Food Fraud: A Supply Network Integrated Systems Analysis

Lead Research Organisation: University of Manchester
Department Name: Law

Abstract

The aim of this research is to develop a predictive computational approach to modeling the food supply network so that the points where food adulteration can occur are identified. By identifying these points of vulnerability to adulteration within the supply chain will allow regulators and retailers to take appropriate action to avoid food adulteration.

This project brings together an interdisciplinary team of researchers from analytical sciences, predictive modeling, law, criminology and business studies. This proposal is highly innovative in that it exploits the dynamics between the diverse disciplines that makes the development of a Supply network integrated systems analysis an exciting and groundbreaking development in tackling issues of food adulteration.

Food fraud is not just a contemporary phenomenon and the deliberate adulteration of food products has a long history and with recent events throughout Europe appears to be an ongoing problem that needs to be addressed. Whilst the link between criminal organisations and food adulteration has not been a specific focus of food adulteration research, the recent incidences of adulteration of the bovine meat supply chain with horsemeat have raised a number of food protection issues related to the potential consequences of unregulated food supply chains. The outcome of this project will be a computational model of risk for predicting fraudulent opportunities for adulteration and points of vulnerability within food supply chains. This will allow for the prediction of potential food adulteration incidences through criminal and other action such as increased monitoring of the identified critical points of vulnerability. Such an analysis allows for an understanding of how the regulatory framework and the prevailing market conditions impact on and create vulnerabilities in food supply chains. These vulnerabilities will be examined in relation to criminal opportunities and market structures. Interactions between these different elements will be explored to identify how they provide criminal opportunities and how legitimate actors within the supply chain are placed to take advantage of such opportunities. The integrative systems analysis model will be dynamic and predictive of possible locations and opportunities for adulteration within the length of the supply chain. Such a dynamic model has application within a number of other arenas and would be of benefit to food regulators, retailers, and ultimately consumers.

The prediction of risk points and vulnerabilities within the food supply chain will provide significant benefit to a wide range of stakeholders and interested parties. Food processors and retailers will be able to maintain their reputations by taking action to ensure that they have protected their food products from adulteration. This will contribute to consumer confidence and trust in UK food supply chains. Finally, regulators will be able to develop well targeted detection and prevention strategies to address unethical behaviour and reduce criminal opportunities.

Planned Impact

Introduction
The food supply chain presents a number of complex problems for regulators in ensuring the robustness of the systems and the integrity of the food offered to the consumer. The adulteration of food is a highly complex issue in terms of detection and a timeline developed by this research team demonstrates the episodic nature of food adulteration incidents. This research team has achieved impact through the work already completed with completion of a qualitative study into the horsemeat scandal and the submission of an article for peer review to British Journal of Criminology. This research team brings together a truly cross disciplinary approach with biological chemists with a focus on food adulteration, criminologists with a range of expertise from across the discipline, a lawyer with a focus on regulation and academics with expertise in business and supply chain integrity. The work of the research team has already gained national prominence with television and radio appearances on the anniversary of the horsemeat scandal. This proposal builds on work completed by all these scholars and it is anticipated that this forward momentum will continue.
Specific users and how they will benefit
The research proposed here represents a step change in detection of critical adulteration points in the food production pipeline and has the potential to impact positively on a wide range of scientists, specifically:
(1) Academic researchers who want to utilise computational systems analysis of complex networks.
(2) Regulatory scientists who conduct and police food fraud analysis, where we shall provide a greater understanding of how interactions across a number of different variable ranges can be modelled to predict certain points of vulnerability
(3) Retailers where the adoption of such an analytic model would also provide effective forms of monitoring different food supply chains.
(4) Government and policy makers, who set policy and are ultimately answerable to the public.
(5) International organisations, including academia and the public and private sectors, which will further improve the recognition of UK funded food fraud detection research worldwide.

Methods for communication and engagement
(1) On completion of the project we will organise a project specific workshop, at the University of Manchester, to inform the academic, governmental, regulatory and industrial community of the major outcomes of this research. The workshop will highlight the new software resource, and provide training in this tool.
(2) A project specific website will be established. We are already have a presence in the 'twittersphere' (@iffnmanchester; The Official twitter account for the UMRI Food Fraud and Adulteration Project, University of Manchester)
(3) The research and software resource will be presented to the scientific community via conferences and publications in international journals. We have a policy of supporting our postdoctoral researchers to present their data at conferences. We support the publication of scientific data and findings in open access journals (all applicants involved). We will also write introductory articles in the 'lay press' (e.g., Meat Press).
(4) Engagement with public, as has already happened recently: Jon Spencer was interviewed on BBC1 News and conducted several local and national radio interviews.

