WPS4956


Policy Research Working Paper                  4956




      The Impacts of International Migration
       on Remaining Household Members
      Omnibus Results from a Migration Lottery Program

                               John Gibson
                             David McKenzie
                             Steven Stillman




The World Bank
Development Research Group
Trade and Integration Team
June 2009
Policy Research Working Paper 4956


  Abstract
  The impacts of international migration on development                             households that were unsuccessful in the ballots. The
  in the sending countries, and especially the effects on                           immigration laws determine which household members
  remaining household members, are increasingly studied.                            can accompany the principal migrant, providing an
  However, comparisons of households in developing                                  instrument to address the second selectivity issue. Using
  countries with and without migrants are complicated by                            this natural experiment, the authors examine the myriad
  a double-selectivity problem: households self-select into                         impacts that migration has on remaining household
  migration, and among households involved in migration,                            members, focussing on labor supply, income, durable
  some send a subset of members with the rest remaining                             assets, financial service usage, diet, and physical and
  while other households migrate en masse. The authors                              mental health. The analysis uses multiple hypothesis
  address these selectivity issues using the randomization                          testing procedures to examine which impacts are
  provided by an immigration ballot under the Pacific                               robust. The findings indicate that the overall impact on
  Access Category of New Zealand's immigration policy.                              households left behind is largely negative. The findings
  They survey applicants to the 2002­05 ballots in Tonga                            also reveal evidence that both sources of selectivity
  and compare outcomes for the remaining household                                  matter, leading studies that fail to adequately address
  members of emigrants with those for members of similar                            them to misrepresent the impact of migration.



  This paper--a product of the Trade Team, Development Research Group--is part of a larger effort in the department to
  study the impacts of migration and remittances. Policy Research Working Papers are also posted on the Web at http://
  econ.worldbank.org. The author may be contacted at dmckenzie@worldbank.org.




         The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development
         issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the
         names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those
         of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and
         its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.


                                                       Produced by the Research Support Team
       The Impacts of International Migration on Remaining Household Members:
                 Omnibus Results from a Migration Lottery Program#

                               John Gibson, University of Waikato
                  David McKenzie, Development Research Group, World Bank *
                  Steven Stillman, Motu Economic and Public Policy Research




Keywords: Emigration; Natural Experiment; Selectivity, Wellbeing; Remittances.
JEL codes: J61, F22, C21



#
  We thank the Government of the Kingdom of Tonga for permission to conduct the survey there, the New
Zealand Department of Labour Workforce Group for providing the sampling frame, Halahingano Rohorua and
her assistants for excellent work conducting the survey, participants at various seminars for helpful comments,
and most especially the survey respondents. Financial support from the World Bank, Stanford University, the
Waikato Management School and Marsden Fund grant UOW0503 is gratefully acknowledged. The views
expressed here are those of the authors alone and do not necessarily reflect the opinions of the World Bank, the
New Zealand Department of Labour, or the Government of Tonga.
*
  Corresponding author: E-mail: dmckenzie@worldbank.org. Address: MSN MC3-307, The World Bank. 1818
H Street N.W., Washington D.C. 20433, USA. Phone: (202) 458-9332, Fax (202) 522-3518.
1. Introduction
        The impacts of international migration on development in the sending countries, and
especially the effects on remaining household members, are increasingly studied. Empirical
analysis is needed because the effect of migration on development in source communities is a
priori unclear. Migrant-sending households and their communities can benefit from
remittance inflows, which now make up 30 percent of total financial flows to the developing
world, but earnings and other household inputs that migrants would have generated locally
are lost. Hence this is a growing area of the literature; for example, out of the 392 journal
articles and working papers with remittances as a title or keyword, 60% were published since
2006. 1 Even more studies are likely in future as new survey data become available and labor
mobility increases in response to growing international wage gaps, rising demand for
services, divergent trends in youth and elderly populations in developed and developing
countries, and catch up from the previously "everything but labor" nature of globalization in
the post-World War II era (Pritchett, 2006).
        The biggest difficulty in measuring impacts of migration on development is posed by
selectivity issues. A common research strategy in this literature is to use household survey
data to compare households who have had at least one member emigrate to those that have
not. Such comparisons are complicated by a double-selectivity problem: first, households
self-select into migration, and second, among households involved in migration, some send a
subset of members with the rest remaining whilst other households migrate en masse.
        In this paper we address these selectivity issues using the randomization provided by
an immigration ballot under New Zealand's immigration policy. We survey applicants to this
random ballot and compare outcomes for the remaining household members of emigrants
with those for members of similar households who were unsuccessful in the ballot. The
policy rules determine which household members can accompany the principal migrant,
providing an instrument to address the second selectivity issue. Since this migration channel
has only recently opened, we measure the short-term impact of migration, which may change
over time. The short term may be when household challenges are greatest, as they adapt to
the absence of household members and have yet to receive large quantities of remittances.
        The particular policy we focus on is the Pacific Access Category (PAC), which was
established in 2001 and allows an annual quota of 250 Tongans to immigrate as permanent
residents to New Zealand without going through the usual channels used for groups such as

1
  Specifically, a search of EconPapers on RePEC (January 27, 2009) reveals that there were 17 papers in all of
the 1980s, 49 in all of the 1990s, 88 between 2000 and 2005, and 87 in 2008 alone.


                                                      2
skilled migrants and business investors. Many more people apply than the quota allows, so a
ballot is used by the New Zealand Department of Labour (DoL) to randomly select from
among the applicants. The probability of success in the ballot is approximately 10 percent.
We evaluate the impact of individuals migrating to New Zealand via the PAC on household
members remaining in Tonga (mainly parents, siblings, and nephews and nieces of the
migrant applicant). We consider a wide range of impacts, including the impact on labor
supply, income, durable assets, financial service usage, diet and physical and mental health.
        Our results suggest that at least in the short run there may be some adverse
consequences for those left behind when a subset of their household migrate to New Zealand.
Income falls by approximately 20-25 percent, whether measured per capita or per adult
equivalent, with a rise in net remittances not offsetting a large fall in labor earnings.
Ownership of livestock, durables, and access to financial services is also lower for the
remaining household members than for the control group. Diets change, with less fruit,
vegetables and fats consumed and more rice and root crops. Beneficial health changes include
falls in the body mass index and waist to hip ratio for working age adults.
        We also use data from a sample of non-applicants, and from ballot losers in
households which would entirely move if they had been successful in the PAC ballot, to
examine the degree of selection of households into migration, and selection among
households with a migrant as to which would partially move and which would move en
masse. We find selection is important in both dimensions. Thus, the non-experimental
estimation of migration impacts results in a biased assessment. In particular, using a sample
of non-applicant households would lead one to conclude that emigration has made remaining
household members wealthier, whereas the natural experiment shows the opposite result.
        These results may have broader applicability since Tongan migrants to New Zealand
under the PAC have characteristics that are quite typical of developing country migrants to
the US, both in terms of their levels of education and the degree of educational self-selection
relative to non-migrants (McKenzie, Gibson and Stillman, 2009). Moreover, although a
stereotype is of a husband migrating alone and leaving a family behind in a developing
country, a majority of married developing country immigrants in the US actually have their
spouse present, similar to our setting. 2 Immigration policies in many countries worldwide
(e.g. Australia, Canada, Ireland, United Kingdom, France, Italy) allow individuals moving on

2
  Specifically, using the 5% public use sample of the 2000 U.S. Census, we find that 59% of married immigrants
from developing countries who arrived in the U.S. in the last year had their spouses also present in the U.S.
Even for Mexico, we find 46% of newly arrived married immigrants have migrated with their spouse.


                                                      3
an employment visa to bring their spouse and dependent children, but not to immediately
bring their parents or adult siblings. The United States also allows for parents to accompany
the migrant, but not adult children or siblings. Consequently, the impacts on household
structure and on other outcomes for the families of those may be quite similar in many other
migrant-sending countries to what we observe among the Tongans left behind when family
members emigrate to New Zealand.
           In the next section we review relevant literature on the impact of emigration on source
areas and discuss channels through which emigration may affect household members left
behind. Section 3 provides background on the immigration program we examine, and Section
4 describes the data from the Pacific Island-New Zealand Migration Study (PINZMS) and
our estimation methods. The impacts on household level outcomes are presented in Section 5
and on individual outcomes in Section 6. Section 7 discusses multiple hypothesis testing,
while Section 8 concludes.

2. Previous Literature

2.1. Channels and Impacts
           The most studied impact of migration on household members left behind has been the
impact of remittances received. There are a variety of reasons that migrants send remittances,
including altruism towards those left behind, exchange for a variety of services provided by
the remaining family members (such as caring for property or other relatives), repayment of
loans made to finance migration or education, and insurance and strategic motives (Rapoport
and Docquier, 2006). These remittances directly contribute to household income, allowing
households to purchase more assets, and buy more normal goods, including education and
health inputs. 3 They can also relax liquidity constraints, enabling greater household
investment in businesses and children's education, and enable households to better mitigate
the impact of domestic shocks.
           If migration purely resulted in an exogenous increase in income for the remaining
household members, the sign of the expected impact on many outcomes of interest would be
easily determined. However, migration can also have a number of other impacts on the
sending household. Most obviously, an absent migrant earns no domestic wage and provides
no time inputs into household production. These effects may counteract the effect of
remittances received, so for example, households have less time to spend educating children,

3
    Remittances may also be received in the form of durable assets, directly increasing household asset stocks.


                                                          4
but perhaps more money to spend on them. Migrants may also transfer knowledge and
attitudes to their remaining family members. For example, Hildebrandt and McKenzie (2005)
find contraceptive knowledge to increase with emigration of household members from
Mexico to the US. Absence of decision-makers may also lead to changes in the bargaining
power of remaining members in the household leading to a reallocation of household
spending priorities (Chen, 2006). Separation from family members may impact on mental
health. Finally, migration of some family members may make it more likely that others will
migrate in the future, changing the incentives to acquire education.
         The result of all of these different potential channels is that the overall impact of
migration on various measures of the welfare of remaining family members is theoretically
uncertain. The effects are also likely to vary with the amount of time the family member is
away. For example, Lucas (1987) finds emigration from Botswana, Lesotho and other
Southern African countries to South Africa decreases domestic crop productivity in the short
run as labor is removed from the farm, but appears to enhance crop productivity and cattle
accumulation in the long run through invested remittances. Many other empirical studies are
unable to control for the length of time migrants have been away, resulting in an averaging of
short-run and long-run effects.

2.2. Selection and Identification
         The main challenge facing empirical analysis of the impacts of migration and
remittances on sending households is a double-selectivity problem. First, households choose
whether to engage in migration. 4 Households which send migrants are likely to differ along a
number of observable and unobservable dimensions from households which do not send
migrants, with some of these characteristics likely correlated with outcomes of interest. For
example, an unobserved asset shock may make the sending household poorer and encourage
emigration. Households with aptitude and knowledge of foreign languages may be more
inclined to engage in migration, and also have children who do better in school. Second,
amongst households which decide to engage in migration, some decide to move with their




4
  To be precise, they choose whether to engage in migration given the existing policy environment. In most
cases this is a policy environment which also involves substantial selection from the receiving country side with
employers and government officials screening interested potential migrants to determine which ones can
actually move. Since there is less screening at destination in our case than in the case of much legal migration,
the degree of self-selection is likely to be less in our example than in cases where employers and governments
are also involved in selecting the migrants. As a result, our results will be, if anything, conservative in terms of
showing the potential bias from self-selection.


