## A. B. Atkinson

Print publication date: 2004

Print ISBN-13: 9780199278558

Published to Oxford Scholarship Online: January 2005

DOI: 10.1093/0199278555.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 27 November 2020

# Remittances by Emigrants: Issues and Evidence

Chapter:
(p.177) 9 Remittances by Emigrants: Issues and Evidence
Source:
New Sources of Development Finance
Publisher:
Oxford University Press
DOI:10.1093/0199278555.003.0009

# Abstract and Keywords

Remittances from migrants are a growing force, and this chapter considers the role that they can play in financing development. To an important extent, they finance consumption, and are an international mechanism of social protection based on private transfers; they can also contribute to financing investment, providing community infrastructure and funds for the financing of new enterprises. The motives for making such remittances, and the problems of measuring their extent are considered, as are the variety of financial entities through which they are channelled, and policies for reducing the cost of remittances and enhancing their development potential. The chapter is organised in five main sections which: discuss global and regional trends in remittance flows and their growing importance as a source of external transfers to developing countries; examine measurement issues and discuss the main micro‐motives for remittances and the implications of their cyclical behaviour for stability; analyse the development impact of remittances (effects on savings, investment, growth, poverty, income distribution); overview the international market for remittances, and provide evidence on the costs of sending remittances to various country groups; and highlights policies for reducing the costs of sending remittances and thus enhancing their developmental impact.

# 9.1 INTRODUCTION

Remittances from migrants are a growing and relatively stable, market-based external source of development finance. Remittances bring foreign exchange, are a complement for national savings, and provide a source of finance for capital formation (mainly small-scale projects). Through these mechanisms, remittances can support economic growth in recipient countries. As remittances depend on flows of people that are often less volatile than capital flows, remittances are expected to be more stable than such capital flows as portfolio investment and international bank credit. Remittances are also an international redistribution from low-income migrants to their families in the home country. These transfers act as the international mechanism of social protection based on private transfers. The sustainability of remittances over time depends on various factors such as the anticipated flow of migration, and whether the migrants come alone or with their family, and how this changes over time.1

It is also important to recognize that benefits from remittances for the receiving countries have to be compared with the potential costs of emigration for the developing countries in terms of the loss of scarce human skills (the so-called brain drain phenomenon). Thus, a certain tradeoff is generated between the inflow of foreign exchange and external savings through remittances and the outflow of skilled individuals.2

Currently, remittances—after foreign direct investment—are the second most important source of external finance for developing countries. Moreover, they surpass foreign aid. There are twenty countries that are the main recipients of remittances. These twenty low- to medium-income developing economies capture around 80 per cent of total worker remittances to the developing world. In terms of value, the (p.178) three main recipient countries are India, Mexico, and the Philippines, while the three main source countries are the United States, Saudi Arabia, and Germany.

The international market for remittances (from a social point of view) is segmented and inefficient, as is reflected by the high costs of intermediation. Money transmitter operators dominating the market charge high fees and use overvalued exchange rates. Commercial banks in both the source and recipient countries have a low share of the global remittances market. Empirical evidence, however, shows that the costs of remittances are lower when sent through banks than through money transfer operators (MTOs).

There is, however, room for leveraging a greater value for remittances if international money transfers were conducted at lower costs. The amount of remittances is below the socially optimal level associated with a more competitive cost structure in the market for remittances (causing, therefore, a deadweight loss for both the sender and the receiver of a remittance). The development potential of remittances is thus diminished under current market realities.

The chapter is organized in seven sections in addition to the introduction. Section 9.2 discusses global and regional trends in remittance flows and their growing importance as a source of external transfers to developing countries. Section 9.3 examines measurement issues and discusses the main micro-motives for remittances and their implications for stability across cycles, while Section 9.4 analyses the development impact of remittances (effects on savings, investment, growth, poverty, income distribution). Section 9.5 overviews the international market for remittances and provides evidence on the costs of sending remittances to various country groups. Section 9.6 highlights policies for reducing the costs of sending remittances and thus enhancing their developmental impact. Section 9.7 concludes.

# 9.2 GLOBAL AND REGIONAL TRENDS IN REMITTANCE FLOWS

In a world of volatile capital flows, remittances3 are a stabilizing component of external resources transfers to the developing world. Remittances are the financial counterpart of the outflow of people, and migration flows have been growing in the last two decades in response to expanding opportunities in advanced economies compared to developing countries. Remittances to the developing world have increased steadily from around US$15 billion in 1980 to 80 billion in 2002. This represents an annual rate of increase of 7.7 per cent (see Table 9.1).4 At the regional level, the highest rate of increase in the flow of remittances is to Latin American and the Caribbean with 12.4 per cent per annum, followed by East Asia and the Pacific with 11 per cent per year. The lowest annual growth rate in remittances is to sub-Saharan Africa with 5.2 per cent. As shown in Table 9.1, in 2002 Latin America and the Caribbean have (p.179) (p.180) the highest level of remittances, totalling US$25 billion, followed by South Asia with 16 billion, the Middle East and North Africa (MENA) with 14 billion, and East Asia and the Pacific with 11 billion. Sub-Saharan Africa has the lowest level of remittances, US$4 billion. Table 9.1 Remittances received by region, 1980–2002 (in billions of US$)

Region

1980

1985

1990

1995

1996

1997

1998

1999

2000

2001

2002 (est.)

