Topic 5: Alternative theories of trade

Topic 5: Additional theories of trade

IntroductionTo this point we have studied the 2 primary “benchmark” theories of what explains international trade and its effects:

1. Ricardian (classical) model: useful for understanding how productivity or technology differences determine comparative advantage and rive trade and for studying the basic gains from trade.

2. Factor-proportions (HO) model: useful for understanding how national economic characteristics (factor endowments) interact with industry variables (factor intensities) to determine comparative advantage and trade. Also important for studying how trade and globalization affect income distribution between scarce and abundant factors.

There is now actually a 3rd “benchmark” theory that focuses on individual firms and how they engage in trade. We will address that briefly when we discuss multinational enterprises and international investment later on.

IntroductionBut here are some problems with the 2 primary theories:

1. They rely on very restrictive assumptions. In economics this is necessary in order to build basic logical frameworks that give us clear theoretical predictions. But it ignores all kinds of interesting issues that clearly are important for trade, such as:

◦ 1. Products are not “homogeneous” (identical in quality). A Japanese car is actually different from an American car and a German car. Products are differentiated.

◦ 2. We often have increasing returns to scale rather than constant returns to scale.◦ 3. We may not have perfect competition. Instead there may be firms with market power.◦ 4. Technologies may be different across countries rather than identical. More than that, technologies are dynamic and

subject to change.

2. Governments do interfere with trade through tariffs, quotas, and regulations.

3. Beyond trade in goods and services, there is also trade in factors (labor and capital mobility across borders).

These notes will focus on some of the elements in point 1. Other items come later.

Addition 1: demand for variety and intra-industry tradeStart with some empirical points that should make us doubt the Ricardo and HO supply-side stories.

The greatest amounts of international trade are among rich countries with similar endowments, not between rich and poor countries with different endowments. The US trades far more with Canada and the EU than it does with Brazil and even with China and Mexico when you control for country size.

◦ Some simple data on this. The top 3 developed countries that the US trades with (exports plus imports) are in order (2014 data): Canada, Japan, Germany. Total 2-way US trade in goods is $1,031 billion. Top 3 developing countries are in order: China, Mexico, and Brazil (in terms of total trade with US, China is number 1 and Mexico number 3). Total 2-way US trade is $1,197 billion. But if you divide by the populations of those 3 countries you get:◦ 2-way US trade per person with Canada, Japan, Germany = $4,200.

◦ 2-way US trade per person with China, Mexico, Brazil = $713.

◦ So trade is considerably higher in proportional terms among rich countries.

This kind of trade is dominated by differentiated products. These are similar versions of particular goods (cars, wines, machines, chemicals, medicines, cosmetics, computers, even bottled water, etc.) that are differentiated in some ways.

Demand and IITHow are products differentiated?

◦ By quality: the design, style, and performance of goods and services. Think of “typical” cars by country. Or “typical” furniture by country. What explains those differences?

◦ By brand loyalty: firms advertise to build brands and trademarks that attach (at least in the consumers’ minds) various desirable characteristics to their goods.

◦ Sometimes these differences are mostly perceived rather than real. Do we really think there are 45 different qualities of diet soft drinks? How about bottled water? Shampoos?

◦ By timing: it’s hard to get fresh fruit from the US in the winter. Maybe Chilean grapes are better then?

What all of this suggests is that consumer demands are more complex (and interesting) than simply thinking they consume corn and wheat or autos and shoes.

Instead, consumers have a clear preference for variety. We like to be able to choose from among multiple versions of types of goods, which may differ in quality.

Put differently: Suppose you have a $20 candy budget. You are better off buying 10 different candy bars (price of $2) than 10 of the same candy bar.

Note this is also true for firms that choose among varieties of inputs in deciding how to produce final goods.

The gains from variety in tradeWhat is the role of trade? If we believe (reasonably) that countries produce different varieties of similar goods, then opening up to trade benefits consumers and firms from a variety gain. International trade offers more choices than a closed economy does.

If countries like to trade with each other in similar goods, we must have a lot of intra-industry trade (IIT). This is the phrase economists use to describe the fact that countries trade back and forth quite a bit of differentiated versions of similar products.

What are obvious examples?◦ Wine and beer;◦ Clothing of different quality levels;◦ Automobiles;◦ Furniture;◦ High-tech precision machinery;◦ Movies and recorded music;◦ Ski-lift operators;◦ You name it.

