Shipping the Good Apples Out: Per-Unit Costs and the Geography of Quality

Posted on October 11, 2026 by nathanalbright

Executive Summary

Travelers often meet a puzzle in reverse depending on where they go. In Greenland, bottled water is scarce while soda and café drinks abound. In Chile, a celebrated wine country, visitors find their friends abroad asking for Chilean wines that cannot be found on Chilean shelves, while instant Nescafé dominates the morning cup. This paper argues that both experiences come from the same economic force: costs charged per unit rather than per unit of value. Such costs reshape which goods travel, which stay home, and which formats win in distant markets. Economists call the core of this the Alchian-Allen effect, or “shipping the good apples out.” The paper sets out the mechanism, applies it to Chile and Greenland, surveys similar cases, and notes where the effect weakens.

1. The Mechanism

The idea originates with economists Armen Alchian and William Allen, who observed in 1964 that a per-unit transaction cost lowers the relative price of the higher-quality version of a good. A simple illustration shows why. Suppose a winery sells a basic bottle for $5 and a fine bottle for $10. At home, the fine bottle costs twice as much. Add $5 of freight and handling to each bottle sent abroad, and the prices become $10 and $15. Abroad, the fine bottle now costs only 1.5 times as much. Distant buyers therefore shift toward the better product, and producers respond by sending their best output where it earns the most. The home market keeps the bulk grades.

The effect has strong empirical support. In a 2004 study in the Journal of Political Economy, David Hummels and Alexandre Skiba used detailed data on traded goods’ prices, quantities, and shipping costs across many countries and found a strong confirmation of the classical Alchian-Allen hypothesis. They found that within narrow product categories, exporters charge prices that vary by destination, rising with shipping costs and falling with tariffs, and that shipping costs act as a quantitative restriction much like quotas. A 2019 study applied the theorem to three quality grades of coffee sold globally and confirmed that a common per-unit charge raises the overall quality of coffee demanded.

A key distinction follows from this research. Per-unit costs, such as freight by weight or a flat fee per container, strengthen the effect. Costs set as a percentage of value, such as ad valorem tariffs, weaken it. Hummels and Skiba found that the Alchian-Allen effect decreases as true ad valorem costs like tariffs grow. A per-unit tax tilts trade toward premium goods. A percentage tax leaves relative prices largely where they were.

2. Two Responses to Per-Unit Costs

Per-unit costs push producers and consumers in two directions, and the two cases in this paper illustrate each.

Quality selection. When the product cannot be made lighter, the market selects for higher-value grades. The expensive version travels, the cheap version stays home. This is the Chilean wine case.

Format concentration. When the product can be made lighter or denser, the market selects for concentrated formats that carry less weight per unit of value, and the bulk ingredient, usually water, is added at the destination. This is the Chilean coffee case and the Greenland café case.

Both responses share one principle: a fixed cost per unit rewards whatever packs the most value into each unit shipped.

3. Case Study: Chilean Wine

Chile is among the most export-dependent wine producers in the world. Using OIV data, one analysis found that in 2020 Chile exported 83 percent of its wine production, with China a leading destination. More recent figures show the same orientation. In 2024 Chile exported 780 million liters of wine, up 14.4 percent from the prior year, while national wine production fell 15.6 percent to just over 930 million liters. Chile ranks as the fourth largest wine exporter in the world, and wine accounts for 18.8 percent of the country’s agricultural exports.

The home market is small by comparison. Chile relies more heavily on exports than Australia because its domestic market is smaller; total wine consumption in Chile in 2017 was 308 million liters, or 15.7 liters per person. Estimates for recent years run lower, with past figures suggesting a national average of around 10.6 liters per person.

These conditions produce exactly what Nathan’s friends experienced. When most of the output goes abroad and foreign buyers favor premium grades because of the per-bottle freight, producers develop labels, blends, and brands built around particular foreign markets and retailers. Some of those wines are never stocked in Chile, since the domestic market leans toward everyday price points where freight from the vineyard to a Santiago store adds little. The wine that foreign friends request may be a genuine Chilean product that functions in practice as an export good only. (This paper did not locate a systematic count of export-only Chilean labels; the pattern is consistent with industry practice but should be treated as an inference.)

4. Case Study: Instant Coffee in Chile

Chile grows no coffee, so it faces the opposite problem from a producer. It must import every cup. Here the format-concentration response dominates. A 2025 study by Corpa found that Chilean respondents preferred instant powdered coffee (66 percent) over whole-bean coffee (28 percent). The analyst attributed the instant preference to convenience and household tradition, especially among middle- and lower-income segments. The market is highly concentrated: Nestlé, with its Nescafé and Nespresso brands, holds an estimated 80 to 90 percent share of instant and freeze-dried coffee.

