The Art of Solving the Same Problem Forever
Why managing problems has become easier than eliminating them
Many years ago, during my career in IT, I became certified in ITIL, the framework for managing IT services. One concept particularly appealed to me: continual improvement. The idea was hardly revolutionary, but its implications were profound. A system was never really finished. Once operational, it would inevitably encounter problems, changing requirements, new technologies, and unexpected circumstances. Keeping it running was important, but ensuring that it continued to improve was equally essential.
ITIL made continual service improvement a major discipline in its own right, alongside practices such as incident and problem management. These address a distinction that may sound trivial but is fundamental to running any complex system.
An incident interrupts a service and needs to be resolved. Problem management looks beyond the immediate interruption, investigating underlying causes and finding ways to reduce the likelihood or impact of future failures. Not every problem can be permanently eliminated, nor would doing so always justify the cost. But repeatedly restoring a failed service without considering why it keeps failing is hardly an example of good management.
That distinction stayed with me throughout my career. It reflects something almost embarrassingly obvious: repeatedly solving the same problem does not imply that you are getting better at your job.
Yet looking at many of the institutions responsible for running our societies, I sometimes wonder whether we have become extraordinarily proficient at precisely that.
A Remarkable Capacity for Handling Problems
Modern institutions are impressive organizational achievements. They employ highly educated professionals, manage enormous budgets, and process millions of transactions within elaborate legal and administrative frameworks. Their activities are monitored, audited, evaluated, and documented. Problems are registered, complaints processed, reports produced, and corrective actions proposed.
When something repeatedly goes wrong, additional procedures may be introduced, responsibilities clarified, or more resources allocated. Sometimes an entirely new department emerges to manage the consequences.
Much of this activity is necessary. But there is a fundamental difference between developing a better capacity to handle recurring difficulties and reducing the need to handle them in the first place.
Imagine a company whose customers repeatedly call because invoices are incorrect. It could invest in a larger call center, improve response times, and proudly report that complaints are now handled within twenty-four hours. Its service department might even receive an award for improved performance.
But the more interesting question would be why so many invoices are still wrong.
Now consider how often citizens encounter administrative procedures requiring repeated submissions of information already available elsewhere, corrections of errors originating within an administration, or interactions between institutions unable to coordinate their responsibilities.
Each organization may handle its part correctly. Each may have excellent procedures for correcting errors, processing appeals, or responding to requests. Yet the overall experience remains unnecessarily complicated because nobody takes responsibility for eliminating the underlying cause.
The recurring problem becomes part of the normal workload, and the institution becomes increasingly sophisticated at handling it.
Who Actually Owns the Problem?
In IT, the people restoring a disrupted service may have neither the authority nor the expertise to correct the underlying problem. That is why problem management requires responsibilities extending beyond the immediate incident.
This raises a fascinating question about public institutions: who owns a problem that crosses organizational boundaries?
Consider a citizen dealing with an inheritance, moving home, or navigating a complicated medical situation. Several administrations, different levels of government, financial institutions, healthcare providers, and legal requirements may be involved. Each participant fulfills its responsibilities, yet the citizen is left coordinating the entire process.
From the perspective of each institution, the work may be properly organized. From the perspective of the person experiencing it, the system is fragmented, repetitive, and unnecessarily demanding.
Where does responsibility for improving that complete experience reside?
This is not simply a question of digitization. Putting existing procedures online may make them more convenient, but it does not necessarily eliminate unnecessary steps. Sometimes it merely transfers work previously performed by employees to citizens.
A genuinely ambitious improvement effort would ask different questions. Why is this information required? Is it already available? Does the procedure remain necessary? Could different organizations coordinate their activities so that citizens no longer need to act as intermediaries between them?
The objective would not merely be to make every transaction faster. It would be to eliminate as many unnecessary transactions as possible.
And this is not some futuristic dream.
Estonia introduced a proactive family-benefits service in 2019. Once a child’s birth is registered, government systems use existing information to assess eligibility and prepare a benefits offer for the parents. Rather than requiring families to discover their entitlements and submit separate applications, the administration initiates the process. Parents can review and confirm the offer.
An evaluation reported by the OECD found that the service reduced direct monthly interactions between officials and beneficiaries by 88%, while automatically analyzing virtually all registered newborns’ files for eligibility.
The significance is not that Estonia has perfected public administration. It is that the government reconsidered the need for an entire category of recurring interactions instead of merely finding ways to process them more efficiently.
The improvement was not a better application form. It was largely removing the need for one.
Reform Is Not Improvement
Governments are certainly not reluctant to change things. Education is continuously being reformed, administrative structures reorganized, regulations revised, and new programs introduced. Entire political careers are built around promises of transformation.
But continuous change and continuous improvement are not the same thing.
A reform is an attempt to make something better. Whether it succeeds is a separate question, requiring observation, evaluation, and sometimes the uncomfortable admission that the original approach was mistaken.
This distinction is familiar in engineering and IT. A technical project may be completed on time, within budget, and according to specification, yet ultimately disappoint its users. Successful implementation does not automatically imply that the expected benefits were achieved.
The OECD’s Government at a Glance 2025 provides an interesting indication of the challenge. Among 30 surveyed countries, 16 reported using service-performance and user-experience data to inform improvements, while only nine reported using such data to measure the impact of reforms.
