Non-farm income and labor markets in rural Ethiopia

ETHIOPIAN DEVELOPMENT RESEARCH INSTITUTE

Non-farm income and labor markets in rural Ethiopia

Fantu Bachewe, Guush Berhane, Bart Minten, and Alemayehu S. TaffesseIFPRI-ESSP

Transformation and vulnerability in Ethiopia: New evidence to inform policy and investments

27 May 2016Getfam Hotel, Addis Ababa

1

2

1. Introduction

• Development of well-functioning labor markets crucial for economic growth and livelihood opportunities, especially for youth

• Rural wage increases strongly linked with poverty reduction

• The poor regularly depend on such wages for their livelihood

• Important to understand the off-farm sector and labor markets

• Ethiopia’s economy is changing fast but unclear how important the off-farm economy and labor income is in rural areas

3

1. Introduction

• Purposes of the presentation:

- How important is the rural off-farm economy?

- What are the associates of off-farm income?

- Are rural wages changing?

- What are the drivers and implications of that change?

• A number of datasets used

• CSA wage data;

• Agricultural Growth Programme (AGP) survey;

• Other large-scale household datasets used.

4

3. Off-farm income in rural areas

• Crop income as source of rural income overwhelming (71%)

• Wage income (10%) as important as livestock income

• Agricultural wages twice as important as non-agricultural wages

• Enterprise income: 8%

0

20

40

60

80

100

All Tigray Amhara Oromiya SNNP

%

Income sources in rural areas (AGP)

Crop Livestock Agricultural wage Non-ag. wage Enterprise

5


• Off-farm income and wage income especially important for the poor

• Enterprise income shares higher for poorest and richest, reflecting different enterprises

0

10

20

30

40

50

60

70

80

90

0

2

4

6

8

10

12

14

Quintile I Quintile II Quintile III Quintile IV Quintile V

Co

ntr

ibu

tio

n o

f cr

op

inco

me

(%)

Co

ntr

ibu

tio

n o

f n

on

-cro

p in

com

e (%

)

Ag wage Non-ag wage Enterprise Livestock Crop

6


• Off-farm income and wage income especially important for the poor

• Enterprise income shares higher for poorest and richest, reflecting different enterprises

0

10

20

30

40

50

60

70

80

90

0

2

4

6

8

10

12

14


Co

ntr

ibu

tio

n o

f cr

op

inco

me

(%)

Co

ntr

ibu

tio

n o

f n

on

-cro

p in

com

e (%

)

Ag wage Non-ag wage Enterprise Livestock Crop

• Prepared food/drinks & crafts more important for the poor

• Agricultural output & merchandise trade more important for the rich

Type of business enterprises engaged, by wealth quintile (%)


0

5

10

15

20

25

30

35

40


Food and local drinks Crafts Trade-agricultural Outputs Trade-merchendise


• Agricultural output & merchandise trade more important for the rich



0

5

10

15

20

25

30

35

40


Food and local drinks Crafts Trade-agricultural Outputs

Trade-merchendise Others

9


• 18% off-farm income in rural Ethiopia:

- Small compared to other African countries

- Small compared to Asia and Latin-America

0

10

20

30

40

50

60

Ethiopia Africa Asia Latin-America

%

Local non-farm Migration income

10

4. Agricultural wage labor

• Hired agricultural wage labor: 7% of all labor

• Relatively more important in Tigray and SNNP

0

20

40

60

80

100

All Tigray Amhara Oromiya SNNP

%

Hired labor in crop production

Family labor Hired labor

11


• Importance of hired labor varies by task

• In teff, less hired labor during production; most at harvest and afterwards

0

20

40

60

80

100

%

Labor by crop activity

Family Hired Exchange

12


• Labor arrangement in teff differ by remoteness

• Monetization/use of hired labor higher when less remote;

• In more remote areas, more reliance on exchange labor

Family, wage, and exchange labor use in teff production

020

4060

80

shar

e (%

)

0 50 100 150distance to Addis (Birr/quintal)

