on the use of differing money transmission methods by mexican immigrants

42
Documento de trabajo E2004/06 On the Use of Differing Money Transmission Methods by Mexican Inmigrants Catalina Amuedo-Dorantes Susan Pozo centrA: Fundación Centro de Estudios Andaluces centr A: Fundación Centro de Estudios Andaluces Consejería de Relaciones Institucionales Consejería de Relaciones Institucionales TURISMO ANDALUZ

Upload: wmich

Post on 03-Dec-2023

0 views

Category:

Documents


0 download

TRANSCRIPT

Documento de trabajoE2004/06

On the Use of Differing MoneyTransmission Methods by Mexican

Inmigrants

Catalina Amuedo-DorantesSusan Pozo

centrA: FundaciónCentro deEstudios AndalucescentrA: Fundación

Centro deEstudios Andaluces

Consejería de Relaciones InstitucionalesConsejería de Relaciones Institucionales

TURISMO ANDALUZ

Las opiniones contenidas en los Documentos de Trabajo de centrA reflejanexclusivamente las de sus autores, y no necesariamente las de la FundaciónCentro de Estudios Andaluces o la Junta de Andalucía.

This paper reflects the opinion of the authors and not necessarily the view of theFundación Centro de Estudios Andaluces (centrA) or the Junta de Andalucía.

Fundación Centro de Estudios Andaluces (centrA) Bailén, 50 - 41001 Sevilla

Tel: 955 055 210, Fax: 955 055 211

e-mail: [email protected]://www.fundacion-centra.org

DEPÓSITO LEGAL: SE-108-2002

Documento de TrabajoSerie Economía E2004/06

On the Use of Differing Money TransmissionMethods by Mexican Immigrants

Catalina Amuedo-Dorantes Susan PozoS. Diego State University y centrA Western Michigan University

RESUMENExiste un creciente interés en los diferentes métodos utilizados por los inmigrantes ala hora de remitir dinero a casa dada la opinión generalizada entre los investigadoresde la inadecuada infraestructura que rodea a la transmisión de remesas. Estapreocupación también se ha reflejado en las autoridades públicas, que coinciden en laimportancia de promover la transferencia segura y eficaz de las remesas a los paísesde origen de los inmigrantes. Este artículo usa datos sobre inmigrantes mejicanos quehan residido en los Estados Unidos con el objeto de analizar los diferentes factores queexplican el uso de determinados métodos a la hora de remitir dinero a Méjico. Elanálisis revela que factores de accesibilidad, al igual que el conocimiento de métodosalternativos para enviar remesas, juegan un papel fundamental a la hora de explicar eluso de determinados métodos de transferencia de capital.

Palabras clave: Remesas internacionales, métodos de transmisión de capital,inmigrantes mejicanos.

Clasificación JEL: F22, J61.

ABSTRACTInterest on the factors shaping migrants’ use of a given money transmittal method hasrecently intensified following researchers agreement on the often inadequateinfrastructure surrounding remittances transfers. This concern has also captured theattention of government officials, who appear more eager to promote more efficientand safe transfers of emigrant’s earnings given the potential that remittances hold forincreasing resources at the disposal of receiving nations. This paper uses data frommexican immigrants who have resided in the United States to examine the variousfactors that shape migrants’ use of the various methods to remit earnings to Mexico.The analysis reveals that accessibility factors and migrants’ awareness of alternativeremitting methods play a key role in explaining their use of the various money transfermechanisms.

Key words: International remittances, money transmission methods, mexicanimmigrants.

JEL classification: F22, J61.

centrA:Fundación Centro de Estudios Andaluces

1

A. Introduction

The rising importance of workers’ remittances has resulted in growing interest in

understanding by what channels immigrant workers repatriate their earnings. The interest in

workers’ remittances arises from the recognition that there is considerable potential for these

flows to positively impact the economic development of the labor-exporting, remittances-

receiving economies. But there is also concern that the channels by which workers traditionally

remit money back home may be sub-optimal in a variety of ways. In this paper, we first give an

overview of the market for transferring remittances. We modify standard payment theory to

account for cross-border monetary flows and then use this framework to derive hypotheses

regarding the factors that might explain migrants’ choice of a money transfer method. These

hypotheses take into consideration the preferences of migrants, the attributes of the differing

transferring mechanisms, and existing institutional constraints. The hypotheses derived are then

tested using data on money transfers to Mexico by Mexican immigrants with U.S. work

experience.

A variety of reasons have been advanced to explain why emigrants remit money to

friends and families back home. Remittances may be sent for altruistic purposes, to meet the

daily consumption needs of non-migrating family members –spouses and children, parents and

siblings (Lianos 1997). Transfers are also sometimes used to finance the emigration of

additional family members and to make human capital investments in family members back

home (Basok 2000; Brown and Ahlburg 1999; Cox Edwards and Ureta 2003). Alternative

reasons for sending remittances advanced by researchers include family risk-sharing (Lucas and

Stark 1985), to accumulate precautionary saving (Amuedo-Dorantes and Pozo 2002), to purchase

Fundación Centro de Estudios Andaluces

2

physical assets in the home communities (Durand et al., 1996), and to pay for goods that have

been received from the family back home (Menjivar 2002).1

1 Menjivar (2002) reports on the use of remittances as payment for medicines sent periodically by Guatemalans to their relatives in the United States.

Just as there are many different motives for sending remittances, there also exist a variety

of avenues by which migrants remit their earnings home. Researchers have observed that the

preferred means seem to differ according to the nationality of the remitter (Meyers 2002),

possibly reflecting the specifics of the market for remitting services available to the migrant’s

family in their home country. Additionally, as we shall discuss in what follows, the migrant’s

choice of a particular means for sending money back home may vary according to the differing

preference functions of each immigrant and the matching of these preferences to the attributes of

each transfer mechanism. Furthermore, a variety of institutional and economic constraints may

also impinge on the transfer mechanism that is used.

Despite the existence of a variety of transmittal modes, almost all researchers agree that

the infrastructure surrounding remittances transfers is inadequate and that the objectives of the

remitters would be better served if more cost effective, secure and efficient means of transferring

sums of money were available. It is estimated that about 70 percent of the total volume of

remittances sent by emigrants to Latin America is transferred to areas with little or no banking

services (Iglesias 2001). As a result of the poor infrastructure, the transactions costs and risks

borne by the migrant for transferring funds are high. While there is some evidence that the

market for money transfer services is becoming more competitive (Lowell and de la Garza 2002;

Handlin et al., 2002), transaction costs continue to be relatively high. Fees charged by non-bank

money transfer firms –such as Western Union and Money Gram– have varied over time and

E2004/06

3

across countries, but Orozco (2002a) quotes 7.54 percent as the average cost of transferring $250

to Latin America. In contrast, he finds that, on average, banks and credit unions charge 4.24

percent and 3.07 percent, respectively.

The observation has also been made that remittances can be an important factor in

equalizing opportunities across countries’ regions (Belo 2001). The poorer regions, with less

saving, investment, and greater levels of unemployment and underemployment, tend to be the

areas experiencing the most out-migration. An important return to labor-exporting regions is the

remittances inflow that eventually takes place. Woodruff and Zenteno (2001) find that

remittances appear to be responsible for over 40 percent of investments in micro-enterprises in

the highest out-migration states of Mexico. The development of small businesses that takes

place in this manner may result in a more evenly developed national economy. But, if the

infrastructure for transferring and saving funds for future investments is not as well developed,

the areas in greatest need are less likely to benefit. Instead of observing that the poorer regions

“catch-up” we may instead observe a growing economic gap across regions of the country as the

more developed areas are better able to capitalize on these inflows. This possibility underscores

the need to better understand the market for transferring remittances.

B. Overview of the Market for Transferring Remittances

The money transmission market appears to be growing very rapidly while, at the same

time, evolving in various directions according to the particulars of the sending and destination

areas (Orozco 2002b). This rapid growth, coupled with non-standardization, makes for

difficulties in generalizing about the market for money transmission services. With this caveat in

mind, we describe and broadly categorize the available options available for transferring

remittances in what follows.

Fundación Centro de Estudios Andaluces

4

In general, there are three categories of “institutions” that individuals use to transfer

funds back home. We will refer to the institutions within each category as: 1) banks, 2) non-

bank money transfer firms (MTFs), and 3) informal channels. In the bank category, we include

all formal financial intermediaries including credit unions. In contrast, we refer to institutions

such as Western Union and Money Gram as non-bank MTFs. MTFs are not categorized as

banks because they do not provide many of the services offered by banks and in particular they

play no formal role in financial intermediation between savers and investors. Instead, MTFs

simply transfer sums of money from one country to another, often combined with exchanging the

sum from one currency to another.2 In our final category (informal channels), we include cash

sent in the mail and the transfer of money by friends and family traveling to the place of

destination.

