How Preferential Trade Agreements Affect the U.S. Economy

CONGRESS OF THE UNITED STATES

CONGRESSIONAL BUDGET OFFICE

CBO

How Preferential

Trade Agreements

Affect the

U.S. Economy

SEPTEMBER 2016

CBO

Note

Numbers in the text may not add up to totals because of rounding.

www.cbo.gov/publication/51924

Contents

CBO

Summary 1

How Does Trade Affect the U.S. Economy? 1

Why Does the United States Establish Preferential Trade Agreements? 1

How Do Preferential Trade Agreements Work? 1

How Much Have Preferential Trade Agreements Affected the U.S. Economy and the

Federal Budget? 2

The Economic Effects of Trade on the United States 2

Productivity 3

Total Employment and Average Wages 3

Output 3

Consumer Spending 3

Outcomes for Workers 4

U.S. Trade Agreements 4

Reasons the United States Establishes Preferential Trade Agreements 4

A Brief History of Trade Agreements 5

Provisions of U.S. Preferential Trade Agreements 7

BOX

1. POTENTIAL U.S. PREFERENTIAL TRADE AGREEMENTS

Effects of Preferential Trade Agreements on U.S. Trade 9

Effects on Total Trade 10

Effects on the Trade Balance 12

Effects of Preferential Trade Agreements on Foreign Direct Investment 13

Indirect Effects of Preferential Trade Agreements on the U.S. Economy 13

Productivity 13

Total Employment and Average Wages 13

Output 14

Consumer Spending 14

Outcomes for Workers 14

Effects of Preferential Trade Agreements on the Federal Budget 15

Appendix: Difficulties in Estimating How Preferential Trade Agreements Affect an Economy 27

About This Document 30

Tables

1. Characteristics of Partner Countries of U.S. Trade Agreements Before the Year of

Implementation 6

2. Estimated Effects of NAFTA on U.S. Trade With Canada and Mexico 11

3. Estimated Effects of NAFTA on Total U.S. Trade 12

CBO

How Preferential Trade Agreements

Affect the U.S. Economy

Summary

Preferential trade agreements (PTAs) are treaties that

remove barriers to trade and set rules for international

commerce between two countries or among a small group

of countries. PTAs directly affect a country’s economy by

altering its flows of trade and investment. Primarily

through trade, PTAs indirectly affect other aspects of a

country’s economy—such as productivity, output, and

employment. As of August 2016, the United States had

established 14 PTAs with 20 of its trading partners. This

report examines the economic literature on trade and

PTAs and summarizes that literature’s findings on how

trade and PTAs have affected the U.S. economy.

How Does Trade Affect the U.S. Economy?

International trade yields several benefits for the U.S.

economy. Trade increases competition between foreign

and domestic producers. That increase in competition

causes the least productive U.S. businesses and industries

to shrink; it also enables the most productive businesses

and industries in the United States to expand to take

advantage of profitable new opportunities to sell abroad

and obtain cost savings from greater economies of scale.

As a result, trade encourages a more efficient allocation of

resources in the economy and raises the average produc-

tivity of businesses and industries in the United States.

Through that increase in productivity, trade can boost

economic output and workers’ average real (inflation-

adjusted) wage. In addition, U.S. consumers and busi-

nesses benefit because trade lowers prices for some goods

and services and increases the variety of products available

for purchase.

Not everyone benefits from trade expansion, however.

Although increases in trade probably do not significantly

affect total employment, trade can affect different work-

ers in different ways. Workers in occupations, businesses,

and industries that expand because of trade may make

more money, whereas workers in occupations, businesses,

and industries that shrink may make less money or expe-

rience longer-than-average unemployment. Such losses

can be temporary or permanent. Nevertheless, economic

theory and historical evidence suggest that the diffuse and

long-term benefits of international trade have outweighed

the concentrated short-term costs. That conclusion has

consistently received strong support from the economics

profession.

Why Does the United States Establish

Preferential Trade Agreements?

The United States establishes preferential trade agreements

for economic and noneconomic reasons. Those agree-

ments enable the United States and its partner countries

to realize the economic benefits of increased trade and

investment. In addition, the agreements sometimes har-

monize laws and regulations which, among other effects,

make the costs of operating businesses in other countries

more similar to those costs in the United States. An

important noneconomic reason for establishing PTAs is

to achieve foreign policy goals. Those goals include sup-

porting the economies of U.S. allies and promoting the

adoption of preferred domestic policies, such as environ-

mental conservation or stronger workers’ rights.

