5D Library › Choice Architecture, Cumulative Prospect Theory and Computational Behavioural Economics of Pricing Interfaces · 中文

Chapter 2 Struck-Through and Reference Prices: Transaction Utility, Credibility and Price-Indication Rules

What a struck-through number can do

Product pages often show a higher price struck through, the lower current price beside it and a line such as “save 32 %”. The author’s framework asks to refute the claim that “simply striking through a former price raises willingness to pay without limit”. This chapter first introduces the theory that explains why the practice might work (transaction utility and reference prices), then looks at how large the reference-price effect is in published evidence and where it stops, then uses an author’s model to show why the gain in willingness to pay has a ceiling, and finally discusses the US and EU rules on “former prices” and what those rules allow and forbid in arithmetic terms.

Transaction utility and reference prices

Evidence

Thaler’s 1985 “mental accounting” framework splits the total utility of a purchase into two parts: acquisition utility, the value of the good to the buyer minus its price; and transaction utility, the “reference price” (what the buyer considers fair or expected) minus the actual price, measuring whether the deal is a good one (Thaler 1985). Reference prices can come from memory of past prices (internal reference prices) or from prices displayed by the seller (external reference prices, such as struck-through or “recommended retail” prices). Mazumdar, Raj and Sinha’s 2005 review organises the literature on how reference prices form, how they are measured and what they do (Mazumdar et al. 2005).

In this framework a struck-through price works by raising the buyer’s reference price and hence transaction utility. If the buyer purchases whenever acquisition plus transaction utility is non-negative, with value VV, reference price rr and weight θ\theta on transaction utility, the most she will pay is

WTP=V+θr1+θ.\mathrm{WTP} = \frac{V + \theta\, r}{1 + \theta}.

With θ=0\theta = 0, WTP is just VV: only the good counts. With θ>0\theta > 0, each dollar added to rr raises WTP by θ/(1+θ)\theta/(1+\theta) dollars. The key question is therefore how far a displayed struck-through price RR can raise rr. If rr rose without limit with RR, so would WTP — exactly the claim the framework wants refuted.

Evidence: the effect exists, but credibility is discounted

Evidence

Urbany, Bearden and Weilbaker (1988) compared “plausible” and “exaggerated” advertised reference prices: exaggerated reference prices were less believable, but, as later literature reports, still raised perceived value relative to showing no reference price (Urbany et al. 1988). Compeau and Grewal’s 1998 integrative review brought together research on comparative price advertising (advertising that displays a reference price) and its public-policy implications (Compeau and Grewal 1998). Krishna and co-authors’ 2002 meta-analysis of 20 marketing studies found that both the percentage and the dollar size of a deal raise perceived savings, the percentage more so; displaying the regular price as an external reference raises perceived value for large plausible deals and for implausible deals, but not for small plausible deals (Krishna et al. 2002).

This evidence says two things, partly differently from what the framework expects. First, reference-price effects do exist, and exaggerated reference prices, though less believable, did not reverse into negative effects in the studies checked. So “implausibly high reference prices backfire” is a possibility, not an established result; the “suspicion” model later in this chapter states it as a hypothesis. Second, “credibility is discounted” is itself enough to refute “without limit”: as long as buyers give ever smaller weight to ever more exaggerated reference prices, the gain in WTP is bounded. A model makes this explicit.

The author’s model: saturating credibility and suspicion

Let the buyer’s internal reference price be r0r_0. The author’s model gives two forms of perceived reference price r(R)r(R):

  • Saturating: r=r0+D(1−e−(R−r0)/D)r = r_0 + D\,(1 – e^{-(R – r_0)/D}). A higher struck-through price raises the reference price by ever smaller increments, at most by DD; exaggeration is discounted but never reverses.
  • Suspicion: as the saturating form, minus b(R−hr0)b\,(R – h\, r_0) once RR exceeds hr0h\, r_0: a very high struck-through price makes the buyer doubt the whole offer and lowers the reference price.

