The current issue and full text archive of this journal is available on Emerald Insight at:

www.emeraldinsight.com/0306-8293.htm

Motives
Motives behind remittances behind
A comparative analysis of single- and remittances
multiple-migrant households in India
Navita Pal and Rama Pal 775
Department of Humanities and Social Sciences,
Indian Institute of Technology Bombay, Mumbai, India Received 6 September 2018
Revised 8 January 2019
Accepted 11 February 2019
Abstract
Purpose – The purpose of this paper is to empirically investigate whether the motives behind sending
remittances differ for the migrants from single- and multi-migrant families in India.
Design/methodology/approach – The paper uses the second round of India Human Development Survey
conducted in the year 2012 to analyse migrant’s motives using the two-part model and the sample selection
model (SSM). Here, the probit model is used to model whether the migrant sends the remittances and then the
amount of remittances is modelled using the ordinary least squares method. The SSM assumes that these two
decisions are interdependent. This migrant-level study is the first to segregate single- and multiple-migrant
settings and compare them for the existence of altruism or inheritance motives.
Findings – The findings indicate altruism as the main motive behind remittances for the single migrants in India.
On the other hand, presence of inheritance motive is evident along with altruism in the multi-migrant setting.
Research limitations/implications – The analysis is restricted due to limited availability of information
on migrant as the data are collected from the migrant’s family at origin.
Originality/value – This is the first study to use a unique nationally representative survey which enables it to
point out differences in the motives behind remittances for the single-migrant and multi-migrant settings in India.
Keywords India, Altruism, Inheritance, Migrant remittances
Paper type Research paper

1. Introduction
In recent years, remittances[1] emerge as one of the main sources of external finance in
developing countries. In 2017, India was the largest recipient of remittances among the
developing countries with US$69bn estimated inflows of remittances (World Bank, 2018).
The share of remittances inflow in India was 2.8 percentage of India’s gross domestic
product in the same year (World Bank, 2018). At the household level, remittances sent by the
migrants may form significant source of financing health and education. For instance,
Mahapatro et al. (2017) showed that Indian households receiving remittances spend more on
health and education as compared to others. In order to understand who benefits from
remittances and the long term sustainability of such remittances, we have to first
understand the micro-economic structure behind them. Given this backdrop, the study
examines motives[2] behind remittances using data for Indian households. For this purpose,
we consider all the migrants from these households who have migrated inside and outside
India. The paper tries to understand whether the motives differ for two settings: when there
is a single migrant and when there are multiple migrants from the family.
The literature based on New Economics of Labour Migration considers remittances as an
outcome of either migrant’s individual decision making or a contractual arrangement
between the migrant and household. Under the individualistic framework, motives behind
remittances are categorised as selfless and self-interest motives. Lucas and Stark (1985) and
Cox (1987) called selflessness as altruism. Here, the selfless care for the family encourages
International Journal of Social
the migrant to send remittances. Lucas and Stark (1985) and Cox (1987) suggested that, Economics
under altruism, the remittances increase with the increase in migrant’s income but decline Vol. 46 No. 6, 2019
pp. 775-791
© Emerald Publishing Limited
0306-8293
The authors would like to thank the editor and two anonymous referees for insightful comments. DOI 10.1108/IJSE-09-2018-0444
IJSE with rise in household’s income. Lucas and Stark (1985) further validated this claim based
46,6 on the findings from Botswana[3].
Under the self-interest motives, the migrant remits to gain certain benefits from
household (Lucas and Stark, 1985). These benefits may be of different forms, such as,
inheritance, and exchange benefits to the migrant. The migrant’s aspiration to inherit
familial assets or a threat from household to disinherit the migrant may increase the
776 remittances (Lucas and Stark, 1985; Hoddinott, 1994)[4]. In this context, Hoddinott (1992,
1994) also reported that sons with more aspiration to inherit familial assets significantly
remit larger amount to Kenyan families than daughters.
Apart from inheritance motive, the migrant may send remittances in order to get some
services from the family. Lucas and Stark (1985) called this as exchange motive. Lucas and Stark
(1985) and Cox (1987) presented the exchange motive as an informal contract between the
migrant and household. Under this contract, household provides its services to the migrant and in
return the migrant sends remittances. These services may include taking care of migrant’s
left-behind familial assets and relatives at home. The remittances may be considered as price paid
of these services and the price will go up with the opportunity cost of the recipient household.
Here, the opportunity cost is the income lost while providing the above mentioned services to the
migrant. Hoddinott (1994), Cox et al. (1998) and De la Briere et al. (2002), based on their studies on
Kenya, Peru and Dominican Sierra, respectively, provided empirical support for this theory
where they find positive relation between remittances and the household’s pre-transfer income.
In addition to informal contract, remittances may result due to mutually beneficial
pre-migration contracts between the migrant and household (Lucas and Stark, 1985; Stark
and Lucas, 1988). There may be two broad types of implicit contracts. Under the first type of
contract, the household sends migrant out to hedge the geographical risk originated in
primary productive activities, such as, crop failure and cattle disease (Stark and Levhari,
1982). Here, the household bears the cost of migration and initial set-up for the migrant.
Migrant in turn provides remittances in the event of financial catastrophe at home. Stark
and Lucas (1988), Rosenzweig (1988) and Akobeng (2017), based on evidence from
Botswana, India and Ghana, respectively, showed that families facing risky encounters in
primary productive activities receive remittances as insurance.
Under the second type of contract, the household invests in education of the migrant and
expects repayment from the migrant in the form of remittances. Good educational
attainment of the migrant results in lucrative earnings and hence is associated with larger
remittances. In this context, Lucas and Stark (1985) and Poirine (1997) suggested a positive
relationship between migrant’s education and the amount of remittances[5]. Regmi and
Tisdell (2002), based on the study on Nepalese migrants, reported a positive relationship
between the amount of remittances and the education level of the household’s own young
migrant as compared to other family members.
Most of the above mentioned studies consider the migrant and household as single
entities. Recently, studies deviate from this formulation and distinguish between two types
of settings: first, where there is only one migrant from the family and second, where there
are multiple migrants from the family. Studies point out that in the multiple-migrant case,
the individual migrants are likely to remit less due to collective sharing of family burden by
all migrants (Agarwal and Horowitz, 2002; Bouoiyour and Miftah, 2015) and less likelihood
of getting inheritance ( Jena, 2016). Apart from these effects, the migrants may also compete
with each other for getting higher share in the inheritance and as a result send more
remittances to increase their own share (Garip et al., 2015).We build on this literature and
test the presence of altruism and inheritance motives in the context of India.
Studies on remittances in India mostly focus either on the flow of remittances (Zachariah and
Rajan, 2007; Tumbe, 2011) or the determinants and implications of remittances at household level
(Bhagat et al., 2013; Mohanty et al., 2014; Parida et al., 2015; Mahapatro et al., 2017). As far as
remittance behaviour of migrants is concerned, there are a limited number of studies for India. Motives
Mitra (2004), based on the study conducted in Delhi-slum area of India, provided evidence of the behind
presence of exchange motive behind sending remittances. Czaika and Spray (2013) and remittances
Mahapatro (2017) examined the motives using a nationally representative survey for India, where
Czaika and Spray (2013) examined the trajectory of remittances over migrant’s time of absence
from home and Mahapatro (2017) focused on the gender differences in remittance behaviour.
We extend this literature by differentiating between the single- and multiple-migrant 777
settings while analysing the remittances. Moreover, the previous studies do not use many
of the relevant migrant characteristics, such as, education, and ties with the household,
due to data limitation. To overcome this obstacle, we use an alternative nationally
representative survey, namely, the India Human Development Survey (IHDS) (Desai et al.,
2015) for the year 2012. This data set allows us to include the above mentioned variables
in the analysis and infer about the presence of altruism and inheritance motives.
The empirical findings provide evidence for altruistic motive behind sending remittances
for the single-migrant. On the contrary, for the multi-migrant case, results show presence
of inheritance and only weak evidence on altruism.
The scheme of this paper is as follows: Section 2 provides the description of data,
variables and descriptive statistics. Section 3 describes the econometric methodology.
Section 4 presents the empirical findings. Finally, Section 5 concludes.

