Agribusiness Management · Agri-sector specialisation · also ABM 530
What quality means for food and farm products, how it is managed and measured, and the national and international standards that govern trade.
In the exams: agribusiness management PhD entrance and fellowship examinations, quality management (present)
Quality management decides what quality a product must have, builds the processes that deliver it, checks that they do, and improves them. For food and farm products it has two faces: market quality (size, colour, purity, moisture, taste, keeping quality), expressed through grades, specifications and brands; and safety (freedom from pathogens, toxins, residues and foreign bodies), which governments regulate and do not compromise on. A mango exporter, a dairy cooperative, a flour mill and a supermarket buyer all live with both.
This manual covers quality concepts and the thinkers who shaped the field, total quality management and the cost of quality, grades and standards with India's AGMARK and BIS and grading systems of other countries, ISO 9001 and the ISO 22000 family, certification and accreditation, the Codex Alimentarius and HACCP, good agricultural and manufacturing practices, acceptance sampling and Six Sigma for agrifood lots, and the WTO SPS and TBT Agreements that link standards to world trade. Control charts and process capability are in the production and operations management manual.
The subject is examined in the ICAR AICE-JRF/SRF agri-business management paper, in UGC-NET Management within operations and TQM, and in agribusiness MBA and PGDM courses. The content is the same worldwide in outline; the national layer (who grades grain, which law makes a standard mandatory, which mark goes on the pack) differs, and where this manual names a national body it says which country.
No single definition of quality satisfies every use, and examiners expect students to know the main ones and who gave them. Joseph Juran defined quality as fitness for use: a product is of good quality if it serves the user's purpose. Philip Crosby defined it as conformance to requirements: once requirements are written down, quality means meeting them, and anything else is a nonconformance. The ISO 9000 vocabulary defines quality as the degree to which a set of inherent characteristics of an object fulfils requirements, where requirements may be stated, generally implied or obligatory. Genichi Taguchi defined poor quality as the loss a product imposes on society from the time it is shipped. Armand Feigenbaum stressed that quality is what the customer says it is.
David Garvin grouped the common views into five approaches. The transcendent approach treats quality as innate excellence, recognised but hard to define. The product-based approach ties quality to measurable attributes, such as protein content in wheat or fat in milk. The user-based approach says quality lies in the eye of the user. The manufacturing-based approach equates quality with conformance to specification. The value-based approach relates quality to price: a product is good if it performs at an acceptable cost.
Garvin's eight dimensions of product quality are performance, features, reliability, conformance, durability, serviceability, aesthetics and perceived quality. They were written for manufactured goods, but most translate to food: performance is how a flour bakes, conformance is meeting the grade, durability becomes shelf life, aesthetics is appearance and flavour, and perceived quality is reputation.
Food scientists group attributes as sensory (appearance, texture, flavour), nutritional, safety (freedom from biological, chemical and physical hazards), functional (behaviour in processing, such as gluten strength in wheat) and process attributes (organic, fair trade, free range). Process attributes cannot be detected in the product, which is why they depend on certification.
For services, the SERVQUAL model of Parasuraman, Zeithaml and Berry measures the gap between expected and perceived service on five dimensions: reliability, assurance, tangibles, empathy and responsiveness. The Kano model classifies attributes as must-be (taken for granted, but their absence causes dissatisfaction), one-dimensional (satisfaction rises with performance), attractive (unexpected delights), indifferent and reverse. Food safety is the clearest must-be attribute: no consumer thanks a processor for safe food, but one incident can destroy a brand.
Economists classify quality attributes by when the buyer can verify them. Search attributes can be judged before purchase: the colour of tomatoes, the size of onions, visible damage. Experience attributes are known only after consumption: taste, tenderness, cooking quality of rice. Credence attributes cannot be verified by the buyer even after consumption: pesticide residues, microbial safety, organic production, geographical origin, animal welfare. The first two categories come from Phillip Nelson; the credence category from Darby and Karni.
This explains why quality institutions exist. When sellers know quality and buyers do not, information is asymmetric. In George Akerlof's market for lemons, buyers offer only an average price, sellers of good quality withdraw, and average quality falls. Ungraded produce markets show the pattern: the farmer who dries and cleans grain carefully gets the same price as the one who does not.
The remedies are signals and third-party verification. Grades turn search attributes into a common language; brands and reputation carry experience attributes across repeat purchases; credence attributes need government regulation (for safety) or independent certification (for claims such as organic or GAP). Each remedy has a cost, and the manager's question is whether the premium or market access it buys exceeds that cost.
Two more distinctions recur. Quality of design is the level of quality intended in the specification; quality of conformance is how closely production meets that specification. A basic-grade product made exactly to its specification has high conformance quality even though its design quality is modest. And food quality is not the same as food safety: a bruised apple is of low quality but safe; a perfect-looking apple with an illegal residue is unsafe. Regulation sets a floor on safety; markets reward quality above it.
Quality management passed through four stages. In the inspection era, inspectors separate from workers sorted good output from bad. In the statistical quality control era, Walter Shewhart at Bell Telephone Laboratories introduced the control chart in the 1920s, and Harold Dodge and Harry Romig developed acceptance sampling tables. In the quality assurance era, after the Second World War, attention moved to preventing defects through planning and design. In the strategic TQM era, from the 1980s, quality became a basis of competition led by top management.
W. Edwards Deming, a statistician, taught that most quality problems come from the system that management controls, not from workers. His 14 points include creating constancy of purpose toward improvement, ceasing dependence on mass inspection, ending the practice of awarding business on price alone, improving constantly the system of production and service, instituting training and leadership, driving out fear, breaking down barriers between departments, eliminating slogans, targets and numerical quotas, and removing barriers to pride of workmanship. His seven deadly diseases of management include lack of constancy of purpose and emphasis on short-term profits. His system of profound knowledge has four parts: appreciation for a system, knowledge of variation, theory of knowledge and psychology. He taught the plan-do-check-act cycle, which he credited to Shewhart and later called plan-do-study-act.