Measuring IITWe can think of this as two-way trade in similar goods within industries. The text shows a formula for computing an index of IIT for a country in this equation (X is exports; M is imports):

𝐼𝐼𝑇 = 100 ∗ 1 − 𝑗=1𝑛 𝑋𝑗−𝑀𝑗

𝑗=1𝑛 (𝑋𝑗+𝑀𝑗)

Here, j indicates particular goods or industries that are groupings of goods. This calculation of IIT ranges between 0 and 100:

If IIT = 0 it means in any good j there are only exports or imports, not any trade in similar goods. This would be pure inter-industry trade, which is what Ricardian and HO models are about.

If IIT = 100 it means that for every good j, exports exactly equal imports and all trade is in similar goods. This would be pure intra-industry trade, driven by tastes for variety.

Measuring IITThe higher is IIT the greater is the proportion of intra-industry trade in total trade.

To illustrate, here are some figures for 2013 for US IIT with Canada, calculated for very aggregated industries. IIT (or 2-way trade) with Canada was 77.2% across all goods and 91.6% for manufacturing.

Measuring IITAnd here are similar figures for US trade with Mexico. IIT with Mexico across all goods was lower at 64.4% and even lower for all manufacturing goods at 59.8%.

Measuring IITBut with either calculation we find that there is a great deal of trade in both directions.

In North America the best example is the automobile industry. Cars within a firm (e.g., Ford, GM, Toyota) may be designed in the US. Designs and high-tech parts are sent to Mexico, where they are combined with other intermediate parts and processed into more advanced parts, then exported back to the US and Canada, where final cars are assembled.

All of this kind of trade tends to get counted in IIT figures. Even at the stage of final cars, the US exports some models to Canada and Mexico and imports other models from Canada and Mexico. These are extremely intricate supply chains and they exist in many manufacturing industries.

Some features of IIT suggested by the Canadian and Mexican IIT numbers.◦ IIT is more common among developed countries than between developed and developing countries.

◦ IIT is higher in manufacturing than in primary goods. This is because manufactured goods are much more differentiated and so there is good reason for 2-way trade between countries.

◦ The high levels of IIT calculated here reflect the very high degree of data aggregation. If we computed it for more detailed industries the indexes would be lower.

Why is IIT important? What explains it?The first to write about this phenomenon was Staffan Linder, a Swedish economist who explained it in terms of the demand similarity hypothesis (or Linder hypothesis):

Countries at similar levels of income have similar tastes for differentiated goods. Firms produce varieties for their home market and then trade permits firms to export their varieties, which are in greatest demand in other countries with similar incomes. This results in intra-industry trade among similar countries of differentiated products.

A simple example: suppose you were asked to explain why the United States exports furniture to Sweden but also imports furniture from Sweden. It seems unlikely that we can find the answer in differences in capital or labor intensity of the furniture industry between the countries.

A more likely explanation: US furniture is different from Swedish furniture in design and consumers care about this. Traditional US furniture is large, heavy, and complex in design. (Why, do you suppose?) Traditional Scandinavian furniture is small, sleek, and simple in design. (Why?) IIT emerges because there are US consumers who like sleek and simple furniture, while there are Swedish consumers who prefer the American style.

Why is IIT important? What explains it?Or think of cars. Is there a typical kind of US car? Italian car? Japanese car? German car? If they all differ somewhat in design and performance characteristics, we will find demand for all of them in each country.

This can hold for consumer goods obviously but it also holds for intermediate goods like chemicals and machinery. Germany produces some varieties of machines and the US produces other varieties of machines and they can trade back and forth.

More recent versions of this theory actually borrow from psychology and marketing studies to argue that consumers actually have a taste for variety. That is, the utility of consumers depends on the number of varieties of various goods available to consume, and the wider is the choice set the greater is welfare.

This seems true: consider the increases in versions of soft drinks, beers, wines, cosmetics, athletic equipment, cars, you name it.

This generates another source of GFT: international trade compared to autarky raises the number of varieties available to consume.

Are gains from greater variety through trade significant? Yes. 2 economists in 2006 estimated these gains (assuming consumers have a preference for variety) to be 2.6% of US GDP from 1972 to 2001.

Such gains are far larger in percentage terms for small economies that cannot produce many domestic varieties.

Addition 2: economies of scaleBut this demand-side theory is not enough. If that’s the whole story we should observe people consuming very small quantities of a very large number of varieties. But we generally don’t see this: there is a limit to the number of varieties available. Why?

