TheEffectofLandInheritance

onYouth

Employm

ent

andMigrationDecisions:

Evidence

from

RuralEthiopia

Presentationby:

HosaenaGhebru

(Ph.D

)Co-authors:

KatrinaKosec,Brian

Holtem

eyer,Valerie

Mueller,andEmily

Schmidt

InternationalFoodPolicy

ResearchInstitute

(IFPRI)

ESSP/IFPRI-EDRIResearchSem

inar

AddisAbab

a,Ethiopia

November

3,2016

ResearchSem

inar

(IFPRI)

November

3,2016

1/26

Introduction

ResearchQuestion

How

doestheamountoflandyo

uth

exp

ect

toinheritaffect

their

migrationandemploym

entdecisions?

Whatis

thedifferentialim

pact

ofexp

ectedinhertance

onyo

uth:

withvariou

sageandgender

grou

ps?

withversuswithou

taccess

tofunctioninglandrentalmarkets?

vspace0.05in

based

ontheirresidential

proxim

ityto

major

urban

centers?

Ove

rall,yo

uth

(aged15–34)arelikely

tomake

decisionsbasedon

inheritance

exp

ectations,

butweknow

little

abouthow

inheritance

impactsdecisions

Kosecet

al.

(2016)

(IFPRI)

November

3,2016

2/26

Introduction

Focus:

RuralEthiopia

InEthiopia,landisform

ally

owned

bythego

vernment,withform

allandmarkets

(sales)ou

tlaw

edandmajor

landaccess

via

administrativeredistribution

Dueto

it’sadversetenure

security

effect,redistribution

has

longbeen

abandon

ed

Thecountrynow

facingacute

prob

lemsof

landscarcity

-especially

inthehighlands

Hou

seholdfarm

sizesaredwindling;

In20

11–1

2,morethan

halfof

ruralhou

seholdscultivated

less

than

onehectare

ofland

You

threly

onperiodic

landredistribution

s,inheritance,and/or

small

rentalmarkets

toaccess

land

Kosecet

al.

(2016)

(IFPRI)

November

3,2016

3/26

Introduction

Preview

oftheResults

Approach:

Use

panel

datafrom

from

2010

and20

14from

1,74

8hou

seholdsin

27

woredas

(districts)of

ruralAmharaandOromia

region

sof

Ethiopia

Focuson

youth

aged

15–3

4;estimateahou

seholdfixedeff

ects

model

Exploitexog

enou

svariationin

thetimingof

landredistributionsto

overcomeendog

enou

shou

seholddecisionsab

outinheritance

size

Findings:

Largerexpectedlandinheritances:

Low

erthelikelihoodoflong-distance

andrural-to-urban

permanent

migrationduring20

10–1

4Increase

thelikelihoodof

employm

entin

agriculture

in20

14Reduce

thelikelihoodof

employm

entin

thenon

-agriculturalsector

Donot

affectthedecisionto

study

Resultsappearto

bedrivenby

males

andolder

youth

Inheritance

plays

amorepron

ouncedrole

inareaswithless-vibrant

landmarkets

andin

moreremoteareas

Kosecet

al.

(2016)

(IFPRI)

November

3,2016

4/26

Background

Accessto

Landin

Ethiopia

Landform

ally

owned

bygo

vernment,withform

allandmarkets

(sales)

outlaw

ed

Individual

landusers

havethelegalrigh

tto

transfer

theirlanduse

righ

tsto

theirchildrenor

other

family

mem

bers

Individualscanalso

renttheirlanduse

righ

tsto

anyperson

Several

landredistribution

shaveoccurred

since

1991

,usuallybased

on

hou

seholdsize

atthetime(w

ithextraweigh

tplacedon

adultmales)

Som

etim

esinvolved

landpreviouslyutilized

bystatefarm

s

Inou

rstudyregion

sofAmharaandOromia,20

ofou

r27

sample

woredas

experiencedsuch

alandredistribution

,thou

ghthereis

substantial

variationin

thetimingof

theredistribution

s

Kosecet

al.

