trade
dodo1995
Lecture-3tadeweek3.pptx3 Why Do Americans Get Their Impalas from Canada?
International Trade
John McLaren
3.1 Impalas on the Horizon 3.2 Increasing Returns More Generally 3.3 How to Tackle Europe: Trade versus FDI 3.4 On a Smaller Scale: Trade and Increasing Returns in Furniture 3.5 Adding Heterogeneity: The Melitz Effect
Consider the Chevrolet Impala.
Classic American car.
Some Chevrolet facts:
All Impalas are now made at Oshawa, Ontario.
Roughly 200, 000 imported into the US per year.
All Cobalts are made in Lordstown, Ohio (similar but smaller Chevrolet).
Why?
Comparative advantage?
Requires Canada to have comparative advantage in Impalas, US to have one in Cobalts.
Let’s list some reasons a country might have a comparative advantage in something, and see if that is reasonable.
The design and know-how originated in the US, so it could be produced in the US (technology is not an issue).
Canada is certainly is not a low-wage country; wages, education, infrastructure, and standards of living are very similar to what they are in the US.
Therefore, it is hardly plausible to argue that Canadians simply have a comparative advantage in producing Impalas, while Americans have a comparative advantage in producing Cobalt.
Comparative advantage doesn’t seem to explain US imports of Impalas.
Alternative: Two factors:
I. Canada-US Auto Pact of 1965
II. Increasing returns to scale
I. Canada-US Auto Pact of 1965
Before 1965, both countries had high tariffs on imported cars and auto parts.
Result: Big 3 automakers mostly produced in Canada for Canadian market.
Little trade; exports from Canada virtually zero.
Duplicated assembly lines on both sides of the border.
Production on smaller scale in Canada: High cost per unit.
Pact provided for free trade between US and Canada.
(Some exceptions: e.g., fire engines.)
Some restrictions applied:
US side: Minimum Canadian content required for duty-free status.
Canadian side: Employment at plants in Canada needed to keep on increasing.
Upshot:
Free trade in autos and auto parts between US and Canada.
But Detroit had to maintain the same level of production in Canada.
Agreement was grandfathered into CUFTA (1988) and NAFTA (1994).
For convenience, U.S. exports to Canada are shown as positive values whereas, U.S. imports from Canada are shown as negative values. This implies that the more positive (negative) the value is, the larger are the exports (imports).
II. Increasing returns to scale.
Setting up and maintaining an assembly line for one model requires huge fixed costs.
E.g., even just to maintain the machines in line to be able to produce 1 car per month.
Induces increasing returns to scale (IRS).
An industry exhibits IRS if and only if an x% increase in all inputs increases the output by more than x%. Equivalently, (assuming that factor markets are competitive) an industry has IRS if an x% increase in output increases cost by less than x%, thus lowering average cost.
Fixed costs and IRS.
Suppose cost of producing Q Impalas at Oshawa is equal to C(Q) = F + waQ.
Then cost of producing 2Q Impalas at Oshawa is equal to F + 2waQ.
LESS THAN TWICE.
Essence of IRS: Double output, less than double cost.
Implications for geographic allocation of production:
Need to maintain same level of production in Canada
But can save on costs by reducing number of models produced in each country.
Suppose that GM needs to produce 11 models.
Each model:
200,000 units for the US market,
20,000 units for the Canadian market.
Assume that in either country, production costs are given by F + waQ for each model, where F is the fixed cost, a is the unit labor requirement for each car, w is the wage and Q is the total number of cars produced.
Pre-Auto Pact.
11 models produced in Canada; 20,000 units each.
Cost for each model = F + wa20,000.
11 models produce in US; 200,000 units each.
Cost for each model = F + wa200,000.
Total cost for GM: 22F + wa11x220,000 = 22F + wa2,420,000.
Post Auto Pact.
Now, GM can concentrate production of each model in one location.
Produce 1 model in Canada, 10 in the US.
Produce 220,000 units at each plant.