The migrant’s choice of a particular money transfer mechanism ultimately depends on the

attributes of the payment media offered by each of the money transmission institutions.3 While

currency and bank money (checks) are the most common payment media used in the United

States for domestic transactions, there are many other payment media including credit cards,

debit cards, traveler’s checks, postal money orders, and e-money (balances embedded in

computer chips) that are also used. The choice of payment media by individuals essentially

involves the matching of the individual’s preferred characteristics in a payment method with the

actual attributes of each available payment media. For example, consider a very simple payment

system involving only two types of payment media –currency and checks. Will individuals

make disbursements and transfer purchasing power using currency or will they instead use

2 MTFs also engage in domestic money transfer services and issue money order type instruments for individuals needing such services. However, MTFs do not provide financial intermediation and, as such, are distinguishable from banks.

E2004/06

5

checks? The answer essentially depends on how the characteristics of each media match the

desired media attributes by the individuals (senders and receivers) involved in the transactions.

Check transactions are generally safer, especially when the sum is transported over long

distances. However, individuals who wish to preserve their anonymity or who distrust the

banking system may prefer to avoid check transactions, using cash instead.

In addition to safety and anonymity, individuals involved in money transfers may also

consider additional attributes of the differing media including transactions costs, its ease of use,

the media’s performance as a store of value, its acceptability as payment, and its geographic

accessibility. The potential users of the payment media will assess how each payment method

rates with respect to each of these attributes.

Table A presents a possible rating of the three payment media considered in this paper –

currency, checks and non-bank money orders– with respect to each of the attributes listed

above.4 For instance, with regard to transmission safety, we rate checks and non-bank money

orders as high (H) in transmission safety, while currency rates low (L). Currency, on the other

hand, is an excellent payment media for individuals interested in preserving their anonymity,

while checks and money orders rate lower on this attribute. Given their fee structure, money

orders of MTFs are quite costly (Orozco 2002a). Similarly, foregone interest opportunities make

currency a costlier media to use when transmitting money relative to checks. Hence, we have

ranked currency and money orders as low in providing an economical means to transmit money,

while checks rank high in economy. Of the three payment media, checks and, to a lesser extent,

money orders are likely to be more intimidating for less educated and older individuals, while

3 See Hancock and Humphrey (1998) for an overview of the different payment media that facilitate the transfer of purchasing power from one entity or individual to another and comprise what is referred to as the payment system. 4 Location-specific differences in the characteristics of these institutions may cause the “accounting” to differ for transfers to different regions in the world.

Fundación Centro de Estudios Andaluces

6

currency is likely to rank highest in desirability and familiarity for the same individuals. Checks

outperform money orders and currency as a store of value, while currency (due to its legal tender

status) is most liquid and ranks highest on that score. Nonetheless, due to their more widespread

geographic dispersion in areas where migrants tend to send money, non-bank MTFs rank higher

than banks with respect to international money transmissions for the majority of the population

we are considering.

Table A – Degree to Which Attributes Are Met by Each Form of Money

Desired Attributes Currency (Informal)

Checks(Banks)

Money orders(Non-bank MTFs)

Transmission Safety L H H Anonymity H L L Economy (for relatively large sums) L H L Ease of Use by Less Educated and Older Remitters H L M Store of value L H M Acceptability as Payment (Liquidity) H M M Geographic Availability (Diffusion) H L M

Notes: H=high, M=medium, L=low.

Using Table A as a guide, we can derive several hypotheses regarding money transfers on

the part of the users of these services. Consider a migrant who sends money home to a remote

region in Mexico. Currency and money orders are better media relative to checks due to the

relatively poor banking infrastructure in the more remote areas. In this case, migrants are more

likely to engage in cash transfer or, alternatively, use the services of MTFs as these institutions

have strived to service these areas. Hence, we hypothesize that migrants who are transporting

sums to more remote regions in Mexico are more likely to use informal (currency) transactions

and the services of non-bank MTFs, ceteris paribus.

By considering the special features of the payment media and the money transmittal

market for money transfers from the United States to Mexico, we derive a set of hypotheses

E2004/06

7

regarding the migrant’s choice of a particular money transfer mechanism. These hypotheses take

into consideration the migrant’s personal characteristics and the attributes of the payment media

offered by the various money transmittal methods. We summarize these hypotheses in Table B.

Table B – Hypothesized Preferred Money Transmittal Methods According to Various Migrant and Payment Characteristics

Migrant and Payment Characteristics Currency (Informal)

Checks(Banks)

Money orders(Non-bank MTFs)

Size of Payment x Undocumented Remitters x x Age x x Educational Attainment x Time or Experience in U.S. x Remitters with Networks of Family and Friends x Rural and Remote Areas as Destinations x x Transfers intended for Saving x Transfer used for Purchasing Goods x x

According to Table B, we hypothesize that the transaction size will influence the choice

of transmission mechanism, with checks taking precedence when payment amounts are large

(Shy and Tarkka 2002; Santomero and Seater 1996). This is anticipated because transactions

costs on a personal check do not vary with the check amount, while fees charged by non-bank

MTFs generally vary directly with the transaction amount. Likewise, currency is less desirable

due to its high opportunity cost in foregone interest. Hence, we hypothesize that a direct

relationship exists between payment size and the likelihood of using checks to transfer money to

the home community.

With respect to documentation status, it is noteworthy that unauthorized immigrants are

often barred from using banking services due to their lack of social security identification.5

Therefore, we hypothesize that undocumented immigrants are more likely to use currency and

Fundación Centro de Estudios Andaluces

8

the services of non-bank MTFs rather than banks. We also hypothesize that older individuals are

less likely to use more complicated and less transparent forms of money, such as checks (Gerdes

and Walton 2002). In contrast, better educated individuals, with more U.S. work experience and

with networks of family and friends in the United States are more likely to understand the

banking system and maintain bank checking account balances. As such, we hypothesize that

these immigrants are more likely to be using banks to transfer sums of money home.

Additionally, institutional constraints may also play a role in the choice of a money

transfer mechanism. If banks are not available in remote regions, currency and, to a lesser

extent, MTFs may be, out of necessity, the only viable disbursement method. Finally, the end

use of payments –whether they are being sent for saving/investment versus for purchasing

goods– may also influence the choice of transmission method (Hancock and Humphrey 1998;

Snellman et al., 2001). We hypothesize that funds that are transmitted for saving or investment

purposes are more likely to be transferred using banks given the availability of saving and

investment instruments in these institutions.

C. Mexican Migrant Transmitting Behavior: Data and Descriptive Evidence

To test the hypotheses outlined above regarding the method used to remit money home,

we take advantage of information contained in a large periodic survey of Mexican migrants

known as the Encuesta sobre Migración en la Frontera Norte de México (EMIF). These data are

collected by the Colegio de la Frontera Norte (COLEF)6 in eight different cities along the United

States-Mexico border: Tijuana, Mexicali, Nogales, Ciudad Juárez, Piedras Negras, Nuevo

Laredo, Reynosa, and Matamoros. These cities account for more than 90 percent of the

5 Grace (2001) notes that, though this is a relatively unrecognized point, financial institutions are not barred from servicing individuals lacking social security numbers so long as their accounts are non-interest bearing.

E2004/06

9

migration flux from Mexico to the United States and vice versa (Secretaría del Trabajo y

Previsión Social 1998). Additionally, the survey methodology is designed to constantly update

the data flow to obtain a sample that properly represents where and when migrants cross the

border into Mexico. Individuals are surveyed who cross on foot, by train, car, bus and plane.

During each survey shift, an interviewer applies a screening form that permits differentiating

migrants from tourists and individuals born in the United States. Once a person is considered

eligible (i.e. a migrant), the EMIF questionnaire is administered anonymously by a trained

interviewer.

We use data from five consecutive waves of the EMIF: 1993-1994, 1994-1995, 1996-

1997, 1998-1999, and 1999-2000. Each wave includes four quarterly surveys administered

separately to four groups of migrants in the border regions: migrants coming from South of the

Northern Mexican border region, migrants in Northern border cities originating from another

Northern border community, migrants returning from the United States to or through the

Mexican Northern border region, and Mexican migrants deported from the United States. Given

the purpose of this study, we focus on one of these four surveys: Mexican migrants returning

(voluntarily) from the United States. This group includes individuals who are 12 years old or

older, not born in the United States, and who migrated to the United States with the purpose of

visiting family, friends, complete some business, or to work for more than one month.

