The effect of remittances on economic growth in India

Student Name: Dick Lorier

Student Number: 5603455
Course Description: Bachelor Thesis
Date Completed: 25-08-2010
Supervisor: Matija Lozej

Bachelor of Science Program in Economics and Business

Faculty of Economics and Business
Universiteit van Amsterdam
Table of contents

Chapter 1 Introduction.3
Chapter 2 Remittances in India
2.1 Definition and measurement of remittances.4
2.2 Macroeconomic determinants of remittances...5
2.3 Why are remittances so high in India?..........................................................7
Chapter 3 Theories of the relation between remittances and economic growth
3.1 Remittances in relation with investment and consumption..8
3.2 Remittances and the development of the financial sector...10
3.3 Remittances as (a)cyclical source of finance..10
3.4 Remittances and trade.11
3.5 Remittances and the labour market.12
3.6 Remittances and political economy effects....12
Chapter 4 Evidence of the relation between remittances and economic growth
4.1 Evidence: Remittances in relation to investment and
consumption..14
4.2 Evidence: Remittances and the development of the financial
sector.16
4.3 Evidence: Remittances as (a)cyclical source of income.17
4.4 Evidence: Remittances and trade17
4.5 Evidence: Remittances and the labour market....18
Chapter 5 Empirical analysis of the cyclicality of remittances.........................19
Chapter 6 Conclusion....21

Bibliography..23
Appendix 1 Remittances, FDI and official development assistance....25
Appendix 2 Remittances in India between 1970 2008..26
Appendix 3 Data on macroeconomic key factors.....27
Appendix 4 Empirical analysis results..28

2
Chapter 1: Introduction
The relation between remittances and economic growth highly depends on the country
characteristics and substantially on migration patterns over time. Without a share of the
countrys people living abroad there will be a limited inflow of remittances. According to the
estimates of the High Level Committee on the Indian Diaspora in 2001 the total stock of
emigrants is 18.5 million-strong diaspora (including descendants of Indian migrants). Besides
their emotional and cultural link with India, the vast majority of the Indian migrants have
economic ties with their Indian family.
Since 2003 India is the top-receiving country of remittances in the world. In 2008
remittances accounted for 49 billion US dollars and 4,3% as a percentage of Gross Domestic
Product (GDP). The magnitude of remittances is substantial as remittances exceed Foreign
Direct Investment (FDI) and development assistance from the 1970s. Particularly since the
1990s remittances exceed FDI and official development assistance by more than 2 percentage
points of GDP in 2000 (see figure in appendix 1). According to the Reserve Bank of India
(RBI) the recent global economic crisis does not have a significant impact on the inflow of
remittances. On the microeconomic level remittances are a vital part of the income of Indian
families and on the macroeconomic level remittances affected the economy of India as a
whole. In this thesis I will discuss the macroeconomic aspects of remittances in India. The
central question in this thesis is the following: What is the effect of the increase of remittances
between 1970-2008 on economic growth in India?
The thesis is organized as follows: Chapter 2 describes remittances in India and is
divided in three paragraphs. Paragraph 2.1 discusses definition and measurement issues of
remittances in India. Paragraph 2.2 describes the macroeconomic determinants of remittances
in India and the final paragraph 2.3 analyses the upsurge in remittances to India. In chapter 3
the theoretical mechanism between the relation of remittances and economic growth are
described. This chapter is divided in six paragraphs. Paragraph 3.1 explains the relations
between remittances and the effects on investment and consumption. Paragraph 3.2 describes
the implications of remittances for the financial sector and paragraph 3.3 describes if
remittances are an (a)cyclical source of finance. Paragraph 3.4 describes the relation between
remittances and trade. In paragraph 3.5 the implications of remittances for the labour market
are described. The final paragraph 3.6 will consider political economy effects of remittances.
In chapter 4 the above mentioned theories are examined in five paragraphs 4.1-4.5 with
evidence from India. In chapter five an empirical analysis will be preformed about the

3
cyclicality of remittances compared with consumption and FDI. Chapter 6 concludes this
thesis.

Chapter 2: Remittances in India

2.1 Definition and measurement of remittances

Remittances are generally described as the transfer of money by a foreign worker to his or her
country of origin. In this paper remittances are referred to cross border remittances, as
transfers of money from a Non-Resident Indian (NRI) towards an Indian resident. The RBI
and the International Monetary Fund (IMF) provide data of remittances. The RBI
distinguishes total remittances by inward remittances towards family maintenance and
remittances in the form of deposits in the NRI deposit schemes in Indian banks (RBI, 2009,
p.407). These deposit schemes are exclusive for NRI. NRI depositors are given the choice of
holding deposits in foreign currency or in Indian rupees. If NRI deposits are locally
withdrawn in rupees the RBI recognizes them as remittances (Chishti, 2007, p. 5). Inward
remittances towards family members are treated as private capital flows and are included on
the current account of the balance of payments. As opposed to NRI deposit schemes on Indian
banks which are treated as capital account transactions (RBI, 2009, p.407). The IMF provides
data for remittances in the annual Balance of Payments Statistics where remittances are
specifically defined in three components, namely the workers' remittances, compensation of
employees and migrant transfers. Workers remittances are transfers from migrants to resident
households in the country of origin. The migrants stay abroad for one year or longer and this
is usually an ongoing flow between family members. Compensation of employees are
remunerations paid to individuals who work in another country, other than where they legally
reside. Migration transfers refer to capital transfers of financial assets made by migrants as
they move from one country to another and stay abroad for one year or longer (Chishti, 2007).
The sum of these three components are the total remittances reported by the IMF and as such
used in this paper. The different definitions used by institutions and countries cause problems
to compare countries statistics. The various definitions have different dynamics with
consequently different outcomes overtime.
Remittances appear on the current account of the balance of payment which reflects
transactions of goods, services and income. In the economic data remittances are captured in
the Gross National Product (GNP). GNP is the sum of the values of produced goods and
rendered services and transferred capital by Indians, both in India and abroad. Remittances

4
are not captured in the GDP, because the GDP encloses the sum of the values of produced
goods, rendered services and transferred capital within the country, both by Indians and
foreigners.
Remittances are difficult to measure, because individuals conduct numerous small
transactions through various formal channels and informal channels. The formal channels
consist of authorized money transfer businesses, such as banks, controlled by the financial
services regulatory officials. These formal flows are captured in the data of the IMF and RBI.
The informal channels are not controlled by the financial services regulatory authorities and
are to a large extent illegal. An informal transfer, for instance to hand over goods from
personal baggage or carried cash, is not necessarily illegal. An important unofficial channel in
India is the Hawala or Hundi system. This trust based system transfers remittances without an
actual movement of cash from abroad to India. The NRI remitter transfers money in a foreign
currency to intermediaries. The intermediaries agree on the commission and organize the
payment to the receiving Indian resident (Sasikumar and Hussain, 2007, p. 30-38).
The volume of unofficial remittances is not captured in the official data of the IMF
and the RBI. Due to this fact, the actual inflow of remittances is underestimated. Another
factor that affects the underestimation of the inflow of remittances is the state of the financial
sector. For instance, if banks are poorly developed, this may raise the costs of remitting and
give an incentive to transfer remittances trough the informal channels (Chishti, 2007). Also
the policy of the government is important, because this influences the incentive to remit via
the formal channels. After the terrorist attacks of September 11, 2001 remittances shifted from
informal to formal channels, because worldwide authorities have strengthened security
measures regarding financial flows. On the other hand, tightening of the border controls may
have increased undocumented migration which may have increased remittances via informal
channels (Sasikumar & Hussain, 2007, p.13).

