International Monetary Fund

Search
Please send us your feedback

Macro Research for Development: An IMF-DFID Collaboration

Topic 4. Macroeconomic Policies and Income Distribution

Last Updated: December 05, 2014

The analysis of the distributional impact of macroeconomic policy choices has been a goal of policy-makers for decades.

Most poverty and social impact policy analysis has relied on computable general equilibrium (CGE) models, carefully calibrated to survey data and used to study detailed distributional impact assessments of changes in taxes, tariffs, or subsidies. The approach is generally one of "comparative statics" and abstracts from dynamic issues in its assessments.

For macroeconomic policy analysis, issues of dynamics are of course central. CGE analyses with dynamic elements and a detailed sectoral breakdown have been done. However, these tend to be based on ad-hoc dynamics and expectations. Paradoxically, these models are at the same time extremely complex and thus not readily used by macroeconomic policymakers or IMF staff.

Modern dynamic general equilibrium analyses of the impact of macroeconomic policies on inequality, with strong micro data underpinnings, exist in the academic literature, but the vast majority deals with advanced economies. These analyses do not incorporate key features of low income economies such as low intersectoral mobility, a large fraction of output produced by unproductive agricultural/family business, and shallow and incomplete financial markets.

The ultimate objective of research on this topic is to allow policymakers to address standard policy questions (e.g. the effects of fiscal policy or public investment scaling up, or the implications of commodity price booms and busts) using tools that are both broadly in use or under development at the Fund and also sufficiently rich in sectoral structure to allow interesting and useful distributional analyses.

Analytical Framework

The model we have developed includes several key features of low income countries such as: 1) disproportionately large and unproductive agricultural sectors; 2) sectoral frictions that prevent resources from being employed in sectors with highest productivity, and 3) financial frictions of different degrees (including no possibility of saving, or borrowing). A summary of the sectors and flows of the model is displayed in figure 1, where c denotes consumption, superindices denote type of good (o being non-food, f being imported food and a denotes domestic food); w are wages.

Commodity Price Booms and Macroeconomic Implications

The Case of Ghana

Global real food prices increased by about 50 percent between 2003 and 2013. This increase occurred after a two-decade period of food price stability, and it is of concern because food prices may disproportionately affect those who are economically vulnerable. Indeed, households in LICs devote half of their budget to food while households in developed nations devote less than 15 percent. It is also true, however, that a larger fraction of households are food producers in LICs and most of the exports of these countries relate to commodities. As a result, the impact of increases in food prices on LICs is not obvious. One of our quantitative experiments is based on the shocks displayed in Figure 2.

We derive the quantitative implications of observed changes in food prices on macroeconomic aggregates and the distribution of income and study the implications of some of the policies currently being debated to mitigate the impact of this shock. We calibrate the model to Ghana, which has unique household panel data. We find this shock can have strong distributional implications that operate through general equilibrium effects by changing the price of domestic food, and thus the purchasing power of farmers vis-a-vis other agents. A simplified summary of results is displayed in Figure 3.

Clearly, different types of households are impacted in a very different way, and the impacts change through time. Entrepreneurs’ consumption increases the most on impact, while urban households (labeled households) benefit the least of the shock. In the long run, entrepreneurs’ consumption ends up falling relative to the initial position, while farmers’ and households’ consumption increases slightly.

Distributional Consequences of Infrastructure Investment Booms

We evaluate the interactions between infrastructure expansions, the distribution of income, and economic performance in a standard dynamic general equilibrium model of structural transformation. The following policy tradeoff emerges: infrastructure investment expansions may have the largest impact on GDP when targeted to the highest productivity sector, but they may also increase income inequality, and even depress the income of the poorest. This is due to the fact that infrastructure expansions require increased production of non-agricultural goods. The latter lowers the relative price and incomes of agriculture so that markets can achieve the necessary reallocation of resources. Wage and relative price movements may be large when output is close to subsistence levels, due to the inelastic demand for agriculture.

Looking Forward

The team is starting to work on the distributional impacts of fiscal policy using the same analytical framework already developed. The objective is to develop a new tool to determine the distributional and macroeconomic impact of alternative consolidation strategies to guide policymakers in choosing the path that balances growth and inclusiveness in line with their policy objectives.

Collaboration

Ongoing Projects: A non-exhaustive list:

  • Louise Fox is working on the paper "Are African Households Heterogeneous Agents? Stylized Facts on Patterns of Consumption, Employment, Income, and Earnings for Macroeconomic Modelers," which presents key empirical facts about consumption, employment, income, and earnings in sub-Saharan Africa.