2_Sultonov_IJEPS2013

The International Journal of Economic Policy Studies Volume 8 2013 Article 2

23

Labour Migration and Import Demand: The Impact of Remittance Inflows for The Case Of Tajikistan

Mirzosaid SULTONOV

Research Fellow

Graduate School of Economics, Kobe University 6578501, Japan, Kobe, Nada-ku, Rokkodai-cho 2-1

E-mail: [email protected]

ABSTRACT

Tajikistan is a low-income country with a small and open economy; it is highly dependent on imports and remittance inflows. The volume of imports and remittances that flow into the country has climbed steadily in the last 10 years, and the overall proportion of these imports and remittances in the economy has increased. For remittances as a percentage of GDP, Tajikistan has been among the top remittance-receiving countries for the last seven years. The volume of remittance inflows is not very large compared with the other top remittance-receiving countries. In 2010, the total remittances that flowed into Tajikistan amounted to only $2.2 billion while 42 countries in the world had remittance inflows of more than $2.2 billion. However, relative to the small size of the Tajikistan economy, the amount of remittances is very large: as a percentage of GDP, the remittance inflows were 6.4% in 2002, 9.4% in 2003, 12.1% in 2004, and 20.8% in 2005; in 2006, 2007, and 2008, they were 36.1%, 45.5%, and as high as 49.3%, respectively; in 2009, the remittances decreased slightly to 35.1%, but then increased again to 40.0% in 2010 and 44.2% in 2011.

The huge inflow of remittances can be expected to affect all macro and microeconomic indicators of the country. This paper focuses on the impact of remittance inflows on imports in Tajikistan. Analysing empirical data and applying an econometric model to quarterly time series, the paper demonstrates the significant impacts of remittance inflows on imports. Imports are a macroeconomic variable very responsive to inflow of remittances. Remittance inflows as a source of foreign currency clearly affect imports. An increasing share of imports in the economy, while export share is decreasing, brings about a trade deficit, negatively affects GDP, and increases dependence on remittances and imports.

The relationship between remittances and imports is one of the main sources of disagreement about the net economic impact of remittances. Import of production goods promotes domestic production and import of consumption goods increases living standards. However, increase in imports of goods which can be produced domestically affects domestic production negatively. Furthermore, an increasing marginal propensity to import decreases the multiplier effect of remittances. This paper examines this relationship for the case of Tajikistan, and tries to derive policy implications for enhancement of positive impacts of remittances and decrease of dependency on them.

Key words: Remittances, Imports, Tajikistan. JEL Classification: F16, F22, F24, F41, J61 Acknowledgement: I would like to express my sincere gratitude to my supervisor Professor Yoshii Masahiko (Kobe University) for his continuous support and guidance. I also want to thank Professor Akiko Sakanishi (Nara Prefectural University), Professor Terukazu Suruga (Kobe University) and other participants in the Labor Economics and Policy Session of the 11th International Conference of the Japan Economic Policy Association (JEPA) at Nagoya Gakuin University, for their valuable and constructive suggestions and remarks on improvement of this paper.

Vol.8 2013

24

Labour Migration and Import Demand: The Impact of Remittance Inflows for The Case Of Tajikistan

1. Introduction

In 2011, the World Bank (WB) estimated total global remittance flows to be $483 billion, which

included remittances of $351 billion into developing countries. Remittances to developing

countries increased by 8% in 2011 over the previous year. India, China, Mexico, the Philippines,

Pakistan, and Bangladesh were the top remittance-receiving countries in 2011. Further, the top

recipients of remittances as a share of GDP were Tajikistan, Lesotho, Nepal, Samoa, and Tonga.

Tajikistan is a low-income country in Central Asia with a small and open economy. The

country gained independence from the Soviet Union, in 1991. In 1996, the collapse of the

economic system of the former Soviet Union and the five years of civil war that followed

independence, resulted in the country’s GDP plummeting by 74% as compared to its 1988 level.

However, the economy began recovering from 1997 onwards after a peace agreement was signed

between the official government and the armed opposition. The recovery period continued until

the end of the first decade of the new millennium.

War and poor economic conditions in Tajikistan caused the displacement and migration of

a huge share of its population. Migration to other countries of the Commonwealth of Independent

States (CIS) continued even after the civil war. The inability of the domestic labour market to

provide sufficient jobs led to a huge share of the labour resources migrating to other countries.

Beginning in 1997, the migration of labour to the CIS, mainly to Russia, became one of the main

features of the economic and social dimension of Tajikistan.

Table 1: Value and Scale of Remittances

Value Scale

Year Million

USD World

ranking %

of GDP World

ranking 2002 78.6 104 6.4 33 2003 146.0 94 9.4 23 2004 252.0 87 12.1 22 2005 466.7 84 20.8 8 2006 1018.8 62 36.1 2 2007 1690.8 49 45.5 1 2008 2544.0 43 49.3 1 2009 1748.2 50 35.1 2 2010 2254.5 43 40.0 1

Source: WB (2011)

International Journal of Economic Policy Studies

25

It has been estimated that 30.4% of Tajikistan’s labour resources migrated abroad in 20101.