2.2 Macroeconomic determinants of remittances

In order to analyze the macroeconomic impact of remittances in India, it is important to look
for factors that determine remittances and in particular the specific economic developments in
India. Various factors play a role in the macroeconomic determinants of remittances in India.
A macroeconomic factor that could determine the inflow of remittances is the number of NRI
workers. The number of NRI workers could be an important factor, because the greater the
volume of NRI workers, the greater the volume of remittances. Estimates on cross border

5
migration in India vary from a stock of 10 million emigrants in 2005 (Ratha and Xu, 2008) to
a stock of 25 million emigrants in 2009, according to the Indian government (Naujoks, 2009).
The economic situation in the host country is a possible important macroeconomic
determinant. Favourable economic conditions increase employment and earnings perspectives,
which could increase remittances to the home country. Also, the economic situation in the
home country could be an important factor, because in bad economic times remittances are
dearly needed. Besides, differences in wage levels of the host country and home country
could be an important determinant (Hagen-Zanker and Siegel, 2007, p.9). If we consider the
background of India a significant share of the Indian workers reside in oil exporting countries
in the Gulf States. The inflow of remittances to India from these countries could be
significantly influenced by fluctuations in the growth of oil exports. Also, fluctuations of the
GDP in the United States (US) could be important in India since a great part of the Indian
migrants reside in the US and the major part works in the US services sector, such as IT.
Especially, the US services sector is sensitive to fluctuations of GDP with a great impact on
demand and supply of the international labour market and consequently for NRI workers.
Because of this, these fluctuations determine the flow of remittances to India. For instance, if
the US is in phase of a cyclical downturn, the demand for labour will be lower and wages will
deteriorate. NRIs receive a lower wage or even become unemployed and due to this fact the
volume of remittances to India shrinks (Jadhav, 2003, p.13-15).
Other possible important determinants of remittances could be differences in exchange
rates and differences in the interest rate between host and home country. Inward remittances
for the sake of family maintenance are expected to be less sensitive for interest fluctuations
and exchange rate differences. However, another way of remitting is via NRI deposit schemes,
since a predominant part of NRI deposit schemes are locally withdrawn in Indian rupees. NRI
deposits may be affected by differences in exchange rates and differences in the interest rate,
because NRI deposits are a capital flow and to a large extent intended for short term
investments with a speculative nature (Sasikumar and Hussain, 2007, p. 14-17). Jadhav
(2003) finds that NRI deposits are significantly influenced by economic activity in the host
country, exchange rates movements and interest rates differentials.
Another determinant of remittances could be the quality of the domestic institutions.
As Catrinescu et al. (2009, p. 14-15) suggest that well-functioning institutions give people the
incentive to invest remittances in the domestic economy. Besides, political risks in the home
county could be an important determinant of remittances. A high political risk reduces

6
favourable conditions to undertake investments. Although, in periods with a high political risk
more remittances are vital and may increase.
Jadhav (2003) finds that the economic activity in the host country and the exchange
rate elasticity have a significant impact on remittances in India. To measure economic activity
in the host country Jadhav uses the variables US GDP and the export price of oil for oil
exporting countries. In addition, Gupta (2005) uses a larger set of variables, a more refined
econometric method and data from the RBI and the IMF statistics. The regression results of
Gupta show that the number of workers, wage rate differences, the economic situation in the
home country and the economic situation in the host country are found to be significant
determinants of the inflow of remittances in India. The other variables are found to be
insignificant including political uncertainty, relative interest rate differences, exchange rate
deprecations and oil prices.

2.3 Why are remittances so high in India?

Since 2003 India is the top-recipient of remittances. Remittances account for 49 billion US
dollars and 4,3% as a percentage of GDP in India. From the beginning of the 1990s the inflow
of remittances started to rise gradually and increased strongly in the last 15 years (see figure
appendix 2). In accordance with this observation, Gupta (2005) finds that remittances exhibit
a strong upward linear trend since the 1990s.
The enormous increase of remittances in the last two decades in India is explained by
several developments. The increase in migration of Indian workers contributed to the increase
of remittances in India. In the 1970s a sharp increase in the migration of Indian workers to the
Gulf States arose as a response to the labour shortages due to the oil discoveries and oil price
shock of the 1970s. The migration from Indian workers to the Gulf States peaked in the
beginning of the 1980s. In the mid 1990s a sharp increase in the (temporary) migration of
Indian workers to the Australia, Canada and particularly the US arose as a response to the
demand for mainly high skilled workers in IT related fields (Chishti, 2007). In the period
between 1990 and March 2001 the Indian population that resided in the United States doubled
and more than tree quarters of the workers had a bachelor degree or better (Desai, Kapur,
Mchale, 2001, p.3). The number of Indian workers abroad has grown and since the mid 1990s
the average earnings of Indian workers are sharply increased, that reflects the high skills of
Indian workers (Gupta, 2005, p.3).
Besides, remittances are increasingly transferred through formal channels instead of
informal channels, which contributed to the high increase in formal remittances to India. In

7
the beginning of the 1990s economic reforms contributed significantly to this shift. Important
reforms are the liberalization of gold imports in 1992, which have eased the tight controls of
the transfer and possession gold, a highly valued good in India 1. And the introduction of a
market based exchange rate in 1993, which enabled NRIs to convert rupees into foreign
currencies and made NRIs less reluctant to remit money to rupee deposit accounts. Between
1991 and 2005 the part of rupee denominated deposits of total deposits of NRIs decreased
from 72% to near 35 % respectively, which reflects that the exchange rate reforms have a
significant impact on the inflow and the currency structure of remittances (Chishti, 2007).
The economic reforms increased the incentives to remit via the formal channels and
consequently contributed to the increase of remittances in India.
Also the Indian government offers several deposit schemes and bonds for NRIs, such
as the Money Transfer Service Schemes (MTSS) and the Rupee Drawing Agreements (RDA),
with attractive interest rates. An extensive part of these NRI schemes and bonds are
withdrawn locally (Mallick, 2008, p.12). In 2002-2003 in India workers remittances
accounted for 66% of private transfers and the remaining 34% transfers are withdrawn locally
from NRI deposit schemes (Jadhav, 2003, p. 6).
Another factor is the increased possibilities to remit in India. The increased
competition between money transfer agencies resulted in lower costs per transaction and
better incentives to remit for migrants. This development has contributed to the increase of
remittances to India. The banking sector accounts for the largest part of the remittances in
India, but remittances via money via internet based agencies has also increased during the last
decade (Chishti, 2007).

