Document [original]

119DIE ERDE · Vol. 147 · 2/2016

Environmental change and migration

in coastal regions: examples from

Ghana and Indonesia

Felicitas Hillmann1,2, Usha Ziegelmayer1

1 Leibniz Institute for Research on Society and Space, Flakenstr. 29-31, 15537 Erkner, Germany, [email protected];

usha.ziegelmayer @leibniz-irs.de

2 Institute of Urban and Regional Planning, Technische Universität Berlin, Hardenbergstr. 40A, 10623 Berlin

Manuscript submitted: 17 August 2015 / Accepted for publication: 9 March 2016 / Published online: 30 June 2016

Abstract

Coastal regions worldwide have been focal points for migration as well as affected by environmental changes for

a long time. In the debate on climate change and migration coastal regions are among the “hot spot” areas that

are supposed to be prone to “climate migration” in the near future. The paper analyses the situation in two differ-

ent regional settings and advocates for a sound regional perspective on the relationship of environmental change

and migration. Based on the conceptual framework of migrant trajectories, the paper shows how popu lations in

Keta (Ghana) and Semarang (Indonesia), affected by similar environmental changes such as flooding and erosion,

react quite differently in terms of migration and mobility. The regional perspective as well as each region’s past

experiences with migration and environmental changes shows to be crucial in order to understand current reac-

tions to environmental degradation. The Keta setting represents a typology that pronounces migration trajecto-

ries as part of long-standing interregional and international migration, the Semarang setting, however, may be

classified as a rather typical modernization-induced migration scheme, linked to rapidly growing urbanisation,

with “trapped populations” on the one hand and in-migration of migrant workers on the other hand.

Hillmann, Felicitas and Usha Ziegelmayer 2016: Environmental change and migration in coastal regions: examples from Ghana and In-

donesia. – DIE ERDE 147 (2): 119-138

DOI: 10.12854/erde-147-9

Vol. 147, No. 2 · Research article

DIE ERDE

Journal of the

Geographical Society

of Berlin

Zusammenfassung

Weltweit sind Küstenregionen seit langem sowohl Ziel- als auch Ausgangspunkt für Migration. Sie gehören au-

ßerdem zu den Gebieten, die eine starke Dynamik hinsichtlich des Umweltwandels aufweisen. In der Diskussion

um Klimawandel und Migration werden sie als „hot spots“ bezeichnet. Es wird zudem angenommen, dass sie

besonders stark von sogenannter „Klimamigration“ betroffen sind bzw. sein werden. Dieser Artikel präsentiert

zwei empirische Fallstudien und plädiert für eine regionale Perspektive auf des Zusammenspiel von Umwelt-

wandel und Migration. Unter Zuhilfenahme des Konzeptes der migrant trajectories untersuchen die Autorinnen,

wie die mit dem Umweltwandel konfrontierten Bevölkerungen in Keta (Ghana) und Semarang (Indonesien), die

von ähnlichen Umweltveränderungen wie Überflutung und Erosion betroffen sind, auf unterschiedliche Mig-

rations- und Mobilitätsformen zurückgreifen. Erst durch die Einbettung des Umweltgeschehens in eine regi-

onale Perspektive, sowie die Einbeziehung der jeweiligen Geschichte von Migration und Umweltwandel in der

Region, lassen sich die aktuellen Mobilitätsdynamiken und Reaktionen auf Umweltwandel verstehen. Der Fall

Keta repräsentiert einen regionalen Typus, bei dem migration trajectories Teil einer schon lange bestehenden

inter regionalen und internationalen Migration sind. Der Fall Semarang dagegen tritt eher als typisches Beispiel

für durch Modernisierungstendenzen ausgelöste Migration hervor, in Verbindung mit rapide voranschreitender

Urbanisierung und ArbeitsmigrantInnen auf der einen Seite sowie „trapped populations“ auf der anderen Seite.

120 DIE ERDE · Vol. 147 · 2/2016

1. Introduction

All over the world, coastal regions are known to

be deeply entangled with regional migration re-

gimes and with population dynamics in general. At

the same time, coastal areas are in the very focus

of environmental change today, with sea level rise,

coastal erosion, salinization, land subsidence as

well as the depletion of mangroves.

This article investigates the relationship between mo-

bility patterns and environmental change in two coast-

al areas that have experienced environmental change

since many decades and have been subject to a vari-

ety of adaptation programs and technologies: Keta in

Southeastern Ghana and Semarang City1 in Central Java

(Indonesia). Both places are well documented in the

scientific literature and promise to function as show-

cases for the larger debate on the relationship between

environmental change and migration, but might not be

regarded as truly representative cases. The chosen re-

gions are distant from each other. But they show a vari-

ety of common traits that make a comparison attractive.

The two regional examples have been selected ex-

actly because of their long history of environmen-

tal change and migration. Both cases illustrate the

increasing role of human impacts on ecosystems in

coastal regions and the feedbacks of environmental

change on urban areas and populations, especially in

terms of mobility and migration. Because they have

been experiencing an active integration into regional

migration systems for a long time, they possibly also

allow for a more detailed and sound analysis of the

presumed interrelations of migration and environ-

mental change. In contrast to many studies in this

field, our research does not refer to projections or

potentials of environmental or climatic change, but

it deliberately focusses on two regions in which mas-

sive environmental changes have already occurred

over the last decades (“rapid change”). This method-

ological approach meets our research interest best,

because it also allows to observe the political and

discursive processing such as narratives within the

normative setting of the region over time.

This article presents the first results of an empirical

study that has been conducted in 2014/15. It echoes

the interpretation of stories that have been told for long

by scientists, local experts and migrants themselves

about these regions. How did and do people in Ghana

and Indonesia perceive climate change and environ-

mental threats themselves, which coping and adapta-

tion strategies did or do they adopt concerning mobil-

ity and immobility? Which narrations go along with the

observable patterns of migration and mobility?

The article is organized along six sections: After this

introduction, we first summarize the main traits of the

debate on environmental change and migration and

then introduce the conceptual framework of migra-

tion trajectories, an emergent field within migration

studies. Built on this, we present the main trends and

pattern of migration and mobility in the two selected

countries in a nutshell and then highlight the main

environmental changes over the past years. Section

three gives an insight into the adopted set of meth-

odologies. Section four presents the first preliminary

results of the empirical work we have undertaken.

The fifth section discusses the results in the light of

theory and the presented literature review. The paper

concludes by highlighting the main outcomes and by

identifying lacunas for future research.

2. State of the art: migration and environmental

change

The question as to whether and how climate change

affects human mobility has been widely debated

both within the policy context and in academia.

Piguet et al. (2011), Gemenne (2010) and Afifi and

Jäger (2010) produced seminal edited volumes,

Hugo (2013) presented a 1000-page best-of reader

on the conceptual debate, and the Foresight Report

(2011) has guided the policy-debate.

2014 was reported as the warmest year since 1880

(NASA 2015). This fact confirmed what the IPCC re-

port (IPCC 2013) put forward as its main message on

the dynamics of climate change: the warming of the

climate system is unequivocal, and the changes are

unprecedented in recent millennia. The effects of

these changes in the climate are expected to increase

the likelihood of internal and international migration

through increased frequencies of droughts and floods,

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

Keywords Migration, environmental change, climate change, migration trajectories, Ghana, Indonesia

121DIE ERDE · Vol. 147 · 2/2016

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

desertification, salinization of soils, coastal erosion

(due to sea-level rise), more storms and extreme

weather events and competition for scarce resources

(Martin 2012). The threat of climate change and en-

vironmental degradation on collective resources such

as fresh water, land and agriculture is enormous and

will increase in the near future. Unscrupulous exploi-

tation of natural resources such as the extraction of

oil and metals – regardless of ecological risks – ad-

ditionally challenge climatic variations (Klein 2014:

215ff.). High pressure on natural resources is further

provoked by rapid demographic growth in many plac-

es in the world and an increasingly extractive global

economy adds to the culmination of environmental

crises in many parts of the world (Sassen 2014).

The debate on the nexus of climate change and migra-

tion already began in the 1990s. The questions as to

when a decision to move is taken, and whether and

under what circumstances a tipping-point for migra-

tion is reached, is among the hottest topics within this

debate, accompanied by estimates and speculation

about the numbers of future migrants. The question

on the quantitative dimension is more easy to ans wer

when concentrating on the effects of sudden climate

events, such as thunderstorms and heavy rains that

create devastating circumstances and threaten local

livelihoods, especially in low-lying coastal areas. The

annual number of newly internally displaced people

as a result of natural disasters oscillated between

31.7 million people in 2010 and 22 million people in

2013 (Gemenne et al. 2014). When it comes to slow-on-

set environmental change the question on the dimen-

sion and the dynamics of migration is more difficult

to answer. Longitudinal studies are lacking and up to

today the process of decision-making itself remains a

black box within many migration studies. The nexus

between environmental change and migration and

mobility is not straightforward, but filtered by a va-

riety of more general developmental issues (cf. Black

2001; Black et al. 2011; Tacoli 2011).

The current debate is best characterized by three

analytical pitfalls: an extremely high degree of po-

liticization, a blurred terminology and a missing spa-

tial differentiation and conflation of various levels of

analysis, i.e. global, national and local (Hillmann et al.

2015b). Some authors also claim that the blurred ter-

minology refers to the field itself because it is thought

to be constructed strongly communicatively. The au-

thors claim that, in many cases the distinction, which

factors are perceived to be environmental ones and

decisive for migration depends very much on the per-

spective applied. In many cases, the discourse and the

narrative going along with migration and environ-

mental change, may be more influential for migra-

tion decisions than the environmental change itself

(Aufenvenne and Felgentreff 2013: 21f).

2.1 Theoretical framework: migrant trajectories

Against the background of these pitfalls in the de-

bate, this paper refers to the conceptual approach

of migrant trajectories. It considers this concept

as one way of bridging the analytical gap between

various scales (global, national, local) by embed-

ding migratory action as one form of social and spa-

tial organization in regional settings that are under

stress, i.e. environmental stress.

