Economic drought management index to evaluate water institutions’ performance under uncertainty*

Reservoir management and intertemporal water allocation are critical issues in semiarid regions where agriculture has to confront highly variable rainfall patterns. In this paper, we derive and propose an economic drought management index (EDMI) to evaluate water institutions’ performance to cope with drought risk. The EDMI is based on the optimal conditions of a stochastic dynamic optimisation problem that characterises reservoir management. The index's main advantages are its ease of interpretation and breadth of scope, as it incorporates information on hydrological processes, structural constraints, water institutions’ rules, and the economic benefits of water use. An empirical application is developed to assess the institutional rules governing water allocation in two different supply systems in Andalusia (southern Spain).     

Evaluation of the Influence of Agricultural Price Policy on Cotton Production in Côte-d'Ivoire

In this paper the impact of the cotton pricing policy of the marketing board of Côte-d'Ivoire is investigated. A theoretical model is developed to derive the price level which maximizes revenue generated from the marketing board's cotton transactions. A dynamic cotton supply function was estimated and used to compute elasticities. The estimates indicate that farmers in the Country response to price changes and that the pricing policy of the marketing board has been consistent with an objective of generating revenue for the government.     

OPTIMAL BUFFER STOCK POLICIES FOR WHEAT AT THE WORLD LEVEL

Alternative wheat storage policies which maximise the expected present value of returns for consumers, producers, a monopoly storage agency and society as a whole are derived using a dynamic programming model. Results are compared with those from an earlier simulation model, and are found to justify higher investment in storage capacity compared with that suggested by the simulation model. The model is extended to derive optimal storage policies if production follows a stochastic cobweb process.     

Agro-Climatic Conditions and Regional Technical Inefficiencies in Agriculture

A survey of applications of the Technical Inefficiency Effects (TIE) model suggests that agro‐climatic and other environment variables are customarily omitted in the model specifications. The justification for such an omission is the assumption that these variables are beyond the control of the farmers and therefore should be treated as random variables. In this paper, we argue that in applications dealing with regional agricultural data, agro‐climatic variables should not be treated as pure random terms. Historical differences in agro‐climatic conditions are known with a reasonable degree of certainty across a larger region. Therefore, omission of such variables from the analysis may lead to inaccurate interregional technical inefficiency comparisons. In order to demonstrate the importance of agro‐climatic variables in such analyses, we estimate the TIE model for Turkey. A translog stochastic frontier production function with agro‐climatic variables such as rainfall and land quality is estimated, and it is shown not only that the agro‐climatic variables are statistically significant but also that their omission substantially affects mean output elasticities and relative technical efficiencies. Une étude sur les applications du modèle de l'effet d'inefficacité technique (EIT) laisse à supposer que les variables agro‐climatiques et les autres variables environnementales sont comme d'habitude omises dans les spécifications du modèle. Une telle omission est justifiée par l'hypothèse selon laquelle ces variables sont en dehors du contrôle des fermiers et devraient être considérées comme des variables aléatoires. Dans ce communiqué, nous affirmons que dans les applications concernant les données agricoles régionales, ces variables agro‐climatiques ne doivent pas être traitées comme de simples termes aléatoires. Les différences historiques dans les conditions agro‐climatiques sont connues avec un degré raisonnable de certitudes pour une grande région. Aussi l'omission de telles variables dans l'analyse peut‐elle donner lieu à de fausses comparaisons interrégionales d'inefficacité technique. Afin de démontrer l'importance des variables agro‐climatiques dans de telles analyses, nous considérons le modèle de l'effet d'inefficacité techniques de la Turquie. II s'agit d'une fonction de production frontalière translogue et stochastique avec des variables agro‐climatiques telles que la pluviosité, la qualité de sol et d'autres variables. Nous démontrons que les variables agro‐climatiques sont non seulement importantes statistiquement, mais que leur omission influence essentiellement les élasticités moyennes de production ainsi que les efficacités techniques relatives.     

