What are "root causes?"
The term "root causes" is often used to describe conditions that help create or maintain a problem.
For relatively straightforward problems, it may be possible to identify a small number of underlying causes.
Many community problems, however, are more complex. They may develop through several interacting causes involving people, organizations, policies, resources, infrastructure, institutions, relationships, and broader social and economic conditions.
For that reason, it is often more useful to think in terms of underlying and contributing conditions rather than assuming that every problem has one root cause.
Understanding these conditions is an important part of the problem-solving process.
Cause analysis can help a group:
- Look beyond the most visible symptom
- Identify several places where change might be possible
- Examine individual, organizational, community, policy, and other system conditions
- Avoid assuming a preferred solution before understanding the problem
- Identify what additional evidence is needed
Cause analysis does not guarantee that the group will identify the correct response.
The explanations developed during the process should be treated as hypotheses to examine with evidence, community knowledge, implementation experience, and evaluation.
What is the "But why?" technique?
The "But why?" technique is a simple method for exploring conditions that may contribute to a community problem.
It is related to an approach often called the "Five Whys," although there is no required number of times that the question must be asked.
The technique begins with a clearly described condition.
Participants ask why that condition may be occurring.
For each proposed explanation, they ask another question such as:
- But why is that happening?
- What conditions make that possible?
- What evidence supports that explanation?
- What else could explain it?
- Could it have been prevented or reduced?
- What conditions would have to change?
The goal is not necessarily to arrive at one final answer.
One outcome may have several causes, and one contributing condition may lead to several additional questions.
The result may therefore look more like a branching map than a single chain.
A simple example
Suppose a community health program notices that evening appointments have a much higher no-show rate than daytime appointments.
The group might ask:
Condition: Evening appointments have a high no-show rate.
Q: Why might that be happening?
A: Some participants report difficulty getting to the site in the evening.Q: Why is transportation more difficult in the evening?
A: Public transportation serving the site operates less frequently after 6:00 p.m.Q: What evidence supports that?
A: Transit schedules and participant interviews show fewer evening routes and longer travel times.
But transportation may be only one pathway.
The group should also ask:
Q: What else could contribute to the no-show rate?
A: Some appointments overlap with work, caregiving responsibilities, or other obligations.Q: Are there other possibilities?
A: Reminder messages may not reach everyone, and some participants may not know how to reschedule.
The analysis now identifies several possible intervention points rather than one supposed root cause.
Why should you identify underlying and contributing causes?
Taking action without understanding what contributes to a problem can lead to an intervention that addresses only a visible symptom or an incorrect assumption.
Cause analysis can help a group direct its resources toward conditions that meaningfully influence the outcome.
It can also reveal that several responses may be needed at the same time.
To distinguish immediate conditions from underlying ones
A community may need an immediate response while also addressing conditions that repeatedly produce the same problem.
For example, emergency food assistance can be essential for people who need food today.
Longer-term analysis might also examine:
- Food affordability
- Household income
- Transportation
- Availability of food retailers
- Eligibility and access to benefits
- Housing and other household costs
- Employment conditions
Immediate relief and efforts to change underlying conditions do not have to compete with one another.
To identify several possible intervention points
A problem may be influenced at several levels.
Depending on the issue, contributing conditions may involve:
- Individual knowledge, skills, behavior, or health
- Family, peer, or other interpersonal relationships
- Organizational practices
- Services and institutions
- Community resources and infrastructure
- Economic conditions
- Policies and laws
- Access to power and resources
- Physical environments
- Social norms and other contextual factors
Do not assume in advance that an individual behavior is the most important cause simply because it is easy to observe.
To avoid blaming people for conditions outside their control
Individual choices matter in many situations, but choices are made within social, economic, organizational, and physical environments.
For example, before concluding that people need more nutrition education, ask whether nutritious food is affordable, accessible, available, and practical within people's actual circumstances.
Before concluding that people need to make better use of a service, ask whether the service is available at appropriate times, physically and financially accessible, trustworthy, understandable, and responsive to community needs.
