Enabling Growth through Agricultural Transformation Economic Growth Forum l Kampala, Uganda| September 14-15, 2017 Holger A. Kray Head | Africa Agriculture Policy Unit hkray@worldbank.org

The weakest point in agricultural transformation is policy, not technology or other enablers.

Agriculture has Unique Importance for Uganda Agriculture contributes substantially to household incomes Source: Nagler & Naudé (2017), Food Policy Stunting affects productivity, lifetime earnings, and economic growth Sectoral annual GDP growth rates are volatile and generally low recently Source: UNHS 2005/06-2012/13. Source: WB 2016, Uganda Poverty Assessment Poverty reduction between 2005/06 and 2012/13 Smallholder agriculture has been crucial for poverty reduction in Uganda Growth AG generates 23% of GDP and over 40% of Uganda’s exports ~2% sector growth p.a. [4% SSA] below population growth of 3.2% [2.7% SSA] Poverty In 2006-2013, Uganda experienced the fastest reduction in extreme poverty in SSA Agricultural households accounted for 79% of poverty reduction Food Despite rapid urbanization, 78% of population still live in rural areas 27% of the population is stunted Given its moderate climate and abundant water, Uganda should have a comparative advantage McMillan, Rodrik, Sepulveda (2017) show that growth through structural transformation less effective than through sector productivity growth Over 78% of the population is below 30 years Around 2/3 of total unemployed persons are 15-29 (Youth) Underemployment and seasonality of agriculture is a persistent cause of rural poverty and low incomes Low labour productivity per year in AG compared with non-AG, but labour productivity per hour worked is not very different  high seasonality in AG Seasonality larger for women with lower agri-business and wage work opportunities Jobs AG employs 72% of the workforce 300,000 Ugandans join the labour market every year

If the poor are in agriculture, is this due to selection? NO Structural transformations (ST) away from agriculture across Africa have not been accompanied by growth in per capita income Source: De Janvry 2017

Most poverty reduction has been achieved within agriculture and rural areas …. not through ST toward urban environment Uganda (2005-09): 70% of poverty reduction in baseline ag population achieved in agriculture Tanzania (1991-2010): 85% of rural poverty reduction achieved in agriculture (34%), in RNFE (25%), and in local towns (30%) Cross-country: Growth through structural transformation is less effective than through sector productivity growth. Hence, income growth in agriculture and rural areas (RNFE, local towns) key for rural poverty reduction Source: De Janvry 2017; Christiaensen et al 2017; McMillan, Rodrik, Sepulveda 2017

Low annual labor productivity the root cause of rural poverty Source: De Janvry 2017; McCullough, 2017 Labor productivity per person/year low in ag compared to non-Ag But labor productivity per person/hour is not very different in ag than in non-ag In non-agriculture: 1850 hours/year (7h/day) In agriculture: 700/year (2.7h/day)

What explains low labor productivity in agriculture What explains low labor productivity in agriculture? Erratic/Spotty labor calendars! Average hours worked per worker per year: non-ag: 1850 h/yr (7h/day); ag: 700/year (2.7h/day) Ag labor calendars are seasonal for rural populations, with insufficient opportunities of access to employment in non-ag business and wage work to smooth out labor calendars Seasonality of labor calendars even larger for women, with much lower Non-Ag business and wage work opportunities Source: De Janvry 2017; McCullough, 2017

What explains low labor productivity in agriculture What explains low labor productivity in agriculture? Erratic/Spotty labor calendars! Compared to urban, rural households in Uganda have more erratic work calendars (hours worked relative to base month) and work on average 22% less For rural households, high seasonality in agriculture, with insufficient off-farm employment opportunities to smooth out labor calendars Urban households work more and have access to some countercyclicality, resulting in smoother labor calendars (and less poverty) Source: De Janvry 2017; McCullough, 2017

Implications for Policy Making Gains in agriculture land/labor productivity key for poverty reduction. “Green Revolution” (land productivity growth in staple foods) necessary but not sufficient. Can help increase the productivity of work in agriculture, but does not directly address the low annual labor productivity issue. First-order of importance to reduce rural poverty is smoothing out labor calendars across the year. This requires an Agricultural Transformation (AT) and a Rural Transformation (RT) beyond a Green Revolution: AT: smooth out labor calendars through the diversification and intensification of agricultural production systems RT: smooth out labor calendars through complementary employment in an emerging rural non-farm economy Rural Transformation • RT to smooth out labor calendars requires: i. Local endogenous ADLI (Agriculture Development-Led Industrialization) and ADLS (Services), where income growth in Ag creates effective demand for non-tradables ii. Decentralization of economic activity to rural areas iii. Population relocation away from areas with low agricultural potential and excessively remote from markets iv. Helping labor migrate (reach other labor markets) during the lean season v. Social protection to help rural households take risks in engaging in AT/RT vi. Observe a strong cross-country correlation between greater participation in RNFE and less rural poverty (Davis, LSMS-ISA data) Source: De Janvry, 2017; Poverty Status Report, 2014

Steering ag transformation requires the right level of incentives Uganda: Public Expenditure on Agriculture Uganda’s agricultural public expenditure trends and levels have been extensively studied thanks to relative ease of access and quality of the data. In addition to several “traditional” agricultural public expenditure reviews undertaken by international or local institutions (e.g., OPM, 2007; EPRC, 2009; World Bank, 2010; FAO, 2014), Uganda was recently selected as a pilot country for testing the new African Union-NEPAD Guidance Note to track government expenditures for agriculture (see AU/NEPAD, 2014; Kakuba, 2016). Several researchers have also analyzed returns to agricultural spending in Uganda, either in an ex-post (e.g., Fan and Zhang 2008; Benin et al. 2011) or ex-ante fashion (e.g., Benin et al., 2012; Pauw and Thurlow, 2015). Source: MAFAP 2017

Steering ag transformation requires the right direction of incentives Rice Coffee Uganda’s agricultural public expenditure trends and levels have been extensively studied thanks to relative ease of access and quality of the data. In addition to several “traditional” agricultural public expenditure reviews undertaken by international or local institutions (e.g., OPM, 2007; EPRC, 2009; World Bank, 2010; FAO, 2014), Uganda was recently selected as a pilot country for testing the new African Union-NEPAD Guidance Note to track government expenditures for agriculture (see AU/NEPAD, 2014; Kakuba, 2016). Several researchers have also analyzed returns to agricultural spending in Uganda, either in an ex-post (e.g., Fan and Zhang 2008; Benin et al. 2011) or ex-ante fashion (e.g., Benin et al., 2012; Pauw and Thurlow, 2015). Source: MAFAP 2017

The weakest point in agricultural transformation is policy, not technology or other enablers.

Enabling Growth through Agricultural Transformation Economic Growth Forum l Kampala, Uganda| September 14-15, 2017 Holger A. Kray Head | Africa Agriculture Policy Unit hkray@worldbank.org