Optimisation of water procurement decisions in an irrigation district: the role of option contracts

Water supply instability is one of the main risks faced by irrigation districts and farmers. Water procurement decision optimisation is essential in order to increase supply reliability and reduce costs. Water markets, such as spot purchases or water supply option contracts, can make this decision process more flexible. We analyse the potential interest in an option contract for an irrigation district that has access to several water sources. We apply a stochastic recursive mathematical programming model to simulate the water procurement decisions of an irrigation district's board operating in a context of water supply uncertainty in south‐eastern Spain. We analyse what role different option contracts could play in securing its water supply. Results suggest that the irrigation district would be willing to accept the proposed option contract in most cases subject to realistic values of the option contract financial terms. Of nine different water sources, desalination and the option contract are the main substitutes, where the use of either depends on the contract parameters. The contract premium and optioned volume are the variables that have a greater impact on the irrigation district's decisions.     

Stochastic Rangeland Use under Capital Constraints

The stocking density on Kazakhstan's extensive rangelands is well below traditional levels. To analyze dynamic flock performance, we develop a stochastic dynamic programming model for livestock systems with stochastic forage production. The model contains continuous five state and 12 control variables, allowing improved characterization of the biophysical relationships and economic tradeoffs inherent in such systems. Most Kazakhstan herders have restricted access to capital. The model indicates that the cost of capital strongly affects flock size and productivity. We conclude that capital constraints are important to explaining the current low stocking density. Improving capital markets in rural areas warrants policy attention.     

Risk aversion and land allocation between annual and perennial crops in semisubsistence farming: a stochastic optimization approach

This article analyzes the effect of production uncertainty on farmland allocation decisions between perennial and annual crops, focusing on a representative farmer's attitude toward risk. A dynamic stochastic optimization model that considers net planting—the difference between new plantings and removals of perennial crops that achieve full production cycle—is used. The effect of uncertainty on the representative farmer's decisions to increase or decrease perennial crops’ acreage, on the optimal path, is examined. Our results reveal that the response of optimal path of net planting to uncertainty related to perennial crop production is highly affected by the farmer's attitude toward risk. A risk‐averse or a low‐risk loving farmer tends to reduce land allocation to perennial crops under uncertainty, while a high‐risk loving farmer will do exactly the opposite. Also, due to disutility of farming, the farmer tends to reduce land allocation to perennial crops when prices are high enough for him to attain a desired income level expectation. One implication of this research is the need for mechanization—in sub‐Saharan countries in particular—that increases per‐acreage yield and output in semisubsistence agriculture.     

Impacts of considering climate variability on investment decisions in Ethiopia

Extreme interannual variability of precipitation within Ethiopia is not uncommon, inducing droughts or floods and often creating serious repercussions on agricultural and nonagricultural commodities. A dynamic climate module is integrated into an economy‐wide model containing a detailed zonal level agricultural structure. This coupled climate‐economic model is used to evaluate the effects of climate variability on prospective irrigation and infrastructure investment strategies, and the ensuing country‐wide economy. The linkages between the dynamic climate module and the economic model are created by the introduction of a climate‐yield factor (CYF), defined at the crop level and varied across Ethiopian zones. Nine sets of variable climate (VC) data are processed by the coupled model, generating stochastic wet and dry shocks, producing an ensemble of potential economic prediction indicators. Analysis of gross domestic product and poverty rate reveal a significant overestimation of the country's future welfare under all investment strategies when climate variability is ignored. The coupled model ensemble is further utilized for risk assessment to guide Ethiopian policy and planning.     