The increased migration resulted in a large increase in remittance inflows. The officially recorded

information on remittance flows into Tajikistan became available only from 2002 onwards; it

showed remittances amounting to only 6.4% of GDP that year, and increasing sharply to 44.2%

of GDP in 2011. Among the top remittance-receiving countries, Tajikistan’s ranking rose from

104th in 2002 to 43rd in 2010; moreover, among the top remittance-receiving countries as a share

of GDP, its ranking rose from 33rd in 2002 to 1st in 2010 (Table 1).

The huge inflow of remittances is anticipated to affect all macro and microeconomic

indicators of the country. This paper focuses on the impact of remittance inflows on imports in

Tajikistan.

As remittances are private money inflows that increase private consumption and domestic

demand, the increase in remittances could cause an expansion in imports. In order to estimate the

impact, the structure of imports is analysed, and a comparative assessment of related

macroeconomic variables and multiple regression analysis are conducted.

Imports are a macroeconomic variable very responsive to inflow of remittances.

Remittance inflows as a source of foreign currency clearly affect imports. The relationship

between remittances and imports is one of the main sources of disagreement about the net

economic impact of remittances.

Import of production goods promotes domestic production, and import of consumption

goods increases living standards. However, an increasing share of imports in the economy, while

the export share is decreasing, thereby bringing about a trade deficit, negatively affects GDP and

increases dependence on remittances and imports. An increase in imports of goods which can be

produced domestically affects domestic production negatively. In the long term, the trade deficit

as an indication of a deteriorating domestic economy may cause a decline in living standards.

Furthermore, an increasing marginal propensity to import decreases the multiplier effect of

remittances. This paper examines this relationship for the case of Tajikistan, and tries to derive

policy implications for enhancement of positive impacts of imports and remittances and decrease

of dependency on them.

2. Migration and Remittances

Data from different sources on the number of migrants from Tajikistan yield different results.

Monitoring of the number of migrants is difficult because of the possibility of free movement of

labour from Tajikistan to most of the former Soviet Union countries.

1 UNDESA (2011).

Vol.8 2013

26

Figure-1: Cumulative Net Migration from Tajikistan (thousand persons)

Source: TAJSTAT (2011) and UNDESA (2011)

According to the Statistical Agency under the President of the Republic of Tajikistan

(TAJSTAT), the annual net migration from Tajikistan to other countries was approximately 6,500

to 14,500 persons during the 2000s, with the cumulative net migration since 1991 totalling

367,159 persons in 2000 and 478,628 persons in 2010. According to the Population Division of

the United Nations’ Department of Economic and Social Affairs (UNDESA), the cumulative net

migration since 1991 was 629,413 persons in 2000 and 1347,211 persons in 2010 (Figure 1).

However, an analysis of TAJSTAT data on the labour force of Tajikistan, and survey data on

migration and remittances show that the UNDESA data are more accurate. The lower number of

migrants reported by the national statistics of Tajikistan was probably motivated by the country’s

intention to hide the vulnerability of its labour market, and the dependence of its economy on

migration and remittances.

The TAJSTAT data show that labour force participation in the domestic labour market

decreased from 78% in 1991 to 56.3% in 2000 and to 51% at the beginning of 2010. The

percentage of labour resources that did not participate in the local labour market, excluding

students, increased from 12.6% in 1991 to 32.9% in 2000 and to 36.9% at the beginning of 2010.

The percentage of labour resources that did not participate in the domestic labour market at the

beginning of 2010 reported by TAJSTAT (36.9%) is very close to the cumulative net migration

reported by UNDESA. On the other hand, during the period 2000–2010, 65%–65.9% of the

employed labour force was engaged in agriculture, where the wages were on average $4.2–$26.6,

equal to only 19.6% of the average wages paid to workers employed by the industrial sector.

Hence, those engaged in agriculture probably found it difficult to maintain their families without

joining the seasonal labour migration.

1347.211

421.839367.159279.2

478.628 296.191

1051.136

629.413

0 200 400 600 800

1000

1200

1400

1600

1990-1995 1990-2000 1990-2005 1990-2010

TAJSTAT UNDESA


27

Figure-2: Migrants’ Destination and Gender

Source: TLSS (2007), WB (2007)

The Tajikistan Living Standards Measurement Survey (TLSS 2007), a household survey

conducted in Tajikistan in 2007 by WB, was one of the most comprehensive surveys, and included

information on migration and remittances. According to the survey, more than 94.9% of migrants

from Tajikistan chose Russia as the destination country. 93.4% of migrants are male and 6.6%

female (Figure 2).

The average age of migrants was 28 years, and only 10% had university degrees; 75% of

the migrants were unemployed before migration. On average, a migrant earns $322 per month

and remits $2,123 per year to his or her family – that is, 54.9% of his or her annual income2.