Instant coffee is the coffee equivalent of the Greenland smoothie. It is a dense, shelf-stable concentrate that ships at low weight per serving and is finished with local hot water. For a country importing all its coffee across long distances, that format offered a cost and convenience advantage early, and habit preserved it.

The pattern is shifting as incomes rise. Chain coffee shops in Chile grew 97.9 percent between 2011 and 2016 according to Euromonitor, while retail whole-bean sales grew 39.5 percent and instant sales 16.7 percent. By 2026, whole-bean coffee had reached 28 percent and was the fastest-growing segment, and 13 Chilean cafés appeared in a ranking of the 100 best in South America. This fits the Alchian-Allen logic: as buyers grow wealthier, the premium grade becomes affordable enough that the relative-price advantage of the per-unit cost begins to pull imported whole beans into the market as well.

5. Case Study Revisited: Greenland

The companion paper on Greenland showed both responses at work at once. Greenland’s beverage packaging import tax is a flat per-unit charge, set at 1 krone 40 øre up to 0.25 liter and 2 kroner 90 øre above that. Combined with weight-based freight through a single shipping concession, this per-unit structure produced the classic outcomes:

  • Quality selection favored higher-priced drinks such as branded soda, which can absorb a fixed fee per bottle, over plain water, which cannot.
  • Format concentration favored coffee, tea, and smoothies, whose concentrated ingredients travel by ship while the bulk of the drink comes from local taps.

Greenland’s case also shows that policy can create Alchian-Allen effects deliberately or by accident. Economists have long studied how specific (per-unit) and ad valorem taxes affect product quality differently, as in John Kay and Michael Keen’s 1991 work. A government that taxes per container rather than per krone of value tilts its shelves toward premium goods, whether or not that was the intent.

6. Other Recurring Examples

The pattern appears across many commodities. The following cases are widely reported and illustrate the principle, though this paper did not verify current figures for each:

  • Tea from Sri Lanka and Kenya: top leaf grades are auctioned for export, while smaller grades are more common in local cups.
  • Coffee in producing countries: for much of the twentieth century, growers in Brazil and Colombia exported their best beans while domestic consumption ran to lower grades or instant. Ethiopia stands as a notable exception, consuming a large share of its own harvest.
  • Premium seafood: Greenland’s shrimp and halibut, Norwegian salmon, and sushi-grade tuna command export prices high enough that the best catch is seldom sold locally at ordinary prices.
  • Planned-economy “export versions”: several centrally planned economies produced export models of manufactured goods held to higher standards than domestic models.

7. Limits of the Effect

The Alchian-Allen effect is a tendency, not a law, and several factors can blunt it:

  1. Value-based charges. Where insurance and handling fees scale with price, they offset the per-unit freight advantage. One study of Japanese and Korean car exports to the United States found that the higher-quality cars were not necessarily relatively cheaper with higher shipping costs, likely because insurance and other charges scale with the car’s price.
  2. Strong domestic demand. France keeps most of its own wine, exporting only about 29 percent of production, presumably because its domestic market is large and attached to its own product. A large, wealthy home market can outbid foreign buyers for premium grades.
  3. Rising local incomes. As the Chilean coffee data show, growing prosperity at home can pull premium goods back into the domestic market.
  4. Supply-side structure. Some theoretical work argues that once suppliers’ decisions are taken into account, the effect holds only in limited settings.

8. Conclusions

The two travel experiences that prompted this paper, unfindable bottled water in Greenland and unfindable Chilean wine in Chile, are not oddities of particular places. They reflect a general rule of trade: when every unit carries a fixed cost to move, the goods that travel are those that pack the most value into each unit. High grades go abroad and low grades stay home; concentrates travel and water is added locally; cheap bulk goods, plain water above all, may not travel at all.

For travelers, the rule offers a practical forecast. In a country that exports a famous product, the best examples are often easier to buy abroad than at the source. In a remote country that imports nearly everything, the shelves will favor premium goods and concentrated formats, and the humblest products will be the hardest to find.

Sources:

Unknown's avatar

About nathanalbright

I'm a person with diverse interests who loves to read. If you want to know something about me, just ask.
This entry was posted in Musings and tagged , , , , . Bookmark the permalink.

Leave a Reply