ECON2915 Economic Growth Lecture 12 : Trade Policy and Growth. - - PowerPoint PPT Presentation

ECON2915 Economic Growth

Lecture 12 : Trade Policy and Growth. Andreas Moxnes

University of Oslo

Fall 2016

1 / 30

Infant-industry protection

Definition

Infant-industry protection: Protection of an industry in which a country does not currently have a comparative advantage, in the hopes that it will thereby gain a comparative advantage in it over time. U.S. textiles 1800s, South Korea, Taiwan, Japan, 1970s-1980s, Argentina, U.S. 2017− → ? IIP is a key feature of Import-Substituting Industrialization (ISI):

◮ A strategy of aggressive import restrictions in order to jump-start

industrial growth.

◮ Still controversial, and criticized by many economists.

We’ll explore the consequences of IIP theoretically and empirically.

◮ Can governments use IIP to jump-start growth? 2 / 30

A case study : Argentina, Singapore and Hong Kong

Argentina:

◮ IIP from the 1940s. Particular sectors were permanently favored with

tariffs > 200%.

Singapore:

◮ Aggressive IIP in 1960s, 70s, and 80s. Targeted “frontier” industries

with temporary tariffs.

◮ After growth in those industries started, the government would drop

protection and move on to new industries.

◮ Not just trade: ⋆ Tax incentives: If a firm was given ‘pioneer status,’ it would be exempt

from profit tax for several years.

⋆ Forced savings: 15% of earnings would be channeled to investment and

development projects.

⋆ Directed credit. Cheap loans for targeted industries. ⋆ Incentives for FDI. Tax breaks, cheap real-estate, etc. Huge inflow of

FDI resulted. 1990: 35.9% of Singaporean GDP went to payments to foreign workers or companies.

Hong Kong:

◮ Always liberalized. 3 / 30

Growth

1960: SG and HK very poor. Real per capita GDP in 1985 dollars: HK: $2,323, SG: $2,409 1960-85: real per capita GDP growth per annum: HK: 6.09% SG: 6.03%.

4 / 30

What role did trade policy play?

But did IIP help at all for Singapore? Would it have had even better performance with free trade? Is there any case to be made for IIP at all? We’ll look at the theory.

5 / 30

Learning by doing

Definition

Learning by doing: Increases in productivity in a particular production activity that are realized through experience in that activity. “Country has a comparative disadvantage in industry X now because it has no experience in it. If people get a chance to work at it, after a few years they’ll have the skills required and the country will have a comparative advantage in industry X.”

6 / 30

Learning by doing

For example, in 1980 Singaporeans had no experience making computer hard drives. Government incentives, protection: Production started up. Rapid LBD. By 1983 Singapore was the largest exporter of disk drives in the world. (Young, p. 27.)

7 / 30

Learning by doing

But does this improve national welfare? Crucial point: If there is no externality or other market failure, LBD in and of itself is not a reason for government intervention (Baldwin, 1969). Intuition: If firms/workers know they will improve productivity by X% in disk drive production, and still choose not to do it (in the absence

• f a tariff), then the tariff cannot be welfare improving.

8 / 30

An example

Consider a worker who can sew together 1,000 pairs of blue jeans per month. Piece rate per pair: 10¢. On the other hand, she can go to work making radios for $2.50 a piece. This takes time to get right: New workers mess up the radios quite a lot.

9 / 30

An example

Goal is to make 50 radios per month. Newbies mess them all up in the first month. Afterward, with t months’ experience, the fraction of defective radios is: Frt =

• 1

1+g t , where g > 0 is growth of expertise. Therefore, in month t, the worker will make Yt = 50(1−Frt) usable radios.

10 / 30

Present discounted value

PDV is the current worth of a future sum of money or stream of cash flow given a specified rate of return. PDV of $100 received 5 years from now: amount to invest today to get $100 5 years from now: $100 = PDV ×(1+r)t ⇐ ⇒ PDV = $100

• 1

1+r t If receiving $100 every year then PDV =

∞

∑

t=0

$100

• 1

1+r t Use the infinite geometric series formula ∑∞

t=0 xt = 1/(1−x) (for

x < 1) to get PDV = $100 1 1−

1 1+r

= $1001+r r .

11 / 30

An example

Suppose the worker faces an interest rate of r (e.g. 0.04). Then her PDV of income from the radio job is: PDV R = $2.50

∞

∑

t=0

Yt

• 1

1+r t PDV in jeans production is PDV J = $0.10

∞

∑

t=0

1000

• 1

1+r t A worker will choose the industry with the highest PDV .

◮ Depends on g and r. 12 / 30

Some observations

If g is high enough that the worker will choose the radio job, then IIP is not needed. If g is low enough that the worker will choose the blue-jeans job, then the economy won’t make any radios, despite the potential long-run comparative advantage in them. A tariff τ on radios:

◮ Will raise the piece rate from $2.50 to $2.50×(1+τ) (and the

domestic radio price).