It is important to recognize that our data lack information on two groups of Mexican

emigrants. First, the EMIF does not include Mexicans emigrants in the United States who never

return to Mexico. In this respect, our sample is not representative of the entire universe of

Mexican immigrants with U.S migration experience. Our sample, instead, represents Mexican

6 COLEF carried out the survey for the Secretaría del Trabajo y Previsión Social and the Consejo Nacional de Población.

Fundación Centro de Estudios Andaluces

10

immigrants returning to Mexico at some point in their lifetimes, either temporarily (e.g. to visit

family, for vacation, etc.) or permanently. A second category omitted from the EMIF survey is

that of Mexican emigrants who return to interior points in Mexico via air.7 While present, these

deficiencies are not likely to greatly compromise our final results for the following reasons.

First, non-returnees are expected to be relatively few due to the proximity of the two countries

(Lowell 1992, Lindstrom 1996, Reyes 1997, and Orrenius 1999).8 Second, both non-returnees

and migrants traveling to interior points of Mexico via air are likely to be more integrated into

the U.S. economic and social fabric. These emigrants are likely to have weaker attachments to

Mexico, making them less likely to remit. As such, their exclusion is less likely to affect

conclusions regarding the question at hand of how migrants remit money home.

We work with a sample of approximately 6,000 Mexican migrants returning from the

United States and remitting a fraction of their last month of U.S. earnings. A description of the

variables used in our analysis, with their means and standard deviations are displayed in Table 1.

Our respondent is on average 34 years old with 413 days of residency in the United States during

his or her last visit. Ninety-five percent of our respondents are male, 60 percent have less than a

middle school education, and about a third are undocumented. Seventy-two percent of our

respondents are wage and salaried workers and 32 percent are employed in the agricultural and

mining sectors. On average, migrants who remitted sent $487 from their last month’s earnings

while in the United States

The EMIF asks migrants about the method used to remit money home, distinguishing

among banks, money orders, family or friends, letter, telegram and other methods. As we

7 Border airports are included in the survey and, hence, air travelers to the border area are incorporated in the survey. 8 This hypothesis is supported by the fact that other surveys—such as the Mexican Migration Project (MMP at www.pop.upenn.edu/mexmig/databases/databases.htm), in their attempt to get a representative sample of Mexican

E2004/06

11

explain in the empirical methodology, we group some of these options together. In particular,

we group family or friends and letter in the ‘informal’ transmission method, while money orders

and telegrams are lumped in the MTF category. In addition to the three categories discussed

earlier: 1) banks, 2) MTFs, and 3) informal methods, we include a fourth category denoted as

‘other.’ A sizable portion (about 5 percent) responded that they remitted via other methods.

While we do not know what other methods refer to, we include these observations in the analysis

under a category designated as ‘other’ instead of simply eliminating them from the study.

The repeated design of the EMIF survey (5 waves over the 1993-2000 period) allows us

to make a few preliminary generalizations regarding immigrant’s remitting methods. An

overview of trends and patterns is evident from the information contained in Table 2. Non-bank

MTFs constitute the preferred money transmission mechanism for most migrants. Nonetheless,

it is interesting how this method has slowly but steadily lost market share to the banking system.

By 1998, banks were the second most important means for sending money back home.

Additionally, the use of informal means to remit money (using friends, family, or cash in the

mail) has decreased over time falling from about 15 percent of transactions in 1993-1994 to less

than 11 percent of transactions in 1999-2000.

These overall results are in-line with a smaller but broader survey undertaken by the

Inter-American Development Bank (IADB 2002) during November and December of 2001. A

sample of 1000 U.S. residents remitting to all areas of Latin America were interviewed. This

survey revealed that only 20 percent of migrants used banks to transfer funds to family and

friends back home while 41 percent of the respondents sent remittances via Western Union and

Money Gram. This corroborating evidence suggests that non-bank MTFs are migrants’ preferred

migrants, interview a relatively small number of Mexican migrant households residing in the U.S. (about 700) while approximately 4,000 households with U.S. migration experience are interviewed in Mexico.

Fundación Centro de Estudios Andaluces

12

means to remit money home whether they are remitting to Mexico or to other areas in Latin

America.

The finding that migrants prefer MTFs is significant because it is well known that the

fees typically charged by MTFs are fairly steep and tend to exceed by a wide margin the fees

charged by banks (Orozco 2002a). Despite these relatively steeper fees, immigrants appear to

feel fairly comfortable and well served by MTFs. The extensive networks of agencies in both

sending and recipient communities apparently compensates for the higher fees. It may also be

the case that differing regulations imposed on banks and non-bank MTFs with respect to

transmitting funds have contributed to the greater ease that migrants appear to enjoy when

remitting via non-bank MTFs. This advantage enjoyed by non-bank MTFs has recently been

eroded by the U.S.A Patriot Act, which places more responsibility on all financial institutions for

better establishing the identity of clients and for putting in place anti-laundering safeguards.

While there is concern that this legislation will further impede the process of sending money

home via formal channels, according to Bair (2002), this regulation will “level the playing field”

for the banking sector by making non-banks subject to the same regulations that banks have

essentially been subjected to all along.

Finally, it is interesting to note that the IADB survey also reports that 29 percent of

respondents claim to either send remittances through the mail or via family and friends traveling

home, what we delineate as the ‘informal’ category. This percentage is considerably higher than

in the EMIF survey, possibly due to differences in the sample and resulting institutional

arrangements for the different receiving regions9.

9 For example, remittances to Cuba have tended to be sent, to a greater extent, via informal means given the set of regulations and restrictions placed on such flows both on the part of the U.S. and Cuban governments. (See Díaz-Briquets and Pérez-López, 1997.)

E2004/06

13

D. Empirical Methodology: Choosing a Particular Money Remitting Method

When modeling migrants’ choice of a particular money transmission method, we can

think of each migrant (i=1…n) as deriving utility Uij from any of the four money transmission

mechanisms with j = 1 corresponding to banks, j = 2 corresponding to MTFs, j = 3 correspondi

to informal channels, and j = 4 corresponding to other mechanism. The migrant considers thes

four alternatives and then chooses the transmission method from which she/he derives the

highest utility. As discussed earlier, the utility derived by each migrant from the remitting

method used depends on a list of personal and payment characteristics listed in Table B. We a

account for the migrant’s gender and industry of employment to control for other socio-

demographic and employment characteristics potentially affecting migrant’s utility and, hence,

final choice of money transmission mechanism. Therefore, we write the utility derived from

choosing a particular money transmission mechanism as follows:

(1) Uij = Vij +εij = ijj x'β + ijε ,

where Vij and εij represent the deterministic and stochastic components of the utility function,

respectively. In this setting, the probability that the ith migrant chooses the jth money

transmission mechanism is equal to the probability of Uij being the largest of: Ui1, …Ui4. That

(2) Pij = Prob (Yi = j) = Prob (Uij > Uik ) = Prob ( ijε - ikε ≤ ijj x'β - ikk x'β ),

where k = 1…4 and k ≠ j. Given our deterministic components, this probability will depend on

the assumptions we make about the distribution of the stochastic error terms: εi1, …εi4. If we

assume that the errors terms are distributed according to a Type I extreme-value distribution an

that there is independence across the εj’s, the multinomial logit can be derived from utility

maximization (McFadden 1974). As a result, the migrant’s choice of a particular transferring

mechanism (such as the jth mechanism) can be described as given by the probability:

Fundación Centro de Estudios Andaluces

14

(3) Prob (Yi = j) = ∑=

4

1)'exp(

)'exp(

kijk

ijj

x

x

β

β ,

where j=1, …4 represents each of the alternative remitting methods. We can then set the

probability of remitting money by the most common method in the data (i.e. MTFs) as the base

category in the estimation, and thereby obtain the relative risk ratios10 for a one-unit change in

the corresponding independent variables (xij):

(4) )'exp( ijjik

ij xPP β= ,

where the category of reference is k=j (i.e. MTFs in our case).

However, one objection that might arise in estimating the multinomial choice model

described by equation (3) is the simultaneity of the choice of remitting method (our dependent

variable) and the dollar amount remitted (one of the independent variables). On the one hand,

migrants’ choice of remitting method may be influenced by the cost associated with remitting a

given dollar sum. On the other hand, it is also possible that the dollar amount remitted back

home may be dependent on the money transmission method used by migrants due to hard to

measure characteristics of the specific remitting method being used. For example, MTFs usually

institute a discrete-step fee structure. The transfer fee may be $15 for remitting amounts up to

$200, jump to $20 for remitting sums from $201 to $300, and jump again to $25 for sums greater

than $300. This would affect migrants’ remitting decision as well as the amount finally remitted.

