C AN LARGE - SCALE AGRO - INVESTMENTS SERVE AS AN ENGINE FOR - PowerPoint PPT Presentation

C AN LARGE - SCALE AGRO - INVESTMENTS SERVE AS AN ENGINE FOR INCLUSIVE GROWTH ? E MPIRICAL EVIDENCE FROM U GANDA & E THIOPIA Paper presented at UNU-WIDER conference on Inclusive Growth in Africa Helsinki, Finnland In country partner: Philipp Baumgartner PhD student | Junior Researcher Center for Development Research pbaumgartner@uni-bonn.de | www.zef.de Ethiopian Economics Makerere University, Association Kampala Sept. 21, 2013 www.eeaecon.org Faculty of Agric. Economics

Phenomenon and problem statement Number of LSLAs globally: Investing countries Source : The Land Matrix , accessed Sept. 9 -2013: Number of deals beyond 200 ha, since 2000. (orange – target country, blue – investing country) Note : Data derived from media reports and validated through various experts and local NGOs, 2 Government officials etc. Accuracy improved since 2010, but stilly approximation

Phenomenon and problem statement Number of LSLAs globally: Target countries Source : The Land Matrix , accessed Sept. 9 -2013: Number of deals beyond 200 ha, since 2000. (orange – target country, blue – investing country) Note : Data derived from media reports and validated through various experts and local NGOs, 3 Government officials etc. Accuracy improved since 2010, but stilly approximation

Phenomenon and problem statement Problem statement Potential risk and opportunities: • Risks: “land grab”, unsustainable resource use, exploitation of labour, etc. • Opportunities: employment generation, market access, improve infrastructure, etc.  Question: Can LSLAs serve as engines for inclusive growth? What do I mean by inclusive ? • pro-poor (reducing poverty AT LEAST proportionally among the poor) In line with Erik Thorbecke (2013) To answer this have to understand how impact on local population’s livelihood situation (conceptual framework) • Impact not direct, but occurs across a number of impact channels 4

Conceptual & Analytical Framework Conceptual framing of the five impact channels • value of land (prices) • transfer of land (transactions) Land Factors of production • access and use (relative importance) Main impact channels • on- and off-farm employment • access to jobs (who gets them) Labour Local population • wage levels • access and use (de facto) Natural • value and price of forest products • who uses when (relative importance) Resources • introduction of new techology Technology & • organisation of production Organisation • diffusion and adoption • property rights (structure/ regime) Institutions & • rules for and governance of transactions Markets • emergence & functioning of markets 5

Analytical approach & Data sources Analytical approach to impact evaluation of LSLAs Biggest challenge for any impact evaluation: attribution problem Two broad categories of impact evaluation (Hemmer 2011, Khandker 2010) • Counterfactual impact evaluation (CIE):  if there was change • Theory-based impact evaluation (TBIE):  how or why there was change Combining ex-ante and ex-post analysis • Early stage project: mathematical optimization ( ex-ante ) (Hazell & Norton 1986) • Older project: analytical narrative ( ex-post ) (Moore 1966, Rodrik 2003) Data sources & mixed methods • Qual’ data: expert interviews, group discussions, + in Uganda semi-structure biographical interviews • Quant’ data: community survey, household survey 6

Case study context Case A: Tilda – Bugiri district, Uganda Local Context: - Eastern Uganda, located at transit corridor to Kenya - multi-ethnic setting - relative poor area within Uganda - small trading points & shops along main road Smallholder fields Investment: to Kampala - 1.200 ha investment (3.900 170km ha including catchment) - irrigated rice ( basmati ) Investment - 4-5h drive to Kampala site to Kenya History of Investment 30km - 1968-88: Kibimba Rice Scheme (Chinese Dev. Proj.) 7 - 1989 -96: Kibimba Rice Company ( State-operated ) - since 1997: Tilda Rice UG Ltd. (UK/Indian investor)

Case study context Case B: Saudi Star - Abobo, Gambella Local Context: - ca. 8.000 people / 1.600 HHs -2 Ethnic groups: Anyuak (ind) & to Highlander (settlers - Derg) Addis - little market integration Affected area 770km Investment: Evolution of Investment (simulation) - Remote area within Ethiopia - 100.000 ha land “affected” ( dark green ) - Saudi-Ethiopian, with Pakistani - 10.000 ha converted by LSLAs ( light green ) farm management & construction - local pop: 25-30 km radius ( settlements – team blue/ green dots ) -10.000 ha investment - irrigated rice (basmati) - Started in 2008/09 8 Investment site

