Optimal taxes in extraction from a spatially explicit aquifer: experimental evidence

Date
2015
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Groundwater resources for drinking water and irrigation are increasingly stressed. They are also common pool resources (CPRs) with incentives leading to potential overuse. Many papers have replicated myopic extraction--in effect, a congestion externality--and the resulting social suboptimality that occurs in intertermporal games. Only a few, however, have produced this in a spatially explicit setting that can be tested in the lab. This paper uses experimental economics techniques and extends this literature with a set of price-based incentive policies. Further, this research considers modified tax policies, which lessen the distributional burdens on users, and also ask research participants their opinions of the different policy instruments at the conclusion of the session. This set of experiments builds on the spatially explicit aquifer first reported in Li et al. (2014). The first task was to derive an optimal intertemporal tax. Numeric optimization was carried out by linking software of Modflow and MATLAB, to derive a single intertemporal tax that incentivizes users to undertake extraction decisions that maximize the value of the aquifer. The experiments were carried out using the optimal marginal tax rate under various distributive configurations. Specifically, the treatments used instruments of thresholds or side payments to generate different distributive outcomes, which theoretically should not have influenced behavior. In addition, a baseline treatment was conducted that did not involve a tax. Participants were recruited to participate in the experiment as groundwater users in May 2014 at the University of Delaware. Seventy-two undergraduate students from economic majors were recruited and 72 × 71 individual choice combinations were collected in the experiment. Each experiment session lasted approximately two hours and participants earned approximately $30 based on their decisions and on the decisions of others sharing a common aquifer. The results show that, compared with baselines, the pumping rates in each round of the tax treatments decrease significantly. This indicates that tax instruments have the potential to reduce groundwater use. However, all of the pumping rates in the tax treatments fall below the optimal pumping path in most rounds. The systematic deviation from the optimal path indicates that the effect of tax policies tends to be slightly excessive. In addition, all of the pumping decisions in each round of tax policies are clustered, supporting the theoretical design that all tax treatments provided the same marginal incentive and participants recognized these incentives. However, the threshold and side-payment treatments produced very different welfare impacts for the respondents. The most important finding was that one side-payment treatment was able to generate the same similar welfare impacts on the users as the baseline, but was also able to drive the respondents closer to the intertemporally socially optimal usage rates. Surprisingly, participants' opinions on the different tax policies did not always match their received welfare. Participants tended to rate the tax treatments without redistribution of tax revenue very low. This can be explained because they gained less profit than the unregulated scenario. Surprisingly, participants also rated the tax treatment with side payments lower than the unregulated scenario in all of the rating aspects although they earned higher profits during these treatments. A potential explanation is that groundwater users simply reject policy interventions such as tax policies. The fixed-effect model addressed the positive impact of profits on participants' rating of a policy as beneficial to individual and group, while the tax policy has a negative effect on rating as in opinions. The final effect is that the positive effect is counterbalanced by a negative effect probably due to the rejection of tax policy. The results also indicate that the negative impact of tax policy on user rating tends to be smaller with higher levels of the tax threshold. The tax treatments with redistribution of tax revenue also decrease the negative impact on rating.
Description
Keywords
Citation