                                                         5
entire families, while in others only some members emigrate. 5 A third form of selection
which also occurs in many contexts is selection into which migrants return. Since we are
examining the short-run impacts of migration, this source of selectivity is not an issue here. 6
         We are not aware of any study of the impact of migration on sending households
which explicitly deals with the second form of selection, since almost all developing country
migrant datasets lack information on entire households that move. The literature has used a
variety of approaches to address the first form of selection. Examples include assuming
selection on observables (e.g. Adams, 1998; Cox-Edwards and Ureta, 2003), parametric
selection correction models (e.g. Barham and Boucher, 1998; Acosta, Fajnzylber and Lopez,
2007), propensity-score matching (Esquivel and Huerta-Pineda, 2006), instrumental variables
methods, predominantly using current migration networks (e.g. Mansuri, 2006, Brown and
Leeves, 2007) or historic networks as instruments (e.g. Woodruff and Zenteno, 2007;
McKenzie and Rapoport 2007) 7 and work by Yang (2008) which uses a natural experiment
provided by exchange rate shocks in destination countries to look at impacts within the group
of households with migrants abroad.
         However, one may question the identification assumptions underlying these non-
experimental approaches to constructing no-migration counterfactuals. There is evidence that
migrants self-select both in terms of observables and unobservables (McKenzie, Gibson and
Stillman, 2009, Akee, 2009), so methods that assume selection on observables (which include
OLS and matching) are likely to be biased. Selection correction methods rely on parametric
structure and dubious excludability assumptions. For example, Acosta et al. (2007) and
Barham and Boucher (1998) assume that household asset holdings predict selection into
migration but do not directly affect earnings or labor force participation, when these assets
could be used to help finance own businesses, or could be the result of labor earnings. The
use of current migration networks as an instrument is subject to concerns about other
variables at the community level which also affect migration and outcomes of interest. For
example, a recent community weather shock such as a drought may have led to both
increased migration and a reduction in agricultural income in the community. Historic

5
  A further issue faced by some of the literature is the attempt to distinguish the impact of remittances from the
overall impact of migration. See McKenzie (2005) for a critique of this approach.
6
  None of the PAC migrants had returned to live in Tonga during the period of our study.
7
  Other instrumental variables have been also been used, but the exclusion restriction underlying these are
perhaps less convincing than the historic network variables. For example, Amuedo-Dorantes and Pozo (2006a)
assume that the number of Western Union branches in a state in Mexico affects labor supply only through
current migration, when these branches are likely to have been established as the result of factors which have
driven migration historically, including the level of development in a state, which likely also impact on labor
supply.


                                                        6
networks are less subject to concerns about recent shocks, but still need to rely on a plausible
story of why networks exogenously formed in one location and not in another, such as the
pattern of development of the railroad system in Mexico, as used by Woodruff and Zenteno
(2007). The natural experiment utilized by Yang (2008) provides the cleanest identification
of the impact of changes in remittance receipts amongst households receiving remittances,
but is unable to address the impacts of other channels through which migration can affect
households.

2.3. On Which Household Outcomes Does the Literature Focus?
        The growing literature on the impact of migration and remittances has examined a
variety of outcomes, all intended to measure the extent to which migration can aid
"development" in the sending countries. However, each study typically focuses on the impact
of migration on only a small number (often one) of outcomes in the sending country,
preventing analysis of the full range of impacts of migration on households in any one
sending country. Common outcomes of interest include income and poverty levels,
employment and business ownership, child health and education, and asset ownership. These
outcomes are both of inherent interest, and also the most commonly available measures in
household surveys.
        Existing evidence paints a generally rosy picture of the impact of migration on the
incomes, asset holdings, and poverty levels of household members left behind (Adams, 2007
provides a recent review). These studies generally attempt to construct a no-migration or no-
remittance counterfactual by estimating what the income of the household would be without
remittances but with the migrant working in the home country (e.g. Barham and Boucher,
1998; Adams, 2006). In earlier work (McKenzie, Gibson and Stillman, 2007) using the same
dataset as this paper, we compare the experimental outcome of migration on per-capita
income and poverty to what would be predicted using such methods. 8 We find that when
these counterfactual earnings are estimated using the unsuccessful lottery applicants the
results are similar to the pure experimental estimates. However, when non-applicants are used
to estimate counterfactual earnings the estimated earnings are much lower, leading to a
spurious finding of migration lowering poverty. This leads us to cast some doubt on the



8
  This earlier paper also looks at impacts on household size and total household income, which are two of the
many outcomes considered in the current paper. The current paper significantly differs from the earlier paper in
a number of ways, including this paper's focus on many different outcomes, its explicit consideration of the
double-selectivity issue, and through consideration of multiple hypothesis testing.


                                                       7
positive impacts of migration on poverty and income seen in earlier studies which are unable
to use a suitable comparison group of non-migrants.
       Among the fewer studies of the impact on child health outcomes, all show positive
effects, although more mixed results on inputs. For example, Hildebrandt and McKenzie
(2005) find lower infant mortality rates and higher birth weights amongst Mexican migrant-
sending families, but also that children in migrant households are less likely to be breastfed or
be vaccinated. Acosta et al. (2007) find higher weight-for-age and height-for-age among
children in migrant families in Nicaragua and Guatemala.
       The existing literature finds ambiguous effects of migration on several other key
outcomes of interest. In terms of the effect on child education, Cox-Edwards and Ureta
(2003) find that migration increases school attendance rates in El Salvador, and Yang (2008)
finds that increased remittances lead to more schooling in the Philippines, both consistent
with higher income alleviating liquidity constraints, whereas McKenzie and Rapoport (2006)
find migration lowers schooling attainment in Mexico, with boys in migrant households more
likely to drop out of school to migrate, and girls undertaking more housework.
       Evidence is also mixed in terms of the impact on adult employment. Funkhouser
(1992) finds remittances to be associated with lower overall labor supply, but higher self-
employment in Nicaragua. Acosta (2006) finds a negative impact on female labor supply in
El Salvador, but no effect on male labor supply. Yang (2008) finds higher remittances lead to
households being more likely to engage in entrepreneurial activities and to spend more hours
in self-employment, but to no significant effect on overall labor supply. Amuedo-Dorantes
and Pozo (2006a) find remittance receipt lowers female labor supply in Mexico, and shifts
male labor supply from formal to informal sector work. Woodruff and Zenteno (2007) find
remittance receipt to significantly increase the amount of capital invested in microenterprises
in Mexico, whereas Amuedo-Dorantes and Pozo (2006b) find a significant negative impact of
remittances on business ownership in the Dominican Republic.
       In this paper, we will consider these outcomes, along with other welfare outcomes
such as diet, anthropometric health measures, and mental health, which are measured less
often in household surveys and for which we have not been able to identify existing literature.
For example, a recent submission to the Global Commission on International Migration states
(Carballo and Mboup, 2005, p. 5) that "for close family and relatives left behind, the
departure of migrants to seek a living elsewhere is also fraught with psychosocial
difficulties", but provides no evidence for this assertion.



                                                8
       In addition, Aggarwal et al. (2006) have recently used cross-country panel data to
show an association between remittances and financial development, with the argument being
that the receipt of remittances paves the way for recipients to demand and gain access to other
financial services, even if the funds themselves are not received through banks. However,
they note that remittances may instead substitute for use of credit and other demands for bank
accounts, so that the direction of causation is unclear. Furthermore, it is possible that
household members who use the banking system are more likely to migrate, reducing
household use of bank accounts when they leave. We will therefore also consider measures of
access to bank accounts as another outcome measure.


3. Context and the Pacific Access Category
3.1 Background
       The Kingdom of Tonga is an archipelago of islands in the South Pacific, a three hour
plane flight from New Zealand. The population is just over 100,000, with a GDP per capita of
approximately US$2,200 in PPP terms. One-third of the labor force is in agriculture and
fishing, with the majority of paid workers in the manufacturing and services sectors, which
are dominated by the public sector and tourism.
       Emigration levels are high, with 30,000 Tongans living abroad, 94% of them in New
Zealand, Australia and the United States. Migration to New Zealand began in sizeable
numbers during the 1960s and 1970s, with Tongans arriving on temporary permits to take up
work opportunities. After their permits expired, some returned to Tonga and others stayed on
in New Zealand illegally. An amnesty in 1976 granted many of these individuals permanent
residence. Migration for work continued in the late 1970s and 1980s. However, in 1991, New
Zealand introduced a selection system for immigration, in which potential migrants are
awarded points for education, skills, and business capital. Few Tongans qualified to migrate
under this system, and so most Tongan migration since this time has been under family-
sponsored categories. For example, in 2004/05 only 20 Tongans were admitted as principal
applicants under the points system, compared to 349 under family categories, the majority
through marriage or as dependent children. Migration to Australia and the United States has
also become much more restrictive and reliant on family reunification categories. Australia
admitted 284 Tongans during the 2004/05 financial year. The United States admitted 324
Tongans in the 2004 calendar year, comprising only 5 under employment-based preferences
and 290 under immediate relative or family-sponsored categories.



                                              9
3.2. The Pacific Access Category
        In 2002, another channel was opened up for immigration to New Zealand through the
creation of the Pacific Access Category (PAC), which allows for a quota of 250 Tongans to
emigrate to New Zealand each year without going through the usual migration categories
used for groups such as skilled migrants and business investors. 9 Specifically, any Tongan
citizens aged between 18 and 45, who meet certain English, health and character
requirements, 10 can register to migrate to New Zealand. Many more applications are received
than the quota allows, so a random ballot is used by the New Zealand Department of Labour
(DoL) to draw from amongst the registrations. The odds of having one's name drawn were
approximately one in ten during the period we study.
        Once their ballot is selected, applicants must provide a valid job offer in New Zealand
(unskilled jobs suffice) within six months in order to have their application to migrate
approved. After a job offer is filed along with their residence application, it typically takes
three to nine months for an applicant to receive a decision. Once receiving approval, they are
then given up to one year to move. The median migrant in our sample moved within one
month of receiving their residence approval.
        The person who registers for the PAC is a Principal Applicant. If they are successful,
their immediate family (spouse and dependent children up to age 24) can also apply to
migrate as Secondary Applicants. The quota of 250 applies to the total of Primary and
Secondary Applicants, and represents about 80 migrant households. Successful applicants
cannot take other members of their household to New Zealand, so anyone living with parents,
siblings, or other relatives will leave household members behind when they migrate.
        These two features of the PAC, random selection amongst applicants and a rule
specifying which family members can and cannot accompany the successful migrant, allow
us to address the double-selectivity issues involved in assessing the impact of migration on
the remaining household. In particular, we can compare the group of households in Tonga
with a PAC emigrant to the group of unsuccessful ballots who would not be eligible to move
their entire household to New Zealand had their principal applicant been chosen in the ballot.


9
  The Pacific Access Category also provides quotas for 75 citizens from Kiribati, and 75 citizens from Tuvalu. A
similar scheme called the Samoan Quota allows 1100 citizens of Samoa to move each year. There have been
some small changes in the conditions for migration under the Pacific Access Category since the period we
examine in this paper ­ here we describe the conditions that applied for the potential migrants studied in this
paper.
10
   Data supplied by the New Zealand Department of Labour for residence decisions made between November
2002 and October 2004 reveals that out of 98 applications only 1 was rejected for failure to meet the English
requirement and only 3 others were rejected for failing other requirements of the policy.


                                                      10
4. Data, Methods and Selection
4.1. Data
       The data are from the Tongan component of the first (and to date only) wave of the
Pacific Islands-New Zealand Migration Survey (PINZMS), which measures multiple impacts
of migration. 11 The PINZMS survey was designed and implemented by the authors in 2005-
06 to allow study of migration through the Pacific Access Category, surveying applicants in
the first four years of the PAC. The survey also covered random samples of non-applicants to
enable comparisons between applicants and non-applicants.
       The survey includes questions on household demographics, education, labor supply,
income, asset ownership and diet, self-reported health status, smoking and alcohol use, and
anthropometric measurements of height and weight for all individuals, and waist and hip
circumference and blood pressure for adults. It also measures mental health for individuals
aged 15 and older using the Mental Health Inventory 5 (MHI-5) of Veit and Ware (1983).
       In a perfect randomized experiment, the impact of the treatment (here, having some
household members emigrate) could then be obtained via a simple comparison of means in
these two groups. However, mean comparisons may be biased if control group members
substitute for the treatment with a similar program or if treatment group members drop out
(Heckman, et al, 2000). For example, substitution bias will occur if PAC applicants who are
not drawn in the ballot migrate through alternative means and dropout bias will occur if PAC
applicants whose name are drawn in the ballot fail to migrate to New Zealand. Substitution
bias is not of serious concern; the low odds of winning the ballot and the limits on eligibility
for other migration channels available to Tongans mean that those with the ability to migrate
via other arrangements would likely have done so previously. But dropout bias is a more
relevant concern because approximately 15 percent of ballot winners do not ultimately move
to New Zealand.
       To adjust experimental estimates for possible dropout bias we use three subsets of the
PINZMS sample (see Appendix Table 1): (i) 61 households, with 283 individuals, in Tonga
with some previous members now PAC migrants in New Zealand; these are the "treatment"
group, (ii) 26 households, with 115 individuals, containing successful participants from the
same PAC ballots who were still in Tonga; these "non-compliers" had not moved when

11
  Further details about this survey and related papers produced from these data can be found at
www.pacificmigration.ac.nz.