Annual rate of growth (%) 1980–2002

East Asia and Pacific

1.1

2.3

3.6

8.3

9.5

14.2

8.3

10.6

10.3

10.4

11.0

11.0

Share (%) of remittances in developing countries

7.1

12.7

12.4

17.3

18.1

22.6

13.9

16.4

15.9

14.4

13.8

Europe and Central Asia

2.1

1.7

3.2

5.5

6.2

7.1

9.2

8.1

8.7

8.9

10.0

7.4

Share (%) of remittances in developing countries

13.5

9.4

11.0

11.5

11.8

11.3

15.5

12.5

13.5

12.3

12.5

Latin America and the Caribbean

1.9

2.6

5.7

12.8

12.8

13.6

14.8

16.9

19.2

22.6

25.0

12.4

Share (%) of remittances in developing countries

12.3

14.4

19.6

26.7

24.3

21.7

24.9

26.1

29.7

31.3

31.3

Middle East and North Africa

3.8

4.6

9.3

8.6

9.1

9.4

10.3

10.5

10.9

13.1

14.0

6.1

Share (%) of remittances in developing countries

24.5

25.4

32.0

18.0

17.3

15.0

17.3

16.2

16.9

18.1

17.5

South Asia

5.3

5.8

5.6

10.0

12.3

14.6

13.3

15.1

13.5

14.9

16.0

5.2

Share (%) of remittances in developing countries

34.2

32.0

19.2

20.9

23.4

23.3

22.4

23.3

20.9

20.6

20.0

Sub-Saharan Africa

1.3

1.1

1.7

2.7

2.7

3.8

3.6

3.5

2.0

2.4

4.0

5.2

Share (%) of remittances in developing countries

8.4

6.1

5.8

5.6

5.1

6.1

6.1

5.4

3.1

3.3

5.0

Developing countries

15.5

18.1

29.1

47.9

52.6

62.7

59.5

64.7

64.6

72.3

80.0

7.7

Industrial countries

n.a.

n.a.

n.a.

37.2

35.7

40.5

41.0

40.2

40.1

39.3

n.a.

n.a.

All countries

n.a.

n.a.

n.a.

85.1

88.3

103.2

100.5

104.9

104.7

111.6

n.a.

n.a.

Note: Remittances are calculated as the sum of workers remittances and compensation of employees; n.a. means not available.

Source: Source: IMF (2003).

In terms of the distribution of remittances by levels of per capita income, the developing-country group received 65 per cent of world remittances. In turn, the lower middle-income and low-income groups received a higher proportion than the upper middle-income countries (Table 9.2).

Table 9.2 Remittancesa received by country groups, 1995–2001 (in billions of US$) Countries 1995 1996 1997 1998 1999 2000 2001 Upper middle income 13.7 13.6 14.3 16.3 15.7 16.6 17.2 Share of remittances in all countries 16.1 15.4 13.8 16.2 15.0 15.9 15.4 Lower middle income 20.7 21.2 24.2 24.1 27.2 28.3 30.0 > Share of remittances in all countries 24.3 24.0 23.5 24.0 26.0 27.0 26.9 Low income 13.5 17.8 24.2 19.1 21.8 19.7 25.1 > Share of remittances in all countries 15.9 20.2 23.5 19.0 20.8 18.8 22.5 All developing 47.9 52.6 62.7 59.5 64.7 64.6 72.3 > Share of remittances in all countries 56.3 59.6 60.7 59.2 61.7 61.7 64.8 Industrial countries 37.2 35.7 40.5 41.0 40.2 40.1 39.3 > Share of remittances in all countries 43.7 40.4 39.3 40.8 38.3 38.3 35.2 All countries 85.1 88.3 103.2 100.5 104.9 104.7 111.6 > Share of remittances in all countries 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Note (a) Remittances are calculated as the sum of workers' remittances and compensation of employees. Source IMF (2003). In 2002, for the developing-country group, worker remittances represented on average 1.3 per cent of GDP, 55.9 per cent of FDI flows, and nearly 140 per cent of the aid flows (Table 9.3). These coefficients vary from region to region. The proportion of worker remittances in GDP is the highest in the MENA region (3 per cent in 2002) and the lowest in the East Asia and Pacific region (0.7 per cent). Remittances as a proportion of FDI are the highest in the MENA region (466.7 per cent in 2002) and the lowest in East Asia and Pacific (19.3 per cent). In turn, the proportion of remittances in foreign aid is the lowest in sub-Saharan Africa, reflecting both lower remittances and high aid flows to this region. Table 9.3 Remittancesa received by developing countries, 1996–2002 Countries 1996 1997 1998 1999 2000 2001 2002 (est.) East Asia and Pacific as % of GDP 1.0 1.3 0.7 0.8 0.7 0.7 0.7 as % of FDI inflows 16.2 22.8 14.4 21.7 23.4 21.3 19.3 as % of aid flows 125.0 215.2 103.8 112.8 128.8 152.9 n.a. Europe and Central Asia as % of GDP 1.4 1.3 1.4 1.1 1.0 0.9 1.0 as % of FDI inflows 38.0 32.6 35.4 28.6 29.8 29.6 34.5 as % of aid flows 89.9 126.8 131.4 84.4 90.6 97.8 n.a. Latin America and the Caribbean as % of GDP 1.3 1.2 1.1 1.1 1.2 1.3 1.4 as % of FDI inflows 28.8 20.6 20.2 19.2 25.3 32.6 59.5 as % of aid flows 232.7 302.2 328.9 359.6 505.3 434.6 n.a. Middle East and North Africa as % of GDP 3.4 3.0 3.1 2.9 2.8 3.0 3.0 as % of FDI inflows 1,300.0 151.6 137.3 328.1 436.0 238.2 466.7 as % of aid flows 171.7 195.8 219.1 244.2 294.6 335.9 n.a. South Asia as % of GDP 3.7 3.8 3.1 3.2 2.6 2.6 2.6 as % of FDI inflows 351.4 298.0 380.0 487.1 435.5 363.4 320.0 as % of aid flows 236.5 339.5 271.4 351.2 321.4 252.5 n.a. Sub-Saharan Africa as % of GDP 1.4 1.7 1.4 1.3 0.7 0.7 1.1 as % of FDI inflows 62.8 46.9 55.4 43.2 32.8 17.4 57.1 as % of aid flows 18.0 28.6 27.1 28.7 16.4 18.9 n.a. Developing countries as % of GDP 1.6 1.7 1.4 1.4 1.3 1.3 1.3 as % of FDI inflows 41.2 37.0 34.1 36.1 40.2 42.1 55.9 as % of aid flows 101.3 134.5 118.3 123.5 127.9 139.0 n.a. Note: (a) Remittances are calculated as the sum of workers' remittances and compensation of employees; FDI is foreign direct investment; Aid flows are official development assistance; n.a. means not available. Source: IMF (2003). In terms of total resource flows, remittances are the second largest component of external resource flows to developing countries after FDI (Table 9.4 and Fig. 9.1). Remittances have been larger than aid flows as a source of external development finance since 1997. In 2001, foreign aid represented 18 per cent of total external finance flows while remittances were 25 per cent. Interestingly, as mentioned earlier, remittances are much more stable than other capital flows. Mainly bank credit and portfolio investment are considered volatile components of external resource flows. (p.181) The quantitative importance of these components of private capital flows is still significant (nearly 30 per cent of total resource flows, on average, to developing countries between 1991 and 2000). These components are an important source of macroeconomic volatility. Often private capital flows do lead the macroeconomic cycles. In contrast, remittances can be even counter-cyclical, as emigrants send money home during bad times to provide income support. Table 9.4 Resource flows to developing countries, 1991–2002 (current US$ billions and %)

