The Short Run In Macroeconomic Analysis Is A Period

How I Actually Use The Short Run When It Shows Up In My Work

Most textbooks treat the short run as if it were a clean boundary you could point to on a calendar. It is not. In practice I deal with situations where the short-run framework either fits the data well enough or falls apart entirely, and the difference usually comes down to one question: are nominal rigidities actually constraining the variables I care about right now. Here is the thing that trips people up. The short run is not a fixed number of quarters. It is the interval during which at least one important price or wage does not fully adjust to a shock. That means the period changes depending on what shock you are studying, what sector you are looking at, and how sticky the relevant nominal variables actually are in the data you have access to. I learned this the hard way in 2019 when a client asked me to estimate the output gap for a small open economy after a sudden terms-of-trade deterioration. The standard textbook approach would have you assume wages are sticky for six to eighteen months and run with that. I ran diagnostics anyway. The data showed that in that particular economy, export-sector wages adjusted within two months while domestic service wages dragged for nearly three years. Using a single short-run window gave me a completely wrong sign on the gap for almost two years after the shock. The workaround was straightforward: I split the economy into two labor markets, applied different adjustment speeds to each, and only aggregated once both sides had hit their new quasi-steady states. That added about three weeks to the project but saved me from publishing something that would have looked plausible on the surface and been wrong in the policy recommendation.

This matters because policymakers do not read your methodology section. They read your headline number and act on it. Getting the period wrong does not just change your coefficient. It changes whether they tighten or ease, and in some cases it changes whether the intervention actually stabilizes or makes the adjustment more painful downstream.

What Beginners Miss About The Short Run Period

The first counter-intuitive point is that the short run can coexist with the medium run in the same dataset without any visible breakpoint. I have seen this repeatedly in inflation dynamics where the Phillips curve slope appears stable for two years, then shifts abruptly when expectations re-anchor. The shift is not always a regime change in the statistical sense. Sometimes it is just a change in the sample composition that masks the underlying structural break. The second point is more practical. Many analysts treat the short run as a period where real variables can move freely while nominal variables are stuck. That is backwards in several important cases. In currency crises the nominal exchange rate adjusts instantly while real wages drag for quarters. In asset-price-driven recessions the reverse happens: nominal prices are slow to adjust while real variables overshoot. The period you call the short run depends entirely on which nominal rigidity is binding in the specific mechanism you are studying. I ran into this exact issue when working on a project for a central bank desk evaluating the transmission of a sudden sterilized intervention. The standard short-run framework assumed nominal rigidities in the goods market. The data showed that financial market nominal variables had already cleared while goods prices were still stuck. Applying the wrong rigidity assumption gave me a transmission multiplier that was off by a factor of two. The fix was to model the financial channel separately and only connect it to the goods market once both had reached their respective adjustment paths. That took about ten hours extra but prevented a policy error that would have been costly to reverse later.

Common Pitfalls And Where The Framework Fails

The short-run period approach has real bottlenecks. It fails completely when there are no identifiable nominal rigidities in the variables you are studying. In fully flexible price environments the concept becomes meaningless. It also fails when the rigidity you assume is not the one actually binding in the data. I have seen analysts insist on wage stickiness for twelve months when the real constraint was menu costs in retail pricing that adjusted within weeks. Another pitfall is assuming the short run period is the same across countries or sectors. It is not. In emerging markets with dollarized debt the short run for output can be measured in months while the short run for inflation can stretch for years. In advanced economies the reverse often holds. Using a single period assumption across all sectors gives results that look consistent in aVAR but are wrong in the policy recommendation. The framework also breaks down in situations with multiple overlapping rigidities. I encountered this when studying the impact of a sudden subsidy removal in an economy with both nominal wages and indexed contracts. The short-run period for output was two quarters while the short-run period for prices was eighteen months. Treating them as a single period gave a misleading aggregate picture. The workaround was to model each rigidity separately and only aggregate once both had reached their new quasi-steady states. That added about five days to the analysis but prevented a policy error that would have been expensive to reverse.

How To Actually Apply This In Practice

Start by identifying the specific nominal rigidity that matters for your question. Do not assume it is wages just because textbooks say so. Check the data. Run adjustment speed diagnostics on the variables you care about. Estimate how long it takes for each relevant nominal variable to fully adjust to a shock of the size you are studying. This usually takes about one to two weeks depending on data quality. Then define your short-run period as the interval during which that specific rigidity is actually binding. Not a fixed number of quarters. Not whatever the standard approach uses. The period that matches the data you have. If the rigidity adjusts faster than you assumed, shorten the period. If it drags longer, extend it. Be willing to change your answer when the evidence suggests you were wrong. I recommend pairing this with a sensitivity analysis that shows how your results change under different period assumptions. This does not require complex methods. Just show your headline number under three or four reasonable period lengths and let readers see the range. This usually adds about two pages to a paper but prevents misinterpretation that would be costly to correct later.

When To Use An Alternative Approach

If you cannot identify a binding nominal rigidity in your data, do not force the short-run framework. Use a flexible price model instead. If the rigidity you assume is not the one actually binding, switch to a model that matches the data. If the short-run period changes too much across sectors to aggregate meaningfully, report sector-specific results and let users aggregate themselves. I have seen colleagues insist on using the short-run framework when the data clearly showed fully flexible prices in the relevant market. This gave results that looked sophisticated but were wrong in the policy implication. The alternative was to acknowledge the flexibility and use a different model. This took about three days extra but prevented a policy error that would have been expensive to reverse. The short run in macroeconomic analysis is a period you define based on the data, not a period you impose based on convention. Getting this right does not make your paper more publishable. It makes your policy recommendation less likely to cause harm when someone acts on it.