How Do Preferential Trade Agreements Work?

Preferential trade agreements facilitate trade and invest-

ment among member countries. To encourage member

2 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY SEPTEMBER 2016

CBO

countries to trade, PTAs reduce or eliminate barriers to

trade, such as import tariffs (taxes that countries impose

on foreign-made goods), restrictions on trade in services,

and other commercial rules that impede the flow of trade.

In addition, PTAs facilitate investment among member

countries by easing regulations on foreign investment and

providing improved legal protections for foreign investors.

Preferential trade agreements also set commercial rules

that, among other effects, narrow differences in the costs

of operations among member countries. For example,

some PTAs establish minimum labor and environmental

standards and protections for intellectual property. If the

costs of compliance are high, those types of rule-based

reforms can impede trade and investment flows, making

some businesses less competitive in foreign markets.

To ensure that member countries comply with the provi-

sions of an agreement, PTAs establish dispute resolution

mechanisms. Those mechanisms can take two forms:

One provides a legal platform for countries to make

claims against other member countries; the other enables

investors in member countries to make claims against the

governments of other member countries.

How Much Have Preferential Trade Agreements

Affected the U.S. Economy and the Federal Budget?

In the Congressional Budget Office’s view, the consensus

among economic studies is that PTAs have had relatively

small positive effects on total U.S. trade (exports plus

imports) and, primarily through that channel, on the

U.S. economy. The effects have been small because

the agreements were mostly between the United States

and countries with much smaller economies and because

tariffs and other trade barriers were generally low when

the agreements took effect. PTAs have had little effect on

the U.S. trade balance (exports minus imports) and have

slightly increased flows of foreign direct investment,

mostly by encouraging additional U.S. investment in the

economies of member countries. As a result, the indirect

effects of PTAs on productivity, output, and employment

in the United States have also been small and positive.

Empirical estimates support that view. But those esti-

mates are uncertain and may be understated, because the

effects of nontariff provisions are hard to measure and

because issues with data keep researchers from identifying

how PTAs affect the service sector.

The effect of PTAs on the federal budget is unclear. In

assessing the budgetary impact of previous preferential

trade agreements, CBO’s cost estimates have indicated

that they would slightly lower the amount of federal reve-

nues received from tariffs. However, those results did not

consider how the macroeconomic effects of PTAs might

alter the federal budget. Nevertheless, the small size of the

effects on output suggests that the overall budgetary

effects have also been small.

The Economic Effects of Trade on the

United States

International trade contributes to the overall economic

well-being of people in the United States in many ways.

Without trade, people in the United States can consume

only those goods and services that U.S. businesses can

produce; some goods and services are unavailable, and the

prices of others that those businesses are ill-suited to pro-

duce are relatively high. By increasing competition in the

markets for tradable goods and services, increases in trade

tend to raise U.S. productivity. Although trade expansion

can reallocate workers across occupations, businesses, and

industries, it has little effect on total employment in the

long term. Higher productivity as a result of increases in

trade tends to boost workers’ average real wages and out-

put.

With additional trade, people can consume goods

and services at a lower cost. Lower prices for traded goods

and services enable consumers to buy more goods and

services—and a greater variety of them—with the same

amount of income. Those same effects can also be seen in

other countries that trade globally.

The benefits and costs of trade expansion are not evenly

distributed: The costs are concentrated among less pro-

ductive workers and businesses in all industries exposed

to greater competition from U.S. trading partners. For

example, laid-off workers in occupations that require less

skill or in industries facing greater competition will expe-

rience spells of unemployment while they search for and

train for new jobs. In a flexible and growing economy,

most workers would be able to take advantage of a greater

demand for workers in expanding sectors. However, in

sectors, regions, and occupations with particularly weak

demand for labor, workers could be unemployed for a

long time. Some workers might permanently lose income

or even stop working altogether.

Productivity

Trade can increase a country’s productivity by reallocating

resources to the most productive businesses within

SEPTEMBER 2016 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY 3

CBO

industries and by increasing the productivity of individ-

ual businesses.

The reallocation of resources occurs in

both import-oriented and export-oriented industries.

Within all of those industries, increased competition

from abroad pushes the least productive businesses to

close.

In contrast, profitable new opportunities to sell

abroad encourage the most productive businesses to

expand and export.