Remittancesa

Aid flowsb

Other official flowsc

FDI

Other private flowsd

Total

US$% US$

%

US$% US$

%

US$% US$

%

1991

33.1

21

49.5

32

11.4

7

35.7

23

26.3

17

156

100

1992

37.2

19

46.4

24

10.1

5

47.1

24

52.2

27

193

100

1993

38.9

15

41.7

16

11.9

5

66.6

26

100.2

39

259.3

100

1994

44.1

16

48.1

18

−0.1

0

90.0

34

85.6

32

267.7

100

1995

47.9

15

61.0

19

8.9

3

105.0

33

99.1

31

322.3

100

1996

52.6

14

51.9

14

−7.8

−2

128.0

34

148.44

40

372.9

100

1997

62.7

15

46.6

11

7.2

2

169.0

41

131.37

31

417.2

100

1998

59.5

15

50.3

12

16.2

4

175.0

43

108.75

27

409.3

100

1999

64.7

19

52.4

15

5.0

1

179.0

52

45.09

13

346.6

100

2000

64.6

19

50.5

15

−3.0

−1

161.0

48

65.15

19

338.0

100

2001

72.3

25

52.0

18

n.a.

n.a.

172.0

60

−11.73

−4

284.3

100

2002 (est.)

80.0

36

n.a.

n.a.

n.a.

n.a.

143.0

64

n.a.

n.a.

223.0

100

Average: 1991–2001

52.51

18

50.04

18

5.98

2

120.75

38

77.32

25

306.04

100

Note (a) Remittances are calculated as the sum of workers' remittances and compensation of employees;

(b) Aid flows are official development assistance and official aid;

(c) Other official flows are total official flows (official development fin.a.nce), net of aid flows;

(d) Other private flows are portfolio flows, and bank and trade; n.a. not available.

Source: IMF (2003) for remittances; World Bank (2003) for all other flows.

Figure 9.1. Long-term resource flows to developing countries, 1991–2001

Source IMF (2003).

(p.182)

