Lag Time
Lag Time: Delays between actions and their effects cause decision-makers to over-correct, creating oscillations. The shower temperature problem: you turn up the heat, feel nothing, turn up more, get scalded. Lags make systems hard to control. Shortening lags or accounting for them explicitly is often the highest-leverage intervention.
What Is Lag Time?β
Lag time (or delay) is the elapsed time between a cause and its observable effect. Lags are ubiquitous in complex systems and are one of the primary sources of counterintuitive, oscillatory behavior.
The shower problem: you feel cold, turn up the hot water, wait for the water to warm up through the pipe, feel nothing, turn up more, wait again, suddenly get scalded, turn the cold back up, overshoot to cold. This oscillation around the target temperature is entirely caused by the delay between your adjustment and the temperature change you can feel. Without the delay, you would reach target temperature smoothly.
Why lags cause oscillation:
- A decision-maker observes a gap between current state and target
- They take corrective action proportional to the gap
- Because of the lag, the corrective action's effects are not yet visible
- The gap still appears to exist (or even worsen, as the lag-delayed first action makes things appear worse during the lag period)
- The decision-maker takes additional corrective action
- Both actions then manifest simultaneously β overshooting the target
- The decision-maker corrects in the other direction β and the oscillation continues
The key insight: In a system with lags, the right corrective action is less than what the current gap suggests β because past actions are already in the pipeline and will reduce the gap once the lag expires. Failing to account for this produces systematic overshoot.
Three Real-World Examplesβ
The Bullwhip Effect in Supply Chainsβ
Supply chains are dominated by lags: orders placed today arrive in weeks; production started today completes in months; hiring started today produces trained workers in months. When a retailer sees increased demand and orders more from the distributor, the distributor orders more from the manufacturer, who ramps production. But by the time the additional inventory arrives, the demand spike may have subsided.
The lag between order and delivery, combined with decision-makers acting on apparent current gaps without accounting for in-transit orders, produces the bullwhip effect: small demand variations create wild inventory oscillations upstream. The solution: visibility into in-transit inventory (shortening the information lag) and smaller, more frequent orders (reducing the magnitude of each corrective action).
Monetary Policy Lagβ
The Federal Reserve raises interest rates to cool inflation. The impact on the economy is lagged: it takes 12β18 months for rate changes to fully propagate through lending costs, investment decisions, hiring, and consumer spending. During this lag, inflation may continue to rise β leading the Fed to raise rates further. When both rounds of increases finally take effect simultaneously, the economy may slow more than intended.
This lag-driven overshoot is a recurring pattern in monetary policy and is a central challenge for central banks. Knowing the lag exists doesn't eliminate it; it requires policy-makers to act on anticipated future states rather than current readings.
Ecological Management of Fish Stocksβ
A fishery appears to be declining. Managers reduce quotas. Fish populations don't recover for years (the lag includes reproduction cycles, juvenile maturation, and population dynamics). During the lag, the population still looks depleted. Managers reduce quotas further. Eventually, multiple quota reductions take effect simultaneously, and the population booms beyond the ecosystem's carrying capacity β triggering a crash.
Lag-aware management: model the population dynamics including reproduction and maturation lags, and set current quotas based on expected future population (accounting for lags), not observed current population.
When to Use Itβ
β Apply Lag Time thinking when:
- Managing any system with a significant delay between action and effect
- Designing feedback systems and decision protocols
- Diagnosing oscillation or overshoot in a system
- Setting expectations about when interventions will show results
| Pairs well with | Why |
|---|---|
| Feedback Loops | Lags are the complication that turns smooth feedback into oscillation |
| Stocks and Flows | Stocks create natural lags β they accumulate and deplete slowly |
| Second Order Effects | Lag-delayed effects are often the "unexpected" second-order effect |
Three Real-World Examplesβ
Federal Reserve Interest Rate Policyβ
The US Federal Reserve estimates that monetary policy changes take 12β18 months to fully transmit into the real economy. When the Fed raises rates in response to rising inflation, the effect on business investment, consumer borrowing, and ultimately prices is delayed by over a year. This lag creates a fundamental policy challenge: by the time the full effect of a rate hike is felt, conditions may have reversed. The Fed is perpetually steering by looking in the rear-view mirror.