(2016)

(IFPRI)

November

3,2016

5/26

Background

YearofMost

RecentLandRedistribution

Year

Shareof

observations(N

=1,98

9)

1992

3.4

1993

9.5

1997

21.8

2003

6.5

2004

4.4

2005

9.5

2006

9.2

2010

4.0

2012

1.7

2013

3.6

non

e26

.6

Notes:

Anobservationsis

anindividual.

Med

ianyear

is2003.

Source:Authors’calculationsbasedonIFPRI’sWatershed

Survey

of2014.

Kosecet

al.

(2016)

(IFPRI)

November

3,2016

6/26

Background

Norm

sofInheritance

Landinheritancesin

Ethiopia

arenot

uniform

across

descendants

of

thehead—

bothdueto

culturalfactorslikenormsassociated

with

gender

andbirth

order

andrestrictionson

landfragmentation

Custom

arynormsfavormen

(Fafcham

psandQuisumbing,

2005):

Son

s(especially

firstborn)traditionally

care

forparents

inoldage

(Kumar

andQuisumbing,

201

2)Custom

arybeliefs

limitthetypeof

agriculturallabor

inwhichfemales

canengage,

necessitatingmalelabor

participationon

anyplot

Siblings

(co-descendants)matteras

well:

Agroom

’snumber

ofbrothers(butnot

sisters)

negativelyaff

ects

land

inheritance

(Fafcham

psandQuisumbing200

5)Havingmoreolder

brothersdim

inishes

agriculturalproductivity(inherit

less

productiveland)(G

ibsonandGurm

u20

11)

You

nger

siblings

also

affectyouth

decisionsto

workou

tsideof

agriculture

(GibsonandGurm

u20

12)

Kosecet

al.

(2016)

(IFPRI)

November

3,2016

7/26

Background

MigrationandEmploym

entin

RuralEthiopia

Migrants

arepredom

inantly‘pushed’from

theirhom

esrather

than

attractedby

anurban

‘pull’

ofhigher

returnson

human

capital

investments

(World

Bank,

2010

;Zelekeet

al.,20

08;Doroshet

al.,

2012

;deBrauw

,20

14;Lee

andMueller,20

16)

Bothem

ploym

entdiversification

(via

non

-agriculturalem

ploym

ent

and/ormigration

)andthepursuitof

education

alop

portunities

representpotential

strategies

ofrisk

diversification

Avarietyof

factorsmay

lead

youth

topursuethesestrategies:

Lackof

sufficientmeansof

subsistence

(poverty)

Agriculturalshocks(e.g.,drough

tsor

floods)

Decliningagriculturalproductivity

Shortage

ofland/siblingcompetitionforinheritable

assets

like

agriculturalplots

Shortage

ofem

ploym

entop

portunitiesin

ruralareas

Kosecet

al.

(2016)

(IFPRI)

November

3,2016

8/26

EmpiricalStrategyandData

Data

Panel

survey

conducted

in20

10(rou

nd1)

and20

14(rou

nd2)

Sam

ple

includes

27kebeles

in9woredas

in2region

s,with

approxim

ately20

0hou

seholdsper

woreda

1,81

0hou

seholdsin

round1and1,74

8in

round2(3.4

percent

hou

seholdattritionover

4years)

Weconstruct

across-sectionaldataset

consistingof

allround1

hou

seholdmem

bersaged

15–3

4;wethen

observetheirspatial

location

andem

ploym

entdecisionsin

2014

Rou

nd2im

portantlycollectsthisinform

ationnot

only

forhousehold

mem

bers,butalldirectdescendants

oftheheador

theirspouse

Permitsusto

analyzeexpectedinheritancesas

wellas

migrationand

employm

entdecisionsmadebetween20

10and20

14of

alldescendants

Kosecet

al.