(p.183) At the individual country level, remittances are relatively concentrated in the group of twenty developing countries that capture around 80 per cent of total remittances to the developing world (Fig. 9.2). In turn, the GDP of these twenty countries represents approximately 60 per cent of the GDP of the developing world (World Bank 2003). In 2001, the main recipient of worker remittances was India, receiving an annual flow of US$10 billion, followed by Mexico with 9.9 billion and the Philippines with 6.4 billion. At the lower end of this group of twenty developing-country recipients of worker remittances are Thailand, China, and Sri Lanka. The country ranking, however, changes, when remittances are measured as shares of GDP, on which measure the top three economies are Tonga, Lesotho, and Jordan with remittances ranging between 20 and 40 per cent of GDP. At the lower end are the Philippines, Uganda, Ecuador and Sri Lanka, with shares between 7 and 9 per cent of GDP (Fig. 9.3). Figure 9.2. Top twenty developing-country recipients of workers' remittances, 2001 (in billions of dollars) source World Bank (2003). Figure 9.3. Top twenty developing-country recipients of workers' remittances, 2001 (as percentage of GDP) Source World Bank (2003). On the other side, the top twenty source countries of remittances in 2001 are headed by the United States with US$ 28.4 billion, followed by Saudi Arabia with 15.1 billion and Germany with 8.2 billion (see Fig. 9.4). At the lower end of the top twenty sending countries are Czech Republic, Venezuela, and Norway (all three with US$0.7 billion in 2001). Figure 9.4. Top twenty country sources of remittance payments, 2001 (p.184) (p.185) Next we turn to the motives for remittances that may shed some light on the empirical behaviour of remittances reviewed in this section. # 9.3 MEASUREMENT, MICRO-MOTIVES FOR REMITTANCES AND CYCLICAL BEHAVIOUR In this section, we review (i) measurement issues; (ii) the micro-motives for remittances, and (iii) the stability of remittances during the cycle.5 ## 9.3.1 Definition and Measurement Issues The economic significance of remittances often goes beyond what is suggested by the official balance-of-payments statistics in the sending and receiving countries. The important concept for measuring the economic impact of remittances is the resource transfer—monetary or in-kind—made by a migrant to his home country. Monetary transfers in dollars directly increase the availability of foreign exchange in the migrant's country of origin, whereas remittances in-kind save foreign exchange for the recipient country. This distinction is important, as there are several modalities for sending remittances. Some of these are recorded while others are not. For example, when remittances are sent through the formal channels, they are recorded in the receiving country's balance-of-payments current account. Conversely, remittances sent informally in cash, for example through couriers, are unrecorded in official statistics. Remittances can be in-kind, for example, goods sent to households in the home country. Only part of the later are recorded as imports. Migrants may also make donations in the host country to institutions like the church or other charitable organizations formed by co-nationals. They can also make numerous payments (insurance premiums, school tuition, international airfares paid directly to airlines) on behalf of relatives or friends from their home country.6 Although most of these payments should, in the economic sense, be treated as ‘remittances’, they rarely are recorded as such. In sum, these considerations should be borne in mind in assessing the true magnitude of remittance transfers based on official statistics, which as noted above, tend to underestimate their full economic impact. In general, data on remittances are available from three items in balance-of-payments reports at country level:7 (i) worker remittances (money sent by workers living abroad for more than one year); (ii) compensation of employees (gross earnings of foreigners residing abroad for less than a year; and (iii) migrant transfer (net worth of migrants moving from one country to another) (see Gammeltoft 2002). ## (p.186) 9.3.2 Microeconomic Motivations to Remit The analytical literature8 on motives for remittances can be summarized in four approaches. ### The altruistic motive Under the altruistic view, the migrant sends remittances home because he cares about the well-being of his or her family in the home country, and a remittance satisfies the emigrant's concern for the welfare of his family. Furthermore, it is an empirical regularity that the migrant generally has a higher education level than family members who stay at home. When a migrant goes to a country where the average wage and per capita income are higher than at home, his income level, once he secures a job, can be expected to be better than that of comparable workers at home. The main prediction of the altruistic motive is that remittances tend to decrease over time,9 as the attachment to family gradually weakens over time when members are in different countries. Furthermore, the migrant may plan to stay abroad for an extended period (or eventually retire there), subsequently bringing his family to his adopted country. This, of course, reduces remittances. The converse case is the return-migration in which the migrant brings fresh funds on his return home, raising remittances the one time. ### The self-interest motive Opposite to the altruistic motive is the emigrant who sends remittances to the home country mainly for economic reasons and financial self-interest. In this scenario, the successful emigrant saves money in a foreign country, creating the dilemma as to how to accumulate wealth (in which assets) and where (in which country). An obvious place to invest at least part of the assets is in the home country, buying property, land, financial assets, etc., where these assets may earn a higher rate of return than in the host country, albeit with a greater risk profile. These assets can be administered on behalf of the migrant by the family, who acts as a trusted agent. Expectations of an inheritance from the emigrant's parents may be another motivation for remittances. In this case, family members who have contributed to the increasing family wealth (e.g., by sending remittances) become the obvious candidates of future inheritance. ### Implicit family contract 1: loan repayment Economic theory has developed explanations of the remittances process, which take the family—rather than the individual—as the main unit of analysis.10 The theory assumes that a family develops an implicit contract with the individual (the migrant) who chooses to live abroad, and those who stay at home. The implicit contract has an intertemporal dimension, say various years or even decades, as the time-horizon, and combines elements of investment and repayment. In the loan repayment theory, (p.187) the family invests in the education of the emigrant and usually finances the migration costs (travel and subsistence in the host country). This is the loan (investment) element of the theory. The repayment part comes after the migrant settles abroad, his income profile starts to rise over time and he is in a position to start repaying the loan (principal and interests) back to the family in the form of remittances. Thus the family invests in a higher yielding ‘asset’—the migrant—who earns a higher income in a foreign country than other family members living and working at home. This model predicts various time profiles of remittances, depending on the length of time it takes for the migrant to become established in the foreign labour market and on the duration of his stay abroad. The quicker the migrant's integration into the labour market of the new country, the faster the flow of remittances. Amounts to be remitted will depend, among other things, on the income profile of the migrant. In this model, remittances do not need to decrease over time as they do in the altruistic model. ### Implicit family contract 2: co-insurance Another variant of the theory of remittances as an implicit family contract between the migrant and those at home is based on the notion of risk diversification. The idea is simple: insurance markets and capital markets in the real world are incomplete, and risks cannot be diversified because of the absence of financial assets that edge risk. In addition, constraints to borrowing, a particularly serious problem for poor migrants, limit the ability to smooth consumption or finance investment. Assuming that economic risks between the sending and foreign country are not positively correlated, then it becomes a convenient strategy for the family as a whole to send abroad some of its members (often the most educated) to diversify economic risks. The migrant, then, can help to support his family in bad times at home. Conversely, for the migrant, having a family in the home country is insurance against bad times that may also occur in the adopted country. Here, emigration becomes a co-insurance strategy, with remittance playing the role of an insurance claim. As in any contract, there is the potential problem of enforcement (e.g. ensuring that the terms of the contract are respected by all parties). However, in principle, enforcement can be expected to be simpler due to the fact that these are implicit family contracts, which are helped by family trust and altruism (a feature often absent in legally sanctioned contracts). ## 9.3.3 Stability of Remittances in the Economic Cycle As mentioned in the previous section, worker remittances are more stable than portfolio investments and bank credit. Remittances can even be counter-cyclical. The different motives to remit reviewed above can shed some light in explaining this behaviour. In the model of remittances based on altruism, the migrant can increase his remittances when there is an economic downturn in the home country (as income of the migrant's family declines). In this case, a remittance would be the equivalent of a private ‘welfare payment’ sent from abroad to help smooth the consumption of the recipient at home. However, business cycles may be internationally synchronized. The growing economic (p.188) interdependencies of globalization make this a more plausible case. In this situation, a recession in the receiving country may be positively correlated with a recession in the source country, so that the ability of the immigrant worker to send remittances may be hampered by economic conditions in the host country. This is a real possibility, although the sender may also draw on existing savings to maintain a steady flow of remittances. If remittances were driven by the portfolio decisions of the migrant (remittances driven by investment), again the relevant issue would be the correlation between the rate of return of the assets in the host country and the rate of return on the assets at home. Here international correlation of the business cycle matters, as does the degree of financial integration between the source and the receiving country. In the model of remittances as mechanisms of co-insurance, risk diversification may call for a steady flow of remittances if business cycles are not fully positively correlated between the source and the receiving country. # 9.4 THE DEVELOPMENT IMPACT OF REMITTANCES Remittances have a potential positive impact as a development tool for the recipient countries. The development effects of remittances can be decomposed into its impact on savings, investment, growth, consumption, and poverty and income distribution. Remittances' impact on growth in the receiving economies is likely to act through savings and investment as well as short-run effects on the aggregate demand and output through consumption. Also the indirect effect of migration on output depends on the productivity level of the emigrant in the home country before departure. The total saving effect of remittances comes from the sum of foreign savings and domestic savings effects. Worker remittances are a component of foreign savings and they complement national savings by increasing the total pool of resources available for investment. Part of the savings effects of remittances takes place in the ‘community’. In fact, migrant associations, often called hometown associations (HTAs) in the United States, organize migrants from various Latin American countries such as El Salvador, Guatemala, Honduras, Mexico, and the Dominican Republic. HTAs regularly send donations to finance investment in community projects and local development in the home countries.11 Migrants' associations of former El Salvadorians send home donations of about US$10,000 per year, while those of Mexicans send home between US$5,000–25,000 per year (see Ellerman 2003). These are small numbers but in the recipient countries these sums can still have an impact. In the Mexican state of Zacatecas, the federal and local government match every dollar (it may be a two-for-one or three-for-one) donated by HTAs to local projects oriented to small infrastructure projects on water treatment, schools, roads, parks, etc. Through this programme, more than 400 projects (p.189) in Zacatecas have been completed in eight years. Total investment made by migrants to these projects amounts to around US$4.5 million (World Bank 2002). In this way, public savings are mobilized to finance small community projects, along with the remittances.

