The first district model addresses the property-enfranchised, assessed-value governance rules that guide most “California water districts.” Board members in these districts respond to political influence based on the assessed-value held by an elector in the district. The second model uses the universal-franchise, popular-vote governance rules that generally direct “irrigation districts.” Board members receive direct political signals of equal weight from each farmer regardless of farm size or tenancy, plus each non-farmers has an equal vote. These differences governance rules lead to predictions about how district resources are managed. The base data set for the empirical analysis is drawn from a survey conducted by the University of California at Berkeley, Department of Agricultural and Resources Policy and Economics, The survey covered 128 districts. The survey methodology and a partial summary of results is included in a department working paper , and an analysis of how the districts altered their behavior during the drought was later published . The survey data set was supplemented with district-specific information from two other sources, The first was the ACW A membership list, which supplied further addresses and contacts,black plastic plant pots wholesale all activities undertaken by the districts, and information on agricultural and municipal customer usage and rates. One-hundred eight districts on the ACWA list were also included in the survey data set The second was the State Controller data on special districts’ financial transactions for the 1991-1992 fiscal year.
One-hundred twenty-seven districts in the survey data set had supplied the State Controller with financial data. This source was also used to pinpoint the primary county and regional location A third source was used to add data on electoral rules. Which voters are eligible in local elections, and how votes are weighted and counted was compiled by district type from the CDWR Bulletin 155-94 . The data set was modified where the Water Code either had provisions specifically relating to a district or making exceptions dependent on the composition of the district . Table 2 shows how the districts in the data set are distributed among twenty-nine counties” and seven regions in California. The largest concentrations of districts are in San Diego , Tulare , Fresno and Kern counties. All other counties have six or less. Most of the districts are located in four regions-the Sacramento and San Joaquin Valleys, the Tulare Lake Basin and Southern California, with 84% of respondents in these regions. Over 60% are located in the Central Valley. This distribution reflects the agricultural orientation of the initial survey since the majority of California’s agricultural activity is located there. Table 3 shows the distribution of land-owner-enfranchised districts in the data set. All but one of the forty-two districts of this type are located in the three Central Valley regions. The Sacramento Valley at the north end has almost the same number at the Tulare Lake Basin at the southern end. Kern county has the largest number, ten, which reflects the seven water storage districts located there. Tulare and Fresno county have five each. The concentration in the Central Valley of both types is apparent, along with the dominance by the region of land-owner-based electoral rules.
Due to large and widespread urban activity in Southern California, land-owner-based electoral rules have difficulty surviving legal and political tests and none are shown in the data set despite the relatively high proportion of all districts located in the region. Average size farm and the acreage irrigated per farm is strongly correlated. Because data on irrigated acreage is probably better than on actual farm size, the irrigated acreage is used the proxy for farm size. Farm size tends to be large in districts with land-owner franchise. This relationship may reflect one of several possibilities. The first could be the desire of larger land owners to better influence district policies. However, the second one is that the more urbanized districts, which tend to have smaller farm operations, are required to use popular-vote electoral rules. Thus, the districts which can use land-owner enfranchisement will tend to have larger farms. Or the relationship may be simply geographical, reflecting the tendency of larger farms to be located in the Central Valley where almost all of the land-owner-enfranchised districts are located. One way to assess the possible source of this relationship is to isolate the analysis to the most agriculturally-dominated districts, irrigation and California water districts, and those located in the three Central Valley regions and the Inland Empire . Table 7 compares the means and correlation coefficients from all districts in the data set to those for irrigation and California water districts located in the Central Valley and Inland Empire. The results differ only slightly with the narrowing of the analysis, indicating the relationship between electoral rules and average farm size appear to be invariant with urbanization or location. This relationship appears to be most consistent with the first proposition that large land owners prefer an electoral system in which they can wield greater direct political influence.
