participation of smallholder farmers in biofuels crop and land rental markets: evidence from south...

18
Participation of Smallholder Farmers in Biofuels Crop and Land Rental Markets: Evidence from South Africa Athur Mabiso International Conference of Agricultural Economists, Foz do Iguaçu, Brazil August 20, 2012 1

TRANSCRIPT

Participation of Smallholder Farmers in Biofuels Crop and Land Rental

Markets: Evidence from South Africa

Athur Mabiso International Conference of Agricultural Economists,

Foz do Iguaçu, Brazil August 20, 2012

1

Motivation

Increased investments in biodiesel and ethanol between 2005 and 2009 Public-Private Partnerships (PPPs); mainly European firms with

government development banks e.g. Industrial Development Corporation (IDC) and Eastern Cape Development Corporation (ECDC)

New feedstock crops for smallholders: Canola, Soy beans, Sunflower, Sugar beets

New R&D investments in cellulosic and algae biofuels – University of Cape town and Nelson Mandela Metropolitan University

2

3

Firm Year Established

Feedstock (Biofuel)

Investment (Million Rand) Investment Type

Relationship with smallholder farmers

Annual Production (litres)

1 2005 Maize, Sugar beets (E-export) 3 Domestic private equity Maize supply contract 400,000

2 2006 Maize (E) Not provided Domestic-private equity Only large scale commercial in 2009

Not provided

3 1995 Sugar beets, sorghum (E) 1,800

PPP with European investors

Only large scale commercial in 2009 0

4 2004 Canola (BD-export) 3,270

PPP with European investors Land rental contract 0

5 2006 Soy beans, sunflower (BD) 2.7

Domestic NGO with Provincial government support

Soybean and Sunflower supply contracts, extension labor contract 420,000

6 2008

Soy beans, Sunflower, waste vegetable oil (BD) 1

Domestic private equity with government grant Multiple feedstock contracts 300,000

7 2007 Algae (BD) Not provided University-based and Domestic private equity None 20,000

8 2007 Waste vegetable oil; soy beans (BD) 50

Domestic Leveraged buyout of foreign multinational Soybean supply contract 4,000

9 2006

Waste vegetable oil, soy beans (BD-domestic) 3

Domestic-European investor partnership Soybean supply contract 70,000

10 2008 Soy beans (BD-export) 1,500

Domestic-Foreign (non-European) investor partnership

Soybean supply contract; sale of animal feed byproduct 0

11 2007

Sunflower, soybeans, waste vegetable oil (BD) 300 Domestic private equity

Land rental and supply contract 120,000

Complex Biofuels Policy Create employment, reduce poverty, increase supply of

renewable energy without compromising food security and environment

Tie biofuels to on-going land redistribution program Subsidies: public investments to crowd in private investments

(PPP) and farm input subsidies to support smallholder farmers Ban maize and jatropha as biofuels feedstock (food security &

environmental concerns) No mandate for biofuels market uptake (market distortion

concern) Retail tax exemption on domestic biodiesel and ethanol

purchases

4

Research Objectives

To assess how smallholder farmers participate in new biofuels investments1/markets and the implications for poverty reduction and food security

5

1. New biofuels developments in South Africa include investments in feedstock production of sunflower, soy beans, canola, sugar beets, algae, maize, sorghum and waste vegetable oil

Data & Methods

• Key informant interviews of biofuels firms, government development banks and relevant ministries (agriculture, department of energy and central energy fund)

• Followed by farm household survey Random samples from 8 villages Purposively chosen based on presence of biofuels

firms 247 Farm households (93 biofuels participants, 154

non-participants)

6

Study Site

7

Eastern Cape (4 firms) Limpopo ( 1 firm) North West (1 firm) Farmers working with firms

in Gauteng, KwaZulu Natal (sugarcane-ethanol), Mpumalanga and Free State were not included

Types of Participation in Biofuels

Biofuels-related market

Participated in biofuels market Did not participate

Land-rental market Rent land to biofuels firm (canola) Rent land to biofuels-producing smallholder farmers (maize, soy beans, sunflower)

Rent land to non-biofuels firm/farmer or leave land unfarmed

Cash crop market Produce biofuels feedstock crop for sale (maize, soy beans, sunflower)

Produce non-biofuels (food/feed/fiber) crop for own consumption and/or sale

Labor market Agricultural extension (Mentoring newly-enrolled biofuels crop farmers)

Provide farm-labor Provide non-farm labor

8

Descriptive Statistics

9

Grow crops for biofuels

firm

Rent land to biofuels firm

Did not participate in

biofuels Total Household size (standard dev.)

