Globalization, Growth, and Trade

Lectures 13-14: Specific Factors Model (SFM)

1

Overview - TodayMotivation: Bringing ‘structure’ of economy into

trade discussion, a quick look at global export shares and comparative advantage

Overview of the Specific Factors Model (SFM)Analytical pieces of the SFM

Production function Production possibility frontierProduction function and implied factor returns4 quadrant model (labor, 2 production functions, possibility

frontier)

Trade, production and factor payments in the SFMAnalysis: winners and losers from global market

shocks

2

3

Primary product export shares

Latin Am: 50% avg., but 85% for Andean countries60% of Mercosur inc. Bolivia and Chile; Over 65%: Argentina, Belize, Bolivia, Chile, Colombia, Nicaragua,

Panama, Paraguay, Peru, Uruguay, Venezuela

Under 40% - Costa Rica and MexicoComparative advantage of most of LA is clear

S.S. Africa: most countries are mainly primary exporters4 countries with <70% (Togo, Senegal, South Africa, Mauritius)6 with 70-80% share (Guinea, Kenya, Madagascar, Niger, Zambia,

Zimbabwe)

The other 25 have > 80% primary export share

4

Comparative Advantage Comparisons

If export shares => comp advantage, then most of L Am & Africa have comparative advantage in primary products.

What implications might this have for development?

Why does it matter to poverty?Recall that Yh = wLh + rKh: how would expanding

primary products shape household incomes? What does your answer depend on?

Why might you be concerned about the poverty implications of comparative advantage in primary products versus manufacturing?

We can get more on this once we dig further into our next trade model

5

Political economy implicationsHO-SS predicts aggregate gains from trade,

but also losses for some groups--> Functional (self-interest) basis for some positions on trade: Owners of abundant factors in favorOwners of scarce factors opposedExamples?

6

Overview of Specific Factors Model (SFM)

2 good model (like H-O)But capital (or natural resource) is specific to

sectors (cannot be moved to other sector)Does this make sense? Can a coffee farm become a clothing factory in short term? In long term?

In SFM, only labor is mobile between sectorsUse SFM to study how changes in trade patterns,

FDI, and technology affect economic structure and incomes when factor-specificity limits adjustment

Helps us to see winners and losers from trade in a slightly different way.

Production function – 1 sectorFactors of productionProduction functionDiminishing returnsVMP and factor payments

8

9

Labor days

Rice (tons)

ƒx(L, K)

• Constant returns to scale in (L,K), so dim. returns to L when quantity of K is fixed

0 10 20 30 40

282725

20

Production Function

10

Labor days

Rice (tons)

ƒx(L, K)

• Constant returns to scale in (L,K), so dim. returns to L when quantity of K is fixed (irrigated paddy)

0 10 20 30 40

282725

20

ƒx(L, K+)

11

Lx0 Lx

X

X0

0

ƒx(K,L)

• Assume: wage = value of labor’s marginal product• Return on labor (= wage) = slope of tangent to function• Return on stock of sector-specific capital is height 0πx = (rx/px)Kx

slope = w/px

X = (w/px)Lx + πx

Revenue = costs

X*px = w*Lx + rx*Kx

or: X = (w/px)Lx + πx

where: πx = (rx/px)*Kx

πx

Calculating Factor Returns

(Derivation of factor returns) Total revenue of the firm: pxX = wL + rxKx

By assumption, the value of output is fully divided between workers and capital owners

Dividing both sides by px:

X = (w/px)L + (r/px)Kx

= (w/px)L + πxNote: w/px is known as the product wage in sector

X

Distribution between L and K: X – (w/px)L = πxHigher wage (steeper slope on w/px) implies lower

profit share. Flatter slope implies higher profit share12

The specific factors model

Assume 2 goods, X and MEach sector uses specific capital, Kx, Km

--> prod’n fns yj = ƒj(L, Kj), j = X, MLabor is ‘mobile’ (can be reallocated) between

X and M productionTotal labor force is fixed and fully employed:

