Bardach’s Step One – Define the Problem

… a core concept used in Implementation and Delivery and Atlas107

Concept description

In his A Practical Guide for Policy Analysis – The Eightfold Path to More Effective Problem Solving (reference below), Eugene Bardach provides tips for defining the problem.

Think of Deficit and Excess

“It often – but not always – helps to think in terms of deficit and excess. For instance … “The demand for agricultural water is growing faster than our ability to supply it at an acceptable financial and environmental cost” …

“It often helps to include the word too in the definition – as in “too big,” “too small,” “growing too slowly,” “growing too fast.” These last two phrases (about “growing”) remind us that problems deserving our attention don’t necessarily exist today but are (at least potentially) in prospect for the future, whether near or distant.”

Make the Definition Evaluative

“Remember that the idea of a “problem” usually means that people think there is something wrong with the world, but note that wrong is a very debatable term. Not everyone will agree that the facts you (or others) have defined as a problem really do constitute a problem, for each person may apply a different evaluative framework to these facts. … A common philosophical as well as practical question is this: “What private troubles warrant definition as public problems and thereby legitimately raise claims for amelioration by public resources?” It is usually helpful to view the situation through the “market failure” lens. … In most – though not all – situations in which no actual market failures can be identified, people’s private troubles cannot typically be ameliorated by even the most well-intentioned governmental interventions. Even when some amelioration is possible, there are usually many adverse side effects. In some cases, it may nevertheless be worthwhile to pay the price of these side effects, but such calculations must be done carefully and scrupulously.”

Quantify if Possible

Your problem definition should, insofar as possible, include a quantitative feature. Assertions of deficit or excess should come with magnitudes attached. How big is “too big”? How small is “too small”? How about “too slowly” or “too fast”? … In many or most cases, you will have to estimate – or, more likely, “guesstimate” – the magnitudes in question. Sometimes you should furnish a range as well as a point estimate of magnitudes – for example, “Our best guess of the number of homeless persons in families is 250,000, although the truth could lie between 100,000 and 400,000.”

“Even if you cannot come up with good numbers yourself, qualitatively defining a metric that might be used to quantify the problem helps you make your problem definition more behavioral and concrete. It is better to say, “Too many people with annual incomes over $60,000 are living in subsidized apartments,” than simply, “Too many relatively well-off people are taking advantage of low-rent public housing.” The $60,000 value provides desirable texture and information about a threshold number that will serve in the promised analysis.”

Diagnose Conditions That Cause Problems

“Some problematic conditions are not experienced as troublesome per se by citizens but are perceived by them, or by analysts working on their behalf, to be causes of trouble. It is sometimes useful to diagnose at least one alleged condition of this type and to define it as a problem to be mitigated or removed – as in, “One of the problems in the air pollution area is that states have not been willing to force motorists to keep their engines tuned up and their exhaust systems in proper order.”

Risky Conditions: “The Odds”

“Referring to “the odds” is a useful way to talk about anything that is uncertain in your analysis, not just the problem definition. It can also refer to the uncertain prospects of an alternative’s working out as planned, or the likelihood that a key political actor will remain in office in order to oversee policy implementation. … The odds formulation can also be used for specifying criteria. For instance, one could say that one criterion is “Maximize the odds that the People’s Party will control the upper chamber following the next election” or “minimize the odds that teens in the catchment area of this program will reject it because it is not ‘cool’.””

Identify Latent Opportunities

“A special kind of problem is an opportunity missed … if latent opportunities are really lying around, it would be a pity to ignore them.” Bardach provides a box of “Generic Opportunities for Social Improvement That Often Go Unnoticed”:

Operations research strategies. By means of sequencing, timing, prioritizing, matching, clustering, and other such rationalizing arrangements, it may be possible to use a fixed stock of resources to achieve higher productivity than is possible otherwise.

Cost-based pricing. Discrepancies between prices and real costs present an opportunity for enhancing social welfare by adjusting prices to better reflect reality.

By-products of personal aspirations. It is possible to structure new incentives or create new opportunities for personal advantage or satisfaction that can indirectly result in social benefit.

Complementarity. Two or more activities can potentially be joined so that each may make the other more productive.

Input substitution. The world abounds in opportunities to substitute less costly inputs in a current production process while achieving roughly equivalent results.

Development. A sequence of activities or operations may be arranged to take advantage of a developmental process.

Exchange. Unrealized possibilities for exchange can increase social value.

Multiple functions. A system can be designed so that one feature has the potential to perform two or more functions.

Nontraditional participants. Line-level employees of public agencies – as well as their customers, clients, or the parties whom they regulate – often have knowledge of potential program improvements that could usefully be incorporated into the agencies’ policies and operations.

Underutilized capacity. Governments sometimes systematically underutilize resources at their disposal.

Avoid Common Pitfalls in Problem Definition

“Problem definition is a step beset by at least two dangerous pitfalls.

Defining the solution into the “problem.” Your problem definition should not include an implicit solution introduced by semantic carelessness. Projected solutions must be evaluated empirically and not legitimated merely by definition. Therefore, keep the problem definition stripped down to a mere description, and leave open where you will look for solutions.

Accepting too easily the causal claims implicit in diagnostic problem definitions. … You have to evaluate the causal chain that goes from the situation itself to the bad effects it is alleged to cause, and to convince yourself that the causal relationship is real.”


“Problem definition is a crucial step. Because it is hard to get it right, however, you may take that same step again and again. Also, your empirical and conceptual understanding will evolve over the course of your analytic work.”

Atlas topic, subject, and course

Designing the Delivery Model (core topic) in Implementation and Delivery and Atlas107


Eugene Bardach (2012), A Practical Guide for Policy Analysis – The Eightfold Path to More Effective Problem Solving, Fourth Edition, Sage, Los Angeles.

Page created by: Ian Clark, last modified 11 February 2018.

Image: From page 1 Bardach book, accessed 11 February 2018.