Three essays on technology choice in rainfed cereal-livestock systems in Morocco

In this dissertation I address (i) a large, but until now not quantified, barrier to adopting certain agricultural technologies facing cereal-livestock farmers, and (ii) a largely overlooked benefit to rainfed farmers of adopting one of these technologies. While the three essays in this dissertation focus on no-till agriculture, the theory and empirics I present have broader applications.;In the first essay I use a household model to estimate the shadow value of crop stubble, a nonmarket byproduct of cereal production. Stubble is an important source of livestock feed, and the full opportunity cost of adopting no-till, or of switching from cereal to cash crops, must include the value of stubble forgone under these technologies. Using household data from Morocco I find that in a normal rainfall year stubble makes up one-third of the total value of feed used by farmers in the sample and over one-fifth of the total value of cereal production-- more than the cost-savings offered by no-till.;In the second essay I address the implications of common stubble grazing on property rights enforcement and technology adoption. The introduction of a technology such as no-till that presents an alternative use for stubble which could increase its value enough so some farmers enforce property rights over the resource, making it more costly for others to forego using stubble as feed in order to adopt the technology. I present a theoretical model and a simulation calibrated with data from Morocco to explain the emergence of property rights enforcement and interdependence of technology adoption between farmers.;In the third essay I demonstrate that rainfed farmers facing the risk of catastrophic drought extract a quasi-option value from technologies that delay input requirements, which I argue include no-till. Using household data from Morocco I investigate the timing of crop abandonment. I then present a two period simulation model that demonstrates how no-till generates a quasi-option value to flexible farmers under drought risk. I find that under the status quo 14 percent of the total cost-savings of using no-till is derived from quasi-option value.