L = Lx + Lm

In equilibrium, same wage offered in both sectorsFor given Kx and Km, when labor is fully employed,

can only increase output (create jobs) in one industry by reducing output (destroying jobs) in the other

13

14

ƒm(Lm,Km)

ƒx(Lx,Kx)

L 50

50 L

M

X

Labor constraint

Prod’n Poss. Frontier, Maps total production possible given PFs and labor

45o

0

General Equilibrium – Supply Side

Production function x

Production function M

15

L

L

M

X

LA

45o

MA

XA

uA

pA

Lm

Lx

0• •

•

•

Autarky (no trade)

16

L

L

M

XXT

MT

LT

p*

uT

0

LA

MA

XA

uA

pA

Lm

Lx

• •

•

•

•

•

•

•

•

From Autarky to Tradep* > pA

MT < MA

LMT < LM

A

… uT = uA

LT = LA

Trade, income, distribution in SFM

Integration with world economy raises aggregate real income & cons. welfare

Structure of production and labor allocation change in predictable ways

What happens to returns to specific factors? (hint: Stolper-Samuelson - see notes from Week 1)

What happens to the real wage?

17

18

L

L

M

X

45o

LA

YT

LT

uA

pA

pT

uT

0

Compare old and new incomes at constant prices!

YA

Aggregate Income

(Aggregate income change)

YT = aggregate income from production of the combination (XT, MT) valued at world prices pT, measured in terms of good X (the value of X that could be bought with that much income)

Compare: YA = aggregate income from production of the combination (XA, MA), valued at world prices pT, measured in terms of good X

YT > YA says the economy is better off in aggregate

19

20

L

L

M

X

45o

LA

LT

pA

pT

0

MA

MT

XA

XT

slope = (w/pM)A

(w/pM)T

(Changes in factor payments)Moving from autarky to trade raises X output and

employment, lowers M output and employmentDemand for KX rises; πX

T > πXA

Demand for KM falls; πMT > πM

A

Demand for L in M falls; with KM fixed , law of diminishing returns says that productivity of remaining workers rises, so (w/pM)A < (w/pM)T

Demand for L in X rises; with KX fixed, (w/pX)A > (w/pX)T

Are workers better off or worse off?

21

22

Capitalists (Km)

Landowners (Kx)

Workers

(L)

Effect of rise in px on nominal inc.

Effect of rise in px on real inc.

(a) When consume mostly X

(b) When consume mostly M

Comparative Advantage in Agriculture

23

Capitalists (Km)

Landowners (Kx)

Workers

(L)

Effect of rise in px on nominal inc.

lose gain gain

Effect of rise in px on real inc.

(a) When consume mostly X

lose gain lose?

(b) When consume mostly M

lose gain gain?

Comparative Advantage in Agriculture

24

Capitalists (Kx)

Landowners (Km)

Workers

(L)

Effect of rise in px on nom. income

Effect of rise in px on real inc.

(a) When consume mostly X

(b) When consume mostly M

Comparative Advantage in Manufacturing

25

Capitalists (Kx)

Landowners (Km)

Workers

(L)

Effect of rise in px on nom. income

gain lose gain

Effect of rise in px on real inc.

(a) When consume mostly X

gain lose lose?

(b) When consume mostly M

gain lose gain?

Comparative Advantage in Manufacturing

Distributional & poverty effects

Real specific factor returns follow own prices: for a rise in px/pm, πx will rise, πm will fall

Real wage change is indeterminate:Wage rises rel. to pm, but falls rel. to px

H’hold welfare: aggregate has risen, but

Gains for owners of capital in XLosses for owners of capital in MWorkers’ welfare change is ambiguous

26

DiscussionIf we have data on asset ownership & cons.

patterns, can compute changes in Rh and poverty for groupsPoverty effects depend on distribution of assets as

well as on changes in payments such as wages and rents

Notice that we have assumed labor is mobile between sectors. Realistic? What if it is not?

SFM vs H-O: which is more realistic? When?

What about more complex models, for example with some endogenous product prices (nontradables?)See next class…

27