diff --git a/gebben.org/content/Areas_of_interst.md b/gebben.org/content/Areas_of_interst.md index e33ecef..2568d06 100644 --- a/gebben.org/content/Areas_of_interst.md +++ b/gebben.org/content/Areas_of_interst.md @@ -6,13 +6,41 @@ menu: title: "Areas of Emphases" weight: 200 --- - ## Ground \& Surface Water +I have a couple of papers in the works that focus on water management and institutional arrangements. I have a particular interest in the development of water rights in the Mountain West. Many of the U.S water rights come from English Common law, or where formed in the Eastern parts of the United States where water is far more abundant than in Mountain West. This lead to conflicts as people moved west facing new economic conditions. This intersection of institutional and resource economics allows for some interesting areas of research. -Test of the heading space -> Test of block quotes +The peculiarities of the water market makes it so that informal trades are easier to engage in than than a pure market, so it is difficult to find good price data to work with. The water rights are ranked, so that early rights have first access to water in drought compared to later right holders meaning a "water right" is heterogeneous and price can not easily be compared even when sales data is available. -## Wind +In my first paper on water management I develop a hedonic model, where the value of the water resource is captured as one element in the value of farm parcels. Generally I shy away from hedonic models as I think selection bias leads to an under evaluation of land values. However I think this approach works well in this case since the farm land is primarily a production input so the change in water value is captured cleanly in the NPV of farms. +### Hedonic Analysis of self governance of ground water management in SLV +In this paper I evaluate a case study of institutional rearrangement in San Luis Valley Colorado, located along the Rio Grand River. This area is a highly agricultural and arid region. Having early settlement of farms there are relatively early water rights for surface water. However, the technology necessary to drill ground water wells came later so the same farmers have a low priority for ground water. There was a legal grey area in pumping ground water as the source of the water was uncertain until better geologic models were created. As a result farmers used ground water to substitute surface water, allowing junior farmers to obtain more water by drilling wells than they would have legally been able to obtain if using river diversions. + +Since this ground water has been more cleanly linked to Rio Grande River flow, the state of Colorado has begun to put pressure on these farmers. During a sever drought in the early 2000's the incentives from the state were higher to evaluate the water use along the river. There was an explicit warning from the state engineer in 2002 that any out of priority wells could be completely shut in if things did not change. + +As a group the farmers in this region have shared interests making collective action easier. The prior apportion system is economically inefficient allocating water not to the best marginal use but to the oldest farm. The legal ramifications of trading water rights under the system opens up lawsuits. As a result the farmers could expect a much higher increase in costs than was necessary if they were forced to follow the strict priority system. On the other hand over pumping does come with externality costs, depleting the reservoir is a common pool resource, and there are additional externalities. By cooperating at a local level the farmers reduce pumping rates in a less costly manner than if the state shut in wells. Allowing existing wells to operate at lower levels and allocating water to the highest uses, but reducing surface water diversions or pumping in lower marginal areas. This incentives helped in the formation of the ground water management Subdistricts. Ground Water Basins are organizations created by the state of Colorado that help manage ground water use within a certain regions, stakeholders are delegated and form policy. The subdistrict was formed when farmers in certain areas of the district voted to create a group to enact policy at a highly local level. In this case regions with the most interconnection with the Rio Grande organized into seven Subdistrict. The first subdistrict is the focus of this analysis, and was picked out as the farmers with the highest legal liability for over pumping. Initially the policy they devised was to charge a fee for pumping ground water, a pigouvan tax. The money from the fee would be redistributed to other members or used for water use reduction programs. The idea being that this is a less burdensome way of managing water than a pure command and control that could result from state intervention + +This subdistrict formation is the quasi-natural experiment I leverage in the paper. Farms in this first subdistrict face similar externalities as other farms in the regions, having the same draw down of the aquifer, but have the highest legal liability of over pumping. This means that during the threat of shut in in 2002 their land values should decrease relative to other farms in the area. However as expectations change in 2006 the local externalities of over pumping, and the risk of being shut in by they state are reduced. I use farms in the same counties but out side of the subdistrict as the control group matching to farm level attributes such as crop choice, to proxy for unobservable. My final data set (which I will explain later) linked sale prices of farms with there location, and USDA crop choice data. This let me identify which farms are located within the subdistrict, and which fall outside of it. After matching the groups are differenced. Since they fall within the same counties the parraell trend assumptions should account for local prices of inputs, economic trends and even highly localized county law. The general result