                                              11
surveyed either because their application for New Zealand residence was not approved
(typically because of lack of a job offer) or was still being processed, and (iii) 124
households, with 654 individuals, containing unsuccessful participants from the same ballots
who were still in Tonga; these are the "control" group and were typically selected from the
same villages that the sampled PAC migrants had lived in prior to moving. The two samples
of successful ballots have a much higher sampling rate than the sample of unsuccessful
ballots (expansion factors of approximately 3.4, 2.5 and 37.9 are needed to weight each
sample up to the relevant population) and all of the analyses take this into account.
        Finally we also use a fourth sample to examine selection into migration, and to carry
out non-experimental estimation of the impacts of migration. This sample consists of 124
households, with 727 individuals, where no member of the household applied for the PAC.
These households were randomly chosen from the same villages as the PAC households, and
administered the same questionnaire.
        At the time of our survey, the sampled Tongan households with PAC emigrants in
New Zealand had a mean (median) time abroad for their former household members of 10
months (8 months). Just over three-quarters (77 percent) of migrant-sending households were
interviewed less than one year after eligible household members had emigrated to New
Zealand. Thus, our analysis is examining the initial impact of sending emigrants. The use of a
homogeneous period of time abroad allows us to avoid averaging short and long run effects
which may differ in sign (as found in Lucas, 1987).


4.2 Movers and Stayers
        We use the age and relationship rules governing which Secondary Applicants can
move with the Principal Applicant to identify household members that would have moved to
New Zealand if the Principal Applicant had been successful and compliant with the
treatment. These rules appear to be the binding constraint since the remaining family of PAC
emigrants are almost all outside the age and relationship eligibility for moving to New
Zealand (see Appendix Table 1). 12 Since the treatment group with migrants does not have
cases where the whole household moved, neither should the control group or non-complier
group. We therefore drop 75 unsuccessful households and 18 non-complier households in


12
  Specifically, just 11 (of the 283 residents of treatment group households) eligible family members stayed in
Tonga rather than immediately move to New Zealand with their principal applicant. Those that did were mainly
very young children and their mothers who eventually moved after our survey, when the children were at a more
suitable age for travel.


                                                     12
which their age and relationship structure would have allowed all members to move to New
Zealand. Note that 60 percent of the unsuccessful ballots fall into this category.
        Individuals in these households, those who would have moved in the control group
and non-complier households and the few eligible ones who did not move to New Zealand,
are all dropped for the individual level analyses, so that only like individuals in the treatment,
non-complier and control group are compared to each other. We define "stayers" to be the
individuals who the legal rules would require to stay behind if their principal applicant had
been successful in the PAC ballot.
        The remaining household members of PAC emigrants typically contain working age
adults who are either the parent and/or the siblings of the Principal Applicant, along with
children who are often their nephews and nieces (Appendix Table 1). Specifically, 46 percent
of migrant households contain a parent of the Principal Applicant, and 52 percent have a
sibling. Just over one-half (57 percent) of other relatives are under 18, and are mostly
nephews and nieces of the Principal Applicant. Very few of these extended family members
appear to have joined the household since the emigrants left, 13 and so as original household
members their welfare is likely to have been impacted by the departure of the PAC emigrants.
        As we have noted in the introduction, these remaining household members are likely
to be similar to the household members remaining in many countries when migrants move to
developed countries through employment categories. With the exception of the United States,
all traditional immigrant receiving countries restrict the relatives that can accompany a
migrant to the spouse and dependent children ­ the U.S. also allows parents. While in some
cases emigrants can later sponsor their parents or siblings, they can not do this until they have
spent several years in the country, and even then there can be restrictions or long waiting
periods. 14 Thus in the short-run, the remaining family of migrants is likely to be anyone apart
from their spouse and dependent children.




13
   We ask about how many of the previous 12 months each person was attached to the household. The number
of recent members who had been attached for less than 12 months was slightly lower (0.48 versus 0.63) for
migrant families than for those with unsuccessful ballots. We do not know for all households who was attached
to the household at the time the ballot result was announced since this is outside the 12 month window for many
of the households. However, given the low turnover in household composition, this does not appear to be a
concern.
14
   For example, several countries employ a "gravity" principle which only allows parents to be sponsored if they
have no remaining children in the home country, and then impose income requirements on the sponsoring
migrant. In general parents are still easier to sponsor than siblings.


                                                      13
4.3. Verifying Randomization
         We first test whether the PAC ballot correctly randomizes "stayer" households into a
treatment and a control group by examining whether the stayer group within the households
containing ballot losers are statistically different than the stayer group in households
containing ballot winners (both the migrant families and the non-compliers). The results in
Table 1 show that most ex-ante pre-migration characteristics are the same for ballot winners
and losers (at 95 percent confidence level). The only exceptions are that stayer adults in
successful ballot households have higher education levels and that there are more children
amongst the stayer group in successful ballot households. We present all regression results
with and without controls for the characteristics of these stayer members to examine the
robustness of our findings to small sample differences in the treatment and control group.

4.4. Calculating Experimental Estimates
         Throughout the remainder of the paper, when we present experimental estimates of
the impact on households and individuals of having household members move to New
Zealand under the PAC, we do not directly compare means of the treatment and control
groups due to concerns about dropout bias from non-compliers. Instead, instrumental
variables regression (IV) models, where ballot success is used as an instrument for having co-
residents emigrate, are used to estimate the treatment effect on the treated. 15 The PAC ballot
outcome can be used as an excluded instrument because randomization ensures that success
in the ballot is uncorrelated with unobserved individual attributes which might also affect
outcomes among the stayer household members and success in the ballot is strongly
correlated with migration.

4.5. Looking for Evidence of Selection
         In addition to obtaining consistent estimates of the impacts of migration, one of the
other goals of this paper is to examine how these estimates might change if we were unable to
correctly control for the double-selectivity. Table 2 and Table 3 examine the evidence for
selection in terms of the household and individual outcomes of interest, respectively. As we


15
   While an IV regression usually estimates the local average treatment effect (LATE), Angrist (2004)
demonstrates that in situations where no individuals assigned to the control group receive the treatment (i.e.,
there is no substitution) the IV-LATE is the same as the average treatment effect on the treated. We focus on the
average treatment effect since this is the parameter we can cleanly identify and which gives the overall impact
of migration. With a larger sample, we could examine the average treatment effect for subgroups of households,
such as those from poorer backgrounds, or single vs married applicants, or those with different household
compositions. We see such analysis of the heterogeneity of migrations' impacts as a fertile area for future
studies should other migration lottery data be able to be collected.


                                                      14
have noted, the appropriate comparison group for the remaining individuals in migrant
households are the group of individuals residing in ballot loser households who would remain
in Tonga even if the principal applicant from their household had been successful in the PAC
ballot. 16 That is, the right comparison group for the parents, siblings nephews, and nieces of a
migrant are the individuals who are the parents, siblings, nephews, and nieces of the would-
be migrant in ballot loser households. Means of the outcomes of interest for this group are
presented in the first column of Tables 2 and 3.
        The second column of Table 2 then presents means for ballot loser households where
everyone in the household would be eligible to move if the household had been successful in
the ballot. We call these households "all move" ballot losers. A comparison of columns 1 and
2 then enables us to examine the evidence for selection among migrant households in terms
of which migrants take their whole household and which do not. Likewise, column 2 of Table
3 presents means for adults and children who would be eligible to move if the principal
applicant in their household was successful in the PAC ballot. We see definite signs of
selectivity. Not surprisingly, whole households which move are smaller than households in
which some individuals would stay. Failure to remove these "all move" households from the
ballot losers will therefore bias estimates of the impact of migration on household size, since
these smaller "all move" households are gone from the ballot winner sample, but still present
in the ballot loser sample.
        We also see other areas where this form of selection is important: "all move"
households have fewer farm animals, are less likely to own an ATM card, and have a diet
with less fruit and vegetables than stayer households. However, the only area where selection
appears important in terms of individual characteristics is mental health, where the
individuals which are eligible to move have worse mental health than those who are not
eligible to migrate if someone in their household is successful in the PAC ballot.
        The third column in Tables 2 and 3 present the means for "stayer" households and
individuals in the sample of non-applicants. Comparing this column to the first allows us to
examine the other channel of selection ­ selection into migration. We see that stayers in
households where someone has applied to migrate are from larger, richer households than
stayers in non-applicant households. Adults in these households are less likely to be studying
than stayer adults in ballot loser households, and their children have fewer years of education.

16
   This is after adjusting for imperfect compliance by instrumenting migration with ballot success. The same
discussion as is applied here to ballot loser households applies equally to non-complier households among ballot
winners.


                                                      15
All these results are consistent with positive selection into migration. 17 Finally, the fourth
column of Tables 2 and 3 includes all non-applicants, thereby combining the two forms of
selectivity bias ­ selection into a migrant household, and selection among migrant households
as to who moves. However, since only 14 percent of non-applicant households are classified
as all movers, the overall effect of the two sources of selectivity is similar to that found for
the impact of selection into migration on its own.

4.6. Calculating Non-Experimental Estimates of the Impact of Migration
         Our experimental estimates of the impacts of migration come from IV regressions for
the group of individuals that the PAC policy rules identify as stayers, with the migration
ballot outcome used as an instrument for migration. We also estimate three other regression
specifications which we use to illustrate the bias caused when the two channels of selection
are ignored. First, we again estimate IV regressions, but do not use the PAC rules to eliminate
ballot loser households and individuals who would have moved if the principal applicant had
had a winning ballot in the PAC lottery. Comparing these estimates to the experimental
estimates illustrates the impact of selection among migrant households as to whom in the
household moves.
         Next, we use OLS regressions to estimate the impact of migration by comparing
migrant stayer households to non-applicant households. 18 In our second comparison
specification, we compare migrant stayer households to only the "stayer" non-applicant
households. This directly isolates the bias caused by household selection into migration.
Then, in our third comparison specification, we examine the combined bias from both
sources of selection by comparing migrant stayer households to non-applicant households.
         In all non-experimental regressions which examine household level impacts, we
condition on location (in Tongatapu or not), the maximum education level in the household,
and the household's labor income earnings in 2004 (the year before the survey). For
individual impacts, we condition on the same covariates as used in the experimental estimates
with controls: location, gender, age, and other basic characteristics which vary for adults and


17
   This positive selection may arise from a combination of the cost of migrating, the provisions in the PAC
which require a basic level of English literacy and the individual to be able to find a job in New Zealand, and
also from higher inequality in New Zealand than in Tonga offering greater opportunities for the educated and
skilled to increase their incomes than for the less skilled.
18
   Alternatively one could consider comparing migrant stayer households to a random sample of the population
of households without migrants, which would consist of non-applicant households, ballot loser households, and
non-complier households. However, since the majority of this population is, in fact, non-applicant households,
the results would be similar for this full population, and we believe it to be clearer for examining selection to
focus solely on a comparison to non-applicant households.


                                                       16
children (see tables for details). Our results are qualitatively robust to controlling for a variety
of other covariates, but given the relatively small sample sizes, we cannot include a huge
number of control variables at the same time.

5. Impacts on Household Level Outcomes

We now turn to estimating the impact of emigration on remaining household members. A
limitation of this analysis is that our surveys do not provide detail on how resources are
allocated within households. While this means some caution must be had in interpreting the
results, it is entirely in keeping with the existing literature, which has also not been able to
look at within household allocation. 19 Moreover, the comparison of experimental and non-
experimental methods will not depend on this limitation.