If my plan is to remit $100 a month, given the fee structure, I might instead remit $200 every

other month to minimize on the transfer fee. In addition to the potential endogeneity of the

decision to remit and the dollar amount remitted with the choice of money transmission method

used by the migrant, we have that a large proportion of migrants in our sample who do not remit

E2004/06

15

any money home. Hence, in order to account for: 1) the potential endogeneity of the decision to

remit and the dollar amount remitted, and 2) the large number of migrants with a value of zero

for the amount remitted, we first estimate a Tobit regression of last month’s remittance amount.

In our Tobit estimation, we use the migrant’s family size back in Mexico as an identifier. The

size of the migrants’ families back in Mexico may affect their remittance amount, whereas there

is no reason for it to affect their choice of remitting method other than through the amount being

remitted itself. From the Tobit estimation, we obtain the unconditional predicted values for the

amount being remitted and use these to instrument for the dollar amount remitted in the

multinomial logit model (equation 3). In this manner, we simultaneously account for: 1) the

endogeneity of migrants’ remitting decision and final amount remitted with their choice of

money transmission method, and 2) the large number of migrants who do not remit money home.

Finally, one of the critical assumptions of the multinomial logit is the fact that the relative

risk ratios in (4) are independent of the other alternatives in the model. This assumption (which

follows from the initial assumption of the disturbances in the four equations in (2) being

independent and homoscedastic) is also known as the independence of irrelevant alternatives or

IIA (Greene 2000). One of the solutions to the rejection of the IIA hypothesis is to estimate a

nested logit model. The latter consists in relaxing the assumption of homoscedasticity in the

multinomial logit by providing an intuitive grouping of the different alternatives or outcomes in

the multinomial logit model and, in this manner, allow for the independence of irrelevant

alternatives.11 While the grouping of some of the alternative methods –such as remitting through

friends and family and remitting through the mail into the so-called ‘informal’ category, or the

grouping of remitting through telegrams and Western Union into the MTF category– should

10 These are also called relative probabilities or odd ratios.

Fundación Centro de Estudios Andaluces

16

provide a greater guarantee of independence of the four alternatives considered in (2), we test

whether this is still a reasonable assumption using the Hausman and McFadden (1984) test.

According to the results from the test (reported at the bottom of Table 4), which proceeds as a

common Hausman specification test, the IIA assumption cannot be rejected. Hence, when a

particular subset of money transferring methods is irrelevant, omitting it from the model

altogether does not significantly alter the estimated coefficients.

E. What Determines Migrants’ Choice of a Particular Money Transmitting Mechanism?

Before turning to our main objective of discerning what influences the choice of

remitting method, we briefly summarize what our data suggest regarding how much migrants

remit. Recall that, due to the potential simultaneity of the amount remitted and the choice of

money transmission method, we first estimate a Tobit model for the amount remitted and use its

predicted value as an instrument for the amount remitted in the estimation of our multinomial

logit. The Tobit coefficients, standard errors, and the marginal effects conditional on remitting

are displayed in Table 3.12 As one would expect, the amount remitted is directly related to the

needs of the family back home. In particular, for those remitting, the amount remitted to Mexico

increases by approximately $5 with every additional family member left back home. Similarly,

the amount remitted by migrants depends on their financial possibilities. Older males, perhaps

owing to their better income opportunities, as well as migrants with more than middle school

education remit more relative to migrants with less than middle school. Likewise, self-employed

family workers who remit, end up sending back home approximately $94 more than their wage

11 In fact, the multinomial logit specified in equation (3) coincides with a nested logit when the decision tree consists of a single level, as in this case. 12 Note that, though we report the conditional on remitting estimates in Table 3 (simply because they are more interesting for perusing), we use the unconditional predicted value for the amount remitted to properly account for the endogeneity of the amount remitted with the method used to remit in the multinomial logit estimation.

E2004/06

17

and salary counterparts. In contrast, migrants with networks of friends and family in the city

they migrated to, remit, on average, $24 less than remitting migrants lacking such social

networks. Lastly, migrants sending remittances for investment/saving purposes seem to remit

approximately $36 more, relative to their counterparts remitting to satisfy consumption needs,

such as rent, food or schooling. These results are consistent with others who have examined

remittance behavior. Menjívar et al. (1998), for example, find that the amount remitted is

positively related to the existence of family obligations back home and the ability to remit

measured by both income and earnings potential.

Table 4 displays the results from estimating the multinomial logit model of the likelihood

that the migrant will choose one of the four money transmitting methods being examined. The

results in Table 4 confirm some of our hypotheses regarding the factors shaping migrants’ use of

banks and informal means (relative to MTFs), while controlling for other variables that may be

affecting the choice of remitting method. Beginning with undocumented workers, our estimates

suggest that the likelihood of using banking services relative to non-bank MTFs is 37 percent

lower among undocumented migrants relative to legal immigrants.13 This finding supports the

hypothesis that undocumented migrants are more likely to be among the so-called “unbanked”

due, in part, to more rigid identification requirements generally imposed for the use of banking

services.

We also find that individuals with greater educational attainment are more apt to use

banks relative to MTFs, as are individuals who are employed in industry rather than in the

omitted agriculture category. In addition, the results indicate that when compared with salaried

13 If legal and undocumented migrants had the same propensity to send via banks, relative to non-bank MTFs, the coefficient on undocumented worker in the first multinomial logit equation would equal 1. However, with a coefficient of Pij/Pik of 0.63 the probability of remitting via banks (Pij) is 37 percent (i.e. (1 - 0.63) percent) lower than via non-bank MTFs.

Fundación Centro de Estudios Andaluces

18

workers, self-employed and specific task workers are more likely to use informal cash transfer

methods relative to MTFs. As we hypothesized, there is a slight preference for using MTFs over

cash as U.S. residency lengthens, and there is also evidence that the presence of social networks

also gives rise to a greater likelihood of bank usage. Immigrants with friends and family in the

U.S. have a 50 percent greater likelihood of remitting via banks relative to using money transfer

firms.

Finally, remittances to rural areas are 38 percent more likely to be sent using cash over

MTFs, supporting the hypothesis that rural areas are still underserved by formal money transfer

institutions. In addition, the time dummies corroborate the earlier finding from Table 2 of an

increasing trend in migrants’ use of banks relative to non-bank MTFs over the second half of the

1990s, possibly in response to the growing efforts by banking institutions to facilitate migrants’

use of the banking system.

While many of the hypotheses outlined in Table B appear to be supported by the data, the

hypotheses regarding the age of the sender, the amount remitted, and the purpose for sending are

not supported by the results. For instance, contrary to what we posited, we find that age does not

significantly influence the migrant’s choice of remitting method. This result may possibly be

due to the lack of information on the age and other characteristics of the recipient of remittances

back home, which, along with the migrant’s age, may also play a role in the choice of money

transmission mechanism. Similarly, we had posited that migrants are more likely to use banks

when remitting larger amounts or when they are remitting for saving and investment purposes.

Instead, we find that migrants rely on ‘other’ unspecified methods when remitting larger

amounts, and use banks and informal methods when remitting money home for purposes other

than consumption, saving or investment. Nonetheless, only 2 percent of remitters send money

E2004/06

19

home for other purposes and only 5 percent of remitters use a money transmission mechanism

other than banks, MTFs, and informal methods, diminishing the significance of these unexpected

results.

At this juncture, it is important to recall that the EMIF survey omits two categories of

Mexican immigrants: those who never return to Mexico and those who fly to non-border

airports. In what ways might our results change if we had been able to incorporate these two

categories of individuals? To the extent that non-returnees are likely to be better integrated into

the U.S. formal economy, these immigrants are more likely to use banking institutions to remit

earnings back to Mexico. Similarly, migrants who return to Mexico using non-border airports

are likely to be more educated and better off economically, making them more likely to use

banking facilities. Therefore, our analysis may yield underestimates of the use of banking

facilities by Mexican migrants in the United States to remit money back home.

However, because both non-returnees and returnees via air to non-border airports are also

likely to have weaker attachments to their home communities, they are less likely to remit and, if

remitting, they are likely to send smaller amounts. Hence, these two omissions are unlikely to

introduce large biases in the analysis of the question at hand, i.e. what motivates migrants’

choice of money transfer method?

F. Discussion and Conclusions

Interest on the factors shaping migrants’ preference for a particular transmittal method

has recently intensified following researchers’ agreement on the sometimes-inadequate

infrastructure surrounding remittances transfers. Furthermore, given governments’ objectives of:

1) making remittances transfers more cost effective, secure, and efficient, and 2) maximizing

their beneficial impacts on receiving areas, it is important to understand how immigrants decide

Fundación Centro de Estudios Andaluces

20

on the remitting method to be used. This paper uses data gathered from Mexican immigrants

during the 1993-2000 period to examine the various factors shaping migrants’ choice of a

particular money transmitting method with respect to remitting their U.S. earnings to Mexico.