2. Data and variables

The study is based on the IHDS. It is a nationally representative, multi-topic longitudinal
survey of households conducted jointly by NCAER and the University of Maryland.
The data cover all the states and union territories of India (Desai et al., 2009; Desai and
Vanneman, 2015), except for Andaman Nicobar and Lakshadweep. We use data from the
second round of this survey (IHDS-II) collected from November 2011 to October 2012.
The present study examines net remittances sent by the migrant to the household of
origin. The information on the non-resident members of the household is provided by the
household members in the data set. We take these non-resident members as the migrants
from the household. The remittances sent by and to these migrants are reported in the data
set. We calculate the net remittances by subtracting the amount sent by the household to the
migrant from the remittances by the migrant to the household[6]. These net remittances are
used to understand differences in motives behind remittances in single-migrant and
multi-migrant settings. Moreover, since children are not expected to send remittances we
concentrate on the migrants above 14 years of age[7].
We consider two different dependent variables, namely, probability of migrant sending
remittances and the amount of remittances sent by the migrant. As literature points out both
household and migrant-level variables affect remittance behaviour. Table I provides the precise
definitions of the explanatory variables considered in the analysis. It also presents the expected
direction of relationship between these variables and remittances; and thus, indicates implied
underlying motives as suggested by the theoretical models. Table II presents the summary
statistics of the explanatory variables in two different settings: single- and multiple-migrant
cases. We discuss below the expected relation between the explanatory variables.

Household characteristics
The theoretical models suggest that the household’s economic status and the need for
additional resources determine remittances from the migrants. To empirically test the relation
between economic status and remittances, we include household income, assets, education of
head and employment status of head in the analysis. On the other hand, to capture the need or
dependency of the household, we consider the following variables: number of migrants, gender
of the head and dependency ratio. Moreover, if the migrant shares close ties with the household
IJSE Expected
46,6 Description sign Motive

Dependent variables
Incidence of ¼ 1 if the migrant sends positive net remittances to
remittances the household; ¼ 0 otherwise
Remittancesa The natural log of the net remittances (in Rs) sent by
778 the migrant to the household in the last 12 months
Explanatory variables
Household characteristics
Per capita 5 quintiles of the per capita pre-transfer income of (+ve)/(−ve) (+ve) exchange and
pre-transfer the household (−ve) altruism
income group
Farm land The natural logarithm of total agricultural farm land (+ve) Inheritance
owned by the household (in acres)
Own house ¼ 1 if the household owns a house; ¼ 0 otherwise (+ve) Inheritance
Dependency ratio Ratio of the number of dependents in the household (+ve) Altruism
by the total number of persons in a household. Here,
dependents include the number of children (under 15
years of age) and elderly persons (above 60 years of
age) in the household
Number of It is the total number of migrantsb from the given (+ve)/(−ve) (+ve) inheritance
migrants household and (−ve) altruismc
Rural household ¼ 1 if the household is from rural area; ¼ 0 otherwise (+ve) Altruism
EAG states ¼ 1 if the household belongs to relatively less (+ve) Altruism
developed states of Uttarakhand, Rajasthan,
Uttar-Pradesh, Bihar, Jharkhand, Orissa,
Chhattisgarh and Madhya Pradesh; ¼ 0 otherwise
Gender of the ¼ 1 if the head of the household is a male and 0 if the (−ve) Altruism
head head is a female
Education of the ¼ 1 if the head has studied beyond primary (5th (−ve) Altruism
head class) education; ¼ 0 otherwise
Employed head ¼ 1 if the head is employed; ¼ 0 otherwise (−ve) Altruism
Migrant’s characteristics
Education of the ¼ 1 if the migrant has completed at least secondary (+ve) Altruism
migrant education; ¼ 0 otherwise
Occupation of the migrant
Unemployed ¼ 1 if the migrant is unemployed; ¼ 0 otherwise (−ve)
(base)
Student ¼ 1 if the migrant is a student; ¼ 0 otherwise (−ve)
Labourer ¼ 1 if the migrant is a labourer; ¼ 0 otherwise (+ve) Altruism
Professional ¼ 1 if the migrant is a professional; ¼ 0 otherwise (+ve) Altruism
Age of the migrant Age of the migrant (in years) (+ve) Altruism
Age square of the Square of the age of the migrant (−ve)
migrant
Destination of the migrant
Same state (base) ¼ 1 if the migrant is living in the same state; ¼ 0 (−ve)
otherwise
Outside state ¼ 1 if the migrant is living outside the state; ¼ 0 (+ve) Altruism
otherwise
Outside India ¼ 1 if the migrant is living outside the country; ¼ 0 (+ve) Altruism
otherwise
Table I.
Description of
variables (continued )
 Expected
Motives
Description sign Motive behind
remittances
Relationship ties with the household
Single (Base) ¼ 1 if the migrant is single and childrend are not (−ve)
staying behind in the household; ¼ 0 otherwise
Married with no ¼ 1 if the migrant is married but neither spouse nor (−ve)
spouse or children are staying behind in the household; ¼ 0 779
children otherwise
Only children ¼ 1 if migrant’s children are staying behind in the (+ve) Altruism
household and spouse is not staying in the
household; ¼ 0 otherwise
Only spouse ¼ 1 if only the spouse (without children) is staying (+ve) Altruism
behind in the household; ¼ 0 otherwise
Spouse and ¼ 1 if the spouse and children of the migrant are (+ve) Altruism
children staying behind in the household; ¼ 0 otherwise
Gender of the ¼ 1 if the migrant is a male; ¼ 0 if the migrant is a (+ve) Inheritance
migrant female
Notes: aThe variable is defined only for migrants with net remittances strictly greater than 0; bsince the study
is restricted to the migrants aged 15 years or more, the number of migrants in the household does not include the
migrants below that age; calthough the amount of remittances are likely to go down, the probability of sending
remittances may go up as all the migrants, under altruism, share the burden and hence send the remittances;
d
here children refers to the sons and daughters of the migrant who are under 15 years of age Table I.