Commercial exploitation of proposed research
Although the software resource will be made available to academic users free of charge, we will seek opportunities to exploit the commercial value of the software package. Specifically we will consult with University of Manchester Intellectual Property (UMIP)
 
Description Food Fraud: A supply Network Integrated Systems Analysis 
Project Findings 
Challenges of Prevention Rather than Strategies of Prediction 
There are a range of findings across the different project streams and it is useful to attempt to bring them together into a coherent presentation. There is a considerable drive to predict food fraud. The project was designed to consider how an integrated systems analysis could be used to predict food fraud episodes. The idea was that by mapping flows through a product supply chain it would be possible to identify nodes in the supply chain that behaved fraudulently - changes in behaviour would be evident in the change of flows through the chain. Once a body of knowledge had been established it would be possible, computationally, to model the supply chain on 'live' information, map the flows through the chain and therefore identify and predict fraudulent practices. This research stream links into the need to be able to predict food fraud events to enable the reduction in fraudulent practices and to ensure and increase food safety, a major concern of the FSA. 
The above is based on a number of assumptions. First, that there exists comprehensive data on supply chain flows in relation to essential elements, product flow and financial transactions. During the project, supply chain product flow data has not been available primarily because it is not held in any one place. Consequently, to put supply chain product flow data together would require back tracing each transaction in each supply chain. As food supply chains are strongly shaped by transactions there is no certainty that the flow will follow the same transactional pattern in each transaction event. This is compounded by large retailers having short term contracts with processors and suppliers on the basis of increasing and maintaining their competitiveness. The second assumption is that actors within the supply chain are willing to share their data. For many processors this data, the transactions between them and retailers, is highly sensitive and to share either the flow or financial data would be exceedingly risky. Many processors employ strategies to prevent the retailer having a full view of either data set on the basis that they need to protect their own trading position. This highlights that many food transactions are based on a sense of mistrust between the different actors. 
The modelling strategies that have attempted to replicate, from our knowledge, supply chain dynamics indicate that with the relevant data a predictive model could be developed to some extent. The other main problem with predictive modelling is that the social environment in which the model is placed is not constant. For example, fraudulent and normal, routine operational business practices are qualitatively similar, meaning a transaction that appears to be honest may well conceal a fraudulent transaction. So, determining what the conditions of stress are that would motivate a fraudulent transaction are also complex and so challenging to model in this context. 
The regulatory stream has also developed a complexity approach to understanding the problems of preventing food fraud under the current legal and regulatory provisions, as well as the practices developing therefrom. The legal and regulatory system creates 'oversight gaps' due to the fragmentation of agencies, authorities and enforcement, as well as of offences and regulatory approaches. Our qualitative data reveals that authorities follow diverse, and at times contradicting, approaches to enforcement due to the different legal interpretations that relevant agencies such as the FSA and DEFRA can provide on the same issue, creating uncertainties both for local authorities that have to prioritise one interpretation over another, as well as for the industry who operate across the jurisdictions of a variety of regulatory agencies. The guidance on food labelling provided by the FSA and DEFRA regarding the European Food Information to Consumers Regulations No. 1169/2011 (FIC) and the Food Information Regulations (FIR) is a case in point. One of our interviewees (local authority) highlighted the problems arising from the definition (or lack of it) of 'pre-packed for direct sale'. Following the FIC, the FSA's guidance establishes that for pre-packed food for direct sale to the consumer, the same requirement of non-prepacked food should be followed insofar as the food has been packed on the same premises from which they are sold and either of the following conditions apply: a) the consumer is able to ask the person who made or packed the product about the ingredients; or, b) the food comprises meat pies, sandwiches and the like that are sold in the premises in which they are made. However, DEFRA's guidance on labelling does not provide a clear definition for 'pre-packed for direct sale' and how these should be treated if they so constitute 'pre-packed for direct sale'. A 2012 document merely establishes that 'pre-packed for direct sale' such as deli counters, bakeries and deli bars should comply with allergen information. The lack of clarity adds to the regulatory complexity which makes both compliance and enforcement more entangled, particularly bearing in mind the fragmentation of regulatory authority. For example, in England the FSA is responsible for labelling policy related to safety but has transferred non-safety labelling issues to DEFRA and nutritional labelling to the Department of Health, whilst in Wales the FSA retains all labelling policy except for health nutritional labelling. However, in Northern Ireland the FSA retains all labelling policy. So, the same agency follows different approaches in three jurisdictions. 