Chapter 3: Theories of the relation between remittances and economic growth

The inflow of remittances may have a broad effect on the decisions of recipients about labour
market participation, consumption, investment, education, migration and shelter with
potential implications for the countrys long run economic performance (Rapoport and
Docquier, 2005, p. 6). According to Chami (2008, p. 38) remittances have a direct effect on
recipients through serving as a source of income, regardless of the motives of the remitter.
Remittances simply drive economic growth through increasing national income measured in
GNP. Beyond the impact on recipients, inflowing remittances may have an economic impact
on non-recipients. Moreover, there could be an economic impact on the receiving economy as

1
Indians have a huge fascination for gold. This is evident in the fact that India is the largest consumer as well as
importer of gold in the world.

8
a whole. Including the investment environment, financial sector, macro economic stability,
trade sector, labour market, and even in a broader sense, political decision making
(Catrenescu, et al. p 1-3).

3.1 Remittances in relation with investment and consumption

The economic impact of remittances on the domestic economy mainly depends on the
utilisation of remittances by recipients. Using remittances for consumption has a different
effect on economic growth than using remittances for investments. If remittances are used for
consumption it is expected to have a limited effect on the economic growth of a country. On
the other hand, for productive investments the effect on economic growth is strong. For
example, when a farmer invests in extra animals in his live stock, this may have a long term
effect on his economic position. Likewise, remittances can finance business activities.
Potential strong evidence for economic growth could be found in the relation between
remittances and small company start ups.
Although, the effect of remittances on economic growth is also affected by the
multiplier effect. The multiplier effect could also work for remittances that are spent on
consumption. The multiplier effect is working if remittances increase consumption;
subsequently this could create extra demand and this could result in economic growth.
Next to the effect of remittances on economic growth, remittances may improve the
basic needs of people and may lift recipients out of poverty. Remittances are generally spend
on consumption necessities, such as food, clothing, medicine and shelter, which directly
improves the prosperity of the recipient and this effect is widely recognized (Barajas et al,
2009, p.3). In addition, remittances tend to flow from relatively rich migrants towards
relatively poor households and remittances recipients in less developed countries spend a
higher portion of the received remittances on consumption (Mallick, 2008, p.9). These two
tendencies enlarge the positive effect of remittances on poverty reduction and subsequent
income equality in the country.
Remittances are part of the savings of workers abroad and increase the total pool of
available resources for investments in the domestic economy (Solimano, 2003, p. 6). If
recipients do not directly spend remittances in the domestic economy, but save remittances on
a deposit account of the bank, it could enhance economic growth in case the additional funds
are used for loans that finance productive investments.
A potential positive effect of remittances on economic growth is that remittances may
increase investments by becoming a substitute for inefficient or non existing credit markets

9
(Guiliano and Ruiz-Arranz, 2006, p.1). Efficient credit markets enable people to obtain credit
in order to undertake investments contributing to economic growth in the domestic country. In
the efficient credit market case, if a large group investors receive remittances, they benefit
from the willingness by banks to provide loans, due to the increased creditworthiness.
Because of this remittances increase the ability of the borrowers to meet the debt obligations
set by the bank. Recipients are able to undertake investments with the remittances and can
also obtain a higher loan from the bank.
But credit markets may also work inefficiently as a result of credit rationing. Credit
rationing is the situation when a bank limits the supply of loans, although the bank has
sufficient funds to provide the loans, even when a borrower is willing to accept the terms of
the bank. The inflow of remittances may enable borrowers with a higher risk to improve their
collateral, so that the risk of the borrower will decrease and consequently the bank will
increase the supply of loans. This may reduce the problem of the credit rationing. Mainly poor
people suffer from the credit rationing problem, because they fall in the high risk pool as they
have limited collateral. The presence of the inflow of remittances can alleviate this credit
market problem and may improve the credit possibilities to undertake potential productive
projects, which can lead to economic growth in the domestic country.

3.2 Remittances and the development of the financial sector

The impact of remittances on the development of the financial sector and the impact on
economic growth is a priori not clear. Several mechanisms are at work in the financial sector
with opposing effects on economic growth.
The development of the financial sector is affected through the increase in
creditworthiness of the recipients. The increase in creditworthiness creates the possibility for
the bank to make more new or augment existing loans. Besides, the banking sector benefits
from the fact that risk on default is diversified. In case that a particular investor defaults, the
bank will suffer a loss, but the bank will not end up in credit problems or even go bankrupt.
The bank is better secured against the risk of default by a single investor, because the large
group of investors, with different investments, compensates the risk of default by a single
investor. This effect is general, but for remittances the effect could be more intense. As
remittances flow to a large group of potential investors, with different risk profiles, the bank is
better secured against default. As a consequence of this, the risk premium set by the bank will
decrease. The interest rate on loans calculated by banks will decrease, because the interest rate
partly consists of the risk premium. In the domestic financial sector the cost of capital will

10
decrease and the demand for loans will increase with positive effects for the economic growth
in the country.
Another mechanism that affects the development of the financial sector is the increase
in financial transactions, through which possible economies of scale arise. As Chami (2008, p.
41-42) notes, the inflow of remittances increases the demand for transactions by the domestic
banking sector. The total volume of transactions in the domestic banking sector will increase
and this may result in a decrease of the cost per transaction for the bank. Phrased differently,
economies of scale may arise from this process. The domestic banking sector will operate
more efficient and this may result in a lower cost of capital in the financial sector of the
country that may have a positive effect on the economic growth.
Remittances may also have a negative effect on the development of the financial
sector. Aggerwald et al. (2005, p. 2) notice that the increase of remittances, as an inflow of
new available funds outside the domestic banking sector, may function as a substitute for the
domestic banking sector. The increase of remittances offers investors an alternative to finance
investments. Therefore, the demand for bank credit will decrease and the development of the
banking sector will be negatively affected.

3.3 Remittances as (a)cyclical source of finance

A positive link between remittances and economic growth is that remittances may increase
macroeconomic stability in the domestic economy. The inflow of remittances is considered to
be a more stable flow of capital than aid, FDI and other capital flows (Ratha, 2003, p. 31).
Subsequently, the inflow of remittances decreases the volatility in the domestic economy.
Therefore the risk premium of investors may decrease and undertaking investments will
become more attractive. Hence more investments will be undertaken and all this together may
lead to economic growth in the domestic economy (Barajas, et al., 2009, p. 6). If recipients of
remittances experience bad times in the home country, for instance due to an economic crisis
or a flood disaster, remittances may increase and work countercyclical for the domestic
economy (Solimano, 2003, p.2).

