Rural Investment and Policy Analysis

Brief model summary

The RIAPA model, developed by IFPRI, integrates survey-based microsimulation with a core Computable General Equilibrium (CGE) model to analyze the impacts of economywide shocks on households and individuals. It features indicators for poverty, diet quality, and women’s inclusion and has expanded to over 30 countries. RIAPA is used for ex ante impact analysis of various socioeconomic policies and investments in agriculture, evaluating their implications for growth and policy prioritization. This model has been instrumental in assessing the economic costs of external shocks like COVID-19, climate change and assessing the implications of agricultural investments and policies across different nations

Input data

The RIAPA model framework relies on a Social Accounting Matrix (SAM) as its primary data source. A SAM is a comprehensive national accounting database that captures the circular flow of receipts and payments associated with economic transactions between agents and markets over a financial year. This information is organized in a square matrix, where each cell represents a transaction between a row account (receipts) and a column account (payments).

IFPRI’s Nexus SAMs adhere to standardized data practices and classification systems, promoting transparency and consistency across countries. A typical Nexus SAM includes 90 economic sectors, with nearly half dedicated to primary agriculture, agri-food processing, and food services. This makes RIAPA distinct in its ability to model national agri-food systems. IFPRI provides and maintains publicly accessible SAMs for approximately 30 countries in Africa, Asia, and Latin America.

Outputs

CGE models are versatile tools for conducting ex ante impact analysis across a wide range of scenarios. IFPRI researchers have used CGE models to examine impacts of on-farm investments such as irrigation, input subsidies, climate smart technologies, as well as off-farm investments, including grain stocks investments and new value chains such as biofuels.

A key focus of RIAPA-based analysis has been assessing the implications of growth in specific agricultural sectors such as teff in Ethiopia, the identification of priority value agricultural value chains, for example in Senegal, or cross-country comparisons of agricultural growth-poverty linkages. The RIAPA framework is also ideally suited to prioritizing among agricultural policies and investments, with studies recently concluded in Rwanda, Mozambique, Ghana, and Uganda.

Methodology overview

The RIAPA data and modeling system is IFPRI’s primary tool for forward-looking, economywide country-level analysis, serving as a simulation laboratory for experimenting with policies, investments, or economic shocks. IFPRI’s static and dynamic economywide models incorporate flexible behavioral features, such as nested production functions, imperfect substitution of imported commodities, and linear expenditure systems of consumer demand. Consumers and producers maximize utility and profits based on factor and product prices, which adjust endogenously to establish market equilibrium. In the dynamic version, population growth and urbanization are set exogenously, affecting labor supplies, while sectoral capital accumulation is endogenously determined based on past investments.

Software requirements

IFPRI’s static and dynamic economywide models are coded in a general algebraic modeling language developed by the GAMS Development Corp. GAMS is a high-level modeling system for mathematical programming and optimization. The modeling system consists of a language compiler and a range of solvers. Running IFPRI’s economywide models requires downloading the GAMS compiler (the latest release is 48.3.0). A GAMS/BASE license is required to run GAMS, while IFPRI’s economywide models additionally require a GAMS/PATH solver license. Users can email sales@gams.com to obtain more information about pricing of different compiler/solver license packages and free trial versions.”

Accessibility

User documentation

Code

Code repository (coming soon):

Data repository:

The RIAPA model framework uses the following data

External model website

RIAPA Data and Modeling System | IFPRI

Acknowledgements

In addition to the support under the Foresight Initiative, the RIAPA data and modeling system was developed over many years with bilateral support from several donors, including USAID (BRFS), BMGF, and IFAD.