It should be noted that beta convergence doesn’t imply sigma convergence and that one looks for beta convergence in testing whether poorer regions are catching up with the wealthier regions. In Table 1, we provide some major papers related to convergence test and how different tests have been applied to test for agricultural total factor productivity convergence in literature. Additionally, literature in convergence initially suffered from clarity of definition, and it led to wrong tests and wrong conclusions . The structure of convergence literature in its initial phase therefore looks like an expedition, where researchers looked for convergence at any country group they could find, using whichever method seemed appropriate at the time. The literature has evolved from identifying convergence in relatively global block to the convergence in regions or a country. Pioneering authors in this literature such as Baumol; Barro; Barro and Sala-i-Martin; De Long; Islam; and Mankiw, Romer and Weil found some form of convergence among OECD countries. Besides those classical studies, some other studies related to the agricultural productivities have been carried out. For example, Thirtle et al. calculated multi-factor agricultural productivity indices for Botswana’s agriculture. Results obtained using a unit root test indicated that there was no regional convergence in agricultural productivity.

Gyawali et al. analyzed income convergence behavior of population in Alabama’s black belt region and found to have a conditional convergence among different census blocks. Garofalo and Yamarik estimated regional convergence by creating a state-by-state capital stock series. This study reconciled the growth empirics’technique of Mankiw, Romer, and Weil as well as with the empirical results of Barro and Sala-i-Martin using the new database covering 1977 96. The results indicated a convergence at the rate of 2% and suggested that the Solow’s neoclassical growth model drives the empirical results of Barro and Sala-i-Martin. In the empirical front,nft hydroponic modeling and testing convergence hypothesis is also the subject far from being settled. Lichtenberg believes that the hypothesis of convergence and mean-reversion are not equivalent and asserts that the lowest initial productivity level followed by the highest subsequent productivity growth does not automatically imply convergence.We compared our results with the existing result from Mccunn and Huffman, which tackles the question we explored in a different way. Our research differs on methodology and clusters formation from theirs. Results from McCunn and Huffman’s approach are presented in Table 6. When we look at regional data, there seems to be some evidence against the null hypothesis of no convergence in these particular regions. The results show Cornbelt and Lake States having a negative and statistically significant parameter estimate for time variable t; suggesting convergence is taking place in these regions.We also ran cluster analysis in which we attempted to find 10 clusters from the data following McCunn and Huffman’s original selection of 10 groups.

When we used data based clustering to find ten clusters from data, the number of states and states within each cluster were quite different from McCunn and Huffman’s original 10 groups and the states included in those groups . For example, in our analysis of 42 states, we found that there are three clusters containing only one state. A cluster identified by the cluster analysis contained 14 states. Compared to McCunn and Huffman, the maximum number of states in a group was eight . In our analysis of 48 states, we found three clusters containing only one state; however, only two clusters contained the same state. In the case of 42 states, we had ND, FL and WV in each cluster itself where as in the case of 48 states, we had IA, FL and WV in a cluster itself. The convergence test is therefore conducted for only seven clusters in both cases. What we have consistently observed in all cases is that FL and WV are unique states that are unlike other states in TFP growth. The results for regional convergence test based on 10 clusters and 42 states are presented in Table 8. The results indicate no regional convergence although three clusters had correct signs associated with the parameter. In two clusters , the sign was positive and significant, indicating that there is a divergence in TFP among states in these two clusters. One way to explain it is by using Bertola’s logic: that in a situation in which technological advances are highly localized and its diffusion is slow, one may see persistent difference or divergence in regional or national productivity. The results for 10 clusters and 48 states are also presented in Table 8. We found that six clusters possessed a correct sign associated with parameter, although none of those were found to be significant. Only one cluster had an opposite sign which indicates divergence among the states within the cluster.