In the saturating form the gain in WTP is bounded by θD/(1+θ)\theta D/(1+\theta), whatever RR is.

Calculation

# Part A: THE AUTHOR'S MODEL of a struck-through ("was") price R. A buyer with internal
# reference r0 and consumption value V pays at most WTP = (V + theta * r) / (1 + theta),
# where theta * (r - p) is transaction utility (Thaler 1985), r the perceived reference.
#   saturating: r = r0 + D (1 - exp(-(R - r0) / D)) -- discounted, never reversed
#   backfire:   the same, minus a suspicion penalty b (R - h r0) once R exceeds h r0
# Either way the gain in WTP is bounded: no struck-through price raises it without limit.
# Part B: the EU rule (Directive 98/6/EC, Art. 6a, inserted by Directive (EU) 2019/2161):
# an announced reduction is measured against the lowest price applied in the 30 days
# before it. A fictitious 90-day price history shows what each reading allows.
import math

R0, V, THETA, D, H, B = 40.0, 40.0, 0.6, 8.0, 1.5, 0.25

def ref(R, backfire=False):
    r = R0 + D * (1 - math.exp(-max(R - R0, 0) / D))
    if backfire and R > H * R0:
        r -= B * (R - H * R0)
    return r

def wtp(R, backfire=False):
    return (V + THETA * ref(R, backfire)) / (1 + THETA)

base = wtp(R0)
print("author's model: WTP gain over 'no reference' (r0 = V = 40) by was-price R")
print("   R (was)   saturating   backfire")
for R in (40, 45, 50, 60, 70, 80, 100, 150, 200, 400):
    print(f"   {R:7.0f}   {wtp(R) - base:+10.2f}   {wtp(R, True) - base:+8.2f}")
cap = THETA * D / (1 + THETA)
print(f"   ceiling of the saturating gain: theta D / (1 + theta) = {cap:.2f}"
      f" ({cap / base:.1%} of WTP)")
best = max((R0 + 0.01 * i for i in range(20001)), key=lambda R: wtp(R, True))
neg = min(R0 + 0.01 * i for i in range(1, 40001) if wtp(R0 + 0.01 * i, True) < base)
print(f"   backfire variant peaks at R = {best:.1f} ({best / R0:.2f} x r0), gain"
      f" {wtp(best, True) - base:+.2f};")
print(f"   its gain turns negative above R = {neg:.1f}")
print()
hist = [39.0] * 45 + [59.0] * 14 + [39.0] * 6 + [29.0] * 25        # days 0..89
day = 65
win = hist[day - 30:day]
print("price history: 39 for 45 days, 59 for 14 days, 39 for 6 days,"
      f" 'sale' at 29 from day {day}")
for lab, r in (("price the day before the sale", hist[day - 1]),
               ("highest price in the 30 days before", max(win)),
               ("lowest price in the 30 days before (Art. 6a)", min(win))):
    print(f"   {lab:46s} {r:5.1f} -> announced cut {1 - 29 / r:4.0%}")
hist2 = [39.0] * 45 + [59.0] * 20 + [29.0] * 25
print(f"59 held right up to the sale: lowest in 30 days {min(hist2[day - 30:day]):.0f},"
      f" announced cut {1 - 29 / min(hist2[day - 30:day]):.0%} (not {1 - 29 / 59:.0%})")
hist3 = [59.0] * 65 + [29.0] * 25
print(f"59 for all 65 days: lowest in 30 days {min(hist3[day - 30:day]):.0f},"
      f" announced cut {1 - 29 / 59:.0%} is allowed by this rule")
# author's model: WTP gain over 'no reference' (r0 = V = 40) by was-price R
#    R (was)   saturating   backfire
#         40        +0.00      +0.00
#         45        +1.39      +1.39
#         50        +2.14      +2.14
#         60        +2.75      +2.75
#         70        +2.93      +1.99
#         80        +2.98      +1.10
#        100        +3.00      -0.75
#        150        +3.00      -5.44
#        200        +3.00     -10.13
#        400        +3.00     -28.88
#    ceiling of the saturating gain: theta D / (1 + theta) = 3.00 (7.5% of WTP)
#    backfire variant peaks at R = 60.0 (1.50 x r0), gain +2.75;
#    its gain turns negative above R = 92.0
#
# price history: 39 for 45 days, 59 for 14 days, 39 for 6 days, 'sale' at 29 from day 65
#    price the day before the sale                   39.0 -> announced cut  26%
#    highest price in the 30 days before             59.0 -> announced cut  51%
#    lowest price in the 30 days before (Art. 6a)    39.0 -> announced cut  26%
# 59 held right up to the sale: lowest in 30 days 39, announced cut 26% (not 51%)
# 59 for all 65 days: lowest in 30 days 59, announced cut 51% is allowed by this rule