To identify structural and institutional conditions
Some problems are strongly influenced by:
- Policies
- Funding arrangements
- Institutional procedures
- Housing
- Transportation
- Employment conditions
- Education
- Discrimination
- Access to services
- Distribution of power and resources
These conditions may be less visible than individual behavior but can strongly shape what options people actually have.
To identify protective factors
Cause analysis should not focus only on what is going wrong.
Ask:
- Where is the desired outcome already occurring?
- What is different there?
- What strengths or resources are protecting against the problem?
- What existing practices should be preserved or expanded?
Understanding why a problem is less severe in some places or circumstances can reveal intervention possibilities that a deficit-only analysis would miss.
To improve community participation in analysis
People affected by a condition often have knowledge about barriers, resources, history, and everyday experience that organizations may not see in administrative data.
Meaningful participation can improve the analysis and help determine whether proposed explanations fit people's actual experiences.
Participation should not be framed as requiring community members to "own" a problem that may be influenced by institutions, policies, markets, or other actors outside their control.
Instead, clarify who has knowledge, who has decision authority, who controls relevant resources, and who can contribute to change.
When should you use the "But why?" technique?
The technique can be useful when:
- A group is describing symptoms but has not explored what contributes to them
- A proposed solution appears to assume a cause that has not been examined
- Participants have different explanations for why a condition exists
- The group wants to identify several possible intervention points
- An operational or organizational problem appears to involve a chain of contributing conditions
The technique is quick and inexpensive, but it should not be treated as sufficient for every problem.
Additional methods may include:
- Surveys
- Interviews
- Community listening sessions
- Administrative data
- Research evidence
- Observation
- Process mapping
- Causal diagrams
- System mapping
- Program evaluation
Use additional methods when the stakes are high, the problem is complex, participants disagree substantially about causes, or the group lacks evidence to support the proposed causal chain.
When immediate action is also necessary
Cause analysis should not delay urgent action needed to protect health, safety, rights, or basic needs.
A group can address an immediate harm while continuing to investigate why the condition occurred and how recurrence might be prevented.
How does the "But why?" technique work?
Technique Guide
1. Start with a clearly defined condition
Begin with a specific description of what is happening.
Avoid defining a population as the problem or embedding a preferred solution in the statement.
For example:
Participation in the program has declined by 25% during the past year.
is more useful than:
People do not care about the program.
The first statement describes a condition that can be investigated. The second assumes a cause.
2. Involve people with relevant knowledge
Depending on the issue, include people who:
- Experience the condition
- Have relevant community or lived knowledge
- Work directly with the program or system
- Have relevant professional or technical expertise
- Control resources or authority related to the problem
- May help implement change
Broad participation can improve the analysis, but it does not guarantee that every cause will be identified.
Consider power differences as well. Some participants may be less willing to challenge an explanation publicly when supervisors, funders, government officials, or other powerful participants are present.
Written, confidential, anonymous, small-group, or separate input may sometimes improve the quality of information.
3. Ask the first "Why?" question
Ask:
What conditions may be contributing to this?
Record several plausible answers rather than accepting the first explanation as the causal pathway.
4. Ask for evidence
For each proposed explanation, ask:
- What evidence supports this?
- Is this something we know, something participants have experienced, or a hypothesis we still need to test?
- Do other credible sources support it?
- Could another factor explain the same pattern?
This helps prevent a plausible story from becoming accepted as fact simply because it sounds convincing.
5. Continue along several branches
Each answer may lead to another question.
For example:
Condition: Evening appointment attendance is low.
Q: Why might attendance be low?
A: Some participants have transportation difficulties.Q: Why are transportation difficulties especially important in the evening?
A: Transit service is less frequent after 6:00 p.m.
At the same time, another branch might be:
Q: What else may contribute?
A: Evening appointments overlap with work and caregiving responsibilities.Q: Why are appointment options concentrated at those times?
A: Staffing and facility schedules limit the hours currently offered.