Human capital has been described as the contributor of growth in Mankiw, Romer and Weil, Lucas, and Shultz. Recent researches on total factor productivity convergence are emphasizing the needs for considering human capital as a factor of growth. For example, Miller and Upadhyay found that human capital has a significant impact on output when it is included as a factor of production. Human capital, when considered as an input, lowers the labor elasticity of output when compared to the production function without human capital. Similar findings were shown in a study by Coulombe and Tremblay. Their analysis indicated that in an open economy with perfect capital mobility, the dynamics of human capital accumulation is the driving force behind the economic growth. According to them, in the process of convergence, physical capital accumulation is driven by accumulation of human capital and per capita income disparities across economies are explained by disparities in human capital stock. Their results indicated that advance education indicator explains roughly 70% of the relative evolution of per-capita income since 1951 across the Canadian provinces. Similarly, Maudos, Pastor, and Serrano had developed Malmquist indices of productivity including human capital as an additional input. Their results indicated the existence of a significant effect associated with human capital and its importance for an accurate measurement of T F P. We take these results as our guide and explore if human capital can describe the disparities in agricultural total factor productivity differences over time across states as seen above in our result. The results from the panel data model are shown in Table 9. We estimated one way fixed effects and two ways fixed effects models. In the one way fixed effects model, we assumed that agricultural productivity differences are caused by state heterogeneity in human capital. The result from the fixed effect model indicates that human capital does play a significant role in determining the total factor productivity. The coefficient associated with human capital in this model is significant at a level of 1%. R2 from the model is 97% indicating that human capital is able to explain most of the difference in productivity difference. Hausman’s test indicated that we failed to reject the state level homogeneity in agricultural total factor productivity. The coefficients associated with each state were found to be significant. The highest coefficient is associated with the state of Florida. The results from the two way fixed effects model indicated similar results, but the coefficient associated with human capital is found to be insignificant. Hausman’s test statistics rejected the homogeneity of the state specific parameters in the model. Results from the random effects models also show the coefficient associated with human capital to be significant. The M-test indicates that we were unable to reject the presence of random effects in the models. In the absence of any assumption related to functional form between total factor productivity and human capital,nft system we should estimate the non-parametric model. The non-parametric model showed that smoothing parameter value equaling to 0:809 should be used to study the relationship. Figure 1 shows the prediction using the non-parametric model. The figure also shows the 90% confidence interval of the predicted value. The non-parametric model has a better fit as indicated by the residual sum of square from the prediction model.

Both parametric and non-parametric models thus show the correlation of human capital with the productivity. With the passage of Proposition 64 , state voters elected to integrate cannabis into civil regulation. The California Department of Food and Agriculture oversees state-licensed cannabis cultivation and defined it as agriculture . Prior to the possibility of state licensure for cultivators, however, counties can decide on other designations and implement strict limitations. In effect, local governments have become gatekeepers to whether and how cultivation of personal, medical or recreational cannabis can occur and the repercussions of noncompliance. When cannabis is denied a consistent status as agriculture, despite being a legal agricultural commodity according to the state, localities can determine who counts as a farmer and who is considered compliant, non-compliant and even criminal. In Siskiyou County’s unincorporated areas, the Sheriff’s Office now arbitrates between the effectively criminal and agricultural. Paradoxically for this libertarian county, the furor around cannabis has seen calls for government intervention, and has led to officials passing highly stringent cannabis cultivation regulations that have been enforced largely by law enforcement, muddying the line between noncompliance and criminality. These strict regulations produced a situation where “not one person” has been able to come into compliance, according to a knowledgeable government official. Nonetheless, at the sheriff’s urging, Siskiyou declared a “state of emergency” due to “nearly universal non-compliance” , branding cannabis cultivation an “out-of-control problem.” Such a strong reaction against cannabis can be understood in terms of cannabis’s potential to reorganize Siskiyou’s agricultural and economic landscape. According to some estimates, there are now approximately twice as many cannabis cultivators as non-cannabis farmers and ranchers in Siskiyou , a significant change from just a few years ago. Although cannabis has been cultivated in this mostly white county for decades, since 2015 it has become associated with an in-migration of Hmong-American cultivators. Made highly visible through enforcement practices, policy forums and media discourses, Hmong-Americans have become symbolically representative of the “problem.” This high visibility, however, obscures a deeper issue, what Doremus et al. see as a nostalgic, static conception of rural culture that requires defensive action as a bulwark against change. Such locally-defined conceptions need to be understood , especially in how they are defined and defended and what effects they have on parity among farmers growing different types of crops. Our goals in this study were to consider the consequences of an enforcement-first regulatory approach — a common regulatory strategy across California — and its differential effects across local populations. Using Siskiyou County as a case study, we paid attention to the public agencies, actors and discourses that guided the formation and enforcement of restrictive cannabis cultivation regulations as well as attempts to ameliorate perceptions of racialized enforcement. This study attends to novel post legalization apparatuses, their grounding in traditional definitions of culture and the ways these dynamics reactivate prohibition. We used qualitative ethnographic methods of research, including participant observation and interviews. In situations of criminalization, which we define not only as the leveling of criminal sanctions but being discursively labeled or responded to as criminal-like , quantitative data can be unreliable and opaque, which necessitates the use of qualitative ethnographic methods . In 2018–2019, we talked to a wide range of people — including cannabis growers from a diversity of ethnic backgrounds, government officials, business people, subdivision residents, farm service providers, medical cannabis advocates, realtors, lawyers, farmers and ranchers, and, with the assistance of a Hmong-American interpreter, members of the Hmong-American community. We also analyzed public records and county ordinances, Board of Supervisors meeting minutes and audio , Sheriff’s Office press releases and documents, related media articles and videos, and websites of owners’ associations in the subdivisions where cannabis law enforcement efforts have focused. Some cannabis cultivators regarded us suspiciously and were hesitant to speak openly, an unsurprising phenomenon when researching hidden, illegal and stigmatized activities, like “drug” commerce . This circumspection was most intense among Hmong-American growers on subdivisions, who had been particularly highlighted through enforcement efforts and local, regional and national media accounts linking their relatively recent presence in Siskiyou to cannabis growing.