(2016)

(IFPRI)

November

3,2016

9/26

EmpiricalStrategyandData

VariableMeasurement:

Outcomes

You

ths’expectedlandinheritance

=am

ountheadreports

that

the

youth

alreadyreceived

(from

family)+

amou

ntheadreports

planning

togive

inthefuture

Threeindicator

variablesformigration

:

Permanentmigrant:

Anyyouth

whowas

ahou

seholdmem

ber

in20

10butnot

2014

Lon

gdistance

permanentmigrant:

Permanentmigrantwhoby

2014

lives

outsideof

theworeda(district)

inwhichthey

resided

in20

10Rural-to-urban

permanentmigrant:

Permanentmigrantwhoby

2014

lives

inan

urban

area

Employm

entou

tcom

es(con

sidersindividual’sprim

aryoccupation):

Agriculture

Non-agriculturalsector

Studying

Kosecet

al.

(2016)

(IFPRI)

November

3,2016

10/26

EmpiricalStrategyandData

Descriptive

Statistics

Mean

S.D

.Dummy-permanen

tmigrant

0.45

0.5

Dummy-permanen

tmigrantoutofworeda

0.21

0.41

Dummy-permanen

tmigrantto

urbanarea

0.28

0.45

Primaryoccupationis

in:

Agricu

lture

0.37

0.48

Non-agricu

lture

0.16

0.36

Studen

t0.29

0.45

Domestic

0.13

0.34

Unem

ployed

0.04

0.19

Landinheritance

(hectares)

0.48

2.54

Loglandinheritance

-1.36

0.92

Dummy-male

0.67

0.47

Age

19.9

4.09

Dummy-ch

ildofhead

0.98

0.13

Dummy-married

0.05

0.23

Dummy-co

mpletedcycle1ofprimaryschool(grade4)

0.68

0.47

Dummy->

1male

descendantim

med

iately

followsin

birth

order

0.25

0.43

Shareofmales18+

attimeoflandredistribution

0.17

0.30

Source:Authors’calculationsbasedonIFPRI’sWatershed

Surveysof2010and2014.

Notes:

N=1,170.

Kosecet

al.

(2016)

(IFPRI)

November

3,2016

11/26

EmpiricalStrategyandData

EconometricSpecification

Weestimatethefollowinglinearprob

ability

model:

Ei=

β0+

β1Li+

γX

i+

αj+

ǫ i

where iindexes

individuals

Eiaremigration

(during201

0–14

)andem

ploym

ent(in20

14)ou

tcom

esLiislogg

edexpectedlandinheritance

Xiisavector

ofindividual-level

control

variables(gender,oldestmale

indicator,age,

marital

status,child

ofheadindicator,prim

aryschool

completionindicator,over

age18

attimeof

mostrecentland

redistribution

indicator,fixedeff

ects

forexactpermutation

ofolder

co-descendants,andindicator

forhavingmorethan

onemale

co-descendantim

mediately

below

one’sselfin

thebirth

order)

αjarehouseholdfixedeff

ects

Standarderrors

areclustered

atthekebelelevel,theadministrativeunitin

whichlandredistribution

policiesareexecuted

Kosecet

al.

(2016)

(IFPRI)

November

3,2016

12/26

EmpiricalStrategyandData

IdentificationStrategy

Exp

ectedlandinheritance

islikelyto

beendog

enou

sto

migration,

employm

ent,andeducation

decisions:

Ahou

sehold’slandendow

ment,wealth,andincomelevelscould

influence

bothmigration

andem

ploym

entdecisionsandexpected

inheritance

Within

ahou

sehold,parents

may

select

descendants

withparticular

unob

served

characteristicsthat

also

influence

migration

and

employm

ent—

such

asaphysical

aptitudeforagriculture—

forlarger

inheritances

Weaddress

thefirstconcern

byincludinghou

seholdfixedeff

ects

Weaddress

thesecondconcern

usingindividual

controlsand

employingan

instrumentalvariables(IV)strategy

Kosecet

al.