The analysis uncovers some of the crucial elements behind migrants’ use of banks and informal

transfer methods relative to the use of MTFs. In particular, we find that migrants are less likely

to use banks (relative to MTFs) when they lack the appropriate immigration and identification

documents. In contrast, more educated and skilled migrants (as proxied by their industry of

employment), as well as migrants with networks of friends and family in the city where they

migrated to, appear more likely to use banks relative to the generally more expensive MTFs.

Furthermore, self-employed and specific-task workers are more likely to use informal transfer

mechanisms (relative to MTFs) than are salaried workers. However, migrants’ preference for the

use of informal (cash) transfer methods (over the use of MTFs) weakens with the increasing

length of their U.S. residency. Finally, migrants are more likely to rely on informal transmittal

methods when remitting money to rural areas perhaps due to the lack of extensive banking

infrastructure in the receiving areas.

Why are these findings of interest? Considerable effort has recently been expended by

the popular press and public interest groups in exposing the differential money transmission fees

typically charged by MTFs relative to banks and credit unions. Since MTFs are generally more

expensive than banks and credit unions, the push has been to find ways to induce immigrants to

use the banking system for money transfer purposes. Using more economical services would

enable migrants to channel a greater volume of funds to their home communities, presumably

benefiting recipient areas. Such discussion has also resulted in the modification of policies by

several banks in the United States allowing migrants to have access to certain banking services

E2004/06

21

using identification cards (la matrícula consular) received from Mexican consulates (Thompson

2002). Such a policy will presumably help alleviate the access problem for senders of

remittances. However, additional attention should be paid to the other end of the spectrum. In

order to induce migrants to use banks more intensively to remit, it is necessary that the banks

make inroads and provide services in the areas that typically receive large volumes of

remittances. Recipients need access to banks in the receiving areas to make it worthwhile for

senders to choose such methods.

Furthermore, another compelling reason to lobby for the greater use of banks on the part

of migrants revolves around the ability of banks to engage in financial intermediation –a service

MTFs do not provide. MTFs do not make loans and they do not take deposits. Banks and credit

unions, on the other hand, do both. When migrants remit funds home using formal banking

institutions, funds that are not immediately used to accommodate the consumption needs of the

recipient and would otherwise remain idle are more likely to be channeled from savers to

investors as banking institutions carry out their role of financial intermediaries. By contrast,

funds transferred to recipients in Mexico by other means, if not immediately consumed or

invested by the recipient, could very well remain as unproductive cash assets.

In our opinion, it would be worthwhile to persuade remitters about the advantages of

using the banking system so that the public good aspects (financial intermediation and resulting

increases in the level of investment) that result from an extensive banking system are realized in

the receiving economies. We know that a portion of remittances that are sent to Mexico do in

fact constitute saving for current or future investment. But where do the savings go when the

owners of these savings do not undertake productive investments immediately? Are these

savings stored as currency? Are they channeled into unproductive investment for lack of other

Fundación Centro de Estudios Andaluces

22

saving instruments? The contribution of financial intermediation to economic development

cannot be overstated and, as such, it cannot hurt to advocate for the greater participation of

financial intermediaries in poorer and rural areas to help alleviate credit constraints. Of course,

the use of banking transferring services will not automatically result in greater use of bank

saving and credit instruments. But it is more likely that, as senders and receivers of remittances

become more familiar with such institutions, demand for the other services that they provide will

rise. In this way, the areas in most need of resources from abroad will be in a better position to

receive and use them.

E2004/06

23

References

Amuedo-Dorantes, C. and S. Pozo. 2002. “Remittances and Insurance: Evidence from Mexican Migrants,” mimeo.

Bair, S. C. February 26, 2002. “Remarks by the Honorable Sheila C. Bair, Assistant Secretary for Financial Institutions Before Multilateral Investment/Inter-American Development Bank,” Second Regional Conference on Impact of Remittances as a Development Tool, http://www.iadb.org/mif/v2/files/bair.doc

Basok, T. 2000. “Migration of Mexican Seasonal Farm Workers to Canada and Development: Obstacles to Productive Investment,” International Migration Review, Vol. 34, pp. 79-97.

Belo, P. J. 2001. Comments at “Remittances as a Development Tool:A Regional Conference.” Inter-American Development Bank, http://www.iadb.org/mif/v2/files/pedrobelo.doc

Brown, R. P.C. and D. A. Ahlburg. 1999. “Remittances in the South Pacific,” InternationalJournal of the Social Economics, Vol. 26, pp. 325-44.

Cox Edwards, A. and M. Ureta. 2003.”International Migration, Remittances, and Schooling: Evidence from El Salvador.” NBER Working Paper No. w9766.

Díz-Briquets, S., and J. Pérez-López. 1997. “Refugee Remittances: Conceptual Issues and the Cuban and Nicaraguan Experiences.” International Migration Review, Vol. 31, pp. 411-437.

Durand, J., W. Kandel, E. A. Parrado, and D. S. Massey. 1996. “International migration and development in Mexican Communities,” Demography, Vol.33, No.2, pp. 249-64.

Gerdes, G. R., and J. K. Walton II. 2002. “The Use of Checks and Other Noncash Payment Instruments in the United States,” Federal Reserve Bulletin, August 2002, pp. 360-374.

Grace, D.. 2000. “The Developmental Potential of Remittances & the Credit Union Difference,” presented at “Remittances as a Development Tool: A Regional Conference,” May 17 & 18, 2001, Inter-American Development Bank, http://www.iadb.org/mif/v2/files/davidgrace.doc

Greene, William H. 2000. Econometric Analysis. Fourth Edition, Prentice Hall, New Jersey.

Hancock, D., and D. B. Humphrey. 1998. “Payment transactions, instruments, and systems: A survey,” Journal of Banking & Finance, Vol.21, pp. 1573-1624.

Handlin, E., M. Krontoft and W. Testa. 2002. “Remittances and the Unbanked,” Chicago Fed Letter, Federal Reserve Bank of Chicago 175A, Special Issue, March, pp. 1-4.

Fundación Centro de Estudios Andaluces

24

Hausman, J., and D. McFadden. 1984. “A Specification Test for the Multinomial Logit Model,” Econometrica, Vol. 52, pp. 1219-1240.

Iglesias, E. V. 2001. Comments at “Remittances as a Development Tool: A Regional Conference,” May 17 & 18, 2001, Inter-American Development Bank, http://www.iadb.org/mif/v2/files/iglesias.doc

Inter American Development Bank. 2002. “Survey of Remittance Senders: U.S. to Latin America,” Second Regional Conference on Impact of Remittances as a Development Tool, http://www.iadb.org/mif/v2/files/Bendixen2002.ppt

Lianos, T. P. 1997. “Factors Determining Migrant Remittances: The Case of Greece,” International Migration Review, Vol. 31 pp. 72-87.

Lowell, B. Lindsay and Rodolfo O. de la Garza. 2002. “The Development Role of Remittances in U.S. Latino Communities and in Latin American Countries.” In Sending Money Home, edited by R. de la Garza and B. L. Lowell, Rowman & Littlefield Publishers Inc., Oxford, England, pp. 3-27.

Lucas R.E. B., and O. Stark. 1985. “Motivations to remit: evidence from Botswana,” Journal of Political Economy, Vol. 93, No. 5, pp. 901-18.

McFadden, D. 1974. “Conditional Logit Analysis of Qualitative Behavior.” In Zarembka, P. (ed) Frontiers in Econometrics. Academic Press, New York, pp. 105-142.

Menjívar, C. 2002.”The Ties that Heal: Guatemalan Immigrant Women’s Networks and Medical Treatment,” International Migration Review, Vol. 36, No. 2, pp. 437-66.

Menjívar, C., J. DaVanzo, L. Greenwell, and R. Burciaga Valdez. 1998. “Remittance Behavior Among Salvadoran and Filipino Immigrants in Los Angeles,” International Migration Review,Vol. 32, No. 1, pp. 97-126.

Meyers, D. W. 2002. “Migrant Remittances to Latin America: Reviewing the Literature.” In Sending Money Home, edited by R. de la Garza and B. L. Lowell, Rowman & Littlefield Publishers Inc., Oxford, England, pp. 53-81.