then remittances are more likely. To account for this factor, we include whether the migrant’s
spouse and children are staying back in the family. We elaborate on these variables below.
Under altruism, the household’s pre-transfer income negatively affects remittances (Lucas
and Stark, 1985). On the other hand, under the exchange motive, household income is
positively related with remittances (Lucas and Stark, 1985; Cox et al., 1998). We include the
five quintile groups of the household’s per capita income in the multivariate analysis. The
descriptive statistics show that the proportion of migrants sending remittances is lower for the
higher income groups for both single- and multiple-migrant settings (Table III). On the other
hand, the average remittances are higher for the richer income groups. The multivariate
analysis will show whether these associations are statistically significant.
Apart from income, household’s physical assets determine remittances under the
inheritance motive (Hoddinott, 1992; Regmi and Tisdell, 2002). To empirically test this motive,
we include household’s ownership of agriculture farm land and house. The literature suggests
that, under inheritance motive, both the probability of sending remittances and amount of
remittances increase with the increase in the physical assets of the household. The descriptive
statistics do not provide any clear indication of the presence of inheritance motive (Table III).
The number of migrants from a given family affects the remittance behaviour of
individual migrant. The literature forecasts negative relation between the number of
migrants and remittances under altruism due to shared burden of the household (Agarwal
and Horowitz, 2002; Bouoiyour and Miftah, 2015). However, under inheritance, the presence
of large number of migrants may positively influence the remittances, at least initially, due
to competition to acquire larger share in the familial assets (Hoddinott, 1994; De la Briere
et al., 2002; Garip et al., 2015). At the same time, more heirs may indicate less share in familial
assets and thus less incentive to remit.
Apart from these variables, we also include other indicators of household’s economic
status and need of remittances, namely, gender, education and employment status of head
and dependency ratio. We capture the location specific variations by including indicator
variables for the households residing in the rural areas and states belonging to the
Empowered Action Group (EAG).
IJSE Single migrant Multi migrant
46,6 Mean SD Mean SD

Household characteristics
Per capita pre-transfer income 26,714.970 51,707.290 34,990.780 70,921.460
Farm land 4.223 12.124 7.125 39.492
With own house 0.935 0.246 0.953 0.210
780 Dependency ratio 0.401 0.294 0.410 0.293
Rural household 0.753 0.430 0.817 0.386
EAG states 0.563 0.496 0.656 0.475
Gender of the head 0.719 0.449 0.735 0.440
Education of the head 0.426 0.494 0.399 0.489
Employed head 0.704 0.456 0.691 0.461
Migrant’s characteristics
Age of the migrant 33.061 15.039 30.670 13.056
Education of the migrant 0.485 0.499 0.474 0.499
Occupation of the migrant
Unemployed (Base) 0.090 0.286 0.104 0.305
Student 0.165 0.372 0.165 0.371
Labourer 0.565 0.495 0.571 0.494
Professional 0.177 0.382 0.159 0.365
Destination of the migrant
Same state (base) 0.555 0.496 0.493 0.500
Outside state 0.378 0.485 0.452 0.497
Outside India 0.066 0.249 0.053 0.224
Relationship ties with the household
Single (base) 0.418 0.493 0.392 0.488
Married (no spouse and children) 0.207 0.405 0.290 0.454
Only children 0.019 0.139 0.080 0.272
Only spouse 0.121 0.326 0.094 0.292
Spouse and children 0.233 0.423 0.141 0.348
Gender of the migrant 0.861 0.345 0.832 0.373
Table II. Observations 6,361 6,033
Descriptive statistics Notes: aDescriptive statistics is reported for original variables without any categorisation or transformation.
of explanatory For this estimation, we consider only the migrant sample
variablesa Source: Estimation based on IHDS-II survey for 2011–2012

Migrant characteristics
Since the data set does not provide information on income of the migrant, we infer about the
migrant’s economic status from migrant’s education and occupation. Education is generally
positively related with income and thus may be considered as a good proxy for the economic
status. In order to capture the effect of education, we include a dummy variable taking value
1 if the migrant has achieved at least secondary education and 0 otherwise. We expect
positive impact of education on remittances under altruism (Lucas and Stark, 1985; Poirine,
1997). However, the descriptive statistics show that the proportion of migrants sending
remittances is lower for the migrants who have completed secondary education as
compared to the others. For instance, in the multiple-migrant setting, only 51.49 per cent of
the migrants with the secondary education send remittances as opposed to 64.24 per cent of
the migrants with lower education (Table III).
We also consider occupation of the migrant as another indicator of the income.
Professionals, such as doctors, lawyers and engineers, are expected to have higher income
than labourers. Moreover, students are likely to be different than other non-working
persons. As a result, we include four different categories, namely, non-working persons[8]
excluding students, students, professionals and labourers. The preliminary analysis
 Incidence of remittances Average remittances
Motives
Single migrant Multi migrants Single migrant Multi migrants behind
remittances
All 58.88 58.19 37,368.68 29,520.62
Household characteristics
Per capita pre-transfer income
1st quintile (base) 79.57 70.34 34,362.20 22,806.53 781
2nd quintile 62.42 66.91 26,905.25 24,100.22
3rdquintile 53.30 61.62 32,787.53 23,738.57
4th quintile 44.32 56.75 42,317.60 31,064.73
5th quintile 33.96 37.56 67,172.30 55,101.38
Farm landa
No land (base) 61.52 60.08 37,063.62 26,837.02
Above 0 to 2.5 acres 65.69 64.76 29,361.21 23,878.27
Above 2.5 to 5 acres 59.91 64.47 41,142.41 29,607.14
Above 5 acres 46.83 48.71 44,294.18 38,357.73
Own house
Without own house (base) 38.10 36.23 36,764.87 32,039.83
With own house 60.32 59.26 37,395.09 29,445.38
Migrant characteristics
Education of the migrant
Below secondary (base) 63.66 64.24 23,939.10 19,256.49
Secondary and above 53.83 51.49 54,169.95 43,696.20
Occupation of the migrant
Unemployed (base) 18.22 22.84 24,168.00 18,013.36
Student 2.16 2.35 38,223.07 13,187.98
Labourers 78.50 77.11 33,303.46 26,162.82
Professional 70.08 71.55 53,567.93 45,477.92
Age of the migrant
0–20 24.77 31.21 15,190.08 15,817.89
20–40 68.01 63.95 38,128.44 29,675.84
40–65 71.76 72.68 43,254.89 35,429.93
65 above 23.61 14.76 17,082.69 18,248.8
Destination of the migrant
Same state (base) 45.98 40.37 27,384.72 21,013.77
Outside state 73.66 74.32 33,764.34 26,610.32
Outside India 82.50 86.39 102,183.70 87,830.53
Relationship ties with the household
Single 42.77 46.13 30,057.12 27,909.22
Married (no spouse and children) 64.63 64.13 24,461.59 19,357.61
Only children 43.48 26.43 10,722.93 11,819.83
Only spouse 51.08 66.40 51,670.49 51,337.67
Both spouse and children 88.00 92.10 48,975.12 38,701.55
Gender of the migrant
Female migrant (base) 21.37 19.00 32,999.75 14,950.39
Male migrant 64.93 66.07 37,600.43 30,363.89
Table III.
Observations 6,361 6,033 3,552 3,258
Incidence of
Notes: aWe have used discrete classes in the descriptive statistics for family land holding to show how remittances (in
incidence and amount of remittances are changing with land holding. However, in the multivariate analysis percentages) and
we have used a continuous variable, i.e., the natural logarithm of land holding average remittances
Source: Estimation based on IHDS-II survey for 2011–2012 (in Rupees)