3.4 Remittances and trade

A potential negative effect of the inflow of remittances on economic growth in the domestic
country is the so called Dutch Disease effect. If the inflow of remittances causes an
appreciation of the real exchange rate and a hampering of the tradable sector, the
competitiveness of the country on the world market will decrease. If remittances flow in a

11
country, the demand of goods by economic agents in the domestic economy will increase.
This demand is typically biased towards non-tradable domestic goods. While prices tend to
rise in the non-tradable sector, the prices for the tradable sector are determined on the world
market (assuming the country does not affect the world market prices). As a result the real
exchange rate will appreciate and this hampers the tradable sector in the domestic economy.
The Dutch Disease effect is a relevant problem for countries that receive a considerable
amount of remittances. India could be a good example for the Dutch Disease effect, because
India is the top-recipient of remittances in the world.

3.5 Remittances and the labour market

If recipients of remittances optimize between labour and leisure, they may substitute the
remitted income flow for labour income and increase the consumption of leisure.
The inflow of remittances, or phrased differently, a non-labour income, will increase the
minimum wage that a worker will accept. This substitution effect will lead to a reduction in
the labour market participation of recipients and will decrease the labour market supply (P.
Acosta, 2006, p.36).
Remittances do not necessarily have a negative effect on the domestic labour market,
because remittances may have indirect effects on the labour market. The inflow of remittances
may increase the demand for goods as extra income is spend on the local market.
Consequently, the demand for domestic labour will increase, this will cause growth of the
domestic labour market and in turn the domestic economy. Also, remittances could be used to
substitute labour for capital in the production process. This process may enhance the market
position of a domestic company by the more efficient production and result in growth of the
company. In this process of growth extra labour is demand and this in turn may generate extra
employment with positive effects for the domestic labour market. The net effect of labour
substitution is however ambiguous.
A positive effect of remittances on the labour market is that remittances may enable to
a small group of extremely poor recipients to participate on the labour market. Normally this
group is inactive on the labour market due to the lack of basic living standards, like nutrition
and shelter, to reach the minimal level of living standard. As well, the income of remittances
may enable households to send their children to school, increasing their educational standards
and this may have positive externalities for the labour market and for economic growth. Even
if remittances tend to flow to relatively rich emigrant households. Even though remittances
decrease the individual labour market participation, the non-market household labour may

12
increase and benefit the household, in for example the parenting, cultural activities etc. with
potential positive effects for the educational level of the children (P. Acosta, 2006, p.37).

3.6 Remittances and political economy effects

Remittances may have a broader influence than solely on the economy. The interaction
between economic policies and institutions may be affected by the inflow of remittances. As
Catrinescu et al. (2009, p 14-16) conclude institutions play an important role in how
remittances affect economic growth.
The inflow of remittances can create a larger group of depositors that in turn may
affect the domestic government in order to undertake beneficial macroeconomic policies
(Chami, 2009, p. 42 & p. 59). For example, the domestic government may undertake reforms
to improve the safety of the domestic banking. On the other hand, Barajas (2009, p. 5-10)
outlines that the inflow of remittances may reduce incentives for recipients of remittances to
monitor the government policy effectiveness. Consequently, potential costs of bad economic
policies of the government are partly alleviated by the permanent inflow of remittances and
this may deteriorate the quality of domestic institutions.
From the perspective of the government the effect of remittances is not clear, because
political economy effects of remittances on the domestic economy are uncertain and depend
on various factors with a complex interaction in the countries economy. The effect on
economic growth is expected to be small. A deep and extensive analysis of the political
economy effects of remittances in the recipient country is beyond the limits of this thesis.

Chapter 4: Evidence of the relation between remittances and economic growth

As described in chapter 3, remittances have an ambiguous impact on several macroeconomic
key factors. The number of empirical studies about the link between remittances and
economic growth for India is limited, partly due to the complexity of the empirical research.
Jadhav (2003) executes a simple plot relating to long run growth of income per capita
and remittances between 1952-2002 to India, but Jadhav does not find a significant relation.
Nevertheless, in India the correlation between growth of income per capita and remittances
improved from 0.09 in the 1970s to 0.22 during the 1990s. A more sophisticated attempt is
done by Mallick (2008) using time series models. Firstly, the author estimates a long run
private consumption model for the period from 1966-1967 to 2003-2004 using the dynamic
ordinary least square method. The author finds that remittances have a significant positive
effect on private consumption. As expected, remittances simply add up to the income of

13
recipients. Then Mallick examines the impact of remittances on private investment and output
growth. The study finds that remittances have an adverse effect on private investment. This
could be due to the increase in withdrawal of resources from investment towards private
consumption. This fact indicates that remittances are principally used for consumption instead
of a result in increased private investments. The study also finds that if the variable public
debt is dropped and augmented with the variable public sector investment, remittances
become an insignificant factor. This may imply that remittances, as part of the private sector
investment, crowd out private fixed investment. Finally, Mallick finds that the growth of
remittances has no effect on the growth rate of output. However, the upsurge in remittances
may cause inflation, due to the increased consumption demand and stock of foreign currency
which increases domestic money supply.

4.1 Evidence of the remittances in relation to investment and consumption

The impact of remittances on the Indian economy is not wide spread, but has a regional
disparity. On the macroeconomic level, for India as a whole, no study is done to the end use
of remittances and subsequent implications for economic growth. However, some evidence
can be found on a regional level, although the drawn conclusion can not simply be extended
to nationwide India.
Kerala is a striking example of a region that benefited enormously from the migration
and related remittances inflow. In 1998 around 10% of the workforce of Kerala was employed
abroad. Remittances to Kerala contributed on average 21% of total income during the 1990s
and the income level exceeded the national income of India on average by 49%. Between
1998-2008 remittances, as share of total income, increased from 25% till 31% respectively.
(Zachariah and Rajan, 2008). This higher income resulted in higher consumption and from the
2000s consumption in Kerala exceeds national consumption by 41%. Until the beginning of
the 1990s the saving rate was similar to the saving rate nationwide around 22%, but in the
period from 1991-1992 till 1999-2000 the saving rate increased till 50%. The increase in
saving rate is also reflected by the increase in total bank deposits which in this period more
than quintupled (Kannan, Hari, 2002).
The economic impact of remittances is importantly determined by the end use by
recipients of remittances. A survey in 1998 in the state of Kerala shows that the major part of
the received remittances where used for consumption purposes; 86% of the households
mentioned living expenses as the main expenditure item of the received remittances. The
other important uses are education (36%), repayment of debt (27%), construction and repairs