SUPPLY RESPONSE TO PRICE UNDER CERTAINTY,WITH REFERENCE TO THE UNITED KINGDOM EGG MARKET, 1954 TO 1966

Hen egg production in the United Kingdom expanded throughout the 1950's and early 1960's in the face of successive reductions in the guaranteed price thus raising questions about the efficacy of the guaranteed price system as a production control mechanism. The objective of this analysis is to estimate the guaranteed price elasticity of supply of hen eggs over the twelve-year period 1954 to 1966. Two models are used, a conventional linear form and a Nerlove type lagged model. The independent variables considered are the guaranteed price, the price of feed, and egg yield as a proxy for technology. The non-lagged equation is the more satisfactory and gives an elasticity value of 1.66, suggesting that flock owners were highly sensitive to changes in the guaranteed price. This sensitivity tended to be obscured by producer eagerness to profit from technological advance, which was the predominant factor affecting the scale of egg production over the period.     

A stochastic dynamic programming framework for weed control decision making: an application to Avena fatua L.

This paper develops a stochastic multi-period decision model to analyse a continuous wheat cropping system infested by wild oats (Avena fatua L.), in southern Australia. The multi-period solutions is obtained by employing a dynamic programming model in conjunction with a bioeconomic simulation model. An empirically estimated dose response function is used to derive the optimal herbicide rate. Uncertainties due to environmental effects on the performance of herbicide and crop yields are modelled and optimal decision rules derived. The results indicate that substantial economic gains can be realised if herbicide dose decisions are taken by considering future profit effects of current decisions, as opposed to the more common approach of only considering the current-period effect.     

Stochastic frontier analysis of productive efficiency in China's Forestry Industry

Forest resources are vital to the development of green economics. Given the booming development of China's forestry industry and its ambitious reforestation efforts in the developing world, this paper is the first to use the output distance function to synthetically consider the economic and ecological outputs of China's forestry industry, and discuss its productive efficiency with a stochastic frontier model. Control and environmental variables are incorporated to capture heterogeneity in China's forestry industry, which helps us get an unbiased estimation. The empirical results show that there was no obvious efficiency disparity among China's economic regions except Northeastern China, and the state-owned forestry structure has a significantly negative effect on productive efficiency in China's forestry industry. Moreover, provinces with poor productive performance in the forestry industry such as Inner-Mongolia, Heilongjiang, and Hebei have been identified and their individual characteristics regarding productive efficiency have also been analyzed. The findings in this paper have targeted and practical implications for the development of China's forest green economy.     

A Maximum Likelihood Cointegration Analysis of Canadian Lumber Exports

Johansen 's cointegration analysis is used to investigate causal relationships between Canadian lumber exports, U.S. lumber price, U.S. housing starts and the Canada-U.S. exchange rate. The results suggest that there is only one stable equilibrium relationship among these variables in the long run. Forecast error variance decomposition and impulse responses from a vector error-correction model are used to examine the short-run dynamic relationships among the variables. The results suggest significant dynamic relationship among the variables. The results also indicate that the pass-through of exchange rate is incomplete in the long run and that the effects of exchange rate on lumber exports vary between the short run and long run L'analyse de cointégration de Johansen est utilisée pour examiner les rapports de cause à effet entre les exportations de bois d'oeuvre canadien, le prix du bois d'oeuvre aux Etats-Unis, les mises en chantier de constructions à usage d'habitation el le taux de change Canada-Etats- Unis. Les résultats semblent montrer qu'il n 'v a à long terme qu'un seul rapport d'équilibre stable entre ces variables. On recourt à la decomposition de la variance de l'erreur de prédiction et aux réponses sur impulsion déhvées à partir d'un modèle de correction d'erreur vectorielpour examiner les rapports dynamiques à court terme entre les variables. Les résultats laissent apparaître un rapport dynamique significatif entre les variables. Ils montrent aussi que la transmission des effets du taux de change canado-américain est incomplète à long terme et que les effets de cette variable sur les exportations canadiennes de bois d'oeuvre ne sont pas les mêmes à court qu'à long terme     

AN EMPIRICAL INVESTIGATION OF THE HARVEST OPERATION USING SYSTEMS SIMULATION

A simple model of the cereal harvest operation was constructed for the Wimmera region of Victoria. The model was used to investigate factors influential in determining the harvesting costs of a machinery system over a period of years. An evaluation of alternative machinery systems was also conducted using the model. The superiority of systems simulation over static analysis was demonstrated by using the model to incorporate stochastic variables and to handle the dynamic nature of the harvest operation. In addition, the model output was shown to provide more information for decision makers than had resulted from previous Australian studies. It was concluded that systems simulation is an appropriate technique for investigating farm management machinery selection problems.     