presented in Table III shows that the average remittances sent by the professionals are the
highest (Rs53,568 and Rs45,478 for single and multiple-migrants settings, respectively)
among the four categories. However, the incidence of sending remittances is the highest
among labourers, with 78 and 77 per cent of the labourers sending remittances in the above
mentioned two settings.
IJSE Migrant’s age can affect remittances. The remittances may initially increase with the age
46,6 of the migrant. For instance, Bouoiyour and Miftah (2015) showed that young migrants send
more remittances to the household. However, they are likely to fall afterwards. Table III
shows the similar pattern when we consider the proportion of migrant’s sending remittances
and amount of remittances. To capture this effect, we include migrant’s age and its square in
the multivariate analysis.
782 Strong ties with the family are likely to increase remittances due to altruism.
Following the literature we include two types of variables; first indicating presence of
close relatives in the household (Brown, 1997; Vanwey, 2004) and second showing how
close the migration location is from the location of origin (Vanwey, 2004). We consider
spouse and children as close relatives in the present analysis[9]. Table III shows that if
both spouse and children are at home then the proportion of migrants sending remittance
is the highest at 88 and 92 per cent for single and multiple-migrants settings, respectively.
Second, to understand closeness of the migration location, we include three dummy
variables for migration in the same state, migration in any other state of India and
migration abroad. The descriptive statistics indicate that if the location is farther away
from the household then the incidence of remitting and average remittances sent go up
(Table III).
The literature shows that the pattern of remittances varies across the gender of migrant
due to both altruism and inheritance motive. While sons show more aspiration to inherit
familial assets and thus, remit larger than daughters (Lucas and Stark, 1985; Hoddinott,
1992, 1994); daughters are more altruistic and insure parents against heath shocks De la
Briere et al. (2002). In the Indian context, the descriptive statistics suggest that, on average,
male migrants tend to send more remittances (Table III).
The next section elaborates on the econometric methodology that we use to empirically
validate these preliminary expectations about the presence of altruism and inheritance
motives in the Indian context.

3. Econometric methodology
In this paper, we examine the determinants of the remittances and indicate possible
motives behind sending remittances in the context of India. The literature uses two
approaches to empirically test the presence of various motives, first, the two-part model
(TPM) and second, the Heckman’s sample selection model (SSM) proposed by Heckman
(1979). Both the approaches separate two decisions, namely, whether to send the
remittances and the amount of remittances. The TPM assumes that the two decisions
are independent of each other and the migrant separately decides on whether to send the
remittances and the amount. The SSM assumes that these two decisions are correlated.
In particular, there is sample selection in the second decision as the amount of remittances
is only observed if the migrant decides to send the positive remittances at the first stage.
There is a long and inconclusive debate in the literature regarding the choice between
these two models (for instance, Manning et al., 1987; Leung and Yu, 1996). The literature
on motives behind remittances also uses both SSM and TPM ( Jena, 2016) to empirically
test the presence of motives. As a result, we use both the models for estimation.
We elaborate on each of these two methods below.
In the TPM, the probability of sending remittances and the amount of remittances are
modelled separately. The reason behind this separate estimation is that the determinants
for the two dependent variables are likely to be different. So, the TPM assumes that the
migrant’s decisions about whether to send remittances and the amount of remittances to
be sent are not connected. First the migrant decides whether to send the remittances and
then, given that he/she decides to remit, determine the amount that should be sent.
 So, in the first part, we use the probit regression to estimate the probability of the Motives
migrant sending positive net remittances[10]. Here, the latent function that determines behind
whether the migrant sends remittances is given as follows: remittances
Rni ¼ bXi þei : (1)
And the probability of positive remittances is given by the equation as follows:
783
ProbðI i ¼ 1jXi Þ ¼ FðXi bÞ; (2)

here, Ii is the indicator variable for positive remittances which equals 1 if Rni 4 0, Xi the
vector of explanatory variables and β the vector of parameters, Φ the cumulative normal
density and ei the error term.
In the second part, we use the OLS method to estimate the following model for the
determinants of the amount of remittances sent:
Ri ¼ gXi þui : (3)
The variable of our interest, Ri, is the logarithm of net remittances sent by the migrant to the
household, Xi the vector of regressors, γ the vector of parameters and ui is the random error
term following normal distribution. It may be noted that we estimate the Equation (3) only
for the migrants who send some remittances and thus, the migrants who do not send
remittances to the household are not considered in the second part.
It may also be noted that the second decision about the amount of remittances may be
affected due to the sample selection bias. Considering this, we also estimate the SSM for the
possible dependence. This dependence between the two equations, Equations (1) and (3), is
incorporated by assuming that the error terms ei and ui follow bivariate normal distribution
with mean 0, varðeÞ ¼ s21 ; varðuÞ ¼ s22 ; and cov (e, u) ¼ ρ. This correlation warrants joint
estimation of the two equations and thus, we use the maximum likelihood method for this
purpose[11]. Moreover, estimation requires presence of an additional variable, instrument, in
the first stage. The instrument that is used in the present study is the proportion of other
households in the village receiving remittances out of all the migrant households. We check
for the possible correlation of this with the second stage dependent variable and find that
this instrument does not have significant impact on the amount of remittances sent for both
the single- and multiple-migrant scenario. The next section elaborates on the empirical
findings of these two models.

