17th ICABR Conference“INNOVATION AND POLICY FOR THE BIOECONOMY”

Ravello - June 18-21, 2013

PRICE CHANGES, WELFARE EFFECTS ANDFOOD SECURITY IN RURAL ETHIOPIA

PASQUALE LUCIO SCANDIZZO1, ADRIANA PAOLANTONIO12, SARA SAVASTANO1

1University of Rome “Tor Vergata”2FAO

Motivations• The recent food prices spikes have increased the concerns on food

security, also in light of the global climate change, population and income growth, and the evolution of nutritional standards and habits.

• Food security is a facet of the wider issues of enhancing nutritional achievements and the performance of the food sector.

• While the impact of the recent crisis on the welfare and poverty of rural population remains controversial (Ataman Aksoy and Hoekman 2010; Barrett 2011; Swinnen 2011; Ivanic and Martin 2011), the impact on nutrition and food security are more limited.

• Lack of a consensus on the core household food security indicators that are needed in order to properly measure and monitor food security.

• Food risk and poverty are related by a double bind.

• Small farmers in developing countries undergo this twofold challenge under often difficult choice conditions and the stress of market forces.

• When commodity prices show the sharp increases experienced lately, several farmers appear to have abandoned the pattern of production characterizing their traditional subsistence status. Has that determined an increase in real well being?

• Alternatively, have more costly changes on the level of nutrition somewhat voided the apparent success of these farms as market agents?

Background

• We extend previous results obtained by Savastano, Weaver and Paolantonio (2013), who analyzed the extent and magnitude of the impact of food price crises on household welfare and agriculture productivity, to the wider issues of food security and nutritional intakes.

• We propose to analyze the effects of commodity price movements on revenues, nutrition and well being of Ethiopian farmers focusing on the role of the adaptation option of the household diet to the patterns of production and sales determined by changing prices, and on the asymmetries that they may determine in production and consumption patterns

• We use 4 rounds of panel survey on 1,239 Ethiopian rural households (ERHS conducted by CSAE and IFPRI) for the years 1994, 1999, 2004, and 2009, to understand differential impact of two different crisis: mid-nineties, and 2008 spike.

• We focus on net producers and net consumers of the main cereals (i.e. tef, maize, wheat, barley, sorghum and millet).

Objectives and Contributions

• According to the existing literature, the impact of a price spike on the rural farmers depends on their net position (i.e. net producer/net consumer). An increase in price results in consumers’ losses and producers’ gains.

• Since Deaton (1989; 1989a; 1997), the Net Benefit Ratio (NBR) approach has been widely applied to analyze the effect of a change in food prices on household welfare (Budd 1993; Barrett and Dorosh 1996; Minot and Goletti 2000; Seshan and Umali-Deininger 2006; Levinsohn and McMillan 2007; Ivanic and Martin 2008; Wodon and Zaman 2008; Zezza et al. 2008; Simler 2010; Karfakis et al. 2011). The impact of price spikes is considered in monetary term.

• However, consumer prices differ from producer prices and nutritional content of own production differs from market purchased food: thus net producers and net consumers distinction may not be decisive.

Some Stylized Facts

Evolution of Poverty and Food Deficiency

• We adopt the NBR metric and define a household as a net producer (consumer) of cereals if it is positive (negative).

• We use the “before-response” concept of the price effect applied to prices of calories and not prices of crops.

• Own consumption and market consumption have different compositions in terms of nutrients and, in particular, calories.

Approach and Hypotheses

• Households first make production and investment plans to maximize the present value of their expected wealth and then choose consumption levels by maximizing utility, given their wealth maximizing production level.

• Households choose the share of own consumption to maximize utility, given consumer and producer prices.

• They also maximize utility by choosing consumption levels conditioned on such a share and their other wealth maximizing choices.

Approach and Hypotheses (cont’d)

• Consider the consumption problem: 

• Solving under the hypothesis of a Cobb Douglas utility implies:

The Model

(1)

(2)

(3)

• Definition:

• Optimum share:

• Optimum consumption:

The Optimal Solution

(4)

(5)

(6)

The Welfare Effect

(8)

(7)

• When there is a price shock, the change in welfare, or change in real income (i.e. change in utility in money-equivalent) has two distinct effects:

1. An increase in the producer price for calories has a positive welfare effect if the household is a net seller of the calories produced and a negative welfare effect if it is a net buyer.

