| dc.description.abstract | Harold Hotelling's (1931) canonical model of exhaustible resource extraction, featuring
resource-owners that maximize profits by trading off extraction today versus extraction in the
future, is widely cited in the literature. However, the mapping of Hotelling's logic to empirical
work in the energy literature has been a limited success. To bridge this gap, Anderson, Kellogg
& Salant (2018) aim to show that a Hotelling model for oil production and drilling can generate
empirical predictions consistent with market observables. They find well-level oil production
(the intensive margin) in Texas to be unresponsive to the oil price, while drilling activity (the
extensive margin) responds strongly. Hence, they reformulate Hotelling's model as a drilling
problem where resource-owners choose when to drill, but the oil flow is restricted by reservoir
pressure. Our thesis aims to contribute to the literature by showing that resource-owners in
fact do manipulate oil production on the intensive margin by injecting water to increase
reservoir pressure.
We first graphically analyze how oil production and the drilling of new wells respond to crude
oil prices in the US between 2000 and 2020. We find that production declines monotonically
over time for a stock of existing wells and that firms respond to increasing crude oil prices by
drilling new production- and water injection wells. These findings are in line with current
literature. However, we also find that the oil production curves for wells that have injection
wells nearby do not decline monotonically.
We then proceed to develop an expression for the marginal cost of water injection and a
function for optimal water injection volumes. Using this function, we estimate a fixed-effect
linear regression model and find that water injection volume responds to oil prices, adding to
the current economic literature. Our findings indicate that firms leverage their ability to
manipulate oil production on the intensive margin by injecting water in response to oil prices.
The evidence contradicts a central assumption made in the model proposed by Anderson,
Kellogg & Salant (2018). | en_US |
