Agro-industry investments, smallholders and workers: evidences on household income effects from...

Preview:

Citation preview

Agro-industry investments, smallholders and workers: evidences on household income effects

from Tanzania

Raoul Herrmann1, 2, Khamaldin Mutabazi 3, Ulrike Grote2

1German Development Institute, Germany2 Leibniz University of Hannover, Germany

3 Sokoine University of Agriculture, Tanzania

Annual World Bank Conference on Land and Poverty, 24 March 2014

Motivation & research question

• Large scale agricultural investments (LSAI) have risks and potential benefits

• Risks are well documented through numerous qualitative case studies

• Little information on actual benefits of such investments

• Can there be any direct positive effects from LSAI on household welfare?

• How do different institutional arrangements affect the welfare outcome?

Case study setting

• SAGCOT (Southern Agricultural Growth Corridor of Tanzania):– Promotion of agri-business clusters along southern road/rail/electricity

network– Focus on inclusive business models– Concerns over land rights

Case study description

Kilombero Sugar Ltd. Kilombero Plantation Ltd.

Ownership of investment

Illovo (British/South African) with government

Agrica (British) with parastatal

History Government scheme in 1960s, privatized end of 1990s Started in 2008

Investment More than 10,000 ha estate, 2 sugar mills, 1 distillery More than 5,000 ha estate

Employment More than 5,000 mainly seasonal & casual workers Several hundred workers

Smallholder integration

Since privatization, strong growth of outgrower scheme: today more

than 10,000 farmers (mostly operate farmers below 1 ha)

~ 2,800 farmers involved in PPP smallholder

intensification project (SRI training, adoption HYV,, fertilizer, microcredit)

Methodology

• Household survey implemented in June 2013 of about 350 hhs • Comparing two potential direct transmission channels with a

control group• Sugar cane outgrower farmer

– Benefits of adopting a new crop, accessing high value markets

• Agro-industry wage labor– Benefits from off-farm income from casual & seasonal employment on

estate & factory

• Non participants– Main economic activities: maize and rice farming, few additional cash crops– Large share of income from off farm employment

• Agricultural & total household income to measure welfare• Descriptive statistics & propensity score matching

Descriptives (I): sugar cane gross margin analysis (preliminary)

 Sugar cane production

(N=129)

Rice production

(N=262)

Maize production

(N=107)

Gross revenue per acre (US$)901

(863)340

(273)228

(170)

Total costs per acre (US$)296

(293)102(98)

30(44)

Net revenue per acre (US$)603

(564)228

(214)189

(165)

Total net revenue (US$)1,788

(1,320)425

(475)219

(362)

Note: mean values, standard deviation in brackets

Descriptives (II): agro-industry wage income(preliminary)

Variables

Sugar survey (median values)

Rice survey(median values)

Agro-industry(N=60)

 Local agric. labor

(N=132)

Agro-industry (N=60)

Local agric. labor (N=74)

Number of months per year 10 4 7 3

Number of person days per year 243 56 162 31

Average income per day (US$) 4.2 4.1 2.9 3.3

Annual wage income (US$) 989 294 564 135

Matching results: simulated effects of participation (preliminary)

Agro-industry wage employment:

Sugar cane outgrower scheme:

    Obs Kernel matching %

Agricultural income per capita (log)Unmatched -0.11

ATT 155 -0.17 -19%

Household per capita income (log)Unmatched 0.81

ATT 160 0.72*** 93%***

Basic needs poverty 

Unmatched -0.44

ATT 160 -0.43*** -37%***

    Obs Kernel matching  %

Agricultural income per capita (log)Unmatched 1.18

ATT 173 1.24*** 215%***

Household per capita income (log)Unmatched 0.80

ATT 175 0.90*** 128%***

Basic needs poverty 

Unmatched -0.42

ATT 175 -0.47*** -40%***

Heterogeneity of simulated effects in outgrower schemes (preliminary)

  Sugar cane outgrower scheme

 Obs. Participation Not participation

Participation effect

Agricultural income per capita (log)Land poor (< 3 acre) 30 12.92 11.99 0.94***Land rich (>3 acres) 45 13.52 12.08 1.44***

Asset poor 32 12.99 12.02 0.97***Asset rich 43 13.50 12.07 1.44***

Household per capita income (log)Land poor (< 3 acre) 30 13.24 12.72 0.52***Land rich (>3 acres) 45 13.88 12.73 1.15***

Asset poor 32 13.36 12.75 0.61***Asset rich 43 13.82 12.71 1.11***

Conclusions

• Indications of positive effects from participating in outgrower schemes & agro-industry wage labor market on household income– But: Heterogeneity of effects in outgrower model benefits seem to

be larger for wealthier farmers

• Participation effects in the agro-industry labor market seem to be influenced by specific conditions:– E.g. capital intensity (low/high labor demand), length of season,…

• Still preliminary results:– More robustness checks needed– Study of additional outcome variables (assets, food consumption…)

• Limitations of this study:– Cross-sectional data and small sample size– Study has focused on two good practice cases (no village land case)

Thank you!