The program uses r0=V=40r_0 = V = 40, θ=0.6\theta = 0.6, D=8D = 8, h=1.5h = 1.5, b=0.25b = 0.25 (illustrative). In the saturating form, raising the struck-through price from 40 to 60 adds 2.75 to WTP; raising it to 400 adds only 3.00, which is the bound θD/(1+θ)=3.00\theta D/(1+\theta) = 3.00, or 7.5 % of WTP. In the suspicion form, WTP peaks at R=60R = 60 (1.5 times the internal reference) and the gain turns negative above 92.0. Both refute “without limit”; they differ on whether extreme exaggeration is “useless” or “harmful”, which is for evidence to decide, and the evidence checked is closer to the saturating form. The left panel of Figure 2 shows both curves.

Note

Where the parameters come from. θ\theta, DD, hh and bb are illustrative values, not estimates. The model shows only the structure: as long as credibility falls fast enough with exaggeration (here, exponential saturation), the gain from any struck-through price has a ceiling. How high the real ceiling is, and whether it reverses, depends on the product category, the buyer’s familiarity with prices and the seller’s reputation, and has to be measured in context.

Where internal reference prices come from

A struck-through price is an external reference; the buyer’s internal reference price comes mainly from the prices she has seen. Mazumdar and co-authors’ review sorts the sources of reference prices into several kinds, and one common class of models writes the internal reference price as a weighted average of past observed prices, adjusting a step towards each new price seen (Mazumdar et al. 2005). In such adaptive models frequent promotions have an easily overlooked consequence: every purchase at a low price pulls the buyer’s internal reference down, so that later purchases at the regular price carry negative transaction utility and the regular price looks “expensive”. In other words, today’s struck-through price raises today’s transaction utility, but using it often may lower tomorrow’s reference price; the author’s model in this chapter describes a single purchase and leaves this dynamic out.

There is an interesting correspondence with the 30-day rule of Article 6a: by defining the “prior price” as the recent lowest price, the rule in effect asks the seller to use a conservative reference point that moves in the same direction as the buyer’s adaptive reference price, rather than a high price the buyer has never seen. Whether the rule was motivated by this, the book does not infer; it notes only that the two point the same way arithmetically.

Anchoring evidence needs qualifying too

Struck-through prices are sometimes described as “anchoring”: an arbitrary number affecting later judgement. In Ariely, Loewenstein and Prelec’s 2003 experiments participants first thought of the last two digits of their social security number and then bid for goods; those with higher numbers bid more, which the authors called “coherent arbitrariness” (Ariely et al. 2003). Maniadis, Tufano and List (2014) repeated similar experiments with larger samples, found much weaker anchoring than the original study and used this to discuss the replicability of single studies (Maniadis et al. 2014). Book 27 treated anchors and the best alternative to a negotiated agreement (BATNA) in bargaining; here we add only one point: a struck-through price differs from an arbitrary anchor in that it claims to be a real former price, so part of its effect comes from the buyer’s trust in that claim, and exploitation of that trust is precisely what the law regulates. Prices ending in 9 (such as 9.99) are another kind of price presentation: Anderson and Simester found in mail-order catalogue field experiments that prices ending in 9 raised demand (Anderson and Simester 2003), but this is a different mechanism from reference prices and is not pursued here.