Joseph Juran defined quality as fitness for use and set out the quality trilogy of quality planning, quality control and quality improvement, parallel to the financial processes of budgeting, cost control and cost reduction. He applied the Pareto principle to quality: a vital few causes account for most of the loss, and the rest are the useful many. He emphasised project-by-project improvement and the cost of poor quality, and edited the Quality Control Handbook.
Philip Crosby, in Quality Is Free, argued that the cost of doing things wrong far exceeds the cost of doing them right. His four absolutes are that quality means conformance to requirements, the system of quality is prevention, the performance standard is zero defects, and the measurement of quality is the price of nonconformance. He set out a 14-step improvement programme beginning with management commitment.
Armand Feigenbaum coined total quality control: quality must be managed across the whole cycle from marketing and design to service, and all departments share responsibility. He also developed the classification of quality costs into prevention, appraisal and failure costs.
Kaoru Ishikawa promoted company-wide quality control, in which everyone from top management to shop floor takes part, and quality circles: small groups of workers from the same area who meet voluntarily to identify and solve quality problems. He popularised the cause-and-effect or fishbone diagram and the seven basic quality tools, and argued that most problems can be solved with these simple tools.
Genichi Taguchi introduced the quality loss function: any deviation from the target value imposes a loss that rises with the square of the deviation, L = k(y - m) squared, where y is the actual value, m the target and k a constant found by dividing the loss at the specification limit by the square of the tolerance. This challenges the goalpost view that everything inside the specification is equally good. Taguchi also developed robust design, in which products and processes are made insensitive to noise factors, using orthogonal arrays and signal-to-noise ratios.
Shigeo Shingo, associated with the Toyota Production System, developed poka-yoke (mistake-proofing devices that make errors impossible or immediately visible) and single-minute exchange of dies, and argued for zero quality control through source inspection. Masaaki Imai popularised kaizen, continuous improvement in small steps involving everyone. Yoji Akao developed quality function deployment (QFD), which translates the voice of the customer into technical requirements through a matrix called the house of quality.
Total quality management (TQM) is an organisation-wide management approach in which quality, defined by the customer, is the basis of strategy, and every person and process is engaged in improving it continuously. Textbooks list its principles in slightly different words, but they converge on customer focus, leadership and top-management commitment, total employee involvement, a process approach, a systematic and integrated approach, continuous improvement, fact-based decision making, and mutually beneficial supplier relationships. These are close to the seven quality management principles on which the ISO 9000 family rests: customer focus, leadership, engagement of people, process approach, improvement, evidence-based decision making and relationship management.
Continuous improvement is organised through the PDCA cycle: plan (study the problem and plan a change), do (try it on a small scale), check or study (measure the result), act (standardise the change if it worked, or begin again). Kaizen is the culture of many small improvements; breakthrough or re-engineering projects seek large step changes. Benchmarking compares one's processes with the best, whether competitors, other industries or other units of the same firm. Employee involvement takes forms such as quality circles, suggestion schemes and cross-functional teams, and depends on training and on management removing fear.
Ishikawa's seven basic tools are the check sheet (a form for recording data at the point of observation), the histogram (the distribution of a measured variable), the Pareto chart (causes ranked by frequency or cost, with a cumulative line, to find the vital few), the cause-and-effect or fishbone diagram (possible causes grouped under headings such as man, machine, method, material, measurement and environment), the scatter diagram (relation between two variables), the control chart, and stratification or a flow chart depending on the author.
A worked example of Pareto analysis: a spice packer records 300 customer complaints in a year: 120 for foreign matter, 80 for torn pouches, 40 for short weight, 30 for caking, 20 for colour and 10 for other causes. The two largest categories account for 200 of 300, or 66.7 per cent. The Pareto chart directs effort first to cleaning (sieving, magnets, destoning) and to pouch sealing, rather than spreading the budget across all six causes.
National quality awards provide frameworks for self-assessment. The Deming Prize was instituted by the Union of Japanese Scientists and Engineers (JUSE) in 1951. The Malcolm Baldrige National Quality Award was established in the United States in 1987. The EFQM Excellence Model, from the European Foundation for Quality Management founded in 1988, is used across Europe and has been revised several times. In India, BIS instituted the Rajiv Gandhi National Quality Award in 1991. These models score enablers such as leadership, strategy, people and processes, and results for customers, people, society and the business; an audit asks whether a system conforms, while an excellence model asks how good the results are.
The cost of quality (COQ) is the total cost of ensuring conformance and of dealing with nonconformance. The prevention-appraisal-failure model, associated with Feigenbaum, has four categories. Prevention costs stop defects: quality planning, training, supplier and farmer development, process validation, preventive maintenance, HACCP plan development. Appraisal costs find defects: incoming inspection, residue and microbial testing, in-process checks, calibration, audits. Internal failure costs arise from defects found before the product reaches the customer: scrap, rework, reprocessing, downgrading produce, lost yield, re-inspection. External failure costs arise after delivery: returns, complaints, discounts, recalls, rejected export consignments, liability claims and lost customers.
Crosby divided the same total into the price of conformance (prevention and appraisal) and the price of nonconformance (internal and external failure). Many firms that measure COQ for the first time find that failure costs dominate and that much of the total is hidden: the visible costs of scrap and recalls are the tip of an iceberg whose underwater part includes lost sales, extra inventory held as a buffer against defects, overtime and management time spent firefighting.