The term migrant trajectory (when referring to mi-

gration itself: “migration trajectories”) calls for an

analysis of all phases of migration, departure(s),

transit(s) and arrival(s). Migration trajectories work

“as the collective and ‘visible’ outcome of many in-

dividually migrating persons” (Spaan and Hillmann

2013: 65) and do pronounce the non-linear charac-

ter of migratory processes, implying “a shift in focus

from the individual to a structural view on migration

processes by concentrating on the outcome of collec-

tive biographical paths or as a repeated spatial ar-

rangement of migrants” (Spaan and Hillmann 2013:

65). Migrant trajectories are defined as “composed

of one or more episodes (change of countries) and

one or more status (periods of residence in differ-

ent countries)” (Castag none 2011: 4f.). The focus of

analysis here is on the process rather than on a sin-

gle event by an individual in time and space.

Inspired by migration theories like transnational-

ism, trans- and transit migration and migration

networks the concept of migrant trajectories aims

at linking (concrete) individual migration projects

with (more abstract) migration regimes. This way,

a new conceptualization of the migration process

is pushed forward – pronouncing the path depend-

ency of migratory action and by taking into account

materiality. “Transmigration is, however, not just a

trajectory but a multiplicity of potential trajectories”

(Grillo 2007: 200) – thus pronouncing fluidity and in-

stability. Vaittinen for example stresses the link be-

tween the macro-level of collective movements and

institutions on the one hand and individual human

122 DIE ERDE · Vol. 147 · 2/2016

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

beings on the other hand, all related through the mi-

grant trajectory (Vaittinen 2014: 8f )2 .

As emphasized by van der Velde and van Naerssen

(2007b: 150f.; 2015) migration trajectories further

refer to spaces of belonging. They take into account

geographical and mental distance as well as percep-

tion as constituting thresholds for migration in a glo-

balized world, putting emphasis on the existence and

importance of “soft” and culturally bounded factors

as contributing to the dynamics of migration.

Accordingly, a growing body of literature on the re-

lationship of migration and environmental change in

developing countries refers to the influence of exist-

ing gender regimes (cf. Awumbila and Tsikata 2010).

Some authors point to existing or changing power re-

lations (cf. Carr 2005) and, more generally, to phases

of immobility (cf. Schapendonk 2010) during the mi-

gration project as contingent of the mobility regime.

Fragmented journeys, known as being typical for the

movement of refugees (Collyer 2010), start becoming

a normal feature also for regular migratory patterns.

We interpret the migrant as part of a larger structure,

a field of knowledge, of perception and integration. So

does Carr, in his study on environmental change and

migration in Central Ghana, bringing in Foucault’s

concept of power/knowledge and hereby stressing

that the interpretation of the world and the environ-

ment is (re)produced by power: “Such an approach,

however, moves beyond the existing literature by

shifting the focus of study from conditions that drive

migration […] toward the local power/knowledge in

which environment, ecology, and politics are under-

stood. The ways migrants negotiate and transform

their context, and the objectives behind such negotia-

tion and transformation, are the condition and result

of this understanding” (Carr 2005: 929).

Here, power relations constitute an important aspect of

migrant trajectories; it admits power of agency to the

migrant him- or herself, definitory power over space. Mi-

grant trajectories position the migrant within migration

regimes, as fed by migration industries in the respective

regions and on different regional scales. A migration re-

gime delineates the “interplay of juridical regulations,

the political handling of migration, migration traditions

and cultures of migration in different parts of the world”

(Hillmann 2016: 160, translated by Ziegelmayer).

Another ‘spatial perspective’ in the debate of migra-

tion and environmental change is put forward by

Black and Collyer (2014). The authors claim that today

“trapped populations” do exist, people that – out of a

lack of resources or other restrictions – are unable to

move when confronted with an environmental shock

situation. Further, Schapendonk stresses the idea of

trajectories as channelling migration. “Migrants’ tra-

jectories are not closed-off corridors but open and

process-like phenomena. They are influenced by,

among others, the trajectories of other people, ob-

jects, capital, rules and information” (Schapendonk

2012: 32). In this paper we frame “trajectories as spa-

tial routes connecting place of origin and places of

desired destination [which] constrain or facilitate the

movement of the actor in space” (van der Velde 2008:

117). Based on the presumption that migration is not

entirely based on rational decision-making this arti-

cle purposefully includes cultural factors like migra-

tion myths and narratives in the analysis. According

to Abu et al. (2013) it is crucial to understand how

people concerned perceive climate hazards and their

own vulnerability towards it in order to understand

migration decisions in the climate change – migra-

tion nexus. In this context the authors argue that a

particular (environmental) event may be perceived

as “normal” when it is experienced over a long period

and therefore might not push people to migrate.

2.2 The relevance of migration in Ghana and Indonesia

As shown above, migration trajectories are rooted

within migration regimes. Indeed, in Ghana there is a

“long history of mobility with migration playing a cen-

tral role on the livelihood and advancement strategies

of both rural and urban populations” (Awumbila et al.

2011: 1; cf. Manuh 2001: 19). Due to the colonial his-

tory of the country, there is a legacy of internal north-

south migration. People migrated first due to “forced

migration through labour recruitment” (van der Geest

2011a: e72), later then, somewhat more voluntarily,

for working in the cocoa plantations in the south. Gen-

erally, the European influence through slavery, mis-

sionaries and colonialism led to still existing ties be-

tween Ghana and the former colonial powers trough

business, religious and family ties.

For a long time Ghana was a net immigration coun-

try – immigrants stemmed from the neighbouring

countries and other ECOWAS (Economic Community

of West African States) member states – before chang-

ing to become a net emigration country in the 1960s

(Awumbila et al. 2011: 1). Nowadays Ghana experienc-

123DIE ERDE · Vol. 147 · 2/2016

es all types of migration: “internal migration, immi-

gration, transit migration and emigration both within

and outside of Africa” (Awumbila et al. 2011: 1). Ac-

cording to Peil (1995: 348f.) economic decline and the

lack of educational and job opportunities led to inter-

national out-migration especially by young Ghanaians

since the mid 1970s. Just after the country’s independ-

ence in 1957, the first president, Kwame Nkrumah, fol-

lowed a rather liberal immigration policy, but already

in 1969 the Aliens Compliance Order was installed

and many immigrants were expelled out of Ghana.

The oil boom in Nigeria and the economic success of

the neighbouring Ivory Coast led to out-migration to

those two as well as other ECOWAS member states. In

1979, ECOWAS installed a Protocol on the Free Move-

ment of People between the respective countries.

Failing economic policies in Ghana brought economic

decline and “migration re-emerged as a tried and test-

ed strategy for many Ghanaians” (Manuh 2001: 19).

High-skilled migration from Ghana especially took

place in the health sector, for example to the US and

the UK (Tankwanchi 2012; Teye et al. 2014).

Like Ghana, Indonesia is highly dynamic in terms

of population and migration. As a former Dutch and

partly Japanese colony, Indonesia gained independ-

ence in 1945 and saw massive forced migration

schemes under the dictatorship of Suharto until

1998. From the 1950s to the 1990s Indonesia ex-

perienced a steady increase in out-migration. The

net migration rate has always been negative since

1950. Between 2010 and 2015 it was around -0.8

(migrants per thousand inhabitants) (IOM 2010).

Immigrants account for not even 0.1 % of the total

population (The World Bank 2011: 24).

Indonesia today is characterized by strong labour

migration to Asian countries, especially Malaysia,

and to the Middle East (IOM 2010). These migra-

tions are often chain migrations and the migrants’

families strongly support the migration decision

in most cases. Half of those Indonesian migrant

workers stay abroad illegally and some are con-

fronted to inhuman treatment by their employers

(Brown and Brownlee 2000).

Migration from Indonesia often takes place undocu-

mented (IOM 2008) with 78 % of labour migrants

being female in 2007 (IOM 2010). Between 1996 and

2007 the number of women migrating overseas has

nearly doubled, while the number of male migrants

decreased – Saudi Arabia and Malaysia being the top

destinations (Platt et al. 2013: 10f.). Skilled emigra-

tion targets towards the OECD countries, especially

to the Netherlands, due to historical colonial ties. A

big share of this international migration is organized

by agencies as part of the above-mentioned migration

industry (cf. Spaan and Hillmann 2013: 68f.).

In Indonesia’s history colonisation and the trans-

migrasi program were used by the ruling powers in

order to prevent social upheaval on the main island

Java and to get rid of potential sources of political in-

stability in other areas (Brown and Brownlee 2000)

The so-called transmigrasi program aimed at reduc-

ing the number of inhabitants on the main island Java

by populating outer islands with Javanese people.

First exercised during Dutch colonization from 1905

to 1941, the second transmigration program was

launched after independence in 1950, initiating also

voluntary migration (Fearnside 1997: 553f.). Many

NGOs criticized the program, since it did not reduce

population pressure in Java, but rather provoked so-

cial and political conflict on the outer islands.

In recent decades, foreign and national investments

in the metropolitan area of Jabotabek (i.e. larger Ja-

karta), as well as in secondary urban centres like

Bandung (West Java) and the study area Semarang

(Central Java), have attracted rural populations from

Java and other islands (Firman 2004: 422ff.). These

migratory movements are mostly circular, leading to

increasingly diversified income bases of both rural

and urban households and reciprocal relationships

between sending and receiving areas (Hugo 2008b:

50ff.). Regarding international migration movements,

the Javanese provinces are among the main sending

areas of official, registered emigration from Indone-

sia, mainly to the USA, the Netherlands and the Middle

East (Hugo 2008a: 51; Spaan 1999). All over Indone-

sia poverty-reduction rates have not been in tandem

with economic growth (Platt et al. 2013: 11f.) and un-

employment rates are significantly higher for women.