Reforestation Programsβ
Trees planted today won't provide meaningful carbon sequestration, habitat value, or timber for 20β80 years depending on species. Conservation programmes that plant millions of trees create a lag of decades between the action and its full benefit. This means investment decisions must be made with very long time horizons, and decision-makers will likely never see the full results of their choices. Deforestation has a much shorter lag to negative consequences β which creates asymmetric political incentives.
Antibiotic Resistance and Prescription Behaviourβ
The overuse of antibiotics creates antibiotic-resistant bacteria β but the lag between prescribing behaviour and resistance emergence can span years or decades. Individual doctors prescribing unnecessary antibiotics today face no immediate negative consequence; the harm accumulates slowly across populations. This lag between cause and effect makes the feedback nearly invisible at the individual level, which is why systemic interventions (prescription guidelines, agricultural antibiotic restrictions) are needed rather than relying on individual behaviour change.
When to Use Itβ
β Apply lag-time thinking when:
- Evaluating why a policy or intervention appears not to be working (the results may simply be delayed)
- Designing feedback systems β ensure measurement frequency matches the system's lag
- Making investment decisions with long payback periods
β Be cautious:
- Not all delayed results are lag effects β some interventions genuinely don't work
- Lag varies by system and context; don't assume a fixed delay
- Impatience with lag can lead to policy reversal before the intended effect materialises
| Pairs well with | Why |
|---|---|
| Feedback Loops | Lag in feedback loops causes oscillation and overshoot |
| Stocks and Flows | Flows change stocks; the change in stock is delayed relative to the flow change |
| Second-Order Thinking | Lag makes second-order effects harder to attribute causally |
Common Misuses and Limitationsβ
Assuming longer lag always means more uncertainty. Sometimes long lags are highly predictable (tree growth, bond maturity, infrastructure depreciation). Uncertainty and lag are separate dimensions. The question is whether the lag is variable (hard to predict) or fixed (predictable).
Using lag as an excuse for inaction. "We won't see results for 10 years" can be used to justify indefinitely deferring decisions. The correct response to long lags is to act earlier, not to wait, since the payoff timing is driven by when action was taken.
Ignoring lag asymmetries. In many systems, the lag for positive effects is longer than for negative effects. Building a reputation takes years; destroying it can happen in days. Investment in culture has long lag; culture damage from a scandal has short lag. These asymmetries have important strategic implications.
Related Modelsβ
| Model | Relationship |
|---|---|
| Feedback Loops | Lag in feedback causes oscillation; too much lag causes instability |
| Second-Order Effects | Lagged effects are often the second-order consequences of actions |
| Unintended Consequences | Lag makes unintended consequences harder to detect and attribute |
Frequently Asked Questionsβ
How do you manage decisions when lags are long and uncertain?
Three strategies: (1) invest in leading indicators β find metrics that move before the lagged outcome so you get earlier signal; (2) run small experiments β shorter-horizon tests that let you validate direction before committing fully; (3) use scenario planning β if lag is variable, plan for multiple outcome timings rather than betting on a single schedule.
Why does lag cause oscillation in systems?
When feedback is delayed, corrective actions overshoot because the system is responding to past conditions rather than current ones. By the time the correction arrives, the system has moved on. The classic example: adjusting a shower to the right temperature with a long lag β you turn it hotter, nothing happens, you turn it hotter still, then suddenly it's scalding. Overshoot triggers overcorrection; oscillation results. Reducing lag or reducing correction aggressiveness both dampen oscillations.
What is the "policy resistance" phenomenon related to lag?
Policy resistance is when the system's homeostatic mechanisms undo policy interventions. Lag contributes because decision-makers see no immediate effect, conclude the policy is failing, and abandon or reverse it β before the lag has elapsed. The policy is actually working; impatience due to lag causes premature reversal. This is especially common in public health, education, and macroeconomic policy, where lags are measured in years.
Further Readingβ
- Meadows, D.H. (2008). Thinking in Systems β the definitive treatment of lag in dynamic systems
- Forrester, J.W. (1961). Industrial Dynamics β the original system dynamics work on oscillation caused by lag
- Sterman, J.D. (2000). Business Dynamics β lag, delay, and feedback in business systems
Apply with AIβ
π Identify and account for lag times in your system with MindMax β
This page is part of the MindMax Mental Models Knowledge Base.