In 2001, the Tajikistan government abolished the 30% tax on remittances and increased the

amount of remittances that could be sent via official channels of transfer. Official statistics on

remittances to Tajikistan are being reported since 2002. The two most frequently referenced

sources of information on remittances to Tajikistan are WB’s annual data and the quarterly data

of the Central Bank of the Russian Federation (CBRF). The data of both sources are very close to

each other. Tajikistan’s official statistics do not present appropriate data on remittances.

The period 2002–2010 witnessed an increase in GDP and remittance inflows. The country’s

GDP witnessed an average annual growth rate of 8.2% and its remittance inflows an average

annual growth rate of 51.1%. A comparison of the remittances with other sources of international

currency shows that remittances as a percentage of GDP had an increasing trend during the 2002–

2010 period, while the percentage of other variables like exports, foreign debt, official

development assistance (ODA), and foreign direct investment (FDI) showed a decreasing trend.

According to the annual data of WB, Tajikistan’s exports decreased from 66.4% of GDP in 2002

2 In this paper TLSS 2007 data are only for the migrants, who were abroad during conduct of the survey; the migrants who already came back to Tajikistan are not included.

Destination Gender

Russia, 94.9% Male, 93.4%

Other countries, 5.1% Female, 6.6%

Vol.8 2013

28

to 28% of GDP in 2010, and the net ODA and FDI decreased from 13.8% and 3% in 2002 to

7.6% and 0.3% of GDP in 2010, respectively. Further, the country’s cumulative public foreign

debt decreased from 73.8% of GDP in 2002 to 32.5% of GDP in 2010.

For the period 2002–2010, the country’s industrial output decreased from 35.1% to 22.8%

of GDP and its agricultural output decreased from 22.2% to 18.7% of GDP.

For the entire period of increase in remittances, there was an increase in internal trade and

negative foreign trade balance: internal trade increased from 11.7% of GDP in 2002 to 19.2% of

GDP in 2010, and the positive foreign trade balance of 0.7% of GDP in 2002 changed to a negative

foreign trade balance of 21.2% of GDP in 2010.

3. Remittances and Imports

Remittances are private money transfers, and support consumption and savings of remittance-

receiving households. The overall impacts of remittances on the economy depend on the

behaviour and preferences of the remittance-receiving households.

According to TAJSTAT the real cash income and expenditure of the population increased

by 3.1 times for 2002–2007, that is the period of a sharp increase in remittance inflows. In 2002,

labour income was 43.2%, income from the sale of agricultural products 42.4%, and other income

including remittances only 14.4% of the total income of population. By 2007, the share of labour

income reduced to 42.3%, and that of income from the sale of agricultural products to 22% of the

total income. On the contrary, the share of other incomes, mainly comprising remittances from

abroad, increased to 35.7% of total income. In 2002, 82.2% of the total income of the population

was spent on the purchase of goods and payment for services, 16.1% was paid as taxes and

gatherings, and 1.7% saved. The increase in income changed the structure of expenditure, with

the share of expenditure on goods and services decreasing to 67.5% of the total expenditure in

2007, the share of taxes and gatherings related payments increasing to 25%, and the share of

savings increasing to 7.5% (Figure 3).

The above analysis demonstrates an increase in the real value of the population’s

expenditure and savings for the period of increase in international remittances. Furthermore, the

data shows an increase in the burden of compulsory payments, taxes and gatherings. Taking into

account the fact that not all households receive remittances from abroad, it is difficult to consider

that the changes in expenditure pattern were caused by remittance inflows from abroad. In order

to assess the impact of remittances on households’ expenditure and savings properly, a two-

sample t-test is conducted for households with remittances from abroad (treatment group) and

households without remittances from abroad (control group).


29

Figure-3: Expenditure Shares of Population

Source: TAJSTAT (2011)

In 2007, average monthly consumption expenditures of households with remittances are

higher than households without remittances. The difference is significant at the 0.01 critical alpha

level. The investment expenditures of households without remittances are higher, but the

difference is insignificant. It seems that remittances significantly increased the consumption of

food and non-food goods (Table 2).

Table 2: Consumption and Investment Expenditures of Households (2007, in somoni)

Expenditures Households

with remittances

Households without

remittancesDifference

Two-sample t-test

Consumption 924.5266 850.0400 74.4866 4.354*** food 695.4544 647.3081 48.1463 4.273*** non-food 229.0722 202.7319 26.3403 2.464***

Investment 167.3764 150.1706 17.2058 0.965 education 50.6771 62.2694 -11.5923 -0.942 health 63.6368 32.3270 31.3098 2.580*** agriculture 22.4030 19.8505 2.5525 1.639* others 30.6595 35.7237 -5.0642 -1.389*

Source: TLSS (2007)

Note: Average monthly data; 3.44 Somoni = 1 USD; *** significant at 1%; ** significant at 5%; *

significant at 10%.