95% CI family labor

95% CI wage labor

95% CI exchange labor

13

5. Rural wages - changes

• Rely on CSA price data from 2004 to 2015

• Use different ways of converting/deflation (exchange rate; CPI)

• Wages in US dollars over 3 times higher in 2015 compared to 2004

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Jul-

04

Feb

-05

Sep

-05

Ap

r-0

6

No

v-0

6

Jun

-07

Jan

-08

Au

g-0

8

Mar

-09

Oct

-09

May

-10

Dec

-10

Jul-

11

Feb

-12

Sep

-12

Ap

r-1

3

No

v-1

3

Jun

-14

Jan

-15

Au

g-1

5

Wages in US dollars

14


• Urban & rural wages deflated by general consumer price index (GCPI)

• Between 2004 and 2015: Real wages increased by 54% in rural areas and by 64% in urban areas

20

25

30

35

40

45

50

Jul-

04

Feb

-05

Sep

-05

Ap

r-0

6

No

v-0

6

Jun

-07

Jan

-08

Au

g-0

8

Mar

-09

Oct

-09

May

-10

Dec

-10

Jul-

11

Feb

-12

Sep

-12

Ap

r-1

3

No

v-1

3

Jun

-14

Jan

-15

Au

g-1

5

Rea

l wag

es in

Dec

emb

er 2

01

1 p

rice

s

GCPI deflated-rural

GCPI deflated-urban

15


• Agricultural wages on average 1.27 USD per day (in AGP zones); however significant variation

• Compared to Asian countries, wages significantly lower; about 1 USD higher in Nepal and Myanmar; 1.59 USD higher in Bangladesh

05

10

15

20

Ob

serv

ations (

pe

rcen

t)

0 2 4 6 8

Agricultural wages (USD/day) 0 3 6 9 12

Nepal (2010)

Sri Lanka (2012)

Myanmar (2004)

Malaysia (2012)

Bangladesh (2010)

Ethiopia (2012)

USD/day

Distribution of wages in AGP zones Comparison of wages in Ethiopia with Asian countries

16

5. Rural wages - drivers

Look at the drivers for wage changes:

• Estimate the following regression: Ln (real wage) = a + b lnY + e

• Growth in unskilled real wage linked with economic growth, particularly with agricultural growth

Economic growth and unskilled real wage elasticity

Real… Elasticity of rural wages

GDP 0.22

Agricultural GDP 0.24

Manufacturing GDP 0.19

Industry GDP 0.17

Services GDP 0.15

Note: All estimates are statistically significant at 1%.

17

5. Rural wages - drivers

Look at the drivers for wage changes:

• Role of the Productive Safety Net Program (PSNP)

20

25

30

35

40

45

19

96

19

97

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

20

16

Real wage PSNP vs non-PSNP areas (real birr/day)

PSNP non-PSNP

18

5. Rural wages - implications

1. Herbicides use increasing• Rapidly taking off in Ethiopia (2013 imports 5 times higher than 2002)• Herbicides is negatively associated or a substitute for weeding labor

2. Mechanization: - Higher wages provide incentives for mechanization in agriculture- In Ethiopia, mechanization is still rather low; In FTF survey, estimated:- 9% of farmers use some form of mechanization

020

40

60

80

Pove

rty

hea

d c

ou

nt

ind

ex

10 20 30 40 50 60

Real wages, December 2011 prices

95% confidence interval

Predicted poverty head count index

3. Poverty:

Real wages and poverty head count index negatively correlated

19

6. Conclusions

• Major findings:

1/ Off-farm income makes up 18% of total income of rural households;

• Wage income contributes 10%, as important as livestock income

2/ Off-farm income especially important for the poorest

• Makes up 26% of their total income (agricultural wage 13%)

3/ Rural wages are rapidly increasing;

• Real wages 54% higher in 2015 compared to 2004; • Improved agricultural performance contributed to that change

4/ Off-farm income and wages significantly lower than other countries

5/ Implications of wage changes on poverty and agricultural production practices (more use of herbicides)

20

6. Conclusions

• Implications:

1/ Low wages have been an asset to attract labor-intensive industries.

