While maintaining base year land productivity would have reduced the number of people in both poverty and extreme poverty, the depth and severity of poverty relative to the extreme poverty line would have increased. The counterfactual increase in poverty severity indicates that observed changes in land productivity over the period of study were beneficial for the poorest of the poor. More generally, the impacts of changing land productivity on poverty are heterogeneous, affecting the moderate and extremely poor in distinct ways.Estimating a stochastic production frontier according to through forms the basis for the decomposition of land productivity. The model is estimated using the sfcross command in Stata, with community-level fixed effects and standard errors clustered at the community level. The lambda, or the ratio of the variance of the inefficiency term to the variance of the error term, is 1.52, indicating that a stochastic frontier model is appropriate. The model is consistent with that of chapter 2, in that it finds evidence of positive technical change in the frontier and increasing technical inefficiency over time. See Appendix C.3 for further details of the model results. Figure 3.5 shows kernel density estimates for the technical inefficiency scores for each year, showing the growth of technical inefficiency over time.Table 3.5 begins to unpack the change in land productivity into its constituent sources by displaying the results of the decomposition described in . Again, this decomposition is akin to that of Oaxaca and Blinder, grow strawberry in containers expressing changes in average land productivity as changes in average observable characteristics and differences in model parameters.

When evaluated at the mean, technical change had a positive contribution to changes in land productivity during the period. Growing average technical inefficiency and the growth in average farm size largely offset that contribution, contributing most to the decline in land productivity. In comparison, changing average input intensities and a changing farm size – frontier productivity relationship contributed relatively little to the observed decline in average land productivity. From a productivity perspective, this suggests that policy helping family farms with the adoption of new technologies and the efficient use of existing resources poses a significant opportunity. The average decompositions, however, do not speak to the heterogeneity with which these productivity sources influence the productivity of family farms. Table 3.6 displays the average change in land productivity attributable to each source, by decile of land productivity in the base year. The changes in land productivity are based upon the counterfactual land productivity distributions generated according to , and follow Shorrocks in using the average change across all possible orders of decomposition. These average changes for each farm from each productivity source are then averaged over deciles. The corresponding counterfactual land productivity distributions associated with each productivity component can be seen in Appendix C.4. An immediate implication of this approach is that the observed change in land productivity exhibits considerable heterogeneity across deciles. As shown in the final column of Table 3.6, those farms with the lowest land productivity in 2002 realized notable gains in land productivity on average, whereas the most productive deciles averaged pronounced losses in productivity. Whereas changes in farm size and inefficiency were contributing to productivity losses for the most productive farms, they contributed to productivity gains among the least productive farms. Technical change, in contrast, had a relatively homogenous and positive relationship with land productivity.

It is clear that going beyond the mean is necessary to assess the links between agricultural productivity and on-farm poverty, and that the channel for productivity gains will likely matter. Tables 3.7 and 3.8 display the counterfactual estimates of the contribution of productivity sources to poverty, in terms of percentage and percentage point changes, respectively. These counterfactual estimates are the average changes to poverty if the productivity source maintained their 2002 levels, averaged over all possible paths of decomposition. For example, the results for technical change are interpreted as the counterfactual estimates of the difference in poverty that would have been realized if there were no technical change during the period – regardless of poverty measure or poverty line used, poverty would have been 4-16% higher absent technical change, depending upon the measure. The changes to the farm size distribution, i.e. the fall in small farm size and the rise in large farm size, contributed to increasing poverty. Absent these changes, poverty would have been 5-9% lower. In conclusion, whereas technical change contributed positively to growth in land productivity and poverty alleviation, changes to the farm size distribution detracted from average land productivity and contributed to an increase in poverty. The relationship between changing technical inefficiency, land productivity, and poverty, appears more nuanced. Yes, productivity would have been higher, on average, and poverty lower in the absence of growing technical inefficiency, but the contribution to changes in poverty is both mixed and relatively muted. Whereas poverty rates would have fallen absent changes to the technical inefficiency distribution, poverty depth and poverty severity relative to the extreme poverty line would have increased. Changes to input intensity on farms made a notably smaller contribution to the decline in average land productivity than did changes to technical inefficiency, but the direct effect on poverty among family farms was more pronounced.