Increasing returns to scale

The second half of this story is increasing returns to scale or economies of scale, which we have not much discussed. We say that a production function has IRS if doubling the inputs more than doubles output. (More generally, any Y% increase in all inputs generates more than Y% in additional output.)

What does this mean in economic terms? The more an industry produces the greater is the gain in additional output. That means the average cost of producing a good falls as you expand output. (At some point average costs could start to rise once you get above some high level of output just because you may run into limits on your machinery or have to pay higher wages or similar factors. But that doesn’t happen in “super increasing returns” industries such as software and building huge jets for airlines.)

Combined effectsWhy is this important for explaining IIT? Because differentiated manufacturing goods (and many services) are generally subject to IRS, at least up to some level of output.

This means within a country it only makes sense to manufacture a limited range of varieties (think cars or refrigerators or furniture), but each at some high scale to make costs low. How many varieties? It depends on the size of the market: small countries would produce very few (and at high cost), large countries would produce more (and at low cost). That is,

With IRS the number of varieties that can be produced in each country is limited by the size of the market.

So we might have 3 car producers in Italy, 2 in France, 4 in Germany, 3 in the US, 4 in Japan, etc. Each produces a limited set of models in autarky.

Now what does trade do? It expands the size of the market by permitting trade between countries. But in autarky you can only consume domestic varieties. So trade has these effects:

It expands the number of varieties available (a welfare benefit).

It permits firms to produce more of each variety, getting larger economies of scale (lower costs, also a benefit).

These are important GAINS FROM TRADE beyond the standard specialization (CA) and lower import prices for consumers.

Combined effectsBut don’t the growing imports raise competition for domestic firms and varieties?

Yes, but let’s get some perspective:

A. This is IIT (2-way trade) so as more varieties come in the home country also has new export markets.

B. It is likely that some firms (and varieties) will end up producing more for the integrated market and taking advantage of economies of scale. These are likely to be the most productive and innovative firms. (This “scale efficiency” raises GDP and is counted as a gain from trade.) It is also likely that some firms (and varieties) will lose out in this competitionand close down. These are likely to be the least productive and innovative firms. Economists also think of this as a good thing (“rationalization efficiency”) though there is room for debate.

C. But consider: imagine 3 countries in autarky, each producing 10 varieties of a good. In autarky consumers in each have 10 varieties. Now the countries trade with each other, and 4 firms close down in each because they are not productive. Now consumers have 6x3 = 18 varieties to choose from.

Trade economists see these tradeoffs as largely beneficial for countries that trade. But one exception: if you are a small and remote country, trade with efficient and larger countries could drive nearly all of your manufacturers and their varieties out of the market. Your country would end up specializing in and exporting, say, primary goods. And it would be difficult to develop domestic versions of goods with economies of scale. If you think producing manufacturing goods is important for dynamic reasons, this could be a concern. So this is sometimes seen as a reason why small countries may gain less from trade than large ones.

IIT: summary Variety gains in combination with IRS (lower costs) mean that international trade generates

additional gains from trade beyond the competitive models, such as Ricardo and HO. IIT is highest among similar developed countries because they have similar preferences for

differentiated goods and tend to have larger domestic markets that support firms with IRS. This suggests that trade in differentiated goods is different from HO trade. In HO trade countries

export goods that are in industries that intensively use lots of capital or lots of labor (or land). Trade in differentiated goods is intra-industry trade while HO trade is inter-industry trade. We can have both of course. (Example: the US exports wheat and imports clothing (inter-industry) while it both imports and exports cars and wine (intra-industry).)

An important observation: with HO (inter-industry) trade, opening to trade makes the scarce factors (unskilled labor in the US) worse off and abundant factors (skilled labor and capital in the US) better off, so there is a redistribution of factor income. But with IIT, trade is not about dealing with factor scarcity, it’s about meeting the demand for variety as firms with IRS expand. This means that all factors can gain income from trade (higher wages, lower prices, more variety).

It also means that opposition to imports of differentiated goods (IIT) is far less than opposition to imports of labor-intensive goods.

In the real world trade generates a complex mix of these things.

Clicker questionWhich of the following statements is NOT true about intra-industry trade (IIT)?

A. IIT generates additional gains from trade due to increases in product variety.

B. International trade in differentiated products generally permits more productive firms to produce larger amounts of each variety, generating economies of scale.

