Every day, from the district magistrate’s office to the Secretariat in Delhi, thousands of decisions shape how governance reaches citizens. Some of these decisions are made in seconds, almost mechanically, while others demand weeks of deliberation, consultations, and judgment calls. What explains this difference? Herbert A. Simon, the Nobel laureate who placed decision-making at the heart of administrative theory, offered a powerful answer through his classification of organizational decisions into two fundamental types: programmed and non-programmed. This distinction is not merely academic jargon – it helps us understand why some administrative tasks move smoothly through established channels while others require novel thinking at the highest levels.
Table of Contents
- Simon’s decision-centred view of administration
- What are programmed decisions?
- Key characteristics of programmed decisions
- Programmed decisions in Indian administration
- What are non-programmed decisions?
- Key characteristics of non-programmed decisions
- Non-programmed decisions in public administration
- Why the distinction matters
- The link with bounded rationality
- How decisions move between categories
- Techniques for handling each type
- Limitations and critiques of the framework
- Relevance in the age of algorithms
Simon’s decision-centred view of administration
Before exploring the two types, it helps to place the distinction in context. In his seminal 1947 work Administrative Behavior, Simon argued that organizations are essentially systems of decision-making. He challenged the assumption of classical economics that decision-makers are perfectly rational agents who evaluate every alternative. Instead, he showed that real administrators operate under limits of memory, attention, and cognitive capacity, which led him to develop the concept of bounded rationality and the idea of “satisficing” – choosing options that are good enough rather than optimal.
Within this broader framework, Simon noticed that not all decisions are alike. Some follow well-worn grooves, while others require breaking new ground. In his 1977 work, Simon proposed that decisions can be arranged along a continuum, with programmed decisions on one end and non-programmed decisions on the other. This classification has since become one of the most widely used frameworks for analysing administrative choices.
What are programmed decisions?
Programmed decisions are the workhorses of any organization. They are routine, repetitive, and follow a clear procedure that has been worked out over time. When a clerk processes a leave application, when the revenue department issues a standard certificate, or when a government warehouse reorders stationery because stock has fallen below a threshold – these are all programmed decisions.
Simon defined these decisions as those where a definite procedure is followed, meaning the decision-maker doesn’t have to treat the problem from scratch each time it arises. The rules, criteria, and even outcomes are largely known in advance.
Key characteristics of programmed decisions
Several features distinguish this category. Repetitiveness: These situations arise frequently, which is precisely why procedures have been developed. Structured nature: The problem is well-defined and the parameters are measurable. Low cognitive effort: The decision-maker relies on habits, standard operating procedures (SOPs), or management information systems rather than creative reasoning. Delegable: Because the rules are explicit, these decisions can be pushed down the hierarchy or automated. Short-term impact: The consequences are usually limited to day-to-day operations rather than long-term strategy.
The decision to order raw materials based on anticipated production and existing stock is a classic illustration. Though the calculation may involve several variables, the criteria are understood, so structure can be applied.
Programmed decisions in Indian administration
Public administration is full of programmed decisions. A Tehsildar approving a routine mutation entry in land records, the Public Distribution System issuing ration cards based on eligibility criteria, Income Tax Department processing standard refunds, or the Railway Reservation System allocating berths – all follow pre-set algorithms. Even the annual confidential report (ACR) process for government employees follows a programmed flow, with defined timelines and formats.
Digital tools have greatly expanded the space of programmed decisions. GST filings, Aadhaar authentication, e-challans for traffic violations, and UPI-based payments to beneficiaries under direct benefit transfer schemes are essentially programmed decisions executed at scale through technology.
What are non-programmed decisions?
On the opposite end of the spectrum sit non-programmed decisions – the ones that keep top administrators up at night. Simon described these as decisions that are novel, unstructured, and unusually consequential. No standard procedure exists because the problem has never arisen before in quite the same form, or because its implications are too significant to be left to routine handling.
Think of the decisions that were required during the onset of the COVID-19 pandemic. When states had to decide whether to impose lockdowns, how to arrange migrant worker transport, which industries to classify as essential, or how to conduct board examinations – none of these had a ready playbook. They required judgment, consultation, creativity, and the willingness to accept significant uncertainty.
Key characteristics of non-programmed decisions
Novelty: The problem is new or appears in an unfamiliar context. Unstructured nature: Criteria are not clearly defined, and information is often incomplete or contradictory. High cognitive demand: The decision-maker must exercise judgment, intuition, and creative problem-solving. Strategic impact: Consequences are usually long-term and affect the organization broadly. Concentrated at the top: These decisions typically reside with senior leadership because they shape direction and policy.
Because no formula exists, non-programmed decisions rely on what Simon called judgment, intuition, creativity, and rules of thumb. Organizations often respond by appointing specially trained managers or setting up dedicated units – expert committees, task forces, or empowered groups – to deal with these issues.
Non-programmed decisions in public administration
Examples abound. The decision to demonetise high-value currency notes in 2016, the rollout of the National Education Policy, the re-organization of states, or the design of a new welfare architecture like MGNREGA – each represented a leap into the unknown where existing procedures offered limited guidance. At the state level, decisions around disaster response, such as the 2018 Kerala floods or the 2023 Sikkim glacial lake outburst, required administrators to improvise in real time.