Ghana and Indonesia both count a strong diaspora

population with 719,404 Ghanaians and nearly 3 mil-

lion Indonesians living outside their country of ori-

gin. The main destination countries are in the region

of the respective countries with 47 % of the Ghanaian

migrants living in other African countries, followed

by the US (21 %), UK (11 %), Italy (6 %) and Germa-

ny (3 %). The Indonesian diaspora is mainly living

in Malaysia (35 %), Saudi Arabia (13 %), the United

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

124 DIE ERDE · Vol. 147 · 2/2016

Arab Emirates (11 %), Netherlands (5 %) and the US

(4 %; all figures from UN-ESA 2013). Both countries

show great flows of remittances (The World Bank

2011: 24) – contributing to local development end re-

inforcing further migration.

2.3 Migration and environmental/climate change in

Ghana and Indonesia

In both countries, migration due to environmental

changes is not a recent phenomenon. The people in the

savannah of West Africa e.g. had to deal with climatic

variability and environmental stress for centuries:

“Human mobility has a very central place in the adap-

tive strategies of farmers and pastoralists” (van der

Geest 2011a: e70) and is an “omnipresent character-

istic of the livelihoods and culture of northern Ghana-

ians” (van der Geest 2011a: e90).

Migration does not necessarily occur immediately af-

ter environmental change takes place (cf. Abu 2011;

Carr 2005; van der Geest 2009, 2010, 2011a). But

people are very aware of climate-related changes.

Migration intentions show to be strongly linked to

socio- demographic factors like age, sex, education,

migration status etc. of the head of the household

(Abu et al. 2013: 1). Moreover, seasonal, circular or

temporary out-migration patterns are already com-

mon where land-dependent residents search for al-

ternative incomes during difficult times and regular

movement has become a part of the regional culture

(CMS 2011: 2). Migration works as one of numerous

strategies of risk diversification, used by households

in an environment challenged by climate change (Abu

et al. 2013: 2; cf. Kwankye et al. 2009; Yaro 2010: 61).

Abu concludes that rural populations are capable of

dealing with just one climate-related event at a time,

e.g. flood or drought. It generally gets difficult for the

resident population when several climate-related

events occur at one point in time and/or cannot be

predicted beforehand (Abu 2011: 9). Van der Geest

shows in his longitudinal analysis that times of great

environmental stress in Northern Ghana were rather

times of reduced out-migration (van der Geest 2011a:

e69). ”In the late 1980s and in the 1990s, [however,]

a time of environmental recovery in northern Ghana,

migration increased again” (van der Geest 2011a: e89).

Like Ghana, also Indonesia shows a combination of

anthropogenic and ‘natural’ environmental change.

Increased landslides and flooding are partly caused

by the transformation of agricultural or forest land

into settlements. According to the World Bank the

economically most important regions of Bali, Java,

Sumatra, and Papua are also most vulnerable to cli-

mate change (The World Bank 2010). In a study on

province-to-province migration in Indonesia over

15 years, Bohra-Mishra et al. (2014) claim that climatic

variations influence permanent migration, while dis-

asters have smaller or no impacts. The study especial-

ly indicates a nonlinear effect of temperature on mi-

gration, “such that above 25°C, a rise in temperature

is related to an increase in outmigration, potentially

through its impact on economic conditions” (Bohra-

Mishra et al. 2014: 9780). Indonesia’s tropical climate

combined with a high dependence on agriculture and

a long tradition of interprovincial migration might in-

crease the effect of temperature rise on interprovin-

cial migration (Bohra-Mishra et al. 2014).

A UNICEF report on migration and climate change

in Indonesia emphasizes that climate change in

Indonesia, as well as in many other parts of the

world, “is likely to exacerbate and aggravate ex-

isting patterns of migration rather than to create

new flows” (Urbano et al. 2011: 32).

As the world’s fourth most populous country and

with a population of over 60 % living in coastal areas

(Purwaka and Sunoto 1999) Indonesia is highly vul-

nerable to climate change. The population density in

Java is the highest in the whole country (Handayani

and Kumalasari 2015), which leads to a close inter-

play of urbanization, population growth, environ-

mental change and migration.

Densely built environments create a “heat islands”

effect, while the burning of forest leads to smoke

over large areas (Urbano et al. 2011: 8, cf. UN Habitat

2010: 6f.). The country is further exposed to geologi-

cal and climatic hazards like “flooding, landslides,

extreme weather events, droughts” (Bohra-Mishra

et al. 2014: 1), sea-level rise and changes in tempera-

ture at land and at sea. These changes occur as rapid

and as slow environmental changes. Due to urbanisa-

tion, coastal cities grow rapidly, despite flooding and

land subsidence, and more than 25 % of the country’s

poor inhabitants live in coastal areas (Handayani and

Kumalasari 2015), which makes them even more vul-

nerable to climate change (Marfai et al. 2008: 237).

Handayani and Kumalasari (2015) state in a study on

Java that out-migration from areas affected by cli-

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

125DIE ERDE · Vol. 147 · 2/2016

mate changes has not yet been chosen as an adapta-

tion strategy. Coastal cities like Jakarta, Semarang

and Surabaya attract migrants despite environmental

hazards, and local populations show a strong attach-

ment to their domicile. This attachment to place as

well as the dependence on the sea for their livelihoods

and the closeness to industrial zones as working place

seem to influence the perception of migration, an im-

portant part of migrant trajectories.

3. Methodology

The research results presented in this paper are

part of on-going research produced by a project con-

sortium named “New Regional Formations – Rapid

Environmental Change in Coastal Regions of Ghana

and Indonesia” (NRF), bringing together a multidis-

ciplinary team. In order to understand the interrela-

tion of migration trajectories with (climate-induced

and man-made) environmental change our empirical

research analyses two regional cases that for many

years have experienced a fragile ecological situation,

characterized by frequent flooding and/or erosion.

The two regions show high population dynamics. In

our methodological framework we decided not to

compare the two situations as such but to confront

two regional settings in order to sharpen the perspec-

tive on common and distinct features regarding the

nexus of migration and environmental change. We de-

liberately turn away from a conceptual approach that

sees migration as caused by environmental change,

respectively environmental change as a root cause for

migration. In order to understand the full picture we

started with explorative studies in the field and first

expert interviews, while reviewing existing docu-

mentation on the two cases. Following the methodo-

logical tradition of triangulation we adopted a mix of

qualitative and quantitative methods during the field-

work. In each of the two regions a quantitative study

was conducted. It consisted in a structured household

survey (274 households were interviewed in Keta in

January 2016; 333 households in Semarang in Decem-

ber 2015). The respondents were composed as fol-

lows: Keta: 58 % male-headed, 42 % female-headed

households, representing 1,345 household members

(47 % male, 53 % female) and Semarang: 82 % male-

headed, 18 % female-headed households representing

1,417 household members (48 % male, 52 % female).

For the purpose of this study we defined “migrants”

as persons who leave their household for more than

three consecutive months in order to live in another

place, at least another sub-district (Ind.: Kecama-

tan) of Semarang City or another community of Keta

Municipality respectively or further away. The re-

spondents in the survey were the household heads

who were asked about all household members who

once used to live in the household and then moved

away for more than three consecutive months.

We also analysed recent census data (2010) that al-

lowed us to understand the bigger picture in terms of

population dynamics.

The qualitative data were gathered via semi-struc-

tured interviews with migrant households from Keta

in the capital Accra and neighbouring Togo, two desti-

nations of migration from Keta, as well as key inform-

ant interviews with traditional and state authorities

in Keta. In Semarang the qualitative interviews con-

sisted of key informant interviews with heads of vil-

lages and sub-districts in Semarang Municipality.

The two case studies selected stand as two exam-

ples for the complex relationship between environ-

mental change and migration in coastal areas in

tropical countries. They are neither representative

for the whole country of Ghana or Indonesia respec-

tively, but rather could be interpreted as showcases

of the broader picture.

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

Fig. 1 Location of Keta in Ghana (all maps: drafts by Ziegel-

mayer and Hillmann, design by Kartographieverbund

TU Berlin)

126 DIE ERDE · Vol. 147 · 2/2016

3.1 Description of study site no. 1: Keta (Ghana)

Keta is a place with an outstanding tradition of migra-

tion and environmental change (Akyeampong 2001).

The local population of the Anlo-Ewe, the dominant

ethnic group in the area, lives in a sometimes ambigu-

ous relationship with nature which on the one hand

provides their livelihoods and on the other hand is a

threat in forms of droughts or floods. They are also

aware of the (negative) consequences of large devel-

opment projects for their environment (Akyeampong

2001: 2). The Anlo-Ewe migrated from Notsie, which

today belongs to Togo to the south-eastern Ghanaian

coast in the 17th century. This ethnic group cherishes

its own migration saga (Kumassah 2009). The Anlo-

Ewe originally were no maritime people and continue

to have a fierce attachment to land (Akyeampong 2001:

215). Local perceptions of environmental change can

be traced back to coastal erosion from the early 20th

century onwards causing loss of property and land in

Keta, experienced as a loss of regional pride and en-

couraging emigration (Akyeampong 2001: 217, 219).

Already in the 1980s two thirds of the town were de-

stroyed by the sea (Akyeampong 2001: 206).

Keta Town is located on a small sand bar between the

Golf of Guinea in the south and the Keta Lagoon in the

north (Fig. 1, Fig. 2). The town has a long history as an

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

Fig. 2 Keta and its environs

9.000

8.000

7.000

6.000

5.000

4.000

3.000

2.000

1.000

2.000

3.000

4.000

5.000

6.000

7.000

8.000

9.000

0-4

5-9

10-14

15-19

20-24

25-29

30-34

35-39

40-44

45-49

50-54

55-59

60-64

65-69

70-74

75-79

80-84

85-89

90-94

95-99

Male

Female

Fig. 3 Keta’s population pyramid, 2010; (source: Keta

Munici pa lity 2010: 35; design by Ziegelmayer)



127DIE ERDE · Vol. 147 · 2/2016

important market place with traders from neighbour-

ing regions and countries and a small port. However,

in the 1960s the city experienced an economic decline,

as a result of a mix of political, economic and envi-

ronmental reasons. One was the opening of the new

harbour in Tema westwards, another the cutting of a

canal trough the sandbar near the market in order to

open the lagoon into the sea and decrease the flood-

ing pressure. This intersection unexpectedly led to a

large flooding in Keta town as the canal grew bigger

than planned for. Additionally, in 1968 the regional

capital was transferred from Keta to Ho and in 1969

the Aliens Compliance Order expelled many foreigners

doing trade at Keta market (Akyeampong 2001: 183f.).