A worked example: a fruit-pulp processor with sales of Rs 800 lakh records prevention Rs 4 lakh, appraisal 6, internal failure 10 and external failure 20. Total COQ is Rs 40 lakh, 5 per cent of sales; failure costs are 75 per cent of COQ. If prevention rises to Rs 10 lakh and appraisal then falls to 5, internal failure to 6 and external failure to 8, total COQ falls to Rs 29 lakh: an extra Rs 6 lakh of prevention removed Rs 17 lakh of other costs, a net saving of Rs 11 lakh.
The classic economic model of quality of conformance, found in older textbooks and associated with Juran's early work, draws two curves against conformance level. The cost of conformance (prevention plus appraisal) rises steeply as conformance approaches 100 per cent, while failure costs fall. Their sum is U-shaped, with a minimum below 100 per cent conformance: on this view some defects are economical.
The modern view, associated with Crosby, Deming and Japanese experience, holds that process improvement makes defect-free output cheaper rather than dearer and that failure costs, especially lost customers, are larger than accounts show; total cost then keeps falling toward zero defects. In food safety the question is settled by law and ethics: no firm may accept an economical rate of Salmonella contamination. In practice, express COQ as a percentage of sales and track the share of failure costs.
The goalpost view treats every unit inside the specification limits as equally good and every unit outside as equally bad. Taguchi argued that loss begins as soon as a characteristic departs from its target and rises with the square of the departure. For a nominal-is-best characteristic the loss per unit is L(y) = k(y - m) squared, where m is the target and k is the loss coefficient. If a customer suffers a loss A when the characteristic reaches the specification limit, at a distance delta from the target, then k = A divided by delta squared.
Example: a 500 g pack of lentils has specification 500 plus or minus 10 g, and a pack at either limit causes a loss of Rs 20. Then k = 20/100 = 0.2 rupees per gram squared, and a pack at 506 g causes a loss of 0.2 x 36 = Rs 7.20 although it is within specification. For a process with mean mu and standard deviation sigma, expected loss per unit is k(sigma squared + (mu - m) squared); a filler running at mean 502 g with standard deviation 3 g loses 0.2 x (9 + 4) = Rs 2.60 per pack. The two routes to lower loss are centring the process on target and reducing variation. For smaller-is-better characteristics such as moisture, L = ky squared.
A standard, in the ISO/IEC definition, is a document established by consensus and approved by a recognised body that provides, for common and repeated use, rules, guidelines or characteristics for activities or their results. In agricultural marketing, grade standards divide a commodity into classes of defined quality; grading is sorting produce into those classes; standardisation is the wider act of setting grade limits, pack sizes, labelling and test methods.
Standards fall into several types. Product standards specify characteristics of the product: moisture, foreign matter, broken grains, size, colour, fat content. Process standards specify how it is produced: hygiene practices, organic methods. Management-system standards specify how the organisation manages quality, safety or the environment. Test-method and sampling standards specify how to measure. Standards may be mandatory, when a law or regulation makes them compulsory (in WTO language these are technical regulations or SPS measures), or voluntary. And they may be public, set by governments and intergovernmental bodies, or private, set by retailers, industry groups or civil-society schemes.
Grades allow trade by description, so a distant buyer can purchase without seeing the lot; they reduce search and inspection costs and disputes; they improve price discovery because quotes for a defined grade are comparable; they allow pooling and storage, as in warehouse receipt systems; they are a precondition for futures contracts, which define the deliverable grade and its premiums and discounts; and they send consumer preferences back to farmers through grade premiums.
A useful grade standard has a small number of grade-determining factors that matter to users and can be measured quickly and cheaply at the point of trade. For grains these are typically moisture, foreign matter, damaged and discoloured grains, broken grains, admixture of other varieties, weevilled grains and test weight; for fruits and vegetables, size or weight, colour, shape, freedom from defects and decay; for livestock and meat, carcass weight, fat cover and conformation; for milk, fat and solids-not-fat and hygienic quality.
Moisture and foreign matter are usually handled by price or weight adjustments. A dry-matter adjustment converts a wet lot to its weight at the standard moisture: equivalent weight = actual weight x (100 - actual moisture)/(100 - standard moisture). Ten tonnes of maize at 16 per cent moisture equals 10 x 84/88 = 9.545 tonnes at 12 per cent, not 9.6 tonnes, the common slip of subtracting four points directly. A foreign-matter deduction removes the excess over a tolerance: 50 quintals with 3 per cent foreign matter and a 1 per cent tolerance, with a one-for-one deduction, is paid as 49 quintals. The contract sets the exact rule, which varies by country and buyer.
Grading pays a seller when the extra revenue from selling graded lots exceeds the cost of grading, which includes labour, equipment, packaging and any certification fee. Suppose a farmer group has 100 quintals of a pulse that sells ungraded at Rs 2,000 per quintal, giving Rs 2,00,000. Grading yields 40 quintals of grade A at Rs 2,400, 45 quintals of grade B at Rs 2,000 and 15 quintals of grade C at Rs 1,600, for Rs 2,10,000. If grading costs Rs 50 per quintal on the whole 100 quintals, net revenue is Rs 2,05,000, a gain of Rs 5,000 or Rs 50 per quintal of produce. The break-even grading cost is Rs 100 per quintal. A seller who ignores grading cost overstates the gain as Rs 10,000.
In the United States, the Agricultural Marketing Service of the Department of Agriculture sets and applies voluntary grade standards for fruits, vegetables, meat, poultry, eggs, dairy products and cotton. Beef carcasses are graded for quality into Prime, Choice, Select, Standard, Commercial, Utility, Cutter and Canner, mainly on marbling and maturity, and separately for yield. Grain is graded under the United States Grain Standards Act by the Federal Grain Inspection Service, with numerical grades (U.S. No. 1 downwards for most grains) plus a sample grade. Meat safety inspection is mandatory and separate from voluntary quality grading, a distinction examiners like to test.