(2016)

(IFPRI)

November

3,2016

13/26

EmpiricalStrategyandData

IVStrategy

Leverageauniquefeature

ofEthiopia:givenpublic

ownership

ofland,

landaccess

isinfluencedby

governmenteff

orts

toredistribute

land

Interviewswithkebeleoffi

cialssugg

estthat

males

over

age18

received

priority

inredistribution

sIn

ourdata,

medianmalelandinheritance

(interm

sof

landarea)is

60percentgreaterthan

that

ofmedianfemale

Sugg

ests

that

HHswithagreatershareof

maledescendants

overage

18at

thetimeof

theredistribution

shou

ldhavemorelandto

allocate

across

descendants

Hou

seholdfixedeff

ects

capture

averageim

pacts

ofredistribution

Within

ahou

sehold,weexpectindividualsat

highrisk

ofreceivinga

smallinheritance

tobenefitmostfrom

havingagreatershareofmale

co-descendants

beover

age18

atthetimeof

redistribution

Datareveal

onesuch

“highrisk”group:thosewith>

1male

descendantim

mediately

below

them

inthebirth

order

Kosecet

al.

(2016)

(IFPRI)

November

3,2016

14/26

EmpiricalStrategyandData

ExcludedInstrument

Weuse

asingleinteractionterm

asan

instrumentforexpectedland

inheritance:

Theshareof

maledescendants

inthehou

seholdwhowereoverage

18at

thetimeof

themostrecentlandredistribution

(callthisr i)

interacted

withadummyforhavingmorethan

onemaledescendant

immediately

follow

theindividual

inbirth

order

(callthism

i)

Thefirststageequationis:

Li=

δ 0+

δ 1r i×

mi+δ 2m

i+θX

i+πj+

ηi

where: πjarehou

seholdfixedeff

ects

r idoes

not

appearin

itslevelform

asitiscollinearwithhou

sehold

FE

ResearchSem

inar

(IFPRI)

November

3,2016

15/26

EmpiricalStrategyandData

IdentifyingAssumption

Thedifferencein

theeff

ectof

havingalarger

shareof

male

descendants

beover

age18

atthetimeof

landredistribution

onthose

withversuswithou

t>

1maledescendantim

mediately

below

them

inbirth

order

only

affects

migration

andem

ploym

entdecisionsthrough

itseff

ecton

expectedlandinheritance

Theindividual

components

oftheexcluded

instrument,m

iandr i,are

included

inou

rmainspecification

andweexplicitly

allow

them

both

todirectlyim

pactou

rmigration

andem

ploym

entou

tcom

es

Thus,weneedon

lybelieve

that

theirinteractionisavalid

instrument—

not

that

either

ofthetwovariablesalon

eis

ResearchSem

inar

(IFPRI)

November

3,2016

16/26

EmpiricalStrategyandData

IVFirst

Stage:Im

pactof

Excluded

Instrumenton

Log

LandInheritence

(1)

(2)

Excluded

instrumen

t:Shareofmale

descendants

18+

attimeofland

redistribution×

Dummyfor>

1male

descendantim

med

iately

followingin

birth

order

2.501***

2.478***

(0.460)

(0.480)

Dummy–>

1male

descendantim

med

iately

followsin

birth

order

-0.082

-0.079

(0.061)

(0.065)

Observations

1,170

1,170

R-squared

0.902

0.902

Number

ofhouseholds

625

625

First

stageFstat

29.59

26.63

Fullsetofindividual-levelco

ntrols?

No

Yes

Source:

Authors’calculationsbased

on

IFPRI’sWatershed

Surveysof

2010and2014.

Notes:

Standard

errors

arein

paren

theses

and

clustered

atthekebele

level.

∗∗∗indicatessignificance

atthe0.01level.

ResearchSem

inar

(IFPRI)

November

3,2016

17/26

Results

Impacts

onPermanentMigration:OLSEstim

ates

Anyw

here

Outofworeda

Tourbanarea

Loglandinheritance

0.025

-0.165**

-0.085**

-0.252**

-0.128***

-0.283***

(0.051)

(0.076)

(0.037)

(0.094)

(0.043)

(0.081)

Additionalco

ntrols

No

Yes

No

Yes

No

Yes

Observations

1,170

1,170

1,167

1,167

1,167

1,167

R-squared

0.001

0.783

0.011

0.788

0.024

0.800

Number

ofhouseholds

625

625

624

624

624

624

Source:Authors’calculationsbasedonIFPRI’sWatershed

Surveysof2010and2014.

