Dror’s Normative-Optimum Model for Public Policy

When governments craft major policies – from healthcare reforms to national education strategies – they are rarely working with complete information or perfectly rational minds. Real decisions are shaped by data and by values, political pressures, gut instincts, and lived experience. This tension between what is measurable and what is merely human sits at the centre of one of political science’s most influential frameworks: Yehezkel Dror’s Normative–Optimum Model. Proposed in his landmark 1968 work Public Policymaking Reexamined, the model offers a sophisticated third way between the extremes of pure rationalism and cautious incrementalism – one that formally accommodates both logic and judgment in the policy process.

The problem Dror was solving 🔗

To understand Dror’s contribution, it helps to know what he was reacting against. By the 1960s, two dominant approaches to public policy-making had emerged, and both left him dissatisfied.

The first was the rational-comprehensive model, championed by Herbert Simon. It held that a policy-maker should identify every possible alternative, calculate all outcomes, and select the single best option. The problem: this assumes perfect information, unlimited time, and an almost computer-like capacity for analysis – none of which exist in the real world.

The second was incrementalism, developed by Charles Lindblom in his famous 1959 essay “The Science of Muddling Through.” Incrementalists argued that, since pure rationality is impossible, policy-makers simply start with existing policy and make marginal adjustments. It’s safe and practical – but, as Dror pointed out, it is also deeply conservative. Incrementalism works well when problems are static, but it is poorly equipped to handle new, fast-moving, or large-scale challenges. Dror found the incremental approach unjust as well, because it tends to entrench existing power imbalances rather than challenge them.

Dror’s answer was to build a model that acknowledges the limits of pure rationality without abandoning the aspiration for better, more rigorous decision-making. He called it normative-optimalism.

What “normative-optimum” actually means 🔗

The name itself carries the model’s philosophy. Normative signals that the model is prescriptive – it describes how policy-making should operate, not merely how it currently does. Optimum means aiming for the best possible outcome given real-world constraints, not a theoretical perfect outcome. Together, these two words signal a model that is ambitious but grounded.

The model’s secret ingredient is its formal recognition of extra-rational elements. Dror argued that effective policy-making also involves intuition, creativity, and judgment that cannot be quantified. These are not flaws in the process to be eliminated – they are essential inputs. Value judgments, tacit knowledge gained from experience, political bargaining, and coalition-building are as legitimate a part of good policy-making as any statistical model.

Dror’s framework thus requires policy-makers to be trained in both rational and extra-rational techniques – a genuinely novel demand for its time. The result is what he called optimal rationality: not perfect rationality (which he viewed as impossible), but the best possible application of analysis within real-world constraints, supplemented by human judgment.

The three phases of the model 🔗

Dror’s optimal model includes three major stages – metapolicymaking, policymaking, and post-policymaking – closely interconnected by communication and feedback channels, broken down into 18 distinct phases. Understanding these three stages reveals how comprehensive and systematic Dror’s vision actually was.

Phase 1: The metapolicy-making stage 🔗

This is the most distinctive and intellectually original part of Dror’s model. Metapolicy-making means making policy about how to make policy. Before any specific problem is addressed, Dror insists that decision-makers must first design and evaluate the decision-making system itself.

This phase covers seven sub-stages: processing values, processing reality, processing problems, surveying and developing resources, designing and redesigning the policy-making system, allocating problems and resources, and determining the policymaking strategy. In practical terms, this involves clarifying the values that will guide decision-making, identifying who should be involved, what institutions need to participate, and how information will be gathered and weighed.

Why is this so important? Because a flawed decision-making process will produce flawed policies regardless of how much data is available. Dror recognised that no amount of good analysis can compensate for a poorly designed policy-making system. Attending to the structure of governance before diving into specific policy questions is what separates Dror’s model from almost every other framework of its era.

Phase 2: The policy-making stage 🔗

Once the metapolicy framework is in place, policy-makers move into the substantive work of developing and selecting policy options. This middle phase covers activities such as sub-allocating resources, establishing operational goals, identifying other significant values, preparing a realistic set of policy alternatives, and then comparing and selecting the best option available.

Crucially, this stage operates simultaneously at two interacting levels. At the rational sub-phase, decision-making involves gathering information on feasibility and opportunity costs. At the extra-rational sub-phase, it involves value judgments, tacit bargaining, and coalition formation.

Consider how this plays out concretely. When a national government evaluates competing approaches to public housing: the rational sub-phase generates data on construction costs, land availability, demographic projections, and budget constraints. The extra-rational sub-phase involves political negotiations about which communities are prioritised, ethical debates about land use rights, and judgments drawn from the experience of planners who have seen previous housing programmes succeed or fail. Dror insists both are necessary. A policy driven only by the numbers may be technically correct but politically dead on arrival. A policy driven only by political horse-trading may win broad support but solve nothing.

Phase 3: The post-policy-making stage 🔗

Many policy frameworks treat implementation as an afterthought. Dror does not. The third phase covers motivating the execution of the policy, monitoring how it is actually carried out, evaluating outputs against intended goals, and feeding that learning back into future decisions.