In the European Union, marketing standards are set under the common market organisation regulation. For fresh fruit and vegetables, specific standards commonly use three classes, Extra Class, Class I and Class II, above minimum requirements that apply to all classes. The United Nations Economic Commission for Europe (UNECE) develops voluntary commercial quality standards for fresh and dried produce, seed potatoes, meat and eggs, with which many countries align; they concern commercial quality, while safety is Codex's domain. In China, mandatory national standards carry the prefix GB and recommended ones GB/T; food safety standards are mandatory. In East Africa, regional East African Standards for staple grains set moisture, foreign matter and aflatoxin limits. The common pattern everywhere is a voluntary commercial grading system, a national standards body and a mandatory food-safety regime; agencies and compulsory requirements vary by country and change over time.
AGMARK is India's certification mark for agricultural and allied produce. Its legal basis is the Agricultural Produce (Grading and Marking) Act, 1937, amended in 1986, which empowers the central government to prescribe grade designations, define quality, specify grade designation marks and authorise persons to grade and mark produce. The Act is administered by the Directorate of Marketing and Inspection (DMI) under the Ministry of Agriculture and Farmers Welfare, with its head office at Faridabad and branch and regional offices across the country.
A packer who wishes to use the AGMARK label obtains a certificate of authorisation from DMI, grades the produce in an approved laboratory, either the packer's own or a state-grading or commercial laboratory, and affixes the grade designation mark. DMI runs a network of AGMARK laboratories, headed by the Central Agmark Laboratory at Nagpur, with regional laboratories, which analyse check samples and develop methods. DMI inspects premises and draws check samples from the market; misgrading can lead to suspension or cancellation of the authorisation. In recent years DMI has moved certification and laboratory management online.
AGMARK certification is voluntary for most commodities. At various times particular regulations, such as food-safety rules for certain products or government procurement conditions, have required AGMARK, and those requirements change; a manager must check the current notification for the commodity in question.
DMI distinguishes grading for internal trade (the packer-level grading of consumer packs described above), grading at the producers' level, and grading for export. Grading at producers' level is meant to help farmers sell on a grade basis in regulated markets: grading units in market yards grade the produce before auction so that the price reflects quality. Export grading under AGMARK was historically more important than it is now; export quality control in India today rests largely with other bodies, described below.
DMI also administers the Agricultural Marketing Information Network (AGMARKNET), a portal of market arrivals and prices, and is involved in marketing infrastructure and market reform schemes. The AGMARK name therefore appears in two senses in Indian examinations: the quality mark and the price-information network.
The Food Safety and Standards Act, 2006 created the Food Safety and Standards Authority of India (FSSAI) under the Ministry of Health and Family Welfare, replacing the Prevention of Food Adulteration Act, 1954 and product orders such as the Fruit Products Order, 1955. FSSAI sets mandatory standards for foods, additives, contaminants, residues and labelling, licenses and registers food businesses, and is India's Codex contact point; a food must meet FSSAI standards whether or not it carries AGMARK or BIS marks. For exports, the Export (Quality Control and Inspection) Act, 1963 established the Export Inspection Council (EIC), which certifies notified commodities. The Agricultural and Processed Food Products Export Development Authority (APEDA) promotes agricultural and processed food exports and operates the National Programme for Organic Production (NPOP) and residue-monitoring and traceability systems, for example for grapes to the European Union.
The Quality Council of India (QCI), set up in 1997, houses the National Accreditation Board for Certification Bodies (NABCB) and runs an India GAP scheme; the National Accreditation Board for Testing and Calibration Laboratories (NABL) accredits laboratories; and PGS-India, run by the Ministry of Agriculture, is a low-cost group-based organic guarantee for the domestic market.
The Bureau of Indian Standards (BIS) is India's national standards body. It was set up under the Bureau of Indian Standards Act, 1986, taking over the Indian Standards Institution founded in 1947, and now operates under the Bureau of Indian Standards Act, 2016. It works under the Ministry of Consumer Affairs, Food and Public Distribution and has its headquarters in New Delhi. BIS formulates Indian Standards (designated IS followed by a number) through technical committees of producers, users, regulators, laboratories and consumers, and represents India in ISO and the International Electrotechnical Commission.
BIS's product certification scheme licenses manufacturers to use the Standard Mark, the ISI mark, when product and production control meet the Indian Standard, followed by surveillance through factory inspections and market samples. Other schemes include compulsory registration (mainly electronics), hallmarking of gold and silver, and management system certification. Certification is voluntary unless made compulsory by a quality control order or regulation; for food, packaged drinking water (IS 14543) and natural mineral water (IS 13428) need BIS certification under food-safety regulations. Because the compulsory list changes, examiners test the principle, not the list. In short: AGMARK certifies grade and purity of produce (Ministry of Agriculture); BIS sets national standards and certifies mainly manufactured goods (Ministry of Consumer Affairs); FSSAI sets the mandatory food-safety floor (Ministry of Health).
Conformity assessment is any activity that shows whether specified requirements are met. Its main forms are testing (determining characteristics by a procedure, done by laboratories), inspection (examining a product, process or installation and judging conformity, often by professional judgement), certification (a third party giving written assurance that a product, process, system or person conforms) and accreditation (formal recognition that a conformity assessment body is competent to carry out specific tasks).
Assessment is first party when the supplier declares its own conformity, second party when the customer audits its supplier, and third party when an independent body does so. A retailer that audits its produce suppliers against its own code is performing second-party assessment; a certification body that audits a packhouse against a GAP standard is third party. A supplier's declaration of conformity is first party and is accepted by regulators in some sectors.