These factors, in combination with the destruction by

the erosion, induced a period of economic decline in

the town of Keta that lost its market functions. This

development was accompanied by out-migration in the

last decades (Interview chief (traditional authority)).

The Volta Region, where the study area Keta is locat-

ed, shows a population growth of 2.5 % from 2000-

2010, corresponding with the national rate (GSS

2012: 21). The age structure of the resident popula-

tion in Keta Municipality with about 150.000 inhab-

itants shows a very young population: 35 % under

the age of 15, 53 % between 15 and 59, and 12 % old-

er than 59 years (Fig. 3). This population structure

seems likely to see strong out-migration of the young

economically active population in the near future.

In comparison with other regions in Ghana, the Volta

region with Keta Municipality features the second

highest out-migration, with a net migration rate (per

1000 inhabitants, from 2000 to 2010) of -258.60, after

the Upper West Region with -302.50 (GSS 2013: 212).

3.2 Description of study site no. 2: Semarang (Indonesia)

In contrast to Keta, Semarang City (Fig. 4, Fig. 5) is a city of

1.5 million inhabitants, which regularly experiences tidal

and river flooding and other environmental changes such

as land subsidence and landslides. Never theless it fea-

tures a population growth rate of 1.4 % per year, “higher

than the surrounding areas” ( Mulyana et al. 2013a: 7)

indicating the city’s “importance in attracting migrants

from surrounding areas” (Mulyana et al. 2013a: 7). The

main livelihoods in Semarang city are industrial work-

ers (25 %) and construction workers (13 %), govern-

ment employees/armed forces (16 %), services (10 %)

and farmers (5 %). Semarang is a city of trade, hotels and

gastronomy. About a third of its popu lation lives in pov-

erty, with substantial local differences. Neighbourhoods

such as Kemijen show a majority of the population living

in poverty (Mercy Corps et. al. 2010: 15f.). Figure 5 shows

the study site indicating the flooded areas as well as the

sub- districts with negative net migration rate near the

coast and in the city centre according to the census.

Semarang City reveals an even higher percentage of

inhabitants in the productive age than the Ghanaian

case: 70 % are between 15-59 years old, 24 % under

15 years and only 7 % older than 59 years (Fig. 6),

indicating on-going in-migration of migrant workers

from surrounding rural areas into the city.

4. Results

This section presents the results revealed by the

quantitative and qualitative research. It highlights

the regional embeddedness of migration trajectories

and environmental change in the two study areas.

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

Fig. 4 Location of Semarang

in Indonesia

128 DIE ERDE · Vol. 147 · 2/2016

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

4.1 Case study Keta (Ghana)

Environmental change in Keta consists of various phe-

nomena. Coastal erosion affects the whole Ghanaian

coastline with Keta as one of the most threatened ar-

eas with 2.66 m of loss of coastline per year (Appeaning

Addo 2014: 9). Additionally, a sea-level rise (SLR) of ap-

proximately 3 mm/year (Boatemaa et al. 2013) affects

Keta. The local population is increasingly aware of SLR

(Interview resident resettlement area and community

elder). Furthermore, changing rainfall patterns influ-

ence agriculture in the area: rainfall is “erratic, it falls

at times you don’t expect it” (Boatemaa et al. 2013).

Apart from this ‘natural’ environmental change there

is a wide range of man-made environmental change

as for example depletion of mangroves (Awumbila and

Tsikata 2010) and the overfishing by big European and

Asian trawlers (Interview resident resettlement area

and community elder; Interview chief (traditional au-

thority). The depletion of mangroves of course affects

the on-going coastal erosion and is an indicator for the

entangled effects of man-made environmental changes

in the area. Additionally, the consequences of the Ako-

sombo Dam led to changing sedimentation (Akyeam-

pong 2001: 170ff.). This affected the fishery livelihoods

in the Lower Volta, including Keta Lagoon, and led to

out- migration towards the Volta Lake (Awumbila and

Tsikata 2010: 125; Interview Keta Municipality).

Fig. 5 Semarang and its environs

(migration data according

to BAPPEDA and BPS Kota

Semarang 2012: xiv)

Fig. 6 Semarang’s population pyramid, 2010; (own calculations,

after BPS 2010, design by Ziegelmayer)

90.000

80.000

70.000

60.000

50.000

40.000

30.000

20.000

10.000

20.000

30.000

40.000

50.000

60.000

70.000

80.000

90.000

0-4

5-9

10-14

15-19

20-24

25-29

30-34

35-39

40-44

45-49

50-54

55-59

60-64

65-69

70-74

75-79

80-84

85-89

90-94

95-99

Male

Female

129DIE ERDE · Vol. 147 · 2/2016

A major sea defence (see Fig. 2) built with US financial

support during 1999-2004 is supposed to protect the

coastline – but is likely to shift problems of erosion

further east (Angnuureng et al. 2013: 107). The resi-

dent population is positive about this intervention and

still hopes that the road that was destroyed by erosion

and rebuilt as part of the project will reopen the way

for traffic to neighbouring Togo and provide access to

more markets (Kumassah 2009). Up to now however,

the Keta market, for example, has not regained its for-

mer economic role. The chief in neighbouring Blekusu,

east of Keta, which is not protected by the sea defence,

claims, in contrast, that high tide affecting his com-

munity has become stronger since the defence was

built (Interview chief in Blekusu) and the community

lost more houses in summer 2015 due to erosion and

flooding. In March 2016 construction works for a sea

defence were supposed to start in the area of Blekusu

(Ketu South District).

Analysing contemporary migrant trajectories in this

changing environment, our survey reveals that 65 %

of all interviewed households have experienced out-

migration of some of their members. 53 % of those

out-migrants were male and 47 % female3. A small

share of these migrants was accompanied even by

children leading to more than 800 out-migrants in-

cluding children in the interviewed households.

Most of the migrants were son or daughter of the

household head (38 %), followed by “other relative”

(24 %) and brother/sister of the household head

(21 %). Most of these migrants had completed mid-

dle school (34 %), followed by graduates from senior

high school (26 %) and those who only had primary

school education (20 %). When leaving most migrants

were between 10 and 19 (45 %) or 20-29 years (37 %)4

old. An interviewed chief refers to the age structure

of out-migrants with mostly youths leaving and puts

migration in relation to lifecycles; “when we retire we

come home, but the active population is outside and

is not helping us to develop our area, because when

you’re young, you can develop your area”.

We claimed that migrant trajectories put forward a

view on migration as a non-linear process. Indeed,

about a sixth of the out-migrants were return mi-

grants who left Keta and then came back for a pe-

riod of more than three consecutive months. Also

qualitative research evidenced the important role

of return migration. Some international migrants

recently came back to invest in tourism, into health

facilities or a “pure water” production factory in

Keta Municipality (Interview resident resettlement

area Kedzi, community elder; Interview community

elder Anloga) thereby contributing to local develop-

ment. Development studies also point to the growing

importance of chieftaincy for regional development

(Kleist 2011), referring to the institutional aspects of

migration trajectories.

The fact that 12 % of all out-migrants in the inter-

viewed households indicated several destinations

also stresses the process-like character of these mi-

grant trajectories. The majority first moved to big cit-

ies, mainly Accra. For 16 % of these multiple migrants

(two destinations or more) the second destination

was another continent, indicating step-wise migra-

tion (cf. Castagnone 2011) with Accra being the start-

ing point for the later international migration.

Looking at single and multiple migrants, the NRF sur-

vey indicates the destinations shown in Table 1. In

spatial terms, these figures clearly show the region-

al embeddedness of migrant trajectories from Keta

with a major orientation towards the capital Accra

and 86 % of all out-migrations being internal within

Ghana. The above-mentioned long migration history

of Keta is also confirmed by the survey, with 27 % of

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

Destination Share of all out-

migrations (%)

Big cities, mainly Accra

Other Ghanaian regions

Other places in Volta Region

Neighbouring countries

(Togo, Ivory Coast)

Other African countries

(mainly Nigeria)

Keta Municipality

Neighbouring municipalities

Other continents

Total

100%

Table 1 Destinations of out-migrations from Keta

(NRF Survey 2015)

130 DIE ERDE · Vol. 147 · 2/2016

the migrants who migrated more than 10 years ago,

18 % migrating 6-10 years ago and a recent peak of

out-migration during 2008-2013 (31 %)5 and only 3 %

migrating less than one year ago.

In contrast to conventional knowledge of mainstream

migration literature (cf. The World Bank 2014), the

remittances of those out-migrants indicated above

did not seem to play a major role within the migration

process: 50 % did not remit, 22 % sent money, 19 %

sent money and goods, 5 % sent goods, mainly food.

Besides this high significance of out-migration to-

wards places outside Keta Municipality, the survey

indicates high internal mobility within the munici-

pality. One reason for this internal mobility is coastal

erosion. 18 households out of a total of 91 households

which moved within Keta Municipality mentioned the

destruction of their house by the sea as the reason to

move. This hints at short distance migration of whole

households as one adaptation strategy to coastal en-

vironmental changes like erosion. Out-migration of

single individuals, however, rarely seems to be direct-

ly related to environmental change. A total of four out-

migrants, a share of 1 % of all out-migrants, left their

former house due to flood or destruction of the house

by the sea according to the interviewed households.

Mobility within Keta Municipality is directly linked

to environmental change when taking into account

the governmental resettlement program, of which

5 % of the interviewed households have been part.

Residents who lost their houses due to coastal ero-

sion get new bungalows on the northern side of the

town. That land there has been reclaimed from the

lagoon, the reclamation thus adding to man-made

environmental change.

The dominant reasons mentioned for the continuous

out-migration from Keta were missing work opportu-

nities and an incomplete educational infrastructure.