This points towards agricultural productivity growth through intensification and technical change being notably more propoor than technical efficiency in terms of the direct contribution to on-farm poverty alleviation. Using a 2002-2009 panel of family farms in Mexico drawn from the Mexican Family Life Survey, this paper contributes to understanding the linkages between productivity and poverty by estimating the direct contribution of changing agricultural productivity to changes in on-farm poverty. The study finds declining average land productivity over the sample period, and while poverty rates increased among family farms, poverty severity declined. Decomposing changing agricultural productivity into five sources – changes in technical efficiency, technical change, input intensification, farm size, and the farm size frontier productivity relationship – the analysis finds increasing inefficiency and changes to the farm size distribution were driving the decline in land productivity in spite of notable technical change. The counterfactual analysis finds evidence of a land productivity – poverty point elasticity of approximately -0.26 to -0.29; poverty would have been approximately 4% lower if land productivity had not changed. Further, the counterfactual analysis suggests that raising land productivity through intensification and technical change would have a larger direct contribution to alleviating on-farm poverty than would increasing technical efficiency. This study can be refined and extended in several ways. First, the counterfactual estimates could be improved with further refinement of the sample’s income variable. There appear to be several areas where marginal components may not have been included, such as on remittances data and profits from non-agricultural businesses. There are other areas where some double-counting of income may be occurring. Similarly, the prices used to value agricultural output are common, taken from the FAO. Whereas a common set of prices was a boon to productivity analysis, where common prices aid comparability, hydroponic nft channel heterogeneous output prices are likely important when assessing incomes, poverty and livelihoods on family farms.There are several important extensions of this line of work. First, a third wave of the survey, conducted in 2005, has not been utilized in this study. Leveraging this third survey year may prove interesting, as the literature has shown evidence of a decline in poverty between 2002 and 2005, with an increase in poverty in the following years. Second, the richness and breadth of the survey make it possible to extend the analysis into alternative measures of poverty. An analysis and comparison of consumption-based poverty, nutrition-based poverty, child poverty, and multidimensional poverty are all possible. Third, a methodological extension may involve utilizing the non-parametric decomposition techniques pioneered in labor economics to complement the parametric decomposition presented here. Lastly, this study should be extended to a more general decomposition of poverty into its constituent components.As climate change advances, the future of agriculture in California becomes more uncertain. For decades, farmers have adapted their practices to adjust in response to unexpected shocks in temperature and precipitation patterns. Additionally, weather conditions in the American Southwest are expected to continue to change, creating conditions that are less hospitable to farming in the coming years . Groundwater is the most crucial buffer resource for irrigated crops during times of unforeseen drought, allowing farmers to have more consistent watering when crop cover decisions cannot be adjusted. Water use is dominated by agriculture in drought or drier years statewide, accounting for over 60% of total water application in a representative dry year . Even in years when precipitation is plentiful, water resources in the state’s agrarian Central Valley are overwhelmingly diverted to agricultural uses.

The percentage of land experiencing severe drought conditions has increased significantly in California since the turn of the century, and will likely not improve in the coming decades . Farmers in California increasingly rely on groundwater for their operational needs, while water table water levels throughout the state continue to trend downward . Depleting water supplies mean higher pumping costs for farmers; as a result of the 2012-2016 drought, farmers experienced an estimated $600 million increase in pumping costs per year . High water costs often prompt farmers to fallow parts of their usable cropland for one or more growing seasons, wasting valuable land and leading to loss of profits and reduced agricultural production . The Sustainable Groundwater Management Act , passed by the California state government in 2014, adds additional pressure to the agriculture sector to reduce water consumption by implementing sustainable water use requirements for regions that derive their water from over drafted basins. The San Joaquin Valley , the lower two-thirds of the Central Valley, is one of these affected regions. The San Joaquin Valley is defined by the two hydrologic regions it is comprised of: San Joaquin River and Tulare Lake, shown in figure 1a above. Figure 1b shows land parcels that have been active for at least one growing year 2008-2023. Most of the irrigated land acreage is concentrated in the western side of the valley, where surface water runoff from the surrounding mountain ranges feeds crops. SGMA classifies 13 of the 19 total groundwater basins that supply the SJV as high priority, and therefore subject to more stringent regulations regarding groundwater usage . Moreover, over half of the entire state’s basins classified as the most severely depleted are in the SJV. Due to these new regulations, it is projected at least 500,000 acres of existing irrigated land fed by the affected basins will have to be removed from irrigation by 2040 to meet sustainable water extraction requirements . The necessary reduction in irrigated farmland acreage could be as much as 1 million acres, meaning around 10-20% of the 5 million acres of existing farmland in the SJV may be unable to be irrigated after 2040 as a result of groundwater regulations alone. Farmers operating agricultural land that may eventually be removed from irrigation, or that has become prohibitively expensive to water, could transition to solar energy production as a profitable alternative to crop cultivation. In many ways, agricultural land is ideal for large-scale solar production. Farmland has often been made relatively flat, receives direct sunlight for multiple hours a day, and provides large swatches of land to build necessary infrastructure. Agricultural land makes utility solar production feasible and more likely to be profitable due to the abundance of sunlight and relatively low cost of land. From a social planner’s perspective, crop-to-energy land transition would be especially ideal for a variety of reasons. Firstly, farmers could turn their idle land into financially productive land, thereby increasing their overall welfare. This would, in turn, decrease irrigated land acreage and avoid other negative environmental externalities that result from agricultural cultivation, like pesticide runoff and fertilizer leaching. Taking irrigated farmland out of crop production would simultaneously help California achieve it’s groundwater management an environmental protection goals. Finally, providing new sources of renewable, zero carbon energy is exceptionally beneficial to California’s clean energy goals. Two recent laws passed in California have encouraged planning and investment in solar energy statewide. Senate Bill 100 , passed in 2018, requires that 100% of electricity sales are renewable or zero-carbon by 2045.