14
of buildings (11%), and bank deposits (8%) (Zachariah, Mathew, Rajan, 1999). In a more
recent survey by Zachariah and Rajan (2008) the end use of remittances by households in
Kerala in 2008 found to be similar as 74,5% of the households used remittances for
subsistence, 38,9% for education, 36.9% for repayment of debt, 14,6% for bank deposits,
9,4% for buying and building houses, 5,6% for land reclamation, 3,1% for dowry payment,
2,6% for purchase of land 0,4% for business and 6,3% for other purposes.
If we range remittances by importance they are primary used for daily consumption.
The above figures show that remittances simply add up to the income and that this income is
consumed with limited direct effects on economic growth. Secondly, a large share of the
received remittances is used for educational purposes. This may have potential high benefits
for the educational level of the children of migrants and consequently positive effects for
economy of India. Thirdly, remittances are used for repayment of debt. A usual way to
finance migration is to loan money from your family. After the migrant is settled, a part of the
remittances are determined to pay back the loan. Money that is left after the previous
mentioned purposes is saved on a bank deposit or invested in housing. Finally, a relatively
small part of the remittances are directly used for investments with positive effects on
economic growth.
Now that we have a clear picture of the end use of remittances in Kerala we can
determine the impact of remittances and compare emigrant with non-emigrant households. A
study of Zachariah et al. (1999) compares emigrant and non-emigrant households in the 1990s
and finds that for an emigrant being abroad for a longer time, the quality of housing of the
emigrant household improves compared with non-emigrant households. As well, household
amenities for emigrant households are better than for non-emigrant households. For instance,
87% of emigrant households have electrified housing compared with 66% of non-emigrant
households. Besides, for emigrant household, the longer the duration of emigration by a
household member, the higher the possession of consumer durable goods by the emigrant
households compared with non-emigrant households. A striking figure is that in 1998 54% of
the emigrant households possess a television compared with 34% of non-emigrant households.
According to the end use of remittances more than a third part is used for education.
However, the authors find that emigrant households have slightly lower average number of
years of schooling than non-emigrants households (Zachariah, Mathew, Rajan, 1999). But the
children of migrants might have a higher educational level due to remittances. Another study
by Zachariah et al. (2003, p.235-236) shows that remittances have a significant positive effect
on the educational level of emigrant households and in particular on the higher levels of

15
education. Remittances seem to have a positive effect on educational levels, although we have
to take into account the selective nature of migration.
The inflow of remittances has a significant impact on the emigrant households with
implications for the economy of Kerala. One remark regarding this conclusion should be
made, because emigrant households could be richer than non-emigrant households before
migration and not as a consequence of migration and forthcoming remittances.
An important observation is that in 1998 in the state of Kerala the greatest part of the
households (82,9%) did not receive any remittances compared with the relatively small part
(17,1%) of households that did receive remittances. And the proportion of recipients (around
20%) and non-recipients (around 80%) stayed the same during the 2000s (Zachariah, Rajan,
2008). Although this large group of non-recipients may benefit indirectly, for instance
through multiplier effects with consequently positive effects on the economy. To the best of
my knowledge no study is done to estimate the multiplier effect of remittances in a state of
Kerala or nationwide India. Cross county experiences indicate that remittances may have
multiplier effects in Kerala. Some evidence can be found in a study by the Institute of
Development Studies in Bangladesh. The study shows that remittances had a multiplier effect
of 3.3% on GNP, 2.8% on consumption and 0.4% on investment. Furthermore, as Kuptsch
and Martin (2004) observe, a 1 US dollar in remittance spending can be doubled or tripled by
the local economic activity and especially if remittances are spent on locally produced goods.
The end use of remittances in Kerala indicates that the majority of the remittances are spent
on consumption which are regularly locally acquired goods and as such may benefit the
economy through the multiplier effect of remittances.

4.2 Evidence: Remittances and the development of the financial sector

The upsurge in remittances had an impact on the balance of payments and the current account.
Since the 1970s remittances contributed significantly to the balance of payments. Remittances
as a part of the private transfers continued to be around 3% form 1999-2000 on and have
offset Indias merchandise trading deficit to a large extent. Especially, from the 2002 onward
remittances, mainly for family maintenances, supported curtailing the deficits of the current
account (RBI, Bulletin, 2008, p. 409 and appendix 3).
Evidence on the national level about the impact of remittances on the banking sector is
not found. However some evidence is found on the regional level of the state Kerala. Eapen
(2003) shows that the lions share of the start-ups of small scale businesses is funded by other
sources than the commercial banking sector. A survey in the Palakkad district of Kerala

16
shows that only 13.5% of the start-up of small scale businesses is funded by commercial
banks. A study by Pushpangadan (2003) shows that the commercial banking sector did not
play a significant role in the growth of the services sector during the 1980s and 1990s in
Kerala. In this period the commercial banks could function as an intermediary between savers
and investors and increase credit in addition to the huge inflow of remittances. An indication
of this role by the commercial banking sector could be an increase in the issue of credit
compared with the issue of debt. However, in the period between 1988-1998 credits increased
with 13.2% and debt increased with 17.7%.

4.3 Evidence: Remittances as (a)cyclical source of income

Next to the positive effect on the balance of payments, are remittances a stable source of
finance for India? Jadhav (2003) finds interest rate differentials do no have a significant
impact on remittances in India. This implies that remittances are a more stable flow than other
financial flows, such as FDI flows. A HP filtered series for remittances by Gupta (2005)
shows that remittances in India are somewhat countercyclical. During drought periods, when
agricultural growth is negative, remittances are higher although not significant in all the
specifications. Sayan (2006) investigates remittances inflow over the business cycles in 12
developing countries during 1976-2003. The author finds that remittances increase during
economic bad times and as such work countercyclical in India. Although the inflow of
remittances lag by one year compared to the business cycles. This implies that in India
remittances are intended to smooth consumption in contrast with an investment motive when
remittances would work procyclical and increase in economic good times. The results show
that one year after an economic crisis in India an upsurge in the inflow of remittances occurs
to support migrant family members at a the microeconomic level. But also at the
macroeconomic level remittances counter macroeconomic shocks, such as high volatile
capital flows and a sharp fall in domestic savings. In chapter 5 I will check if remittances are a
stable source of income for the economy of India.

4.4 Evidence: Remittances and trade

Does the economy of India suffer from the Dutch Disease effects due to remittances?
According to Harilal and Joseph (2003) remittances have a significant and negative impact on
the tradable sector in Kerala. The boom in remittances from the 1970s had a dampening effect
on the goods producing sector in Kerala. The production shifted towards the non-tradable
sector and increased the demand in the non-tradable sector. As wages in this period increased

17
and prices of the non-tradable tended to move up compared to tradable prices in Kerala, the
competiveness for the tradable sector was more hampered than the non-tradable sector. The
inflation rate in Kerala was relatively higher compared with the national average of India,
which resulted in unfavourable real exchange rates in Kerala. The Dutch Disease effect is
present in Kerala. However, Harilal and Joseph state that overall remittances have a positive
effect on the economy of Kerala. Nevertheless, the non-tradable sector appeared to benefit
more than the tradable sector suffered from the dampen effects of the remittances boom in
Kerala.

