in this equation as a separate variable because a direct
deflating of fixed capital yields unstable estimates. A similar approach we can find in Benedictow et al., eq.7.
41
See, for example, Pindyck and Rubinfeld (2009), p.197.
42
If we just not use so-called “frontier” production functions; see, for example, Aigner e. a. (1977)
43
An illustration of the fact that this gap was significant in Russian economy is the statistics of low-capacity
utilization in it since 1995. This trend can be calculated on the basis of Rosstat (2015a) data, see page
http://www.gks.ru/free_doc/new_site/business/prom/moch.htm; the calculation of capacity utilization is
complicated by the fact that labor input in Russian statistics is calculated as the number of employees, not as the
number of hours worked, and doesn’t reflect the intensity of labor use therefore. If we pay attention on the
specifics of Russian labor market where employer respond on the fall of demand usually not by layoffs but by
reduction of salaries actually paid (including the hidden, illegal one) the formally employed workforce is mostly
not fully utilized.
A discussion of the problem of “export-led growth” and terms of trade as a factor of growth is presented in
Findlay (2007, particularly pp. 215-26). The necessity to consider the role of demand in economic growth was
specifically mentioned by Solow (2005). We say here about the direct impact of demand on production; its
indirect impact may take place via production factors’ input.
The share of import can also explain the part of the gap as the remaining 1/3 of import consists of
consumer goods which can be partly substituted by domestic production. The corresponding variable has
negative sign that means that Russian producers have a chance to increase production (and capacity utilization,
consequently) in the conditions of decline in competition from imports.
45
Here we use the formula given in Jonston and DiNardo (2007), pp.245-247.
44
46
The indicator of unemployment turned to be insignificant in this equation that can be considered as a
manifestation of the specifics of Russian labor market.
47
About 20 % of employees in Russia work in state-owned companies. Detailed discussion on the subject one
can find in World Bank (2013, 2014b).
48
The real interest rate was insignificant in this equation; see a more detailed comment about it below.
Other financial sources of investment in fixed assets such as stocks and bonds, households’ money invested in
residential construction under sharing scheme and non-government investment are rather small and are equal
together to 1.5 % of total investment in fixed capital. That’s why when we calculate the model we simply
multiply the sum of the right side of identity 16 on 1.015. In 2015 this coefficient increased to 1.02.
49
The definition of net corporate income is conditional here as we don’t detract the depreciation from the gross
value due to the absence of data.
50
42
51
If we explore the three parts of indirect taxes their dynamics was contradictory. Import duties as a share of the
value of import increased from 3.9 % in 2000 to 6.8 % in 2009 but then fall to 3.3 % in 2015. Export duties as a
share of the value of export increased steadily from 4 % in 2000 to 22 % in 2011 but then reduced to 16 % in
2015. The share of other indirect taxes in the value of the GDP at first reduced from 13 % in 2000 to 10 % in
2011 but then increased to 11.5 % in 2015.
52
Presence of national savings as a separate variable in this equation may serve as an evidence of presence of
other than retained profit (probably hidden) sources of savings that serve to finance private investment. The
recorded statistics of profits may be incredible besides.
53
The same result was found by Barro (1991, 1996, and 1997) and by Sachs and Warner (1995) in cross country
regressions. Alesina et al. (1998, 2005) showed a presence an “economies of scale” in government consumption
(that is their decrease as a share of economy) as result of economic growth. In Alesina et al. (2005) cross-country
model government consumption is a significantly negative variable for growth.
54
See e.g. La Porta, Lopez-de-Silanes and Shleifer (2000) on this subject.
55
The impact of the banking system on economic growth was studied in many works. We mention here only
those we consider the most important such as Levine (2005), Beck and Levine (2004), Orphanides and Solow
(1990), World Bank (2002).
56
Levine et al. (2000) studied such type of regression on cross-country data.
57
That’s why recent attempts to restrict interest rates may have negative impact on bank loans to business.
58
It is a proxy as a) some taxes and non-tax revenues are not included in the model; b) the social payment was
taken here as 25 % of gross wages which is an approximate ratio; c) expenditures here are equal to the sum of
government purchases and social transfers that is not exactly equal to budget expenditures in reality. But such
variable is useful as regressor in some equations.
We use so-called “national determination” of money mass in the model which includes currency and ruble
deposits and is analog of conventional M2 aggregate. The Bank of Russia also use so-called “wide
determination” of the money mass which includes deposits in foreign currency.
59
60
A discussion devoted different measures of the evaluation of predictive accuracy of models see in Intriligator,
Bodkin and Hsiao (1996, pp. 523-7), in Fair (2005, pp. 1984-93) and in Pindyck and Rubinfeld (1998, ch.13-14).
61
We calculated the percent change of the variables as the model is non-linear and we couldn’t use the
traditional reduced form.
As these ‘pseudo endogenous’ variables are calculated via factor to some other endogenous variable (see
Appendix 2) their 1 % increase means the following increase of this factor.
62