The previous discussion suggests that the direct effects of remittances on investment are bound to be on small community projects. Ratha (2003) cites instances of positive effects on investment in such receiving countries as Mexico, Egypt, and sub-Saharan Africa, where remittances have financed the building of schools, clinics, and other infrastructure. In addition, return-migrants bring fresh capital that can help finance investment projects.

Remittances also finance consumption; thus, private savings will increase proportionally less than an increase in income from external remittances. A study of remittances for Ecuador (Bendixen and Associates 2003) shows that around 60 per cent of remittances to that country are spent on food, medicine, housing rent, and other basic commodities; less than 5 per cent of remittances are used for the acquisition of residential property.

The combined effects of remittances on investment and consumption can increase output and growth. The sustainability of this effect is an open discussion. If remittances are a response to recent migration, remittances may be transitory and thus their effect on investment, consumption and growth can be more of a temporary nature. In contrast, if migrants form associations and their commitment to their home country becomes ‘institutionalized’, then the positive developmental effects of their remittances may become more permanent.

The indirect growth effect of remittances on growth (or output) depends on the type of emigrant leaving home, the state of the labour market, and the productivity of the emigrant. If the emigrant is unskilled with low productivity, or an unemployed worker, reflecting slack and excess supply in the labour market, then the effect of emigration on output in the home country is bound to be small. In contrast, if the emigrant is a highly skilled worker, an information technology expert or an entrepreneur with a high direct and indirect contribution to output, the adverse growth effect of high-skilled emigration is bound to be large (see Solimano 2001, 2002a).

One negative effect of (substantial) remittances is the possibility that they produce the so-called ‘Dutch disease’ effect.12 In countries receiving substantial amounts in remittances, there is a tendency for the real exchange rate to appreciate which then penalizes non-traditional exports and hampers the development of the tradable goods sector.

Remittances may also have a poverty reducing and income distribution effect . As mentioned above, the recipient of remittances is often a low-income family whose offspring has left the country to work abroad. In a way, emigration is an effort to escape poverty (p.190) at home13 and to improve the income-earning capacity of the emigrant by attempting to enter a foreign labour market in a richer country. At the same time, remittances serve to alleviate the poverty of the migrant's family in the home country by supplementing its income through transfers. The negative side of this is that remittances may create a ‘culture of dependence’ on the income from remittances. This, in turn, can impair efforts to escape poverty through education and work. The distributive effect of remittances is another dimension of the development effects of remittances.14 Stark (1991) studies the effects of remittances on domestic inequality in two Mexican villages near the US border in which villagers engage in internal rural–urban migration as well as in migration to the United States. The study finds that remittances from internal migrants are correlated more with the years of schooling than remittances from international migrants to the United States, as the later often go to low-skilled labour-intensive jobs. Stark (1991) generalizes that the inequality impact of changes in remittances depends on the recipients' position in the village income distribution scale, the share of remittances in the village incomes and the distribution of the remittances themselves. These variables in turn depend on the distribution of human capital (education and skills) among the villagers and the migration opportunities of the villages. Another piece of evidence is provided by Ratha (2003) who reports that a household data survey for Pakistan shows that the share of income originating from external transfers increases with the income level (the households with the highest incomes receive the largest shares of their income from remittances). So remittances might appear to be increasing local inequality However, income distribution between countries may eventually improve with remittances, as income is redistributed from source countries with a higher income level to receiving countries with a lower income per capita. As seen in Section 9.2, remittances represent a very significant share of GDP in several low-income countries.

A final remark here: the development effect of remittances depends on the ‘life-cycle’ of the whole migration process at the country level. In fact, in instances where the source countries have expanding economies with rising per capita incomes, the differentials across countries in the income per head will diminish, reducing the incentives for emigration. Thus, the relative importance of remittances is likely to decline as a country moves up the development ladder. This is valid mainly for remittances from low-skilled migrants, however. In the case of highly skilled well-educated individuals, migration flows are likely to continue at the high per capita income levels, as has been seen within the European Union or between Europe and the United States. In this case, remittances may continue although their economic effects are probably quite different than those discussed earlier when the recipients of remittances are the developing countries.

# (p.191) 9.5 THE INTERNATIONAL MARKETS FOR REMITTANCES

Remittances are channelled through financial entities such as MTOs, post offices, travel agencies, couriers, informal financial institutions, etc. MTOs owned and run by immigrants (or naturalized citizens of the same ethnic or national group) are denominated as ‘ethnic stores’. Commercial banks are also in the remittances business, but are generally not important players (see Table 9.5). These financial intermediaries often charge fees for money transfers well above their marginal costs (see Orozco 2003). The most important MTO at the global level is Western Union with branches in many countries, followed by MoneyGram and Thomas Cook. The less competitive, more concentrated and more segmented the market for remittances, the higher the costs of the remittances. There are a number of reasons why the international market for remittances tends to be a thin and poorly competitive (only few players dominate the market and costs of intermediation are high). First, the legal status of the migrant sending the remittance is not always regularized. Some migrants have resident (working) visas, others are waiting for their visas to be processed and others are simply ‘illegal’. Commercial banks are reluctant to enter the financial services market for low-income migrants whose immigration status often is not regularized.15 The result is a less competitive market, where furthermore migrants are not well integrated as customers in the formal banking circuits. Second, it is important to note that worker remittances are small-scale transactions. In Latin America, the typical remittance per migrant is in the range of US$200–300 per month.16 As individual transactions (remittances) are small, service standardization is needed for the remittances market to become a profitable activity at competitive fees. In this context, high fees may compensate for the cost of small transactions.17 Finally, other factors that affect the market for remittances include exchange rate risk, government regulations for foreign exchange transactions in the receiving country and regulations in the sending country such as licensing costs. Table 9.5 Sources of remittances: countries and transfer agents Receiving country Remittances sent from Number of companies reviewed All businesses Banks MTOsa Other Philippines United States 5 14 5 24 Egypt United States 2 2 Greece Germany and USA 4 2 6 India Saudi Arabia, USA, UK 7 11 18 Pakistan Saudi Arabia, USA, UK 7 1 8 Portugal France, USA 3 2 5 Turkey Germany, USA 3 2 5 Mozambique South Africa, USA 1 1 Zimbabwe South Africa, USA 7 7 Bangladesh UK 1 3 4 Ghana UK 7 7 Note (a) Money transfer operators. Source Orozco (2003). ### Costs of remittances Let us turn now to the efficiency of the market for remittances to the Andean region. If the costs of remittances are above the marginal cost (including a normal return to capital) of sending money, then the amount of the remittances is below the socially optimal level. As a consequence, consumption, investment, and output opportunities foregone in the receiving country cannot be realized. (p.192) The work by Orozco (2001, 2002) highlights two main cost components of sending remittances: $Display mathematics$ Companies charge a (explicit) fee that can be a percentage of the amount remitted or a fixed amount (often in dollars). The fee usually depends on the services offered (speed and type of delivery, etc.). The exchange rate spread is the difference between the exchange rate applied by the MTO to convert dollars into local currency and the market (e.g. interbank) exchange rate. MTOs usually offer the sender a less favourable exchange rate than the market rate. This is an additional source of profits for companies transmitting money and an additional cost component for the user. The average cost of sending a remittance of US$200 through a commercial bank to selected non-Latin American countries is 7 per cent compared to the 12 per cent charged by such MTOs as Western Union and MoneyGram (Table 9.6).18 Clearly, sending money through the bank is less expensive than sending it through the MTOs. Banks also offer a variety of money transfer services and charges decline substantially when the remittance is deposited with the same bank at both source and destination (p.193) countries. Foreign exchange spreads represent around 14 per cent of the total costs of remittances to non-Latin American countries. However, the country averages mask significant cross-country differences in the costs of sending remittances. For example, according to Table 9.7 drawn from Orozco (2003), the cost of sending money through banks is the lowest for Pakistan and the highest for the Philippines. These costs are much more uniform but also higher when money is sent through the major MTOs (in the range of 9.5–13.8 per cent).

