Delay-speedup asymmetry
Also known as: Delay-speedup effect
It costs more to delay a reward than you'd pay to speed the same reward up.
What it means
The delay-speedup asymmetry is the finding that people demand far more to accept postponing an expected outcome than they will pay to bring the same outcome forward by an identical interval. The gap turns on framing: a fixed reference point — the expected timing — makes a delay read as a loss and a speedup as a gain. The leading explanation is that delaying an expected reward is coded as a loss and speeding one up as a gain, so loss aversion inflates the delay premium relative to the speedup price — though a rival query-theory account attributes the gap to the order in which reasons for and against waiting are retrieved, and experimentally reversing that order can shrink or eliminate it. The direction reverses for unpleasant events, where people pay more to postpone than to advance them.
The original demonstration
Loewenstein (1988) produced the effect by importing prospect theory's reference point into the timing of an outcome. People who expected a reward on a given date were asked either what compensation would make postponing it acceptable, or what they would pay to have it arrive earlier. The amount demanded to accept the delay dwarfed the amount offered to buy the speedup, even though the interval was identical. He found the mirror image for aversive outcomes: people paid more to push an unpleasant event back than they demanded to have it brought forward. The common thread is the reference point — the expected timing — against which each change is coded as a gain or a loss. What moves the price is the frame, delay versus advance, not the interval itself.
Two accounts of the mechanism
The standard explanation, formalized by Loewenstein and Prelec (1992) and Scholten and Read (2010), is loss aversion on the reward: not receiving it when expected feels like a loss, and losses loom larger than equivalent gains, so the delay premium exceeds the speedup price. A rival process account, query theory (Weber et al., 2007), locates the effect in the order of memory retrieval. The delay frame prompts you to generate reasons to consume now first and reasons to wait second, tilting the balance toward impatience; the speedup frame reverses that order. In their studies the thoughts generated first predicted the valuation, and experimentally reversing the querying order shrank or eliminated the asymmetry. That is evidence the gap is partly constructed on the spot, not read off a fixed underlying preference.
Where it shows up
The asymmetry surfaces wherever a schedule can be renegotiated against a baseline. In contracts, late-delivery penalties are set far higher than early-completion bonuses for the same shift in dates, because one party frames the change as a loss and the other as a gain. In negotiation, whoever establishes the reference deadline controls whether a later move reads as a costly concession or a cheap favor. In products and subscriptions, keeping a benefit you already expect is defended harder than gaining a new one is rewarded. Project and public-budget planning inherits the same pattern: stakeholders demand outsized compensation for slippage while valuing acceleration modestly. Recognizing which date counts as the reference point tells you in advance which direction of change will feel expensive.
Limits and what to watch
The framing is real but bounded, and the loss-aversion story is not the whole account. Some studies find the gain-loss asymmetry only for smaller amounts and not larger ones, so a clean "delay equals loss" reading is incomplete. The query-theory result that reordering the reasons people generate can erase the gap means the effect is malleable and context-dependent rather than a fixed constant. Its size also moves with magnitude, with the length of the interval, and with whether the outcome is a receipt or a payment — the sign of the outcome flips the direction. The practical implication is to treat the delay premium as a designable quantity, not a law of nature: how you set the reference point and sequence the framing can deliberately amplify it or largely neutralize it.
Examples
Told a delivery already due will arrive a week later, you demand a big discount; told the same item can come a week earlier, you'll pay only a little for it.
A team asked to push a launch back a month wants extra budget as compensation; asked to pull the identical launch forward a month, they will trade away far less to get it.
Told your tax refund will land a fortnight late, you feel robbed. Offered the same refund a fortnight early for a small fee, you decline — fourteen days, priced twice.
A software vendor that slips a promised feature by a quarter faces churn and refund demands; shipping the same feature a quarter early wins only mild goodwill and little extra customers will pay for.
An airline charges a steep change fee plus fare difference to push a booked flight to a later date, yet moving the same trip earlier by the identical number of days often costs little or nothing — the later departure is framed as a loss to be penalized, the earlier one as a favor barely worth pricing.
First described in Loewenstein (1988).
Key references
- Scholten, M., & Read, D. (2013). Time and outcome framing in intertemporal tradeoffs. Journal of Experimental Psychology: Learning, Memory, and Cognition, 39(4), 1192-1212. doi.org/10.1037/a0031171
- Scholten, M., & Read, D. (2010). The psychology of intertemporal tradeoffs. Psychological Review, 117(3), 925-944. doi.org/10.1037/a0019619
- Weber, E. U., Johnson, E. J., Milch, K. F., Chang, H. H., Brodscholl, J. C., & Goldstein, D. G. (2007). Asymmetric discounting in intertemporal choice: A query-theory account. Psychological Science, 18(6), 516-523. doi.org/10.1111/j.1467-9280.2007.01932.x
- Frederick, S., Loewenstein, G., & O'Donoghue, T. (2002). Time discounting and time preference: A critical review. Journal of Economic Literature, 40(2), 351-401. doi.org/10.1257/002205102320161311
- Loewenstein, G. (1988). Frames of mind in intertemporal choice. Management Science, 34(2), 200-214. doi.org/10.1287/mnsc.34.2.200