That expansion also allows those

businesses to take advantage of economies of scale and

become even more productive. As a result, the average

productivity of businesses within each industry rises,

boosting the nation’s overall productivity.

Over the past

few decades, productivity in the U.S. manufacturing sec-

tor has improved substantially as competition from trade

increased.

Increased competition from trade encourages businesses

to invest more in physical capital, improved production

processes, and innovative techniques to remain profitable.

Closer trade relationships can promote the transmission

of information and new technologies across borders.

(Some investments arise from such exchanges, particularly

outside of the United States. Businesses’ greater produc-

tivity from such investments may have contributed to the

growth of overall productivity in global economies, espe-

cially for countries with less-developed economies.)

Trade can also increase a country’s productivity by reallo-

cating resources to the industries within that country that

can produce goods and services more cheaply than those

in its trading partners. Trade encourages labor and capital

resources to flow between industries to best take advan-

tage of a country’s natural resources, labor force, stock of

physical capital, and technical knowledge. If resources are

shifted from industries that are less productive to industries

that are more productive, then overall productivity rises.

Total Employment and Average Wages

Trade expansion can have different effects on labor mar-

kets in the short term and the long term. In the short

term, the competitive pressures from increased trade real-

locate productive labor resources. Therefore, total

employment can fall below its maximum sustainable

level, at least temporarily, as workers who lose their job

search for a new job.

In the long term, total employ-

ment should return to its maximum sustainable level as

those displaced workers find jobs in expanding export-

oriented industries and in other industries. However, the

maximum sustainable level of employment may be

greater or smaller than it was before trade expanded.

Insofar as greater trade raises average real wages, the max-

imum sustainable level of employment may be greater

than before the trade expansion because changes in wages

affect people’s desire to work. In addition, to the extent

that displaced workers cannot find new jobs over the long

term, total employment may be smaller. Although it is

difficult to estimate how trade affects total employment,

the prevailing consensus in the literature is that, over the

long term, trade expansion has had little effect on total

U.S. employment and has raised workers’ average real

wages.

Output

In the short term, the effects of trade expansion on out-

put can be difficult to discern. That is because the effects

from disruptions surrounding the reallocation of resources

may be greater than the effects from increases in produc-

tivity. But in the long term, by increasing competition

and domestic productivity, trade generally boosts real

economic output. Evidence from economic studies gen-

erally supports the claim that trade has contributed to the

growth of economic output in the United States over

the long term.

Consumer Spending

Trade benefits consumers mainly by lowering the prices of

some traded goods and services, which raises consumers’

purchasing power.

When countries trade, consumers pay

less for goods produced more cheaply abroad. But domes-

tic consumers pay more for the goods and services that

their country exports (because of higher demand from

foreign consumers), partially offsetting that effect. In

addition, domestic productivity growth and competition

between businesses put downward pressure on prices.

Since the mid-1990s, relatively low rates of inflation in

the prices of imports as a result of trade have increased

consumers’ purchasing power around the world.

Fur-

thermore, low-income consumers in the United States

have benefited more from trade than high-income con-

sumers, because low-income consumers tend to spend a

larger portion of their income on imported goods whose

prices are more likely to fall as a result of trade—

specifically, food and clothing.

4 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY SEPTEMBER 2016

CBO

Trade also benefits consumers by increasing the variety

and the availability of products.

Trade is an important

channel for expanding the number of products available

for purchase by domestic businesses and consumers.

That greater variety improves consumer welfare, although

measuring the precise extent to which consumers benefit

from more variety is difficult.

Outcomes for Workers

Trade affects different workers in different ways. Greater

trade benefits some workers, such as those employed in

export-oriented industries or highly skilled occupations.

But it harms other workers, who tend to lose income

when they are displaced from their jobs for trade-related

reasons. Estimating how many workers have lost or

gained jobs from trade is hard because of the challenge of

distinguishing the effects of trade from other factors that

influence labor markets.

Some workers directly benefit from trade. As demand for

labor rises because of trade in export-oriented or fast-

growing industries and businesses, relative wages rise—

benefiting those workers.

In the United States, another

group of workers who tend to see their wages rise as a

consequence of trade are highly skilled and more-educated

workers whose services are in higher demand.

Similarly,

those who work in occupations that require fewer routine

or easily automated tasks are likely to see their wages rise

as a result of trade.