Price-indication rules: what is a “former price”?

Evidence

Section 233.1 of the US Federal Trade Commission’s Guides Against Deceptive Pricing (16 CFR Part 233) states that a former-price comparison is legitimate if the former price is the actual, bona fide price at which the article was offered to the public on a regular basis for a reasonably substantial period of time; if an artificial, inflated price was established in order to offer a large reduction later, the “former price” is fictitious and the advertised bargain is false (U.S. Federal Trade Commission 1964). Article 6a of EU Directive 98/6/EC (inserted by Directive 2019/2161) requires any announcement of a price reduction to indicate the prior price, meaning the lowest price applied by the trader during a period of not less than 30 days before the reduction (European Parliament and Council of the European Union 1998, 2019); member states may provide differently for perishable goods, goods on the market for less than 30 days and progressively increasing reductions. On 26 September 2024, in Case C-330/23 (Verbraucherzentrale Baden-Württemberg v Aldi Süd), the Court of Justice of the EU confirmed that a reduction announced as a percentage or by other wording that stresses an advantageous price must be calculated from this 30-day lowest price, not from the price on the day before the offer (Court of Justice of the European Union 2024).

The second part of the program uses a fictitious 90-day price history to show the arithmetic of the two kinds of rule. An item sells at 39 for 45 days, is raised to 59 for 14 days, returns to 39 for 6 days and from day 65 is “on sale” at 29. Taking the price on the day before the sale as the former price, the claimed reduction is 26 %; taking the highest price of the last 30 days, it could be 51 %; under Article 6a the prior price is the lowest of the last 30 days, 39, so only 26 % can be claimed. Even if 59 had been kept right up to the day before the sale, the 30-day window would still contain 39, and the claim would still be 26 % (Figure 2, right). But if 59 had been charged for all 65 days, the 30-day low would be 59 and the 51 % claim would be allowed under this rule. So the 30-day-low rule blocks “raise briefly, then cut”; on its own it cannot exclude “list for a long time at a price nobody pays”, which needs a substantive test like the US Guides’ “bona fide … on a regular basis”.

Left: in the author’s model, the gain in willingness to pay against the struck-through price (saturating and suspicion forms, internal reference 40). Right: a fictitious price history and the 30-day window of Article 6a; under this rule the sale on day 65 can only use 39 as the prior price. Illustrative.

Note

Other jurisdictions. The author’s framework mentions Malaysian rules. This round of checking found no rule text that could be verified directly, so the book does not cite one. Rules differ from place to place and are revised; this chapter uses two verified sets of rules, the US and the EU, only to illustrate the arithmetic, and it is not an interpretation of the law anywhere.

Hypothesis · not established · Exaggerated struck-through prices damage trust

The author’s hypothesis (not established): when a struck-through price exceeds the buyer’s internal reference price by more than some multiple (about 2.3 times in the author’s model, 92.0/40; 1.5 times is where the gain peaks), a page displaying it has a lower purchase rate than a page without one (the suspicion form). What would show it wrong: in a pre-registered experiment randomising “no struck-through price”, “struck-through price at 1.2 times the reference” and “at 2.5 times” (each struck-through price must be a real former price, so that the experiment itself is not a false indication), the purchase rate in the 2.5-times arm is not lower than with no struck-through price, or the confidence interval for the difference includes zero.