5.1. Household Size and Composition
         We begin by examining the impact of emigration on household size and composition,
since one immediate effect is that there are "fewer mouths to feed". The impact of having
some household members migrate to New Zealand on household size and composition is
shown in Table 4. Emigration leads to a significant reduction in household size. The mean
household has 6.7 people, and emigration is estimated to reduce this by 2.2 people.
Emigration leads to households having, on average, 1.5 fewer prime-age adults and 0.7-0.8
fewer children. There is no change in the number of older adults (>45 years), which is
reassuring since they are not eligible to move as Secondary Applicants.
         Table 2 showed that whole households which move are smaller than stayer
households. Failure to remove "all move" ballot losers would therefore cause us to understate
the fall in household size from migration. 20 Similarly, households which apply to migrate are
larger than those which do not. Thus, panel 3 shows that all the non-experimental estimates,
which ignore this selectivity, therefore understate the fall in household size arising from
migration. Instead of correctly estimating a fall of 2.2 people, we now find that household
size decreased by only 0.6 to 0.8 people, and is insignificant for two of the three non-


19
   The only paper we are aware of which collects data on within household allocation and migration is de Vreyer
et al. (2008) who consider large polygamous households in Senegal and look at allocation within subunits of the
household.
20
   For household level impacts, the first set of non-experimental estimates do not include any controls. They are
thus directly comparable to the estimates in Panel A in each Table examining household impacts. The reason for
not including controls in these regressions is that the household level controls are defined in terms of stayer
characteristics, and are thus not defined for all move households.


                                                       17
experimental estimates. The non-experimental estimates also incorrectly indicate that
migration led to a statistically significant rise in the number of adults aged over 45.

5.2. Household Income
        We next examine the impact on total household income, which can be disaggregated
into four sources, i) household earnings (annualized from individual reports for the previous
week), ii) net returns from sales of fish, crops, livestock, tapa cloth and mats (annualized
from household reports on an average month), iii) the imputed value of own-produced or
own-captured food consumed by the household (annualized from household reports for the
previous week), and iv) net remittances of money and goods.
        Since households in Tonga who have had some members move to New Zealand under
the PAC have fewer members, we examine the impact on per capita incomes and
alternatively on adult equivalent incomes. 21 The results in Table 5 for log total income
indicate that the families of migrants have 22-23% lower incomes than the families of non-
migrants, when no control variables are included regardless of whether income is per capita
or per adult equivalent. 22 The estimated impact is a 20-21% decline in income when control
variables are added, but the per capita estimate is no longer statistically significant. If we
instead estimate a linear model, which is more sensitive to outliers, we find that income
declines by $1,000 per capita or $1,250 per adult equivalent (19-20% of the mean for
treatment group households) for families of migrants when there are no controls and by $635
per capita or $910 per adult equivalent (12-14%) when controls are included; in neither case
are the estimates significantly different from zero.
        Examining the four components of household income, we find that having household
members migrate to New Zealand under the PAC leads to significant reductions in household
labor income per capita ($1,030-$1,280) or per adult equivalent ($1,260-$1,560). This is the
main opportunity cost to the household of sending a migrant ­ the income the individual
would have earned had they not migrated. These falls are partially offset by significant
increases in remittances received of $465-$500 per capita or $560-$590 per adult equivalent.




21
   Nutrition-based equivalence scales are not available for Tonga. We therefore follow Deaton and Paxson
(1994) and define the number of adult equivalents as the number of adults 18 and over, plus 0.5 times the
number of children 17 and under. As households in Tonga who have had some members move to New Zealand
under the PAC have fewer children, equivalence scales which are based on children needing less food and other
resources than adults will raise per-person resources more for the control group than for the migrant group.
22
   Percent changes are calculated as 100×[exp (-0.259) ­ 1] and 100×[exp (-0.253) ­ 1].


                                                     18
There is no significant change in either agricultural or subsistence income per person. 23 Thus,
while households with PAC migrants receive more remittances and have fewer mouths to
feed, this does not compensate for the large reduction in labor earnings faced by these
households.
        Table 2 showed little selection between stayer and "all move" ballot loser households
in terms of income variables. As a result, the first non-experimental specification (which uses
all ballot losers, including "all move" households) produces results similar to the
experimental results in Panel A. In contrast, Table 2 indicated significant positive selection
on income in terms of being an applicant household. Thus, the non-experimental estimates in
Table 5 that use the non-applicant sample understate the fall in income from migration. In
fact, the point estimates here indicate that migration leads to a 15 to 18 percent increase in
total income per capita for the remaining household members rather than the correct finding
of a 22 to 25 percent fall that the experimental estimate yields. Essentially, non-applicant
households are substantially poorer than migrant households would be (even conditional on
prior labor earnings), and thus using them as a control group causes one to think that migrant
households are relatively well off. Similarly, the use of these non-experimental control
groups causes one to miss the fall in household labor earnings per capita that comes with
household members migrating.

5.3. Durable Assets and Financial Access
        We next examine changes in other measures of household resources, including three
types of durable assets; i) the dwelling, ii) durable goods, and iii) livestock. We also examine
the impact on the financial access of each household, in particular, whether any household
members have bank accounts or ATM cards. 24
        Among our survey questions, we ask whether the dwelling is owned by anyone in the
household and whether, in the last 12 months, any renovations have been done. Our survey
also asks whether household members own any of 24 durable assets, including household
appliances, entertainment equipment and motor vehicles. We aggregate these responses into a
single index using the prices of durable goods we collected from stores in Tonga. 25 A


23
   It is worth nothing that both overall agriculture and subsistence income decline significantly for households
with PAC migrants, but the decline in the number of individuals in these households offsets these declines at
least using per capita and per adult equivalent measures of household income.
24
   An ATM card allows access to a remittance channel with transactions costs ten percentage points lower than
the usual money transfer fees in the New Zealand to Tonga corridor (Gibson, McKenzie and Rohorua, 2006).
25
   We also used principal component analysis to create a single dimensional index of wealth based on the first
principal component and found similar results.


                                                      19
separate question is asked on the number of automobiles that household members have
available for their regular use. The final asset questions concern holdings of domestic
livestock (pigs, chicken, cattle, goats and horses). We examine the impact of having
household members emigrate to New Zealand on the value of durables, the number of cars,
and the holdings of the main livestock; pigs, chickens and cattle.
       Table 6 reports the estimated impact of having household members migrate to New
Zealand on each of these outcomes. Again, impacts are estimated both without any control
variables and with controls to deal with small sample differences between treatment and
control groups. Although some outcomes are discrete, we continue to present estimates from
linear instrumental variable regression models. We also estimated treatment effects for the
discrete outcomes using the equivalent simultaneous equations probit models and found
nearly identical marginal effects as those presented for the models with no covariates, but had
difficulty getting the models with control variables to converge (a small number of covariates
perfectly predict whether households are in the non-complier group).
       We find that having household members migrate to New Zealand leads to the
remaining members having fewer cars and livestock and being less likely to have a bank
account or ATM card. When control variables are added, these effects persist for chickens
and the financial access variables. The impacts are large, with the remaining family of
emigrants having half as many chickens as non-migrant households and being 17% less likely
to have a bank account and 31-34% less likely to have an ATM card. We also find negative,
albeit insignificant, impacts on home ownership, the likelihood of having renovations, and
the value of durable goods.
       It is worth emphasizing that all of these results merely reflect changes in household
level assets and/or financial access. These changes may be occurring for a number of reasons:
i) households may have sold-off assets so the proceeds could used by the individuals moving
to New Zealand; ii) the lower incomes caused by having these family members move to New
Zealand may have caused a reduction in assets and financial access relative to unsuccessful
stayer households; iii) the individual in the household who used a bank account may have
been the person who migrated, or iv) the change in household composition (eg the moving
away of working-age household members) reduces needs for particular assets (such as cars
and computers) or financial access. Only 10 percent of migrant households in New Zealand
with family members remaining in Tonga report selling livestock, vehicles or other assets
before moving to New Zealand, suggesting that the first explanation is not the main channel.



                                              20
       We saw in Table 2 that stayer ballot losers tend to be wealthier than non-applicant
households, and in some dimensions, also hold more assets than entire households which
move. The non-experimental estimates at the bottom of Table 6 reflect the consequences of
this selection. Using the non-applicant sample, one would conclude that migration has made
remaining household members wealthier, with a higher value of durable assets, more cars,
greater home ownership, and increased likelihood of having a bank account. All of these
results just reflect selection, and if anything, our experimental results indicate that migration
of some household members has reduced the wealth of those left behind.

5.4 Diet
       We next examine the impact of having some household members emigrate to New
Zealand on the diet of those who stay. Specifically, we ask about the number of meals at
which each of 30 different foods were eaten at by any household resident the previous day.
To focus our analysis, we examine the cumulative number of meals in which seven foods are
consumed, six of which are composites. These foods are: rice, roots, fruits and non-root
vegetables, fish, fats, meats and milk.
       The results presented in Table 7 indicate that having family members emigrate leads
to a significant increase in the consumption of rice and roots and a significant decrease in the
consumption of fruits and vegetables. These changes in diet are large; consumption of rice
doubles, consumption of roots goes up by 20-25%, and consumption of fruits and vegetables
declines by 38-40%.
       The question on diet asks which foods anyone in the household ate yesterday. We
would thus expect larger households to be more likely to have someone eating any given food
group. Since this question is asked of the entire household, and not just of the members who
would stay if the household had a PAC ballot winner, the smaller size of migrant households
should lead to a tendency to find lower likelihoods of consuming any particular food group.
The significant negative result for fruits and vegetables may therefore just reflect that migrant
households do not have as many prime age adults who are likely to eat these foods. However,
this mechanical effect of household size cannot explain the increase in rice and root
consumption amongst migrant households.
       The main selectivity in terms of diet is on rice, roots, and fruits and vegetables.
Consistent with stayer ballot loser households being wealthier, they have a diet richer in more
expensive fruits and vegetables, and eat less rice and roots, which are cheap sources of
calories. As a consequence of this selectivity, failure to exclude whole households which


                                               21
would have moved or using the non-applicant sample leads one to understate the fall in fruits
and vegetables, and the rise in root and rice consumption. This parallels the understatement of
the fall in wealth and income seen in the previous subsections.

6. Impacts on Individuals
       Comparison of household level outcomes for migrant and non-migrant families is
complicated by the issue of how to account for differences in household size. Furthermore, a
unitary household model may not necessarily apply to a large household with siblings or
parents living with the potential migrant. For example, while we find non-migrant households
to be more likely to have a car, it is possible that the car is used only by the potential migrant,
and that stayer members of this non-migrant household never use the car. We therefore turn
to comparing individual level outcomes, which do not require assumptions about distribution
of resources within the household.

6.1. Working-Age Adults
       Table 8 examines the impact of migration on the labor supply, employment activity,
and health of 18 to 45 year old stayer adults ­ the age range eligible to apply for the PAC.
Since the literature has found that the impact of migration on labor supply varies by gender
we split the employment results by gender. The point estimates suggest a negative effect for
females and positive impact on males, but neither is significant. There is also no significant
impact on business ownership/self-employment or on whether the adult is currently studying.
However, we do find that individuals are less likely to be working in agriculture, significant
at the 10 percent level.
       Self-assessed general health status does not change with migration of other household
members. We do see significant impacts on some health behaviors and anthropometric
measures. Individuals in migrant households do not change smoking behavior, but consume
significantly more alcoholic drinks per month, although the significance disappears when we
add controls. The Tongan population has one of the highest levels of obesity in the world,
with 60 percent of the population classified as obese (Colagiuri et al., 2002). Migration is
found to reduce BMI, significant at the 10 percent level, and to reduce the waist-to-hip ratio,
significant at the 5 percent level. The waist-to-hip ratio is a marker of central obesity which is
more strongly associated with the risk of myocardial infarction among many ethnic groups
than BMI (Yusuf et al., 2005). Reductions in BMI and waist-to-hip ratios thus represent




                                                22
health improvements for the remaining adults in emigrant households. 26 The point estimates
also suggest lowered blood pressure, but this effect is insignificant.
         In related work (Stillman et al., 2009), we have found migration to improve the
mental health of the Tongans who move to New Zealand under the PAC. The last column of
Table 8 shows that their remaining family members do not receive the same improvement.
The MHI-5 ranges from 5 to 25, with higher scores indicating better mental health. The point
estimates thus suggest that, if anything, migration lowered the mental health of remaining
family members, although this effect is insignificant.
         The two individual characteristics for which evidence of selection was greatest in
Table 3 were mental health and currently studying. Stayer household members in ballot loser
households have better mental health than individuals who would move if the household had
been successful in the ballot, and also better mental health than non-applicants. The point
estimates in Panel C of Table 6 reflect this selectivity, with the results that do not control for
migrant selectivity indicating that migration improves mental health of those left behind,
whereas the experimental estimate suggests it lowers mental health. However, these
differences are not statistically significant. The currently studying variable is an example
where controlling for individual characteristics eliminates selection bias ­ here we are
controlling for years of education, which likely corrects for much of the selectivity in whether
an individual is currently studying.