Since the economy of Tajikistan, characterized by poor business conditions and migration

of the most active proportion of the labour force, was not able to meet the increasing domestic

82.2 80.3 81.7 79.4 73.4 67.5 67.7 68.0 62.3

16.1 15.9 16.1 18.519.4 25.0 28.4 28.6

29.8

7.2 7.5 7.93.43.92.12.23.81.7

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2002 2003 2004 2005 2006 2007 2008 2009 2010

Increase in deposits and securities Compulsory payments, taxes and gatherings Purchase of goods and payment for services

Vol.8 2013

30

demand, the huge inflow of international currency (remittances), increase in domestic demand,

and increase in domestic prices led to a large increase in imports. Imports as a share of GDP

increased from 65.6% of 2002 to 70.9% of 2007. At the same time, exports as a share of GDP

decreased from 66.4% to 48.5% causing a trade deficit equal to 22.4% of GDP in 2007.

During the period of a rapid increase in inward remittances (2002–2007), the real value of

imports (both consumer and industrial goods) increased by 2.5 times, but the share of consumer

goods increased, while that of industrial goods decreased.

Table 3: Imports Classified with Standard International Trade Classification (SITC)

Imports 2002 2007

mln. USD % mln. USD %

Food and live animals 86.6 8.4 270.1 10.6

Beverages and tobacco 1.4 0.1 11.5 0.5

Crude materials, inedible, except fuels 12.3 1.2 78.1 3.1

Mineral fuels, lubricants and related materials 319.5 31.0 499.2 19.6

Animal and vegetable oils, fats and waxes 6.3 0.6 21.8 0.9

Chemicals and related products 316.0 30.7 497.9 19.5

Manufactured goods classified chiefly by material 55.3 5.4 276.2 10.8

Machinery and transport equipment 182.4 17.7 321.8 12.6

Miscellaneous manufactured articles 50.0 4.9 568.6 22.3

Commodities/transactions not classified in the SITC 0.0 0.0 0.6 0.0

Other unspecified products 0.1 0.0 1.4 0.1

Total 1029.9 100 2547.2 100


Note: The value of imports in 2002 is the real value at 2007 prices.

Mineral fuels, lubricants and related materials were 31% of imports in 2002 and 19.6% of

imports in 2007. Chemicals and related products were 30.7% of imports in 2002 and 19.5% of

imports in 2007. Machinery and transport equipment were 17.7% of imports in 2002 and 12.6%

of imports in 2007. Despite the fact that the real value of the above mentioned goods, mainly

industrial goods, increased their share in imports significantly decreased.

Food, live animals, beverages, tobacco, vegetable oils, fats and waxes were 9.1% of imports

in 2002 and 12% of imports in 2007. Manufactured goods and articles were 10.3% of imports in

2002 and 33.1% of imports in 2007. These goods are mainly consumption goods, and their share

in imports has significantly increased (Table 3).


31

4. Model Specification

Most of the econometric studies of import demand such as Leamer and Stern (1970), Murray and

Ginman (1976), Goldstein and Khan (1985), Dornbusch (1988), Faini et al. (1988), Bertola and

Faini (1991), Hooper and Marquez (1993) and Carone (1996) have defined aggregate imports as

a function of real domestic income and the relative price of imports. Mathematically, this function

can be written as

, (1)

M is the quantity of real imports, Y the real domestic income, and RP the relative price or

the ratio of imports’ price to the domestic price level. t stands for time. Usually, the log-linear

form of equation 1 is used in the calculation as shown below:

(2)

Based on the purpose of the research, some studies have incorporated additional variables

into the model, such as lagged value of the variables3, GNP4, foreign exchange inflows5, foreign

exchange reserves6, real exchange rate7 and import tariffs8.

Most of the research related to macroeconomics of remittances has used a simple

Keynesian type multiplier to estimate the impact of remittances on consumption, investment and

imports, and remittances’ multiplier effect on the overall economy9. Unlike previous research,

this paper incorporates remittances into the import demand function, and applies the function to

quarterly time series, focusing on the impact of remittances on imports.

The main equation of the model is constructed by extending equation 2 by inclusion of

remittances and other related macroeconomic variables like the ratio of foreign trade to domestic

income (trade openness) as a proxy for import tariffs and the real effective exchange rate (REER).

The model has the following form10

∆ ∆ ∆ ∆ ∆ ∆ (3)

Real domestic income (Y), remittances (R), trade openness (OP) and appreciation of REER

are expected to have a positive impact on imports (M). The expected impact of an increase in the

relative price of imports (RP) should be negative; however, for some developing countries like

3 Leamer and Stern (1970) and Faini et al. (1988). 4 Leamer and Stern (1970), Murray and Ginman (1976). 5 Faini et al. (1988). 6 Leamer and Stern (1970) and Faini et al. (1988). 7 Dornbusch (1988), Bertola and Faini (1991). 8 Bertola and Faini (1991). 9 Stahl and Habib (1989), Nishat and Bilgrami (1991), and Glytsos (1993). 10 Considering the presence of unit root in the levels and the non-stationarity of the levels, the first differences are used in estimations.