In contrast, trade can be costly for workers who become

unemployed when their occupations, businesses, or

industries shrink. In the United States, trade-displaced

workers tend to work in industries subject to more com-

petition from imports (such as the textile industry), rela-

tively less productive businesses, or occupations involving

easily automated or routine tasks (such as data entry or

customer service). Those workers also are typically less

educated, older, or longer tenured at their previous posi-

tions.

Most trade-displaced workers eventually find

other jobs, but they are unemployed longer than workers

who lose their job for other reasons.

In addition, trade-

displaced workers tend to earn markedly less once

reemployed.

Such workers who switch occupations

when reemployed can lose even more income.

Even

though most displaced workers eventually find new

employment, some workers displaced as a result of trade

cannot, and they eventually stop working.

Attributing a lost or gained job to any single factor is dif-

ficult; for that reason, researchers have trouble determin-

ing how many jobs have been lost or gained from trade.

Nevertheless, several studies have tried to estimate how

many workers have been affected by trade. In one study,

the Organisation for Economic Co-operation and Devel-

opment examined all available U.S. data but could not

precisely estimate the total number of U.S. workers dis-

placed for trade-related reasons. The analysis suggested

that although only a small share of job losses in the

United States between 1997 and 2003 could be confi-

dently attributed to trade-related factors, it is possible

that increased competition from foreign firms was an

important factor in a much larger share of all job dis-

placements in the United States over those years.

Another study found that growth in U.S. imports from

China from 1999 to 2011 may have led to significant

U.S. job losses—accounting for nearly 10 percent of the

decline in U.S. employment in the manufacturing sector

over that period.

Researchers also have been unable to

precisely estimate how many jobs have been created or

how many workers have seen their wages rise as a result of

international trade.

U.S. Trade Agreements

Trade agreements are treaties that stimulate trade and

cross-border investment and set international commercial

rules and standards. Since World War II, the major trade

agreements have focused on lowering import tariffs and

removing quotas on imported goods. Those large, older

agreements—which had many participating countries—

are known as multilateral trade agreements. In recent

decades, however, the scope of U.S. trade agreements has

narrowed to include fewer countries but has broadened to

address other barriers to trade and investment. Those

smaller, more recent agreements are known as preferen-

tial trade agreements. Recent PTAs have addressed other

issues affecting international commerce, such as product

standards and intellectual property rights.

Reasons the United States Establishes

Preferential Trade Agreements

The United States establishes PTAs for economic and

noneconomic reasons. By increasing trade flows, PTAs

SEPTEMBER 2016 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY 5

CBO

can help the United States and other member countries

realize many economic benefits. In addition, trade agree-

ments can be used to achieve foreign policy goals.

Economic Reasons. The United States establishes PTAs

to realize the economic benefits of increased trade and

investment. PTAs lower barriers to trade, such as tariffs

and quotas, among participating nations, and they can

increase cost efficiencies by specifying and unifying

commercial rules and investment codes. In addition, the

United States can use PTAs to promote the spread of free

trade among all countries. As those agreements proliferate,

countries not party to them will have greater incentives to

either join existing agreements or negotiate multilateral

agreements. Conceivably, the largest PTAs could expand

or merge to eventually include all countries.

The United States establishes PTAs that try to harmonize

laws among member countries which, among other effects

mentioned below, can make the costs of operation in

other member countries more similar to the costs of oper-

ation in the United States. One way in which the agree-

ments do that is by strengthening labor and environmental

standards in countries where they are weak. Imports from

those countries tend to become more expensive as a result,

whereas exports from countries with stronger standards,

like the United States, tend to become more competitive

in the markets of other member countries.

Noneconomic Reasons. The main noneconomic reason

that the United States establishes PTAs is to achieve for-

eign policy objectives. By establishing such agreements,

governments can support the economies of allied

nations—bolstering the economic relationship between

nations to possibly yield stronger political ties. PTAs also

can serve as a valuable bargaining chip to attain foreign

policy goals. The United States can offer PTAs to partner

countries in exchange for adopting favored domestic pol-

icies—such as environmental conservation or stronger

rights for workers. Furthermore, the United States can

use the promise of becoming part of a PTA to obtain

concessions or reforms from nonmember countries.

A Brief History of Trade Agreements

Several major agreements involving many countries

helped expand trade from the end of World War II

through the mid-1990s. Since then, only agreements

between smaller numbers of countries have been ratified.