Even at the district level, a Collector deciding how to manage communal tension during an unusual festival combination, or how to allocate relief when a drought overlaps with an unseasonal flood, is making a non-programmed decision.
Why the distinction matters
Simon’s classification is more than a descriptive exercise – it has practical implications for how organizations design themselves. Different levels of an organization handle different types of decisions, with higher levels concentrating on non-programmed matters and lower levels executing programmed ones. This is why bureaucratic hierarchies are structured the way they are: field officers handle standardized cases, while secretaries and ministers grapple with policy novelty.
The framework also informs training. A probationer at the Lal Bahadur Shastri National Academy of Administration spends considerable time mastering SOPs, rules of business, and financial regulations – the toolkit for programmed decisions. As officers rise through the ranks, training shifts toward strategic thinking, crisis management, and negotiation – the competencies required for non-programmed decisions.
The link with bounded rationality
The distinction between the two decision types is closely linked to Simon’s theory of bounded rationality. Because humans have limited cognitive capacity and time, organizations reduce decision burden by converting as many problems as possible into programmed routines. Every time a recurring issue is codified into a procedure, cognitive energy is freed up for genuinely novel challenges. This is why, as Simon suggested, administrators should strive to convert non-programmed decisions into programmed ones over time, even though they should never stop making difficult decisions for fear of being wrong.
How decisions move between categories
One of the most fascinating insights in Simon’s framework is that the two categories are not static. Decisions migrate from the non-programmed zone to the programmed zone as organizations learn from experience. This movement is the essence of organizational learning.
Consider digital governance in India. When Aadhaar enrolment began in 2009, every aspect – from biometric standards to grievance redressal – was a non-programmed decision area. Over the years, as experience accumulated, most of these decisions were codified into clear procedures that UIDAI offices follow today. Similarly, the initial COVID vaccination rollout was a non-programmed decision space, but by the second wave, platforms like CoWIN had converted much of the logistics into a programmed, rule-based system.
Importantly, movement can also go the other way. A decision that was once routine may suddenly become non-programmed when circumstances change. Election Commission procedures for conducting polls are largely programmed, but a situation like conducting elections during a pandemic pushed many of those decisions back into the non-programmed space, requiring fresh thinking.
Techniques for handling each type
Simon also suggested that different techniques suit different types. For programmed decisions, organizations rely on habits, clerical routines, standard operating procedures, organizational structures (clearly defined roles and responsibilities), and increasingly, information systems and algorithms. The goal is efficiency and consistency.
For non-programmed decisions, the approach is different. Techniques include expert judgment, stakeholder consultation, scenario planning, pilot projects before full-scale implementation, and careful review of analogous situations from the past. The Indian practice of referring complex policy questions to committees or commissions – whether the Finance Commission for fiscal matters or a Group of Ministers for inter-ministerial issues – reflects this approach.
Limitations and critiques of the framework
While Simon’s typology is widely used, it isn’t without critics. The binary distinction can oversimplify the messy reality of administrative life, where most decisions are actually semi-programmed – partly routine, partly novel. A district magistrate approving a protest permit may follow a standard procedure but must also exercise judgment about local sensitivities.
The framework also tends to underplay the political and ethical dimensions of decisions. Public administration is not only about technical efficiency; it involves value choices, power relations, and questions of justice. The Indian Institute of Public Administration notes how bureaucratic risk aversion can push officers toward routine, well-documented decisions even when the situation genuinely requires bold, non-programmed thinking. This shows that decision categories are shaped not just by problem structure but also by incentives, accountability, and institutional culture.
Relevance in the age of algorithms
In an era where artificial intelligence and data analytics are reshaping governance, Simon’s distinction is more relevant than ever. Machine learning systems are extremely good at handling programmed decisions at scale – think of automated flagging of tax returns for scrutiny or AI-assisted crop insurance claim processing. But when decisions involve genuine novelty, human judgment remains irreplaceable. The real skill for the administrator of tomorrow is knowing which decisions to delegate to algorithms and which to hold close for human deliberation.
What do you think? Can you recall an administrative or policy decision in recent news that started as non-programmed but has since become routine? And where would you draw the line between the decisions that should remain with human administrators and those that can safely be handed over to automated systems?
References
- https://onlinelibrary.wiley.com/doi/10.1111/puar.13540
- http://samuellearning.org/decisionmaking/handout%202.pdf
- https://www.cvs.edu.in/upload/Rational%20Decision%20Making.pdf
- https://openstax.org/books/principles-management/pages/2-3-programmed-and-nonprogrammed-decisions
- https://ils.unc.edu/daniel/614/decisiontheory.html
- https://banotes.org/administrative-thinkers/decision-making-administration-herbert-simon/
- https://en.wikipedia.org/wiki/Bounded_rationality
- https://plutusias.com/elearning/wp-content/uploads/2025/07/Simon.docx.pdf
- https://www.iipa.org.in/cms/public/training_course/49
Leave a Reply