Findings Uganda case Today’s situation in Bugiri, Uganda: smallholder operated wetlands “Kibimba started in 1971 and farmers who went there for employment, acquired skills to cultivate rice, and currently no wetland idle.” (Older Farmer, Igogo Village) 9

Findings: Uganda case 4 drivers of conversion of wetland to rice fields Point of departure: until mid-20 th century few fields had individualized rights, and wetlands only used for grazing and cultivation during drought (little value) 1 st driver: Pioneers had acquired skills from Kibimba and started growing at own fields. Opened wetland close to their own land. (1970s-1980s) 2 nd driver: Restructuring at the farm (leaving of Chinese and privatization) cause laying-off of worker. They apply their skills on remaining wetland (1988-92 & 1997-2000) In addition : Population increases significantly & relative prices of cash crops change (early 2000s) Pull factors Push factors Related to 1) Training in skills & demand for 2) laying-off  lack of source investment output of income 3) Change in relative price of cash 4) Population growth  land External driver crops scarcity  “Today all land is taken, but expansion is possible through rental markets.” (First generation rice growing farmer, Buwuni village) 10

Findings: Uganda case Source of knowledge about growing rice by point in time started to grow (farmers growing rice in 2010/11 season) Year started Source of knowledge on growing rice Total growing rice (frequencies of total group, in %) Working Working Parents Neighbours Extension Other N % at KRS at Tilda 1st Generation - 33.3 33.3 - - 33.3 3 100 (before 1988) 2nd Generation 12.5 50.0 12.5 - - 25.0 8 100 (1988-1997) 3rd Generation-a 42.9 50.0 - - 7.14 - 14 100 (1998-2003) 3rd Generation-b 50.0 26.9 11.5 3.9 3.9 3.9 26 100 (2004-2011) Total 39.2 37.3 9.8 2.0 4.0 7.8 51 100 Source: HH-Survey (2011), N=170  Friends and family main source of knowledge  However, interviews often revealed peers’ relation to Kibimba/Tilda 11

Findings Uganda case Price trend for wetland in selected villages (1990-2013) based on community survey and recall questions (biographic interviews)  Price increased over past 20 years (nominal price even Note : Prices were deflated and standardized for 2010 prices (World Bank, 2013); steeper) : Today: 120 USD/ acre per season Exchange rate from Jan 2011 (oanda, 2011).  Land for expansion only through rental or farer away 12

Findings Ethiopia case Future situation in Ethiopia: Large-scale operated rice scheme (Picture: construction work & test fields; early 2011) 13

Set up of the model “All models are wrong, but some are useful.” (Box & Draper 1987, p.424) Concept: each group = 1 representative large- farm Livelihood activities 1. Agriculture using hand tools 2. Agriculture using draught animals 3. Land clearing /preparation for cultivation 4. Hunting * (*only indigenous) 5. Gathering of wild fruits, roots, and fuel wood 6. Self-employment : e.g. beer brewing or small businesses 7. Off-farm employment paid in cash on a monthly or daily basis Daily labour Boy with firewood Maize field Handmade crafts Antilope / game meat (weeding for 9Birr/day) 14 Late -2010

Set up of the model Required resources and constraints Endowments/ Inputs Market constraints 1. agricultural land ; 1. Market constraint : limited demand for locally/ self-produced services & 2. open access land / forest goods (isolation) 3. labour during peak season 2. Labour market constraint : Limited 4. labour during off-peak season jobs 5. draught animal (Ox) 6. cash and assets Open access land – similar to Example of indigenous the one cleared for homestay investment Rice field on the nursery – for seed multiplication 15

Simulations Ethiopia case Possible future impacts: four scenarios Base-run: Situation prior to investment’s arrival Showing mix of income strategies and initial levels of income • 1 st Scenario: Forest cleared Taking away 10.000 ha of prior open access land (forest/savannah) • 2 nd Scenario: Evolution of big investment (10.000 ha size) Jobs created and partly taken up by locals (ca. 1/3) • Increasing demand for locally produced goods/services • 3 rd Scenario: Smaller investment + inclusive rural development plan ‘Only’ 5.000 ha investment size (with same effects on employment + demand) • PLUS: Public investment in infrastructure • PLUS: Investment in extension services + improved inputs + availability of • draught animals/ tractor service 16