Specification and Estimation of Production Functions Involving Damage Control Inputs: A Two-Stage, Semiparametric Approach

Productivity assessment of damage control inputs (such as pesticides) is complicated because their effect depends on the exposure to damage agents (such as pests). We discuss some open specification and estimation issues. The contribution is threefold. First, we elaborate the separability conditions for the damage control function and production function. Second, we employ a two-stage semiparametric technique to combine attractive features of both nonparametric and parametric approaches. Third, we model the interaction between pest exposure and damage control inputs by using slope dummies. An application to productivity analysis of small scale Bt cotton production in China illustrates the approach.     

Spatial Spillovers on Input‐specific Inefficiency of Dutch Arable Farms

Traditional benchmarking implicitly assumes that decision making units operate in isolation from their peers. For arable production systems in particular, this assumption is unlikely to hold in reality. This paper quantifies spatial spillovers on input‐specific inefficiency using data envelopment analysis and a second‐stage bootstrap truncated regression model. The bootstrap algorithm is extended to allow for the estimation of the parameter of the spatial weight matrix, which captures the proximity between producers. The empirical application concerns Dutch arable farms for which latitudes and longitudes are available. The average inefficiency across years was 3.87% for productive inputs and 2.98% for damage abatement inputs under variable returns to scale. For productive inputs technical inefficiency, statistically significant spillover effects from neighbours’ age and their degree of specialisation depended on the type of the spatial weight matrix used (inverse distance or k‐nearest neighbours). Statistically significant spillover effects of subsidy payments were adverse while statistically significant spillover effects from insurance payments were beneficial. For damage abatement inputs technical inefficiency, statistically significant adverse effects were found for neighbours’ age and subsidy payments and beneficial effects from neighbours’ insurance payments and their degree of specialisation.     

How do Agricultural Programmes Alter Crop Production? Evidence from Ecuador

Evaluating agricultural programmes requires considering not only the programmes’ influence on input and output indicators, but also considering the relationship between these indicators as embodied in the production technology. This article examines the impact on production of an intervention in the Ecuadorian Sierra designed to improve returns to potato production through training and through linking smallholders to high‐value markets. Critical to identifying the impact of the programme is the careful construction of a counterfactual and meticulous data collection. To assess the impact of the programme on production, a weighted estimation, where weights are constructed through propensity score matching, is employed to estimate a production function within a damage abatement framework. The function incorporates a series of interaction terms to assess the impact of the programme on the production technology. The findings provide evidence that the programme enhances yields both through a general shift in technology as well as increased input use. The results suggest that the use of effective farming techniques that are learned through the programme induce this technological shift.     

Rural credit and agricultural supply in Brazil

The objective of this study is to determine the financing impact of total expenditure on the use of agriculture inputs (fertilizers, labor, and pesticides), and the output of cotton, rice, beans, corn, soybean, and wheat in Brazil. We study the period 1976–2005. The analysis is based on duality applied to the production theory. The output supplies and conditioned input demands are estimated from a translog multi‐output, multi‐input restricted profit function, where the total production credit is used as proxy of the total expenditure. Farmer expectations with respect to crop prices are incorporated to the estimation based on the quasi‐rational expectation model. The output and input responses to the total expenditure are positive and statistically significant except for cotton, wheat, fertilizer, and pesticides. The short‐run output supply response to own prices is inelastic, except for wheat, which presents elastic response to its price. Acreage has a positive impact on the output supply and it is influenced by land productivity. The main conclusion is that farmers face budget restrictions to purchase inputs, and a government credit program might increase the agricultural supply.     

Estimation of input-oriented technical efficiency using a nonhomogeneous stochastic production frontier model

Technical inefficiency can be modeled as either input‐oriented (IO) or output‐oriented (OO). However, in the estimation of parametric stochastic production frontier models which use maximum likelihood method only the OO measure is used. In this article we consider a simple nonhomogeneous production function and estimate it with both IO and OO specifications. A sample of 80 Spanish dairy data (1993–1998) is used to estimate both models. We consider one output (liters of milk) and four variable inputs (viz., number of cows, kilograms of concentrates, hectares of land, and labor [measured in man‐equivalent units]). We find that returns to scale (RTS) and technical efficiency results derived from these models are different because either estimated technologies are different, or they are evaluated at different points. Using a Monte Carlo analysis we show that if RTS is close to unity differences in the estimates of RTS and technical efficiency are smaller. This holds true for estimates of both RTS and technical efficiency.     