2.8 (2.2)

4.6 (2.9)

4.6 (2.5)

4.2 (2.6)

Age in years (standard dev)

52.9 (13.7)

58.3 (11.8)

57.5 (13.1)

56.7 (13.1)

Age: 21-40 years 20.4% 2.3% 11.7% 11.7% 41-60 years 46.9% 59.1% 45.5% 48.2% 61-89 years 32.7% 38.6% 42.9% 40.1%

Sex: female 24.5% 25.0% 42.5% 35.8% Education: Grade 1-9 21.7% 56.8% 55.2% 49.2% Grade 10-Grade 12 32.6% 15.9% 20.8% 22.1% Landholding size (ha) 16.1 10.9 6.0 8.8 Distance to market (km) 50.98 50.05 31.4 48.9

Econometric Modeling • Descriptive and qualitative • Multinomial probit for participation decision • Mixed multinomial treatment effects model (Deb & Trivedi,

2006)

1. lease land to biofuels firm I = 2. sell crops to biofuels firm 3. do neither 1 nor 2 Y1 = Household cash income Y2 = Household monthly expenditure on groceries

10

Econometric Modeling X = household characteristics, community-level market variables Had to worry about:

Selection bias, hence use of treatment effects model Exclusion restrictions: Last year’s savings/access to credit, distance to main market and

membership in a cooperative or group How we measured income (hence tested effects on both cash income and

consumption expenditure) Functional form – explored multiple specifications with little changes in

parameter estimates Sensitivity of estimated parameters to variations in sample coverage –

estimated models using subsamples of provinces Distributional assumption - attempted beta distributional assumption but did

not converge; potentially overspecified model Estimated alternative GMM models using subsamples to see which variables

influenced incomes between participant types

11

Determinants of biofuels market participation: Multinomial probit model

Prob (Grow crops for biofuels firm)

Prob (Rent land to biofuels firm)

Received price information 0.18 *** 0.27 *** Price of maize 4.92E-06 1.73E-04 Land rental rate -2.35E-03 * 1.71E-03 Off-farm wage rate -4.96E-08 1.28E-03 ** Distance to main market 1.25 E-05 * 6.3E-04 Savings last year/access to credit 1.27E-06 ** -7.05E-05 Cooperative or group member 2.18E-05 *** 0.02 *** Dependency ratio 0.0047 0.19 * Sex: female -0.016 * -0.09 * Age:

41-60 years 2.39E-04 * 0.13 * 61-89 years -1.95E-03 0.33

Landholding size 3.67E-06 0.01

12 *** Significance at 1% level; ** Significant at 5% level, * Significant at 10% level

Association of biofuels participation with household incomes and expenditure

Variable Household

income Grocery

expenditure Sample Mean

or % Household size -230.88 40.78 *** 4.27 Dependency ratio -126.54 ** -28.66 ** 0.32 Grade 1-9± -606.71 ** 21.27 49.2% Completed Grade 12 - diploma± 862.90 ** 7.32 * 13.6% Bachelor's degree of higher± 325.95 3.09 2.5% Marital status: Married civil/religious± 431.69 -56.63 49.2% Married traditional/polygamous± -301.10 -22.69 29.3% Landholding size 73.68 *** 4.29 *** 8.82 Received price information± -12.64 -10.42 57.5% Price of maize -4.91 ** -0.03 *** 1398.68 Off-farm wage rate 7.63 *** 0.13 ** 2340.52 Land rental rate 16.13 -0.33 * 421.57 Grow biofuels crop 620.21 18.68 19.8% Rent land to biofuels firm -93.82 18.15 17.9%

13 ± If variable is continuous the marginal effect is evaluated at the mean shown in the fourth column otherwise for categorical variables the marginal effect is evaluated as the percentage change due to a change in the dummy variable from 0 to 1 *** Significant at the 1% level, ** Significant at the 5% level, * Significant at the 10% level.

Discussion • Price information and membership to

cooperative/farmers group increases smallholder participation in biofuels markets

• Savings or access to credit enable participation in biofuels feedstock crop production for sale – Biofuels firms providing input credit under contract further

enable participation of smallholder farmers • Increases in off-farm wage rate increase land rentals by

resource poor households enabling them to realize income from land rent and non-farm employment – Complementary initiatives to create non-farm rural

employment may be critical as land rental markets takeoff

14

Discussion • Biofuels did not significantly create non-farm rural employment

– Limited potential due to capital intensive structure of production at processing level

– Feedstock supply bottleneck implies jobs potential in farm production – Limited farming skills and knowledge of new crops chosen by investors

creates challenges for smallholder farmers – Government ban on maize probably justified for food security reasons

in 2008/9 though this is the crop farmers are familiar with • Land rental markets appear to be helping certain groups, e.g.

women’s labor constraints relaxed somewhat – However, fewer women leased out land – may be due to low access to land in the first place and/or fear to lose

land if found not to be farming

15

Discussion

• While incomes are different between participants and non participants, the difference was not statistically significant

• Study limitation is we did not use panel data, nonetheless results suggest biofuels not associated with increased incomes but, Increased choices for farm households Spillover effects, especially through input and information markets

may explain lack of differences in incomes

16

Summary and Conclusion • Biofuels developments have complex land and labor market

implications for households (expanded choice set and tradeoffs) – Skills and knowledge of new crops needed for farm production – Complementary non-farm employment is needed to draw labor saved due

to land rental market – Women benefit less with current structure, but land rental markets

appear to relax farm labor constraints • Market variables, particularly market information and farmer group

coordination, are critical to participation – Provision of information on land rental rates and land values may be

useful – Important to provide information on prices of crops new to the area e.g.

canola and sugar beets

17

Muito Obrigado!

18