Agreements Involving Many Countries. In 1948, the

United States and 22 other countries established the

General Agreement on Tariffs and Trade (GATT) to for-

mulate and enforce rules that govern international trade.

A major focus of GATT has been to reduce barriers to

trade on a nondiscriminatory basis. Before GATT, tariffs

among countries were high, averaging between 20 per-

cent and 30 percent.

By 1994, GATT had expanded to

include 128 member countries, and most tariffs rarely

exceeded 10 percent, especially among developed nations.

Trade liberalization under GATT is widely referred to as

multilateral liberalization primarily because of the many

countries participating in that agreement. A crucial ele-

ment of GATT is the concept of most-favored nation

(MFN) status. Under GATT, all members agree to grant

one another MFN status, meaning that any favorable

trade terms extended to the imports from any GATT

member country must be extended to imports from all

GATT member countries. GATT has been an important

factor behind the substantial increase in global trade and

has contributed to global economic growth since its

inception.

In 1994, the Uruguay Round of GATT negotiations estab-

lished the World Trade Organization (WTO) to serve as a

permanent forum for negotiating further trade liberaliza-

tion and to monitor members’ compliance. Because

GATT reduced tariffs and eliminated quotas, multilateral

trade negotiations under the WTO expanded to include

rule-based reforms. Those reforms address legal issues such

as investment protections, market access for trade in ser-

vices, intellectual property rights, and labor and environ-

mental standards. The WTO instituted those reforms to

further facilitate trade and investment among its members

by creating standardized rules for businesses operating in

one another’s countries. For example, in 1995, the WTO

established minimum protections for intellectual property

rights for member countries through the Agreement on

Trade-Related Aspects of Intellectual Property Rights

(commonly known as TRIPS). In addition, WTO negotia-

tions in 1995 addressed liberalization of trade in services

through the Global Agreement on Trade in Services.

Perhaps most significantly, the Uruguay Round established

a dispute-settlement mechanism that enables aggrieved

countries to place punitive tariffs on the imports of

countries judged to have violated the WTO’s rules.

6 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY SEPTEMBER 2016

CBO

Table 1.

Characteristics of Partner Countries of U.S. Trade Agreements Before the Year of Implementation

Percent

Source: Congressional Budget Office.

Data on average trade-weighted MFN import tariff rates come from the United Nations Conference on Trade and Development (UNCTAD) Trade Analysis

Information System (TRAINS). The database contains historical data on tariff rates. For agreements with more than one partner country, the average

among all partner countries is calculated by weighting each partner country’s average tariff rate by the relative volume of U.S. exports to each country.

Data on U.S. exports to partner countries that are used to calculate those weights come from the U.S. Census Bureau. Data on GDP come from the World

Bank’s World Development Indicators database. GDP data are in constant 2010 U.S. dollars. Dollar figures for GDP are converted from domestic

currencies by using official exchange rates in 2010. Data on agreement implementation dates come from U.S. Customs and Border Protection.

CAFTA-DR = Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, and the Dominican Republic; GDP = gross domestic product; MFN = most-favored

nation; NAFTA = North American Free Trade Agreement; n.a. = not applicable; * = between zero and 0.05 percent.

a. Includes only trade in goods; does not include trade in services.

b. Data for average trade-weighted MFN import tariffs for Canada are only available starting in 1989.

c. Includes Canada and Mexico. In 1994, NAFTA superseded the 1989 agreement between the United States and Canada.

d. CAFTA-DR was implemented by the United States, El Salvador, Guatemala, Honduras, and Nicaragua in 2006. Implementation of the agreement by

the Dominican Republic and Costa Rica did not occur until 2007 and 2009, respectively. This table presents data for CAFTA-DR countries in 2005, the

year before the agreement’s initial implementation.

In the late 1990s, multilateral trade negotiations under the

WTO began to stall. The Doha Round of negotiations

began in 2001 but still has not concluded, even after sev-

eral attempts to reach agreement. Global trade talks have

proven particularly difficult in this round of negotiations

because rule-based reforms tend to be more complicated

and contentious than simply eliminating tariffs on indus-

trial goods. For example, the WTO’s negotiations to

increase intellectual property protections have failed in

part because less-developed countries are reluctant or

unable to offer additional protections.

Agreements Involving Several Countries. Trade liberaliza-

tion over the past few decades has shifted toward preferen-

tial trade agreements among a few countries spread over a

broad area.