Estimating the Marginal Productivity of Agricultural Pesticides: the Case of Tree-Fruit Farms in the Okanagan Valley

This paper uses input and output data from a cross-sectional sample of tree-fruit farms in British Columbia to derive an estimate of the marginal productivity of agricultural pesticides. The statistical techniques used in the analysis of the data include Ordinary Least Squares and Factor Analysis Regression. The analysis indicates that the value of the marginal dollar's worth of pesticides is considerably higher than its private profit-maximizing level of unity. It is suggested that the reason for the existence of such a divergence is the presence of a constraint on farmers' total expenditure on variable inputs. This high estimate of the value of the marginal dollar's worth of pesticides inputs indicates that public policies aimed at reducing the level of pesticides inputs in the interests of environmental protection will have a relatively high opportunity cost in terms of forgone agricultural output. Cette étude réalisée à partir d'un échantillon d'aboriculteurs de la Colombie Britannique se base sur des données de production et de consommation pour estimer la productivité marginale des pesticides agricoles. L'analyse factorielle de la regression et la méthode des moindres carrés font partie des techniques statistiques employées. L'analyse indique que la valeur marginale d'une unité monétaire de pesticide est considerablement plus elevée que l'unité qui est la valeur obtenue quand l'on maximise la fonction de profit privé. L'explication suggérée pour justifier l'existence d'une telle divergence est la presence d'une limite sur les depenses variables de consommation des agriculteurs. Cette valeur élevée de l'estimateur de la valeur marginale d'une unité monetaire de pesticide indique que les politiques destinées à reduire l'usage des pesticides dans le but de proteger l'environnement auront un coût relativement élevé en terme de production agricole perdue.     

Estimating Bird Damage to Rice in Africa: Evidence from the Senegal River Valley

Granivorous birds, mainly the Red‐billed Quelea, have subsisted on cereal crops in Africa for centuries and have caused substantial damage. There is, however, limited recent evidence on their impact. We propose an indirect method to estimate bird‐inflicted crop losses by fitting a production function with a damage abatement component and pest intensity slope dummies on a panel database of rice farmers in the Senegal River Valley. This allows us to estimate both bird damage and marginal productivity of bird control at different levels of bird pressure. Annual bird damage is found to average around 13.2% of the potential rice production during the wet seasons of 2003–2007, which translates into an average annual economic loss of 4.7 billion FCFA (€7.1 million). Our results are consistent with farmers’ perceived bird‐inflicted crop losses, averaging 15.2%. More alarmingly, we observe declining marginal productivities of bird control under increasing bird pressure. Farmers indicate that at high bird pressure, the efficacy of traditional bird scaring methods is inadequate, which suggests that predictive (monitoring), preventive (population control) and protective (insurance) measures against massive invasions are more urgent than improving the average efficacy of curative measures (pest control). These findings are especially relevant to farmers and policy‐makers who are currently struggling to implement an ambitious food self‐sufficiency programme in Senegal.     

AN INTERMEDIATE INPUT MODEL OF DAIRY PRODUCTION USING COMPLEX SURVEY DATA

Agricultural production is often characterised by multiple inputs and multiple outputs to multiple production processes. Where an output from one process is used as an input to another, this output is called an intermediate product. This is a common situation when a farm produces both crops and livestock. The analysis of production efficiency is important for the evaluation of agricultural policy, but until recently, no methods have explicitly included intermediate products. This study applies a non-parametric technique of efficiency measurement which includes intermediate products. The data set is a sample of dairy farms drawn using a complex survey design. The use of non-parametric efficiency measurement and the subsequent application of bootstrapping and kernel density estimation to the results allow inferences to be drawn concerning the whole population from which the sample was drawn. We find that the decomposition of production into subproduction processes reduces the dimensions of problem specification, with the effect that a larger number of variables may be usefully included in the model.     

Determinants, differentiation, and development of short-term marginal costs in dairy production: an empirical analysis for selected regions of the EU

This work provides evidence on the determinants, cost differentiation, and development of short‐term marginal costs of dairy farms in important production regions of the European Union. The empirical study is based on the estimation of multi‐input multi‐output Symmetric Generalized McFadden cost functions using an unbalanced panel data set of the European Farm Accountancy Data Network. The results show considerable regional differences in the impact of the outputs, input prices, and fixed factors on marginal costs. Strong evidence can be found that marginal costs decrease over time and is further underlined by the development of derived regional aggregated short‐term supply curves. Marginal cost elasticities and correlation coefficients validate the hypotheses that a high degree of farm specialization, large milk output, and low milk prices are associated with lower marginal costs. Furthermore, the marginal cost spread in the data sample is analyzed. We show that milk output, milk yield, herd size, labor input, and fodder production can be attributed to significant marginal cost differentiation of farms, whereas for crop and animal output, grassland, stock of other animals, and depreciation only minor differentiation can be found.     