4. Empirical findings
The empirical findings suggest that the remittances to the single-migrant household in India
are mostly influenced by altruism. On the other hand, both, altruism and inheritance
motives affect remittances in the case of multi-migrant households. The results are
qualitatively similar from both the TPM (Table IV ) and the SSM (Table V ). The results of
the SSM indicate the presence of sample selection for the multiple-migrant case. However,
for the single-migrant scenario, we did not find any sample selection. We discuss the
empirical findings below in detail.

Household characteristics
The sole migrant is more likely to send remittances if the recipient household is from lower
income groups, has a female head, and has an unemployed head as compared to their
respective counterparts. Also, the probability of sending remittances for the sole migrant
goes up with the number of dependents in the household of origin. All these household
characteristics suggest the presence of altruistic motive. At the same time, if the household
IJSE Single migrant Multi migrant
46,6 Probit Probit
(marginal (marginal
Explanatory variables effects) p-value OLS p-value effects) p-value OLS p-value

Household characteristics
Per capita pre-transfer income
784 2nd quintile −0.047 0.009 −0.054 0.399 0.016 0.359 0.492 0.000
3rd quintile −0.067 0.000 0.116 0.090 0.001 0.956 0.546 0.000
4th quintile −0.089 0.000 0.213 0.003 −0.012 0.622 0.651 0.000
5th quintile −0.134 0.000 0.315 0.000 −0.072 0.000 0.876 0.000
Farm land −0.0001 0.434 0.0004 0.291 −0.0004 0.005 0.001 0.008
With own house 0.045 0.058 0.165 0.130 0.064 0.012 0.388 0.041
No. of migrants – – – – 0.011 0.079 −0.081 0.005
Dependency ratio 0.046 0.037 −0.343 0.000 −0.025 0.259 −0.163 0.052
Rural household 0.026 0.072 −0.119 0.026 0.049 0.003 −0.218 0.002
EAG state −0.042 0.000 −0.039 0.435 −0.054 0.000 0.146 0.011
Head’s characteristics
Gender of the head (Male) −0.060 0.000 −0.063 0.253 −0.010 0.497 0.069 0.328
Education of the head
(beyond primary
education) −0.011 0.356 0.170 0.001 −0.042 0.003 0.147 0.061
Employed head −0.074 0.000 −0.294 0.000 −0.048 0.001 −0.329 0.000
Migrant’s characteristics
Education of the migrant
(secondary) 0.050 0.000 0.461 0.000 0.010 0.458 0.349 0.000
Occupation of the migrant
Student −0.329 0.000 −0.422 0.414 −0.337 0.000 −1.065 0.031
Labourers 0.223 0.000 0.278 0.102 0.166 0.000 0.007 0.952
Professional 0.192 0.000 0.448 0.008 0.168 0.000 0.134 0.282
Age of the migrant 0.011 0.000 0.022 0.033 0.020 0.000 −0.001 0.902
Age square of the
migrant −0.0001 0.000 −0.0002 0.019 −0.0002 0.000 0.00005 0.754
Destination of the migrant
Outside state 0.067 0.000 0.149 0.001 0.106 0.000 0.254 0.000
Outside India 0.092 0.000 0.973 0.000 0.139 0.000 1.136 0.000
Relationship ties with the household
Married (no spouse and
children) −0.043 0.027 −0.391 0.000 −0.051 0.007 −0.390 0.000
Only children −0.206 0.000 −1.124 0.000 −0.262 0.000 −0.728 0.000
Only spouse −0.115 0.000 0.269 0.005 −0.052 0.022 0.601 0.000
Spouse and children −0.005 0.794 0.571 0.000 0.085 0.000 0.514 0.000
Gender of the migrant
Male migrant 0.089 0.000 0.169 0.053 0.133 0.000 0.319 0.001
Constant 8.769 0.000 8.547 0.000
Observations 6361 3,552 6,033 3,258
Table IV. (pseudo) R2 0.417 0.359 0.425 0.331
Determinants of Wald F ¼ 61.83 Wald F ¼ 51.97
remittances χ2 ¼ 1,239.88 0.000 0.000 χ ¼ 1,168.74
2
0.000 0.000
(two-part model) Source: Estimation based on IHDS-II survey for 2011-12

of origin belongs to the EAG states then the probability of the sole migrant sending
remittances is lower by 4.2 percentage points as compared to the case where the household
does not belong to the EAG states (Table V ). This result does not support the altruistic
motive as migrants from these states are expected to send remittances with greater
probability if they are altruistic. However, overall, based on the household characteristics,
we may say that there is evidence of altruism motive behind the probability of sending
remittances by the sole migrant.
 Single migrant Multi migrant
Motives
First stage First stage behind
(marginal Second (marginal Second remittances
Explanatory variables effects) p-value stage p-value effects) p-value stage p-value