Another branch might examine reminders, rescheduling procedures, accessibility, or other conditions.
The purpose is to see the network of plausible contributing factors rather than forcing every observation into one chain.
6. Ask what could have prevented or reduced the condition
Prevention questions can reveal intervention points.
Ask:
- What could have prevented this?
- What could have reduced its likelihood or severity?
- What existing protective factors already reduce it in some places or circumstances?
7. Look beyond individual error or behavior
If the analysis reaches an answer such as:
Someone made a mistake.
continue asking:
- What conditions made the mistake more likely?
- Were instructions clear?
- Was training sufficient?
- Was workload reasonable?
- Did technology or equipment contribute?
- Did the process make errors easy to make or difficult to detect?
Individual accountability may still matter, but stopping at "human error" may hide system conditions that can be changed.
8. Examine several levels of the system
Depending on the issue, ask about:
- Individual conditions
- Interpersonal relationships
- Organizational processes
- Community resources
- Physical environment and infrastructure
- Economic conditions
- Policies and laws
- Distribution of power and resources
Not every problem requires analysis at every level. Use the levels that are relevant to the condition.
9. Look for relationships and feedback
Causes do not always operate in one direction.
A condition may create a response that later strengthens or weakens the original condition.
Ask:
- Does this factor affect another factor?
- Does that second factor eventually affect the first?
- Could an intervention create an unintended effect elsewhere?
- Does the importance of the factor change over time?
A systems map or causal diagram may be more useful than a simple chain when these relationships become complicated.
10. Identify what the group can influence
Once several contributing conditions have been identified, distinguish among factors the group can:
- Change directly
- Influence through partnership or advocacy
- Monitor but not directly control
This helps turn causal analysis into realistic intervention planning without pretending every underlying condition is within the group's authority.
11. Document uncertainty
At the end of the exercise, identify:
- Contributing conditions strongly supported by evidence
- Plausible conditions that need more investigation
- Explanations for which evidence is weak or conflicting
- Important information that is still missing
Do not force agreement on a causal explanation simply to complete the exercise.
12. Use additional analysis when necessary
The "But why?" technique is a starting point, not proof.
Follow it with other methods when necessary, including:
- Research
- Surveys
- Interviews
- Administrative data
- Observation
- Process mapping
- Systems mapping
- Evaluation
13. Compare possible intervention points
Cause analysis may reveal several possible places to act.
Compare them using criteria such as:
- Evidence
- Expected effectiveness
- Equity
- Safety
- Accessibility
- Feasibility
- Cost
- Organizational authority and capacity
- Sustainability
- Potential unintended consequences
Several interventions at different levels may be more appropriate than one solution directed at one supposed root cause.
14. Test the analysis during implementation
A causal analysis is a working explanation of how a problem operates.
Once action begins, monitor whether the expected changes actually occur.
Ask:
- Did changing the identified factor affect the outcome?
- Did another barrier become more important?
- Were there unintended effects?
- Did the problem affect different groups differently?
- Does the causal explanation need to be revised?
In Summary
The "But why?" technique is a useful way to move beyond an immediate symptom and begin exploring why a community problem may be occurring.
Its strength is its simplicity. Repeated questioning can reveal conditions that were not obvious when the problem was first described.
Its simplicity is also its limitation.
Complex community problems often result from several interacting causes rather than one root cause. A single chain of "why" questions can therefore produce an explanation that is too narrow or overly linear.
Use the technique to generate possible causal pathways, not to prove them. Ask what evidence supports each explanation, consider competing explanations, examine individual and system conditions, identify protective factors, and look for feedback and unintended consequences.
Include people affected by the condition along with others who hold relevant knowledge, responsibility, resources, or decision authority. Pay attention to power and create additional ways to contribute when people may not be comfortable challenging an explanation publicly.
Finally, connect cause analysis to action and learning. Compare possible intervention points, select responses using explicit criteria, monitor what happens during implementation, and revise the analysis when new evidence shows that the problem works differently than expected.