Stochastic efficiency analysis with risk aversion bounds: a simplified approach

A method of stochastic dominance analysis with respect to a function (SDRF) is described and illustrated. The method, called stochastic efficiency with respect to a function (SERF), orders a set of risky alternatives in terms of certainty equivalents for a specified range of attitudes to risk. It can be applied for conforming utility functions with risk attitudes defined by corresponding ranges of absolute, relative or partial risk aversion coefficients. Unlike conventional SDRF, SERF involves comparing each alternative with all the other alternatives simultaneously, not pairwise, and hence can produce a smaller efficient set than that found by simple pairwise SDRF over the same range of risk attitudes. Moreover, the method can be implemented in a simple spreadsheet with no special software needed.     

Application of mean–Gini stochastic efficiency analysis

This article adds to the information base concerning the applicability of mean–Gini stochastic efficiency analysis in agriculture. The mean–Gini efficient set of decisions is characterized rigorously in terms of its corresponding absolute risk aversion. In an empirical analysis, the mean–Gini efficient set of decisions is derived for four studies from the literature and compared to the second degree stochastic dominance efficient set. An alternative quantitative measure of risk aversion is used to gain insight in a visceral sense to the risk preferences associated with mean–Gini efficient decisions.     

Tests for the Role of Risk Aversion on Input Use

For risk-averting agents, risks alter production decisions while the existence of institutions to insure against adverse states of nature will likely restore decisions toward levels under risk neutrality. In this article, conditions are identified on a stochastic technology to test   H^rn _0,≤  : that risk averters choose smaller input levels than risk neutral agents, and   H^ra _0,≤  : that an increase in risk aversion reduces input use. A robust statistical method to test for dominance is adapted to stochastic production relations. It is found that   H^rn _0,≤  is likely true for nitrogen application on Iowa corn. Weaker evidence is found in favor of   H^ra _0,≤  .     

Risks and returns from soil conservation: evidence from low-income farms in the Philippines

This paper examines risks and returns associated with soil conservation on hillside farms in the Philippines. Stochastic efficiency analysis is combined with a heteroskedastic regression model to assess the impacts of contour hedgerows on low-income corn farms. Regression analysis indicates that, over time, contour hedgerows can improve yields up to 15% compared with conventional practices. The analysis also provides weak support tor a hypothesis that hedgerows are variance reducing. However, results show that the reduction in yield variability afforded by hedgerows is modest, and that yield variability may increase by as much as 5% as hedgerow intensity rises. Tests for stochastic dominance show that, compared with the conventional tillage system, hedgerows do not constitute an unambiguously dominant production strategy. Stochastic elticiency with respect to a function is used to identify a range lor the coefficient ol relative risk aversion within which hedgerows dominate conventional tillage. Results suggest this range would be rather high; hedgerows dominate the conventional cropping strategy only lor decision-makers with relative risk aversion coefficients in the range 3-5.5. Implications for soil conservation adoption in low-income settings are discussed.     

Estimating Technical Efficiency and Production Risk under Contract Farming: A Bayesian Estimation and Stochastic Dominance Methodology

We investigate production risk, technical efficiency and risk attitudes amongst contract and independent farmers. We use a Bayesian parametric approach and stochastic dominance quantile regression methods to compare technical efficiency and risk attitude of smallholders in Nepal. Using farm‐level data, we find that contract farmers appear to show lower inefficiency and lower production risk. Additionally, contract and independent farmers can increase output by reducing the scale of operation. Regardless of the commodity produced and farming arrangement (contract or independent production), we find that labour, land and other inputs are risk‐augmenting, while the role of capital is mixed. We find a second order stochastic dominance (SSD) for lentils, and first order stochastic dominance (FSD) for tomatoes, ginger and HYV paddy seed commodities. Finally, contract farmers are more risk averse than independent farmers, regardless of the commodity produced.     

Robust Comparisons of Malnutrition in Developing Countries

We use Demographic and Health Surveys (DHS) to make international and inter–temporal welfare comparisons. While most poverty analyses rely on expenditures or income, we use anthropometric measures of nutrition as indicators of living standards. The advantages are that we observe individual—not household—well–being, deflators and exchange rates are unnecessary, and measurement techniques are similar across surveys. We test the robustness of the headcount results, and find that applying higher order Foster–Greer–Thorbecke poverty measures adds little information; although stochastic dominance testing of nutrition distributions reveals that changes in malnutrition are sensitive to the choice of the "nutrition poverty line."     

