Danilo Gambelli

Assessing the risk of non-compliance in UK organic agriculture: An econometric analysis

Purpose – The purpose of this paper is to provide an analysis of the risk factors influencing non-compliance in UK organic farming. Design/methodology/approach – The paper uses a formal econometric model of risk analysis to provide empirical evidence on the determinants of non-compliance in organic farming. A panel of data from the archives of the largest control body in the UK for 2007-2009 is used, and specific analyses are performed for two types of non-compliances. A zero inflated count data model is used for the estimation, taking into account the fact that the occurrences of non-compliance are very sparse. Findings – Results show the existence of strong co-dependence of non-compliant behaviours (i.e. the occurrence of major and critical non-compliance increases the probability of occurrence of the minor one; similarly the probability of occurrence of major non-compliance increases when minor non-compliance occur). Besides, livestock production and farm size are relevant risk factors. Research limita...

Exploiting Qualitative Information for Decision Support in Scenario Analysis

The development of scenario analysis (SA) to assist decision makers and stakeholders has been growing over the last few years through mainly exploiting qualitative information provided by experts. In this study, we present SA based on the use of qualitative data for strategy planning. We discuss the potential of SA as a decision-support tool, and provide a structured approach for the interpretation of SA data, and an empirical validation of expert evaluations that can help to measure the consistency of the analysis. An application to a specific case study is provided, with reference to the European organic farming business.

Un sistema di certificazione risk-based per i controlli in agricoltura biologica: un’applicazione tramite Bayesian networks

The existing method of certification in the organic agriculture system, which requires periodic inspection of all operators, is inefficient due to the high cost of these controls. A risk-based decision support system, which could assist the inspection body during the planning of the annual inspection visits, is advocated as being more cost-effective and efficient. The risk-based decision support system is constructed as a Bayesian network; the models incorporate the factors that influence risk of irregularity and analyse their effects by determining probability of noncompliance. Empirical findings, using a sample of Italian data regarding inspection of organic farms, support the idea that the current risk categories used by control bodies in Italy are reasonable, but could be recursively updated by using a Bayesian network model and incremental inspection evidence.

Assessing Risk Factors in the Organic Control System: Evidence from Inspection Data in Italy

Certification is an essential feature in organic farming, and it is based on inspections to verify compliance with respect to European Council Regulation-EC Reg. No 834/2007. A risk-based approach to noncompliance that alerts the control bodies to activate planning inspections would contribute to a more efficient and cost-effective certification system. An analysis of factors that can affect the probability of noncompliance in organic farming has thus been developed. This article examines the application of zero-inflated count data models to farm-level panel data from inspection results and sanctions obtained from the Ethical and Environmental Certification Institute, one of the main control bodies in Italy. We tested many a priori hypotheses related to the risk of noncompliance. We find evidence of an important role for past noncompliant behavior in predicting future noncompliance, while farm size and the occurrence of livestock also have roles in an increased probability of noncompliance. We conclude the article proposing that an efficient risk-based inspection system should be designed, weighting up the known probability of occurrence of a given noncompliance according to the severity of its impact.

Estimating missing data for organic farming by multiple imputation: the case of organic fruit yields in Italy

The availability of timely and good-quality data on the organic farming sector is a crucial factor for the development of the organic food market. While data on hectares and farms are now widely available in Europe, data on organic yields are still relatively sparsely reported by official statistical sources for most European countries, including Italy. Information on organic yields is crucial to determine the volumes of organic production and supply. Issues such as the potential of organic farming for feeding the world, the understanding of the optimal conditions for conversion and the appropriate policy measures for supporting the organic sector are all dependent on the knowledge of organic productivity. In this study, we show how a statistical method known as multiple imputation can contribute to the improvement of the availability of organic data, through systematic exploitation of data from different sources. We apply the method to estimate missing data on organic fruit crop yields for the central regions of Italy, based on data from official national statistics and expert assessments. The results illustrate the advantages and limitations of such methods for estimating missing data on organic crops.

Improving controls in organic farming by timely inspections: a statistical analysis

Abstract This paper investigated if the inspection procedures in organic certification may be improved exploiting information that is generally available from the control bodies involved in the certification process. The analysis was based on data from the archives of one of the largest Italian organic control bodies, containing information on operators’ characteristics, and including: risk scores for farmers, inspectors’ characteristics, type of inspection and the outcome of the inspection in terms of the type of non-compliance detected. The analysis considered both irregularities, i.e. mainly formal or bureaucratic non-compliance, and infringements, i.e. more substantial non-compliance. A bivariate probit model with random parameters was used to estimate the likelihood of, presumably correlated, irregularities and infringements, conditional to a set of covariates concerning risk assessment of the operators, inspector’s characteristics, and modalities of the inspections, including the period of the year scheduled for the inspections. The results showed that irregularities and infringements were actually correlated and that there is scope for improving the effectiveness of inspections, particularly using an appropriate timing for inspections and taking samples more frequently during inspections.