The ISO/IEC 17000 series sets requirements for the bodies involved. ISO/IEC 17025 covers the competence of testing and calibration laboratories. ISO/IEC 17020 covers inspection bodies. ISO/IEC 17065 covers bodies that certify products, processes and services, which is the standard under which GAP, organic and many food product schemes operate. ISO/IEC 17021-1 covers bodies that audit and certify management systems, such as ISO 9001 and ISO 22000. ISO/IEC 17011 covers accreditation bodies themselves. ISO 19011 gives guidance on auditing management systems.
Accreditation answers the question: who checks the checkers? A certification body's certificate is only as credible as its competence and impartiality. Accreditation bodies assess certification bodies, laboratories and inspection bodies against the ISO/IEC 17000-series standards and accredit them for a defined scope. Normally there is one national accreditation body per country: in India, NABCB accredits certification and inspection bodies and NABL accredits laboratories; other countries have their own national bodies.
International recognition comes through multilateral arrangements. The International Accreditation Forum (IAF) runs a multilateral recognition arrangement for accreditation of certification bodies, and the International Laboratory Accreditation Cooperation (ILAC) runs a mutual recognition arrangement for laboratories and inspection bodies; regional cooperations feed into them. The aim is tested once, certified once, accepted everywhere: a residue test from an accredited laboratory in one signatory country should be accepted in another without retesting.
The International Organization for Standardization (ISO), founded in 1947 and based in Geneva, is a non-governmental federation of national standards bodies, one per country; India's member is BIS. ISO standards are voluntary and developed by technical committees through consensus. ISO does not itself certify anyone: certification is done by independent certification bodies, accredited by national accreditation bodies.
The ISO 9000 family deals with quality management systems (QMS). ISO 9000 sets out fundamentals and vocabulary; ISO 9001 specifies requirements and is the only certifiable standard in the family; ISO 9004 gives guidance on sustained success; ISO 19011 guides auditing. ISO 9001 was published in 1987 and revised in 1994, 2000, 2008 and 2015; a 2024 amendment added climate change to the issues an organisation must consider, and a further revision has been in preparation, so the current edition must be checked. The 2000 revision merged the earlier 9001, 9002 and 9003 into one standard and introduced the process approach.
ISO 9001:2015 is built on the seven quality management principles (customer focus, leadership, engagement of people, process approach, improvement, evidence-based decision making and relationship management), the process approach, the PDCA cycle and risk-based thinking. Like other modern ISO management-system standards it follows a common high-level structure of ten clauses: 1 scope, 2 normative references, 3 terms and definitions, 4 context of the organisation, 5 leadership, 6 planning, 7 support, 8 operation, 9 performance evaluation and 10 improvement. The common structure makes it easier to run an integrated management system combining quality, environment (ISO 14001), occupational health and safety (ISO 45001) and food safety (ISO 22000).
ISO 22000 specifies requirements for a food safety management system (FSMS) for any organisation in the food chain, from feed and primary production through processing, storage, transport and retail to packaging and equipment suppliers. First published in 2005, its current edition is 2018. It combines four key elements: interactive communication, system management, prerequisite programmes and HACCP principles. The 2018 edition follows the common high-level structure and uses two PDCA cycles, one for the management system and one for operational planning and control.
ISO 22000 distinguishes three kinds of control measure. Prerequisite programmes (PRPs) are basic conditions and activities needed to maintain a hygienic environment, such as cleaning, pest control, water quality, personnel hygiene and maintenance. Operational prerequisite programmes (OPRPs) are control measures identified by the hazard analysis as needed to control a significant hazard, with an action criterion and measurable or observable monitoring, but not managed as a CCP. Critical control points (CCPs) are steps where a control measure with measurable critical limits is applied to prevent or eliminate a significant hazard or reduce it to an acceptable level. The OPRP concept is one of the main differences between ISO 22000 and classical HACCP.
Beyond quality and food safety, several ISO standards are common in agribusiness. ISO 14001 specifies requirements for an environmental management system and is often used by exporters and plantation companies. ISO 45001, which replaced the earlier OHSAS 18001, covers occupational health and safety. ISO 50001 covers energy management, relevant to cold chains and processing plants. ISO 26000 gives guidance on social responsibility and is explicitly not intended for certification. ISO/IEC 17025 for laboratories matters to every firm that relies on residue, aflatoxin or microbiological test results. ISO 2859-1 (attributes) and ISO 3951 (variables) set sampling schemes for inspection, discussed in the SQC chapter. ISO also publishes many test methods for cereals, oilseeds, spices, milk and meat, which national standards often adopt as identical standards.
The Codex Alimentarius, Latin for food code, is the collection of international food standards, guidelines and codes of practice adopted by the Codex Alimentarius Commission (CAC), established by FAO and WHO, which held its first session in 1963 and implements the Joint FAO/WHO Food Standards Programme. Its purposes are protecting consumers' health and ensuring fair practices in the food trade. It has close to 190 members, including the European Union, and its secretariat is hosted by FAO in Rome.
Subsidiary bodies include general subject committees (for example food hygiene, food labelling, food additives, contaminants, pesticide residues, veterinary drug residues, methods of analysis and sampling, import and export inspection and certification, and general principles), commodity committees, and regional coordinating committees, each hosted by a member country. India hosts the Codex Committee on Spices and Culinary Herbs.
Scientific advice comes from independent joint FAO/WHO expert bodies: the Joint Expert Committee on Food Additives (JECFA), which also evaluates contaminants and veterinary drug residues; the Joint Meeting on Pesticide Residues (JMPR); and the Joint Expert Meetings on Microbiological Risk Assessment (JEMRA). This separation reflects the risk-analysis framework Codex uses: risk assessment is done by scientists, risk management by the Commission and governments, and risk communication runs between assessors, managers and stakeholders.