Now costs are equal to 11x(F + wa220,000) = 11F + wa2,420,000.
Thus, GM has saved 11F.
Same number of each type of car produced, but costs are lower.
Note: Before, there was no trade in cars.
Now, every car is traded.
200,000 cars exported from Canada to the US.
200,000 cars exported from the US to Canada.
Intra-industry trade.
The point:
IRS provides a reason for trade, by creating an incentive to concentrate production of each product in one location.
Kinds of IRS.
Internal IRS.
Double the firm’s inputs and more than double that firm’s output.
Can result from fixed cost.
Can also result from learning by doing.
Implies a downward-sloping AC curve.
Note: Incompatible with perfect competition.
To see this: Assume Total Cost (TC) = F + cQ, where F is fixed cost, c is marginal cost and Q is the total quantity produced. Because of the presence of the fixed cost, average cost (AC) is always larger than marginal cost (MC):
AC = TC/Q = F/Q + c and MC = dTC/dQ = c
So, AC > MC.
Because average cost is always greater than marginal cost, it is not possible to have a perfectly competitive equilibrium (since price = marginal cost implies negative profits).
External, national IRS.
Double all inputs employed by all firms in an industry in one country, and more than double total industry output.
Can result from learning-by-doing spillovers.
Can also result from shared infrastructure improvements.
External, international IRS.
Case when external, national IRS are not present, but if you double the inputs of all producers in the industry worldwide it will more than double worldwide output.
IRS sheds light on a number of additional topics in trade:
I. How to tackle a foreign market.
II. Role of monopolistic competition in trade.
III. Intra-industry trade.
Tackling a foreign market.
You are the CEO of GM.
You want to break into the European market.
Two options: Produce here and export, or produce over there.
Producing in Europe: FDI Option.
Set up a plant in Spain.
Fixed cost: F.
After that, each unit requires a* units of labor.
Each unit of labor costs w*.
Choose P to maximize (P - a*w*)Q(P) - F.
This yields maximum profit from FDI option.
Producing here: Export option.
No additional fixed cost. (Important.)
Each unit requires a units of labor.
Labor costs w per unit.
Transport cost of k(d) per car, where d is distance to market.
Tariff of t per car.
Choose P to maximize (P-wa-k(d)-t)Q(P).
Compare the two.
FDI option is more likely to be attractive if:
w*a* is small;
t is high;
d is high;
F is small.
Brainard (1997) study examined US exports and foreign sales of foreign affiliates of US multinationals.
Key variable: Export share of total foreign sales of each industry.
Found pretty good evidence for the last three of these points.
(Didn’t really look at the first one; that’s harder.)
How does GM actually serve Europe?
Mostly through OPEL subsidiary (FDI option).
Cars are made in Europe.
BUT: Each model is made in only one location in Europe (e.g., Zaragoza, Spain).
Thus, intercontinentally -- FDI option; within Europe, export option.
Monopolistic competition in trade.
IRS matters not only for giant firms like GM.
Pervasive in manufacturing, including small-scale manufacturing.
E.g., furniture.
A pretty good description of this type of industry: Monopolistic competition.
Example: Baronet and Thos. Moser
Two medium-sized firms.
Each has a tiny share of the total furniture market.
Baronet is Canadian; Moser is American.
Fixed cost from production and design.
Distinctive styles.
Baronet Java dining set.
Thos. Moser Hawthorne dining set.
Key features of monopolistic competition:
Large number of firms; each small compared to the whole market.
Each produces a unique product; hence, monopoly power.
Free entry, hence zero profits in equilibrium.
Baronet’s demand curve conditional on the number
of other furniture firms in the industry
If more firms to enter, at a given price Baronet would
lose some customers to them, so its demand will shift
to the left. If some firms shut down, its demand will
shift to the right.
Since in equilibrium, P* = AC given by the tangency condition, it must necessarily occur at the downward sloping portion of the AC curve.