Against the background of Keta’s past as a prosper-

ous town, today’s lack of industries in the region is

reported to be one main driver for out-migration. One

of the reverends stated that youths often consulted

him before leaving and asked him to pray for them

so that they would find a job in Accra. According to

him, many youths would prefer to stay in Keta if there

were jobs for them (Interview reverend).

In addition to this lack of job opportunities an in-

terviewed chief draws a closer look on education

as a motivation for migration. Although there are

educational facilities up to the secondary level in

Keta and some tertiary institutions, this “education

makes us un-useful in our own environment“, as the

chief states. Students are not trained to be success-

ful in fishing or agriculture, the dominant economic

activities in the area, and those who want to pur-

sue their education at university level have to leave

for the bigger cities. There is neither a university in

Keta nor a training institute for fishing or agricul-

ture that could provide adequate education for the

area (Interview chief (traditional authority)).

The fact that international migration outside Africa

plays a minor role in the survey supports the argu-

ment that there is a weak link between ongoing en-

vironmental change at the coast and international

migration. Migration in a context of environmental

change rather seems to be internal. As one respond-

ent clearly stated, “those who have left for Europe

and the US went there to work. They did so as a

result of work and not because of the destruction

of the sea” (Interview resident Kedzi resettlement

area, community elder).

As put forward under 2.1, migration myths and nar-

ratives are pivotal in order to understand migration

trajectories in Keta. The Anlo-Ewe belong to the Ewe

group that is known for a long tradition of fishermen

migration along the West African coast (cf. Akyeampong

2001: 133ff.; Odotei 2002a, 2002b). Until today, fishing

is one of the main economic activities in the area and

respondents in the interviews often refer to the migra-

tion tradition of fishermen from Keta: “Our people here

are fishermen, a lot of them do fishing. And they migrate

a lot. They migrate towards the west coast of West Afri-

ca. Some even move as far as to Canary Islands, Senegal,

Gambia and all that, they fish, they stay over a year, then

they come back. Some finally settle in Abidjan and all

other places” (Interview Keta Municipality).

This long tradition of (fishermen) migration in the

area explains partly today’s rather positive percep-

tion of migration as indicated in the survey. Migration

is seen as bringing (economic) benefits to the indi-

vidual household (64 %) as well as to the community

(76 %), but people are also aware of conflicts within

the community due to migration (55 %). Referring to

the gender regimes inherent in migration trajectories,

the survey indicates a balanced picture: 44 % agree

that migration is dangerous for women while 48 %

disagree with this opinion.

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

131DIE ERDE · Vol. 147 · 2/2016

4.2 Case study Semarang (Indonesia)

Environmental change in Semarang combines anthro-

pogenic and ‘natural’ factors: tidal and river flooding

(cf. Marfai et al. 2008; Anita and Latief 2013), man-

groves destruction since the 1990s, urban growth lead-

ing to suburbanisation processes (cf. Wilonoyudho 2010,

Handayani and Rudiarto 2014, Handayani and Kumalasa-

ri 2015) and extensive groundwater extraction worsen-

ing the land subsidence situation (Marfai and King 2007,

Marfai et al. 2008). In addition to these anthropogenic

influences, the topography of Semarang City with a very

low coastline and a steep mountain area (Helmi et al.

2014: 27) makes it prone to flash floods. Further, man-

agement of tidal flood has been one of the main goals

of the Central Java Provincial Government. In coopera-

tion with the Dutch government, the Banger Pilot Polder

is constructed (Mercy Corps et al. 2010). The already

difficult situation of a fragile water regime gets worse

through the increasing conversion of forestland to set-

tlements in the hilly areas so that water can no longer

drain away. 36 % of the interviewed households had

already experienced river flooding (= banjir), 17 % had

experienced tidal flood (= rob). 8 % of the interviewed

households have experienced land subsidence.

Photo 1 shows a house where the floor has been raised

by its owners in order to prevent (tidal) flooding in

the house. Many owners lift up either the whole house

or the floor depending on their financial means. 46 %

of the interviewed household had already lifted up

the floor of their house at least once, in average about

five years ago and for the majority about one to two

metres. Those who cannot afford these protection

measures either abandon the house as it is shown on

the right side of the picture, or they are obliged to live

with repeated flooding inside their house.

Figures for migration in Semarang Municipality indi-

cate a negative net migration rate for half of the 16 sub-

districts (Bappeda and BPS Kota Semarang 2012: xiv).

The net migration rate for the whole municipality, how-

ever, is positive: 38.13 (recent migration) and 167.86

(lifetime migration)6 for 2010 (Fig. 5). Despite its envi-

ronmental challenges, Semarang Municipality has the

highest positive net migration rate in Central Java (own

calculation based on BPS 2012: 21)7, a province with

high out-migration. These official migration data, how-

ever, have to be handled with care as they do not show

the high numbers of undocumented migrations. The

projected population for 2030 is 2.1 million (Handayani

and Rudiarto 2014: 80). The three neighbouring regen-

cies show negative net migration rates8 indicating rural-

urban migration movements towards Semarang City.

Our survey indicates a rather sedentary population

with more than half of the interviewed households

having lived at their current place of living for more

than one generation, and for those who moved to the

current place of living, 28 % moved internally within

Semarang City. We counted 265 out-migrants in 39 %

of the interviewed households and 129 in-migrants in

28 % of the households.

The out- and in-migrants belong overwhelmingly to the

economically active population. The majority of out-

migrants was likely to be the son or the daughter of the

interviewed household head (65 %). The in- migrants

were either son-/daughter-in-law (27 %), other relative

(23 %) or spouse (14 %) which makes us think of mar-

riage constituting one major reason for in-migration.

Photo 1 Those who can afford it lift up their house to protect it

from flooding, Tambak Lorok, Kelurahan Tanjung Mas

(Photo: Ziegelmayer, 09/08/2015)

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

132 DIE ERDE · Vol. 147 · 2/2016

The majority of the migrants was married9 and the

out-migrants’ main occupation was mainly employee

(22 %), unpaid family worker (16 %), self-employment

(12 %) and factory worker (9 %). The in-migrants were

employees (20 %), unpaid family workers (16 %) or

factory workers (14 %). Most of the out-migrations are

quite recent: 70 % moved within the last 10 years. In-

migration is also quite recent with 72 % within the last

10 years. An important aspect of migration movements

are remittances to the families back home. The survey

points at 42 % of the migrants remitting, mostly mon-

ey. The survey in Semarang indicates family as a ma-

jor reason for out-migration (55 %), followed by work

(30 %) and education (7 %), the same order of impor-

tance holds true for in-migration to Semarang.

One important aspect of migration trajectories are

narratives around migration. In Semarang the per-

ceptions of migration by the respondents are rather

ambiguous: 49 % perceived migration rather negative

and stated that “you should never leave your place of

origin”, about half of our sample see no further bene-

fits for the household in the migration of household

members. This indicates that the migration of one

household member is not a major strategy to improve

the situation of the whole household. However, 60 %

agreed that migration brings benefit to the community.

A major part of mobility in Semarang is taking place

within the municipality. 24 % of in-migrants moved from

another sub-district within Semarang to their current

place of living. The out-migrations are given in Table 2.

These results might be explained by the perception of

internal migration within Indonesia being better than

international migration – as stated by 67 % of the re-

spondents. Handayani and Kumalasari (2015) refer in

this context to a Javanese value of staying with the fam-

ily (mangan ora mangan ngumpul). In this regional set-

ting attachment to place turns out to be an important

denominator for the outline of migrant trajectories.

According to our survey Semarang hosts only a few

migrants who experienced multiple journeys. The in-

migrants mostly come from other places in Central Java

(36 %), neighbouring regencies (17 %), other islands

(11 %) or other provinces in Java (8 %). Studies on the

neighbouring regencies show cases of fishermen who

migrated to Semarang’s industrial centre looking for

work after a decline in fish catch due to environmental

hazards like “flood, coastal inundation and subsidence”

(Handayani and Kumalasari 2015: 123).

As for the organisation of the individual migration

projects, only 29 % of the in-migrants received as-

sistance for their migration, most of them by house-

hold members of the household they are now living

in or from other migrated relatives. Only 28 % of

the out-migrants had received assistance by house-

hold members or other relatives.

5. Discussion

As shown by the two case studies the interrelation of

environmental change and migration cannot be re-

garded as mono-causal. In order to understand current

migration trajectories as well as immobility in contexts

of environmental change, the integration of past migra-

tion experiences as well as the perception and expecta-

tion of former environmental changes among the local

population into the analysis are crucial.

Referring to one of the study regions, the southeastern

coast of Ghana around the Keta Lagoon, Arthur and Ar-

thur identify flooding during the rainy season as a ma-

jor source for temporal migration along the Ghanaian

coast (2011: 10). In a World Bank Report Yaro (2010)

states that some fishermen from the coastal zone have

migrated to Atakpamé in neighbouring Togo or to

Accra because fishing or farming in their home com-

munities did not secure their livelihood any longer.

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

Destination Share of all out-

migrations (%)

Other sub-districts of Semarang City

Other Indonesian islands

Other places in Central Java

Neighbouring municipalities

Big cities, esp. Jakarta

Other provinces in Java

Other countries

unknown

Total

100

Table 2 Destinations of out-migrations from Semarang

(NRF Survey 2015)

133DIE ERDE · Vol. 147 · 2/2016

This corresponds to our findings concerning regional

migration along the West African coast.

In the case of this coastline, the observable man-made

environmental change due to big development projects

or infrastructural projects with the aim to adapt to

climate change, sometimes goes along with govern-

mentally controlled population resettlement policies.