Table 9.6 Average costs of sending money to selected non-Latin American countries

Type

For a remittance of US$200 FX % Fee % Total % Bank 1.0 6.5 7.0 Major MTO 1.7 10.9 12.0 Source: Orozco (2003). Table 9.7 Charges by types of operators for sending a remittance of US$200 to selected countries (%)

Country

Bank

Ethnic store/exchange house

Major MTO

Egypt

13.8

Philippines

8.0

10.1

10.3

India

6.0

2.5

13.8

Greece

6.8

9.5

Pakistan

0.4

3.0

13.0

Portugal

3.4

12.3

Turkey

3.1

9.5

Mozambique

1.0

Mean

7.0

6.0

12.0

Source: Orozco (2003).

The cost of sending money from the United States to Latin America is in the range of 8–9 per cent (see Table 9.8). Interestingly, as a share of total costs, the component of exchange rate spreads is twice as high for remittances to Latin American than to non-Latin American destinations. In fact, the exchange rate spread component for the latter is around 14 per cent of total costs of sending a remittance while it is nearly 28 per cent for Latin American recipients. Finally, let us look at the costs of remittances for the Andean region. Table 9.9 provides the average cost or charge of sending remittances of US$200, 250, and 300 to Bolivia, Colombia, Ecuador, Peru, and Venezuela. The (p.194) (p.195) data are based on a survey conducted in January 2003 of MTOs and ethnic stores in the United States that are engaged in the remittances industry with these countries.19 Table 9.9 gives the costs of a money transfer to be delivered in dollars and in local currency. The percentage charges are systematically lower across countries for dollar remittances than local currency remittances, ranging from 3 to 5 percentage points. For remittances of US$200–250, the costs vary from 5.6 to 13.8 per cent, and for amounts of US$300, between 5.1 and 12.7 per cent. In general, charges decline with the amount remitted, but there are significant differences in individual countries. Ecuador has the lowest charges and Venezuela the highest. An important factor explaining the lower charges for money remitted to Ecuador is that the exchange rate spread component of total costs (for the sender) is eliminated because the country's official currency is the US dollar. This is an important result: Ecuador, the Andean economy to have adopted the US dollar, enjoys lower costs for remittances than an economy with a national currency.20 Table 9.8 Average charges for sending a remittance of US$200 from the United States to Latin America (in US$dollars, and as %) Charges November 2001 November 2002 US$

%

US$% Total charge 20.06 10.10 17.02 8.50 FX charge 4.73 2.44 2.97 1.48 Fee charge 15.33 7.66 14.05 7.02 source: Orozco (2003). Table 9.9 Cost of remittances from the United States to the Andean Countries (in local currency versus US$, averages per country)

Amount

Country

Currency

Exchange

Fee charge

Total charge

Level

%

Level

%

Level

%

US$200 Colombia Local 9.30 4.65 10.67 5.33 19.96 9.98 Dollar 0.00 0.00 12.33 6.17 12.33 6.17 Ecuador Dollar 0.00 0.00 11.23 5.62 11.23 5.62 Bolivia Local 6.50 3.25 21.00 10.50 27.50 13.75 Dollar 0.00 0.00 16.80 8.40 16.80 8.40 Peru Local −3.54 −1.77 18.50 9.25 14.96 7.48 Dollar 0.00 0.00 13.00 6.50 13.00 6.50 Venezuela Local 12.04 6.02 15.00 7.50 27.04 13.52 Dollar 0.00 0.00 21.00 10.50 21.00 10.50 US$250