By including post-implementation stages, Dror’s model adopts a cyclical view of policy-making, where evaluation findings feed back into future policy development – creating a learning-oriented approach to governance that can adapt to changing circumstances and new information. This feedback loop is not merely procedural; it is the mechanism by which the entire model improves over time. Evaluation results might reveal that original value priorities were misplaced, or that the policy-making system itself needs redesign, pushing decision-makers back to Phase 1.

The two-level operation: Rational and extra-rational working together 🔗

The most analytically rich feature of Dror’s framework is how it handles the relationship between rational and extra-rational thinking across all three phases. Through his normative-optimum model, Dror seeks to accommodate qualitative rather than merely quantitative aspects of policy, while still aiming to increase the overall rational content of decision-making.

The processing of values in Phase 1 is the clearest illustration. At the rational sub-level, this means gathering feasibility data and analysing opportunity costs – hard evidence about what is actually achievable. At the extra-rational sub-level, it means holding value discussions with stakeholders, bargaining between competing interests, and building the political coalitions necessary to make any policy viable. Neither level operates independently. The rational analysis constrains what is proposed; the extra-rational process determines what is accepted and implemented.

Dror’s concept of normative optimalism argues that policy analysis must acknowledge the role of extra-rational understanding based on tacit knowledge and personal experience, with the goal of expanding decision-makers’ thinking to deal with a complex world. This is not an endorsement of gut instinct over evidence. It is an insistence that evidence alone is never sufficient, and that wisdom – accumulated through experience but difficult to quantify – is a legitimate and necessary input into governance.

A cyclical and comprehensive framework 🔗

One of Dror’s most enduring contributions is his insistence that policy-making is not a linear event but an ongoing, adaptive cycle. The feedback and communication channels that connect all three phases mean that every policy evaluation is simultaneously the beginning of the next policy design. What is learned from implementation informs revised values, updated problem definitions, and redesigned decision processes.

Dror combined the approaches of policy analysis, behavioural science, and systems analysis in his examination of the reality of public policymaking and his suggestions for its reform. This interdisciplinary ambition – drawing on economics, political science, organisational theory, and decision science – gives the model its breadth. It was designed not just as a theoretical framework but as a practical guide for improving real governance systems.

The model also makes a strong case for the inclusion of policy experts and specialist advisors throughout the process. Rather than treating expertise as something consulted only during technical sub-phases, Dror sees professional knowledge – including the tacit knowledge of experienced practitioners – as woven through every stage of policy-making.

Strengths and limitations 🔗

Dror’s framework has attracted considerable praise precisely because it takes complexity seriously. Its major strengths are its comprehensive scope (covering the full policy lifecycle, not just the decision moment), its intellectual honesty about the limits of rationality, and its explicit treatment of values as something to be processed deliberately rather than hidden.

Its limitations are equally clear. A recurring challenge is finding the right balance between thorough analysis and timely action. In crisis situations, policy-makers may not have the luxury of working through all 18 stages before making decisions. The model is also resource-intensive: implementing it fully requires significant time, expertise, and institutional capacity – things that are unevenly distributed, particularly in developing governance contexts.

There is also a democratic tension embedded in the model. As scholar critics have noted, Dror appears to have limited regard for direct public participation in policy-making, placing greater weight on expert judgment than on citizen input. This sets him apart from more participatory approaches that emerged in later decades.

Nonetheless, many of today’s most pressing policy challenges – from climate change to technological disruption to global health crises – are characterised by exactly the complexity, uncertainty, and value conflicts where Dror’s comprehensive approach proves most valuable. In an era where governments are routinely confronted with problems that resist simple data-driven solutions, the model’s insistence on integrating rigorous analysis with human judgment remains remarkably relevant.

What do you think? If governments were to formally adopt Dror’s two-level approach – requiring both rational analysis and explicit value bargaining at every stage – would that produce better policies, or simply more complicated bureaucracy? And in your view, should citizen participation be a formal part of the extra-rational phase, or does effective governance ultimately require insulating decision-makers from popular pressure?

References

1. https://www.routledge.com/Public-Policy-Making-Reexamined/Dror/p/book/9780878559282
2. https://academic.oup.com/policyandsociety/article/30/1/29/6422231
3. https://banotes.org/administrative-thinkers/yehezkel-dror-normative-optimal-policy-making/
4. https://www.rand.org/pubs/papers/P4030.html
5. https://www.taylorfrancis.com/chapters/mono/10.4324/9781315127774-14/phases-optimal-model-yehezkel-dror
6. https://discuss.forumias.com/uploads/FileUpload/69/47a35d56ad0c6595a22e7b1ba7fe21.pdf
7. https://www.ebookbou.edu.bd/Books/Text/SOB/CEMBA-CEMPA/scom_3612/Module-6.pdf
8. https://iasexamportal.com/courses/ias-mains-pub-ad-model-papers-test-45
9. https://pubadmin.institute/public-policy-and-analysis/normative-optimum-model-rational-decision-making
10. https://www.studocu.com/row/document/bishop-stuart-university/system-administration-ii/models-of-public-policy-making/90432680