Fixes that fail is a systems-archetypes describing solutions that address problem symptoms effectively in the short run but create unintended consequences that, after a delay, recreate or worsen the original problem. The structure is simple: a problem symptom triggers an intervention that reduces the symptom. But the intervention has side effects that, with delay, circle back to worsen the original problem condition. Because of the delay, practitioners do not connect their fix to the eventual worsening — they experience the deterioration as evidence that the problem is getting worse, which motivates more of the same fix, which creates more side effects, in an escalating cycle.
The delay between the fix and its unintended consequences is what makes the pattern so difficult to recognize. If the side effects appeared immediately, the feedback would be clear: this fix makes things worse. But in most organizational systems, the delays are long enough that the fix and the consequence are separated by months or years, which breaks the intuitive causal connection. Practitioners experiencing the pattern typically explain the worsening as evidence of a harder problem or external forces, not as the result of their own intervention. This is one of the primary reasons peter-senge emphasizes the importance of causal-loop-diagrams that explicitly include delays — without representing delays in the feedback structure, the pattern is genuinely invisible.
Organizational examples span management practice. Price cuts to address declining market share may attract customers initially, but trigger competitor price responses and erode brand positioning, which worsens the underlying competitive situation. Layoffs to address cost pressure reduce costs in the short run but damage organizational knowledge, morale, and customer service capacity in ways that reduce competitiveness over time. Emergency hiring to address workload surges brings in underprepared people who require training that overloads the most experienced staff, reducing quality and productivity — and potentially increasing workload.
The archetype differs from shifting-the-burden in a subtle but important way. In shifting the burden, there is an alternative fundamental solution available that is not being pursued because the symptomatic solution is good enough. In fixes that fail, there may be no alternative; the issue is purely the unintended consequences of the solution. The leverage-points are similarly different: for shifting the burden, the leverage is in building fundamental solution capacity; for fixes that fail, the leverage is in identifying and mitigating the side effects loop before it has propagated too far. This often requires mapping the system explicitly — drawing out the causal-loop-diagrams that make the side effect pathway visible — and building monitoring systems that watch for early signals of the unintended consequences, rather than waiting for them to become undeniable.