Optimal Insurance Against Climatic Experience

An optimal insurance contract against a climatic risk is derived in the presence of an uninsurable and dependent aggregate production risk. The optimal design depends on the stochastic dependency between both sources of uncertainty and on the producer's attitude towards risk, especially on his prudent behavior. Rational weather insurance purchasing decisions are also derived. The prudent producer responds to actuarially fair weather insurance by increasing his exposure towards risk.     

THE EFFECT OF FERTILISER ON RISK: A HETEROSCEDASTIC PRODUCTION FUNCTION WITH MEASURABLE STOCHASTIC INPUTS

The sources of production risk are many and diverse in nature. Estimating risk as a black box, without explicit recognition of its sources, can lead to inferior estimates of optimal inputs under risk aversion. In this paper, a method is presented for estimating production functions with measurable stochastic inputs and for generating the parameters of the probability distributions of yield for various environments and input levels. Based on this method, it appears that moderate risk aversion can account for a 6.7 per cent to 16.7 per cent reduction in nitrogen use (relative to the risk-neutral solution) for selected rice producing areas of the Philippines. Estimating optimal inputs without environment specific information about the sources of risk leads to large errors. This underscores the value of collecting information about the sources of risk and of exercising caution when such information is not available.     

Off‐farm work,technical efficiency,and rice production risk in Taiwan

This article investigates the differences in yield production, production efficiency, and yield risk for farmers both with and without off‐farm work. Using a nationwide survey of rice farmers in Taiwan, we estimate two stochastic production frontier models that accommodate technical inefficiency and production risk simultaneously for farmers both with and without off‐farm work. The stochastic dominance criterion is then applied to compare the differences in the distributions of the estimated technical efficiency and yield risk between groups. The empirical results indicate that these two groups of farmers use resources in different ways, and off‐farm work is not necessarily associated with lower technical efficiency. For farmers in the lower percentiles of the efficiency distribution, those with off‐farm work are more efficient than their counterparts without off‐farm work. In addition, farmers with off‐farm work face higher production risk and this result is robust for the entire distribution.     

Modelling embedded risk in peasant agriculture: methodological insights from northern Malawi

Using a linear-programming model of farming systems in northern Malawi, the conditions under which peasant farm-household models may need to allow for embedded risk are investigated, Tactical, sequential responses to uncertainty are found to be more important to labour-scarce households with limited access to capital and to credit markets. Compared with semi-sequential programming, discrete stochastic programming (DSP) provided more efficient solutions for problems involving embedded risk. There may be intuitive advantages in presenting results from DSP models in terms of a semi-sequential strategy.     

Optimal supplier testing and tolerance strategies for genetically modified (GM) wheat

A stochastic optimization model was developed to determine optimal testing strategies, costs, and risks for dual marketing of genetically modified (GM) and non‐GM wheat in an export supply chain. The optimal testing strategy is derived that minimizes disutility of additional system costs due to testing and quality loss. Cost components were estimated including those related to testing, quality loss, and a risk premium to induce shippers to undertake dual marketing as opposed to handling only non‐GM crops. Uncertainties were incorporated for adventitious presence and commingling, variety declaration, and test accuracy. Sensitivities were performed for effects of variety risks and declaration, penalty differentials, buyer tolerances, risk aversion, and GM adoption. Results indicate testing and segregation can be performed at a relatively low cost and risk to buyers.     

Should farmers invest in financial assets as a risk management strategy? Some evidence from New Zealand*

This study explores the potential for risk reduction by New Zealand farmers through the diversification of their farm asset portfolios to include financial investments such as ordinary industrial shares, government bonds and bank bills. Low correlations between rates of return on farm and these financial assets suggest that significant reduction of income variability might follow their inclusion in farmers’ portfolios. Stochastic efficiency analysis is used to analyse alternative portfolios of ordinary shares, government bonds and bank bills and New Zealand farmland, using coefficients of absolute risk aversion derived from a negative exponential utility function. The results suggest that those farmers showing high degrees of risk aversion would gain utility by including financial assets in their portfolios. Deregulation of the New Zealand economy in the 1980s appeared to reduce the potential gains from diversification. Bonds rather than ordinary shares are the main contributors to portfolios which maximise utility for individuals classified as ‘somewhat’ risk averse.     

Demand Estimation in the Presence of Stochastic Trend and Seasonality: The Case of Meat Demand in the United Kingdom

If budget shares have stochastic trend or seasonality or both, then demand equations based on the assumption of deterministic trend and deterministic seasonality will be mis-specified. We test this proposition by estimating a Linearized Almost Ideal (LAI) demand system for meat demand in the United Kingdom using Harvey's structural time series methodology. We demonstrate that the model specification allowing for stochastic trend and deterministic seasonality performs best in terms of diagnostic tests and goodness of fit measures. It is also shown that the model with stochastic trend is better at out-of-sample forecasting.     