Large-scale infrastructure projects such as the Ako-

somobo and Bui dam led to resettlement programs

that did not correspond to the people’s needs. Over

half of the resettled population left the new resettle-

ment townships within four years after project imple-

mentation (Hart 1980, cited in Miescher and Tsikata

2009/2010: 25). The construction of the dam led to en-

vironmental and social decline for the communities at

the Lower Volta. Awumbila and Tsikata (2010) show in

a detailed case study on gender, land tenure and man-

grove exploitation in South Tongu District (Lower Vol-

ta), on the western border of the study area Keta, how

the dam construction severely affected the livelihoods

in the communities in the Lower Volta. It remains to be

shown how the construction of the Sea Defence which

is protecting part of the coastal populations for now,

will affect livelihoods in the surrounding areas due to

changing water streams and sedimentation and there-

by stronger erosion further east.

Regarding the Indonesian case the results of our NRF

survey suggest higher out- than in-migration for Sema-

rang, contradicting official migration data. This fact

hints at a high number of undocumented migrations,

but could also be attributed to the rather neat defini-

tion of migration as movement from at least one sub-

district (Kecamatan) to another10. In comparison with

the Ghanaian case and putting into account the vari-

ous environmental changes, Semarang – even if being

subject to frequent and increasing floodings – does not

experience strong out-migration related to environ-

mental change. In contrast, the city presents itself as

a highly dynamic urban centre that still attracts mi-

grants (workers) and where the local population tries

to adapt to the degradation. Its inhabitants raise the

floors of their houses, build small dams etc. (Marfai et

al. 2008: 237; Handayani and Kumalasari 2015); “[…]

for many houses in the coastal area in Semarang, the

houses have a particular room under the roof called as

tataban to keep their valuable goods during the flood”

(Kumalasari 2014, cited in Handayani and Kumalasari

2015). As Mulyana et al. point out, this strong in-mi-

gration into an area with various environmental chal-

lenges might lead to difficulties, especially for newly

arriving migrants. Recent migrants might be less re-

silient to climatic changes due to their limited integra-

tion in the community and therefore less social capital

to adapt to challenges due to climate change (Mulyana

et al. 2013b: 4). As Marfai et al. (2008) show in a study

on two coastal villages of Semarang City, one reason

for less out-migration than expected from coastal ar-

eas threatened by flooding might be a lack of financial

resources which does not allow to move away. This

situation hints to the existence of a large proportion of

the population that the recent literature has classified

as “trapped populations” (Black and Collyer 2014).

In both case studies the role of remittances, re-

presenting one link between local communities and

migrants, was not as significant as we expected it

to be within a highly dynamic migratory setting.

Also, the perception of what migration might bring

as benefits to the household and to the community

was challenged in our survey: While two thirds of

the Ghanaian respondents saw out-migration as a

rather positive event, only a quarter of the Indone-

sian respondents welcomed migration as a possible

strategy to improve the household’s situation. How-

ever, more than half of the Indonesian respondents

agreed that migration brings benefit to their com-

munity, a result that points to a regionally highly di-

versified perception of migration as a coping strat-

egy concerning environmental stress.

With regard to the assumptions of the theoretical

literature on migrant trajectories (cf. Spaan and Hill-

mann 2013; Vaittinen 2014; van der Velde and van

Naerssen 2007b; Schapendonk 2012), in both cases our

interviews revealed that the ‘mental distance’, here

understood as the emotional involvement attributed

to eventual mobility and migration, played a consid-

erable role in shaping migratory flows. Migration and

mobility, especially international migration, proved

to be an acknowledged part of survival options in the

Ghanaian case. In the Indonesian case people showed

to be much more attached to their place of residence

by feelings of belonging as well as belief-systems.

Here the concept of mental thresholds as put forward

by van der Velde and van Naerssen (2007b) seemed

to be at work. Additionally, immobility and “trapped

populations” stood out as a constitutional part of the

nexus between environmental change and migration.

Missing resources among the resident population

were one explanation for the reluctance of house-

holds to consider out-migration of their current (fre-

quently flooded) neighbourhood. Here, in contrast to

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

134 DIE ERDE · Vol. 147 · 2/2016

the Ghanaian situation, resettlement was neither an

option for the resident population, nor a solution that

was pushed forward by governmental programs.

6. Conclusions

The two case studies presented in this paper show

that Ghana’s and Indonesia’s coastal areas are both ex-

posed to environmental changes but that populations

concerned react quite differently. The pronounced

regional analysis, which sets the focus on the role of

migration trajectories within the changing migration

regimes can be helpful to better understand this com-

plex relationship. By using the two cases as examples

of different contexts of coastal environmental chang-

es, a tentative typology of how populations react in

terms of migration and mobility can be deducted.

Keta in Southern Ghana may work as an example of a

small town that is embedded into a longstanding mi-

gration history with its own migration saga and linked

until today to international, regional and internal mi-

gration networks. The town is still characterized by

fishing and agriculture as main livelihoods and could

not compensate the former loss of regional functions.

In this case migrant trajectories are embedded into

a culture that perceives migration as part of life and

welcomes the mobility of women and men. Due to the

long history of migration in the area today’s migrants

can often rely on contacts at the destinations facilitat-

ing the migration project. Strong international ties and

skilled migration are frequent. Additionally, former

emigrants cooperate with the local institutions such as

the municipality or the traditional chiefs and do set up

local infrastructure. In the context of environmental

change, short-distance migration within the munici-

pality seems to be one adaptation strategy.

Semarang City on the other hand represents a boom-

ing urban centre with growing industries and an ex-

panding service sector. The city attracts (labour) mi-

grants from the surrounding areas. The place shows a

rather sedentary population and some out-migration

mostly within Indonesia. International migration is an

exceptional feature. The populations that live in the

areas most affected by flooding are extremely vulner-

able and dependent on immediate incomes in the in-

formal sector or neighbouring industries. The places

in which they live have been subject to seasonal flood-

ing for a very long time. In this setting out-migration,

especially international migration, does not seem to

be an (adaptation) option but is rather perceived as a

threat by the local population.

In both regional cases, a straightforward causality

between environmental changes at the coast and out-

migration could not be identified. Migrants, however,

were able to perform as actors within changing re-

gional realities and, depending on their perceptions

and interpretations, migration became an option, not

only in reaction to environmental change. The Keta

setting represents a typology that pronounces migra-

tion trajectories as part of long-standing interregion-

al and international migration, the Semarang setting

might be classified as a rather typical modernization-

induced migration scheme, linked to rapidly growing

urbanisation, with “trapped populations” on the one

hand and in- and out-migration of (labour) migrants

on the other hand. The authors see an important la-

cuna in research in the further exploration of the en-

tanglement of man-made environmental change, ur-

banization patterns and migrant trajectories.

Notes

1 The research area is mainly Semarang City (Kota Sema-

rang) and not the neighbouring Regency with the same

name (Kabupaten Semarang) which is located south of

Semarang City in inland Central Java.

2 Vaittinen refers here to Kynsilehto’s note on skills (2011:

1548).

3 The figures refer to all household members who left the

household to live elsewhere for more than three consecu-

tive months. The NRF survey did not apply any time con-

straints concerning the year of out-migration but deliber-

ately asked for all out-migrants who have left the household.

4 N.b.: Unfortunately, there are 38 % of the migrants where

the age when leaving is missing because the interviewed

household heads did not know it.

5 The explanation of this recent peak of out-migration re-

mains one focus of the ongoing research.

6 The Indonesian census differentiates between “lifetime mi-

gration” (district of birthplace differs from district of cur-

rent residence) and “recent migration” (district of residence

five years ago differs from district of current residence).

7 The other regencies/municipalities with positive net mi-

gration rate (no. of migrants per year per 1,000 inhabit-

ants) in Central Java are Sukoharjo (15.19 recent migra-

tion), Salatiga (11.59 recent migration) and Karanganyar

(5.45 recent migration).

8 Net migration rates: Regency Semarang -294.64 (lifetime

migration), -34.93 (recent migration); Demak: -78.98 (life-

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

135DIE ERDE · Vol. 147 · 2/2016

time migration), -14.13 (recent migration); Kendal: -81.10

(lifetime migration), -5.46 (recent migration)

9 Out-migrants: 72 % married, 25 % never married, 2 % divorced;

in-migrants: 62 % married, 35 % never married, 3 % widow(er).

10 We used this definition of migration in order to include

population mobility within Semarang City which might be

very relevant in the case of environmental change. This does

not conform with the Indonesian census data that count

only movement to outside Semarang city as migration.

Acknowledgments

This article is based on the research results within the

project “New Regional Formations: rapid environmental

change and migration in coastal areas in Ghana and Indone-

sia”, funded by Volkswagen Stiftung. The project consorti-

um was coordinated by the University of Bremen (lead) and

the Leibniz Institute for Research on Society and Space in

Erkner near Berlin. The overall project is divided into five

subprojects that were run by the ZMT (Centre for Marine

Tropical Ecology, Bremen), the Jacobs University (Bremen)

and the KWI in Essen in the years 2014-2016. The results

presented here were gathered within subproject 3, titled:

Migrant trajectories. The authors want to express their

gratitude for the generous funding by Volkswagen Stiftung.

References

Abu, M. 2011: Migration as an adaptation strategy to climate

change: evidence from Buoku and Bofie-Banda in the

Wenchi and Tain Districts of Ghana. – Paper for the Sixth

African Population Conference, held in Ouagadougou, Bur-

kina Faso, 5-9 December 2011. – Accessed online: http://

uaps2011.princeton.edu/papers/110974, 7/16/2014

Abu, M., S.N.A. Codjoe and J. Sward 2013: Climate change

and internal migration intentions in the forest-savannah

transition zone of Ghana. – Population and Environment

35 (4): 341-364

Afifi, T. and J. Jäger (eds.) 2010: Environment, forced migra-

tion and social vulnerability. – Berlin et al.

Akyeampong, E.K. 2001: Between the sea and the lagoon:

an eco-social history of the Anlo of southeastern Ghana,

c. 1850 to recent times. – Athens, Ohio et al.