Governments are not devoid of improvement initiatives. Lean management, service design, quality management, and increasingly sophisticated digital services are well established in parts of the public sector. The OECD also reports that most member countries have introduced government-wide service standards. The weakness appears less in the recognition that improvement matters than in making it a consistent operating discipline, particularly across organizational boundaries. The result is a curious contradiction: institutions may be better organized to implement decisions than to discover whether those decisions produced the intended benefits. And when the results disappoint, another reform may follow.
Measuring the Wrong Success
One of the most important lessons from operational management is that performance indicators can be misleading.
A department may shorten its processing time by transferring work elsewhere. A customer service center may improve its statistics by closing complaints more quickly, without dealing with the problems that generated them. An administration may reduce its own costs while requiring citizens to spend considerably more time completing its procedures. The indicators improve, but the system does not.
This becomes particularly troublesome when organizations are evaluated within their own boundaries rather than according to the complete service they collectively provide.
Imagine several administrations involved in a single procedure. Each reports that it meets its processing targets, yet the citizen must submit information repeatedly, wait for sequential approvals, and resolve inconsistencies between departments.
From an institutional perspective, everything may appear to be functioning according to plan. From the citizen’s perspective, the plan itself is the problem. A process that saves an administration ten minutes but costs thousands of citizens an additional hour may improve internal efficiency while making society less efficient overall.
This suggests some rather different measures of success. Instead of concentrating exclusively on transactions completed, requests processed, and complaints resolved, why not measure the frequency of avoidable contacts, repeated errors, unnecessary information requests, and the total effort required from citizens? These measurements would not capture every aspect of public value. Fairness, accessibility, legal certainty, and the protection of vulnerable people remain essential. But they would expose costs and inconveniences that conventional institutional performance reports often overlook. A well-designed system should become better at providing necessary services while progressively reducing unnecessary work.
When Improvement Threatens the Organization
There is an uncomfortable organizational dimension to all this.
Recurring problems generate work. Work requires people, budgets, procedures, responsibilities, and management structures. Over time, institutions develop considerable expertise in managing the difficulties they encounter. But eliminating the causes of those difficulties may challenge the structures built around them.
This does not mean that employees deliberately preserve problems to protect their jobs. Many professionals are deeply frustrated by the unnecessary work they are required to perform. Nor are private companies immune to the same phenomenon. Large corporations can become remarkably successful at preserving activities whose original purpose has disappeared.
But organizational incentives matter. A manager facing increasing workloads can usually justify requesting additional resources. A manager who permanently eliminates a substantial part of that workload may instead face questions about the department’s budget or future responsibilities.
In public institutions, responsibilities are often established by law, budgets are negotiated politically, and organizational structures depend on agreements extending far beyond operational efficiency. Resources freed through improvement cannot necessarily be redirected toward new activities as easily as they might be in a competitive business. This makes improvement more complicated, but certainly not less important.
The real challenge is to create conditions in which eliminating unnecessary work becomes a recognized achievement, even when it reduces an organization’s own workload or influence.
Improvement Is Not Another Department
It might be tempting to address these difficulties by creating dedicated continuous improvement departments throughout government. Some public administrations already have comparable teams, and their work can produce valuable results.
But there is something faintly ironic about responding to excessive organizational complexity by adding another organizational layer to manage it.
Continuous improvement is not primarily a department, a reporting obligation, or an annual improvement plan. It is a way of operating. The people encountering recurring problems must have meaningful opportunities to identify them. Someone must have responsibility for investigating their causes, including those extending beyond departmental boundaries. Those with the authority to change procedures must be willing to reconsider them, and the consequences of those changes must be evaluated. Crucially, improvement must be able to make existing activities unnecessary.
This becomes particularly relevant as artificial intelligence and increasingly capable digital systems offer new opportunities. Automating an inefficient procedure may make it faster and cheaper, but eliminating the need for that procedure can be far more valuable.
The most impressive digital transformation may be one citizens barely notice, because something that previously demanded their time and attention now happens automatically, correctly, or no longer needs to happen at all. Yet even here, technology is merely an instrument. Genuine improvement requires the organizational willingness to reconsider the work itself.
The Ambition to Make Problems Disappear
Not every societal problem can be permanently eliminated. Human needs evolve, unexpected circumstances arise, and public responsibilities frequently involve conflicting objectives that cannot be resolved through technical optimization alone.
Some recurring activities are necessary precisely because society values fairness, accountability, and the protection of individual rights. Eliminating a safeguard simply because it consumes resources would hardly qualify as improvement.
But none of this changes the underlying principle. Organizations should never become so comfortable with their existing activities that they stop questioning why those activities are necessary. Recurring difficulties should stimulate curiosity. Unnecessary complexity should provoke investigation. And improvements should be evaluated by their actual consequences, not merely by the successful completion of another project.
I still find it remarkable that we developed sophisticated methods for managing the lifecycle of technological systems decades ago, while the institutions responsible for much of our collective life often remain more visibly organized around responsibilities, procedures, and periodic reforms than around continuous learning.
Perhaps we should begin asking public institutions some rather different questions. Not merely how many requests they processed, how quickly they answered complaints, or how successfully they implemented the latest reform. But how many recurring problems they prevented, how much unnecessary work they eliminated, and how much easier they made life for the people they serve.
Because the ultimate achievement of a service organization is not necessarily to become indispensable. Sometimes it is to make part of its own work unnecessary.
And perhaps that is the simplest test of continual improvement: not how efficiently we keep solving the same problems, but how determined we are to ensure that we don’t have to solve them again.