Intertemporal and Interfirm Heterogeneity: Implications for Pesticide Productivity

This paper presents evidence concerning the existence and implications of heterogeneity effects for estimates of the marginal productivity of pesticides, The paper extends the work of Calpentier and Weaver (1996a), which considered fured effects that homothetically shift the conditional mean of output. This paper considers the implications of time-varying marginal productivity, allowing for nonhomothetic shifts in production. The extent and structure of heterogeneity found in French crop agriculture is first presented to illustrate that persistent heterogeneity exists across these firms. Next, we review the implications of such heterogeneity for econometric estimates and show that estimates of marginal productivity of pesticides will be biased when based on models that do not explicitly account for these types of heterogeneity across firms and time. An alternative estimation approach is proposed, based on Chamberlain (1982 and 1984) and implemented to produce asymptotically unbiased estimates and to statistically validate the model specification. Existence of fixed firm effects is statistically confirmed and temporal variation of marginal productivity is found to be substantial.     

Impacts of Norwegian Milk Quotas on Output Growth: A Modified Distance Function Approach

Quota regulations that prevent output expansion of farms and reallocation of output between farms can cause lower growth in output and productivity. The aim of this study was to explain the output growth rate of Norwegian dairy farms since 1976, and to decompose it into output, input, socioeconomic and technical change components. Instead of using the standard distance function approach for multi‐output technologies, we use a growth rate formulation, which automatically removes the farm‐specific effects. This formulation also helps to impose non‐negativity constraints on marginal products of inputs (input elasticities), which are often violated for many observations, especially when flexible functional forms are used. The farm‐level panel data cover three periods: before the quota scheme was introduced (1976–1982); the period with the most output‐restricting quota scheme (1983–1996); and the period with a more flexible quota scheme (from 1997 onwards). Results show that the milk quota regulations had a significant constraining effect on output growth, in particular on milk output in the period 1983–1996. Furthermore, the output mix has shifted towards meat production for the average farm. What emerges from this study is that output growth and technical change are negatively influenced by policy aims where productive performance has not been the primary objective, and that there is scope for increased farm growth if the quota regime is liberalised.     

Measuring pesticide overuse and its determinants: Evidence from Vietnamese rice and fruit farms

Pesticides have long been important for the development of agricultural production. However, improper use of pesticides may result in inefficiency with respect to farm profitability, in addition to external effects of pesticide use on environmental and human health. This paper employs a production function that explicitly accounts for the role of damage abatement inputs (i.e. pesticides) in the production process, to examine the optimal use of pesticides. It then investigates determinants of pesticide overuse versus underuse and the intensity of overuse. The empirical application uses data on Vietnamese rice and fruit farms drawn from the 2016 Vietnamese Household Living Standards Survey. Results show about 95% of farmers overused pesticides for both rice and fruit farming systems. The Mekong Delta, also known as the ‘Rice Bowl’ of Vietnam, has higher levels of overuse on rice farms than two other regions. Overuse intensity is lower for female and poorer rice farmers while intensity is higher for those with more income and more family members. For fruit farms, younger farmers or those with more family members were more likely to overuse versus underuse pesticides.     

Marginal abatement costs of greenhouse gas emissions: broadacre farming in the Great Southern Region of Western Australia

Broadacre agriculture is a major emitter of greenhouse gases (GHG). To improve efficiency of climate change policies, we need to know the marginal abatement costs of agricultural GHG. This article combines calculations of on‐farm GHG emissions with an input‐based distance function approach to estimate the marginal abatement costs for a broadacre farming system in the Great Southern Region of Western Australia. The results show that, in the study region, the average marginal abatement cost for the 1998–2005 periods was $29.3 per tonne CO₂‐e. Farms with higher crop output shares were found to have higher marginal abatement costs. Overall, our results indicate that broadacre agriculture is among the lowest cost sources of GHG mitigation.     

Measurement errors in agricultural data and their implications on marginal returns to modern agricultural inputs

The low uptake of modern agricultural technologies in sub-Saharan African countries has encouraged researchers to revisit the returns to (or profitability of) these agricultural inputs. A related strand of literature is exploring the allocative efficiency of these factors of production in African agriculture. However, all these studies rely on self-reported agricultural data, which are prone to nonclassical measurement errors, the errors in these data are correlated with the true values of variables of interest. In this article we investigate the implication of measurement errors in self-reported agricultural input and production data on marginal returns to these modern agricultural inputs. We consider a generic two-sided measurement error problem where both production and inputs can be measured with error, and these errors can be correlated. We employ both self-reported and objective measures of production and plot size to compute output elasticities under these alternative measurement scenarios. We find that using self-reported production and plot size overestimates output elasticities and hence marginal returns to modern agricultural inputs (including chemical fertilizer and improved seed). These results are noteworthy in terms of informing conventional technology diffusion strategies as well as in view of revisiting existing presumptions about the profitability of modern agricultural inputs.     

