As a very general rule, the richer a country is, the less likely any given consumer purchase is to involve haggling. In the 19th century, general stores in the Western US often didn't post prices, both because haggling was the norm and because some of their business was barter—currency was scarce, but since land was so cheap there were plenty of productive workers who could basically barter with one another indirectly through the store. Once that same town has a bank and is getting products from big cities rather than surrounding farms and ranches, and once people's income and thus opportunity cost per hour went up, it made more sense to just put up a sign saying how much a dozen eggs cost rather than sizing up customers one by one to see what they'd trade for them.

And that's a decent sketch of why lunch prices are more negotiable at a small restaurant in a poor country than they are at, say, a Chipotle in midtown Manhattan. The poorer the customer is, the more price-sensitive they have to be, and the more likely it is that there's some reserve price at which it's worth doing business with them even if it's not a great deal, while better-off customers might be indifferent (or, if they're tourists, oblivious) to the possibility of saving a little money.

But, in another sense, that Chipotle is a hotbed of haggling! Every time a customer says that the scoop of barbacoa was a little light, or expresses annoyance about the delay and gets comped a free soda, it's basically lightning-fast structured haggling, with the company or store manager setting some general policy for how much to flex their offerings. This doesn't feel like haggling because the dollar prices are fixed, but that's because it's easier to adjust the bundle of goods and services offered than to override the preset prices in the point-of-sale system. And, given the markup on ingredients, this is value-maximizing for both sides! Getting a dollar worth of free guac because you're a regular costs Chipotle a lot less than a dollar in ingredients and labor (especially on the ingredient side, where you're dealing with a substance that has an incredibly short shelf life). Chipotle would rather keep you a regular, and maximize your lifetime value, than turn you off the store completely by shorting you on guac. Guac as CAC! 

You'll expect more haggling in a context where information is scarce, and one where the return on time spent haggling is high for both sides. The latter condition often prevails in poor countries, where there's scarce capital and abundant labor, so there simply isn't as much productive work to do because there are so many people competing to do it. Or, in other cases, it's sufficiently physically demanding work that it isn't realistic to do all day, a situation that might make someone very sensitive to the dollar/calorie tradeoff but comparatively indifferent to the time cost required to optimize it.

The information-scarcity component shows up in haggling, too, and in particular in the contexts where it still persists in rich countries: big-ticket purchases like houses and cars, and business-to-business transactions. One way to frame the negotiation between a software vendor and a seven-figure customer is that both sides are sizing up the potential surplus the other side will generate. If my clever anti-fraud tool saves your company $10m a year, I want to capture as much of that as possible. But the fact that I'm going to the effort to sell it implies that it's not a slam dunk for everyone who tries it. That's pretty much a necessity for any given complex sale—even if the owners of a gas turbine, a 787, a GB200 NVL72 server, or an EUV machine all get positive returns, it requires a very specific owner to get any positive return from using these. And because these specific circumstances vary so much, the return on spending a given amount on this equipment—or, more to the point for the seller, the price point at which that return hits whatever the relevant hurdle rate is—becomes very valuable information.

In these cases, haggling goes back to its classic model, just more explicitly and with more zeroes tacked on. The process ends up measuring cases where each side's elasticities don't line up—maybe the turbine owner has priced out third-party repair and maintenance services and knows they aren't giving up much when they buy this directly from the manufacturer; perhaps the 787 buyer is flush with cash, the seller has some upcoming debt maturities, and thus it’s mutually-beneficial for the buyer to wire a slightly smaller-than-sticker-price sum the second the deal is agreed to rather than stretching out payments.

Even on the consumer side, we engage in a sort of individualized branch of a collective haggling process any time we consume content that is monetized with ads, because we're basically engaged in a continuous debate over exactly how much commercial distraction we're willing to tolerate for a given amount of entertainment.¹ Meanwhile, platforms are haggling in the other direction. On a like-for-like basis, ad load almost always goes up, though aggregate ad load growth is throttled because the most shameless sites die off. The reason they can raise the frequency of ads over time is that as they scale, it's easier to find something to keep users engaged, and also that the more they scale, the better-targeted the ads can be.

So, if haggling is the literal process of shouting out offers and counteroffers until you converge on something, or don't, it is indeed a feature of economics that is poor in wage and information terms. But it doesn't entirely go away. In fact, it's arguably bigger: competitive markets amount to continuous haggling at scale, and by systematizing that process, they make it even more economically efficient.

We’ve covered price discrimination and information asymmetry across many scales and contexts in The Diff, including: 

• The limits to using ads as a tool of price discrimination ($).

• What works in SaaS pricing.

• The potential rise of the AI sandwich model of outsourced haggling ($).

• The economics of fast-food price discrimination are surprisingly complex ($).

• In praise of the economic thinking of Ronald Coase.

Join the discussion!

1  This actually pushes scripted entertainment in a strange direction: towards having a pretty steady amount of excitement over the course of an episode so there isn't a specifically boring point to tune out. So the Fly episode of Breaking Bad might have been A/B tested out of existence given sufficiently ubiquitous ads.