PTAs are a less effective method of increasing

global trade flows than multilateral trade liberalization

because, by nature, they include fewer countries—but that

smaller scope has also made it possible to conclude nego-

tiations. The United States has established more than a

dozen PTAs in the past three decades (see Table 1). The

largest PTA—as measured by the sum of members’ real

gross domestic product, or GDP—is the North American

Partner Countries

Year in Which

Agreement Was

Implemented

Israel 1985 0.7 n.a. n.a. n.a.

Canada 1989 9.7 19.9

8.2

n.a.

NAFTA

1994 16.3 28.0 n.a. 4.7

Jordan 2001 0.1 * 18.9 2.8

Australia 2004 5.3 1.0 4.0 2.8

Chile 2004 0.9 0.3 6.0 2.8

Singapore 2004 0.9 1.6 * 2.8

CAFTA-DR

2005 0.9 1.5 6.6 2.5

Bahrain 2006 0.1 * 5.6 2.4

Morocco 2006 0.5 * 19.9 2.4

Oman 2006 0.2 * 4.7 2.4

Peru 2007 0.6 0.3 6.8 2.4

Colombia 2012 1.4 1.0 9.1 2.6

Panama 2012 0.2 0.3 6.8 2.6

South Korea 2012 8.2 2.7 7.3 2.6

Cumulative GDP of

Partner Countries

(Percentage of

U.S. GDP)

Share of

Total U.S. Trade

Trade-Weighted

Average MFN Import

Tariff Rates of

Partner Countries

Trade-Weighted

U.S. Average MFN

Import Tariff Rates

SEPTEMBER 2016 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY 7

CBO

Free Trade Agreement (NAFTA), an agreement among

the United States, Mexico, and Canada that took effect

in 1994. Apart from NAFTA (and its predecessor, the

Canada–U.S. Free Trade Agreement between the United

States and Canada), the United States belongs to 13 PTAs

with 18 other countries.

Currently, the United States is pursuing two major PTAs:

one with selected Pacific Rim countries (the Trans-Pacific

Partnership) and one with member countries of the Euro-

pean Union (the Transatlantic Trade and Investment

Partnership). For a discussion of those potential PTAs,

see Box 1.

Provisions of U.S. Preferential Trade Agreements

PTAs generally include three types of provisions. The

first type facilitates trade and investment by lowering

import tariffs, easing quotas, and improving market

access for service providers and investors. The second

type sets commercial rules and standards. Although that

type of provision also often stimulates trade and cross-

border investment, its costs of compliance (if they are

high) can impede trade among partners. The third type

of provision establishes legal mechanisms through which

parties can resolve disputes to ensure that partners satisfy

the terms of the agreement.

Provisions That Facilitate Trade and Investment. Some

provisions directly reduce the costs of trade or investment,

and other provisions offer partner markets preferential

access to goods or services. More recent agreements have

included provisions that facilitate trade and investment

by loosening regulations on services and cross-border

investments. Among the topics that PTAs have addressed

are the following:

B Import Tariffs. PTAs lower or eliminate taxes on

foreign-made goods among trading partners, making

it less costly for those countries to trade.

B Tariff Rate Import Quotas. Under GATT, limits on

how many goods can be imported from abroad were

largely eliminated. Instead, those traditional import

quotas have been replaced by tariff rate import quotas,

which limit how many goods another country may

import at a low (or zero) tariff rate; all imported goods

in excess of that threshold are subject to a higher rate.

PTAs often loosen or remove those tariff rate quotas

on imported goods from member countries.

B Customs Procedures. PTAs have streamlined customs

procedures for imports from member countries.

B Government Procurement. PTAs have sometimes

eliminated rules that obligate governments of member

countries to buy goods and services from domestic

businesses.

B Market Access for Services. Some countries protect their

domestic service sectors from international competition

by prohibiting foreign businesses from selling those

services in local markets. PTAs can loosen those

restrictions by granting service-sector businesses in

member countries improved access to the local

markets of other member countries.

B Foreign Direct Investment. Countries commonly

restrict foreign ownership of domestic businesses in

certain industries and regulate whether foreign

businesses can open local branches or facilities. To

encourage flows of foreign direct investment (FDI)

among member countries, PTAs sometimes include

provisions that relax those rules.

B Investor Protections. PTAs establish rules to prevent

discrimination against foreign businesses that operate

in domestic markets. By granting foreign businesses

what is known as national treatment, those provisions

reduce the risks that foreign investors face when

operating in member countries.