Chapter summary

Thaler’s transaction utility explains a struck-through price as raising the reference price. The evidence checked shows that reference-price effects exist and that exaggerated reference prices lose credibility, without reversing in those studies; Krishna and co-authors’ meta-analysis shows the regular price helping large plausible deals and implausible ones, but not small plausible ones. The author’s model shows that as long as credibility saturates with exaggeration, the gain in WTP is bounded (3.00, or 7.5 %, in the example); with suspicion added, very high struck-through prices lower WTP, but that is a hypothesis. The US Guides require the former price to be bona fide and regularly offered; EU Article 6a requires the prior price to be the lowest price in at least the 30 days before a reduction, which blocks “raise briefly, then cut” but cannot by itself exclude a long-standing inflated list price. “Striking through a former price raises willingness to pay without limit” therefore fails in theory, in evidence and in law.

References

Anderson, Eric T., and Duncan I. Simester. 2003. “Effects of $9 Price Endings on Retail Sales: Evidence from Field Experiments.” Quantitative Marketing and Economics 1 (1): 93–110. https://doi.org/10.1023/A:1023581927405.
Ariely, Dan, George Loewenstein, and Drazen Prelec. 2003. “‘Coherent Arbitrariness’: Stable Demand Curves without Stable Preferences.” Quarterly Journal of Economics 118 (1): 73–105.
Compeau, Larry D., and Dhruv Grewal. 1998. “Comparative Price Advertising: An Integrative Review.” Journal of Public Policy & Marketing 17 (2): 257–73. https://doi.org/10.1177/074391569801700209.
Court of Justice of the European Union. 2024. Judgment of 26 September 2024, Verbraucherzentrale Baden-württemberg v Aldi süd, Case C-330/23. Preliminary ruling on Article 6a of Directive 98/6/EC.
European Parliament and Council of the European Union. 1998. Directive 98/6/EC of 16 February 1998 on Consumer Protection in the Indication of the Prices of Products Offered to Consumers. Official Journal of the European Communities, L 80, 18 March 1998; Article 6a as inserted by Directive (EU) 2019/2161.
European Parliament and Council of the European Union. 2019. Directive (EU) 2019/2161 of 27 November 2019 Amending Council Directive 93/13/EEC and Directives 98/6/EC, 2005/29/EC and 2011/83/EU as Regards the Better Enforcement and Modernisation of Union Consumer Protection Rules. Official Journal of the European Union, L 328, 18 December 2019, p. 7.
Krishna, Aradhna, Richard Briesch, Donald R. Lehmann, and Hong Yuan. 2002. “A Meta-Analysis of the Impact of Price Presentation on Perceived Savings.” Journal of Retailing 78 (2): 101–18.
Maniadis, Zacharias, Fabio Tufano, and John A. List. 2014. “One Swallow Doesn’t Make a Summer: New Evidence on Anchoring Effects.” American Economic Review 104 (1): 277–90.
Mazumdar, Tridib, S. P. Raj, and Indrajit Sinha. 2005. “Reference Price Research: Review and Propositions.” Journal of Marketing 69 (4): 84–102. https://doi.org/10.1509/jmkg.2005.69.4.84.
Thaler, Richard H. 1985. “Mental Accounting and Consumer Choice.” Marketing Science 4 (3): 199–214.
Urbany, Joel E., William O. Bearden, and Dan C. Weilbaker. 1988. “The Effect of Plausible and Exaggerated Reference Prices on Consumer Perceptions and Price Search.” Journal of Consumer Research 15 (1): 95–110.
U.S. Federal Trade Commission. 1964. Guides against Deceptive Pricing, 16 c.f.r. Part 233. Code of Federal Regulations, Title 16, Part 233 (§ 233.1 Former price comparisons).

Fan, Y. (2026). Choice Architecture, Cumulative Prospect Theory and Computational Behavioural Economics of Pricing Interfaces: From Loss-Aversion Value Functions and Asymmetric Dominance to Hyperbolic Discounting, Read Beside the Qingzhong Chapters of the Guanzi [选择架构、累积前景理论与销售界面价格锚定的计算行为学] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23836260 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).