Demand Displacement

What it is

Sales that are displaced are moved, not made. A shopper who stays at home during a storm may make the trip the following week, and the month's total is the same. The shopper is an illustration, but the mechanism is the one behind the economic model in Starr-McCluer's study.

Martha Starr-McCluer's 2000 Federal Reserve paper proposes that weather changes how productive people's time is for activities outside work, such as shopping or recreation. Through time and budget constraints, that can shift spending among goods and across periods. Using monthly retail sales and National Weather Service data, she found that unusual weather had a modest but significant effect on monthly sales. Lagged effects often offset the initial effect, so the influence largely disappeared at a quarterly frequency.

On this site the term is used for that specific question: whether a weather effect is new demand or moved demand.

In effect

Brigitte Roth Tran's 2022 letter tests displacement on daily data. The sample was daily sales at more than 100 stores of one national apparel and sporting goods brand between April 2010 and December 2013, matched to weather station records. The pattern was asymmetric. Good-weather gains were offset by lower sales in the following weeks and were no longer statistically significant by four weeks. Bad-weather losses persisted: a moderately bad day cost a net quarter of a day's sales, and a very bad day an average net loss of about 40 per cent of a day's sales, excluding days when stores closed.

Two channels might have absorbed the loss and did so only slightly. Online sales did not offset the store losses from bad weather, and shoppers moving between indoor and outdoor shopping centres offset only a small share of the store response. Online sales of the brand rose on good-weather days.

For a planner the asymmetry carries a rule of thumb, which the article What a Weather Forecast Does for Sales states with its limit: a forecast of good weather should raise this week's order and trim the following weeks', and a forecast of very bad weather should cut the order without any later top-up. The data come from one brand's stores, so the rule is a pattern to test.

What it does not say

It does not say weather creates no sales. Over a day or a week the effect is real and sometimes large. Displacement is a claim about how long it lasts.

It does not say the offset is complete. Starr-McCluer found that lagged effects often offset the first effect, and Roth Tran found bad-weather losses persisted. The two results describe different data and different frequencies.

It does not generalise from one brand. Roth Tran's letter covers the stores of one apparel and sporting goods brand and says nothing formal about forecasting or seasonal adjustment. It also notes that the online estimates apply to local sales, not aggregate sales.

And it was read here partly at the level of an abstract. Starr-McCluer's abstract gives no figure for the size of the effect, which is why none is printed.

How to use it

For a planner the concept reduces to one question asked of every weather lift: was this demand created or only moved? The answer changes what to order. If a warm weekend moved a week's sales forward from the next few weeks, the sensible response is to raise this week's stock and trim the following ones. If a storm destroyed sales that nobody recovers, a later top-up would overstock. The evidence in What a Weather Forecast Does for Sales says the two cases are not symmetrical, which is why a single smoothing rule applied in both directions would be wrong.

The question can be tested in a retailer's own data without any new method. Compare the weeks after a good-weather spike with the weeks after an ordinary one, store by store and range by range, and see whether they run lower. Starr-McCluer's monthly national series and Roth Tran's daily store sales are different objects, so neither supplies the answer for a particular shop. They show where to look and how long the effect might last, which is a few weeks on the good side.

Where the idea is weakest

The term hides a measurement problem. A displaced sale is not observed, because the shopper who waits is recorded only when the sale finally happens, and attributing that sale back to the weather needs a model. Both studies use one, and both were read here at the level of their published summaries, so the attribution cannot be checked from this page. The concept also says little about price. A retailer that discounts after a bad week may pull sales forward by its own action, and the weather studies do not separate the two.

---