Orozco, M. 2001. “Family Remittances to Latin America: The Marketplace and Its Changing Dynamics,” presented at “Remittances as a Development Tool: A Regional Conference,” May 17 & 18, 2001, Inter-American Development Bank, http://www.iadb.org/mif/v2/files/orozco.doc

Orozco, M. 2002a. “Remittances and Markets: New Players and Practices,” in “Sending Money Home.” edited by R. de la Garza and B. L. Lowell, Rowman & Littlefield Publishers Inc., Oxford, England, pp. 125-154.

E2004/06

25

Orozco, M. 2002b. “Attracting Remittances: Market, money and reduced costs,” Report commissioned by the Multilateral Investment Fund of the Inter-American Bank, http://www.iadb.org/mif/v2/files/Orozco2002.ppt

Orrenius, P. M, 1999, Return Migration from Mexico: Theory and Evidence. Ph.D, Dissertation, University of California, Los Angeles.

Reyes, B. 1997, “Dynamics of Immigration: Return Migration to Western Mexico,” Public Policy Institute of California.

Shy, O. and J. Tarkka. 2002. “The Market for Electronic Cash Cards,” Journal of Money Credit and Banking, Vol. 34, No. 2.

Snellman, J. S, J. M. Vesala, and D. B. Humphrey. 2001. “Substitution on Noncash Payment Instruments for Cash in Europe,” Journal of Financial Services Research, Vol. 19, No. 2/3, pp. 131-145.

Sontomero, A. M, and J. J. Seater. 1996. “Alternative Monies and the Demand for Media of Exchange,” Journal of Money, Credit, and Banking, Vol. 28, No. 4, Part 2, pp. 942-960.

Thomson, G. “Migrants to U.S. are a Major Resource for Mexico”, New York Times, March 25, 2002.

Woodruff, Christopher and Rene Zenteno. 2001. “Remittances and Microenterprises in Mexico,” mimeo University of California, San Diego.

Fundación Centro de Estudios Andaluces

26

Table 1 Description of Variables Used in the Analysis

Variable Name Description Mean S.D.

Dependent Variables:

Banks Dummy indicating the use of banks or credit unions to remit 0.0842 0.2777

MTF Dummy indicating the use of non-bank money transfer firms to remit 0.7385 0.4395

Informal Dummy indicating the use of friends, family, or cash in the mail to remit 0.1296 0.3359

Other Dummy indicating the use of other unspecified means to remit 0.0469 0.2113

Independent Variables:

Dollar Amount Sent Dollar amount remitted last month 487.47 408.59 Undocumented Worker Undocumented Worker 0.3293 0.4699 Age Age of respondent 34.4522 10.6838 Male Gender dummy 0.9507 0.2166 Family Size in Mexico Number of family members left in Mexico 2.3207 2.9320 Less than Middle School Less than a middle school education 0.5905 0.4918 Middle School Middle school education 0.2498 0.4329 High School Complemented HS but no college 0.1199 0.3249 More than High School More than a HS education 0.0398 0.1954 Agriculture & Mining Industry dummy 0.3230 0.4676 Industry Industry dummy 0.1226 0.3279 Construction Industry dummy 0.2127 0.4093 Commerce Industry dummy 0.0570 0.2319 Services Industry dummy 0.2842 0.4511 Wage and Salary Worker Wage and salary worker 0.7202 0.4489 Self-employed/Family Worker Self employed or family worker 0.0263 0.1599 Specific-task Worker Worker with a specific-task contract 0.2535 0.4351 Days Residing in the United States Last Number of days in the United States last visit 413 863

Social Networks Friends in host city in the United States 0.8158 0.3877 Sent to Low Standard of Living Areas Standard of living dummy 0.1138 0.3176 Sent to Rural Areas Geographic area dummy 0.3707 0.4830 Sent Remittances for Consumption Subject to remitting, the proportion remitting for consumption 0.7279 0.4451 Sent Remittances for Investment/Saving

Subject to remitting, the proportion remitting to purchase physical/financial assets 0.2494 0.4327

Sent Remittances for Other Purposes Subject to remitting, the proportion remitting for unspecified purposes 0.0228 0.1492

1993 Year dummy 0.1969 0.3977 1994 Year dummy 0.0405 0.1972 1995 Year dummy 0.2401 0.4272 Year 1996 Year dummy 0.1450 0.3522 Year 1997 Year dummy 0.0920 0.2891 Year 1998 Year dummy 0.0743 0.2622 Year 1999 Year dummy 0.1645 0.3707 Year 2000 Year dummy 0.0467 0.2110

E2004/06

27

Table 2 Percent of Individuals Using Each Remitting Method in Each Survey Wave

Survey Wave Banks Non-bank MTFs Informal Other

First Wave (93/94) 3.84 76.57 15.00 4.58

Second Wave (94/95) 5.60 75.20 13.87 5.34

Third Wave (96/97) 8.54 77.08 11.46 2.92

Fourth Wave (98/99) 12.3 69.19 11.42 7.09

Fifth Wave (99/00) 16.81 66.05 11.72 5.42

Fundación Centro de Estudios Andaluces

28

Table 3 Coefficient Estimates from the Tobit Regression for the Dollar Amount Remitted

Variables Coefficient Robust S.E. Marginal Effects

Family Size in Mexico 8.1081*** 1.9767 4.9820 Undocumented Worker -19.1876 12.7356 -11.7483 Age 1.1451* 0.5966 0.7036 Male 142.0638*** 25.1718 81.2721 Middle School 55.8125*** 13.7759 34.8147 High School 91.0783*** 18.1131 58.0693 More than High School 137.0555*** 28.8623 90.0276 Industry 7.1151 18.9707 4.3847 Construction 65.4831*** 15.7488 41.0660 Commerce -14.5562 25.5254 -8.8806 Services -20.5022 14.9916 -12.5384 Self-employed/Family Worker 142.7931*** 34.2309 94.2274 Specific-task Worker -0.7893 12.7330 -0.4848 Days Residing in the United States -0.0054 0.0063 -0.0033 Social Networks -38.6901*** 14.6133 -24.0980 Sent to Rural Areas -17.8377 11.5229 -10.9332 Sent to Low Standard of Living Areas 23.8530 17.1891 14.8037 Sent Remittances for Investment/Saving 57.7214*** 12.9530 36.0326 Sent Remittances for Other Purposes 43.9126 35.7455 27.5998 Year 1994 -60.9492* 30.8273 -36.2933 Year 1995 -13.6204 17.3912 -8.3335 Year 1996 16.7668 19.7748 10.3720 Year 1997 19.7102 22.7707 12.2198 Year 1998 -153.2020*** 22.8652 -87.7051 Year 1999 -191.3838*** 18.3551 -110.3215 Year 2000 -184.4176*** 26.6131 -103.3785

No. Observations 6622 LR Chi2(26) 427.88 Prob>Chi2 0.0000

Notes: The following variables were used as reference categories: sent remittances for consumption purposes, less than middle school, agriculture & mining, wage and salary worker, and the year 1993. *** Signifies statistically different from zero at the 1 percent level or better, **signifies statistically different from zero at the 5 percent level or better and *signifies statistically different from zero at the 10 percent level or better. The marginal effects are computed conditional on remitting a positive amount.

E2004/06

29

Table 4 Relative Risk Ratios for the Multinomial Logit Model of the Likelihood of Sending Remittances by Each

Remitting Method Relative to non-bank MTFs

Banks Informal Other Variables RRR S.E. RRR S.E. RRR S.E.

