Projects Funded for Thibault Fally


Farms, Firms, and Fixed Costs: Clustering and Returns to Scale in Agricultural Exporting

  • James Sayre
  • Thibault Fally


Estimation of Demand and Price Aggregators: Application to US and California Household Scanner Data

  • Thibault Fally


Specific Objectives of the Project
1. Combine detailed data on nutrients and food expenditure (Nielsen, FCD-USDA and Label insight).
2. Model and estimate the functional form of demand with price aggregators.
3. Examine implications for price elasticities and substitution effects.
4. Publish the final article in a top refereed journal, as well as a summary in the ARE Update.

Project Report/Summary of Results
We have cleaned and matched three main sources of data: Nielsen Home scanner data, FDC-USDA data, and Label Insight data. The Nielsen data provide information on shopping trips and items purchased at the barcode and household level. FDC-USDA data provide information on ingredients and other product characteristic at the barcode level. Label Insight data covers similar information for a larger availability of barcodes. Merging the three data sources has been challenging due to non-conformity of barcode definitions (e.g. number of digits, presence of check digit) but we managed to cover more than 85% of household food expenditures for some of the years (excluding generic store brands). We are still working on constructing price aggregators and estimating demand using these aggregators.


The Cost of Gluten Avoidance: Evidence from US Scanner Data

  • Thibault Fally


Specific Objectives of the Project:

  1. Provide new stylized facts to document issues and consumption patterns related to gluten consumption and gluten-avoiding diets across US households
  2. Quantify the costs of gluten avoidance, substitution patterns, as well as the recent changes in the supply of gluten-free food products in the US
  3. Quantify the effects of these changes in demand for the supply side
  4. Publish the final article in a top refereed journal, as well as a summary in the ARE Update and one ERS/USDA publication.

Project Report/Summary of Results:
The project was delayed due to having access to the IRI scanner data only starting August 2019. We are still working on the project, but we have a series of new stylized facts on gluten avoidance (objectives 1 and 2). Contrary to our expectations, richer households are not more likely to avoid gluten, relative to poorer households; and gluten-avoiding households are very similar to other households in terms of expenditure shares across a hundred broad categories of food items (suggesting an elasticity close to unity across broad categories). But gluten-avoiding households spend much more on individual products (within broad categories) having a “gluten-free” label. In addition, we confirm that gluten-free-labeled products tend to be 20~30% more expensive. We are still working on estimating elasticities of substitution before we are able to compute our final estimates of the “cost of gluten avoidance."


International Trade in Commodities Matters

  • Thibault Fally


Specific Objectives of the Project
-Compile data on the production, trade, input-output linkages, and supply and
demand elasticities for many commodities, including agricultural commodities; no one
has previously gathered these statistics about commodities within one dataset,
particularly one that features commodities listed at the Harmonized System level and
countries listed using a standardized code system (ISO)
- Review the vast, but highly scattered, existing literature examining the
sensitivity of supply and demand for these commodities in response to changes in prices
- Develop a general-equilibrium model of consumption, production and inputoutput
linkages to determine role of crucial primary commodities in international trade
-Quantify the welfare gains from trade and the winners and losers from the
distribution of natural resources across space
- Publish data on website so they can be used by researchers and practitioners
interested in studying the importance of price sensitivities of supply and demand for
agricultural commodities and their implications for foreign and domestic agricultural
producers and consumers, including those in California
- Publish the final article(s) in a top refereed journal, as well as a summary in the ARE update.

Project Report/Summary of Results
Primary commodities account for approximately 16 percent of world trade, yet they are
used extensively as intermediate inputs into many production processes. We show that
ignoring several key features of trade in commodities leads to a large understatement of
aggregate gains from trade despite their relatively small share of world trade. We
quantify the welfare gains from international trade when we account for specific
characteristics of most primary commodities: i) a low price elasticity of demand as a
result of difficulty in finding substitutes, ii) a low price elasticity of supply, and iii) a high
concentration of natural resources and production among a few countries. For instance,
copper is difficult to replace in the electronic equipment industry, the supply and demand
for copper vary only slightly with changes in prices, a large share of its supply comes
from Chile and copper accounts for half of Chilean total export revenues. We explicitly
account for these features in a general-equilibrium model of consumption, production,
and input-output linkages. In our simulations, we confirm that ignoring these specific
features of commodities leads to a wide understatement of the aggregate gains from trade.


Trade and Food Grocery Shopping Across the U.S.

  • Thibault Fally


Specific Objectives of the Project
Examine consumption choices of US consumers for food products from California versus products from other states and countries, and how trade costs and income affect these choices, using home and store scanner data.

Project Report/Summary of Results
Relative to our initial proposal, we have worked until now on a simpler way to approximate the location of production using the distribution of sales across counties and states. Our method can be applied to a wide range of product, except those with high import penetration ratios.
• Thus, we are focusing on goods that are mainly produced in the US, focusing on the geography of sales and prices within the US. Most of our preliminary results focus on dairy products.
• Our method works well for medium-sized producers, but not as well for the largest producers. To our surprise, we found that sales by the largest producers are geographically very dispersed (for the largest producers, we will manually search for the actual location(s) of production).
• Based on our estimate of the location of production, we then examine how prices and sales vary with distance (current work in progress).
• We also examine within-US trade across households. We find that richer households tend to purchase products that are distributed more widely, across regions and distributors.