Vol.8 2013

32

Tajikistan it can be positive. Since one quarter is not enough time for remittances to show an

impact on other macroeconomic variables, the one-period lagged value of remittances is

incorporated in the model.

REER is used as a proxy for relative foreign prices in some of the literature. In the

calculation of REER, the weight for each foreign trade partner country depends on the share of

foreign trade (exports and imports) with that country. Considering the fact that import and export

partners of Tajikistan are different countries we use relative prices of imports in the calculation,

which is based on the share of imports from each foreign trade partner country.

In order to test the exogeneity of remittances, remittances are regressed on their

macroeconomic determinants, and the residuals from the regression are used as an additional

independent variable in equation 3. The coefficient of the residuals is not significantly different

from zero showing that remittances are exogenous in the model.

5. Data Specification

The data used in calculations are quarterly time series data for the period from the first quarter of

2003 to the fourth quarter of 2011. The National Bank of Tajikistan (NBT) and TAJSTAT are the

main sources of data. WB and CBRF are the sources of data on remittances. The data for imports,

GDP and remittances are real values at constant prices of the first quarter of 2010. REER, trade

openness and relative price of imports are also based on the first quarter of 2010. All data are

seasonally adjusted (with Census X12, multiplicative).

Table 4: Summary Statistics

Variables Measure Observations Mean Std. Dev. Min. Max.

Imports mln. USD 36 564.733 187.681 233.854 889.180

GDP mln.USD 36 1008.193 381.552 398.840 1742.421

Relative price Rate 36 0.933 0.056 0.833 1.031

Remittances mln. USD 36 364.296 238.226 34.358 737.812

Trade openness Rate 36 0.922 0.192 0.558 1.202

REER Index 36 132.260 28.240 90.080 187.770

Source: NBT (2011), TAJSTAT (2011), CBRF (2012) and WB (2011)

Summary statistics for the data are presented in Table 4. Imports, GDP and remittances are

in million USD. Relative price of imports and trade openness are rates. REER is index.


33

Considering large standard deviation for imports, GDP and remittances the logarithmic form of

them will be used in estimations.

Figure-4: Model Variables

Imports GDP

Relative price Remittances

REER Trade openness


Note: The graphs show quarters on the horizontal axis for all variables; million USD on the vertical

axis for imports, GDP and remittances; an index for REER; and rates for relative price and trade

openness.

0

100

200

300

400

500

600

700

800

900

1000

2003

q1

2003

q3

2004

q1

2004

q3

2005

q1

2005

q3

2006

q1

2006

q3

2007

q1

2007

q3

2008

q1

2008

q3

2009

q1

2009

q3

2010

q1

2010

q3

2011

q1

2011

q3

0

200

400

600

800

1000

1200

1400

1600

1800

2000

2003

q1

2003

q3

2004

q1

2004

q3

2005

q1

2005

q3

2006

q1

2006

q3

2007

q1

2007

q3

2008

q1

2008

q3

2009

q1

2009

q3

2010

q1

2010

q3

2011

q1

2011

q3

0

0.2

0.4

0.6

0.8

1

1.2

2003

q1

2003

q3

2004

q1

2004

q3

2005

q1

2005

q3

2006

q1

2006

q3

2007

q1

2007

q3

2008

q1

2008

q3

2009

q1

2009

q3

2010

q1

2010

q3

2011

q1

2011

q3

0

100

200

300

400

500

600

700

800

2003

q1

2003

q3

2004

q1

2004

q3

2005

q1

2005

q3

2006

q1

2006

q3

2007

q1

2007

q3

2008

q1

2008

q3

2009

q1

2009

q3

2010

q1

2010

q3

2011

q1

2011

q3

0

20

40

60

80

100

120

140

160

180

200

2003

q1

2003

q3

2004

q1

2004

q3

2005

q1

2005

q3

2006

q1

2006

q3

2007

q1

2007

q3

2008

q1

2008

q3

2009

q1

2009

q3

2010

q1

2010

q3

2011

q1

2011

q3

0

0.2

0.4

0.6

0.8

1

1.2

1.4

2003

q1

2003

q3

2004

q1

2004

q3

2005

q1

2005

q3

2006

q1

2006

q3

2007

q1

2007

q3

2008

q1

2008

q3

2009

q1

2009

q3

2010

q1

2010

q3

2011

q1

2011

q3

Vol.8 2013

34

Figure 4 shows the evolution of the variables used in the model. All variables showed

increasing trends, except REER, which had a decreasing trend and trade openness, which was

fluctuating and decreasing slightly after 2008.

Imports, GDP and remittances fell in the second half of 2008 and the first half of 2009 as

the result of the world financial crises. Considering the decline in imports caused by the financial

crises of 2008, a dummy variable is incorporated into the model.