Notes:Standard

errors

arein

paren

theses

and

clustered

atthekebelelevel.

∗∗∗,∗∗,and

∗

indicate

significance

atthe0.01,0.05,and0.10levels,respectively.

ResearchSem

inar

(IFPRI)

November

3,2016

18/26

Results

Impacts

onSectorof

PrimaryEmploym

ent:

OLSEstim

ates

Agricu

lture

Non-agricu

lture

Study

Loglandinheritance

0.309***

0.262**

-0.059

-0.180

-0.142***

-0.050

(0.044)

(0.109)

(0.042)

(0.126)

(0.042)

(0.201)

Additionalco

ntrols

No

Yes

No

Yes

No

Yes

Observations

1,167

1,167

1,167

1,167

1,167

1,167

R-squared

0.095

0.815

0.006

0.753

0.021

0.778

Number

ofhouseholds

625

625

625

625

625

625

Source:Authors’calculationsbasedonIFPRI’sWatershed

Surveysof2010and2014.

Notes:Standard

errors

arein

paren

theses

and

clustered

atthekebelelevel.

∗∗∗,∗∗,and

∗

indicate

significance

atthe0.01,0.05,and0.10levels,respectively.

ResearchSem

inar

(IFPRI)

November

3,2016

19/26

Results

Impacts

onPermanentMigration:IV

Estim

ates

OLS

IV

(1)

(2)

(3)

(4)

(5)

(6)

Dummy-migrated...

Anyw

here

Outof

Tourban

Anyw

here

Outof

Tourban

woreda

area

woreda

area

Lnlandinheritance

-0.165**

-0.252**

-0.283***

-0.198

-0.855***

-0.508***

(0.076)

(0.094)

(0.081)

(0.199)

(0.173)

(0.173)

Observations

1,170

1,167

1,167

1,170

1,167

1,167

R-squared

0.783

0.788

0.800

0.783

0.727

0.791

#ofhouseholds

625

624

624

625

624

624

First

StageF-Stat

21.73

21.73

21.73

Source:Authors’calculationsbasedonIFPRI’sWatershed

Surveysof2010and2014.

Notes:Standard

errors

arein

paren

theses

and

clustered

atthekebelelevel.

∗∗∗,∗∗,and

∗

indicate

significance

atthe0.01,0.05,and0.10levels,respectively.

ResearchSem

inar

(IFPRI)

November

3,2016

20/26

Results

Impacts

onSectorof

PrimaryEmploym

ent:

IVEstim

ates

OLS

IV

(1)

(2)

(3)

(4)

(5)

(6)

Dummy-primarilyem

ployedin

...

Agricu

lture

Non-

Study

Agricu

lture

Non-

Study

agric.

agric.

Lnlandinheritance

0.262**

-0.180

-0.050

0.655***

-0.427***

-0.171

(0.109)

(0.126)

(0.201)

(0.168)

(0.095)

(0.140)

Observations

1,167

1,167

1,167

1,167

1,167

1,167

R-squared

0.815

0.753

0.778

0.799

0.742

0.776

#ofhouseholds

625

625

625

625

625

625

First

StageF-Stat

22.61

22.61

22.61

Source:Authors’calculationsbasedonIFPRI’sWatershed

Surveysof2010and2014.

Notes:Standard

errors

arein

paren

theses

and

clustered

atthekebelelevel.

∗∗∗,∗∗,and

∗

indicate

significance

atthe0.01,0.05,and0.10levels,respectively.

ResearchSem

inar

(IFPRI)

November

3,2016

21/26

Results

Resultsby

Gender

(1)

(2)

(3)

(4)

(5)

(6)

Dummy-migrated...

Dummy-em

ployedin

...

Any-

Outof

Urban

Agric.

Non-

Study

where

woreda

area

agric.