Predicted Dollar Amount Remitted 1.0002 0.0022 0.9984 0.0017 1.0067*** 0.0026Undocumented Worker 0.6307*** 0.0802 1.1374 0.1093 1.0684 0.1676 Age 0.9922 0.0064 1.0067 0.0051 1.0006 0.0078 Male 1.0217 0.4085 0.9028 0.2846 0.1836*** 0.0835Middle School 1.1027 0.1883 0.9117 0.1258 0.6098** 0.1315 High School 1.3808 0.3405 1.1208 0.2269 0.7537 0.2247 More than High School 1.8621* 0.6809 1.2162 0.3885 0.8455 0.3656 Industry 1.3906** 0.2280 1.0255 0.1442 1.2664 0.2640 Construction 1.0811 0.2232 1.2038 0.1935 0.6384* 0.1627 Commerce 1.0941 0.2485 1.1236 0.2034 1.1693 0.3404 Services 1.2110 0.1781 1.0818 0.1227 1.2864 0.2372 Self-employed/Family Worker 1.4257 0.5981 2.1094** 0.7172 0.7415 0.3595 Specific-task Worker 1.0812 0.1234 1.2498*** 0.1106 1.0267 0.1530 Days Residing in the United States 1.0000 0.0001 0.9999** 0.0001 0.9997** 0.0001 Social Networks 1.4986** 0.2631 1.1705 0.1481 1.4289* 0.2829 Sent to Rural Areas 0.8697 0.0976 1.2132** 0.1051 1.0659 0.1507 Sent to Low Standard of Living Areas 0.9636 0.1550 1.3844*** 0.1674 0.5912** 0.1407 Sent Remittances for Investment/Saving 0.8557 0.1396 1.0765 0.1423 0.8718 0.1711 Sent Remittances for Other Purposes 1.5936* 0.4383 1.7545** 0.4470 1.4789 0.5306 Year 1994 0.4735 0.2354 0.8584 0.1948 1.6886 0.6247 Year 1995 1.4844** 0.2819 0.8651 0.1011 1.3463 0.2570 Year 1996 1.3812 0.2974 0.7760* 0.1072 0.4851*** 0.1283Year 1997 3.1630*** 0.6423 0.6814** 0.1154 0.6206* 0.1824 Year 1998 3.1135*** 1.2737 0.7760 0.2452 6.4531*** 2.9921Year 1999 4.0231*** 1.8863 0.5857 0.2134 3.7047** 2.0544 Year 2000 5.7447*** 2.7160 0.5998 0.2340 3.1713* 1.9132

No. Observations 5979 LR Chi-square (78) 382.23 Log Likelihood -4833.4342 Prob > Chi-square 0.0000

Hausman tests of IIA assumption Ho: Odds (Outcome-J vs Outcome-K) are independent of other alternatives.

Omitted chi2 df P>chi2 evidence _________________________________Banks 1.264 51 1.000 for Ho Informal -2.704 51 1.000 for Ho Other -2.795 51 1.000 for Ho

Notes: non-bank MTFs are used as the reference category in the multinomial logit. The following variables were used as reference categories: sent remittances for consumption purposes, less than middle school, agriculture & mining, wage and salary worker, and the year 1993. *** Signifies statistically different from zero at the 1 percent level or better, **signifies statistically different from zero at the 5 percent level or better and *signifies statistically different from zero at the 10 percent level or better.

Fundación Centro de Estudios Andaluces

centrA:Fundación Centro de Estudios Andaluces

Documentos de Trabajo

Serie Economía

E2001/01 “The nineties in Spain: so much flexibility in the labor market?’’,J. Ignacio García Pérez y Fernando Muñoz Bullón.

E2001/02 “A Log-linear Homotopy Approach to Initialize the ParameterizedExpectations Algorithm’’, Javier J. Pérez.

E2001/03 “Computing Robust Stylized Facts on Comovement’’, Francisco J.André, Javier J. Pérez, y Ricardo Martín.

E2001/04 “Linking public investment to private investment. The case of theSpanish regions”, Diego Martínez López.

E2001/05 “Price Wars and Collusion in the Spanish Electricity Market”, JuanToro y Natalia Fabra.

E2001/06 “Expedient and Monotone Learning Rules”, Tilman Börgers,Antonio J. Morales y Rajiv Sarin.

E2001/07 “A Generalized Production Set. The Production and RecyclingFunction”, Francisco J. André y Emilio Cerdá.

E2002/01 “Flujos Migratorios entre provincias andaluzas y entre éstas y elresto de España’’, J. Ignacio García Pérez y Consuelo GámezAmián.

E2002/02 “Flujos de trabajadores en el mercado de trabajo andaluz’’, J.Ignacio García Pérez y Consuelo Gámez Amián.

E2002/03 “Absolute Expediency and Imitative Behaviour”, Antonio J.Morales Siles.

E2002/04 “Implementing the 35 Hour Workweek by means of OvertimeTaxation”, Victoria Osuna Padilla y José-Víctor Ríos-Rull.

E2002/05 “Landfilling, Set-Up costs and Optimal Capacity”, Francisco J.André y Emilio Cerdá.

E2002/06 “Identifying endogenous fiscal policy rules for macroeconomicmodels”, Javier J. Pérez y Paul Hiebert.

E2002/07 “Análisis dinámico de la relación entre ciclo económico y ciclo deldesempleo en Andalucía en comparación con el resto de España”,Javier J. Pérez, Jesús Rodríguez López y Carlos Usabiaga.

E2002/08 “Provisión eficiente de inversión pública financiada conimpuestos distorsionantes”, José Manuel González-Páramo yDiego Martínez López.

E2002/09 “Complete or Partial Inflation convergence in the EU?”, ConsueloGámez y Amalia Morales-Zumaquero.

E2002/10 “On the Choice of an Exchange Regime: Target Zones Revisited”,Jesús Rodríguez López y Hugo Rodríguez Mendizábal.

E2002/11 “Should Fiscal Policy Be Different in a Non-CompetitiveFramework?”, Arantza Gorostiaga.

E2002/12 “Debt Reduction and Automatic Stabilisation”, Paul Hiebert,Javier J. Pérez y Massimo Rostagno.

E2002/13 “An Applied General Equilibrium Model to Assess the Impact ofNational Tax Changes on a Regional Economy”, M. AlejandroCardenete y Ferran Sancho.

E2002/14 “Optimal Endowments of Public Investment: An EmpiricalAnalysis for the Spanish Regions”, Óscar Bajo Rubio, CarmenDíaz Roldán y M. Dolores Montávez Garcés.

E2002/15 “Is it Worth Refining Linear Approximations to Non-LinearRational Expectations Models?” , Alfonso Novales y Javier J.Pérez.

E2002/16 “Factors affecting quits and layoffs in Spain”, Antonio CaparrósRuiz y M.ª Lucía Navarro Gómez.

E2002/17 “El problema de desempleo en la economía andaluza (1990-2001): análisis de la transición desde la educación al mercadolaboral”, Emilio Congregado y J. Ignacio García Pérez.

E2002/18 “Pautas cíclicas de la economía andaluza en el período 1984-2001: un análisis comparado”, Teresa Leal, Javier J. Pérez yJesús Rodríguez.

E2002/19 “The European Business Cycle”, Mike Artis, Hans-Martin Krolzig yJuan Toro.

E2002/20 “Classical and Modern Business Cycle Measurement: TheEuropean Case”, Hans-Martin Krolzig y Juan Toro.

E2002/21 “On the Desirability of Supply-Side Intervention in a MonetaryUnion”, Mª Carmen Díaz Roldán.

E2003/01 “Modelo Input-Output de agua. Análisis de las relacionesintersectoriales de agua en Andalucía”, Esther Velázquez Alonso.

E2003/02 “Robust Stylized Facts on Comovement for the SpanishEconomy”, Francisco J. André y Javier Pérez.

E2003/03 “Income Distribution in a Regional Economy: A SAM Model”,Maria Llop y Antonio Manresa.

E2003/04 “Quantitative Restrictions on Clothing Imports: Impact andDeterminants of the Common Trade Policy Towards DevelopingCountries”, Juliette Milgram.

E2003/05 “Convergencia entre Andalucía y España: una aproximación a suscausas (1965-1995). ¿Afecta la inversión pública al crecimiento?”,Javier Rodero Cosano, Diego Martínez López y Rafaela PérezSánchez.

E2003/06 “Human Capital Externalities: A Sectoral-Regional Application forSpain”, Lorenzo Serrano.

E2003/07 “Dominant Strategies Implementation of the Critical PathAllocation in the Project Planning Problem”, Juan Perote Peña.

E2003/08 “The Impossibility of Strategy-Proof Clustering”, Javier PerotePeña y Juan Perote Peña.

E2003/09 “Plurality Rule Works in Three-Candidate Elections”, BernardoMoreno y M. Socorro Puy.

E2003/10 “A Social Choice Trade-off Between Alternative FairnessConcepts: Solidarity versus Flexibility”, Juan Perote Peña.

E2003/11 “Computational Errors in Guessing Games”, Pablo Brañas Garzay Antonio Morales.

E2003/12 “Dominant Strategies Implementation when Compensations areAllowed: a Characterization”, Juan Perote Peña.

E2003/13 “Filter-Design and Model-Based Analysis of Economic Cycles”,Diego J. Pedregal.

E2003/14 “Strategy-Proof Estimators for Simple Regression”, Javier PerotePeña y Juan Perote Peña.

E2003/15 “La Teoría de Grafos aplicada al estudio del consumo sectorial deagua en Andalucía", Esther Velázquez Alonso.

E2003/16 “Solidarity in Terms of Reciprocity", Juan Perote Peña.

E2003/17 “The Effects of Common Advice on One-shot Traveler’s DilemmaGames: Explaining Behavior through an Introspective Model withErrors", C. Monica Capra, Susana Cabrera y Rosario Gómez.

