Agricultural Growth and Poverty Reduction: A Cross Country Synthesis Roles of Agriculture Project International Conference.

Agricultural Growth and Poverty Reduction:

A Cross Country Synthesis

Roles of Agriculture Project

International Conference

Outline of the presentation

Motivations, objectives Review of the theoretical background Implementation Summarizing the results by channels Elements of a synthesis Conclusions and what next

Why a poverty module?......

Poverty reduction at core of international debate on national development strategies and development aid

Declining trend in share of resources devoted to agriculture at the national and international level

Q: Is agriculture important in reducing poverty? If yes, why and how much?

Q: Should we care? Markets, social goals, and the role of economists

… and what should we get out of it?

Which channels are most critical in translating agricultural growth into less poverty?

Identifying constraints: the role of policies, institutions, and natural factors in molding the link between agricultural growth and poverty reduction.

Derive policy implications

Theoretical background in few slides Ramon Lopez, 2002

I. Wages, employment, general equilibrium effect

Agriculture uses intensively unskilled labor; A high share of unskilled labor works in agriculture; Agricultural wage is opportunity cost for (unskilled)

labor sector-wide; Agricultural growth raises wages and/or employment; Interact labor demand and supply elasticities, L/K; Translate labor market effects into poverty changes.

II. Real food Prices

High share of food consumption Real food prices influence poverty line

computation Ideally, separate tradable and non-tradable in

food basket Role of real exchange rate and trade

openness Other factors: non-agricultural output, non

food prices.

III. Smallholders’ income

Direct effect: raising farm income, household income

Key issue 1: share of farm income into total household income

Key issue 2: the income gap for the average smallholder might be too great

Access to markets and the impact of agricultural productivity on farm revenues

IV. Growth Multipliers

Conditions for relevance (Delgado et al. 1998) Agriculture occupies a large share of

aggregate employment Ag growth benefits should be wide-spread Consumption patterns, high MBS of non-

tradables Elastic supply of underused local resourcesThen: tradables income shock regional

growth spurred by demand for non-tradablesQuestion: how do farm and non-farm tradables

multipliers compare?

V. Composition of growth and poverty reduction

Ravallion and Datt (1996) India: Agricultural, services growth lead in poverty

reduction Cross sectoral spill-over (rural into urban) Elasticity of agricultural growth increases from

HCR to PG to SPG, i.e. growth reaches poorest of the poor

Implementation

Data …..but also country’s specific conditions

matter I guess (a help in final synthesis)

I. Wage & employment channel

Chile, Mexico: estimates of labour demand f(Ws,Wu,r,Qa,Qn)

Definition issue: (1) agriculture (primary+ agro-industry); (2) skilled labour

Chile: we saw before: 4.5% Ag growth (ext. definition) leads to –5.5%/-8.6% in HCR (compensated case)

Mexico (as Chile): Agriculture (strictly), higher labour intensity similar results to Chile, weaker due to different sector boundaries

Dominican Republic: Data show that higher Ag. Y/L did not lead to higher Wu: Labour supply very elastic due to Haitian immigration

I. Wage & employment channel: tentative conclusions

Labour supply elasticity (DR, Chile): inter-industry labour mobility, labour market rigidities, regulations can limit the impact of Ag.growth on poverty

Output mix: Policy Reforms, commodity booms, Tech. Change output mix, farming systems

Labour supply elasticitywage, employment Wage effectHCR …. but employment effectPG

II. Food Price Channel: Key results

Cointegration analysis: Chile (NT Food Prices): Ag GDP Elasticity of

FPI is –0.8 in short run; -0.18 in long run Mexico (FPI): very small elasticity, RER

dominates DR: elasticity -0.43 (LR), -0-80 (SR) Ghana (no cointegration analysis). Ag Labour

Productivity Elasticity of urban food prices –1.0 (1 Year) and –0.89 (2 Years)

II. Food Price Channel: Key results (cont.)

Studies exploring role of infrastructure: South Africa: Pass-through to farm gate prices

(e.g. wheat and maize) substantial But farm gate to retail disconnected: thus little

role for ag growth to reduce food prices in urban areas. Infrastructure bottleneck.

China: 1% Ag. Growth -0.1% food retail prices. Investment raising productivity of food marketing (farm gate to retail) affects food retail prices

II. Food price channel:tentative conclusions

In many countries food prices are critical in moving large shares of population in and out of poverty (e.g. results from Mali, DR, Morocco, India)

Including small poor farmers (e.g. Ethiopia, DR) Integration into world markets weakens relationship

b/w food prices and Ag. Growth Marketing infrastructure might be as well critical as

ag growth Issue (DR) of broad price stability,food prices, and

poverty: does efficiency of food system affect economy wide inflation, hence wage dynamics?