Valuing Agricultural Insurance

This paper proposes a method for estimating a farmer's stochastic discount factor that is independent of his or her risk preferences, and shows that that stochastic discount factor is appropriate for calculating a farmer's willingness to pay for a crop insurance product. An empirical example illustrates how production and price data might be combined with returns data to permit econometric estimation of the stochastic discount factor, and the implications of those illustrative results are briefly discussed.     

Farms’ technical inefficiencies in the presence of government programs*

We focus on determining the impacts of government programs on farms’ technical inefficiency levels. We use Kumbhakar's stochastic frontier model that accounts for both production risks and risk preferences. Our theoretical framework shows that decoupled government transfers are likely to increase (decrease) DARA (IARA) farmers’ production inefficiencies if variable inputs are risk decreasing. However, the impacts of decoupled payments cannot be anticipated if variable inputs are risk increasing. We use farm‐level data collected in Kansas to illustrate the model.     

The State-Contingent Properties of Stochastic Production Functions

The state-contingent properties of the most frequently used representations of stochastic production in the agricultural-economics literature are examined. Particular attention is paid to the cases of multiplicative uncertainty, additive uncertainty, and the Just–Pope production function. State-contingent technologies and their associated cost functions are reviewed, and that theory is applied to the stochastic production function. A generalization of the Just–Pope technology that has desirable state-contingent characteristics is proposed. Cost functions based upon state-contingent technologies are compared with cost functions based on a parametrized distribution representation of production uncertainty.     

On the co-existence of spot and contract markets: the delivery requirement as contract externality

A contract between an upstream and a downstream party consistsof a contract price and a delivery requirement. Contract formationentails an externality. It changes the probability distributionof the spot market price by removing high reservation pricebuyers and various sellers from the spot market. The first effectdecreases the expected spot market price when the number ofcontracts is small, whereas the decrease in the number of sellersand additional residual contract demand increase the expectedspot market price beyond a certain number of contracts. It impliesan endogenous upper bound on the number of contracts. Contractprices are positively related to the number of contracts. Finally,additional contract formation reduces the variance of the spotmarket price when the number of contracts is sufficiently large.     

Urban Influence on Costs of Production in the Corn Belt

This article determines the relative technical efficiency of rural- and urban-influenced crop/livestock enterprises in the Corn Belt. Farmers in urban-influenced locations are less technically efficient than farmers in rural locations. During 1998–2000, stochastic production frontier procedures indicate that a 10% increase in urban influence leads to a close to 4% decrease in technical efficiency. The most successful urban-influenced farms have controlled costs as effectively as rural farms. They have tended to de-emphasize that nondairy livestock activities—particularly beef and hogs—do not rely extensively on off-farm income, and have relatively large, less residential/lifestyle operations compared to less successful urban-influenced farmers. However, our statistical analysis clearly bears out the refrain in popular literature that urban proximity raises the cost for, and decreases the viability of, traditional farms.     

Evaluating the Potential of Index Insurance Schemes to Reduce Crop Yield Risk in an Arid Region

We evaluate yield risk reduction through weather index, area yield index and farm yield insurance contracts for wheat farms in Kazakhstan by employing data from 1980 to 2002. We use the usual mean variance (MV) approach and also a second‐degree stochastic dominance (SSD) criterion. While MV is not necessarily consistent with the expected utility (EU) theory, SSD results only in a minimum but EU‐consistent benefit from insuring. Differences in the estimation results for both approaches underline the advantage of applying both criteria to analyse the risk‐reducing potential of crop insurance. Bootstrapping results show that none of the analysed insurance schemes provides statistically significant risk reduction for every single farm. In addition, weather‐based index insurance is found to provide less risk reduction than area yield insurance based on the rayon (county) yield. Moreover, rayon yield index insurance can reduce yield risk more effectively for Kazakhstan's wheat producers than farm yield insurance with a low strike yield.     

Decent rural employment and farm production efficiency: empirical evidence from Tanzania and Ethiopia

Promoting decent rural employment, by creating new jobs in rural areas and upgrading the existing ones, could be one of the most efficient pathways to reduce rural poverty. This article systematically investigates the impact of decent rural employment on agricultural production efficiency in Ethiopia and Tanzania. The analysis applies an output‐oriented distance function approach with an estimation procedure that accounts for different technological, demographic, socioeconomic, institutional, and decent rural employment indicators. Data of the 2011 round of Living Standards Measurement Study‐Integrated Surveys on Agriculture for the two countries are used, and a set of indicators is derived to proxy core dimensions of decent rural employment. The findings of our analysis show that decent rural employment contributes to agricultural production efficiency.