6.2. Children
         Table 9 examines the impacts of migration of other household members on the
education and health of children aged 17 and under. Recall that the migrants here are
typically aunts and uncles of these children 27 and their parents are not the ones migrating. For
this reason we might expect less of the potential negative effects of parental absence on
education and health, and that the main channel through which migration would affect these
children is through income effects. Tonga has good basic education and health services, and
is ranked by the UN as high in terms of human development, with an adult literacy rate of
98.2 percent. 28 As such, liquidity constraints appear unlikely to be of large importance in


26
   These health improvements have occurred along with a decline in household economic resources, suggesting
that one pathway is that less food leads to less over-nutrition but behavioural change could also be a factor (e.g.
more walking if migrants sold the family car). Evidence from developed countries suggests that people adopt
healthier lifestyles when income declines, such as during recessions (Ruhm, 2005).
27
   72% are classified as "other relative" in terms of their relation to the principal applicant, and 23% are a sibling
of the principal applicant.
28
   http://palaceoffice.gov.to/content/view/124/95/ [accessed December 18, 2007].


                                                         23
determining access to schooling and health, so that changes in income may have relatively
small impacts on health or education outcomes.
       The results in Table 9 are consistent with this hypothesis. Migration is not found to
significantly affect the likelihood of currently studying, years of education attained, Tongan
literacy, or either parental-assessed or anthropometric health measures. The only marginally
significant effect is greater English literacy among children in migrant households, which is
an impact that is missed when the non-experimental control groups are used. Similarly, the
point estimates using the non-applicant sample would lead one to think that migration is
increasing years of education of children remaining in migrant households, which contrasts
with the zero or negative effect found with the experimental estimates.

6.3. Older Adults
       Finally in Table 10 we report the results of migration of household members on adults
aged 46 and older. The majority of these older household members are parents of the migrant,
with a mean (median) age of 60 (59) years. The point estimates suggest that both older males
and females are less likely to be employed when their children migrate. The magnitudes are
sizeable relative to the mean, corresponding to a halving of the employment rate. However,
the results are not significant when we examine men and women separately, and are only
significant at the 10 percent level when we combine males and females and do not include
additional controls in the regression. The point estimates also show large negative effects on
the likelihood of being a business owner, but again these are statistically insignificant.
       As with younger adults, there is a tendency for older adults to be less involved in
agriculture when they are in migrant households, although this difference is insignificant.
Older adults are marginally less likely to view themselves as being in very good health when
other household members have migrated, but we find no significant impacts on health
behavior, BMI, waist-to-hip ratios, blood pressure, or mental health.
       Since, under the PAC rules, none of these older adults would be eligible to move, the
only channel for selection is in terms of whether adults aged over 45 in households where
someone else applies to migrate are different from adults aged over 45 in households where
no one applies to migrate. There appears to be some selection in terms of mental health ­
older adults in non-applicant households have better mental health. This results in a negative
point estimate when using the non-experimental control group, as opposed to the positive
experimental point estimate, although neither estimate is statistically significant.




                                               24
7. Omnibus Effects and Multiple Hypothesis Testing
           Our analysis so far has followed the existing literature and tested for the impact of
migration on particular outcomes on a one-by-one basis. An advantage of this approach is
that we can directly compare our results for any particular outcome, for example, the impact
of migration on business ownership, to those from other studies examining the same
outcome. However, in total over Tables 4 through 10, we are examining the impacts of
migration on 62 different outcomes, and for each outcome, we consider the result with and
without controls. This raises questions about multiple hypothesis testing. In this section, we
examine which of our results are robust to different corrections for multiple testing.
           One approach sometimes used with multiple outcomes is to aggregate them into
particular groupings to examine whether the overall impact of the treatment on a family of
outcomes is different from zero. 29 This is useful if the intention is to see whether the global
impact of a particular treatment is generally positive. This is not our focus here, since we are
interested in the individual channels through which migration impacts family members left
behind. For example, we are interested in whether household labor earnings and subsistence
earnings go down with migration and remittances go up, more than whether the average
effect over all types of income is positive or not.
           We instead consider the significance of individual coefficients when viewed as part of
a family of n hypotheses. For example, we could consider all outcomes related to diet as a
family. The family-wise error rate is then defined as the probability of at least one type I error
in the family (Shaffer, 1995). Then, we can maintain the family-wise error rate at some
designated level , such as 0.05 or 0.10, by adjusting the p-values used to test each individual
null hypothesis in the family. The simplest such method is the Bonferroni method, which uses
as critical values /n. Thus, with 10 outcomes in a family, we would need to use a cutoff of a
p-value less than 0.01 when testing each individual outcome to maintain the family-wise error
rate at 10 percent.
           Several refinements to the Bonferroni method offer greater power. 30 Ranking the n
outcomes in increasing order of their p-values for testing a null effect, so that
 p1  p 2  ...  p n . Then, the Holm's (1979) sequentially-rejective bonferroni method is

applied as follows. In the first step, a null effect for outcome 1 is rejected if p1   / n . If we
cannot reject this outcome, we cannot reject null effects for all other outcomes. Otherwise,


29
     See, for example, O'Brien (1984) and Kling and Liebman (2004)
30
     The description of methods here is based on Shaffer (1995).


                                                      25
reject a null effect for outcome 2 if p 2   / (n - 1) , and at step j, reject a null effect for
outcome j if and only if null effects have been rejected for all outcomes i<j, and
 p j   / (n - j + 1) . Hochberg (1988) provides a "step-up" modification of this procedure,

which rejects null effects for all outcomes i  j if p j   / (n - j + 1) for any j=1,2,...n.

           The disadvantages of these approaches are that the larger the number of outcomes in
the family, the smaller the average power for testing each individual outcome. Furthermore,
these tests are conservative, as they are based on the assumption of independence between
outcomes. This is certainly not the case in our application, where most outcomes within
families are very closely related to one another. We therefore follow Katz, Kling and
Liebman (2007) in calculating bootstrapped estimates of adjusted p-values using a
modification of the free step-down algorithm of Westfall and Young (1993). 31 This approach
uses the correlation across test statistics to increase the power for testing each individual
outcome.
           Table 11 shows the 18 outcomes for which the experimental estimates are significant
(with controls added) at the 10 percent level when examined individually. If a researcher
were examining migration papers for evidence of a significant impact of migration on, say
working age adult BMI, then the p-value of 0.052 is strongly suggestive that migration
lowered BMI. However, if the researcher is reading our paper to see what the significant
effects of migration are, one should have a lot more caution in interpreting these BMI results
­ they may just be the outcome observed by chance to be significant among a whole host of
health outcomes that are being examined at the same time. In contrast, the adjusted p-value of
0.268 for adult BMI means that if one were to search for an effect among the 12 different
working age adult outcomes in Table 8, at least one effect this large would be observed 26.8
percent of the time.
           Given the loss of power involved in multiple testing and our small sample sizes, we
fix the family-wise error rate at 10 percent. If we were to consider all 62 outcomes as a
family, the Bonferroni p-value is thus 0.0016. The only outcomes that are significant at this
level are total household size, and the number of adults aged 18 to 45. The Holm and
Hochberg and Westfall-Young adjustments do not reveal any other outcomes to be
significant. That is we can be very confident that migration lowers the size of the sending
household, a none too startling result.


31
     See Appendix A of Katz et al. (2007) for a detailed description of how this is implemented.


                                                         26
         A slightly less conservative approach is to consider the outcomes in each table as a
family of outcomes. The second, third, and fourth columns of Table 11 provide adjusted p-
values for this family-wise comparison. Doing this again reveals that in addition to lowering
household size, we can be very confident that migration raises the amount of remittances the
sending household receives. After the Westfall-Young adjustments, there are five other
outcomes that are marginally significant (in the 0.10 to 0.14 adjusted p-value range). 32 These
are the fall in household labor earnings per capita and per adult, the decline in household
ATM card ownership, the decline in adult waist-to-hip ratio, and a drop in the number of
meals of fruits and vegetables eaten by the household. None of the individual level outcomes
for older adults or children (Tables 9 and 10) are significant when we adjust for multiple
hypothesis testing.

8. Conclusions
         In this paper, we have made two innovations to advance the literature on estimating
the impact of migration on development. First, we have used an unusual randomized
migration policy along with data collected specifically to exploit the experimental variation
provided by this policy to estimate the true short-run impact of migration on household
members remaining in a developing country after some family members have migrated to a
developed country. We have also demonstrated that both the selection of households into
migration and the decision among migrant households whether to send a subset of members
or to migrate en masse biases non-experimental estimates of the impact of migration on
development. Second, in contrast to most studies, which examine the impact of migration on
at most a few outcomes, we examined the impact on a comprehensive set of household and
individual level development indicators, including income, asset ownership, labor supply,
business ownership, physical and mental health, and child education.
         Our results suggest that family members remaining in Tonga may initially be made
worse off in several respects after some of their household members emigrate to New
Zealand, and that failure to account for the double-selectivity of migration would miss most
of this impact. Households sending migrants are smaller in size, and receive more remittances
per capita. However, the amount received in remittances and the reduction in household size
is not enough to compensate for the lost labor earnings of the migrants, leading to sizeable
32
   Note that when the treatment tends to operate in the same direction on the different outcomes in a family, the
Westfall-Young p-values are smaller than the more conservative Bonferroni and Holm p-values. However, in
some of the families of outcomes examined here, there can be a negative correlation. For example, the more
positive and significant is the increase in remittances from migration, the less negative (and hence less
significantly negative) is the fall in total income.


                                                      27
reductions in household income per capita. Migrant sending households also appear to have
fewer durable assets and livestock, and are less likely to have access to banking services,
such as ATM cards. The impacts on individual level outcomes are imprecise, with sizeable
point estimates accompanied by large standard errors in many cases. Adults in migrant-
sending households are less likely to be obese, although the significance of this result is not
robust to corrections for multiple hypothesis testing.
       Overall, our findings give a less rosy picture of the (immediate) impact of migration
on the incomes and wealth of household members left behind than is provided by much of the
existing literature. We show that the failure to account for both the selection of households
into migration and the decision among migrant households whether to send a subset of
members or to migrate en masse leads to estimates which incorrectly indicate that PAC
migration has increased income and wealth among Tongan households. This result suggests
that nearly all previous papers in this large literature should be viewed with some suspicion
since it is nearly impossible to control for both sources of migrant self-selection in
observational studies.
       It is worth emphasizing that we are only examining short-run impacts. While this has
the advantage that we are not mixing together households with different lengths of time in the
receiving country, as some papers in the literature appear to do, it is possible that the impacts
will differ in the long-run. For example, there are a number of significant costs that emigrants
face in moving to New Zealand, so it is possible that their remaining family in Tonga will
receive greater remittances in the future, once the migrants have repaid their moving costs.
However, there is no guarantee of this occurring, and, in fact, expectations questions that we
ask of both migrants and their remaining family members suggest that remittances will
decline over time.
       Finally, it must be noted that the treatment studied in this paper is the combination of
emigration and restriction on which family members can accompany the principal migrant.
Other research with PINZMS data shows that those who move to New Zealand experience
large gains in income (McKenzie et al., 2009) and improvements in mental health (Stillman
et al., 2009). These positive impacts would likely extend to the remaining family if they were
also allowed to move with the migrant. Since almost all migrant-destination countries impose
age and relationship rules blocking certain family members from accompanying migrants,
there may be millions of migrant-sending households in the developing world whose
remaining members become worse off. The methodological comparisons in our current paper



                                               28
suggest that these negative impacts would be unlikely to be detected by the conventional
methods and data used in previous studies of the development impact of migration.