The relationship of popular-vote districts and appropriative rights probably is indicative of the fact that these districts were formed before land-owner enfranchised districts since these districts would be better able to access appropriative rights. That Central Valley Project Exchange contractors, who relinquished their appropriative and pre-1914 rights to the u.s. Bureau of Reclamation in exchange for favorable water supply contracts, also slightly tend to be popular-vote districts is consistent with this observation. Receiving water project supplies, either from the CVP or SWP is negatively correlated with infrastructure size. This might reflect the fact that much of the delivery infrastructure for these contractors is paid for through project charges rather than by direct district investment. Irrigation efficiency generally has the expected strong positive correlations with orchard and nursery crops, which have the highest product value per acre , and negative with field and pasture crops. Produce crops, such as vegetables, berries and melons, show no relationship with irrigation efficiency, which is somewhat surprising given the relative output value per acre. The relationship of crop patterns and irrigation efficiencies to electoral rules also is interesting. Orchard crops tend to be located in districts using popular-vote rules while field crops tend to be in land-owner enfranchised districts. In addition, larger farms tend to grow more field crops, consistent with the fundamental economics of these various crops. As a result of these two relationships, irrigation efficiency is positively correlated with popular-vote rules. At first glance, this would seem to be inconsistent with Proposition 6 which states that popular-vote districts will tend to set lower water-use charges,black plastic plant pots bulk which in tum should encourage lower, not higher, efficiencies. However, Green,et al , found that water pricing had a relatively small effect on irrigation choices. According to Proposition 5, if orchard fanning requires the use of more local inputs such as equipment, fertilizer and labor relative to field crops, then district managers will tend to set rates which encourage this crop choice. This is consistent with past findings that orchard crops have substantially higher employment rates per acre-foot of water applied and a regional economic analysis of the Sacramento Valley found a higher ratio of in-region purchases for the “fruit and nuts” sub-sector than for “feed grains” . The improvement in irrigation efficiency would simply be a byproduct of this tendency toward local-input-intensive crops in popular vote districts. A set of eight propositions are developed from analyzing the theoretical model presented here. Propositions 1 and 2 define decisions rules for a theoretical constrained optimal cooperative. Proposition 3 compares the conditions under which an assessed-value-weighted voting district will arrive at the same decision rules as constrained optimal cooperative. The subsequent five propositions present hypotheses which could be tested with empirical data and analysis. However, the presently available data is only sufficient to test to of the propositions, Propositions 5 and 6. Some preliminary inferences can be drawn for Propositions 4 and 8, and the data is sufficiently confounding preclude any assessment of Proposition 7. This study sets out a sense of propositions about how governance rules affect the management incentives and decisions in special districts that supply water to agricultural customers.
The first two propositions set out decision criteria that a aggregate net-wealth maximizing cooperative facing a non-profit budget constraint would use to determine the optimal level of water -use charges and property-based taxes and assessments. A third proposition says that a water district which uses Land-owner-franchise / assessed-value-weighted voting rules, under conditions consistent with empirical economic data, will also tend to use these rules because this voting scheme is consistent with incentives and benefit distribution in the constrained cooperative. A fourth proposition states that larger landholders will tend to prefer relatively higher water use charges than smaller landowners. The empirical analysis contradicted this statement, but this may have resulted from one of two causes. The data set was disproportionately drawn from districts which contract with the CVP for water supplies. USBR rules require that “farms” be smaller than 960 acres to receive the lowest-prices supplies, so these districts show smaller farms, which in fact may be managed jointly in larger “management units.” A second cause might be from an economy of scale for conveyance to larger farms. This scale economy may be decreasing per farm delivery costs faster than the desire of larger landowners to see water-use rates rather than property taxes. The next two propositions compare incentives for managers between AVV and universal franchise / popular-vote rule districts. The fifth proposition says that managers in PV districts will tend to set water rates to encourage greater use of local inputs for farming, and as a result foster the growth of local-resource-intensive crops, such as fruit and nut trees. Econometric analysis supports this proposition. The sixth proposition makes a fundamental comparison of how much district managers rely on water sales to cover district expenditures. Empirical analysis of the hypothesized model supports the proposition that PV districts will tend to rely less on water sales than AVV districts. The seventh proposition states that as irrigation efficiency increases in a PV district, managers will tend to rely more on property taxes and assessments. The complicated relationship of irrigation choice and institutional structure could not be disentangled using the data available here. In the last proposition, two conditions were set out for when a PV district might have more or less investment in water-supply infrastructure than an AVV district. While not analyzed empirically, the data could be supplemented to assess the likely type of scale economies that these decision rules imply. In general the empirical analyses support the propositions that the rules governing district elections influence the decisions that district board members and managers make. These differences in institutionally-derived incentives have several important policy implications. First, AVV districts are more likely to rely on water sales and water-use charges. Given that the recent trend to encourage agricultural water conservation through increased water rates, , this means that AVV districts will be more likely to adopt these types of measures. That West lands and Broad view Water Districts, which are AVV districts, are at the forefront in adopting agricultural BMPs is consistent with this finding. Conversely, PV districts, such as irrigation districts, are likely to be more resistant to adopting BMPs, particularly ones that shift district revenues toward water sales. Another implication is that AVV districts are likely to be willing to participate in water transfers outside of the district boundaries. These districts’ members view water sales revenues, no matter the source, as beneficial. PV districts are more likely to encourage input-intensive orchard crops. This means that local communities are more dependent on agricultural activity for their livelihood. These crops also tend to use more efficient irrigation technologies. These two effects tend to amplify the local influences from water transfers out of the district. These operations cannot easily reduce their water use due to the already high levels of efficiency without either fallowing or turning to groundwater. If either the land is fallowed or water costs increase, use of local resources is likely to decrease. Because of the tighter local linkage, this reduction will be felt more severely in these PV district communities. The empirical analysis presented here is somewhat limited in scope. It assesses only one of the propositions developed in this study and it looks at data from only one year, the 1991-1992 fiscal year.