International Trade
A country exports the good it makes at lower opportunity cost, and the terms of trade must lie between the two countries' opportunity costs or one country refuses to trade. The gain shows up as the CPC lying outside the PPC.
Where comparative advantage comes from
- Same idea as Unit 2
- Comparative advantage, now between countries instead of people.
- Factor abundance
- A country has a lot of some input, so goods that use it heavily are cheap for it to make. Minerals drove early US industrialization; climate and soil decide where coffee grows; capital and skilled labor shape the US today; abundant less-skilled labor gives developing countries low-tech manufacturing.
- Dynamic advantage
- Acquired by getting an early start, not inherited.
- Terms of trade
- The world relative price of your exports — here, machines per ton of soybeans. Soybeans $400 and machines $300 gives 400/300 = 1⅓ machines per ton.
The two-country example
| Output per worker-day | Soybeans (tons) | Washing machines | Opp. cost of 1 ton soy | Opp. cost of 1 machine |
|---|---|---|---|---|
| U.S. | 2 | 2 | 1 machine | 1 ton |
| China | 1 | 2 | 2 machines | ½ ton |
The US exports soybeans, China washing machines. US PPC slope −1, China's −2. Terms of trade must sit between 1 and 2 machines per ton. At ⅔ machines per ton — soybeans $200, machines $300 — China would buy but the US would refuse to sell, and excess demand would bid the relative price of soybeans back up.
Linear versus bowed-out PPC
| PPC | Opportunity cost | Tangency | Specialization |
|---|---|---|---|
| Linear | Constant | At a corner | Complete — the simple model |
| Bowed out | Rises as you specialize | Interior | Incomplete — the honest version |
Does trade actually help? The evidence chain
- Sachs and Warner
- Always-open economies grew much faster than always-closed ones, 1966–90.
- The objections
- Reverse causation — getting rich may make you trade more. Omitted variables — open countries adopt other growth-friendly policies too.
- Frankel and Romer
- An instrument: predict trade from geography — size, distance from partners, landlocked — which plausibly affects income only through trade. The geography-predicted component of trade is still positively associated with income per worker.
- Dollar and Kraay
- Developing-country "globalisers" that opened from the late 1970s grew substantially faster afterwards.
Rules that earn points
- Build the opportunity cost table: both goods, both countries, four numbers.
- Assign each good to its low-opportunity-cost producer.
- State the range the terms of trade must fall in, and why outside it one country refuses.
- On the graph: PPC, then the CPC as a line with slope = −terms of trade tangent to it, then production and consumption points.
- Terms of trade upside down. It is the price of your exports relative to your imports.
- Drawing the CPC inside the PPC. If it is not outside, there are no gains.
- Claiming complete specialization with a bowed PPC. That is the linear case only.
A US worker-day produces 2 tons of soybeans or 2 washing machines; a Chinese worker-day produces 1 ton or 2 machines. Then:
- AThe US exports soybeans; China exports washing machines
- BChina exports soybeans; the US exports washing machines
- CThe US exports both, because it is at least as productive in both goods
- DNeither country gains, since both make 2 machines per worker-day
Answer
A — US soybeans, China machines.
In the US a ton of soybeans costs 1 machine; in China it costs 2. The US is the low-opportunity-cost soybean producer, so China is the low-opportunity-cost machine producer.
C is the trap: it reads absolute advantage off the output numbers. Trade patterns follow opportunity cost, and a country cannot have a comparative advantage in everything.
For both countries in that example to want to trade, the terms of trade must be:
- ABetween ½ and 1 machine per ton of soybeans
- BBetween 1 and 2 machines per ton of soybeans
- CExactly 1⅓ machines per ton of soybeans
- DAnything, since both countries gain at any world price
Answer
B — between 1 and 2 machines per ton.
Below 1 the US would rather build its own machines; above 2 China would rather grow its own soybeans. The world price must lie between the two opportunity costs.
C is the trap: 1⅓ (soybeans $400, machines $300) is a perfectly good example of a workable price, but the question asks for the range.
Suppose world prices were soybeans $200 and machines $300, so the terms of trade are ⅔ machines per ton. What happens?
- ABoth countries trade, with the US exporting soybeans
- BThe US refuses to sell soybeans, and excess demand bids their relative price back up
- CChina refuses to buy, and the relative price of soybeans falls further
- DChina exports soybeans and the US exports machines
Answer
B — the US refuses to sell and the price is bid back up.
⅔ machines per ton is below the US opportunity cost of 1. Selling soybeans abroad would get the US less than making machines itself, so it produces for itself. China still wants soybeans at that price, and the resulting excess demand pushes the relative price of soybeans back into the 1-to-2 range.
A is the trap: the direction of comparative advantage is right, but at this price trade does not happen at all.
On a PPC/CPC diagram, the gains from trade appear as:
- AThe consumption possibilities curve lying outside the production possibilities curve
- BThe production possibilities curve shifting outward
- CThe production point moving to a corner of the PPC
- DThe CPC lying inside the PPC but with a steeper slope
Answer
A — CPC outside the PPC.
The country consumes a bundle it could not possibly have produced. That gap is the gain.
B is the trap. Trade does not change what the country can produce — its resources and technology are unchanged. It changes what the country can consume.
Draw a bowed-out production possibilities curve with a consumption possibilities curve. Mark the production point, the consumption point, exports and imports, and say where the gains from trade are.
Model answer
Draw the bowed-out PPC. The CPC is a straight line with slope equal to minus the terms of trade, drawn tangent to the PPC.
Point A is the tangency: the country produces there because that bundle has the highest value at world prices. Point B is elsewhere along the CPC, where it consumes after trading. The horizontal gap between A and B is exports of the good on that axis; the vertical gap is imports of the other.
The gains are visible as the CPC lying outside the PPC — the country consumes a bundle it could not have produced.
- CPC drawn as a straight line tangent to the PPC, not a curve.
- Slope of the CPC named as minus the terms of trade.
- Production at the tangency, consumption elsewhere on the CPC — two distinct labelled points.
- Exports and imports marked as the horizontal and vertical gaps.
- Gains identified as CPC outside PPC, with a bowed PPC giving an interior tangency and therefore incomplete specialization.
Sachs and Warner found that economies open to trade from 1966 to 1990 grew much faster than closed ones. Give two reasons not to read that as causal, and explain how Frankel and Romer address the problem.
Model answer
Reverse causation: being rich may cause a country to trade more, rather than trade causing income. Omitted variables: countries that adopt free trade tend to adopt other growth-friendly policies at the same time, so openness may just be a marker for good policy generally.
Frankel and Romer use geography as an instrument. They predict how much a country trades from its size, its distance from trading partners, and whether it is landlocked. Geography is not caused by income and plausibly affects income only through trade. The geography-predicted component of trade is still positively associated with income per worker, which is much harder to explain away. Dollar and Kraay add that developing-country "globalisers" that opened from the late 1970s grew substantially faster afterwards.
- Both objections named with the right labels: reverse causation and omitted variables.
- Calling the geography variable an instrument.
- The two conditions: geography is not caused by income, and affects income only through trade.
- Stating the result — the predicted component of trade still correlates with income per worker.