Household characteristics
Per capita pre-transfer income
2nd quintile −0.047 0.010 −0.053 0.407 −0.011 0.585 0.435 0.000 785
3rd quintile −0.066 0.000 0.118 0.086 −0.029 0.170 0.511 0.000
4th quintile −0.092 0.000 0.216 0.003 −0.041 0.119 0.656 0.000
5th quintile −0.144 0.000 0.318 0.000 −0.096 0.000 1.032 0.000
Farm land −0.00009 0.463 0.0004 0.287 −0.0003 0.038 0.002 0.000
With own house 0.045 0.057 0.163 0.132 0.033 0.239 0.168 0.380
No. of migrants 0.009 0.134 −0.108 0.000
Dependency ratio 0.047 0.057 −0.345 0.000 −0.025 0.259 −0.109 0.255
Rural household 0.023 0.110 −0.120 0.024 0.041 0.016 −0.325 0.000
EAG state −0.042 0.000 −0.038 0.449 −0.055 0.000 0.271 0.000
Head’s characteristics
Gender of the head (Male) −0.059 0.000 −0.062 0.262 −0.012 0.382 0.089 0.209
Education of the head
(beyond primary education) −0.011 0.376 0.170 0.001 −0.042 0.003 0.234 0.011
Employed head −0.073 0.000 −0.293 0.000 −0.030 0.046 −0.218 0.001
Migrant’s characteristics
Education of the migrant
(secondary) 0.049 0.000 0.460 0.000 0.027 0.101 0.348 0.000
Occupation of the migrant
Student −0.335 0.000 −0.405 0.428 −0.369 0.000 0.494 0.454
Labourers 0.320 0.000 0.270 0.132 0.233 0.000 −0.423 0.011
Professional 0.280 0.000 0.440 0.013 0.254 0.000 −0.292 0.073
Age of the migrant 0.011 0.000 0.021 0.037 0.020 0.000 −0.053 0.008
Age square of the migrant −0.0001 0.000 −0.0002 0.023 −0.0002 0.000 0.0007 0.004
Destination of the migrant
Outside state 0.068 0.000 0.147 0.001 0.107 0.000 0.033 0.695
Outside India 0.090 0.000 0.970 0.000 0.143 0.000 0.886 0.000
Relationship ties with the household
Married (no spouse and
children) −0.043 0.025 −0.389 0.000 −0.041 0.030 −0.269 0.003
Only children −0.234 0.000 −1.117 0.000 −0.300 0.000 0.022 0.923
Only spouse −0.121 0.000 0.272 0.005 −0.053 0.028 0.699 0.000
Spouse and children −0.004 0.848 0.572 0.000 0.054 0.008 0.438 0.000
Gender of the migrant
Male migrant 0.011 0.000 0.166 0.056 0.020 0.000 −0.070 0.582
Instruments
Proportion of other
households getting
remittances in village/
neighbourhood 0.157 0.021 0.101 0.050
Constant 8.792 0.000 10.864 0.000
λ −0.014 −1.061
Wald test of independent Table V.
equations (rho ¼ 0) χ2 0.140 0.709 22.32 0.000 Determinants of
Observations 6,361 6,033 remittances using
Source: Estimation based on IHDS-II survey for 2011–2012 Heckman selection
IJSE When we consider the amount of remittances for the single-migrant setting, we find that
46,6 many of the variables show opposite impact as compared to the probability of remittances.
For instance, the highest two income quintiles and the education of head have positive
impact, whereas the dependency ratio and the rural origin have negative impact on the
amount remittances (Table V ). A possible explanation for these results is that migrants
from richer and more educated households are likely to earn more and thus, the
786 remittances (if sent) are likely to be higher than the migrants from the poorer and less
educated background. Second, in the rural sector, the amount of remittances might be
lower because of less costly standard of living in the rural areas as compared to the urban
counterparts. At last, we also find that the land holdings and ownership of house do not
affect the amount of remittances. This result along with the fact that a majority of the
coefficients of the income classes are statistically insignificant indicates lack of
inheritance motive in the case of sole migrant.
When we compare these results with the multiple-migrant setting, we find some
evidence of inheritance motive based on the household characteristics. The income classes
and land holding all show positive impact on the amount of remittances sent by the
migrant in the SSM (Table V ). Similarly, the probability of sending remittances is higher
by 2 percentage points for the male migrants vis-à-vis the female migrants (Table V ). This
result is similar to studies from the countries where the male migrants are more likely to
send remittances as they are the main recipients of the inheritance (Hoddinott, 1992, 1994).
These results indicate the presence of inheritance motive in the multiple-migrant case.
On the other hand, altruism is not as prominent in the multiple-migrant setting as in the
single-migrant case. The probability of sending remittances reduces only in the case of the
richest income classes as compared to the poorest income class. Moreover, the dependency
ratio and the gender of the head do not affect the probability of sending remittances.
At the same time, we find that if the household of origin is from the rural area, has a head
educated at most up to primary education and has an unemployed head then the
probability of sending remittances to the household is higher as compared to their
counterparts in the multiple-migrant case (Table V ).
Apart from the determinants discussed above, in the multi-migrant case, we also want to
understand how the presence of other migrants in the household affects remittance
behaviour. The empirical results show that as the number of migrants in the household goes
up the amount of remittances sent by the given migrant goes down (Table V ). This result
does not give any definite indication of either altruism or inheritance. On the one hand, it
may indicate complementarity in the remittances by migrants of the same household under
altruism, and on the other, it may signify that migrants remit less as their share in
inheritance goes down.

Migrant characteristics
Results based on the migrant characteristics also support the presence of altruism in the
case of single-migrant. Here, migrants who are employed and have higher education
are more likely to send remittances as compared to their counterparts. When we compare
the amount of remittances among migrants who are sending remittances, we find that, on
average, students, unemployed and labourers send similar level of remittances. Only the
professionals send higher average remittances than the other three categories. In particular,
the proportional increment in average remittances for professionals is 1.553, i.e., if the
average remittance for unemployed is Rs100 then the professionals, on average, send
Rs155.3 to the household.
As opposed to these results, we find a weaker evidence of altruism in the case of
multi-migrant based on education and employment. Completion of the secondary education
does not affect the probability of sending remittances. However, the higher education does
increase average remittances sent by each migrant when there are multiple migrants in the Motives
family (Table V ). Moreover, employment status and occupation only affect the probability behind
of sending remittances positively in the multiple-migrant case. The average remittances sent remittances
by the migrants are actually lower for the labourers and professionals as compared to the
unemployed in the multiple-migrant scenario.
The strong ties of the migrant with family also prompt altruistic behaviour from the
migrant. We find (Table V ) that the amount of remittances is higher for the sole migrant if 787
the spouse and children stay back at home as compared to the single (unmarried) sole
migrant. For the multi-migrant situation, we find that if the spouse and children are staying
back in the household then both the probability and amount of remittances sent by the
migrant are higher as compared to the single migrant. These results suggest that
the migrant cares more for the immediate family and shows altruistic behaviour in the
presence of other migrants.
Along with the above factors, place of migration, migrant’s age and gender, affect the
probability and amount of remittances. The effect of these variables is similar in both
the sole-migrant and multiple-migrant settings. In particular, the migrants staying outside
the state of origin are more likely to send remittances and the amount of remittances sent is
also higher as compared to the amount sent by the migrants from the same state. The age
and gender of migrant also affect the probability and amount of remittances. Older migrants
are more likely to send remittances, however, the probability increases at a decreasing rate
with age of the migrant. Male migrants are likely to send more remittances as compared to
the female migrants.
To sum up, the empirical results support presence of altruism for the sole migrants.
In the multiple-migrants setting, we find evidence of the inheritance motive where the
migrants send higher remittances to household with larger income and assets. As opposed
to the single-migrant case, the results do not provide strong evidence of altruism in the case
of multiple migrants. Overall, the findings suggest that different motives govern the
remittance behaviour for single- and multiple-migrants situations. When there is only one
migrant in the family, the migrants behave altruistically. It is possibly due to the fact that
there is no uncertainty or competition regarding inheritance. On the other hand, when there
are more than one migrants in the family, the results indicate that any given migrant is
likely to remit due to both, the inheritance and altruistic motives. Also, rather than overall
characteristics of the left-behind family, it is the presence of immediate family members that
prompts the altruistic behaviour.