III. Growth Multipliers: Key Findings

Chile: use of CGE to compare alternative definitions of multiplier (Hirschman, HD, CGE-flex price)

1. Except livestock (2.6), multipliers of ag sub-sectors become 2nd order under flex-prices (0.71-1.84)

2. Difference b/w CGE and HD small for traditional ag sectors (except fruits). Higher for agro-industry. Question: is scepticism (e.g. Delgado et al, 1998) over ag growth multipliers warranted in case of LAC?

III. Growth Multipliers: Key Findings (cont.)

Ghana: multipliers by income; farm and non-farm tradables multipliers compared

1. Poorest show highest MBS for non-tradables

2. Multipliers are higher for Non Farm Tradables shocks than for Farm Tradables (3.2 vs 2.5)…

3. …but FNT more important than NFNT in overall size of multipliers

DR: suggests that NFNT are more important than FNT in size of multipliers.

1. 1% increase in income of 40% poorest households 1.1% increase in non food items

III. Growth Multipliers: Key Findings (cont.)

South Africa: Communal areas in Kwa-Zulu Natal

1. MBS are high for consumer durables fruits, meat and vegetables, not for staple crops (e.g. maize); NB: impact on nutrition

2. Growth multipliers smaller than expected (1.28 – if discounted by Delgado’s 30% almost no impact)

3. FNT contribute less than NFNT to multiplier

4. Linkages are stronger in case of export industry of agricultural processed and non products (e.g. chemical, transport, wood and paper)

III. Growth Multipliers: Tentative conclusions

Results appear consistent with literature. In particular, multipliers are weaker when:

1. Few underemployed resources in rural areas2. MBS for NT are low, especially FNT3. Non tradables are small share of hh budget

Are production linkages more important than consumption linkages depending on trade openness? See case of Chile’s linkage analysis

IV. Small-holders’ income channel

South Africa: while agriculture provides buffer against poverty, it is a means of accumulating wealth only when it is done seriously (e.g. a minimum level of commercialisation)

Ethiopia: Average farm size is 1.02 (no irrigation) Farm income growth would have great impact on poverty

but: …agriculture is key because it employs 85% of labour. …are there conditions for sustained growth? Demographic pressure cum land policy: maybe there

more role for urban growth while reforming land policy

IV. Smallholders’ income channel (cont.)

China: The constrained importance of ag growth Rural areas lack infrastructure Little investment in rural education Institutional bias against out-migration Rural poverty->Gov’t investment, but no impact More investment in rural public goods would

reduce importance of ag growth for poverty reduction

Chile, Mexico: small farmers’ Y does not affect poverty much

IV. Small-holders’ income channel (cont.)

India: agricultural growth (per capita income or yield) is important in reducing poverty, but increasing land/man ratio has greatest impact.

IV. Small-holders’ income channel: Tentative Conclusions

To lift small holders out of Pov Ag growth may be important but not enough

Where unequal land distribution, ag Y share could be lower for smallholders (Mexico, Chile): agriculture competes with non-ag income sources for limited assets

Incomplete or missing land markets may constrain small holders, e.g. land: but…..

Land reform is expensive if done seriously, and land rental may not increase productivity by much

Where land is equally distributed (Mali, Ghana, Ethiopia) ag growth has strong poverty reduction potential

V. Growth Composition

Indonesia. We saw: Elasticity for total poverty is at -1.86

Mexico. Total food poverty elast:-1.32 (-0.88) Total moderate poverty elast: -0.56 (-0.48) No cross-sectoral spill-over Rural growth reaches the poorest of the poor

more than urban growth (inequality reducing)

V. Growth Composition (cont.)

China: Ag growth reduces rural poverty significantly No effect of services,industries: data quality,

maybe consistency with smallholders story

Preliminary considerations

Agriculture is important in reducing both rural and national poverty,

..but effect on urban poverty might fade as agriculture becomes minor industry

Constraints limit some channels and enhance others:

were land is more unequally distributed wage and employment channel might be important depending on labour supply elast.; direct effect is smaller

Preliminary considerations (cont.)

Where land is more equally distributed, direct effect is important, wage channel is weaker

Food price channel is constrained by infrastructure bottlenecks, trade policy

Trade policy and deregulation change output mix and farming systems, production linkages become more important than consumption linkages within multipliers

Elements for further discussion

Land administration reform and land ownership reforms: how would channels be affected (e.g. shadow price of family labour in rural areas, consumption patterns)

Labour market policies: is there scope for analysing functioning of rural labour markets?

Trade policies, more broadly policies affecting incentives:

1. how do these transform farming systems in the long run 2. How do forward and backward linkages evolve in the

process?3. how do the roles of agriculture change with that? How important are food prices in affecting inflation and

labour market dynamics?

Q: Should agricultural policies incorporate a poverty reduction concern/target or should poverty reduction be left to social welfare policies?