Markets for tradable emissions permits are the economically preferred tool for reducing pollutants such as GHG. Researchers will conduct an assessment of the performance of existing markets of this type with the goal of drawing lessons for their potential application in California water; proposing to synthesize and apply the lessons drawn from this literature to begin practical design of California‐based and/or regional markets for water.The Regional Offices of the DWR perform land surveys of every field in the Central Valley every two or three years. These surveys indicate an incredible diversity of Central Valley agriculture, including high‐value fruit and vegetable crops and low‐value field crops and cotton. Crop data from the GIS dataset for the San Joaquin Valley illustrates the distribution of high‐ and low‐value crops, as shown in Figure 2‐7. Although most crops may be grown in almost any part of the Central Valley, high‐value crops tend to predominate in the east side of the San Joaquin Valley. Comparing this information to the water delivery data, it appears that the high‐ value cropping occurs primarily in areas with better access to surface water deliveries.The database of farmland value for this study was derived from the USDA June Agricultural Survey. This survey is conducted in June of each year to construct forecasts of expected yields of most crops. The survey includes a random sample of the Census of Agriculture and a stratified sample of farms based on geographic location.

The stratified sample is geo‐referenced by latitude and longitude. The sample includes self‐reported farmland value per acre for the years 1998–2003. The distribution of farmland values reported in the survey for the California Central Valley is shown in Figure 2‐8.Similar to water supply data, farmland values vary widely across the Valley,hydroponic channel but tend to be highest in districts located in the central and eastern portions of the Central Valley .This section examines the potential damage to California agriculture from predicted changes in water supply reliability under climate change. Researchers compiled a unique micro‐level dataset of farms in California that allows them to test how different water rights capitalize into farmland values. These capitalized values are hence the implicit market prices for the water rights of the land. Although there have been theoretical studies that outlined the value of varying water rights with different seniority in the Southwest , there are only a few empirical studies that have examined whether and how access to irrigation water is capitalized into farmland value in practice. Hartman and Anderson consider land sales within an irrigation district in Colorado; and Crouter considers land sales within a different irrigation district in Colorado; Faux and Perry consider land sales in four irrigation districts in Malheur County, Oregon. Finally, Moreno et al. 2005 evaluate agricultural land values in a single water district in California. Three of these studies find that water availability is a significant determinant of farmland value. However, all four studies cover a much smaller area than this study’s sample, which extends to over 150 irrigation districts in 39 counties in California. This study’s larger spatial coverage permits us to allow for the effect on farmland value of climate variables that are not likely to vary much within the small scale covered by these other studies. The only other study that incorporated surface water use on a larger scale relied on average farmland values in a county, where both dryland and irrigated farmland values are averaged . 

The analysis proceeds as follows. Section 2 discusses the history of water projects in California that motivate the study, Section 3 introduces the reduced form hedonic model, Section 4 describes the unique dataset, Section 5 describes and discusses the study’s empirical results, Section 6 provides a sample calculation of the potential impact of climate change on farmland value, and Section 7 presents the conclusions. At the beginning of the twentieth century, California was still very much an agricultural state. Under appropriative water rights, users could file a claim for water rights with the SWRCB, as long as the water was put to a beneficial use. Water rights are a prime example of first‐order stochastic dominance: because the runoff of rivers is stochastic, individual water rights are filled by seniority . With decreasing seniority, each claimant can only get water from the remaining water resources after the entitlements of more senior water right holders have been satisfied. The random nature of water availability was exacerbated by the fact that limited storage capacity was available at the beginning of the twentieth century. Precipitation occurs almost exclusively during the winter and in the northern part of the state, therefore, storage and conveyance facilities are necessary to bring surface runoff to the south and to farms during the growing season. In sharp contrast, the use of groundwater is virtually unregulated, which, similar to all common access problems, gives a disincentive to conserve the groundwater table for future periods. About one‐and‐a‐half million acres were under irrigation in the San Joaquin Valley by 1930 and almost all of them relied on groundwater as the source of irrigation . However, extensive overdraft of the unregulated groundwater resources had resulted in a drop of the water table of as much as 300 feet.