Under a current legal and regulatory system that was framed as a response to food crises, regulators are challenged in keeping pace with a constantly changing and adapting regulatory target. This complexity is exacerbated by an increasing amount of private standards and certification schemes - the extent to which local authorities can rely on these mechanisms, mostly used by powerful industry actors, remains to be established. As such, regulators are under-resourced not only in terms of financial resources, but also in terms of coordination mechanisms between authorities and agencies, sufficient expertise of the industry, the necessary enforcement powers to pursue investigations, and of sanctions. Although sanctions might exist, the high level of evidence required for a successful prosecution on fraud is reflected in the low number of food fraud prosecutions as opposed to administrative breaches. 
When asked about the complexities of regulation in dealing with food fraud, one interviewee (local authority) responded '[f]raud, that we haven't use, that would be a high level, we have to prove that it was for gain and they have that knowledge of it, so I have not use that yet so I don't know the complexities'. A more striking problem is evident. Certainly, dealing with food fraud investigations and prosecutions seems difficult under a legal and regulatory system that is framed on safety and public health concerns. 
Therefore, there is a mismatch between the complex social and criminal organisation of food fraud and the complex legal and regulatory framework that undermine the capacity for an effective regulation and the protection of consumers. Any regulatory intervention is meaningless without firstly fully understanding the nature of the problem or phenomenon to be regulated - that is, the proximal (i.e. necessary aspects of the crime commission process) and the distal (i.e. the organisational and industry cultures that are conducive to tolerate, encourage and/or turn a blind eye to deviant practices) circumstances underpinning food fraud. Such an understanding is fundamental to develop reduction strategies and to move from a responsive into a proactive regulatory approach which is more targeted and efficient. 
The criminological approach has developed through the use of the application of 'script analysis' and utilising the 'crime as enterprise' approach that provides insights into the illicit behaviours within the food industry. The significant findings are that the majority of food crimes are organised within the food industry. There is little evidence, especially in the UK, of organised crime groups external to the food industry being involved in the food crime. This finding is supported by the NFCU. This finding requires a change of approach into understanding food fraud. The qualitative data repeatedly makes clear that the food industry is one that is prone to fraud because of the unethical behaviours that are common practice within the industry. The lack of ethical business behaviour suggests that fraudulent practices become embedded and normalised within the day to day routines of business within the food industry. A review of food fraud incidents reveals that they occur across all food sectors and are relatively regular occurrences, and that they implicate 'big business', middle-level entrepreneurs, as well as lower level 'rogue' businesses. The development of a script analysis highlights the way in which 'offenders' co-operate across the industry to undertake fraudulent practices. The development of the script analysis also highlights how it is necessary for there to be a network of actors linked together to undertake the actions necessary to commit the fraud. 
This moves to the final question; how do we prevent food fraud? Importantly, we do not try to find it, doing so is like looking for a needle in a haystack. As an interviewee (local authority) stated: '[Testing] is a postcode lottery in terms of sampling being taken'. Testing is an 'after the fact' detection mechanism that would not prevent food fraud insofar as it is implemented in every business through the whole supply chain - an unrealistic and costly mechanism. Prevention will come from two actions; first reform the market (long term). This is not as straightforward and easy as it sounds. There are many vested interests in the market but until food production and processing moves away from extremely tight profit margins, 
with some taking losses, the motivation for fraud to ensure survival will never be far away. Reform has to ensure a fair price for those other than consumers and retailers. Reform within the Brexit era will prove to be even more demanding and there is probably no political will to do so. However, Brexit may offer the opportunity for food market and industry reform. The second approach is that of crime prevention (short and medium term). Analyse each market sector in depth and identify the potential locations of fraud. It is clear in the olive oil industry that adulteration can only happen at certain points in the process. Therefore a criminological understanding of how food sector markets operate will provide increased intelligence and assist in the targeting of resources. Finally, do away with conventional assumptions. Do not assume that the financially stressed firm will behave criminally, how come the profitable ones in a marginal sector of profitability are so profitable? 
Exploitation Route There is a possibility to continue from the theoretical modelling work.
The criminological work is well developed and can be used in exploring other food sectors, for example the food service sector,
Sectors Agriculture, Food and Drink,Other