Environmental Emissions and Production Economics: Implications of the Materials Balance

In modeling emissions, the literature has usually specified an explicit emission function, or treated emissions as a production input. We examine the validity of these approaches, taking into account the materials balance principle. We show that a technology can equivalently be described by (i) a production function with material and nonmaterial inputs and bounded marginal product of the material input, (ii) a well-behaved production function with emissions as an input, and (iii) a well-behaved emission function, if the materials balance is accounted for as an additional condition. We offer a forma derivation of common, but not rigorously established modeling approaches.     

PRODUCTION INTERRELATIONSHIPS IN SRI LANKAN PEASANT AGRICULTURE

Technical change and the extent to which commodity supplies and input demands are interrelated in Sri Lankan peasant agriculture are explored in this paper. Using a multiple-product dual model, a seemingly unrelated system of product supply and input demand equations is estimated for four crops and four variable inputs. Restrictions based on competitive behaviour and a twice continuously-differentiable production function are maintained in the non-linear least squares estimation. A number of important interrelationships in individual product supplies and input demands are identified, further documenting the need to account for intercommodity production relationships in econometric and simulation studies and in policy formulation. Non-joint production and Hicksneutral technical change are both rejected.     

On the nature and magnitude of cost economies in hog production

In this article, we assess the impact of farm size on production cost and evaluate the marginal costs and margins by considering that input prices may change with the scale of production. By using French hog farm data, we estimate a system of equations including a feed price function, input demand functions, and an output supply function based on a technology approximated by a combined generalized Leontief‐Quadratic form. Our results suggest that the marginal costs are over‐estimated when the adjustment of the feed unit prices to a change in farm size is not controlled for. More specifically, the cost economies for large farms (enjoying the highest profits) arise primarily from lower feed prices, with technological scale economies having little impact. In contrast, farms with no hired labor exhibit technological scale economies and reach higher price‐cost margins compared to larger farms.     

Input and output technical efficiencies of wheat production in Kerman, Iran

The Iranian government encourages farmers to produce wheat (a common agricultural enterprise) by increasing farm productivity and efficiency. In this paper, using a Cobb–Douglas frontier production function, a simple relationship between a farm‐level output‐based technical efficiency measure (the Timmer index) and an input‐based measure (the Kopp index) is first developed. Then, using 1995 data from 164 farms in Kerman province, Iran, the average Timmer and Kopp indexes were estimated at 0.93 and 0.91, respectively, and were found to be similarly affected by farm size (positively up to about 9 ha) and by input ratios, though with rather small explanatory power. Thus, there seems some but limited scope to increase the profitability of Iranian wheat production either by increasing the product, given input levels, or by decreasing inputs for the current level of wheat production. However, since wheat producers may be able to adapt their production process more easily and quickly by implementing new techniques, i.e. by more efficient combination of inputs, than by adopting new technology, correction of input over‐use can be regarded as a policy with speedy if limited effect in this case.     

Input use under cost-of-production crop insurance: Theory and evidence

There have been a number of previous studies that examined the effects of yield- or revenue-based crop insurance products on input use of farmers. However, no study has specifically investigated the input use impacts of a cost-of-production (COP) crop insurance policy, even though this type of crop insurance is the predominant one used in several other countries outside of the United States (such as the Philippines and China). This article aims to theoretically and empirically examine the effect of a COP crop insurance product on farmers’ chemical input use. Our theoretical model suggests that the effect of COP insurance on input use can either be positive or negative, with the resulting impact depending on the strengths of (a) the traditional moral hazard effect of insurance (i.e., an input use decreasing effect); versus (b) the marginal incentives to apply more inputs due to input levels being the main determinant for expected indemnity amounts in this type of insurance (i.e., an input use increasing effect). A survey data set from corn farmers in the Philippines is then used to empirically illustrate how a particular COP insurance product influences input use in a real-life context. In this case, we find that COP insurance increases the use of chemical inputs (e.g., fertilizers and total chemical expenditure), implying that the positive marginal incentive to apply more inputs dominates the negative moral hazard effect.     

TECHNOLOGICAL CHANGE AND SURPLUS LABOUR IN EGYPTIAN AGRICULTURE, 1952–1972

This paper has two main purposes. First to determine if technical innovation has occurred in Egyptian agriculture and whether or not it has been labour-using in nature. Secondly, to determine whether or not the marginal product of labour in Egyptian agriculture has been negative (surplus labour). The time period covered by the study is 1952 to 1972. The above objectives were met by estimating a weak disposability of inputs (WDI) production function. The characteristics of the weak disposability of inputs function were discussed in great detail, since much of the profession is unfamiliar with this functional form. The function was then estimated and the results seemed to indicate that this functional form was appropriate for Egyptian agriculture. The results of the estimation process indicated that labour-using technical innovation had indeed occurred, although at a very slow rate. In addition, for most of the time period covered the marginal product of labour was indeed negative.