E2003/18 “Multi-Criteria Analysis of Factors Use Level: The Case of Waterfor Irrigation", José A. Gómez-Limón, Laura Riesgo y ManuelArriaza.

E2003/19 “Gender Differences in Prisoners’ Dilemma", Pablo Brañas-Garzay Antonio J. Morales-Siles.

E2003/20 “Un análisis estructural de la economía andaluza a través dematrices de contabilidad social: 1990-1999", M. Carmen Lima,M. Alejandro Cardenete y José Vallés.

E2003/21 “Análisis de multiplicadores lineales en una economía regionalabierta", Maria Llop y Antonio Manresa.

E2003/22 “Testing the Fisher Effect in the Presence of Structural Change:A Case Study of the UK", Óscar Bajo-Rubio, Carmen Díaz-Roldány Vicente Esteve.

E2003/23 "On Tests for Double Differencing: Some Extensions and the Roleof Initial Values", Paulo M. M. Rodrigues y A. M. Robert Taylor.

E2003/24 "How Tight Should Central Bank’s Hands be Tied? Credibility,Volatility and the Optimal Band Width of a Target Zone", JesúsRodríguez López y Hugo Rodríguez Mendizábal.

E2003/25 "Ethical implementation and the Creation of Moral Values", JuanPerote Peña.

E2003/26 "The Scoring Rules in an Endogenous Election", Bernardo Morenoy M. Socorro Puy.

E2003/27 "Nash Implementation and Uncertain Renegotiation", PabloAmorós.

E2003/28 "Does Familiar Environment Affect Individual Risk Attitudes?Olive-oil Producer vs. no-producer Households", FranciscaJiménez Jiménez.

E2003/29 "Searching for Threshold Effects in the Evolution of BudgetDeficits: An Application to the Spanish Case", Óscar Bajo-Rubio,Carmen Díaz-Roldán y Vicente Esteve.

E2003/30 "The Construction of input-output Coefficients Matrices in anAxiomatic Context: Some Further Considerations", Thijs ten Raay José Manuel Rueda Cantuche.

E2003/31 "Tax Reforms in an Endogenous Growth Model with Pollution",Esther Fernández, Rafaela Pérez y Jesús Ruiz.

E2003/32 "Is the Budget Deficit Sustainable when Fiscal Policy isnonlinear? The Case of Spain, 1961-2001", Óscar Bajo-Rubio,Carmen Díaz-Roldán y Vicente Esteve.

E2003/33 "On the Credibility of a Target Zone: Evidence from the EMS",Francisco Ledesma-Rodríguez, Manuel Navarro-Ibáñez, JorgePérez-Rodríguez y Simón Sosvilla-Rivero.

E2003/34 "Efectos a largo plazo sobre la economía andaluza de las ayudasprocedentes de los fondos estructurales: el Marco de ApoyoComunitario 1994-1999", Encarnación Murillo García y SimónSosvilla-Rivero.

E2003/35 “Researching with Whom? Stability and Manipulation”, JoséAlcalde y Pablo Revilla.

E2003/36 “Cómo deciden los matrimonios el número óptimo de hijos”,Francisca Jiménez Jiménez.

E2003/37 “Applications of Distributed Optimal Control in Economics. TheCase of Forest Management”, Renan Goetz y Angels Xabadia.

E2003/38 “An Extra Time Duration Model with Application toUnemployment Duration under Benefits in Spain”, José MaríaArranz y Juan Muro Romero.

E2003/39 “Regulation and Evolution of Harvesting Rules and Compliance inCommon Pool Resources”, Anastasios Xepapadeas.

E2003/40 “On the Coincidence of the Feedback Nash and StackelbergEquilibria in Economic Applications of Differential Games”,Santiago J. Rubio.

E2003/41 “Collusion with Capacity Constraints over the Business Cycle”,Natalia Fabra.

E2003/42 “Profitable Unproductive Innovations”, María J. Álvarez-Peláez,Christian Groth.

E2003/43 “Sustainability and Substitution of Exhaustible NaturalResources. How Resource Prices Affect Long-Term R&D-Investments”, Lucas Bretschger, Sjak Smulders.

E2003/44 “Análisis de la estructura de la inflación de las regionesespañolas: La metodología de Ball y Mankiw”, María ÁngelesCaraballo, Carlos Usabiaga.

E2003/45 “An Empirical Analysis of the Demand for Physician ServicesAcross the European Union”, Sergi Jiménez-Martín, José M.Labeaga, Maite Martínez-Granado.

E2003/46 “An Exploration into the Effects of Fiscal Variables on RegionalGrowth”, Diego Martínez López.

E2003/47 “Teaching Nash Equilibrium and Strategy Dominance: AClassroom Experiment on the Beauty Contest”. Virtudes AlbaFernández, Francisca Jiménez Jiménez, Pablo Brañas Garza,Javier Rodero Cosano.

E2003/48 "Environmental Fiscal Policies Might be Ineffective to ControlPollution", Esther Fernández, Rafaela Pérez y Jesús Ruiz.

E2003/49 "Non-stationary Job Search When Jobs Do Not Last Forever: AStructural Estimation to Evaluate Alternative UnemploymentInsurance Systems", José Ignacio García Pérez.

E2003/50 “Poverty in Dictator Games: Awakening Solidarity”, PabloBrañas-Garza.

E2003/51 “Exchange Rate Regimes, Globalisation and the Cost of Capital inEmerging Markets” Antonio Díez de los Ríos.

E2003/52 “Opting-out of Public Education in Urban Economies”. FranciscoMartínez Mora.

E2004/01 “Partial Horizontal Inequity Orderings: A non-parametricApproach”, Juan Gabriel Rodríguez, Rafael Salas, Irene Perrote.

E2004/02 “El enfoque microeconómico en la estimación de la demanda detransporte de mercancías. Análisis desde una perspectivaregional”, Cristina Borra Marcos, Luis Palma Martos.

E2004/03 “El marco del SEC95 y las matrices de contabilidad social:España 19951”, M. Alejandro Cardenete, Ferrán Sancho.

E2004/04 “Performing an Environmental Tax Reform in a RegionalEconomy. A Computable General Equilibrium Approach”,Francisco J. André, M. Alejandro Cardenete, E. Velázquez.

E2004/05 “Is the Fisher Effect Nonlinear? Some Evidence for Spain, 1963-2002”, Óscar Bajo-Rubio, Carmen Díaz-Roldán, Vicente Esteve.

E2004/06 “On the Use of Differing Money Transmission Methods byMexican Immigrants”, Catalina Amuedo-Dorantes, Susan Pozo.

centrA:Fundación Centro de Estudios Andaluces

Normas de publicación de Documentos de TrabajocentrA Economía

La Fundación Centro de Estudios Andaluces (centrA) tiene como uno de sus objetivosprioritarios proporcionar un marco idóneo para la discusión y difusión de resultadoscientíficos en el ámbito de la Economía. Con esta intención pone a disposición de losinvestigadores interesados una colección de Documentos de Trabajo que facilita latransmisión de conocimientos. La Fundación Centro de Estudios Andaluces invita a lacomunidad científica al envío de trabajos que, basados en los principios del análisiseconómico y/o utilizando técnicas cuantitativas rigurosas, ofrezcan resultados deinvestigaciones en curso.

Las normas de presentación y selección de originales son las siguientes: 1. El autor(es) interesado(s) en publicar un Documento de Trabajo en la serie de

Economía de centrA debe enviar su artículo en formato PDF a la dirección de email:[email protected]

2. Todos los trabajos que se envíen a la colección han de ser originales y no estarpublicados en ningún medio de difusión. Los trabajos remitidos podrán estarredactados en castellano o en inglés.

3. Los originales recibidos serán sometidos a un breve proceso de evaluación en elque serán directamente aceptados para su publicación, aceptados sujetos arevisión o rechazados. Se valorará, asimismo, la presentación de¡ trabajo enseminarios de centrA.

4. En la primera página deberá aparecer el título del trabajo, nombre y filiación delautor(es), dirección postal y electrónica de referencia y agradecimientos. En estamisma página se incluirá también un resumen en castellano e inglés de no más de100 palabras, los códigos JEL y las palabras clave de trabajo.

5. Las notas al texto deberán numerarse correlativamente al pie de página. Lasecuaciones se numerarán, cuando el autor lo considere necesario, con númerosarábigos entre corchetes a la derecha de las mismas.

6. La Fundación Centro de Estudios Andaluces facilitará la difusión electrónica de losdocumentos de trabajo. Del mismo modo, se incentivará económicamente suposterior publicación en revistas científicas de reconocido prestigio.