Codex texts carry prefixes: standards (CXS), including commodity standards and general standards such as those for labelling of prepackaged foods, food additives, and contaminants and toxins in food and feed; guidelines (CXG); and codes of practice (CXC), the most important being the General Principles of Food Hygiene (CXC 1-1969), which contains the HACCP text. Codex also adopts maximum residue limits (MRLs) for pesticides and veterinary drugs. Standards pass through a step procedure, usually eight steps, and are adopted by consensus where possible.
Codex standards are voluntary: Codex has no power to enforce them, and a country need not adopt them. They gain force in three ways. Countries adopt them into national law, often as the basis of their own standards, as FSSAI does in harmonising many limits. The WTO SPS Agreement names Codex as the reference international standard-setting body for food safety, so a measure based on a Codex standard is presumed consistent with the Agreement. And the TBT Agreement requires technical regulations to be based on relevant international standards, which for food labelling and composition often means Codex. Each member designates a Codex contact point; India's is FSSAI.
Hazard Analysis and Critical Control Point (HACCP) is a science-based, preventive system that identifies, evaluates and controls hazards significant for food safety. It originated in the 1960s in work to supply safe food for the United States space programme and was set out internationally by Codex, most recently in the 2020 revision of the General Principles of Food Hygiene, which also introduced the concept of food safety culture. A food hazard is a biological, chemical or physical agent with the potential to cause an adverse health effect; allergens are treated as chemical hazards. HACCP is built on good hygiene practices, the prerequisite programmes that must be in place first.
The seven principles in the current Codex text are: (1) conduct a hazard analysis and identify control measures; (2) determine the critical control points; (3) establish validated critical limits; (4) establish a system to monitor control of CCPs; (5) establish the corrective actions to be taken when monitoring indicates a deviation from a critical limit at a CCP; (6) validate the HACCP plan and then establish procedures for verification to confirm that the system works as intended; and (7) establish documentation concerning all procedures and records.
These principles are applied through a 12-step logic sequence. The five preliminary steps are: assemble the HACCP team and identify the scope; describe the product; identify the intended use and users; construct a flow diagram; and confirm the flow diagram on site. Steps 6 to 12 then apply the seven principles in order.
A critical control point is a step at which a control measure essential to control a significant hazard is applied. A critical limit separates acceptability from unacceptability and must be measurable or observable in real time: time and temperature in pasteurisation, pH in acidified vegetables, detection sensitivity of a metal detector. Validation asks whether the control measures can control the hazard; verification asks whether the plan is followed and working, through record review, calibration, sampling and audit. A decision tree helps identify CCPs but does not replace judgement.
Two examples show how HACCP works across products. In fluid milk processing, the significant biological hazards are vegetative pathogens; pasteurisation is the CCP, the critical limit is a time-temperature combination set by national regulation, monitoring is by a continuous recording thermometer, and the corrective action, a flow-diversion valve returning under-heated milk, is automatic. Receipt of raw milk is typically controlled by prerequisites such as cooling and by antibiotic residue screening.
In groundnut and maize handling, aflatoxin is the key chemical hazard. It is formed by moulds in the field and in storage, and no single processing step reliably removes it, so control depends mostly on prerequisite and upstream measures: timely harvest, rapid drying to safe moisture, clean dry storage, and sorting of damaged kernels, with testing at receipt to reject contaminated lots. This is an example of a hazard controlled largely through the chain rather than at a single CCP, which is why the ISO 22000 concept of an operational prerequisite programme is useful.
Good agricultural practices (GAP), as FAO describes them, are practices that address environmental, economic and social sustainability for on-farm processes and result in safe and quality food and non-food agricultural products. In certification schemes GAP usually covers site history and management, soil and substrate management, fertiliser use, water quality and irrigation, integrated pest management and plant protection product use (only approved products, applied by trained operators at label rates, with pre-harvest intervals observed and records kept), harvest hygiene, worker health, safety and welfare, traceability and record keeping, and environmental protection.
GAP exists at three levels: national GAP codes published by governments; regional and national certification schemes such as ASEAN GAP and an India GAP scheme run through the Quality Council of India; and private international standards, of which the best known began in 1997 as a European retailer initiative, EUREPGAP, renamed GLOBALG.A.P. in 2007. Its integrated farm assurance standard covers crops, livestock and aquaculture. Producers may be certified individually (option 1) or as a producer group with a quality management system and internal inspections (option 2); group certification spreads audit costs and is the main route by which smallholders meet export requirements. Control points are graded by severity, and the exact compliance rules change between versions. GAP certification is usually a buyer requirement, not a legal one, and private standards fall largely outside WTO disciplines, which bind governments.
Good hygiene practices (GHP), in Codex terms, are the fundamental measures and conditions applied at any step in the food chain to provide safe and suitable food: primary production, establishment design, training, maintenance, cleaning and pest control, personal hygiene, control of operation, product information and transport. Good manufacturing practices (GMP) is the equivalent term in many national laws and in feed and pharmaceutical industries. Related codes include good veterinary, laboratory, storage and distribution practices. General conditions are controlled through these prerequisites so that HACCP can concentrate on the few hazards needing specific control.
Process standards certify credence attributes such as organic production, fair trade, animal welfare, sustainability and geographical origin. Organic standards are set in national law in many markets, such as the European Union's organic regulation and the United States National Organic Program, and in India by NPOP for exports and by food-safety regulations for the domestic market; international guidance comes from Codex guidelines on organically produced foods and from the international organic movement's standards. Equivalence agreements between countries allow organic produce certified in one to be sold as organic in the other.
Traceability underpins all of these. The EU General Food Law requires food businesses to identify their immediate supplier and immediate customer, the one-step-back, one-step-forward rule. With each scheme the manager asks whether the premium or market access exceeds the cost of compliance. A GAP-certified mango exporter earning an 8 per cent premium on Rs 60 per kg across 20 tonnes gains Rs 96,000; if certification and compliance cost Rs 1,00,000, the scheme pays only if it secures a market that would otherwise be closed.