• This might be changing and Ethiopia might slightly lose that edge

• Ensure that the youth upgrade skills towards higher labor productivity

2/ Push for the adoption of labor-saving technologies;

• Important that Ethiopia pro-actively implements policies that allow appropriate technologies at low costs

3/ Ensure flexible labor markets so that people can benefit from these opportunities

Thank you.

Thank you.

23


• ERSS (national representative) paints a rather similar picture

• Differences: 1/ Livestock more important than in AGP areas; 2/ In the 4 regions, higher wage and enterprise income in high potential areas; 3/ Wage and enterprise income in four regions lower than in other regions

0

20

40

60

80

100

Fourregions

Tigray Amhara Oromiya SNNP Fourregions

Afar &Somalie

BG, DD,Gambella,& Harari

%

AGP (2010/11) ERSS (2013/14)

Crop Livestock Ag. Wage Non-ag. Wage Enterprise

Crafts16%

Trade-grains16%

Trade-livestock11%

Trade-non agricultural

10%Milling

1%

Tella (local ale)5%

Araqi (local liquor)

19%

Enjera/Dabbo (bread)

3%

Other19%

Type of business enterprise engaged, (%)


25


• Major results on diversification (Tobit model with diversification index as dependent variable):

1. Younger heads of households are more likely to be associated with off-farm income (esp. agricultural wages and enterprise income)

2. Education associated with higher enterprise and non-ag. income

3. Gender link: Women make up 1/3rd of hired labor; men 2/3rds.

4. More and better quality agricultural assets associated with less diversification

5. Distance to cities an important associate of non-farm income. Households 100 kms farther from Addis have 11% lower share of off-farm income.

26


• Associates of agricultural wages:

1. Time of year: compare to land preparation,

• 13 % lower at planting, 12% higher at weeding; 17% higher at harvesting

2. Gender: men earn 8% more

3. Age: older people earn less (0.2% less per year extra)

4. Remoteness: 100 kms away from Addis reduces wage by 7%

5. Poverty in the zone: The higher the poverty, the lower the wage

6. Regions: higher wages in Tigray and Amhara

27


4. Impact of off-farm income on modern input use

- Off-farm income might lead to relaxation of credit constraints in period of input needs

- Study finds positive association between fertilizer use and off-farm income (using probit model on AGP data):

• Households that had off-farm income were 7% more likely to use chemical fertilizer

- No association with improved seeds and agro-chemicals

Contribution of income sources by wealth quintile (all assets included)

0

10

20

30

40

50

60

70

80

0

2

4

6

8

10

12

14

16

18


Co

ntr

ibu

tio

n o

f cr

op

inco

me

(%)

Co

ntr

ibu

tio

n o

f n

on

-cro

p in

com

e (

%)

Livestock Wage Enterprise Crop

29


2. Increasing use of herbicides

0

5

10

15

20

Herbicides imports (millions USD)

30


2. Increasing use of herbicides

Herbicides is a substitute for weeding; rapidly taking off in Ethiopia

01

23

log

(valu

e h

erb

icid

es/h

a)

0 2 4 6Log(weeding labor/ha)

95% interval correlation

02

04

06

08

01

00

% o

f fa

rmers

0 50 100 150Transport costs to Addis (Birr/quintal)

95% CI herbicides 2012

95% CI herbicides 2002

31


3. Mechanization

- Higher wages provide incentives for mechanization in agriculture

- In Ethiopia, mechanization is still rather low; In FTF survey, estimated:

1. 9% of farmers use some form of mechanization

• 5% in plowing; 3% in harvesting; 2% in threshing

2. Strong threshold effect (Berhane et al., 2016)

Rural wages and use of

mechanization


• Agricultural outputs & merchandise trade more important for the rich



0

5

10

15

20

25

30

35

40

45

Quintile I Quintile II Quintile III Quintile IV Quintile VFood and local drinks Crafts Trade-agricultural OutputsTrade-merchendise Others

Government & Nonprofit

Non-farm income and labor markets in rural Ethiopia