The correlation matrix for imports and other variables used in the model is presented in

Table 5. Imports are positively correlated with GDP, relative price of imports and remittances;

and negatively correlated with trade openness and real effective exchange rate.

An increase in GDP increases import demand: that is why the correlation of imports with

remittances is similar to theoretical expectations.

Table 5: Correlation of Imports with Explanatory Variables

Variables Imports GDP Relative

price Remittances

Trade

openness REER

Imports 1.000

GDP 0.888 1.000

Relative price 0.727 0.928 1.000

Remittances 0.951 0.952 0.835 1.000

Trade openness -0.498 -0.815 -0.868 -0.644 1.000

REER -0.866 -0.965 -0.941 -0.940 0.767 1.000


Under the condition that other variables are constant the positive correlation of imports

with the relative price of imports is controversial. Theoretically an increase in the relative price

of imports should affect imports negatively. However, the information about changes in volume

and prices of the main imported products (Table 6) show that the imported volume of some

products can increase despite the increase in the prices. On the other hand, a decrease in the

volume of imports of some products coincides with a decrease in prices.

The positive correlation of imports with remittances may demonstrate the positive impact

of remittances on imports. Confirmation of this suggestion through the use of the above model

and empirical data is the main purpose of this paper.


35

Table 6: Imports of the Main Industrial and Consumer Goods

Goods Imports share Price Import volume

2005 2010 2005 2010 2005 2010

Alumina 27.20% 11.9% 453.0$ 450.0$ 798691 708218

Oil 9.48% 16.3% 409.4$ 691.9$ 308069 627687

Grain 2.30% 3.07% 106.3$ 184.8$ 288577 442811

Flour 3.41% 3.7% 131.2$ 264.5$ 346334 370292


Note: Import volume is in tons.

The negative correlation between imports and trade openness is related to the increase in

GDP and decrease in exports as a percentage of GDP for the period of increase in imports. Trade

openness is defined as the ratio of foreign trade to GDP. That is why while imports are increasing

trade openness falls if GDP increases or exports decrease.

While REER appreciates, imports are expected to increase. That is why the correlation

between these variables should be positive. The negative correlation for the case of Tajikistan can

be explained by the increase in money supply, which was high for the period of increase in

remittance inflows.

In order to properly estimate the impact of remittances and other explanatory variables on

imports, the time series are checked for the presence of unit root and autocorrelation, and adjusted

for OLS estimation.

Table 7: Unit Root and Autocorrelation Tests

Variables Levels First differences

Lags: 0 Lags: 1 Lags: 0 Lags: 1 ADF DW ADF DW ADF DW ADF DW

ln Imports -2.064 1.52 -2.570 2.00 -4.395*** 1.89 -3.417** 2.01

ln GDP -1.542 2.59 -1.546 1.85 -8.155*** 1.86 -4.356*** 1.74

Relative price -0.602 2.22 -0.203 2.12 -6.642*** 2.13 -7.214*** 2.04

ln Remittances -2.391 2.25 -2.384 1.97 -5.817*** 1.98 -3.162** 2.05

Trade openness -1.178 2.19 -0.617 1.94 -7.097*** 1.98 -4.870*** 1.97

REER -1.430 1.94 -0.617 1.98 -6.656*** 2.00 -4.345*** 2.00

Note: ***Smaller than the critical value at 1% significance level; **Smaller than the critical value at

5% significance level; *Smaller than the critical value at 10% significance level.

Vol.8 2013

36

Considering the presence of a unit root in the levels, the first differences of all variables are

used in ordinary least squares (OLS) regression. The first differences are co-integrated and

stationary. The results of the Augmented Dickey-Fuller (ADF) test for unit root and the

autocorrelation test reject the null hypothesis of the existence of unit root and show that the series

are stationary (Table 7).

6. Empirical Findings

OLS regression is run in order to test the effect of each explanatory variable on the dependent

variable. Ultimately, the regression model is tested for serial correlation, co-integration, and

presence of heteroskedasticity. The results of the regression for the model are presented in Table

8.

Ignoring stationarity of the data shows that income and relative price significantly affect

imports. However, the co-integration approach shows that the impact of real income on imports

is positive and statistically significant at the 1% level, while that of the relative price of imports

is insignificant. A 1% increase in GDP growth is associated with 0.7%–0.8% increase in imports.

Considering the positive impact of remittances on private consumption (see Table 2) and GDP

(see Appendix), and the positive and significant impact of GDP on imports, we can say that the

overall impact of remittances on imports is stronger than we derived in our estimation.

The impact of the present-period remittances on imports is positive but statistically

insignificant. This can be explained by the fact that it takes time for remittances to be exchanged,

become available to foreign trade agents and to be used. Consequently, the impact of one-period

lagged remittances on imports is positive and statistically significant (at the 10% level).

