The foundation of SISMod analysis is a series of modules, taking into account household incomes and expenditure and estimating demand/supply/price transmission elasticities based on data from household surveys, national food price collection systems, and other assessments. The process determines the interaction between production and income-generation decisions (income effects) and consumption decisions (price effects), which quantify the impacts of price changes and income changes on household food consumption.
SISMod-Light adopts the Agricultural Household Model approach developed by Singh et al. (1986). In this model, household consumption decisions are based on household income, which comprises agricultural profits as well as wages, remittances and any other type of income. Income generation and the allocation of income to expenditure are based on separable decisions, which maximize income and utility in a two-step process. The Agricultural Household Model incorporates both household production and consumption. It integrates price effects – which are presumed to be exogenous – and takes interactions between them into account.
Unlike in pure consumer models, in the Agricultural Household Model the household budget is not fixed but endogenous and depends on production decisions that contribute to income through agricultural profits. This implies that the additional effects of agricultural profits, which can be simultaneously positive and negative, are added to the standard Slutsky equation (Taylor, 2003). For example, an increase in the price of staples unleashes to two opposing forces: the traditional price effects (where household food demand decreases when the price rises), and the opposing effect of agricultural profits (which, by contributing to household total income, lift budget constraints thereby increasing household food demand). Therefore, a change in price of a given commodity affects both supply and demand decisions.
The key concept of the model is shown in the following diagram showing the framework:
Two-Stage Food Demand System
1st Stage – Linear Expenditure System (LES)
The first stage of the demand system models consumption decisions by allocating total expenditure to broad commodity groups. For example, the broad commodity groups include food, clothing, utilities, housing, transport, communication, and personal and health care. A Seemingly Unrelated Regression is used to estimate a Linear Expenditure System (LES) for the first stage of the demand system. LES is a widely used traditional approach functional form derived from maximization of the utility function subject to the expenditure restriction. It was derived by Stone (1954) and expressed as:
2nd Stage – Linear Almost Ideal Demand System (LAIDS)
In the second stage, a Linear Almost Ideal Demand System (LAIDS) (Deaton, 1986) is estimated for the food groups: rice, cassava, other cereals, pulses, vegetables and fruits, meat, eggs and dairy products, sugar and oils, and other food. LAIDS models consumption decisions by allocating food expenditure to food groups.
Per Capita Dietary Energy Consumption (DEC)
Per Capita Dietary Energy Consumption (DEC) is the amount of food, in kilocalorie (kcal) per day, for each individual in the total population (FAO, 2008). DEC is converted from food consumption in quantities, which derived from the 2nd stage demand system, by using energy conversion factors. It would be used to estimate food security indicators in the model.
Key Outputs of SISMod
Proportion of food-energy-deficient population
dietary energy consumption below an acceptable threshold;
Depth of hunger measured in kcal/person/day
deficit in absolute terms between the average dietary energy consumption and the dietary energy threshold;
Food gap measured in kg/person/month
the cereal equivalent of the depth of hunger;
Total food assistance needed to meet the needs
measured in metric tons/year.
SISMod outputs for different population groups can be visualized as charts or tables in the SISMod Tool.