I just finished up a new working paper with Glynn Tonsor that shows how to determine the extent to which a change in price (or quantity) results from a change in supply and/or demand. For some time, Glynn has been reporting updated retail demand indices for meat products. In this new paper, we show how to calculate an analogous supply index, which might provide a useful way to determine how much productivity is changing over time. The basic idea is that we want a way to separate changes in quantity demanded (or supplied) versus a shift in the demand curve (or the supply curve). We also show how the two indicies can be used to determine changes in consumer and producer economic well-being over time.
Here’s the motivation:
“In 2015, per-capita beef consumption in the U.S. reached a record low of 54 lbs/person, falling almost 20% over the prior decade from 2005 to 2015 alone (USDA, Economic Research Service, 2020). Why? Some environmental, public health, and animal advocacy organizations heralded the decline as an indicator of their efforts to convince consumers to reduce their demand for beef; others argued, instead, the change was a result of supply-side factors such as drought and higher feed prices (e.g., Strom, 2017). Per-capita beef consumption subsequently rebounded, and in 2018 was almost 6% higher than in 2015. Dramatic fluctuations in corn, soybean, and wheat prices in the late 2000s through the mid-2010s led to similar heated debates about whether and to what extent price rises were due to demand (e.g., biofuel policy and rising incomes in China) or supply (e.g., drought in various regions of the world) factors (e.g., Abbot, Hurt, Tyner 2019; Carter, Rouser, and Smith, 2016; Hochman, Rajagopal, and Zilberman, 2010; Roberts and Schlenker, 2013). These cases highlight the challenge of interpreting market dynamics and the need for metrics that can decompose price or quantity changes to reveal underlying drivers and consequences.”
We calculate the supply and demand indicies for a number of agricultural markets and time periods. First, consider changes in supply and demand in the fed cattle market since the 1950s, as shown in the figure below. The demand index trended positively from 1950 through the mid 1970s. The demand index peaked at a value of 204 in 1976, and it hasn’t been as high since. Demand fell through the 1980s and early 1990s before rebounding. Since 2010, the demand index has been at values just below the 1970’s peak. The supply index trend was positive from 1950 up till about 2000, but has been stagnant except for the past couple decades. Nonetheless, the 2018 supply index value is the highest of the entire time period since 1950. The figure shows a significant drop in the supply index that began in 2013 and bottomed out in 2015, which is likely a result of drought in the great plains and from high feed prices. The fact that the supply index dropped during this period while the demand index remained relatively flat helps provide insight into the debate discussed in the quote above.
One can also calculate changes in producer and consumer surplus over time. The following figure calculates the year-to-year changes. On average, from 1980 to 2018, producer surplus increased $2.7 billion each year and consumer surplus increased $0.58 billion each year. Despite these averages, there is a high degree of year-to-year variability. The largest annual change in producer surplus was $34 billion from 2015 to 2016; the largest decline in producer surplus was -$28 billion from 2013 to 2014.
Here’s how changes in the supply index compare for the three main meat categories. Chicken supply shifts have far outpaced that for hogs or cattle. The 2018 chicken supply index value is 380, meaning chicken supply is (380-100) = 280% higher than in 1980. By contrast, hog and beef supply are only 66% and 28% higher, respectively, than in 1980. These differences are likely explained by differential productivity patterns in these sectors. The rise in hog productivity since 2000 corresponds with a time period over which the industry became increasingly vertically integrated, increasingly mirroring the broiler chicken sector. The much longer biological production lags in beef cattle (which range from two to three years from the time a breeding decision is made until harvest) and less integrated nature of the beef cattle industry help explain the smaller increases in the supply index in this sector as compared to pork and chicken. We also show, in the paper, that these supply indicies correlate in intuitive ways with changes in factors like feed prices, drought, and aggregate U.S. agricultural total factor productivity.
One of the useful aspects of the supply and demand indicies is that they can be applied for highly disaggregated geographic units. To illustrate, we calculated U.S. county-level supply indicies. Here are the changes in U.S. supply indicies in the past couple years relative to 2000. Perhaps surprisingly, many areas of Ohio, Indiana, and southern Illinois have experienced negative corn supply shocks in 2016-2018 relative to 2000. The expanded geographical area of U.S. corn production (e.g more acres in the Dakotas) over this period helped mitigate national corn market effects of the adverse Eastern Cornbelt supply shocks. Note that corn yields and total production have increased significantly in many of the red counties over time, and this illustrates the importance of calculating a supply index rather than just looking at yield or production. The supply index gives us a feel for how much more (or less) is produced in 2018 relative to what we would have expected if the level of technology, weather, etc. were the same as in the year 2000.
There is a lot more in the paper.