Dynamic Programming and Learning Models for Management of a Nonnative Species

Nonnative invasive species result in sizeable economic damages and control costs. Because dynamic optimization models break down if controls depend in complex ways on past controls, nonuniform or scale‐dependent spatial attributes, etc., decision‐support systems that allow learning may be preferred. We compare two models of an invasive weed in California's grazing lands: (i) a stochastic dynamic programming model and (ii) a reinforcement‐based, experience‐weighted attraction (EWA) learning model. We extend the EWA approach by including stochastic forage growth and penalties for repeated application of environmentally harmful controls. Results indicate that EWA learning models offer some promise for managing invasive species. Les espèces non indigènes envahissantes entraînent des dommages économiques et des coûts de lutte considérables. Compte tenu que les modèles d'optimisation dynamique échouent lorsque les moyens de lutte dépendent, de façon complexe, de moyens de lutte antérieurs, d'attributs spatiaux influencés par l'échelle ou non uniformes, etc., l'utilisation de systèmes d'aide à la décision permettant l'apprentissage pourrait être préférable. Nous avons comparé deux modèles dans le cas d'une plante adventice envahissant les pâturages de la Californie: 1. un modèle de programmation dynamique stochastique; 2. un modèle d'apprentissage experience‐weighted attraction (EWA), fondé sur le renforcement. Nous avons élargi le modèle EWA en y incluant la croissance stochastique des fourrages et des pénalités imposées pour l'utilisation répétée de moyens de lutte dommageables pour l'environnement. Selon les résultats obtenus, les modèles d'apprentissage EWA semblent prometteurs pour la gestion des espèces envahissantes.     

Econometric analysis of noncompliance with organic farming standards in Switzerland

Applying the economics of crime theory, we model the decision of an opportunistic and/or careless organic farmer and derive hypotheses to explain noncompliance. Where empirical data are available, hypotheses are tested. We use data for the years 2007 through 2009 of organic farms certified by Bio Suisse. Imposed sanctions are used as a proxy variable for noncompliance and farm characteristics as explanatory variables. Random effects logit models show that processing activities and livestock diversity significantly increase a farm's sanction probability. Past noncompliances also indicate a higher present sanction probability. Finally, we discuss some methodological issues and suggest a way to organize risk‐based inspections more effectively.     

The Economics of Managing Infectious Wildlife Disease

We use a two-state linear control model to examine the socially optimal management of disease in a valuable wildlife population when diseased animals cannot be harvested selectively. The two control variables are nonselective harvests and supplemental feeding of wildlife, where feeding increases both in situ productivity and disease prevalence. We derive a double singular solution which depends on the initial state and does not require bang-bang controls. The case of bovine tuberculosis among Michigan white-tailed deer is analyzed. In the base model, the disease is optimally maintained at low levels, with intermittent investments (via feeding) in deer productivity.     

WELFARE IMPLICATIONS OF MORE ACCURATE RATIONAL FORECAST PRICES

Models are constructed to assess the welfare effects for producers, consumers and society of producers using forecast prices based on more accurate estimates of variables causing shifts in the demand for and supply of commodities. The basic model is a stochastic cobweb model in which producers' forecast price is the rational forecast price. The model is extended for many commodities, for partial producer response to more accurate forecast prices, and to include stock holding. In terms of economic surplus, producers and consumers gain from more accurate estimates of demand shift variables, producers gain and consumers lose from more accurate estimates of supply shift variables, and in both cases there is a net society gain.     