4.5 Evidence: Remittances and the labour market

Remittances may work as a substitute for labour income and consequently reduce labour
market supply. Conversely, remittances may increase labour market supply because of
multiplier effects and positive spillovers from the increase in education of the household
members of emigrants. Evidence on the national level is not found. However, on the regional
level of the state Kerala some evidence is found. The economy of Kerala technologically
stagnated in the 1980s and beginning 1990s, but survived due to the expansion of the service
sector as a result of the inflow of remittances. A labour market mismatch between the demand
and supply of labour occurred, because the labour force is reasonably educated but a large
part is not employed (Kannan, 1998). A survey by Zachariah et al. (1999) shows that worker-
population ratio is 55% for non-emigrant households compared to 31.6% for emigrant
households. This figure may indicate that emigrant households may be more selective in the
selection of employment. As well the rate of unemployment for emigrant households is 20.8%
higher compared to 8.12% for the non-emigrant households. Again we have to be cautious in
the interpretation of these figures since they could be biased due to the selective nature of
migration. Nevertheless, both figures strengthen the argument that recipient households of
remittances have less incentive to work and reduce their labour market supply due to the fact
they are financially supported through remittances.
Besides the reduction in labour supply of recipients, remittances have a different effect
on the specific sectors in the Indian state of Kerala. An indication of the effect of remittances
on the labour market is the increase in demand for construction workers. The end use of
remittances reveals that recipients use a fair part of the received remittances for construction
and reparation of housing. From the 1970s a higher demand for especially house construction
and reparations by recipients of remittances increased the demand for construction workers
(Prakash, 1998, p. 3211).

18
Chapter 5: Empirical analysis of the cyclicality of remittances
In paragraph 4.3 I conclude that remittances in India seem to work countercyclical and that
they are used for consumption. By using time series I attempt to check if remittances are
countercyclical and compare the cyclicality of remittances with consumption and FDI. Also, I
will check the stability of remittances compared with consumption and FDI. The data is
gathered from the World Development Indicators & Global Development Finance database
provided by the World Bank. The data spans from 1970-2008. The results of the analysis are
provided in appendix 4.
In the first place, I will define the used indicators: remittances, FDI and consumption.
Remittances are defined as the sum of workers' remittances and compensation of employees.
They comprise current transfers by migrant workers and wages and salaries earned by non-
resident workers. FDI shows net inflows (new investment inflows less disinvestment) in the
Indian economy from foreign investors. Consumption is the sum of household final
consumption expenditure and general government final consumption expenditure. All
indicators in the dataset are given in US dollars.
In several steps I modified the data. Firstly, I deflate all the indicators with the US
GDP deflator with base year 2000 to transform the data in real terms. Subsequently I take logs
of the real indicators and after I apply the Hodrick Prescott filter (HP filter). The HP filter
separates the cyclical components from the trend. I use three different penalties on the
sensitivity for short run fluctuations from the trend, namely =400, =100 and =6,25. These
values are often used in the literature. Finally, I take the difference from the log indicator and
the HP filtered indicator. Through the fact that we take logs of the indicators we can
interpreted the deviations as percentage deviations from the trend. In figures 4a-4i the results
for these modifications are plotted comparing GDP with remittances, consumption and FDI.
Figures 4j-4l show the correlation coefficients between remittances, consumption and FDI,
compared to GDP. In figure 4m the standard deviations of the cyclical components of
remittances, consumption and FDI, relative to the standard deviation of the cyclical
component of GDP are shown.
Some evidence for the cyclical behaviour of remittances is shown in the figures 4a-4c.
The differences from the trend of remittances range between 30% and -20%. Compared with
GDP, in some periods remittances are countercyclical, such as the period 1970-1976 and
1986-1992 in figure 4a. But in other periods remittances seem to work procyclical, such as in
the periods 1976-1986 and 1994-2000. Also the level of the fluctuation matters. For instance,

19
in the period 1970-1976 the largest fluctuation is shown of 30% from the trend. Figure 4b and
4c show similar results, but with a higher intensity and a lower level of fluctuation from the
trend, because the fact that the lower value decreases the penalty on fluctuations from the
trend. The figures 4d-4f show that consumption clearly follows GDP, which implies that
consumption and GDP have practically the same cycle. Figures 4g-4i show that FDI
fluctuates enormously between 40% and -100% around the trend. Compared with GDP, the
figures indicate that FDI is procyclical in most of the period between 1970-2008.
Another indication of the cyclical behaviour of remittances, consumption and FDI is
to compute the correlation coefficients with different time lags between -2 and 2 years. In
figure 4j the correlation coefficients between remittances and GDP are shown with three
different values for the parameter : =400, =100 and =6,25. The three different parameters
show the same convex shape. This indicates that the results are robust to which smoothing
parameter is used. The figure ranges from -0,07 at a lag of 2 years and peaks at a correlation
coefficient of 0,22 without lag. The correlation between remittances and GDP is low. The
correlation between consumption and GDP is shown in figure 4k. Also this figure shows a
convex shape, but the correlation ranges from -0,2 at a lag of 2 years and a correlation
coefficient of 0,9 without lag. The correlation between remittances and GDP is considerably
lower compared with the correlation between consumption and GDP. This is an indication
that remittances are less procyclical than consumption. A high correlation between
remittances and consumption may suggest that remittances are used for consumption. Since
the correlation between consumption and GDP is nearly perfect, the correlation between
consumption and remittances is virtually the same as the correlation between GDP and
remittances. Because if this, figure 4j is comparable to the figure that shows the correlation
between consumption and remittances. It seems that the correlation between remittances and
consumption is low and this does not provide us with much evidence that remittances are used
for consumption.
I also compared the correlation coefficients between FDI and GDP and plotted them
in figure 4l. The figure shows a clear convex shape and ranges between -0,15 and 0,5. The
highest correlation of 0,5 is found if GDP lags by one half a year compared with FDI. An
increase in the inflow of FDI could predict that GDP will increase half a year later. The
correlation is no direct evidence for a causal relation between FDI and GDP. However, the
correlation could indicate a potential causal relation between FDI and GDP and provide some
evidence that FDI is somewhat procyclical. In fact, as the inflow of FDI increases, the figure

20
suggests that FDI could increase GDP half a year later on. This is an indication that FDI is a
forerunner for GDP and that FDI may enforce the cyclicality of GDP.
To find some evidence for the stability of remittances compared to consumption and
FDI figure 4m provides us some evidence. In figure 4m the standard deviations of the cyclical
components of remittances, consumption and FDI, relative to the standard deviation of the
cyclical components of GDP are shown. The figure shows that remittances relative to GDP
are less deviating (by a factor between 2-4) than FDI relative to GDP. But consumption
relative to GDP is less deviating (by a factor between 1-2) than remittances relative to GDP.
This provides evidence that remittances are more stable than FDI. As well some evidence is
found that consumption is more stable than remittances. However, the difference between the
consumption and remittances is fairly small.
I conclude that remittances have a higher stabilizing effect on GDP than FDI. And
remittances have a stabilizing effect on GDP close to the stabilizing effect of consumption. It
looks as if remittances smooth out cyclical fluctuations in GDP.

