Supermarket Loss Estimates for Fresh Fruit, Vegetables, Meat, Poultry, and Seafood and Their Use in the ERS Loss-Adjusted Food Availability Data
by
Jean C. Buzby,
Hodan Farah Wells, Bruce Axtman, and Jana Mickey
Economic Information Bulletin No. (EIB-44) 26 pp, March 2009
ERS maintains the web-based Loss-Adjusted Food Availability data
series, an important resource for estimating trends in the amount
of food available for consumption over time. By tracking food
loss-food made inedible by moisture loss, spoilage, and other
causes-analysts can estimate how much food is eaten per person over
a given period.
What Is the Issue?
Some of the food-loss assumptions used in the ERS Loss-Adjusted
Food Availability data were seemingly simplistic and may not have
reflected current manufacturing, retail, and food preparation
practices. Retail food loss was particularly difficult to estimate.
Prior to this study, the per capita food loss estimates at the
retail level were, across the board, 12 percent for every fresh
fruit and vegetable commodity (e.g., fresh strawberries, fresh
spinach) and 7 percent for every type of meat, poultry, and seafood
(i.e., fish and shellfish, both farm-raised and wild-caught)
covered in the database. More precise estimates for each fresh
commodity are desirable to reflect physical differences in spoilage
rates and other reasons that influence food loss, such as use of
innovative packaging to prolong shelf life. Using new estimates for
each commodity could affect ERS calculations of the amounts of
different foods available for consumption.
What Did the Study Find?
This report and the accompanying ERS Loss-Adjusted Food
Availability data give analysts, for the first time, national
estimates of the food loss percentage at the supermarket level for
each fresh fruit, vegetable, meat, and poultry commodity in the
data set. The average loss rates for 2005-06 for individual fresh
fruit, vegetable, meat, and poultry commodities at the supermarket
level, as estimated by the Perishables Group, Inc., varied from 0.6
percent for sweet corn to 63.6 percent for mustard greens. The
study also provided new average estimates for all fish and all
shellfish. When the study incorporated the new loss estimates into
the ERS Loss-Adjusted Food Availability data series, the impact on
per capita estimates varied broadly among commodities within a food
group (e.g., among all fresh fruit). The largest annual impacts,
per capita, were for fresh potatoes, chicken, beef, pork, bananas,
and sweet corn-all of which have high shares of food available for
consumption for their respective food groups.
However, as a whole, the new food loss estimates had little
impact on average food loss rates for each food group in the ERS
Loss-Adjusted Food Availability data series or on per capita
estimates of the quantity of the different food groups available
for consumption at the retail level because the newer estimates
were generally close to the earlier loss assumptions. Compared with
the earlier ERS per capita food loss estimates of 12 percent for
each type of fruit and vegetable and 7 percent for each type of
meat, poultry, and seafood covered in the database, ERS found that
annual supermarket losses for 2005 and 2006 averaged 11.4 percent
for fresh fruit, 9.7 percent for fresh vegetables, and 4.5 percent
for fresh meat, poultry, and seafood. The new estimates would
increase per capita estimates at the retail level in 2005 by 0.7
pounds (0.6 percent) for fresh fruit, 4.2 pounds (2.7 percent) for
fresh vegetables, and 4.8 pounds (2.7 percent) for fresh meat,
poultry, and seafood. Dividing these annual changes in per capita
estimates by 365 days results in very small daily per capita
changes.
How Was the Study
Conducted?
ERS obtained updated food loss estimates at the retail level for
individual fresh fruits, vegetables, meat, and poultry and
aggregate estimates for all fish and all shellfish from the
Perishables Group, Inc., and applied them to update some of the
assumptions used in constructing ERS Loss-Adjusted Food
Availability data to see how they affected per capita estimates of
the food available for consumption. The Perishables Group, Inc., an
independent consulting firm, used a sample of data from six large
national and regional supermarket retailers from their proprietary
database. The sample did not include convenience stores,
megastores, club stores, and mom-and-pop grocery stores. For each
store in the sample, supplier shipment data for 2005 and 2006 was
paired with point-of-sale data to identify food loss percentages
for each covered commodity. For fresh meat, poultry, and seafood,
data were supplemented by the Perishables Group with qualitative
information from more than 10 retailers. The study also compared
loss estimates for 2005 and 2006 as a validation of methods
used.