                                            29
References

Acosta, Pablo (2006) "Labor supply, school attendance, and remittances from international
       migration: The case of El Salvador", World Bank Policy Research Working Paper No.
       3903
Acosta, Pablo, Fajnzylber, Pablo and Humberto Lopez (2007) "The impact of remittances on
       poverty and human capital: evidence from Latin American household surveys" in C.
       Özden and M. Schiff (eds.) International Migration, Economic Development and
       Policy, The World Bank and Palgrave Macmillan, Washington DC (2007), pp. 59-98.
Adams, Richard (1998) "Remittances, Investment, and Rural Asset Accumulation in
      Pakistan", Economic Development and Cultural Change 41(1): 155­73
Adams, Richard (2007) ""International Remittances and the Household: Analysis and Review
      of Global Evidence" World Bank Policy Research Working Paper No. 4116
Aggarwal, Reena, Asli Demirgüç-Kunt and Maria Soledad Martinez Peria (2006) "Do
      Workers' Remittances Promote Financial Development?", World Bank Policy
      Research Working Paper No. 3957.
Akee, Randall (2009) "Deciphering Immigrant Self-Selection: New Evidence from a
      Developing Country", Economic Development and Cultural Change, forthcoming.
Amuedo-Dorantes, Catalina and Pozo, Susan. (2006a). "Migration, Remittances and Male
     and Female Employment Patterns," American Economic Review Papers and
     Proceedings 96(2): 222-226.
Amuedo-Dorantes, Catalina and Pozo, Susan. (2006b). "Remittance Receipt and Business
     Ownership in the Dominican Republic", World Economy 29(7): 939-956.
Angrist, Joshua (2004) "Treatment Effect Heterogeneity In Theory and Practice", Economic
       Journal 502: C52-C83.
Barham, Bradford and Stephen Boucher (1998) "Migration, remittances and inequality:
      estimating the net effect of migration on income distribution", Journal of
      Development Economics, 55: 307-31.
Brown, Richard and Gareth Leeves (2007) "Impacts of international migration and
      remittances on source country household incomes in small island states: Fiji and
      Tonga" Discussion Paper No. 347, School of Economics, University of Queensland.
Carballo, Manuel and Mourtala Mboup (2005) "International Migration and Health", Paper
       submitted to the Global Commission on International Migration.
Chen, Joyce J. (2006), "Identifying Non-Cooperative Behavior Among Spouses: Child
      Outcomes in Migrant-Sending Households" Mimeo. Ohio State University
Colagiuri, Stephen, Ruth Colagirui, Siva Na'ati, Soana Muimuiheata, Zafirul Hussain and
       Taniela Palu (2002) "The Prevalence of Diabetes in the Kingdom of Tonga", Diabetes
       Care 25: 1378-1383.
Cox Edwards, Alejandra and Manuelita Ureta (2003) "International migration, remittances
      and schooling: evidence from El Salvador", Journal of Development Economics, 72,
      2: 429-61.
Deaton, Angus and Christina Paxson, (1994). "Intertemporal choice and inequality". Journal
      of Political Economy, 102(3), 437­ 467.



                                           30
Department for International Development [DFID]. (2007). Moving out of Poverty ­ Making
      Migration Work Better for Poor People, DFID, London.
De Vreyer, Philippe, Sylvie Lambert and Abla Safir (2008) "Remittances and Poverty : Who
      benefits in the Household ? ", Mimeo. Paris School of Economics.
Esquivel, Gerardo and Huerta-Pineda, Alejandra (2006) "Remittances and poverty in Mexico:
       a propensity score matching approach" mimeo Inter-American Development Bank,
       Washington DC.
Funkhouser, Edward (1992) "Migration from Nicaragua: Some Recent Evidence", World
      Development 20(8): 1209-1218.
Gibson, John, David McKenzie and Halahingano Rohorua (2006) "How cost elastic are
      remittances? Evidence from Tongan migrants in New Zealand", Pacific Economic
      Bulletin, 21(1): 112-128.
Heckman, J., Hohmann, N., Smith, J. and Khoo, M. (2000) "Substitution and Dropout Bias in
     Social Experiments: A Study of an Influential Social Experiment", Quarterly Journal
     of Economics 115(2): 651-694.
Hildebrandt, Nicole and David McKenzie (2005) "The effects of migration on child health in
       Mexico", Economia, 6(1): 257-289
Hochberg, Yosef (1988) "A sharper Bonferroni procedure for multiple tests of significance",
      Biometrika 75: 800-3.
Holm, Sture (1979) "A simple sequentially rejective multiple test procedure", Scandinavian
      Journal of Statistics 6: 65-70.
Katz, Lawrence, Jeffrey Kling, and Jeffrey Liebman (2007) "Experimental Analysis of
      Neighborhood Effects" Econometrica 75(1): 83-119.
Kling, Jeffrey and Jeffrey Liebman (2004) "Experimental Analysis of Neighborhood Effects on
        Youth", Working Paper 483, Industrial Relations Section, Princeton University.
Lucas, Robert E.B (1987) "Emigration to South Africa's mines", American Economic
       Review, 77(3): 313-30
Mansuri, Ghazala (2006) "Migration, Sex Bias, and Child Growth in Rural Pakistan" World
      Bank Policy Research Working Paper No. 3946
McKenzie, David (2005) "Beyond Remittances: The effects of Migration on Mexican
     households", pp123-148 in C. Özden and M. Schiff (eds) International Migration,
     Remittances and the Brain Drain, The World Bank: Washington D.C. 2005
McKenzie, David, John Gibson and Steven Stillman (2009) "How Important is Selection?
     Experimental Vs Non-experimental Measures of the Income Gains from Migration",
     Journal of the European Economic Association, forthcoming.
McKenzie, David, John Gibson and Steven Stillman (2007) "Moving to Opportunity,
     Leaving Behind What? Evaluating the Initial Effects of a Migration Policy on
     Incomes and Poverty in Source Areas", New Zealand Economic Papers 41(2): 197-
     224.
McKenzie, David and Hillel Rapoport (2006) Can Migration reduce Educational Attainment?
     Evidence from Mexico (with Hillel Rapoport), World Bank Policy Research Working
     Paper No. 3952.



                                            31
McKenzie, David and Hillel Rapoport (2007) "Network effects and the dynamics of migration
     and inequality: theory and evidence from Mexico" Journal of Development Economics,
     84(1): 1-24.
Milanovic, Branko (1987) "Remittances and income distribution", Journal of Economic
      Studies, 14(5): 24-37.
O'Brien, P. C. (1984): "Procedures for Comparing Samples with Multiple Endpoints,"
      Biometrics, 40(4): 1079­1087.
Pritchett, Lant (2006) Let Their People Come: Breaking the Gridlock on Global Labor
        Mobility Center for Global Development, Washington DC.
Rapoport, Hillel and Frederic Docquier (2006) "The Economics of Migrants' Remittances",
      pp. 1135-1199 in S.C. Kolm and J.M. Ythier (eds.) Handbook of the Economics of
      Giving, Altruism and Reciprocity: Volume 2, North-Holland: Amsterdam
Ruhm, Chris (2005) "Healthy living in hard times", Journal of Health Economics 24(2):
      341-63.
Shaffer, Juliet Popper (1995) "Multiple Hypothesis Testing", Annual Review of Psychology
       46: 561-84.
Stark, Oded, J. Edward Taylor and Shlomo Yitzhaki (1986) "Remittances and inequality",
       Economic Journal, 96(383): 722-40.
Stillman, Steven, David McKenzie and John Gibson (2009) "Migration and Mental Health:
       Evidence from a Natural Experiment", Journal of Health Economics, forthcoming.
Veit, C. and Ware, J. (1983) "The structure of psychological distress and well-being in
       general populations", Journal of Consulting and Clinical Psychology, 51(5): 730-742.
Westfall, Peter H., and Stanley Young, (1993). Resampling-Based Multiple Testing:
       Examples and Methods for P-value Adjustment. New York: Wiley.
Woodruff, Christopher, and Rene M. Zenteno (2007): Migration networks and
     microenterprises in Mexico, Journal of Development Economics, 82(2): 509-28.
Yang, Dean (2008) "International Migration, Remittances, and Household Investment:
      Evidence from Philippine Migrants' Exchange Rate Shocks", Economic Journal
      118(April): 591-630.
Yang, Dean (2006) "Remittances and poverty in migrants home areas: evidence from the
      Philippines" in C. Özden and M. Schiff (eds). International Migration, Remittances
      and the Brain Drain Washington DC: World Bank, pp. 81-121.
Yusuf, Salim and fourteen others (2005) "Obesity and the risk of myocardial infarction in
       27,000 participants from 52 countries: a case-control study", The Lancet 366: 1640-
       49.




                                            32
                         Table 1: Tests of Randomization

                                                     Successful Unsuccessful    T-Test
                                                       Ballot         Ballot   P-Value
                           Stayer Household Characteristics (n=118)
Size of the Stayer Household                             4.2            3.3     0.068
Number of Adults 18-45 Among Stayers                     1.5            1.5     0.928
Number of Children <18 Among Stayers                     1.6            0.8     0.005
Number of Adults >45 Among Stayers                       1.1            1.0     0.726
Proportion of Adults 18-45 Who Are Female               0.53           0.52     0.949
Annual Labor Earnings of Stayers in 2004               4,118          5,337     0.419
                           Characteristics of Stayer Children (n=146)
Proportion Female                                       0.45           0.58     0.150
Mean Age in Months                                       91            114      0.189
                    Characteristics of Stayer Working-Age Adults (n=176)
Proportion Female                                       0.53           0.49     0.578
Mean Age                                                29.4           27.5     0.172
Mean Height                                             167            168      0.693
Born on Tongatapu                                       0.79           0.68     0.394
Mean Years of Education                                 10.9            9.7     0.035
Visited New Zealand Prior to 2000                       0.14           0.10     0.388
Weekly Labor Earnings in 2004                            46             48      0.903
                        Characteristics of Stayer Older Adults (n=121)
Proportion Female                                       0.57           0.55     0.777
Mean Age                                                61.2           58.5     0.177
Mean Height                                             167            165      0.750
Born on Tongatapu                                       0.76           0.72     0.659
Mean Years of Education                                  9.7           8.6      0.090
Visited New Zealand Prior to 2000                       0.40           0.27     0.141
Note: T-tests account for household level clustering




                                                   33
        Table 2: Selectivity in Household Characteristics Among Who Applies and Who Moves with Their Whole Household

                                                                                                                      P-value for T-test of Equality with
                                                                          Sample Means                                       Stayer Ballot Losers
                                                   Stayer        All Move      Non-Applicant All Non-Applicant  All Move       Non-Applicant           All
                                                Ballot Losers   Ballot Losers    Stayer HHs      Households    Ballot Losers       Stayers      Non-Applicants
Total Household Size                                6.65             4.37            6.04           5.90           0.000            0.227             0.128
Adults Aged 18 to 45                                3.08             1.84            2.62           2.55           0.000            0.085             0.035
Children Aged Under 18                              2.57             2.53            2.88           2.89           0.930            0.406             0.387
Adults Aged over 45                                 1.00             0.00            0.50           0.43           0.000            0.001             0.000
Log Total Income Per Capita                         8.36             8.41            7.93           8.00           0.784            0.002             0.007
Total Income Per Capita                            5,400            6,508           3,626          3,896           0.238            0.001             0.005
Household Labor Earnings Per Capita                2,683            3,359           1,712          1,851           0.274            0.004             0.013
Agricultural Income Per Capita                      282              141             124            113            0.283            0.126             0.079
Subsistence Income Per Capita                      2,192            2,621           1,659          1,789           0.360            0.103             0.206
Remittances Per Capita                              243              373             130            130            0.133            0.179             0.156
Home Ownership                                      0.53             0.35            0.29           0.32           0.043            0.004             0.011
Improve Home                                        0.06             0.01            0.06           0.05           0.142            0.899             0.734
Value of Durables                                  7,672            7,456           6,042          6,250           0.611            0.001             0.003
Number of Cars                                      1.24            1.01            0.83           0.82            0.047            0.001             0.000
Number of Pigs                                      5.96            4.12            5.36           5.22            0.010            0.607             0.495
Number of Chickens                                  8.49            3.84            6.28           6.07            0.001            0.148             0.097
Number of Cattle                                    1.71            0.87            0.71           0.73            0.042            0.004             0.003
Has Bank Account                                    0.89            0.88            0.64           0.64           0.792             0.002             0.002
Has ATM Card                                        0.76            0.57            0.44           0.49           0.034             0.000             0.001
# Meals Rice                                        0.08            0.29            0.23           0.25           0.004             0.030             0.016
# Meals Roots                                       1.57            1.85            1.81           1.83           0.022             0.048             0.029
# Meals Fruits/Vegetables                           3.27            1.79            2.77           2.66           0.000             0.151             0.068
# Meals Fish                                        0.55            0.63            0.62           0.62           0.435             0.493             0.450
# Meals Fats                                        0.84            0.65            0.81           0.78           0.149             0.868             0.695
# Meals Meats                                       1.02            0.96            1.00           1.02           0.608             0.875             0.977
# Meals Milk                                        0.35             0.43            0.20           0.26           0.447            0.073             0.305
Note: See the paper for variable definitions.