Chapter 6: Conclusion
This paper analyses the link between remittances and economic growth in India. From
theoretical point of view remittances may have a wide impact on the economy of India. Based
on the evidence in chapter four, remittances directly supplement the GNP of India. Besides,
remittances stabilized the balance of payments and supported the current account in India.
The results of the empirical analysis of the cyclical behaviour of remittances show that
remittances seem to have a stabilising effect on the economy of India. If we compare
remittances with FDI it seems that remittances are stabilizing the economy. And the
stabilizing effect of remittances on the economy is close to the stabilizing effect of
consumption. Nevertheless, the empirical results do not show that remittances smooth out
cyclical fluctuations in consumption. The results also show that FDI is a forerunner for GDP
and that FDI may enforce the cyclicality of GDP.
On the regional level of the state Kerala evidence is found that remittances have a
positive impact on emigrant households consumption, housing, amenities and education.
Particularly the relative large share of remittances towards educational purposes improves the
quality of labour with positive effects for the labour market and subsequently economic
growth. By contrast, remittances have a limited impact on business activities in Kerala and the
commercial banking sector did not play a significant role to supplement remittances with
credit. Non-recipients of remittances may benefit through multiplier effects. As lion share of

21
the remittances is used for local consumption of goods and as such may increase employment
as demand increases, but strong evidence is not found. Besides, the Dutch Disease effect is
present in Kerala. However, the overall welfare effect for the economy of Kerala is positive,
because the non-tradable sector appeared to benefit more than the tradable sector suffered
from the dampen effects of the remittances. The labour market of Kerala suffered from
remittances, since recipient households have less incentive to work and reduced their labour
market supply. As well, remittances seem to have a different effect on economic sectors in
Kerala. Or at least the construction sector benefited from the upsurge of remittances.
On the state level of Kerala I concluded that overall remittances seem to have a
positive effect on the economy. However, this conclusion is not drawn on the nationwide
level of India. From the empirical studies in the literature overview I concluded that it is not
found that remittances have a significant effect on economic growth in India. This could be
partly due to the complexity of the empirical research that no clear link is found or that the
negative effects cancel out the positive effects of remittances on economic growth.

22
Bibliography
Aggerwall, R., Demirg-Kunt, A., and Peria, M. S. M. (2005). Do Workers
Remittances Promote Financial Development? World Bank Research Paper
2006 No. 3975.
Acosta, P., (2006). Labor Supply, School Attendance, and Remittances from the
International Migration: The Case for Salvador. Washington, DC: The
World Bank.
Barajas, A. et al. (2009). Do Workers Remittances Promote Economic Growth?
Washington, DC: International Monetary Fund.
Catrinescu, N. et al. (2009) Remittances, Institutions, and Economic Growth. World
Development, 37, (1).
Chami, R. et al. (2008). Macroeconomic Consequences of Remittances. Washington,
DC: International Monetary Fund.
Chishti, M. (2007). The Rise in Remittances to India: A Closer Look. Washington
D.C.: Migration Policy Institute.
Desai M. A., D. Kapur and McHale J., (2001). Fiscal Impact of the Brain Drain: Indian
Emigration to the U.S. NBER-NCAER conference December 17/17 2001.
Docquier F. and Rapoport, H. (2005). The economics of Migrants Remittances.
Bonn: Institute for Study of Labor.
Eapen M. (2003). Rural industrialisation in Kerala: Re-Examining the issue of rural
growth linkages. Centre of Development Studies Working Paper No. 348.
Guiliano O. and Ruiz-Arranz M. (2006). Remittances, Financial Development and
Growth. Bonn: Institute for Study of Labor.
Gupta P. (2005). Macroeconomic Determinants of Remittances: Evidence from India
IMF Working Paper No. 224.
Hagen-Zanker J., Siegel M. (2007). The determinants of remittances: A review of the
literature. Maastricht Graduate School of Governance Working Paper Series
No. 003.
Harilal K. N., and Joseph K. J. (2003). Stagnation and Revival of Kerala Economy:
An Open Economy Perspective. Economic and Political Weekly, 38, (23), 2286-2294.
Jadhav, N. (2003). Maximising Developmental Benefits of Migrant Remittances: The
Indian Experience. Paper presented at the joint conference of DFID-
Worldbank, London.
Kannan K. P., Hari K. S. (2002). Keralas Gulf Connection Emigration, Remittances
and their Macroeconomic Impact 1972-2000. Centre for Development Studies,
Trivandrum Working Paper No. 328.
Kannan, K. P. (1998). Political Economy of Labour and Development in Kerala. Centre for
Development Studies Trivandrum.
Kuptsch, C., Martin, P. (2004). Migration and Development: Remittances and Cooperation
with the Diaspora. International Institute for Labour Studies.
Mallick, H. (2008). Do Remittances Impact the Economy? Some Empirical Evidence
from a Developing Country. Trivandrum: Centre of Developing Studies.
Naujoks D., (2009). Emigration, Immigration, and Diaspora Relations in India.
Washington D.C.: Migration Policy Institute.
Prakash B. A. (1998). Gulf migration and its economic impact: The Kerala experience.
Economic and Political Weekly.
Pushpangadan K. (2003). Remittances, consumption and economic growth in Kerala:
1980-2000. Centre for Development Studies Trivendrum Working Papers No.
343.

23
Ratha D., Xu Z. (2008). Migration and Remittances Factbook 2008. Washington, DC:
The World Bank.
Ratha, D. (2003). Workers Remittances: An important and stable source of external
Finance. Washington, DC: The World Bank.
Reserve Bank of India (2009). Invisibles in Indias Balance of Payments: An Analysis
of Trade in Services, Remittances and Income. Monthly Bulletin, 3.
Sasikumar S.K., Hussain Z. (2007) Migration, Remittances and Development:
Lessons from India. National Labour Institute.
Sayan S. (2006). Business Cycles and Workers Remittances: How Do Migrant
Workers Respond to Cyclical Movements of GDP at Home? IMF Working
Paper No. 52.
Solimano, A. (2003). Remittances by Emigrants Issues and Evidence. Oxford University
Press.
Zachariah K.C., Rajan S. I. (2008). Migration and Development: The Kerala
Experience. Daanish Publishers, New Delhi,
Zachariah, K.C., Mathew, E. T., and Rajan S. I. (1999). Impact of Migration on
Kerala's Economy and Society. Blackwell Publishing International Migration,
39, (1), 63-87.
Zachariah, K.C., Mathew E. T., and Rajan S. I. (1999). Migration in Kerala State,
India; Dimensions, Determinants and Consequences. Centre For Development
Studies Trivandrum Working Papers No. 1-2.