This means in equilibrium, we cannot reach the minimum point of the AC curve (also named as minimum efficient scale).
This can be interpreted loosely as the cost of providing more variety. If the number of firms was reduced by 10% and each firm produced 10% more output, total output will be unchanged and average costs would go down, but there would be less variety available for consumers.
Opening up trade flattens Baronet’s demand curve, so it becomes more sensitive to price changes.
Raising its price will send some of its customers to a US competitor;
Lowering its price will grab some customers from US competitors.
Thus, each firm’s demand is now more elastic.
Thus, Baronet now has an incentive to lower its price and sell more dining sets.
But at the same time all other firms have the same incentive.
All other furniture makers therefore cut their prices and increase their sales, shifting Baronet’s demand curve down.
If industry-wide price cutting goes far enough, firms start to lose money, some will exit until remaining firms just break even.
As a result, each country has fewer furniture makers but each consumer can access to a greater variety of furniture (since they can access to varieties in the other country as well).
Implications:
Plenty of trade even between identical countries.
Once again, trade results from IRS.
Trade is intra-industry.
Trade reduces number of products produced in each country (“shake-out”).
But increases variety available to each consumer.
In addition, trade tends to increase elasticity of demand for each product: P closer to MC.
Intra-industry trade.
Important feature of trade between similar countries where monopolistic competition is important.
Exports of k from country i to country j:
Total trade between i and j:
Net trade, or inter-industry trade,
in industry k:
Net trade as a fraction of total trade:
Intra-industry trade as a fraction of total trade:
Adding heterogeneity: The Melitz effect.
It’s actually a stretch to assume that all furniture makers are equally productive.
Suppose there are high-cost and low-cost producers in the same industry -- heterogeneous firms?
Question was explored in an influential paper by Marc Melitz (2003).
The constant f is a fixed labor requirement, and is the same for all firms.
Therefore, the fixed cost is equal to wf, where w is the wage.
Suppose that to produce q units of output, a firm must hire f + q/Φ units of labor.
The parameter Φ represents the marginal product of labor, which is a constant for each firm, but varies from firm to firm. Thus, Φ is a measure of firm’s productivity. More productive firms have higher values of Φ.
The marginal cost for each firm is equal to w/Φ, where w is the wage.
Suppose that to produce q units of output, a firm must hire f + q/Φ units of labor.
Autarky equilibrium.
Firms enter until the profits for the marginal firm are equal to zero.
Only the most efficient firms enter.
More productive firms (higher Φ) produce more and make higher profits than less productive firms.
Now open up trade.
If a firm wants to export, it must pay an additional fixed cost (e.g., setting up a distribution network).
As a result, only the most productive firms choose to export.
Less productive firms are hit by imports but don’t benefit from exports.
Therefore, less productive firms produce less and have lower profits than before trade; some drop out.
Result: Effects of trade.
More productive firms benefit from less productive firms dropping out; their output and profits go up.
Market share of less productive firms falls; market share of more productive firms rises.
Average productivity of industry therefore rises.
Call this the ‘Melitz effect.’
The outcome is important: Trade causes the most productive firms to export and expand, while less productive firms serve the domestic market and shrink, and the least productive firms drop out entirely.
This all implies that globalization raises productivity for two reasons:
The least productive firms drop out.
Among the surviving firms the market share of the more productive firms rises at the expense of the market share of the less productive firms.
Increasing returns to scale and trade: Main points.
IRS is a reason for trade: Motive to concentrate production of each product in one spot.
Three types of IRS: Internal, external national, and external international.
In serving a foreign market, stronger IRS argues for exporting, but higher tariffs or transport costs argue for local production via FDI.
Many industries have small-scale IRS, many producers, differentiated products: monopolistic competition.
Explains intra-industry trade.
Increasing returns to scale and trade: Main points.
Finally, heterogeneous firms plus IRS yields the Melitz effect: Trade raises industry productivity by increasing market share of more productive firms at the expense of less productive firms.
Where we are.
