Modeling migration flows in the Mekong River Delta region of Vietnam: an augment gravity approach

Modeling migration flows in the Mekong River Delta region

of Vietnam: an augment gravity approach

Prepared by Dr. Huynh Truong HUY Prof.dr. Walter NONNEMAN

Faculty of Applied EconomicsDepartment of Economics

June 28th, 2013

2

1. Basic facts about Vietnam and the MRD region2. Research question and “Gravity” approach3. Data and statistical description4. Empirical findings and forecast5. Concluding

Outline of the presentation

3

Basic facts about Vietnam

• Vietnam is an emerging economy with a population 86 million people• Economic policy since the mid-1990s changed the structure of GDP

shifting agriculture towards industry and services• These changes are accompanied with important migration flows

* Relative rate (per 1000 population) is shown in the parentheses

Migration (million) 1989 1999 2009

-Vietnam 1.3 (25) 2.0 (29) 3.4 (43)

-MRD 0.32 (23) 0.63 (39) 1.15 (67)

Vietnam’s GDP structure (%) 1990 2000 2010

-Agriculture 38.7 24.5 20.6

-Industry 22.7 36.7 41.1

-Services 38.6 38.8 38.3

4

Basic facts about the Mekong River Delta region

• The MRD region is one of the 6 regions in Vietnam and has 13 provinces/cities

• MRD has a population of 17 million • The region is predominantly agricultural

(40% of GDP; 50% of the labor force) and rural (75% of the population in rural areas)

• The MRD is the major exporting region of Vietnam of rice & pangasius fish

• The region’s economy rapidly shifted from agriculture to industry and services (e.g. the share of agriculture dropped from 51% of GDP in 2000 to 40% in 2010).

5

Internal migration in Vietnam and in the MRD

• Internal migration in Vietnam is important and increasing– Emigration rate between regions of Vietnam in 1999 was 29 per 1000

population and increased to 43 per 1000 in 2009• Internal migration flows from the MRD to other regions are

large and increasing– Emigration rate from the MRD in 1989 was 14 per 1000 population; in

1999 it increased to 25 per 1000 and to 52 per 1000 in 2009.– About 75% of MRD migrants moved out of the region, mainly to Ho

Chi Minh city (45%), Binh Duong (21%).

6

Research question and application of Gravity model

• How well do models – such as “gravity” models – explain internal migration flows between MRD provinces?

• Earlier works concerning Gravity models (Ravenstein, 1885; Zipf, 1946; Fields, 1979; Hunt & Greenwood, 1985)

taking into account of economic theories to explain migration flows between MRD provinces and 3 major cities (Ho Chi Minh, Binh Duong, Ha Noi) and the rest by an enhanced gravity model.

7

Different gravity models

• A simple form:

• A modified form (Zipf, 1946)

• A modified form with relative income (Fields, 1979; Todaro, 1969)

• A more generally modified form (Hunt & Greenwood, 1985)

i jij

ij

y yM k

d

i jij

ij

PPM k

d

0 1 2 3ln ln ln( ) ln( )ij ij i j i j ijM a a D a PP a Y Y

ijm

mjmn

ninjiijij XXPPDM lnlnlnlnlnln 3210

8

Theories of migration & hypotheses

• The classic urban-rural migration model (Harris & Todaro 1970) stresses differences in expected labor earnings

• Migration investment requires sufficient funds to overcome the cost of migration (the liquidity trap hypothesis) (Lucas 1997)

• The theory “relative deprivation” (Stark 1991) stresses relative income as driving force

• Some theories are “testable” with an enhanced gravity model

9

Data and statistical description

• Migration flows (dependent) from the Vietnam population census, 2009.

• Some variables: distance (km), relative average income, population, urbanization, poverty rate, Gini coefficient, unemployment.

Note: M ij is the number of migrants from i province to j province (in the period 2004-09)

pii and pij are the share of non-migrant and migrant in each province

- Since the share of non-migrant shows little variation, so we can use the variable: the number of migrants (Mij) as dependent.

Variable Mean Std.dev. Min MaxMij 8973.2 45016.2 4.000 567049

pii 0.997 0.066 0.955 0.999

pij 0.003 0.066 0.000 0.045

Some variables used in the analysis

Variable Mean Std.dev. Min MaxDij Distance source-destination (km) 337.7 563.0 13.7 2070.0

Yj/Yi Relative average income destination/source

1.086 0.466 0.361 2.850

EYj/EYi Relative expected income destination/source

1.085 0.464 0.352 2.838

POPi Population (in 1000 units) source (destination)

4925 12406 754 54105

URBi Share of urban population (%) 27.49 18.70 9.57 82.57

POVi Poverty rate (%) 11.12 5.69 0.40 21.45

GINIi Gini coefficient 0.485 0.058 0.317 0.572

UNEMPi Unemployment rate (%) 4.289 0.390 3.763 5.004

10

Empirical analysis & major findings (1)

11

05

1015

ln(M

ij)

12 14 16 18 20ln(POPi*POPj) 2004

Fitted values ln(Mij)

05

10

15

ln(M

ij)2 4 6 8

ln(DIS)

Fitted values ln(Mij)

Population size induces migration

Distance reduces migration due to costs

• Bivariate results

Empirical analysis & major findings (2)

• However, some others are expressed in weak or no significant relations with migration, including urbanization, Gini, poverty rate.

12

05

10

15

ln(M

ij)

-1 -.5 0 .5 1ln(Expected Income j/i

Fitted values ln(Mij)

Expected income gap drives people in migration

Empirical analysis & major findings (3)

Multivariate results with the application of Gravity models • Basic model: Key determinants of are

– Distance (-)– Population size (+)– Income gap between destination and origin (+)– Expected income gap two areas (+)

• Augmented model: number of migrants relates – Distance (-), Population size (+), Income gap (+)– Costs (i.e. distance) restrict poor people in migration (-)

In sum, these results confirm the earlier findings in the bi-variate analysis.

13

Empirical analysis & major findings (4)

Forecasting the migration flow towards 2014• The forecast done is based on the estimated coefficients from

augmented model regarding some main variables– Distance– Distance combined with poverty rate (costs)– Population size – Income disparity between destination and source

• The forecast shows some figures: – The migration flows in this region would reach at 3 million in the period 2009-

2014, an increase by 0,5 million (or 22%) in comparison with the earlier 5-year period.

– Binh Duong will be the most attractive destination for migrants, an alternative for Ho Chi Minh due to policies relating migrant controls, economic development, etc.

14

15

Concluding

• This study has contributed to improve better understanding of internal migration in Vietnam, particularly in the MRD region.

• This study has explored this topic using the population census data in the period 2004-2009 to describe the pattern of migration flows in the MRD region.

• Application of economic theories and empirical models of migration (Gravity approach) contributes to explain the migration flows between the region and the rest of country.

• The forecasted result also indicates a fact that an increasing pattern of migration flows in the MRD region would continuously take place in near future.

Thank You Very Much For Your Kind Attention!