5. Concluding remarks
The present paper focuses on the difference in the motives behind remittances for the
single- and multiple-migrants setting. Here, we concentrate on two particular motives, namely,
altruism and inheritance. Using the IHDS data set, the study examines various determinants
of the probability and amount of remittances sent by the migrants. These determinants
indicate presence of either altruism or inheritance in the above mentioned two cases.
The empirical results suggest the presence of altruistic motive in the case of
single-migrant, whereas migrants from multi-migrant household are likely to send
remittances due to both altruistic and inheritance motives. In particular, we find that
single-migrants behave altruistically as they are more likely to send remittances to poorer
households and with unemployed head. Also, since the probability of sending remittances is
not affected by the household’s ownership of assets, the inheritance motive does not govern
the remittance behaviour of the single-migrants. On the other hand, for the multiple-migrant
case, family income and assets have positive impact on the amount of remittances indicating
presence of the inheritance motive. Moreover, the remittances respond to the presence of
immediate family members than the general characteristics of the household.
IJSE From these empirical findings, we may say that the households with multiple migrants
46,6 may not necessarily be in a better position as compared to the single-migrant households.
In the case of multi-migrant families, since the migrants behave with self-interest, it is not
clear how the number of migrants affects the remittances received by the household.
Further research is required to explore this question in detail.
Nonetheless, the study points out that the Indian single migrants are likely to send
788 remittances out of altruism and thus benefit the left-behind families. Moreover, the migrants
who have migrated outside state or India are more likely to send remittances and the
averages remittances sent are also higher as compared to the migrants migrating within
the same state. Thus, improving mobility across states and outside India is likely to be
beneficial to the left-behind families in terms of higher remittances.
The study cannot be concluded without possible limitations. The main short-coming of the
present study is the limited information about the migrant, since the data are collected from
the household at origin. As a result, the IHDS data set does not provide information on certain
variables which may influence remittance decision, such as, migrant’s income, duration of stay
at the destination, intention to return back and its social standings at the destination.
Even though the data set has some limitations, it is one of the very few data sets available
which provide information on migration in the case of India. Also, it provides information
which is not available from other data sources such as migrant’s family ties, income and asset
holding of the household. For the present study, these variables are crucial as the presence of
inheritance motive may be tested mainly using the asset holding of the household. Thus, to
test altruism against inheritance the IHDS data set provides most of the relevant variables and
enables us to point out differences in the motives for the single- and multiple-migrant settings.

Notes
1. It is the amount of money that the migrants send to the family that is staying behind.
2. Since the study focuses on determinants of remittances at individual level, we discuss the motives
behind remittances elaborated in the micro-economic theoretical models. The relation of
remittances with macro-economic factors such as growth and development of country is not
discussed. Rapoport and Docquier (2006) provided extensive literature review on both
micro-economic determinants and macro-economic factors associated with the remittances.
3. Many studies from developing countries have found altruism as a motive behind remittances.
For instance, Agarwal and Horowitz (2002) provided evidence for altruism from Guyana. Osaki
(2003) and Vanwey (2004) found supportive evidence for altruism from Thailand. Piracha and
Saraogi (2011) and Bouoiyour and Miftah (2015) found altruism as a motive behind sending
remittances based on their studies from Moldova and Morocco, respectively. Depoo (2014)
supported altruism as the main motive behind sending remittances from the study of Guyanese
migrants living in New York City. Alia et al. (2017) found evidence of altruism as a primary
motive behind sending remittances, based on their study on Burkina Faso.
4. De la Briere et al. (2002), Schrieder and Knerr (2000) and Osili (2004) provided similar empirical
evidence on inheritance motive.
5. Stark and Lucas (1988), Lillard and Willis (1997) and Regmi and Tisdell (2002), based on their
studies from Botswana, Malaysia, and Nepal, respectively, provided supportive empirical
evidence in this context.
6. Only 1.82 per cent cases report remittances from both migrant and household, 55.58 per cent of
cases report remittances from migrant to household and 25.22 per cent of cases report remittances
from household to the migrant.
7. The definition of children, in the Indian labour laws and Right to Education Act (The Child
Labour (Prohibition & Regulation) Amendment Act 35, 2016), provides the upper bound age of
children as 14 years. We follow this definition while selecting the minimum age of the migrant.
 The study does not consider any upper age limit for migrants. However, we tried the same Motives
analysis on a subsample of migrants under the age group of 15–60 years but did not find any behind
change in the findings.
remittances
8. Non-working persons are labelled in the data set, however, the category is mentioned as “others”
in the questionnaire.
9. Presence of parents at home may also influence remittances, however, the data set does not
provide information on whether parents of the migrants are residing in the household. 789
10. Here, positive net remittances mean that the remittances sent by the migrant to the household are
greater than the remittances received by the migrant from the household.
11. For the detailed discussion of the SSM, refer Greene (2003) and Wooldridge (2010).