Angnuureng, B.D., K. Appeaning Addo and G. Wiafe 2013:

Impact of sea defense structures on downdrift coasts:

the case of Keta in Ghana. – Academia Journal of Envi-

ronmental Sciences 1 (6): 104-121

Anita, J. and H. Latief 2013: Coastal flooding adaptation by

housing adjustment in coastal settlements. Case stud-

ies: Muara Angke, North Jakarta and Tambak Lorok,

Semarang. – Conference paper; ITB, Bandung. – Ban-

dung, Indonesia

Appeaning Addo, K. 2014: The threat of sea level rise on

delta coastal communities in Ghana: engineering or mi-

gration? – Department of Marine and Fisheries Sciences,

University of Ghana. – Accra

Arthur, J.L. and I.A.Y. Arthur 2011: Movement under environ-

mental disasters: the case of flooding and bushfires for

selected periods in Ghana. Paper presented at the ESF-

UniBi-ZiF research conference on ‘Environmental Change

and Migration: From Vulnerabilities to Capabilities’,

Bad Salzuflen, Germany, December 5-9, 2010. – Center

on Migration, Citizenship and Development (COMCAD),

Working papers 97. – Bielefeld. – http://www.ecosia.

org/url?url=http%253A%2F%2Fwww.uni-bielefeld.

de%2Ftdrc%2Fag_comcad%2Fdownloads%2Fworking

paper_97_arthur_arthur.pdf&v=0&i=5&q=Ghana +envir

onmental+change+mobility&p=2, 4/17/2014

Aufenvenne, P. und C. Felgentreff 2013: Umweltmigranten

und Klimaflüchtlinge – zweifelhafte Kategorien in der

aktuellen Debatte. – In: Felgentreff, C. und M. Geiger

(Hrsg.): Migration und Umwelt. IMIS-Beiträge 44. –

Osnabrück: 19-44

Awumbila, M. and D. Tsikata 2010: Economic liberalisation,

changing resource tenures and gendered livelihoods: a

study of small-scale gold mining and mangrove exploita-

tion in rural Ghana. – In: Tsikata, D. and P. Golah (eds.):

Land tenure, gender, and globalisation: research and

analysis from Africa, Asia, and Latin America. – New

Delhi, Ottawa: 98-144

Awumbila, M., T. Manuh, P. Quartey, T.A. Bosiakoh and C. Addo-

quaye Tagoe 2011: Migration and mobility in Ghana:

trends, issues, and emerging research gaps. – Accra

Badan Pusat Statistik (BPS) 2010: Population census

data – population by age group and sex, Semarang Mu-

nicipality. – http://sp2010.bps.go.id/index.php/site/

tabel?tid=336&wid=3374000000, 09/06/2016

Badan Pusat Statistik (BPS) 2012: Statistics of migration,

Jawa Tengah. Results of the 2010 population census. –

Jakarta

Bappeda Kota Semarang and Badan Pusat Statistik (BPS)

Kota Semarang 2012: Profil Kpendudukan Kota Sema-

rang [Demographic profile of Semarang]. – Semarang

Black, R. 2001: Environmental refugees: myth or reality? –

New Issues in Refugee Research Working Paper 34. – Geneva

Black, R. and M. Collyer 2014: Populations ‘trapped’ at times

of crisis. – Forced Migration Review 45: 52-56. – Accessed

online: http://www.fmreview.org/crisis/black-collyer,

8/7/2015

Black, R., S.R.G. Bennett, S.M. Thomas and J.R. Beddington

2011: Climate change: migration as adaptation. – Nature

478: 447-449

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

136 DIE ERDE · Vol. 147 · 2/2016

Ghana Statistical Service (GSS) 2012: 2010 Population &

housing census. Summary report of final results. – Accra,

Ghana

Ghana Statistical Service (GSS) 2013: 2010 Population &

housing census. National analytical report. – Accra,

Ghana

Graham, P. (ed.) 2008: Horizons of home: nation, gender and

migrancy in island Southeast Asia. – Annual Indonesia

lecture series 25. – Clayton

Grillo, R. 2007: Betwixt and between: trajectories and pro-

jects of transmigration. – Journal of Ethnic and Migra-

tion Studies 33 (2): 199-217

Handayani, W. and N.R. Kumalasari 2015: Migration as

future adaptive capacity: the case of Java – Indone-

sia. – In: Hillmann, F., M. Pahl, B. Rafflenbeul and H. Sterly

(eds.): Environmental change, adaptation and migration:

bringing in the region. – Basingstoke: 117-138

Handayani, W. and I. Rudiarto 2014: Dynamics of urban

growth in Semarang Metropolitan – Central Java: an

examination based on built-up area and population

change. – Journal of Geography and Geology 6 (4): 80-87

Hebinck, P., S. Slootweg and L. Smith (eds.) 2008: Tales of

development: people, power and space. – Assen

Helmi, M., R.P. Irwani and D.N. dan Sunaryo 2014: Illustra-

tion of hazard and issues on Semarang Coastal Area. –

Paper for the International Workshop Urbanization and

Marine Sciences; How to Develop Resilience Processes to

Cope with Natural Hazard” in Semarang

Hillmann, F. 2016: Migration. Eine Einführung aus sozialgeo-

graphischer Perspektive. – Sozialgeographie kompakt

4. – Stuttgart

Hillmann, F., M. Pahl, B. Rafflenbeul and H. Sterly (eds.) 2015a:

Environmental change, adaptation and migration: bring-

ing in the region. – Basingstoke

Hillmann, F., M. Pahl, B. Rafflenbeul and H. Sterly 2015b: In-

troduction: (Re-)locating the nexus of migration, envi-

ronmental change and adaptation. – In: Hillmann, F., M.

Pahl, B. Rafflenbeul and H. Sterly (eds.): Environmental

change, adaptation and migration: bringing in the re-

gion. – Basingstoke: 1-17

Hugo, G. 2008a: International migration in Indonesia and

its impacts on regional development. – In: van Naerssen,

A.L., E. Spaan and E.B. Zoomers (eds.): Global migration

and development. – Routledge Studies in Development

and Society 15. – New York et al.: 43-65

Hugo, G. 2008b: Migration in Indonesia: recent trends and

implications. – In: Graham, P. (ed.): Horizons of home:

nation, gender and migrancy in island Southeast Asia. –

Annual Indonesia lecture series 25. – Clayton: 45-70

Hugo, G. (ed.) 2013: Migration and climate change. – The

International Library of Studies on Migration 15. –

Cheltenham et al.

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

Boatemaa, M.A., K. Appeaning Addo and A. Mensah 2013: Im-

pacts of shoreline morphological change and sea level

rise on mangroves: the case of the Keta coastal zone. –

Journal of Environmental Research and Management 4

(11): 359-367. – Accessed online: http://www.e3jour-

nals.org/cms/articles/1387371004_Kwasi%20et%20al.

pdf, 09/06/2016

Bohra-Mishra, P., M. Oppenheimer and S.M. Hsiang 2014:

Nonlinear permanent migration response to climatic

variations but minimal response to disasters. Support-

ing information. – Proceedings of the National Academy

of Sciences of the United States of America 111 (27):

1-7. – Accessed online: http://www.pnas.org/content/

suppl/2014/06/19/1317166111.DCSupplemental/

pnas.201317166SI.pdf, 09/06/2016

Brown, S. and P. Brownlee 2000: Introduction. – In: Sukamdi,

A. Haris and P. Brownlee (eds.): Labour migration in In-

donesia: policies and practice. – Yogyakarta. – Accessed

online: http://www.unesco.org/most/apmrlabo.htm,

14/06/2016

Carr E.R. 2005: Placing the environment in migration: en-

vironment, economy, and power in Ghana’s Central Re-

gion. – Environment and Planning A 37 (5): 925-946

Castagnone, E. 2011: Transit migration: a piece of the com-

plex mobility puzzle. The case of Senegalese migra-

tion. – Cahiers de l’Urmis, Unité Recherche Migrations

et Société 13. – Accessed online: http://urmis.revues.

org/927, 14/06/2016

Centre for Migration Studies CMS, University of Ghana 2011:

Dynamics of migration in West Africa. – Migration factsheet

series 1 (1). – Accesssed online: www.cms.ug.edu.gh/cms-

docs/downloads/cms_factsheet.pdf, 3/17/2014

Collyer, M. 2010: Stranded migrants and the fragmented

journey. – Journal of Refugee Studies 23 (3): 273-293

Fearnside, P.M. 1997: Transmigration in Indonesia. Lessons

from its environmental and social impacts. – Environ-

mental Management 21 (4): 553-570

Firman, T. 2004: Demographic and spatial patterns of In-

donesia’s recent urbanisation. – Population, Space and

Place 10 (6): 421-434

Foresight (2011): Migration and global environmental

change: future challenges and opportunities. Final project

report. – The Governance Office for Science. – London

Gammeltoft-Hansen, T. and N. Nyberg Sørensen (eds.) 2013:

The migration industry and the commercialization of

international migration. – Routledge Global Institutions

69. – London et al.

Gemenne, F. (ed.) 2010: Migrations et environnement. Spe-

cial issue. – Hommes et Migrations (1284)

Gemenne, F., P. Brücker and D. Ionesco (eds.) 2014: The state

of environmental migration 2014: a review of 2013. –

IOM. – Geneva

137DIE ERDE · Vol. 147 · 2/2016

Environmental change and migration in coastal regions: examples from Ghana and Indonesia

IOM 2008: Migration in Indonesia. Facts and figures. – Jakarta

IOM 2010: Labour migration from Indonesia: an overview

of Indonesian migration to selected destinations in Asia

and the Middle East. – Jakarta

IPCC 2013: Summary for policymakers. – In: The Inter-

governmental Panel on Climate Change (IPCC) (ed.): Cli-

mate change 2013: the physical science basis. Contribu-

tion of Working Group I to the Fifth Assessment Report

of the Intergovernmental Panel on Climate Change. –

Cambridge, UK, New York, USA

Keta Municipality 2010: Keta Municipality profile. – Keta

Klein, N. 2014: This changes everything: capitalism vs. the

climate. – New York et al.