                                                                                    34
                   Table 3: Selectivity in Individual Characteristics Among Who Applies and Who Moves with Their Whole Household

                                                                                                                                             P-value for T-test of Equality with
                                                                                  Sample Means                                                     Stayer Ballot Losers
                                                     Stayer                Mover           Non-Applicant             All              Mover            Non-Applicant              All
                                                  Ballot Losers         Ballot Losers         Stayers         Non-Applicants      Ballot Losers           Stayers           Non-Applicants
                                                                                      Adults Aged 18 to 45
Currently Employed (Males)                            0.46                  0.59                0.57                0.56               0.161               0.243                 0.284
Currently Employed (Females)                          0.33                  0.47                0.25                0.28               0.152               0.308                 0.537
Business Owner                                        0.10                  0.12                0.13                0.11               0.666               0.463                 0.679
Main Activity is Agriculture                          0.30                  0.25                0.26                0.25               0.436               0.506                 0.343
Currently Studying                                    0.20                  0.13                0.10                0.10               0.173               0.025                 0.027
Very Good Health                                      0.34                  0.36                0.34                0.34               0.760               0.974                 0.967
Currently Smokes                                      0.14                  0.21                0.19                0.18               0.261               0.374                 0.477
Alcoholic Drinks per Month                             3.3                   9.1                 6.6                 6.0               0.203               0.210                 0.269
Body Mass Index                                       32.4                  34.2                33.0                33.0               0.136               0.558                 0.577
Waist-to-Hip Ratio                                    0.92                  0.92                0.93                0.93               0.317               0.668                 0.733
Diastolic Blood Pressure                              86.3                  83.7                83.7                83.7               0.126               0.187                 0.169
Mental Health                                         20.4                  19.3                20.0                19.8               0.000               0.118                 0.035
                                                                                            Children
English Literacy                                      0.45                  0.57                0.44                0.49               0.148               0.952                 0.629
Tongan Literacy                                       0.61                  0.64                0.54                0.59               0.675               0.454                 0.881
Currently Studying                                    0.63                  0.72                0.61                0.64               0.263               0.871                 0.889
Years of Education                                    1.50                  1.74                0.70                0.88               0.683               0.064                 0.177
Very Good Health                                      0.68                  0.52                0.55                0.54               0.070               0.143                 0.112
BMI for Age                                           1.18                  1.49                1.32                1.31               0.332               0.680                 0.706
                                                                                    Adults Aged 46 and Over
Currently Employed (Males)                            0.36                   n.a.               0.38                0.38                n.a.               0.938                 0.938
Currently Employed (Females)                          0.30                   n.a.               0.28                0.28                n.a.               0.869                 0.869
Business Owner                                        0.16                   n.a.               0.19                0.19                n.a.               0.740                 0.740
Main Activity is Agriculture                          0.43                   n.a.               0.42                0.42                n.a.               0.845                 0.845
Very Good Health                                      0.37                   n.a.               0.23                0.23                n.a.               0.121                 0.121
Currently Smokes                                      0.26                   n.a.               0.15                0.15                n.a.               0.179                 0.179
Alcoholic Drinks per Month                            3.61                   n.a.               4.72                4.72                n.a.               0.784                 0.784
Body Mass Index                                       35.6                   n.a.               34.8                34.8                n.a.               0.770                 0.770
Waist-to-Hip Ratio                                     0.92                  n.a.               0.94                0.94                n.a.               0.153                 0.153
Diastolic Blood Pressure                               87.1                  n.a.               90.3                90.3                n.a.               0.216                 0.216
Mental Health                                          19.3                  n.a.               20.2                20.2                n.a.               0.036                 0.036
Note: T-tests account for household level clustering. n.a. denotes not applicable, since individuals aged over 45 can not be migrants under the PAC. Given this, the group of non-applicant
stayers is identical to the group of all non-applicants for this age group. Mover Ballot Losers include both individuals who would move with their entire households, as well as individuals
who would move when some of their members remain in Tonga.



                                                                                                 35
                       Table 4: Impact of Migration on Household Composition

                                                          Total           Adults          Children          Adults
                                                       Household           Aged             Aged            Aged
                                                           Size          18 to 45         under 18         over 45
                                    Panel A: Experimental Estimates without Controls
Impact of Migration                                     -2.23***         -1.54***          -0.80*            0.06
                                                         (0.62)           (0.33)           (0.44)           (0.18)
                                      Panel B: Experimental Estimates with Controls
Impact of Migration                                     -2.19***         -1.47***           -0.68            -0.09
                                                         (0.70)           (0.34)           (0.52)           (0.20)
Mean for Unsuccessful Stayer Households                   6.65             3.08             2.57             1.00
Sample Size                                                118              118              118              118
                               Panel C: Estimates using Non-Experimental Control Groups
Including All Move Ballot Losers                          -0.85          -0.76***         -0.78**          0.64***
                                                         (0.53)           (0.23)           (0.39)           (0.15)
Compared to Non-Applicant Stayers                        -0.72*          -0.60***         -0.74**          0.61***
                                                         (0.41)           (0.18)           (0.32)           (0.15)
Compared to All Non-Applicants                            -0.64          -0.54***         -0.78**          0.68***
                                                         (0.40)           (0.18)           (0.31)           (0.15)
Note: Experimental Estimates are IV estimates where migration is instrumented with the PAC ballot outcome. Controls
are labor earnings of stayers in 2004, the proportion of adult stayers who are female, highest education level of stayer
adults, and whether the household lives on Tongatapu. See text for more detail on the estimates using non-experimental
control groups.
*, **, and *** indicate significance at the 10%, 5% and 1% levels.




                                                                                    36
                   Table 5: Impact of Migration on the Household Income of Remaining Household Members

                                                        Log Total                        Household       Agricultural     Subsistence
                                                                       Total Income                                                      Net Remittances
                                                         Income                        Labor Earnings       Income           Income
                                                Panel A: Experimental Estimates, Per Capita without Controls
Impact of Migration                                      -0.259*           -1,007         -1,281***           -197               5            466***
                                                         (0.149)            (720)            (421)           (165)            (461)            (156)
                                                 Panel B: Experimental Estimates, Per Capita with Controls
Impact of Migration                                       -0.217             -635          -1,031**            45              -150           501***
                                                         (0.150)            (726)            (436)           (136)            (531)            (163)
                                           Panel C: Experimental Estimates, Per Adult Equivalent without Controls
Impact of Migration                                      -0.253*           -1,246         -1,556***           -233              -18           560***
                                                         (0.143)            (782)            (497)           (175)            (481)            (199)
                                             Panel D: Experimental Estimates, Per Adult Equivalent with Controls
Impact of Migration                                      -0.237*             -908          -1,257**            25              -266           589***
                                                         (0.141)            (766)            (501)           (144)            (560)            (206)
Mean PC for Unsuccessful Stayer Households                                  5,400            2,683            282             2,192             243
Mean PAE for Unsuccessful Stayer Households                                 6,377            3,224            322             2,546             285
Sample Size                                                 118              118              118             118              118              118
                                           Panel E: Estimates using Non-Experimental Control Groups, Per Capita
Including All Move Ballot Losers                         -0.243*          -1,397*         -1,478***            -90             -195           366**
                                                         (0.145)            (844)            (553)           (118)            (418)            (150)
Compared to Non-Applicant Stayers                          0.178             585              -206             -45             259            576***
                                                         (0.134)            (537)            (231)            (71)            (319)            (159)
Compared to All Non-Applicants                             0.149             490              -220             -32             167            581***
                                                         (0.129)            (529)            (225)            (69)            (313)            (156)
Note: Experimental estimates are IV estimates where migration is instrumented with the PAC ballot outcome. Controls are labor earnings of stayers in 2004,
the proportion of adult stayers who are female, highest education level of stayer adults, the number of stayers who are children and adults 18 to 45, and
whether the household lives on Tongatapu. See text for non-experimental details.
*, **, and *** indicate significance at the 10%, 5% and 1% levels.




                                                                                    37
                                       Table 6: Impact of Migration on Durable Assets and Financial Access

                                                       Home        Improve     Value of Number of Number of Number of Number of Has Bank Has ATM
                                                     Ownership       Home      Durables       Cars        Pigs    Chickens      Cattle     Account       Card
                                                           Panel A: Experimental Estimates without Controls
Impact of Migration                                    -0.022        -0.043       -615      -0.288*     -1.339*  -4.639***     -0.860*     -0.172** -0.340***
                                                      (0.103)       (0.041)      (508)      (0.153)     (0.807)    (1.711)     (0.493)      (0.078)     (0.095)
                                                            Panel B: Experimental Estimates with Controls
Impact of Migration                                    -0.058        -0.038       -306       -0.236      -1.593   -3.860**       -0.81      -0.167*   -0.306***
                                                      (0.123)       (0.049)      (637)      (0.163)     (0.995)    (1.900)     (0.515)      (0.090)     (0.108)
Mean for Unsuccessful Stayer Households                0.531         0.061      7,672         1.24        5.96       8.49        1.71        0.891       0.761
Sample Size                                             118           118          117        118         118        118          118         115         115
                                                     Panel C: Estimates using Non-Experimental Control Groups
Including All Move Ballot Losers                       0.117         0.018        -328       -0.181      -0.207     -2.015      -0.438     -0.147**    -0.219**
                                                      (0.093)       (0.034)      (403)      (0.125)     (0.814)    (1.230)     (0.376)      (0.072)     (0.090)
Compared to Non-Applicant Stayers                     0.170**     -0.068*** 1,150***       0.295**       -0.081     -1.231      0.260       0.137*       0.039
                                                      (0.080)       (0.024)      (388)      (0.115)     (1.155)    (1.170)     (0.277)      (0.074)     (0.080)
Compared to All Non-Applicants                         0.149*     -0.057*** 1,066*** 0.318***            0.054      -1.063      0.263       0.146**      0.002
                                                      (0.079)       (0.021)      (380)      (0.112)     (1.088)    (1.109)     (0.270)      (0.072)     (0.078)
Note: Experimental estimates are IV estimates where migration is instrumented with the PAC ballot outcome. Controls are labor earnings of stayers in 2004, the
proportion of adult stayers who are female, highest education level of stayer adults, the number of stayers who are children and adults 18 to 45, and whether the
household lives on Tongatapu. See text for description of non-experimental control groups.
*, **, and *** indicate significance at the 10%, 5% and 1% levels.