Provisions That Set Commercial Rules and Standards.

Some provisions that set commercial rules and standards

can either stimulate or impede flows of trade and invest-

ment. They can impede flows if the compliance costs of

the provisions make some businesses less competitive in

foreign markets. Those types of provisions address the

following issues:

B Intellectual Property Rights. Some PTAs establish rules

that determine how long international holders of

copyrights, patents, and trademarks may sell their

products exclusively in member countries. Those

provisions typically grant the intellectual property of

foreign businesses stronger protections from domestic

competition than were previously afforded.

8 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY SEPTEMBER 2016

CBO

Box 1.

Potential U.S. Preferential Trade Agreements

Both the Trans-Pacific Partnership (TPP) and the Transatlantic Trade and Investment Partnership (TTIP) are

intended to be comprehensive preferential trade agreements. As such, they lower tariffs and establish rule-

based reforms among member countries. Those reforms would, for example, harmonize labor and environ-

mental standards, protect intellectual property, and loosen regulations on international investment. The rule-

based reforms are central to both agreements and are likely to generate the largest economic effects for the

United States. Negotiations on the TPP concluded on October 5, 2015, but the President has not yet submitted

the agreement to the Congress for ratification. Terms of the TTIP are still being negotiated.

Trade-Weighted Average Tariff Rates on Imported Goods for

Countries Negotiating the TPP and TTIP, 2013

Percent

Sources: Congressional Budget Office; World Trade Organization.

Tariffs are taxes paid on the value of imported goods. The trade-weighted average tariff is equal to the average of the most-

favored-nation (MFN) import tariff rates set by a country, weighted by the volume of imports into each country (including both

agricultural and nonagricultural products). More weight is given to products with larger import flows. MFN tariffs are the standard

rates charged on imports from all members of the World Trade Organization, excluding preferential rates, or lower rates charged

within quotas. The agricultural tariff is equal to the trade-weighted average of all MFN tariffs on all agricultural products. The

nonagricultural tariff is equal to the trade-weighted average of all MFN tariffs on all nonagricultural products.

Tariff data are from 2013, the latest year for which such data are available.

TPP = Trans-Pacific Partnership; TTIP = Transatlantic Trade and Investment Partnership.

a. The European Union comprises Austria, Belgium, Bulgaria, Croatia, Republic of Cyprus, Czech Republic, Denmark, Estonia,

Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Poland, Portugal,

Romania, Slovakia, Slovenia, Spain, Sweden, and the United Kingdom.

Continued

B Labor and Environmental Standards. PTAs have

included provisions that establish minimum labor and

environmental standards. Such provisions can improve

the competitive advantage of businesses that already

meet higher standards because they will face lower

compliance costs.

B Phytosanitary and Sanitary Standards. PTAs can

establish rules for both plant (phytosanitary) and

animal (sanitary) products.

B Data Transfer and Data Localization Rules. More

recent PTAs have tried to define rules regarding the

cross-border transfer of personal data—specifically,

how and where foreign businesses can store data on

their customers and employees.

B State-Owned Enterprises. PTAs have included rules

intended to prevent governments from favoring

state-owned enterprises; those provisions prohibit state-

owned businesses from gaining competitive advantages

over businesses from other member countries.

Country

Australia 2.5 4.2 4.1

Brunei 0.1 1.7 1.5

Canada 14.2 2.2 3.1

Chile 6.0 5.9 5.9

European Union

22.3 2.3 3.6

Japan 12.8 1.2 2.1

Malaysia 14.0 3.6 4.4

Mexico 26.6 3.3 5.0

New Zealand 2.3 2.3 2.3

Peru 1.7 1.8 1.8

Singapore 15.7 0 0.5

United States 4.1 2.1 2.2

Vietnam 7.3 4.9 5.1

Agricultural Nonagricultural Overall

SEPTEMBER 2016 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY 9

CBO

Box 1. Continued

Potential U.S. Preferential Trade Agreements

The TPP is an agreement among 12 Asian and American nations (including the United States). It aims to reduce

trade barriers and establish international standards and regulations for various economic and commercial

issues.

The United States already has preferential trade agreements with about half of the TPP’s member

countries—but not with Japan, its fourth-largest trading partner.