Kleist, N. 2011: Modern chiefs: tradition, development and

return among traditional authorities in Ghana. – African

Affairs 110 (441): 629-647

Kumassah, A. 2009: The migration saga of the Anlo-Ewes of

Ghana. – 2009 edition. – Keta, Ghana

Kwankye, S.O., J.K. Anarfi, C.A. Tagoe and A. Castaldo 2009:

Independent north-south child migration in Ghana: the

decision making process. – Development Research Cen-

tre on Migration, Globalisation and Poverty, University

of Sussex, Working Paper T 29. – Brighton

Kynsilehto, A. 2011: Negotiating intersectionality in highly

educated migrant Maghrebi women’s life stories. – Envi-

ronment and Planning A 43 (7): 1547-1561

Manuh, T. 2001: Ghanaian Migrants in Toronto, Canada: care

of kin and gender relations. – Research Review NS 17 (2):

17-26. – Accessed online: http://archive.lib.msu.edu/

DMC/African%20Journals/pdfs/Institue%20of%20Af-

rican%20Studies%20Research%20Review/2001v17n2/

asrv017002003.pdf, 13/06/2016

Marfai, M.A. and L. King 2007: Monitoring land subsidence

in Semarang, Indonesia. – Environmental Geology 53

(3): 651-659

Marfai, M.A. and L. King 2008: Coastal flood management in Sema-

rang, Indonesia. – Environmental Geology 55 (7): 1507-1518

Marfai, M.A., L. King, J. Sartohadi, S. Sudrajat, S.R. Budiani and

F. Yulianto 2008: The impact of tidal flooding on a coastal

community in Semarang, Indonesia. – The Environmen-

talist 28 (3): 237-248

Martin, S.F. 2012: Environmental change and migration:

legal and political frameworks. – Environment and Plan-

ning C: Government and Policy 30 (6): 1045-1060

Mercy Corps, CCROM Southeast Asia and Pacific and URDI

2010: Vulnerability and adaptation to climate change:

a community-based vulnerability assessment. – Urban

Bulletin Series 9. – Jakarta

Miescher, S.F. and D. Tsikata 2009/2010: Hydro-power and

the promise of modernity and development in Ghana:

comparing the Akosombo and Bui Dam projects. – Ghana

Studies 12/13: 15-53

Mulyana, W., I. Setiono, A. Kracker Selzer, S. Zhang, D. Dodman

and D. Schensul 2013a: Urbanisation, demographics and

adaptation to climate change in Semarang, Indonesia. –

Urbanisation and Emerging Population Issues Working

Paper 11. – London

Mulyana, W., D. Dodman, S. Zhang and D. Schensul 2013b:

Climate vulnerability and adaptation in the Semarang

Metropolitan Area: a spatial and demographic analysis.

Technical briefing. – London

National Aeronautics and Space Administration (NASA) 2015:

NASA, NOAA find 2014 warmest year in modern record. –

Release 15-010, January 16, 2015. – Accessed online: http://

www.nasa.gov/press/2015/january/nasa-determines-

2014-warmest-year-in-modern-record, 23/05/2016

Odotei, I.K. 2002a: Sea power, money power. Ghanaian migrant

fishermen and women in the Republic of Benin. – Institute

of African Studies, University of Ghana. – Legon, Ghana

Odotei, I.K. 2002b: There is money in the sea. Ghanaian

migrant fishermen and women in the Ivory Coast. –

Legon, Ghana

Peil, M. 1995: Ghanaians abroad. – African Affairs 94 (376):

345-368

Piguet, E., A. Pécoud and P. de Guchteneire (eds.) 2011: Migra-

tion and climate change. – Cambridge et al.

Platt, M., B.S.A. Yeoh, G. Baey, C.Y. Khoo, T. Lam, D. Das and

M. Ee 2013: Financing migration, generating remit-

tances and the building of livelihood strategies: a case

study of Indonesian migrant women as domestic work-

ers in Singapore. – Migrating out of Poverty, University

of Sussex, Working Paper 10. – Brighton. – Accessed

online: http://migratingoutofpoverty.dfid.gov.uk/files/

file.php?name=wp10-platt-et-al-final.pdf&site=354,

13/06/2016

Purwaka, T.H. and Sunoto 1999: Coastal and marine re-

sources management in Indonesia: legal and institu-

tional aspects. – International Center for Living Aquatic

Resources Management, Working paper 02. – Manila. –

Accessed online: http://pubs.iclarm.net/Pubs/institu-

tional_sea/pub_insea4.pdf, 13/06/2016

Sassen, S. 2014: Expulsions: brutality and complexity in the

global economy. – Cambridge, Mass.

Schapendonk, J. 2010: Staying put in moving sands. The

stepwise migration process of sub-Saharan African mi-

grants heading north. – In: Engel, U. and P. Nugent (eds.):

Respacing Africa. – African-Europe Group for Interdisci-

plinary Studies 4. – Leiden, Boston: 113-139

Schapendonk, J. 2012: Turbulent trajectories: African mi-

grants on their way to the European Union. – Societies

2 (2): 27-41

Spaan, E. 1999: Labour circulation and socioeconomic

transformation: the case of East Java, Indonesia. – PhD

thesis, Rijksuniversiteit Groningen

138 DIE ERDE · Vol. 147 · 2/2016

Spaan, E. and F. Hillmann 2013: Migration trajectories and

the migration industry. Theoretical reflections and em-

pirical examples from Asia. – In: Gammeltoft-Hansen, T.

and N. Nyberg Sørensen (eds.) 2013: The migration in-

dustry and the commercialization of international mi-

gration. – Routledge Global Institutions 69. – London et

al.: 64-86

Sukamdi, A. Haris and P. Brownlee (eds.) 2000: Labour mi-

gration in Indonesia: policies and practice. – Yogyakarta

Tacoli, C. 2011: Not only climate change: mobility, vulner-

ability and socio-economic transformations in environ-

mentally fragile areas of Bolivia, Senegal and Tanzania. –

Human Settlements Working Paper Series, Rural-Urban

Interactions and Livelihood Strategies 28. – London

Tankwanchi, A.B.S. 2012: Doctors beyond borders: data

trends and medical migration dynamics from sub-

Saharan Africa to the United States. – Dissertation,

Vanderbilt University, Nashville, Tennessee

Teye, J.K., M.B. Setrana and A.A. Acheampong 2014: Migra-

tion of health professionals from Ghana: trends, drivers

and emerging issues. – In: Roscoe, C. (ed.): Ghana: social,

economic and political issues. – New York: 253-274

The World Bank 2010: Mainstreaming climate change for

sustainability. – Jakarta. – Accessed online: http://www-

wds.worldbank.org/external/default/WDSContent-

Server/WDSP/IB/2010/03/30/000333038_201003300

05825/Rendered/PDF/534750BRI0710c10Box345611B

01PUBLIC1.pdf, 13/06/2016

The World Bank 2011: Migration and remittances factbook

2011. – 2nd edition. – Washington, DC. – Accessed online:

http://siteresources.worldbank.org/INTLAC/Resourc-

es/Factbook2011-Ebook.pdf, 13/06/2016

The World Bank 2014: Remittances to developing coun-

tries to stay robust this year, despite increased deporta-

tions of migrant workers, says WB. – World Bank press

release, April 11, 2014. – Washington, DC. – Accessed

online: http://www.worldbank.org/en/news/press-re-

lease/2014/04/11/remittances-developing-countries-

deportations-migrant-workers-wb, 13/06/2016

UN Habitat 2010: Colored water. Assessment of climate

change vulnerability in Kelurahan Pabean Pekalongan,

Central Java. – Jakarta

United Nations, Department of Economic and Social Affairs,

Population Division 2013: Trends in international mi-

grant stock: the 2013 revision. Migrants by destination

and origin. – United Nations database, POP/DB/MIG/

Stock/Rev.2013. – New York

Urbano, M., M. Deasey, N. Kusumawardani and J. Pardosi

2011: Children and climate change. The impacts of cli-

mate change on nutrition and migration affecting chil-

dren in Indonesia. Report commissioned by UNICEF. –

UNICEF East Asia and Pacific Regional Office. – Bangkok

Vaittinen, T. 2014: Reading global care chains as migrant

trajectories: a theoretical framework for the under-

standing of structural change. – Women’s Studies Inter-

national Forum 47 B: 191-202

Van der Geest, K. 2009: Ghana case study report: Migration

and natural resources scarcity in Ghana. – Amsterdam. –

Accessed online: http://dare.uva.nl/document/2/62270,

13/06/2016

Van der Geest, K. 2010: Local perceptions of migration from

north-west Ghana. – Africa 80 (4): 595-619

Van der Geest, K. 2011a: North-south migration in Ghana:

what role for the environment? – International Migra-

tion 49, Supplement 1: e69-e94

Van der Geest, K. 2011b: The Dagara farmer at home and

away. Migration, environment and development in

Ghana. – PhD thesis, University of Amsterdam. – African

Studies Collection 33. – Leiden

Van der Velde, M. 2008: People, borders, and trajectories. –

In: Hebinck, P., S. Slootweg and L. Smith (eds.): Tales of

development: people, power and space. – Assen: 111-122

Van der Velde, M. and T. van Naerssen (eds.) 2007a: Migration

in a new Europe: people, borders and trajectories. – IGU

Home of Geography Publication Series 8. – Rome

Van der Velde, M. and T. van Naerssen 2007b: People, borders

and trajectories: a model to approach migration in the

enlarged European Union. – In: van der Velde, M. and T.

van Naerssen (eds.): Migration in a new Europe: people,

borders and trajectories. – IGU Home of Geography Pub-

lication Series 8. – Rome: 145-153

Van der Velde, M. and T. van Naerssen (eds.) 2015: Mobility

and migration choices: thresholds to crossing borders. –

Farnham et al.

Wilonoyudho, S. 2010: The urbanization in Semarang City. –

Indonesian Journal of Geography 42 (2): 181-194

Yaro, J.A. 2010: The social dimensions of adaptation to

climate change in Ghana. – Economics of adaptation to

climate change. – The World Bank. Discussion Paper

15. – Washington, D.C.

Environmental change and migration in coastal regions: examples from Ghana and Indonesia