Colombia

Local

11.62

4.65

13.25

5.30

24.87

9.95

Dollar

0.00

0.00

15.39

6.16

15.39

6.16

Dollar

0.00

0.00

13.96

5.58

13.96

5.58

Bolivia

Local

8.12

3.25

27.00

10.80

35.12

14.05

Dollar

0.00

0.00

20.80

8.32

20.80

8.32

Peru

Local

−4.42

−1.77

22.50

9.00

18.08

7.23

Dollar

0.00

0.00

16.25

6.50

16.25

6.50

Venezuela

Local

15.05

6.02

18.75

7.50

33.80

13.52

Dollar

0.00

0.00

25.00

10.00

25.00

10.00

US$300 Colombia Local 13.95 4.65 14.88 4.96 28.82 9.61 Dollar 0.00 0.00 17.22 5.74 17.22 5.74 Ecuador Dollar 0.00 0.00 15.38 5.13 15.38 5.13 Bolivia Local 9.75 3.25 27.00 9.00 36.75 12.25 Dollar 0.00 0.00 22.40 7.47 22.40 7.47 Peru Local −5.31 −1.77 24.00 8.00 18.69 6.23 Dollar 0.00 0.00 17.83 5.94 17.83 5.94 Venezuela Local 18.05 6.02 20.00 6.67 38.05 12.68 Dollar 0.00 0.00 29.00 9.67 29.00 9.67 source: Solimano (2003). # 9.6 POLICIES TO REDUCE COSTS OF REMITTANCES AND ENHANCE THEIR DEVELOPMENT IMPACT As we have documented in this chapter, the cost of sending money transfers to developing countries is high, and this leads to an inefficient level of transfers. How to reduce the costs of sending money abroad? How to increase competition in the international market for transfers? How to enhance the development impact of remittances in the receiving countries? Measures are needed on both the sending side as well as the recipient side. ## 9.6.1 The Sending-Country Perspective ‘Formalization’ of the legal status of the migrant would certainly promote greater access by the migrant to a variety of bank services, including remittances services. This should lower the costs of remittances. For example, the use of ATM cards for making transfers rather than the current, more costly methods can be an effective mechanism for reducing the costs of remittances. Another factor that apparently prevents a competitive atmosphere in the remittances business in the United States is the cost of procuring a license for becoming an MTO, which is about US$100,000 per each state where operations are to be conducted. Prospective money operators find this cost high.

It is important also to avoid increases in transaction costs, or to add to the regulations governing worker remittances to reflect the mounting controls on financial intermediaries for preventing money laundering or the financing of terrorism.

(p.196) In sum, we believe that increasing the efficiency of the market for remittances requires:

1. (i) The costs of licensing for new operators to be contained or reduced so as to make the process of certification of new financial intermediaries in the remittances business less costly and more expedite.

2. (ii) The process of granting residence visas and/or citizenship to be expedited so as to avoid long visa processing periods for migrants (which currently can take up to several years, at least in the United States). This would help to regularize the immigrant sector, inviting commercial banks to target the financial needs of the migrants.

3. (iii) Domestic banks (particularly those with an international scope) to be encouraged to develop new product lines for migrants such as chequeing or savings account, remittances services, etc. The creation of ‘banks for migrants’ is an idea worth exploring.

The remittance-receiving nations would benefit from a more efficient and less costly market for remittances. Currently, a significant slice of remittances goes to operators as profits rather than to families of migrants in developing countries. This has adverse efficiency effects and is socially regressive.

## 9.6.2 The Recipient-Country Perspective

From the viewpoint of recipient countries, leveraging remittances and enhancing their productive use for development are two important issues. There are various mechanisms for leveraging remittances in the receiving countries: governments and local financial institutions can issue bonds targeted for emigrants, who would thus earn interest, and it would create a more attractive instrument for channelling remittances.

In addition, housing and education accounts can be created to channel remittances to various productive uses in the home country, such as investment in durables (housing) and education (investment in human capital).

The development of alliances between domestic banks in the receiving countries and banks, credit unions and MTOs in the sending nations can help to increase efficiency and reduce costs in the remittances market. Mechanisms to ensure a productive use of remittances include the mobilization of HTAs similar to those that have evolved in the United States in recent years (Mexican migrants have been very active in creating HTAs and are being helped by their government in this effort).

Finally, taxing remittances (mainly worker remittances) in the sending countries or in the receiving economies does not seem to be a good idea.21 These are transfers, sent in general by and to low-income groups. So, it is doubly inequitable that such flows, based on income that has already subject to the income tax system of the sending country, should be taxed. In receiving countries, remittances are a source of foreign exchange, a complement of national savings and a transfer to low- to medium-income (p.197) groups. It is unclear what the social gain would be if governments were to interfere directly with these income flows in any way to diminish them.

# 9.7 CONCLUDING REMARKS

This chapter examines the developmental and financial dimensions of remittances from international migrants. Remittances are currently the second most important source of development finance at the global level after FDI. Also, they are more stable than private capital flows such as portfolio investment and bank credit. The sustainability over time of remittances as a source of income for developing countries depends also on the cycle of migration (recent versus older migration) and the expected flows of migration. Remittances have become a very significant source of development finance for several developing countries. They are a source of foreign exchange; they support the consumption levels of low-to-middle-income families and constitute a direct source for funding small, community–oriented investment projects through the migrants' associations that send donations home in support of these projects (the so-called ‘community remittances’). From a social point of view, remittances can have a positive poverty-reducing effect, as families receiving remittances from migrants are often low-income people. However, the recipient syndrome of relying on remittances for income should be avoided. Properly mobilized remittances can contribute to increased investment in basic infrastructure such as water, roads, low-income housing, school-buildings, investment in human capital (education), and help to finance micro and small-scale firms. For remittance-sending countries, remittances represent a market-based international transfer to developing countries that indirectly reduces the demand for ODA.

Still, we have to recognize that earning foreign exchange through remittances entails an implicit tradeoff in the form of an outflow of skilled manpower from the sending countries.

Currently, the potential development impact of remittances is impaired, in part, by the existence of a costly, concentrated and poorly competitive international market for remittances. Empirical evidence shows that the costs of remittances are above what the marginal costs of (electronically) transferring funds provided that electronic transfers are possible. Although the involvement of commercial banks in the remittances market is still small, evidence shows that the costs of sending remittances tend to be lower when transferred through banks rather than through international MTOs. In addition, there are differences in the costs of sending remittances to non-Latin American countries compared to Latin American countries: the exchange rate spread component is higher for remittances sent to Latin American countries than to non-Latin American countries. Our empirical analysis, based on a detailed survey of MTOs in the United States who operate within the Andean region, shows that the total cost of remittances for these countries vary from 5 to 12 per cent of the value remitted depending on the type of currency to be delivered, destination country, type of financial operator involved, and other factors. Reducing the costs of sending remittances by, say, 5 percentage points could increase the amount of remittances received by developing countries by a few billion.