Lninheritance

(women

)0.097

0.024

-0.022

0.154***

-0.078

-0.082

(0.073)

(0.044)

(0.053)

(0.047)

(0.047)

(0.065)

Lninheritance

(men

)-0.003

-0.155***

-0.188***

0.226***

-0.163***

-0.058

(0.062)

(0.047)

(0.051)

(0.047)

(0.042)

(0.061)

Observations

1,170

1,167

1,167

1,167

1,167

1,167

R-squared

0.418

0.449

0.415

0.563

0.436

0.525

#ofhouseholds

625

624

624

625

625

625

P-valueofdifferen

ce0.11

0.002

0.004

0.093

0.022

0.614

Source:Authors’calculationsbasedonIFPRI’sWatershed

Surveysof2010and2014.

Notes:Standard

errors

arein

paren

theses

and

clustered

atthekebelelevel.

∗∗∗,∗∗,and

∗

indicate

significance

atthe0.01,0.05,and0.10levels,respectively.

ResearchSem

inar

(IFPRI)

November

3,2016

22/26

Results

Resultsby

Age

(1)

(2)

(3)

(4)

(5)

(6)

Dummy-migrated...

Dummy-em

ployedin

...

Any-

Outof

Urban

Agric.

Non-

Study

where

woreda

area

agric.

Lninheritance

(20-34)

-0.010

-0.104*

-0.149**

0.220***

-0.139***

-0.011

(0.068)

(0.055)

(0.058)

(0.053)

(0.040)

(0.058)

Lninheritance

(15-19)

-0.024

-0.028

-0.107

0.221***

-0.061

-0.082

(0.09)

(0.056)

(0.066)

(0.046)

(0.048)

(0.077)

Observations

1,170

1,167

1,167

1,167

1,167

1,167

R-squared

0.407

0.425

0.393

0.542

0.424

0.501

#ofhouseholds

625

624

624

625

625

625

P-valueofdifferen

ce0.799

0.096

0.317

0.984

0.069

0.181

Source:Authors’calculationsbasedonIFPRI’sWatershed

Surveysof2010and2014.

Notes:Standard

errors

arein

paren

theses

and

clustered

atthekebelelevel.

∗∗∗,∗∗,and

∗

indicate

significance

atthe0.01,0.05,and0.10levels,respectively.

ResearchSem

inar

(IFPRI)

November

3,2016

23/26

Results

Resultsby

LandRentalMarketActivity(Low

vs.High)

(1)

(2)

(3)

(4)

(5)

(6)

Dummy-migrated...

Dummy-em

ployedin

...

Any-

Outof

Urban

Agric.

Non-

Study

where

woreda

area

agric.

Lninheritance

(low)

-0.337***

-0.181*

-0.514***

0.410***

-0.661***

0.418**

(0.108)

(0.094)

(0.080)

(0.099)

(0.120)

(0.196)

Lninheritance

(high)

-0.179

-0.378**

-0.267***

0.443***

0.006

-0.361**

(0.119)

(0.151)

(0.077)

(0.143)

(0.143)

(0.155)

Observations

1,170

1,167

1,167

1,167

1,167

1,167

R-squared

0.830

0.848

0.850

0.869

0.826

0.850

#ofhouseholds

625

624

624

625

625

625

P-valueofdifferen

ce0.335

0.280

0.035

0.853

0.001

0.004

Source:Authors’calculationsbasedonIFPRI’sWatershed

Surveysof2010and2014.

Notes:Standard

errors

arein

paren

theses

and

clustered

atthekebelelevel.

∗∗∗,∗∗,and

∗

indicate

significance

atthe0.01,0.05,and0.10levels,respectively.