There was heightened concern that accessible groundwater would vanish in the next couple of decades. In 1933 the state legislature approved the Central Valley Project , which was designed to capture two thirds of the state’s runoff. Almost all of the water that is captured in the Sierra Nevada is collected in the Sacramento River and the San Joaquin River which meet at the Delta and empty into the ocean. The CVP collects water from these river basins, and transports it from the northern part of the state to the southern part for several hundred miles through canals and by reversing the natural flow of some rivers. When voters finally approved a $170 million bond measure to build the project, the country was in the Great Depression and the State of California was not able to sell the bonds. President Franklin Roosevelt ordered the U.S. Bureau of Reclamation to take over the project in December 1935. The original project was constructed between 1937 and 1951, with several newer features being added later. By 1990, the CVP had 20 dams and reservoirs capable of storing 12 million acre‐feet of water and 500 miles of major aqueducts and canals. The three largest dams are Shasta Lake, with a capacity of 4.5 MAF , Clair Engle Lake , and New Melones Reservoir , with storage capacities of about 2.5 MAF each. Water that would otherwise flow into the Delta is pumped into the Delta Mendota Canal at the Tracy pumping plant. The pumping capacity of this plant is 6.34 acre‐feet per minute. The total annual contracting quantity of the CVP is 9.3 MAF, where 4.8 MAF are project water and 4.5 MAF are water rights settlements . The growing urban demand for water in Southern California led to the construction of another large surface water storage and distribution system that is owned by the State of California–the California State Water Project with yearly contracts averaging 4.2 MAF of water. The SWP consists of 22 dams and reservoirs,hydroponic dutch buckets by far the biggest of which is Oroville Dam with a storage capacity of 3.5 MAF. The SWP was constructed between 1961 and 1973 and delivers about 2.5 MAF of water to Southern California, depending on wetness conditions. It also supplies water to irrigation districts, . These water deliveries are not subsidized and the wholesale cost in Kern County is about $70/AF. The SWP has only about 60% of the supply capacity that was originally planned in 1960. Completion of the remainder has been blocked since 1982, when voters rejected Proposition 9 to build the Peripheral Canal.5 If the system were now to be built out, current estimates are that the new water storage facilities would cost on the order of $500–$1,000/AF , which is much larger than historic cost estimates and hence historic water rights result in rents for farmers. Continued conflict and expensive legal battles over water rights demonstrate that these rents must be of significant magnitude. One would hence expect that that these rents capitalize into farmland values. The next section presents a brief model to motivate the reduced form analysis used to estimate how water rights capitalize into farmland values.The dependent variable, farmland value per acre, was derived from the June Agricultural Survey . This survey is conducted in June of each year to construct forecasts of expected yields of most crops.

The survey is split into two parts: the first is a random sample of the Census of Agriculture, while the second is a stratified sample of farms based on geographic location. This study relies on the second part, as it is a geo‐ referenced sample of all farms . The hedonic regression uses the self‐reported farmland value per acre as the dependent variable. The dataset includes observations for the years 1998–2003, and all farmland prices were adjusted by the gross domestic product implicit price deflator to be in 2000 dollars. Figure 3‐1 illustrates the locations of farmland values reported in the survey. In the future, the researchers will attempt to test this assumption by gathering other farmland value data sources.This study uses a 103‐year, high‐resolution temperature and precipitation climate dataset for the coterminous United States. This small‐scale climate series was developed by Spatial Climate Analysis Service at Oregon State University for the National Oceanic and Atmospheric Administration . Researchers at Oregon State University developed the PRISM model that is employed by almost all professional weather services and regarded as one of the most reliable interpolation procedures for climatic data on a small scale.6 The existing economics literature has generally represented the effect of climate on agriculture by using the monthly averages for January, April, July, and October. However, from an agronomic perspective, this approach is less than optimal. First, except for winter wheat, most field crops are not in the ground in January; most are planted in April or May and harvested in September or October. Second, plant growth depends on exposure to moisture and heat throughout the growing season, albeit in different ways at different periods in the plantʹs life cycle; therefore, including weather variables for April and July can produce a distorted representation of how crops respond to ambient weather conditions. The agronomic literature typically represents the effects of temperature on plant growth in terms of cumulative exposure to heat, while recognizing that plant growth is, in part, a nonlinear function of temperature. Agronomists postulate that plant growth is a linear function of temperature only within a certain range, between specific lower and upper thresholds; there is a plateau at the upper threshold beyond which higher temperatures become harmful.This agronomic relationship is captured through the concept of degree‐days, defined as the sum of degrees above a lower baseline and below an upper threshold during the growing season. Here, the study followed the definition of Ritchie and NeSmith and set the lower  bound equal to 8°C and the upper bound to 32°C .8 In other words, a day with a temperature below 8°C results in zero degree‐days; a day with a temperature between 8°C and 32°C contributes the number of degrees above 8°C ; and a day with a temperature above 32°C contributes 24°C degree‐days. Degree‐days are then summed over all days in the growing season. Researchers derived the sum of degree‐days during the main growing season . Boundaries of all major irrigation districts were obtained from the DWR, which made it possible to link individual farms to irrigation districts. Water deliveries of the CVP between the years 1992 and 2002 are available in the Operations Report from the CVP and DWR Bulletin 132. Researchers matched this data with estimated water deliveries and water prices obtained from ACWA. It should be noted that the available data on water rights in California are often incomplete, and it is not easy to obtain comprehensive and accurate information about water rights. Researchers are still in the process of expanding and updating the database on water rights.