                                                                                 38
                          Table 7: Impact of Migration on the Diet of Remaining Household Members

                                                      # of Meals # of Meals # of Meals # of Meals # of Meals # of Meals # of Meals
                                                         Rice          Roots     Fruits / Vegs       Fish           Fats       Meats          Milk
                                                  Panel A: Experimental Estimates without Controls
Impact of Migration                                    0.189**       0.392***      -1.291***        0.177         -0.213       -0.133        -0.054
                                                        (0.072)       (0.142)        (0.434)       (0.111)       (0.149)      (0.137)       (0.116)
                                                    Panel B: Experimental Estimates with Controls
Impact of Migration                                      0.130        0.314*        -1.277**        0.201        -0.337*       -0.152         0.014
                                                        (0.087)       (0.171)        (0.498)       (0.143)       (0.180)      (0.169)       (0.118)
Mean for Unsuccessful Stayer Households                  0.082         1.571           3.265        0.551          0.837        1.020         0.347
Sample Size                                               118           118             118          118            118          118           118
                                              Panel C: Estimates using Non-Experimental Control Groups
Including All Move Ballot Losers                         0.052       0.356**          -0.511        0.142         -0.069       -0.005        -0.024
                                                        (0.078)       (0.144)        (0.344)       (0.092)       (0.128)      (0.122)       (0.104)
Compared to Non-Applicant Stayers                        0.026         0.135          -0.221       0.145*         -0.100       -0.129         0.023
                                                        (0.077)       (0.142)        (0.290)       (0.085)       (0.117)      (0.114)       (0.077)
Compared to All Non-Applicants                           0.010         0.120          -0.136       0.150*         -0.080       -0.147        -0.019
                                                        (0.076)       (0.138)        (0.282)       (0.083)       (0.114)      (0.111)       (0.077)
Note: Experimental estimates are IV estimates where migration is instrumented with the PAC ballot outcome. Controls are labor earnings of stayers in
2004, the proportion of adult stayers who are female, highest education level of stayer adults, the number of stayers who are children and adults 18 to
45, whether the household lives on Tongatapu, and day of the week fixed effects. Roots include taro (swamp taro), taro taruas (chinese taro), kumara
(sweet potato), taamu/kape, yams, cassava/manioc, and potato. Fruits and vegetables include other vegetables, coconut (fresh and dry), banana,
mango, pawpaw, and other fruits. Fish includes tinned fish and fresh fish. Fats include corned beef, mutton, and coconut (fresh and dry). Meats
include corned beef, mutton, fresh beef, chicken, pork, and other meat (eg. sausage).
*, **, and *** indicate significance at the 10%, 5% and 1% levels.




                                                                                     39
                                    Table 8: Impact of Migration on Outcomes for 18-45 Year-Old Adults Remaining in Tonga

                                    Currently    Currently                  Main                                             Alcoholic                               Diastolic
                                                              Business                  Currently   Very Good    Currently              Body Mass       Waist to                     Mental
                                    Employed     Employed                 Activity is                                        Drinks Per                               Blood
                                                               Owner                    Studying     Health       Smokes                  Index         Hip Ratio                    Health
                                     (Males)     (Females)                Agriculture                                         Month                                  Pressure
                                                                     Panel A: Experimental Estimates without Controls
Impact of Migration                   0.084        -0.103       -0.001      -0.178*     -0.085       -0.005       0.009        7.476**       -0.565     -0.028**       -2.686         -0.624
                                     (0.123)      (0.100)      (0.059)      (0.092)    (0.072)      (0.072)      (0.068)       (3.426)      (1.776)      (0.011)      (2.402)        (0.417)
                                                                       Panel B: Experimental Estimates with Controls
Impact of Migration                   0.057        -0.052       0.072       -0.178*     0.011        0.019        0.026         6.550       -2.151*     -0.029**       -2.164         -0.457
                                     (0.165)      (0.093)      (0.047)      (0.098)    (0.058)      (0.099)      (0.078)       (4.268)      (1.092)      (0.012)      (2.707)        (0.556)
Mean for Unsuccessful
Stayer Households                     0.459        0.333          0.097        0.300        0.197        0.338        0.143         3.31       32.4      0.925          86.3          20.4
Sample Size                             85           91            175          170          174           171         135           134       157        159           144           172
                                                                   Panel C: Estimates using Non-Experimental Control Groups
Including All Move Ballot Losers          -0.038   -0.155**      -0.044      -0.164**       -0.059       -0.028       -0.057       -1.150     -1.803   -0.021**        1.958          -0.003
                                         (0.092)    (0.075)     (0.043)       (0.065)      (0.052)      (0.069)      (0.065)      (5.336)    (1.095)    (0.011)       (1.801)        (0.414)
Compared to Non-Applicant Stayer -0.077              -0.075      -0.019        -0.090       -0.002       -0.013       -0.031        2.259    -1.537*  -0.027***        0.350           0.169
                                         (0.065)    (0.066)     (0.034)       (0.057)      (0.043)      (0.062)      (0.061)      (3.696)    (0.927)    (0.009)       (1.580)        (0.362)
Compared to All Non-Applicants            -0.037     -0.077      -0.003       -0.094*       -0.013        0.012       -0.008        3.863     -1.398  -0.026***        0.332           0.315
                                         (0.064)    (0.063)     (0.032)       (0.051)      (0.041)      (0.058)      (0.056)      (3.470)    (0.852)    (0.009)       (1.362)        (0.332)
Note: Experimental estimates are IV estimates where migration is instrumented with the PAC ballot outcome. Controls are gender, age, years of education, height, labor earnings in   2004, and
whether the household lives on Tongatapu. Standard errors account for household-level clustering. See text for non-experimental description.
*, **, and *** indicate significance at the 10%, 5% and 1% levels.




                                                                                            40
          Table 9: Impact of Migration on Outcomes for Children Remaining in Tonga

                                               English    Tongan      Currently     Years of Very Good BMI for
                                               Literacy   Literacy     Studying Education     Health       Age
                                    Panel A: Experimental Estimates without Controls
Impact of Migration                             0.062      -0.059        -0.027      0.019     -0.190     0.123
                                               (0.123)    (0.122)       (0.111)     (0.754)   (0.141)    (0.374)
                                      Panel B: Experimental Estimates with Controls
Impact of Migration                            0.183*       0.015        0.038       -0.466    -0.011     0.164
                                               (0.106)    (0.097)       (0.060)     (0.743)   (0.117)    (0.430)
Mean for Unsuccessful Stayer Households         0.447       0.605        0.629        1.50     0.676       1.18
Sample Size                                      146         146          143         146       142        123
                               Panel C: Estimates using Non-Experimental Control Groups
Including All Move Ballot Losers                0.044       0.051        0.024       -0.059    -0.157     -0.161
                                               (0.080)    (0.059)       (0.047)     (0.562)   (0.100)    (0.313)
Compared to Non-Applicant Stayers               -0.005     -0.041        -0.022      0.463     -0.020     -0.049
                                               (0.071)    (0.051)       (0.036)     (0.329)   (0.092)    (0.238)
Compared to All Non-Applicants                  -0.023     -0.055        -0.028      0.328     -0.020     -0.020
                                               (0.074)    (0.052)       (0.036)     (0.333)   (0.090)    (0.232)
Note: Experimental estimates are IV estimates where migration is instrumented with the PAC ballot outcome. Controls
are gender, age in months, age in months squared, birth order, log(household income in 2004 +1), maximum education
level in household, and whether the household lives on Tongatapu. Standard errors account for household-level
clustering. See text for non-experimental estimates.
*, **, and *** indicate significance at the 10%, 5% and 1% levels.




                                                                                41
                                    Table 10: Impact of Migration on Outcomes for Adults Aged 46 and above in Tonga.

                                       Currently     Currently                   Main                                  Alcoholic                               Diastolic
                                                                   Business                Very Good      Currently               Body Mass Waist to                          Mental
                                       Employed      Employed                  Activity is                             Drinks Per                               Blood
                                                                    Owner                   Health         Smokes                   Index   Hip Ratio                         Health
                                        (Males)      (Females)                 Agriculture                              Month                                  Pressure
                                                                 Panel A: Experimental Estimates without Controls
Impact of Migration                       -0.159       -0.163      -0.083      -0.130       -0.104      -0.102           0.463         -0.469       0.007        1.258        0.224
                                         (0.129)      (0.109)     (0.069)     (0.104)      (0.093)     (0.097)          (3.838)       (3.576)      (0.012)      (3.360)      (0.484)
                                                                  Panel B: Experimental Estimates with Controls
Impact of Migration                       -0.182       -0.113      -0.127      -0.111      -0.212*      -0.074           2.124         1.902        0.004        -1.144       0.146
                                         (0.140)      (0.156)     (0.112)     (0.125)      (0.116)     (0.128)          (3.997)       (2.073)      (0.014)      (3.561)      (0.537)
Mean for Unsuccessful
Stayer Households                         0.364         0.296         0.163        0.435         0.367         0.263          3.61        35.6        0.918        87.1         19.3
Sample Size                                 53            68           121          117           121            85            85         105          104         104           120
                                                             Panel C: Estimates using Non-Experimental Control Groups
Compared to Non-Applicant Stayers          -0.013       -0.164        -0.069       -0.206        0.103         0.051         -0.703      0.400        0.005       -2.506       -0.298
                                          (0.126)      (0.114)       (0.072)      (0.129)       (0.103)       (0.093)       (6.254)     (1.528)      (0.015)     (2.753)      (0.532)
Note: Experimental estimates are 2SLS estimates where migration is instrumented with the PAC ballot outcome. Controls are gender, age, years of education, height, labor earnings
in 2004, and whether the household lives on Tongatapu. Standard errors account for household-level clustering. See text for non-experimental estimates. Individuals of this age are
not eligible to migrate under the PAC, so there is only single selection (into migrant families), and not the selection into whose entire family migrates and who migrates for this age.
*, **, and *** indicate significance at the 10%, 5% and 1% levels.




                                                                                          42
          Table 11: P-Values and Family-Wise Adjusted P-Values for Models with Controls

                                                                           Adjusted P-value for Family-Wise Comparison
                                                       Single Variable                                    Westfall-Young
                                                           P-value          Bonferroni        Holm       Step-Down MinP
Adults Aged 18 to 45                                        0.000             0.000           0.000             0.000
Household Size                                              0.000             0.000           0.000             0.000
Net Remittances Per Capita                                  0.003             0.032           0.032             0.045
Net Remittances Per Adult                                   0.005             0.062           0.057             0.069
Household Has ATM card                                      0.006             0.050           0.050             0.129
# Meals of Fruits and Vegetables                            0.012             0.083           0.083             0.114
Household Labor Earnings Per Adult Equivalent               0.014             0.163           0.135             0.107
Adult Waist-to-Hip Ratio                                    0.016             0.197           0.197             0.111
Household Labor Earnings Per Capita                         0.020             0.237           0.178             0.138
Number of Chickens                                          0.045             0.402           0.358             0.369
Adult BMI                                                   0.052             0.629           0.629             0.268
# Meals of Fats                                             0.064             0.446           0.382             0.307
Household Has Bank Account                                  0.068             0.614           0.477             0.434
# Meals of Roots                                            0.069             0.480           0.343             0.307
Adult Main Activity Is Agriculture                          0.072             0.867           0.578             0.316
Older Adult in Very Good Health                             0.073             0.801           0.801             0.358
English Literacy for Children                               0.091             0.637           0.637             0.690
Log Total Income Per Adult Equivalent                       0.096             1.000           0.674             0.348
Note: This table shows the 18 outcomes for which the experimental estimates are significant (with controls added) at the 10
percent level when examined individually. See the text for more information about how the adjusted p-values are calculated.




                                                                                   43
                                      Appendix Table 1: Sample Size

                                             Full          Dropping        Percent      Dropping      Percent
                                            Sample        All Movers      All Movers Indv Movers Indv Movers
                                                     Individuals
Unsuccessful Ballots                           654            326            50%           160         51%
Successful Ballots - Non-Compliers             115             31            73%            11         65%
Successful Ballots - Migrants                  283            283                          272          4%
Non-Applicants                                 727            641            12%           604          6%
                                                     Households
Unsuccessful Ballots                           124            49             60%
Successful Ballots - Non-Compliers             26              8             69%
Successful Ballots - Migrants                  61             61
Non-Applicants                                 124            107            14%
Note: Successful Ballots - Migrants only includes Successful Ballots with members remaining in Tonga.

Relationship to Principal Applicant in        All                    Dropping Indv
                                                          Percent                         Percent
Migrant Households                       Individuals                     Movers
Principal Applicant                             5          1.8%              5             1.8%
Spouse                                          7          2.5%              5             1.8%
Own/Adopted child                              20          7.1%             11             4.0%
Son/Daughter-in-law                             6          2.1%              6             2.2%
Parent                                         45         15.9%             45            16.5%
Parent-in-law                                   4          1.4%              4             1.5%
Brother/Sister                                 73         25.8%             73            26.8%
Other Relative                                122         43.1%            122            44.9%
Non-Relative                                    1          0.4%              1             0.4%
Individuals                                   283                          272              11
Note: The non-dropped principal applicants, spouses and own/adopted children are outside the age
range eligible for the PAC




                                                                                   44