According to the regression results, a 1% increase in remittances growth in the previous quarter

increases imports in the current quarter by at least 0.10%. Considering the positive and significant

impact of remittances on GDP growth, and the similar impact of GDP growth on imports, the

impact of the present quarter’s remittances might be stronger11.

Trade openness affected imports positively. The impact is statistically significant at the 1%

level. A unit increase in trade openness is associated with an average of 86.1%–91.7% increase

in imports. The positive impact of trade openness implies that while the domestic economy is not

able to meet a significant share of domestic demand, and there are no considerable barriers

restricting imports, increasing domestic demand will lead to a considerable increase in imports.

Trade openness is a good proxy for import tariffs, especially for the case of Tajikistan, where

11 For more information, see Table A1 in Appendix.


37

imports of the main industrial and consumer goods is monopolised, and officially reported tariffs

do not reflect reality.

Table 8: Regression Results, Dependent Variable – Imports

Independent variables

Equations 1 2 3 4 5 6

Coefficients Coefficients Coefficients Coefficients Coefficients Coefficients

(t value) (t value) (t value) (t value) (t value) (t value)

Δ ln GDP 0.7252

(4.44***) 0.7455

(4.49***)0.6996

(4.37***) 0.7707

(5.28***)0.7751

(5.41***) 0.741

(5.36***)

Δ Relative price 0.3487 (0.47)

0.3690 (0.49)

0.4942 (0.70)

0.4954 (0.71)

Δ ln Remittances 0.0438 (0.65)

0.0394 (0.53)

0.0551 (0.78)

Δ ln Remittances (lagged)

0.1171 (1.93*)

0.1259 (1.88*)

0.1170 (1.70*)

0.1140 (1.91*)

0.1181 (2.04*)

0.1097 (1.86*)

Δ Trade openness

0.8826 (7.34***)

0.8885 (7.21***)

0.8605 (7.59***)

0.9174 (8.60***)

0.914 (8.73***)

0.8914 (9.00***)

Δ REER 0.0012 (0.53)

0.0013 (0.54)

0.0009 (0.39)

0.0009 (0.39)

D2008 0.0050

(0.13) 0.0024 (0.06)

Constant 0.0031 (0.22)

-0.0018 (-0.11)

0.0044 (0.24)

0.0040 (0.29)

0.0013 (0.11)

0.0077 (0.61)

R-squared 0.7846 0.7825 0.7829 0.7812 0.7801 0.7774 F-statistics 16.39 16.19 16.23 20.00 25.71 25.32

DW d-statistics 1.93 1.90 1.90 1.96 1.93 1.93 EG-ADF test -5.529*** -5.448*** -5.422*** -5.640*** -5.557*** -5.530*** RESET (F value)

0.94 (0.4355)

0.88 (0.4636)

1.06 (0.3832)

0.33 (0.8053)

0.34 (0.7955)

0.40 (0.7529)

Observations 36 36 36 36 36 36 Note: *** significant at 1%, ** significant at 5%, and * significant at 10%.

The impact of REER is positive and insignificant. It seems that the fluctuations in imports

are not affected by changes in the real effective exchange rate of the national currency. It might

be related to the high level of money supply growth for the period 2003–2011.

Despite the fact that imports fell when the 2008 financial crises occurred, the effect of the

dummy variable for the crises is not significant. This can be explained by the fact that there was

a similar fall in other macroeconomic variables, like GDP and remittances, for the period of crises.

That is why the effect of the dummy variable for the 2008 financial crises on imports was

insignificant.

The value of R-squared is equal to 0.77–0.78, thereby indicating that the percentage of

variance in imports which is explained by the independent variables is sufficiently high, and the

outcomes are well predicted by the model. Moreover, the F-statistics are significant.

Vol.8 2013

38

The Durbin-Watson d-statistics show that all values are remarkably close to 2, and the time

series used in the equations are not affected by autocorrelation.

Engle and Granger (1987) suggested a two-step process (an OLS regression and an ADF

unit root test on the residuals) to test for co-integration, which shows that the hypothesis of co-

integration cannot be rejected for all equations, and all the equations adequately explain the long-

run relationship between imports and the explanatory variables.

The results of the Ramsey Regression Equation Specification Error Test (RESET) show

that all the equations in the model are free of misspecification problems, and the model has no

omitted variables.

An increase in imports caused by remittances can be interpreted in different ways. An

increase in imports of industrial goods promotes development of the domestic economy, while an

increase in imports of consumer goods enables the meeting of domestic demand. On the contrary,

an increase in the marginal propensity to import decreases the multiplier effect of remittances on

the overall economy. Furthermore, an increasing negative foreign trade balance has an

unfavourable effect on the economy in the long run (see Appendix).

7. Conclusions and Policy Implications

This paper examines the impact of remittances on imports. It demonstrates the increase in imports

caused by an increase in real domestic income and remittance inflows. A comparative study and

regression analyses prove that remittances have significant impacts on imports. The impact of

remittances on imports might be stronger if to take into account the impact of remittances on real

domestic income and its indirect impact on imports via the increase in real domestic income.