The Application of Multivariate Stochastic Dominance Criteria to Agricultural Economic Problems

This paper provides a discussion of multivariate risk analysis and multivariate stochastic dominance criteria. The assumptions and conditions required to extend the use of stochastic dominance criteria to multivariate problems are examined. The assumptions of additive utility or statistical independence are required to apply true multivariate stochastic dominance criteria. Implications for the application of several multivariate stochastic dominance criteria to empirical problems are discussed. An empirical application is presented to demonstrate the importance of recognizing the assumptions underlying alternative multivariate stochastic dominance criteria. Alternative structural adjustments for a representative upper midwestern dairy farm are examined and assessed using multivariate stochastic dominance. The results of the analysis indicate that, if utility is assumed to be additive, dairy expansion and specialization may not be optimal. If utility is assumed to be defined in terms of overall wealth, however, dairy expansion and specialization are preferred to the present firm structure. Cet article analyse le risque des variables aléatoires multiples et les critères de prédominance des variables aléatoires stochastiques multiples. Les suppositions et les conditions requises pour étendre l'usage des critères de prédominance stochastiques aux variables aléatoires multiples sont étudiées. Les suppositions d'utilité additive ou d'indépendance statistique sont exigées pour appliquer les critères de prédominance des variables aléatoires stochastiques multiples. L'étude examine également les implications de l'application de plusieurs critères de prédominance des variables aléatoires stochastiques multiples sur des problèmes empiriques. Une application empirique est utilisée pour démontrer qu'il important de reconnaître les suppositions qui sont à la base des critères alternatifs de prédominance des variables aléatoires stochastiques multiples. Les variables aléatoires multiples de prédominance stochastique sont utilisées pour examiner et évaluer des ajustements structurels alternatifs pour une ferme laitière representative du haut-midwest. Les résultats de l'analyse indiquent que, dans les conditions d'utilityé additive, l'expansion et la spécialisation des fermes laitières pourraient ne pas être optimales. Par contre si l'utilityé est défine en termes de fortune, l'expansion et la spécialisation laitiéres sont préferables à la structure actuelle des fermes.     

Optimal Integrated Strategies to Control an Invasive Weed

Although there have been numerous studies on the economics of weed control in agriculture, relatively few studies have focused on weeds in natural ecosystems. A stochastic dynamic simulation model and a stochastic dynamic programming model are developed to: (A) identify the combination of control options that is optimal for blackberry (Rubus anglocandicans) in Australian natural ecosystems, (B) assess whether an integrated control strategy is superior to chemical‐only strategies, (C) evaluate the net benefits of biological control (a rust and grazing by goats), and (D) determine how changes in model parameters affect the optimal control strategy. The results indicate that, while an integrated strategy combining chemical and nonchemical control methods may be optimal in certain circumstances, it is not necessarily superior to a chemical‐dominant strategy in all cases. The results show that grazing goats for control of blackberry can be optimal despite uncertainty about its effectiveness. Policy makers need to be aware of the trade‐offs between choosing more effective control options that may impose environmental and health risks versus less effective control strategies that are safer to the environment and human health.     

Estimating random effects production function models with selectivity bias: an application to Swedish crop producers

In this paper, the estimation of production functions and measurement of the rate of technical change is performed when selectivity bias is expected. A sample selection model consisting of a selection and a regression equation is estimated using Heckman's two-stage method. It is discussed in the context of a production function where the underlying technology is represented by a translog functional form. For the regression, a random effects model with heteroscedastic variances is assumed. This model and an alternative conventional model retaining heteroscedasticity without considering selectivity bias are estimated using the Generalized Least Squares method. The data used are a large rotating panel data set from Swedish crop producers over the period 1976–1988. The empirical results from the comparison between these two models show that the introduction of heteroscedasticity and the integration of sample selection in the production relationship is important. The impact of a correction for selectivity bias on the results, in terms of input elasticities and returns to scale is found to be significant.