The two main chapters of this thesis study how mineral leasing, drilling decisions, and learning by firms interact over time in unconventional U.S. oil and gas plays. The first chapter uses a stylized, theoretical model to study how the paths of aggregate mineral leasing and drilling investment are jointly determined in equilibrium by forward-looking landowners and firms. This longer-run, market-level view abstracts away from the particular geological characteristics and mineral lease-terms of each location, focusing instead on the dynamics of the general equilibrium and a theory of how the quantity and price of mineral lease transactions are determined over time. The model uses search costs to explain the empirical delay between when firms acquire leases and drill them. Drilling appears to be delayed because firms accelerate leasing during early periods when a large supply of unleased acreage makes search costs low.

The second chapter studies what firms know and learn about the geological quality of the particular locations they drill. The paper does this by estimating a dynamic discrete choice model of firms drilling decisions that incorporates learning about quality over space. Forecasts from the model show how the depletability of oil and gas deposits, the evolution of firms' information about geology, and firms' incentive to drill better locations first affect average output per well over time. Results suggest that firms may have been able to improve average output of wells drilled by around 25% since 2008 by learning about geological quality over space and then targeting better locations. Firms will exhaust the most prolific locations first and transition to drilling less productive areas; however, the model implies that the associated annual decline in average output per well drilled over the next ten years will be mild—a little less than 0.5%.