(p.198) What can be done to increase competition and reduce costs in the remittances market? In the sending countries, facilitating the process of opening bank accounts for immigrants could be an important step for integrating the migrant community into the financial system of the host country. This would increase competition and reduce the costs of remittances. On the other hand, the costs of licensing for new operators and other regulations on the part of banks and non-bank intermediaries wishing to provide services for migrants should be minimal to as to avoid creating entry barriers into the this market. Also, efforts for controlling money laundering or the financing of terrorism should not unnecessarily increase the costs to emigrants of sending home remittances. On the recipient side, the issuance of remittance bonds, the opening of foreign currency accounts for migrant workers in the home country, the creation of facilities for voluntary donations to projects are all measures to leverage remittances for development. In turn, the creation of education and housing accounts at home for migrants could help to enhance the productive and social use of remittances proceeds. Also encouraging the return of emigrants—bringing fresh capital, new ideas, and international contacts—can be a promising way to attract remittances for growth and development in receiving countries.

REFERENCES

Bibliography references:

Barham, B. and S. Boucher (1998). ‘Migration, Remittances and Inequality: Estimating the Net Effects of Migration on Income Distribution’. Journal of Development Economics, 55: 307–31.

Beckerman, P. and A. Solimano (eds) (2002). Crisis and Dollarization in Ecuador: Stability, Growth, and Social Equity (Directions in Development). Washington, DC: World Bank.

Bendixen and Associates (2003). ‘Receptores de Remesas en Ecuador. Una Investigacion del Mercado’. Presentation at Multilateral Investment Fund—Pew Hispanic Center conference on Remittances and Development, Quito, Ecuador, May.

Brown, R. (1997). ‘Estimating Remittances Functions for Pacific Island Migrants’. World Development, 25(4): 613–26.

Ellerman, D. (2003). ‘Policy Research on Migration and Development’. WB Policy Research Working Paper 3117. Washington, DC: World Bank.

Gammeltoft, P. (2002). ‘Remittances and other Financial Flows to Developing Countries’. Working Paper 02.11. Copenhagen: Centre for Development Research.

International Monetary Fund (IMF) (2003). Balance of Payments Statistics Yearbook 2003. Washington, DC: IMF.

Orozco, M. (2001). ‘Globalization and Migration: The Impact of Family Remittances in Latin America’. Washington, DC: Multilateral Investment Fund-IADB.

—— (2002). ‘Attracting Remittances: Markets, Money and Reduced Costs’. Washington, DC: Multilateral Investment Fund-IADB.

—— (2003). ‘Workers Remittances: The Human Face of Globalization’. Washington, DC: Multilateral Investment Fund-IADB.

Poirine, B. (1997). ‘A Theory of Remittances as an Implicit Family Loan Arrangement’. World Development, 25(4): 589–611.

(p.199) Ratha, D. (2003). ‘Workers' Remittances: An Important and Stable Source of External Development Finance’, Global Development Finance, 2003. Washington, DC: World Bank, chapter 7.

Smith, A. (2003). ‘Leveraging “Mobile” Human Capital for Development. Migration and Development Finance’. Mimeo, UN-DESA.

Solimano, A. (2001). ‘International Migration and the Global Economic Order: An Overview’. WB Policy Research Working Paper 2720. Washington, DC: World Bank.

—— (2002a). ‘Globalizing Talent and Human Capital: Implications for Developing Countries’. Series Macroeconomics of Development No. 15. Santiago: UN-ECLAC.

—— (2002b). ‘Development Cycles, Political Regimes and International Migration: Argentina in the 20th Century’. Paper presented at the UNU-WIDER Conference on Poverty, International Migration and Asylum, 27–28 September. Helsinki: UNU-WIDER.

—— (2003). ‘Workers Remittances to the Andean Region: Mechanisms, Costs and Development Impact’. Santiago: UN-ECLAC. Mimeo.

Stark, O. (1991). The Migration of Labor. Oxford: Basil Blackwell.

World Bank (2002). ‘Migrants Capital for Small-Scale Infrastructure and Small Enterprise Development in Mexico’. Project Report. Washington, DC: World Bank.

—— (2003). Global Development Finance, 2003. Washington, DC: World Bank.

## Notes:

Very useful comments by Tony Atkinson are greatly appreciated, as are the comments received from participants at the workshop of the UNU-WIDER project ‘Innovative Sources for Development Finance’ in Helsinki on 17–18 May 2003. Efficient assistance by Claudio Aravena is greatly appreciated.

(1) Leaving their families at home, new immigrants may initially go to the foreign country alone. Later, as their employment situation in the host country is stabilized, they tend to bring their families. This may have implications for the flow of remittances and their persistence over time, as families are often the main recipients of remittances.

(2) See Ellerman (2003) and Solimano (2002a) for a discussion of these issues.

(3) Remittances are defined as the sum of workers remittances and compensation of employees.

(4) As remittances are also sent through informal and un-recorded channels, official data may underestimate actual remittances (see Section 9.3).

(5) This section draws largely on Solimano (2003).

(6) See Brown (1997).

(7) Data for different countries are compiled in the IMF (International Monetary Fund) Balance of Payments Statistical Yearbook.

(8) References of this literature are Stark (1991), Brown (1997), Poirine (1997), Smith (2003).

(9) See Stark (1991, chapter 16).

(10) See Poirine (1997) and Brown (1997) for an elaboration on this specification of remittances.

(11) See chapter by Micklewright and Wright in this volume on the role of private donations, mainly from foundations and other vehicles, as a source of development finance.

(12) This effect is extensive to all kinds of transfers, not only to remittances.

(13) However, extreme poverty may also impede emigration, as the very poor may not be able to finance the costs of migrating to a foreign country.

(14) The distributive effects of remittances in the home country are more ambiguous. The issue is investigated in Barham and Boucher (1998).

(15) In the United States, banks request people (migrants) to provide a tax identification number, TIN, as a requisite for opening a bank account. In addition, some banks have recently accepted consular identification cards for opening bank accounts. Many migrants are totally compliant with tax returns even if their immigration status is not regular.

(16) See Orozco (2002) and Solimano (2003).

(17) In the aggregate, however, this is a sector that mobilizes a large volume of resources: aggregate remittances for Latin America were on the order of US\$32 billion in 2002 for the main twelve recipient countries in Latin America (see IMF 2003).

(18) Table 9.5 reports the countries and companies studied to determine the costs of remittances according to major source/destination countries and type of financial operator.

(19) See Solimano (2003).

(20) See Beckerman and Solimano (2002) for an analysis of the macroeconomic and social impact of official dollarization in Ecuador.

(21) Another possibility is to make remittances tax deductible.