Discussion of Commodity Price Booms: Macroeconomic and Distributional Implications by Mendes-Tavares, Peralta-Alva, and Telyukova

Transcription

1 1/14 Discussion of Commodity Price Booms: Macroeconomic and Distributional Implications by Mendes-Tavares, Peralta-Alva, and Telyukova Diego Restuccia University of Toronto IMF Washington DC September 18, 2014

2 2/14 What MPT Do International price of food has increased by 50% in the last decade Question: what are the macro and distributional implications of this price increase on a small-open low-income economy Develop a rich model structure to study the implications of this price change and assess policy alternatives Model calibrated to micro data for Ghana Perform two experiments: An increase in the international price of food of 50% A policy that subsidizes domestic food prices by 5% (revenues from lump sum taxation to business)

3 3/14 What MPT Find A 50% increase in international food price generates strong distributional implications but direction of results depend on the elasticity of substitution of imported and domestic food The price change also generates a negative impact on aggregate consumption and output A subsidy to domestic food prices can partially undo the effect of the shock

4 4/14 Comments (1) MPT are to be commended for pursuing such an ambitions, difficult, and important question Great to see quantitative macro applied to the issues of food security and income inequality in poor countries Most quantitative macro literature on inequality has focused on rich economies, where incomes/earnings at the top 1% of the distribution has been emphasized (2) While a rich model structure may be required for this question, the paper could do more to motivate and justify many modelling choices, results could exploit more the given structure

5 5/14 Comments (3) How sensitive are allocations in poor countries to changes in international food prices? What are the key allocations to focus on? Paper assumes perfect segmentation of occupation and sectoral choices, abstracting from changes in the allocation of labor across sectors, land across farm households, etc. It can be argued these changes are key for development and not clear how sensitive they are to price changes (extreme example) Paper could do more to motivate the importance of international prices for domestic consumers in poor countries, including subsistence farmers! The elasticity of allocations to prices is likely to vary a lot across countries

6 6/14 Comments (4) Model abstracts from non-homothetic preferences Non-homothetic preferences are likely to play a role in the reaction of poor households to changes in prices/income Have been shown to be important in the allocation of time across sectors, across activities such as leisure, home production, and schooling, etc.

7 7/14 Comments (5) Calibration of λ (share of domestically produced food) and ρ (elasticity of substitution between domestic and imported food) Do not use poor country data to get these preference parameters, allocations are likely to be influenced by barriers and other distortions not in the model Poor countries observe negligible trade flows in agriculture, but this may be a consequence of broad barriers to trade, including geographical dispersion of rural population, poor public infrastructure, etc.

8 8/14 Share of Expenditures on Food across Countries 0.6 CAF Nominal Share of Food (in GDP) ZAR TZA NGA ARM KHM ZWE MOZ TGO SDN TJK ETH MDGBGD KGZ NPL PAK RWACIV SEN TCD BEN SLEGHA AZE CMR EGY BDI NERGNB MLI BFA GIN YEMPHLSYR IDN KEN MRT LAO UGA LKA SWZ COG UKR PRY BLR LSO BOL IND MNG PER MAR MKD MDA GEOTUN BIH GNQMUS RUS GAB GMB VNM VEN LBN MWI IRQ ECU ALB JOR COL ROM ARGLTU NAM TUR URY SRB MEX AGO ZAF POL LBR BGR KAZ BRAMYS CHL IRN LVA EST OMN HUN CHN THA CZE BWA SVK GRC PRT SAU ISRITA NZL TWN SVN KOR ESP FRA FIN JPN KWT DEU BEL AUS AUT CHE SWE ZMB GBR DNK HKG CAN NLD SGP NOR IRL USA GDP per capita (PPP adjusted, log)

9 9/14 Limitations of the share of expenditure picture: Food prices are not the same as agricultural prices Most households in low income countries live and work in rural agriculture, produce close to subsistence levels, i.e., consume most of what they produce

10 10/14 Share of Employment in Agriculture across Countries Share of Employment in Agriculture GDP per Worker relative to the U.S.

11 11/14 Labor Productivity in Agriculture across Countries 1 Relative GDP per Worker in Agriculture(log scale) 1/10 1/50 1/100 1/200 1/200 1/100 1/50 1/10 1 Relative GDP per Worker (log scale)

12 12/14 Different Perspective Prices of agricultural goods are a factor of 7-fold higher in poor countries relative to rich, mostly a reflection of their low productivity in agriculture This dwarfs the 50% increase in international commodity prices Mitigating the international price gap requires reducing the productivity gap in agriculture across rich and poor countries To do so most often will involve poor countries eliminating established policies... What explains the productivity gap in agriculture? Farm size and misallocation of factors due to restricted land markets

13 13/14 Farm Size across Countries Small Farms (<5Ha) Large Farms (>20Ha) Holdings 0 5 Ha (fraction of total) Holdings 20+ Ha (fraction of total) Q1 Q2 Q3 Q4 Q5 Quintile of Income Distribution 0 Q1 Q2 Q3 Q4 Q5 Quintile of Income Distribution

14 14/14 Conclusion Excellent idea of bringing quantitative macro to study issues on inequality in poor countries Exploit micro data to connect better response of household choices to price changes Explore changes in sectoral allocations to connect better to recent quantitative work on macro development