is that farms in the subdistrict did see a large and highly statistically significant decrease in land values with a dummy for being sold after 2002. This should only account for the perception of a shut in from the state, as the expectation in NPV of the land will shift for the control if the aquifer is depleted. In 2006 there is again a significant effect but in the opposite direction, in this explanation of the data the self organization into subdistrict one changes expectations. Now people purchasing farmland in the subdistrict think that self management of water will decrease the threat of being shut in and perceive a higher future return on farm land, with access to cheaper water. + +Unfortunately this region is mostly evenly intersected by three counties. This required scripts to be used to collect data from three separate county assessors offices. In some cases I reverse engineered the web API and was able to download the PDF's of all sales records through a web call. In other cases I coded a script to walk through the web page and sequentially download each item through keyboard bindings. + +Another problem was that the data was not tied to geographic location. The data set available cost \$10,000 to obtain which was above my graduate school budget. So I created a python script that would zoom in on the legal parcels on a map provided for the region, then it would take screen shot of the map, save it locally with the parcel ID number name and loop through. I set this up on multiple virtual machines since they made it very difficult to automate. Eventually I had images of each legal parcel with a reference point for lat long. I then created a image processing procedure that used the lat long point as a reference, and the axis scale on the scale to draw a shape of the parcel. These shapes were then loaded into QGIS where some clean up was done that minimized the overlap of parcels and smoothed any erroneous ragged edges. In QGIS I intersected the parcels with USDA map data of crop growth in different years. The results is a time series of parcels with associated attributes, crop choice, and a history of sales records. + + +I think my findings line up well with some qualitative data I have collected. Initially I was focusing primarily on the externality of pumping. I was aware of the legal concerns but was considering the subdistrict formation as a response to local pumping cost increases and not a legal threat. However I read through the archive of state engineer records and found that wells were threatened with being shut in at the this time. In another water management district wells end up being shut in making this threat credible. In conversations with farmers in the area, they pointed to the legal threat as the key concern in the area. I think this qualitative analysis strengthen this statistical conclusions. + +## Wind +I have some engineering work and economic work around wind energy. On the engineering side I designed a wind turbine in CAD software for a company seeking assistance from Colorado School of Mines students. The design included simulations of various wind speeds and the blade depth could be adjusted based on conditions. +My economic investigation of wind started when I learned about policies meant to encourage completion in the energy sector. Small power plants are allowed to receive the consumer rate of electricity where as larger firms, receive a lower price per unit produced. The idea is to encourage market competition in the energy production. Wind however opened up a new way to game the system. The size of the power plant is determined by the plant capacity owned by the same company in a contiguous area. Wind turbines can be easily separated so a large firm can place a few wind turbines in one location, then buy a plot a mile away and place a few more, each segment will receive a higher rate of return. + +The problem arises from some subtle externalities in wind generation. The ideal conditions for wind turbines is a constant, (relatively) fast, and smooth wind flow. The angle of wind turbines are adjusted to extract the most power from a given wind speed, they can often be adjusted as the wind speed changes. If wind is "bumpy" the optimal angle must be changing optimal. For example if the wind speed is always 7mph the angle can be set to optimize to that speed and the turbine will generate say 1kwh. If the same wind turbine is used and the average wind speed is 7mph but the variation in speed ranges from 1mph to 20mph, the generation may be bumped down to 0.5kwh. + +A wake is formed behind a operating turbine creating local variation similar to the effect described. Over a large enough distance this "bumpiness" subsides and the wake externality approaches zero. + +The incentives to space wind farms in small alternating plots exacerbates the externality. Over capitalization occurs because the upstream farm spaces turbines more densely than if they considered the cost to downstream farms. + +I created a model of wind turbine spacing that accounts for this effect. The micro economic model then has "small firms" and "large firms" purchase plots under the incentive structure of the law optimizing profits. The model predicts that given I high enough wind speed large firms will purchase the leading wind and not receive the bonus, however at cut off point small firms purchase land in alternating plots. Depending on the subsidy size there may be more or less wind energy generated, but in all cases the capital costs are higher than the subsidy required to induce the same power generation by a single large firm. + +My next step in the model will be to account for game theory considerations. The game would involve each having imperfect strategic information of other firms, until they reveal there strategy when building on the plot. It is possible that a collaborative Nash equilibrium is possible from this signaling, but the threat may no be credible when the firm owns too little land to punish in later rounds. ## Oil \& Gas