2. Conversely, an increase in the consumer prices of calories has always a negative effect on real income, which is larger the larger the share of consumption that comes from the market.

Kcalorie Impact ofFood Price Changes

• Notice that this is a different result from saying that the effect of a price change is positive if the household is a net seller and negative if it is a net buyer.

• In reality, even if it is a net seller, the effect on real income may be negative, if the increase in consumer prices is sufficiently larger than the increase in producer price.

• Vice versa, the household may be a net buyer and yet the effect of a price increase may be positive, if the difference between the increase in producer and consumer prices is sufficiently large, as shown by equation:

Kcalorie Impactof Food Price Changes (cont’d)

(8)

Estimation Strategy

• Estimate first the demand equation for own consumption share.• Obtain predicted value of the share.• Use this predicted value as a regressor in a second stage estimation of

total demand.• Model predicts negative relation of share with consumer-producer

price ratio and negative relation of consumption with share weighted index of producer and consumer price, as suggested by equations:

and:

(5)

(6)

Results from First Stage Estimation

t statistics in parentheses p<0.05 ** p<0.01 *** p<0.001aPanel nested binomial with logit link function and RE (Papke and Wooldridge, 2008)

Results from Second Stage Estimation

t statistics in parentheses p<0.05 ** p<0.01 *** p<0.001

Quintiles of cereals

consumption per AE  

DIRECT WELFARE EFFECT (as % of total HH real income)Two prices

1994/1999 1999/2004 2004/2009

Net Consumers

Net Producers

Net Consumers

Net Producers

Net Consumers

Net Producers

1 -68.2 -52.4 8.4 12 8.6 28.42 -54.5 -28 13.7 21.4 17.5 353 -54.5 -11.6 19.5 32.5 22.9 32.84 -44.3 - 23 28.7 2 35.35 -44.3 - 22.7 27.4 25 37.9

Average -53.2 -30.7 17.4 24.4 19.8 33.9

Quintiles of cereals

consumption per AE in

caloric terms 

DIRECT WELFARE EFFECT (as % of total HH real income)IN CALORIC TERMS

Two prices1994/1999 1999/2004 2004/2009

Net Consumers

Net Producers

Net Consumers

Net Producers

Net Consumers

Net Producers

1 -3.3 -1.5 2.2 3.7 -2.3 1.52 -9.3 0 3.9 4.7 -1.7 1.53 -13.2 -0.3 5.2 14.2 -0.3 13.54 -9.2 -0.2 5.6 7.6 -1.4 1.15 -12.2 -0.4 9.2 46.7 -0.6 4.1

Average -9.4 -0.5 5.2 15.4 -1.3 4.3

Comparison of Welfare Effectsusing Two Prices

Comparison of Welfare Effectsusing a Single Price

Quintiles of cereals

consumption per AE

DIRECT WELFARE EFFECT (as % of total HH real income)Single prices

1994/1999 1999/2004 2004/2009

Net Consumers

Net Producers

Net Consumers

Net Producers

Net Consumers

Net Producers

1 -31.3 -16.3 11.4 16.2 -4.9 -1.92 -23.3 3 17.4 25.9 -4.7 -1.73 -24.6 9.9 25.4 44.3 -4.5 -3.44 -13.1 - 29.6 43 -4.9 -0.75 -12.1 - 29.5 39.6 -3.7 -1.6

Average -20.8 -1.1 22.7 33.8 -4.5 -1.9

Quintiles of cereals

consumption per AE in

caloric terms

DIRECT WELFARE EFFECT (as % of total HH real income)IN CALORIC TERMS

Single prices1994/1999 1999/2004 2004/2009

Net Consumers

Net Producers

Net Consumers

Net Producers

Net Consumers

Net Producers

1 -1.8 -1 1.9 1.9 -1.5 1.12 -6.4 0.4 2.3 -0.7 -1.8 1.13 -6 0.1 0.8 0.4 -1 5.74 -6.9 0.9 3.7 -4.6 -0.9 1.25 -5.8 -0.7 6.1 28.6 0 3.5

Average -5.4 -0.1 2.9 5.1 -1 2.5

Conclusions

• Allocation of an optimal share of production to own consumption is significantly affected by the consumer- producer price ratio.

• Total consumption instead depends on a weighted average of both prices.

• Once this mechanism is taken into account and calories rather than cereals are considered, the welfare effects of even large price changes are relatively small.

• Thus households appear to be able to maintain a reasonably constant energy intake even under adverse price changes.