 
Description Our findings have informed the approach of the Food Safety Authority Ireland and the National Food Crime Unit (FSA). The findings from the criminological side of the project have informed investigative decision making.
First Year Of Impact 2016
Sector Agriculture, Food and Drink
Impact Types Policy & public services

 
Description Distributiona nd Consumption of Counterfeit Alcohol: Getting to Grips with Fake Booze
Amount £60,000 (GBP)
Organisation Alcohol Research UK 
Sector Charity/Non Profit
Country United Kingdom of Great Britain & Northern Ireland (UK)
Start 09/2016 
End 08/2018
 
Description University of Manchester Research Institute (UMRI)
Amount £24,500 (GBP)
Organisation University of Manchester 
Sector Academic/University
Country United Kingdom of Great Britain & Northern Ireland (UK)
Start 02/2016 
End 07/2016
 
Description Development of Food Fraud Approaches in ensuring food integrity 
Organisation Fera Science Limited
Country United Kingdom of Great Britain & Northern Ireland (UK) 
Sector Multiple 
PI Contribution Collaborations on developing a social science approach to food integrity
Collaborator Contribution Sharing of ideas and information concerning food fraud
Impact Attendance and workshop at Food Integrity 2016
Start Year 2015
 
Description Fair Trade 
Organisation Traidcraft
Country United Kingdom of Great Britain & Northern Ireland (UK) 
Sector Charity/Non Profit 
PI Contribution Discussion concerning regulation of the grocery sector
Collaborator Contribution Access to issues and concerns in grocery retail
Impact Evidence to the Government consultation on the grocery adjudicator
Start Year 2016
 
Description Food Crime 
Organisation Food Standards Agency (FSA)
Department National Food Crime Unit
Country United Kingdom of Great Britain & Northern Ireland (UK) 
Sector Public 
PI Contribution Sharing of ideas and approaches to food crime and its detection and prevention
Collaborator Contribution Contribution to research seminars
Impact Review of articles and discussion papers.
Start Year 2016
 
Description Food Safety 
Organisation Food Safety Authority of Ireland
Country Ireland, Republic of 
Sector Public 
PI Contribution Analysis of crime files in relation to particular food incidents
Collaborator Contribution Sharing of data Provision of meeting rooms and hospitality
Impact Development of approaches to understanding distribution networks of counterfeit product.
Start Year 2015
 
Description Operation OPSON VI 
Organisation Europol
Country Netherlands, Kingdom of the 
Sector Public 
PI Contribution Sharing of information and approaches to understanding a scripts analysis in relation to food fraud
Collaborator Contribution Inclusion at relevant OPSON report seminar
Impact Inclusion in OPSON Seminar
Start Year 2016
 
Description Problems of Modelling 
Organisation Organic Services GmbH
Country Germany, Federal Republic of 
Sector Private 
PI Contribution Discussion of the issues of food fraud and how to model possible food fraud cases.
Collaborator Contribution Visit to Manchester, sharing of information and ideas concerning modelling of food fraud.
Impact Contribution of a criminological analysis to the development of computer based modelling in relation to food fraud. This was part of the wider 'Food Integrity' project
Start Year 2016
 
Description Supply chain analytics to improve resillience 
Organisation NOFIMA, Ås Norway
Country Norway, Kingdom of 
Sector Private 
PI Contribution Contribution to a large research proposal - Primefood
Collaborator Contribution Development of supply chain analysis
Impact Full research proposal
Start Year 2016
 
Description ASC 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Study participants or study members
Results and Impact Presentation within a Food Fraud Panel at the American Society of Criminology 2015 Conference. Washington DC
Year(s) Of Engagement Activity 2015
 
Description British Library 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Industry/Business
Results and Impact Presentation - Issues in Preventing Food Fraud
Year(s) Of Engagement Activity 2016
 
Description Colloquium: Blurred Boundaries: Compliance to Non-Compliance - Understanding the threat of to sustaing food integrity 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact Meeting across industry, practitioners and academics to explore the notion of blurred boundaries and securing of compliance
Year(s) Of Engagement Activity 2015
 
Description Conference 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Public/other audiences
Results and Impact Presentation - Challenges for Prevention rather than Strategies for Prediction
Year(s) Of Engagement Activity 2017
 
Description Conference 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Other audiences
Results and Impact Presentation - The Dynamics of Food Fraud:The interactions between criminal opportunity and market (dys)functionality in legitimate businesses
European Society of Criminology - Muenster Germany
Year(s) Of Engagement Activity 2016
 
Description Conference Presentation - Tilburg _Regulation and Governance (ECPR Standing Group on Regulatory Governance) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Policymakers/politicians
Results and Impact Paper presented - Complexity in Food Regulation
Year(s) Of Engagement Activity 2016
 
Description ESC 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Other audiences
Results and Impact Presentation at the European Society of Criminology Conference - Porto
Year(s) Of Engagement Activity 2015
 
Description ESRC/FSA Food Fraud Findings 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Public/other audiences
Results and Impact Presentation to a dissemination event
Year(s) Of Engagement Activity 2017
 
Description FDA China Programme 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Policymakers/politicians
Results and Impact Presentation - Complexity in Food Fraud regulation
Year(s) Of Engagement Activity 2017
 
Description Food Fraud Task Force 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Policymakers/politicians
Results and Impact Presentation of criminological analysis of case file to Food Fraud Task Force (Ireland)
Year(s) Of Engagement Activity 2016
 
Description Food Integrity 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact Organisation of a Workshop at the Food Integrity EU Conference in Prague April 2016.
Year(s) Of Engagement Activity 2016
 
Description Food Integrity 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Industry/Business
Results and Impact Issues in Supply Chain Modelling for Fraud Detection
Food Integrity Prague
Year(s) Of Engagement Activity 2016