Statistical quality control (SQC) has three branches: descriptive statistics of quality characteristics, statistical process control (SPC) with control charts, and acceptance sampling. Control charts and capability indices are worked through in the production and operations management manual. Two points bear repeating: control limits come from the process's own data, while specification limits come from the customer; and capability means nothing until the process is in control. In agrifood trade, where buyers receive lots from many suppliers whose processes they cannot control, acceptance sampling is often the more important tool, and contaminants such as aflatoxin are so unevenly spread that sampling error often exceeds analytical error.
A single sampling plan by attributes is defined by sample size n and acceptance number c: inspect n units and accept the lot if c or fewer are defective. For a lot with fraction defective p, the probability of acceptance is binomial, Pa = the sum over d from 0 to c of C(n, d) p^d (1 - p)^(n - d); the Poisson with mean np approximates it when n is large and p small; for small lots the hypergeometric is exact.
Worked examples. Plan n = 20, c = 0, lot 5 per cent defective: Pa = 0.95^20 = 0.358. The same lot under n = 20, c = 1: Pa = 0.358 + 20 x 0.05 x 0.95^19 = 0.736. A plan n = 100, c = 2 on a lot 1 per cent defective, by Poisson with np = 1: Pa = e^-1 (1 + 1 + 1/2) = 0.920. For a lot of 20 cartons containing 2 defective cartons, a sample of 5 with c = 0 accepts with probability C(18, 5)/C(20, 5) = 8,568/15,504 = 0.553; the binomial with p = 0.1 would give 0.590, overstating acceptance slightly.
The operating characteristic (OC) curve plots Pa against p. The acceptable quality level (AQL) is the quality the plan is designed to accept most of the time, and the producer's risk (alpha) is the probability of rejecting a lot at the AQL. The lot tolerance percent defective (LTPD) is the poor quality the consumer wants rejected, and the consumer's risk (beta) is the probability of accepting a lot at the LTPD. For n = 50, c = 1: at p = 1 per cent, Pa = 0.911, so alpha is about 0.089; at p = 8 per cent, Pa = 0.083, so beta is about 0.08. With rectifying inspection, average outgoing quality AOQ is approximately Pa x p x (N - n)/N: for N = 2,000, n = 100, c = 2 and p = 0.01, AOQ = 0.920 x 0.01 x 1,900/2,000 = 0.874 per cent.
Food microbiology uses plans from the International Commission on Microbiological Specifications for Foods (ICMSF), adopted in Codex guidance and many regulations. In a two-class plan, n units are tested against a limit m, and the lot is accepted if no more than c exceed it; for pathogens such as Salmonella, c = 0 is usual. A plan with n = 10, c = 0 accepts a lot in which 10 per cent of units are contaminated with probability 0.9^10 = 0.349, which shows why testing cannot replace process control.
A three-class plan adds a second limit M. Units at or below m are acceptable, units between m and M are marginal, and units above M are defective; the lot is accepted if no unit exceeds M and no more than c units are marginal. With n = 5, c = 2, and a lot 80 per cent acceptable, 15 per cent marginal and 5 per cent above M, Pa = the sum over i from 0 to 2 of C(5, i) x 0.15^i x 0.80^(5 - i) = 0.750. The choice of n, c, m and M depends on hazard severity and on whether later handling will reduce or increase the hazard.
Six Sigma is a disciplined, data-driven improvement approach developed at an American electronics company in the 1980s and spread widely in the 1990s. A process at six-sigma quality has specification limits six standard deviations from the mean; allowing for a long-term drift of the mean by 1.5 standard deviations, which is a convention, this corresponds to 3.4 defects per million opportunities (DPMO). Under the same convention, 5 sigma is about 233 DPMO, 4 sigma about 6,210, 3 sigma about 66,807 and 2 sigma about 308,538.
DPMO = defects / (units x opportunities per unit) x 1,000,000. A fruit packhouse that finds 45 defects in 1,500 crates, each with 4 defect opportunities (grade, weight, label, damage), has 45/6,000 x 1,000,000 = 7,500 DPMO. The sigma level is the standard normal value whose upper tail equals the defect rate, plus 1.5: z for 0.0075 is 2.43, so the process runs at about 3.93 sigma. Forgetting the 1.5 shift gives 2.43, and forgetting the opportunities gives 30,000 DPMO. Defects per unit (DPU) is defects divided by units; under a Poisson assumption the first-pass yield is e^-DPU. Rolled throughput yield (RTY), the probability that a unit passes every step without defect, is the product of step yields: steps of 98, 95 and 97 per cent give 0.98 x 0.95 x 0.97 = 90.3 per cent, not the 96.7 per cent average.
Projects follow DMAIC: define, measure, analyse, improve, control; new designs use DMADV (define, measure, analyse, design, verify). Roles include champions (senior sponsors), master black belts, black belts (full-time project leaders) and green belts (part-time). Compared with earlier TQM, Six Sigma insists on measurable financial results, trained project leaders and rigorous statistics. Typical agrifood projects tackle fill-weight giveaway, yield loss, drying-moisture variability and complaint rates.
The World Trade Organization, established in 1995, administers two agreements that discipline food standards as non-tariff measures. The Agreement on the Application of Sanitary and Phytosanitary Measures (SPS Agreement) covers measures to protect human or animal life or health from additives, contaminants, toxins or disease-causing organisms in food and feed, human health from diseases carried by animals and plants, animal and plant life from pests and diseases, and a country from damage by pests. The Agreement on Technical Barriers to Trade (TBT Agreement) covers technical regulations, standards and conformity assessment procedures that are not SPS measures: for food, typically labelling, composition, quality, packaging and grades. A pesticide MRL is SPS; a rule on how a product name appears on the label is TBT; a labelling requirement directly related to safety, such as an allergen warning, is SPS.