◮ PDV R ↑ −

→ Possible that PDV R > PDV J so workers move into that industry.

◮ IIP “success” in terms of changing the stucture of the economy. 13 / 30

Welfare

But is welfare higher under IIP than under free trade? NO. What matters for real purchasing power is income at world prices, not tariff-altered domestic prices. Imagine that the country would produce jeans under free trade and radios under IIP: PDV J > PDF R. Purchasing power (in terms of radios)

◮ Free trade:

W FT = PDV J pR where pR is the world market price of radios.

◮ IIP:

W IIP = PDV R (1+τ) pR (1+τ) = PDV R pR

But if PDF J > PDV R, then W FT > W IIP!

14 / 30

Welfare

Shows that what matters for welfare is output Yt valued at world prices. Government intervention can change the pattern of specialization but this will decrease welfare. LBD is not a reason for government intervention unless there is a market failure.

15 / 30

Possible market failures

A number of market failures may give a potential reason that government intervention may be called for. Examples:

1

Credit market failures.

2

LBD spillovers, or positive externalities.

3

Assymetric information. The goverment knows that LBD will be stronger than what firms/workers expect.

16 / 30

Credit market failures

Perhaps the worker would get a higher PDV from the radio job, but needs a loan during the learning process, until she starts making money. If she can’t get a loan, she may get stuck with the blue-jeans job. Inefficient outcome due to malfunctioning credit market. In this case, IIP might help:

◮ Raise the domestic price of radios. ◮ Therefore raise the piece rate. ◮ Enable the worker to break-even earlier so she doesn’t need the loan.

Trade policy as imperfect substitute for fixing the credit market.

17 / 30

Credit market failures

Of course, the best thing is to fix the credit market. For example, in the past a lot of Third World governments have put a ceiling on interest rates.

◮ Supply < demand of credit (credit-market rationing). ◮ Often called “financial repression.”

In this case, IIP can help at the margin, but: The first-best policy is to get rid of the financial repression!

18 / 30

Agglomeration externalities (LBD spillovers)

Now, suppose that every new worker in the radio industry raises the productivity of everyone else. Thus, the more workers there are in the industry, the faster they all learn and the more productive they all become.

◮ External increasing returns to scale. 19 / 30

Mechanisms

How can larger industries become more productive?

◮ Easy access (low price, on short notice) to suppliers and subsuppliers. ◮ Allows firms to become more specialized. ◮ Labor market pooling.

Many emprically important cases:

◮ Silicon Valley. ◮ Shenzhen, China: The world’s manufacturing center. 20 / 30

China, 1992

21 / 30

China, 2010

22 / 30

Agglomeration externalities

In the presence of externalities, the market returns to working in the radio industry will typically be too low, because the market does not capture the full returns to working there. Therefore, the market system generally won’t provide the optimal level

• f employment in the industry.

23 / 30

An example (Ethier, 1982)

Suppose we’ve got 1,200 workers in total. Suppose that 1 worker can produce 1,000 pairs of blue jeans per month. 1 worker can also produce (50 × LR/1,000) radios per month, where LR is the number of workers making radios in the country. I.e., my productivity in making radios depends on the number of other people also making radios, because I can learn from them.

24 / 30

PPF

Bc of external increasing returns, PPF is convex: Slope is the OC of radio production QR. OC is decreasing in QR

25 / 30

Autarky equilibrium

Suppose that in autarky, 800 workers produce radios. Then income is

◮ PJ ×1,000 from blue jeans. ◮ PR× (50 × 800/1,000) = 40PR.

Relative price: In equilibrium, the worker is indifferent between which industry she works in − → income the same. PR PJ = 1000 40 = 25. Output:

◮ QJ = 1,000×(1200−800) = 400,000. ◮ QR = 40×800 = 32,000. 26 / 30

PPF

Budget line must: Pass through production point A. Have slope PR/PJ = 25.

27 / 30

Autarky and trade

Budget line (BL) not tangent to the PPF.

◮ Slope of PPF < BL −

→ The social OC of radio production < the private cost.

◮ Bc your actions have positive impact on others.

Consumption in point A.

◮ Worker could achieve higher utility if moving down the PPF. ◮ The economy is not producing enough radios. 28 / 30

Trade

Suppose that with trade, fewer workers produce radios (e.g. foreign captures market share). More jeans production. Bad for welfare because we get even less output of the good with the positive externality. Trade barriers (IIP) can counteract this. But again, better to directly target the sector with the positive externality.

29 / 30

Summing up

You can justify IIP as a growth strategy, but:

◮ It must be based on a market failure.

And even then – it’s treacherous terrain.

◮ Need to know which sectors have/have not externality etc.

Empirical studies on IIP find very small negative/positive effects on welfare (Head, 1994, Irwin, 2000). More generally, limited empirical evidence that trade restrictions are good/bad for growth.

◮ Measurement problems. ◮ Countries package trade policy with other policies. 30 / 30