New EHES working paper
Manifested in historical accounts, songs, and family tales, the boll weevil (Anthonomus grandis), an approximately one-fourth inch long beetle with a very long snout, is considered as the most well-known agricultural pest in the American South.
 |
| Anthonomus grandis |
Arriving near Brownsville, Texas, from Mexico in 1892, the boll weevil started to impair the main economic engine of the South: cotton production. Depending on prevailing wind and weather conditions, the boll weevil could cover from 40 to 160 miles a year such that thirty years after its arrival the whole Cotton Belt was almost completely infested.
 |
| Map showing spread of boll weevil 1892 to 1922 |
The recent EHES working paper by Ager, Brückner and Herz (2014) focuses on the Cotton Belt counties of the American South that were infested by the boll weevil during the late 19th and early 20th centuries. As the boll weevil adversely affected cotton production and hence the demand for labor, the authors exploit the arrival of the boll weevil as agricultural productivity shock to identify the response of labor supply to changes in labor income. The central message of Ager, Brückner and Herz’s article is that labor income shocks had a significant effect on labor supply at the extensive margin in the United States during the 1880-1940 period. For a panel of 903 counties the authors estimate that a one percent increase in labor income increased the labor force participation rate by around 0.2 percentage points. The finding is based on an instrumental variables approach that carefully addresses endogeneity issues.
The instrumental variables approach exploits that in the beginning of the 1890s, counties located in the Cotton Belt of the American South were hit by an agricultural plague, the boll weevil, that adversely affected cotton production and hence the demand for labor. The impact of the boll weevil on output per worker varied across US counties depending on the initial importance of cotton production in a particular county. Counties with a greater initial cotton share experienced a significantly larger drop in output per worker due to the incidence of the boll weevil. Ager et al. therefore use the interaction between the incidence of the boll weevil and counties’ 1880 cotton share as an instrument for labor income.
It is notable that instrumental variables estimates of labor supply are significantly larger than those produced by least squares regressions. An explanation for the larger two-stage least squares estimates is a reverse causal effect, which downward biases the least squares estimates. The negative reverse causal effect arises in the least squares regressions because increases in labor supply decrease output per worker as well as wages (under the standard assumption of decreasing returns to scale in labor).
Ager et al. also explore alternative adjustment mechanisms to the labor income shock. They find that there were significant effects on immigration and emigration, as well as on non-market labor. Decreases in output per worker due to the boll weevil lead to significant decreases in immigration, significant increases in emigration, and significant increases in the share of housekeepers.
The instrumental variables approach exploits that in the beginning of the 1890s, counties located in the Cotton Belt of the American South were hit by an agricultural plague, the boll weevil, that adversely affected cotton production and hence the demand for labor. The impact of the boll weevil on output per worker varied across US counties depending on the initial importance of cotton production in a particular county. Counties with a greater initial cotton share experienced a significantly larger drop in output per worker due to the incidence of the boll weevil. Ager et al. therefore use the interaction between the incidence of the boll weevil and counties’ 1880 cotton share as an instrument for labor income.
It is notable that instrumental variables estimates of labor supply are significantly larger than those produced by least squares regressions. An explanation for the larger two-stage least squares estimates is a reverse causal effect, which downward biases the least squares estimates. The negative reverse causal effect arises in the least squares regressions because increases in labor supply decrease output per worker as well as wages (under the standard assumption of decreasing returns to scale in labor).
Ager et al. also explore alternative adjustment mechanisms to the labor income shock. They find that there were significant effects on immigration and emigration, as well as on non-market labor. Decreases in output per worker due to the boll weevil lead to significant decreases in immigration, significant increases in emigration, and significant increases in the share of housekeepers.
This blog post was written by Philipp Ager, assistant professor of Economics at University of Southern Denmark, Markus Brückner, Associate Professor at the National University of Singapore and Benedikt Herz, PhD student in Economics at Universitat Pompeu Fabra.
The working paper can be downloaded here. https://www.ehes.org/EHES%2068.pdf