The SPS Agreement recognises the right of members to take measures necessary to protect life or health, provided they are based on scientific principles, are not maintained without sufficient scientific evidence, do not arbitrarily or unjustifiably discriminate between members where similar conditions prevail, and are not disguised restrictions on trade.
Harmonisation (Article 3) encourages members to base measures on international standards: for food safety those of the Codex Alimentarius Commission, for animal health and zoonoses those of the World Organisation for Animal Health (WOAH, formerly OIE), and for plant health those of the International Plant Protection Convention (IPPC), the three sisters. Measures conforming to them are presumed consistent with the Agreement; a higher level of protection needs scientific justification or a risk assessment. Equivalence (Article 4) requires an importer to accept an exporter's different measures if they achieve the importer's appropriate level of protection. Risk assessment (Article 5) requires measures not based on international standards to rest on an assessment of risk and be no more trade-restrictive than required; where evidence is insufficient a member may adopt a provisional measure but must seek more information and review it (Article 5.7). Regionalisation (Article 6) requires recognition of pest- and disease-free areas. Transparency (Annex B) requires notification of new measures and a national enquiry point.
Disputes have shaped these rules. In EC hormones, brought by the United States and Canada, the Appellate Body found in 1998 that the EU ban on beef from hormone-treated cattle was not based on a risk assessment. Australia's salmon ban and Japan's apple measures for fire blight also failed risk-assessment obligations, and the United States successfully challenged India's avian influenza import restrictions.
The TBT Agreement distinguishes technical regulations, which are mandatory, from standards, which are voluntary, and covers their conformity assessment procedures. Technical regulations must give national and most-favoured-nation treatment, must not be more trade-restrictive than necessary to fulfil a legitimate objective, and should be based on relevant international standards; Annex 3 is a Code of Good Practice for standardising bodies. In the EC sardines dispute, Peru successfully challenged a European rule reserving the name sardines for one species, because it was not based on the Codex standard that allowed other species to be sold as sardines with a qualifier. Private standards set by retailers and industry often matter more to exporters than official measures; they bind through contracts, and whether they fall under WTO disciplines remains unresolved.
For developing countries standards are both barrier and catalyst: compliance costs fall heavily on smallholders, yet meeting standards has upgraded whole export sectors through group certification, contract farming and public investment in laboratories. The Standards and Trade Development Facility, a partnership including FAO, WOAH, the World Bank, WHO and WTO, funds SPS capacity projects. An exporter can compare expected costs: a 4 per cent chance of border rejection costing US$6,000 is an expected loss of US$240; pre-shipment testing at US$150 that cuts the chance to 0.5 per cent gives an expected cost of US$180.
Learn the definitions and who gave them, because examiners match thinkers to ideas. Be able to list HACCP's seven principles and 12 steps in order, the ISO 9001 principles and clause structure, the four cost-of-quality categories with an agrifood example of each, and the difference between AGMARK, BIS and FSSAI. Practise the numericals until they are routine: cost of quality as a share of sales, Taguchi loss, probability of acceptance for a sampling plan, DPMO and sigma level, rolled throughput yield, and the net gain from grading. Finally, be clear on SPS versus TBT, and on why Codex standards are voluntary yet decisive in trade disputes.
A quality manager in a processing plant, packhouse or procurement business uses this subject daily. Typical tasks: running the HACCP team and reviewing CCP records, preparing for ISO 22000 or GFSI-recognised certification audits, setting incoming-material specifications and sampling plans with suppliers, handling complaints through root-cause analysis, and reporting the cost of quality to management. Typical decisions: whether a failing lot is rejected, reworked or downgraded; whether a new supplier's certificate and test reports come from accredited bodies; whether a buyer's premium for GAP or organic certification covers the cost of compliance.
A senior manager with 25 years' experience sets quality strategy across plants and markets. The decisions are which certifications the business will carry and which markets they unlock, how much to invest in prevention (supplier development, laboratories, traceability) against expected failure costs such as recalls and border rejections, how to manage a crisis when a contamination or residue incident occurs, and how to build a quality culture that survives staff turnover. Such managers also represent their industry in national standards committees, negotiate with regulators on new limits, and judge when to lead a standard rather than follow it.
A researcher with long experience asks questions the textbook treats as settled. What is the causal effect of certification on smallholder incomes, after correcting for selection into schemes? Do grade premiums in traditional markets fail because of weak enforcement or because the grades do not track value? How do SPS measures and private standards affect trade flows, measured through gravity models and ad valorem equivalents of non-tariff measures? What sampling designs detect heterogeneous contaminants such as aflatoxin at acceptable cost? These questions sit where quality management meets development economics, trade policy and food science, and they inform the standards that governments and Codex adopt.
This manual does not reproduce commodity-by-commodity AGMARK, BIS or Codex limits, because they are numerous and revised often; it gives the principles and says where to look. It does not list which products are under compulsory BIS certification or compulsory AGMARK at present, since those notifications change. Control charts and process capability are covered in the production and operations management manual. It does not cover detailed laboratory methods, the internal clause-by-clause content of GFSI-recognised private schemes, the latest version rules of GAP standards, or the details of national food laws beyond India's main institutions and brief notes on the United States, European Union, China, Brazil and East Africa. Membership counts and edition dates of standards should be checked against the issuing body's website.
Original practice questions written to the syllabus and pattern of the exams named in each question's examTags. They are not past papers and do not reproduce any board's questions.
These notes are free to read here, with no account.
What the Laisha app adds:
Free for ten days, then a subscription. We would rather write that here than have you find it later.