References
Agarwal, R. and Horowitz, A.W. (2002), “Are international remittances altruism or insurance?
Evidence from Guyana using multiple-migrant households”, World Development, Vol. 30 No. 11,
pp. 2033-2044.
Akobeng, E. (2017), “Safety net for agriculture: effect of idiosyncratic income shock on remittance
payments”, International Journal of Social Economics, Vol. 44 No. 1, pp. 2-20.
Alia, D.Y., Ponty, N. and Tchwaket, I.K. (2017), “Altruism or exchange: what motives inter-household
transfers in Burkina Faso?”, International Journal of Social Economics, Vol. 44 No. 12,
pp. 2257-2270.
Bhagat, R.B., Keshri, K. and Ali, I. (2013), “Emigration and flow of remittances in India”, Migration and
Development, Vol. 2 No. 1, pp. 93-105.
Bouoiyour, J. and Miftah, A. (2015), “Why do migrants remit? Testing hypotheses for the case of
Morocco”, IZA Journal of Migration, Vol. 4 No. 2, pp. 1-20, doi: 10.1186/s40176-014-0027-2.
Brown, R. (1997), “Estimating remittance functions for pacific Island migrants”, World Development,
Vol. 25 No. 4, pp. 613-626.
Cox, D. (1987), “Motives for private income transfers”, Journal of Political Economy, Vol. 95 No. 3,
pp. 508-546.
Cox, D., Eser, Z. and Jimenez, E. (1998), “Motives for private transfers over the life cycle: an analytical
framework and evidence for Peru”, Journal of Development Economics, Vol. 55 No. 1, pp. 57-80.
Czaika, M. and Spray, J. (2013), “Drivers and dynamics of internal and international remittances”,
Journal of Development Studies, Vol. 49 No. 10, pp. 1299-1315.
De la Briere, B., Sadoulet, E., De Janvry, A. and Lambert, S. (2002), “The roles of destination, gender,
and household composition in explaining remittances: an analysis for the Dominican sierra”,
Journal of Development Economics, Vol. 68 No. 2, pp. 309-328.
Depoo, T. (2014), “Guyanese remittance motivations: altruistic?”, International Journal of Social Economics,
Vol. 41 No. 3, pp. 201-212.
Desai, S. and Vanneman, R. (2015), “India Human Development Survey-II”, User’s Guide, New Delhi,
pp. 1-45.
Desai, S. and Vanneman, R., National Council of Applied Economic Research, New Delhi (2015), India
Human Development Survey-II (IHDS-II), 2011-12. ICPSR36151-v2, Inter-university Consortium
for Political and Social Research, Ann Arbor, MI, available at: http://doi.org/10.3886/
ICPSR36151.v2 (accessed 27 April 2016).
Desai, S., Dubey, A., Joshi, B.L., Sen, M., Shariff, A. and Vanneman, R. (2009), “India Human
Development Survey: design and data quality”, IHDS Technical Paper 1, pp. 1-27.
Garip, F., Eskici, B. and Snyder, B. (2015), “Network effects in migrant remittances: evidence from
household, sibling, and village ties in Nang Rong, Thailand”, American Behavioral Scientist,
Vol. 59 No. 9, pp. 1066-1082.
IJSE Greene, W.H. (2003), Econometric Analysis, 5th ed., Pearson Education.
46,6 Heckman, J.J. (1979), “Sample selection bias as a specification error”, Econometrica, Vol. 47 No. 1,
pp. 153-161.
Hoddinott, J. (1992), “Modelling remittance flows in Kenya”, Journal of African Economies, Vol. 1 No. 2,
pp. 206-232.
Hoddinott, J. (1994), “A model of migration and remittances applied to Western Kenya”, Oxford
790 Economic Papers, Vol. 46 No. 3, pp. 459-476.
Jena, F. (2016), “The remittance behaviour of Kenyan sibling migrants”, IZA Journal of Migration,
Vol. 5 No. 11, pp. 1-19, doi: 10.1186/s40176-016-0059-x.
Leung, S.F. and Yu, S. (1996), “On the choice between sample selection and two-part models”, Journal of
Econometrics, Vol. 72 Nos 1-2, pp. 197-229.
Lillard, L.A. and Willis, R.J. (1997), “Motives for intergenerational transfers: evidence from Malaysia”,
Demography, Vol. 34 No. 1, pp. 115-134.
Lucas, R.E. and Stark, O. (1985), “Motivations to remit: evidence from Botswana”, Journal of Political
Economy, Vol. 93 No. 5, pp. 901-918.
Mahapatro, S. (2017), “Why do they remit? Examining factors influencing migrant remittances in
India”, Journal of Development Policy and Practice, Vol. 2 No. 2, pp. 225-235.
Mahapatro, S., Bailey, A., James, K.S. and Hutter, I. (2017), “Remittances and household expenditure
patterns in India and selected states”, Migration and Development, Vol. 6 No. 1, pp. 83-101.
Manning, W.G., Duan, N. and Rogers, W.H. (1987), “Monte carlo evidence on the choice between sample
selection and two-part models”, Journal of Econometrics, Vol. 35 No. 1, pp. 59-82.
Mitra, A. (2004), “What motivates transfer of resources: altruism or principle of exchange?”,
The Journal of Developing Areas, Vol. 37 No. 2, pp. 31-48.
Mohanty, S.K., Dubey, M. and Parida, J.K. (2014), “Economic wellbeing and spending behavior of
households in India: does remittance matter?”, Migration and Development, Vol. 3 No. 1,
pp. 38-53.
Osaki, K. (2003), “Migrant remittances in Thailand: economic necessity or social norm?”, Journal of
Population Research, Vol. 20 No. 2, pp. 203-222.
Osili, U.O. (2004), “Migrants and housing investments: theory and evidence from Nigeria”, Economic
Development and Cultural Change, Vol. 52 No. 4, pp. 821-849.
Parida, J.K., Mohanty, S.K. and Raman, K.R. (2015), “Remittances, household expenditure and
investment in rural India: evidence from NSS data”, Indian Economic Review, Vol. 50 No. 1,
pp. 79-104.
Piracha, M. and Saraogi, A. (2011), “Motivations for remittances: evidence from Moldova”, Discussion
Paper No. 5467, IZA, Bonn.
Poirine, B. (1997), “A theory of remittances as an implicit family loan arrangement”, World
Development, Vol. 25 No. 4, pp. 589-611.
Rapoport, H. and Docquier, F. (2006), “The economics of migrants’ remittances”, in Kolm, S.-C. and
Ythier, J.M. (Eds), Handbook of the Economics of Giving, Altruism and Reciprocity, Vol. 2,
Elsevier, New York, NY, pp. 1135-1198.
Regmi, G. and Tisdell, C. (2002), “Remitting behaviour of Nepalese rural-to-urban migrants:
implications for theory and policy”, Journal of Development Studies, Vol. 38 No. 3, pp. 76-94.
Rosenzweig, M.R. (1988), “Risk, implicit contracts and the family in rural areas of low-income
countries”, The Economic Journal, Vol. 98 No. 393, pp. 1148-1170.
Schrieder, G. and Knerr, B. (2000), “Labour migration as a social security mechanism for smallholder
households in Sub-Saharan Africa: the case of Cameroon”, Oxford Development Studies, Vol. 28
No. 2, pp. 223-236.
Stark, O. and Levhari, D. (1982), “On migration, remittances and risk in LDCs”, Economic Development
and Cultural Change, Vol. 31 No. 1, pp. 191-196.
Stark, O. and Lucas, R.E. (1988), “Migration, remittances, and the family”, Economic Development and Motives
Cultural Change, Vol. 36 No. 3, pp. 465-481. behind
The Child Labour (Prohibition & Regulation) Amendment Act 35 (2016), “Ministry of Law and Justice”, remittances
Government of India, New Delhi.
Tumbe, C. (2011), “Remittances in India: facts and issues”, Indian Journal of Labour Economics, Vol. 54
No. 3, pp. 479-501.
Vanwey, L.K. (2004), “Altruistic and contractual remittances between male and female migrants and 791
households in Rural Thailand”, Demography, Vol. 41 No. 4, pp. 739-756.
Wooldridge, J.M. (2010), Econometric Analysis of Cross Section and Panel Data, MIT Press, England.
World Bank (2018), “Migration and development brief”, Vol. 29, World Bank, Washington DC.
Zachariah, K.C. and Rajan, I. (2007), “Migration, remittances and employment: short-term trends and
long-term implications”, CDS Working Paper 395, Thiruvananthapuram.

Corresponding author
Navita Pal can be contacted at: navita.eco@gmail.com

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com
Reproduced with permission of copyright owner. Further
reproduction prohibited without permission.