C. IIT is largest among developed countries with similar income levels.

D. IIT tends to reduce the real incomes of scarce factors and raise the real incomes of abundant factors.

Addition 3: Product cycles in trade (and foreign direct investment)

The product cycle is a very simple idea but with important implications.

The theory is that each product goes through a typical life cycle from newly innovated to fully standardized. Along the way it changes from being highly skill-intensive and engineering-intensive to unskilled-labor-intensive, which means that the location of comparative advantage changes over time.

But since there is always new innovation going on there are always new product cycles starting in new goods.

This notion is very different from basic static Ricardian theory (unchanging technology differences) and HO theory (unchanging endowment differences). The PC model is inherently dynamic, involving innovation, technology transfer, and trade.

Product cycleConsider a product such as a laptop, tablet computer, automobile brakes, or an advanced machine tool (a machine used to produce other goods in a factory). Each can be expected to go through the following stages of the PC. This is pretty loose theory and details will be different for each type of good.

Stage 1 (Innovative good): the product is innovated in one of the skill-abundant, engineering-abundant economies such as US, Germany, UK or Japan. Why there? Because of the skills and also because those countries generally have high incomes and demands for new goods. During this time the new good has to go through testing and initial production and it makes sense to locate production in the place where it was innovated (or close, e.g., in Canada for a US innovation). Let’s suppose this is the US.

Trade and FDI flows: Probably none or very little as the good is sold only or primarily to home consumers or industrial users.

Characteristics of the good: new, uncertain in demand, still needs engineering and testing.

Product cycleStage 2 (New good): Consumers and users in other developed countries learn about the good and demand it, so the innovator starts to export.

Trade and FDI flows: Expect exports to go to Canada, Japan, EU, and other developed countries (DCs).

And after some time we can expect US MNEs to invest in local facilities or acquire affiliates in those countries, so production starts moving to these DCs through foreign direct investment.

Characteristics of the good: Still skill-intensive and capital-intensive, demand in DCs is established.

Stage 3 (Maturing good): How to produce the good is better understood and it is possible to move production abroad through FDI or licensing the rights to produce. But in this stage you would locate production in economies that also have high skill and capital supplies since the good remains skill-intensive in production.

You would also do this to make it cheaper to supply the good to local markets (e.g., in Europe or Asia). So-called “platform FDI”.

Demand for the good arises in emerging countries like China and Brazil and Mexico.

Trade and FDI flows: Expect FDI to other DCs who produce it in rising quantities and can even export back to the US. They would also export to emerging countries (ECs). At this stage DCs would export the good to ECs.

Characteristics of the good: Still skill-intensive and capital-intensive but no more engineering is needed.

Product cycleStage 4 (Standardized good): Good is fully standardized and labor-intensive. It becomes cheapest to produce in the ECs (and maybe even poorer developing countries). This would happen through even further FDI (and outsourcing or offshoring). FDI involves technology transfer.

Or firms in those countries might try to imitate the good and produce their own versions. This is an uncompensated form of technology transfer.

Meanwhile, the US and other DCs have moved on to innovating a new version of the good.

Trade and FDI flows: production shifts to ECs, who then begin to export it and continue to export it as long as there is demand for the good. The US and other DCs are net importers of this good.

We can imagine a diagram like the one on the next page depicting a typical product cycle.

Product cycle: important observations Not all goods are like this; the PC really refers to industries where you expect continuous innovation and

technology transfer. Because within an industry different versions of a good (e.g., a tablet or machinery) will exist we can expect IIT

to happen, even between rich and poor countries. Generally, innovative goods and new goods offer high profits to firms because they are new and the firms have

(temporary) monopolies on producing them and can charge high prices. But over time as goods mature and become labor-intensive profits would likely fall.

This means there is an important tradeoff to consider:o The rate of innovation in advanced countries determines how high their profits are in the early stages of a good.o The rate of diffusion or imitation to ECs determines how quickly firms in those countries gain comparative advantage and

produce and export the good. This process raises wages in ECs (like China) but tends to reduce wages in the US and other DCs.

So there is a dynamic “race” of sorts: the higher the rate of innovation the higher are profits and real wages in DCs. The higher the rate of diffusion and imitation the lower are profits and real wages in DCs but the higher are real wages in ECs. Generally, firms in the “North” want high innovation but limited imitation. This is the main reason we see firms in the “North” push for stronger patent and trademark protection in the “South”.

This is an extremely interesting and very current area of research in trade and development economics.