24
Appendix 1: Remittances, FDI and official development assistance

5,0%
4,5%
4,0%
3,5%
3,0%
% of GDP

2,5%
2,0%
1,5%
1,0%
0,5%
0,0%
-0,5%
70

72

74

76

78

80

82

84

86

88

90

92

94

96

98

00

02

04

06

08
19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

20

20

20

20

20
Year

Workers' remittances and compensation of employees, received (current US$, as % GDP )

Foreign direct investment, net inflows (current US$, as % GDP)
Net official development assistance received (current US$, as % GDP)

Source: Worldbank

25
Appendix 2: Remittances in India between 1970 - 2008

$40.000

$35.000
Remittances (US$ million)

$30.000

$25.000

$20.000

$15.000

$10.000

$5.000

$0
70
72
74
76

78
80
82
84
86
88

90
92
94
96
98
00

02
04
06
08
19
19
19
19
19
19

19
19
19
19
19
19

19
19
19
20
20
20
20

20
Year

Workers' remittances, compensation of employees, and migrant transfers, credit (in real terms, US$
million)

Source: Worldbank

26
Appendix 3: Data on macro economic key factors

Year GDP (real, in Inflation, CA balance Exports of goods Gross

US$ billion) consumer prices (BoP, real, in and services (% savings (%
(annual %) US$ billion) of GDP) of GDP)
1970 222,19 5,10 - 3,8 -
1971 229,18 3,08 - 3,7 -
1972 236,00 6,46 - 4,1 -
1973 267,90 16,93 - 4,3 -
1974 284,68 28,60 - 4,9 -
1975 255,46 5,74 -0,39 5,7 18,1
1976 251,74 -7,63 3,93 6,8 20,3
1977 280,37 8,32 4,96 6,5 19,5
1978 296,21 2,52 1,51 6,4 21,3
1979 304,76 6,25 0,10 6,8 21,3
1980 340,40 11,37 -3,31 6,2 17,5
1981 322,44 13,12 -4,57 6,0 21,1
1982 315,43 7,89 -4,03 6,1 20,4
1983 330,30 11,87 -2,97 5,9 18,5
1984 310,21 8,32 -3,42 6,4 20,7
1985 330,07 5,56 -5,94 5,3 21,8
1986 345,94 8,73 -6,41 5,3 21,7
1987 377,14 8,80 -7,07 5,7 21,0
1988 387,29 9,89 -9,44 6,1 22,1
1989 372,83 6,16 -8,67 7,1 22,5
1990 389,03 8,97 -8,62 7,1 22,3
1991 316,76 13,87 -5,08 8,6 21,8
1992 284,20 11,79 -5,19 8,9 23,2
1993 312,30 6,36 -2,12 9,9 21,8
1994 358,45 10,21 -1,86 10,0 24,5
1995 386,90 10,22 -6,04 11,0 26,6
1996 413,81 8,98 -6,35 10,5 22,9
1997 430,67 7,16 -3,11 10,8 25,2
1998 431,47 13,23 -7,16 11,2 22,6
1999 460,30 4,67 -3,30 11,7 26,0
2000 460,18 4,01 -4,60 13,2 25,1
2001 466,61 3,68 1,38 12,8 25,5
2002 486,74 4,39 6,77 14,5 26,7
2003 563,29 3,81 8,24 14,8 28,5
2004 640,24 3,77 0,71 18,1 31,9
2005 716,62 4,25 -9,10 19,9 34,2
2006 784,05 5,80 -7,97 22,2 35,8
2007 982,17 6,37 -9,42 21,2 37,8
2008 947,01 8,35 -29,48 22,7 38,0

Source: Worldbank

27
Appendix 4: Empirical analysis results

Cyclicality of Remittances (fig 4a)

0,4

0,3
% deviations from the trend

0,2

GDP (HP filtered, =400)

0,1

Remittances (HP filtered,

0,0
=400)
70
73
76
79
82
85
88
91
94

97
00
03
06
-0,1
19
20
20
19
19
19
19
19
19
19
19
19

20
-0,2

-0,3
Year

Cyclicality of Remittances (fig 4b)

0,3

0,25

0,2
% deviations from the trend

0,15

0,1
GDP (HP filtered, =100)
0,05
Remittances (HP filtered,
0
=100)
1970

1973

1976

1979

1982
1985

1988

1991

1994

1997

2000
2003

2006

-0,05

-0,1

-0,15

-0,2
Year

28
 Cyclicality of Remittances (fig 4c)

0,15

0,1
% deviations from the trend

0,05 GDP (HP filtered, =6,25)

Remittances (HP filtered,

0 =6,25)
1970
1973
1976
1979
1982
1985
1988
1991
1994
1997
2000
2003
2006
-0,05

-0,1
Year

Cyclicality of Consumption (fig 4d)

0,10

0,05
% deviations from the trend

0,00 GDP (HP filtered,

=400)
70
73

76
79

03
82

85

88

91

94

97

00

06

Consumption (HP
19

19

19

19

19
19

19

19

19

19
20

20

20

-0,05 filtered, =400)

-0,10

-0,15
Year

29
 Cyclicality of Consumption (fig 4e)

0,08

0,06

0,04
% devations from the trend

0,02

0 GDP (HP filtered,

=100)
-0,02
70

73

76

79

82

85

88

91

94

97

00

03

06
Consumption (HP

19
19

19

19

19

19

19

19

19

19

20

20

20
-0,04 filtered, =100)

-0,06

-0,08

-0,1

-0,12
Year

Cyclicality of Consumption (fig 4f)

0,06

0,04
% deviations from the trend

0,02

GDP (HP filtered,

0
=6,25)
1970
1973

1976
1979

1982
1985

1988
1991
1994
1997
2000
2003

2006

Consumption (HP
-0,02 filtered, =6,25)

-0,04

-0,06

-0,08
Year

30
 Cyclicality of FDI (fig 4g)

0,6

0,4
% deviations from the trend

0,2
GDP (HP
0
filtered, =400)
78

80

82

84

86

88

90

92

94

96

98

00

02

04

06

08
-0,2
19

19

19

19

19

19

19

19

19

19

19

20

20

20

20

20
FDI (HP
-0,4
filtered, =400)
-0,6

-0,8

-1

-1,2
Year

Cyclicality of FDI (fig 4h)

0,6

0,4

0,2
% deviations from the trend

GDP (HP filtered,

0
=100)
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008

-0,2 FDI (HP filtered,

=100)
-0,4

-0,6

-0,8

-1

-1,2
Year

31
 Cyclicality of FDI (fig 4i)

0,4

0,2
% deviations from the trend

0
1978
1980
1982
1984

1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006

2008
-0,2
GDP (HP filtered,
=6,25)
-0,4 FDI (HP filtered,
=6,25)
-0,6

-0,8

-1
Year

Correlation Remittances and GDP (fig 4j)

0,25

0,2

0,15
Correlation Coefficient

0,1 COR =400

COR =100
0,05 COR =6,25

0
REMt,Yt-2 REMt,Yt-1 REMt,Yt REMt,Yt+1 REMt,Yt+2
-0,05

-0,1
Lag

32
 Correlation Consumption and GDP (fig 4k)

0,8

0,6
Correlation Coefficient

0,4 COR =400

COR =100
0,2 COR =6,25

0
Ct,Yt-2 Ct,Yt-1 Ct,Yt Ct,Yt+1 Ct,Yt+2
-0,2

-0,4
Lag

Correlation FDI and GDP (fig 4l)

0,6

0,5

0,4
Correlation coefficient

0,3
COR =400
0,2 COR =100
COR =6,25
0,1

0
FDIt,Yt-2 FDIt,Yt-1 FDIt,Yt FDIt,Yt+1 FDIt,Yt+2
-0,1

-0,2
Lag

33
 Standard deviation relative to GDP (fig 4m)

12

10

8
=400
6 =100
=6,25

0
REM CON FDI

34