This paper studies how the economic conditions in rural counties affect the quality of teachers. I make use of quasi-experimental research methods to study the degree to which the changing well-being of farm-dependent communities affects their ability to staff classrooms with highly-qualified teachers.

The educational data in this study come from the School and Staffing Survey (SASS) and the related Teacher Follow-up Survey, which span from the 1987/88 school year to the 2011/12 school year. The economic data in this study come from the National Agriculture Statistics Service. The study uses a reduced form instrumental variables approach to estimate the annual value of the agriculture bundle in a county. This bundle is more important to the local economy in counties where farming is a large share of the local economy; therefore, the study is limited to nonmetropolitan counties that the Economic Research Service classified as farm-dependent. The annual value of the bundle is calculated based on state-year average prices of goods and on annual production amounts prior to the SASS. To compare trends in different types of counties, I classify counties by the modal crop produced in the pre-SASS period.

I first show that the real value of the agriculture bundle in farm-dependent counties fell dramatically during the SASS timeframe and compare the profile of teachers in farm-dependent counties to teachers in all counties. In farm-dependent counties the average age of teachers is increasing faster than the national trend. Teachers are staying at their schools longer, whereas nationally teachers are staying at their schools for a shorter period of time. This means that farm-dependent counties have a lower proportion of teachers who are new to their school than is observed nationally.

The Energy Independence and Security Act of 2007 dramatically impacted the price of corn and corn substitutes. It did not have as large of an impact on the price of agriculture products that are not corn substitutes.  More than any other event in the past 25 years, this has differentially impacted the real value of the goods produced within farm-dependent counties. I will use a differences-in-differences model to compare teachers in counties that gained the most to teachers in counties that did not benefit as much from the energy bill.

Preliminary results using SASS data from 1987/88 to 2007/08 suggest that the different types of counties are staffing their classrooms differently. However, the survey a year after the Energy Bill, these trends are statistically indistinguishable from no difference. New results will use 2011/12 SASS data to examine the impact of the 2007 energy bill.