Analysis of data on households’ expenditure, and comparison of expenditure patterns of

households with remittances and households without remittances, showed that savings make up

only a small percentage of populations expenditure share, and remittances affect consumption

expenditures significantly while their impact on investment expenditures is insignificant.

The paper demonstrates decreases in the labour force participation rate, the share of exports

in the economy, industrial output and agricultural output, and an increase in imports of consumer

goods which could be produced domestically. Such a situation has brought about a huge trade

deficit which affects the economy negatively.

Dependence on imports and remittances makes the economy of the country weak and

vulnerable to external shocks. That is why the use of remittances to enhance the economy and to

decrease dependence of the economy on remittances and imports should be main components of

the country’s economic policy.


39

In order to maximize the positive impacts of remittance inflows on the overall economy,

the government should establish a long-run policy of stimulation of private savings and support

for domestic production of goods and services. Such a policy would increase the profitability of

remittances, and might be a key solution to the gradual decrease of dependence on remittances

and imports.

Appendix

Table A1 OLS Regression Results, Dependent Variable – GDP

Independent Variables

Equations 1 2 3

Coefficients (t value)



Δ ln Remittances 0.1503

(2.58**) 0.1344

(2.30**)

Δ ln Trade Deficit -0.0420

(-2.24**)

Δ REER Constant

0.0207 (1.78*)

0.0036 (1.59) 0.0266 (1.57)

0.0097 (0.82)

R-squared 0.6349 0.6495 0.5750 F-statistics 7.83 5.79 8.46

LM test chi2 (P>chi2) 0.000

(0.9874) 0.214

(0.6434) 1.342

(0.2467)

Ramsey RESET test 1.35

(0.2830) 1.17

(0.3437) 1.03

(0.3972) Observations 34 34 30

Control variables Δ ln GDP lagged, Δ ln Investment, Δ ln Government Consumption, Δ Trade Liberalization, Δ Labour, Δ Prices, Δ ln FDI and Δ ln M2

Note: *** significant at 1%, ** significant at 5%, and * significant at 10%.

Vol.8 2013

40

REFERENCES

Bertola, G. and R. Faini. 1991. Import demand and non-tariff barriers: the impact of trade liberalization. Journal of Development Economics 34: 269–286.

Carone, G. 1996. Modeling the U.S. demand for imports through cointegration and error correction. Journal of Policy Modeling, Elsevier 18(1): 1–48.

CBRF. 2012. Cross-border Remittances via Money Transfer Operators in Breakdown by Countries. The Central Bank of Russian Federation. Moscow.

Dornbusch, R. 1988. Exchange rates and inflation. Cambridge, MIT Press.

Engle, R. and C. Granger. 1987. Co-integration and Error Correction: Representation, Estimation and Testing. Econometrica 55(2): 251–276.

Faini, R., Pritchett, L. and F. Clavijo. 1988. Import demand in developing countries. Policy Research Working Paper Series 122, the World Bank.

Glytsos, N. 1993. Measuring the income effects of migrant remittances: A methodological approach applied to Greece. Economic Development and Cultural Change. University of Chicago Press. 42(1): 131-68.

Goldstein, M. and M. S. Khan. 1985. Income and price effects in foreign trade, Handbook of International Economics, in: R. W. Jones & P. B. Kenen (ed.), Handbook of International Economics, edition 1, volume 2, chapter 20, pp. 1041–1105, Elsevier.

Hooper, P. and J. Marquez. 1993. Exchange rates, prices, and external adjustment in the United States and Japan. Board of Governors of the Federal Reserve System. International finance discussion papers. Number 456.

Leamer, E. E. and R. M. Stern. 1970. Quantitative International Economics. Boston: Allyn and Bacon.

Murray, T. and P. J. Ginman. 1976. An Empirical Examination of the Traditional Aggregate Import Demand Model. The Review of Economics and Statistics, MIT Press, 58(1): 75–80.

NBT. (2011). Statistics and balance of payments. Dushanbe.

Nishat, M. and N. Bilgrami. 1991. The impact of migrant worker’s remittances on Pakistan economy. Pakistan Economic and Social Review, 29: 21–41.

Stahl, Ch. and A. Habib. 1989. The impact of overseas workers’ remittances on indigenous industries: Evidence from Bangladesh. The Developing Economies 27 (3): 269-85.

TAJSTAT. 2011. Tajikistan in figures. The Statistical Agency under the President of the Republic of Tajikistan. Dushanbe.

UNDESA. (2011). Net migration. Population Division. New York.

World Bank. 2011. Migration and Remittances Factbook. Washington D.C.

World Bank. 2011. Workers' remittances and compensation of employees. World Development Indicators. Washington D.C.

World Bank. 2007. Tajikistan Living Standards Measurement Survey (TLSS). Washington D.C.

Documents

2_Sultonov_IJEPS2013