ResearchSem

inar

(IFPRI)

November

3,2016

24/26

Lninheritance

(low)

-0.337***

-0.181*

-0.514***

0.410***

-0.661***

0.418**

(0.108)

(0.094)

(0.080)

(0.099)

(0.120)

(0.196)

Lninheritance

(low)

-0.337***

-0.181*

-0.514***

0.410***

-0.661***

0.418**

(0.108)

(0.094)

(0.080)

(0.099)

(0.120)

(0.196)

Lninheritance

(high)

-0.179

-0.378**

-0.267***

0.443***

0.006

-0.361**

(0.119)

(0.151)

(0.077)

(0.143)

(0.143)

(0.155)

Lninheritance

(high)

-0.179

-0.378**

-0.267***

0.443***

0.006

-0.361**

(0.119)

(0.151)

(0.077)

(0.143)

(0.143)

(0.155)

Results

Resultsby

Proximityto

Urban

Center(C

lose

vs.Far)

(1)

(2)

(3)

(4)

(5)

(6)

Dummy-migrated...

Dummy-em

ployedin

...

Any-

Outof

Urban

Agric.

Non-

Study

where

woreda

area

agric.

Lninheritance

(close)

0.430***

0.173

-0.071

0.473**

0.090

-0.610***

(0.079)

(0.131)

(0.064)

(0.204)

(0.121)

(0.104)

Lninheritance

(far)

-0.126

-0.330**

-0.317***

0.234*

-0.319**

0.116

(0.100)

(0.158)

(0.073)

(0.125)

(0.148)

(0.289)

Observations

1,170

1,167

1,167

1,167

1,167

1,167

R-squared

0.841

0.849

0.871

0.892

0.828

0.849

#ofhouseholds

625

624

624

625

625

625

P-valueofdifferen

ce0.000

0.021

0.018

0.327

0.042

0.026

Source:Authors’calculationsbasedonIFPRI’sWatershed

Surveysof2010and2014.

Notes:Standard

errors

arein

paren

theses

and

clustered

atthekebelelevel.

∗∗∗,∗∗,and

∗

indicate

significance

atthe0.01,0.05,and0.10levels,respectively.

ResearchSem

inar

(IFPRI)

November

3,2016

25/26

Lninheritance

(close)

0.430***

0.173

-0.071

0.473**

0.090

-0.610***

Lninheritance

(close)

0.430***

0.173

-0.071

0.473**

0.090

-0.610***

Lninheritance

(close)

0.430***

0.173

-0.071

0.473**

0.090

-0.610***

(0.079)

(0.131)

(0.064)

(0.204)

(0.121)

(0.104)

(0.079)

(0.131)

(0.064)

(0.204)

(0.121)

(0.104)

(0.079)

(0.131)

(0.064)

(0.204)

(0.121)

(0.104)

Lninheritance

(far)

-0.126

-0.330**

-0.317***

0.234*

-0.319**

0.116

(0.100)

(0.158)

(0.073)

(0.125)

(0.148)

(0.289)

Lninheritance

(far)

-0.126

-0.330**

-0.317***

0.234*

-0.319**

0.116

(0.100)

(0.158)

(0.073)

(0.125)

(0.148)

(0.289)

Lninheritance

(far)

-0.126

-0.330**

-0.317***

0.234*

-0.319**

0.116

(0.100)

(0.158)

(0.073)

(0.125)

(0.148)

(0.289)

Conclusion

Conclusion

Approach:

Use

panel

datafrom

from

2010

and20

14from

1,74

8hou

seholdsin

27

woredas

(districts)of

ruralAmharaandOromia

region

sof

Ethiopia

Focuson

youth

aged

15–3

4;estimateahou

seholdfixedeff

ects

model

Exploitexog

enou

svariationin

thetimingof

landredistributionsto

overcomeendog

enou

shou

seholddecisionsab

outinheritance

size

Findings:

Largerexpectedlandinheritances:

Low

erthelikelihoodoflong-distance

andrural-to-urban

permanent

migrationduring20

10–1

4Increase

thelikelihoodof

employm

entin

agriculture

in20

14Reduce

thelikelihoodof

employm

entin

thenon

-agriculturalsector

Donot

affectthedecisionto

study

Resultsappearto

bedrivenby

males

andolder

youth

Inheritance

plays

amorepron

ouncedrole

inareaswithless-vibrant

landmarkets

andin

moreremoteareas

ResearchSem

inar

(IFPRI)

November

3,2016

26/26