If the TPP is ratified, Japan’s inclusion would

give U.S. exporters better access to a large and wealthy consumer market, which represents the TPP’s most

significant advantage for the United States. However, because trade-weighted average tariffs among the

member countries are already so low (see the table), the rule-based reforms—not tariff reductions—are

the mechanism through which the TPP would have its largest effect on the flow of trade among member

countries. Those potential reforms cover many issues, including labor standards, intellectual property

protections, regulations for state-owned enterprises, cross-border investment rules, and market access for

foreign businesses in service industries.

The TTIP is a potential agreement between the United States and the European Union. Much like the TPP, the

TTIP is expected to increase trade and economic integration as a result of rule-based reforms.

For example,

unifying product safety standards across TTIP countries would reduce the costs that potential exporters face in

trying to comply with disparate sets of product safety codes. Because average import tariffs between the

United States and the European Union are already low, the lower tariffs stipulated in the TTIP would have less

of an effect on the U.S. economy than the other terms of the agreement.

1. TPP member countries are Australia, Brunei, Canada, Chile, Japan, Malaysia, Mexico, New Zealand, Peru, Singapore, the United

States, and Vietnam. South Korea has expressed an interest in joining the agreement after its ratification. For an economic

analysis of the TPP, see United States International Trade Commission, Trans-Pacific Partnership Agreement: Likely Impact on

the U.S. Economy and on Specific Industry Sectors, Publication 4607 (May 2016), www.usitc.gov/publications/332/pub4607.pdf

(7.57 MB). Other researchers have also conducted analyses of the agreement. For example, see Peter A. Petri and Michael G.

Plummer, The Economic Effects of the Trans-Pacific Partnership: New Estimates, Working Paper 16-2 (Peterson Institute for

International Economics, January 2016), http://piie.com/publications/wp/wp16-2.pdf (2.43 MB).

2. In addition to the North American Free Trade Agreement (the agreement that the United States has with its two largest trading

partners, Canada and Mexico), the United States has separate free trade agreements with four other TPP countries—Australia,

Chile, Peru, and Singapore.

3. The countries of the European Union trade heavily with the United States. In 2014, the sum of trade flows from all European

Union countries to the United States accounted for nearly 16 percent of all U.S. trade. By contrast, trade flows from Japan to the

United States accounted for only 5 percent.

Provisions That Establish Dispute-Resolution

Mechanisms. PTAs also contain provisions for resolving

disputes among member countries. Those provisions

include the following:

B State–State Dispute Settlement. PTAs typically provide

a legal platform for countries to make claims against

other member countries for breaching a trade

agreement.

B Investor–State Dispute Settlement. PTAs have included

provisions that provide a mechanism through which

investors can make claims against the governments of

other member countries for breaching a trade

agreement.

Effects of Preferential Trade Agreements on

U.S. Trade

In all likelihood, most preferential trade agreements that

have involved the United States have boosted total U.S.

trade, but by relatively small amounts. The effects have

been small because the agreements generally were with

countries with much smaller economies than that of the

United States and because the trade barriers before estab-

lishment of those PTAs were already low. Even NAFTA,

the largest U.S. trade agreement (by share of U.S. trade)

to date, probably only modestly affected total U.S. trade.

Nevertheless, estimates of the effects of NAFTA and

other PTAs are subject to considerable uncertainty, espe-

cially because they omit the effects of certain aspects of

the agreements.

10 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY SEPTEMBER 2016

CBO

The evidence also indicates that PTAs have not signifi-

cantly affected the U.S. trade balance, because the small

increases in exports and imports that have occurred have

probably been similar in size.

Effects on Total Trade

PTAs can change total trade for a member country

directly by creating new trade among all member coun-

tries and indirectly by diverting trade from countries out-

side the agreement. When trade barriers fall, exports and

imports among member countries tend to rise because

the goods and services that the agreement covers become

cheaper and more readily available.

That change in

trade sometimes comes at the expense of trade with coun-

tries outside the agreement—for example, imports from a

partner country replace some goods previously imported

from a country outside a trade agreement—a phenome-

non known as trade diversion. As a result, the change in

total trade among an agreement’s member countries may

be larger than the change in total trade among all trading

partners.

The potential for a trade agreement to stimulate addi-

tional trade among member countries depends on many

factors. Two of the most important factors are the relative

size of the member countries and the extent to which

PTAs reduce barriers to trade. Establishing a trade agree-

ment with a much larger country (or a group of countries

whose combined size is large) gives domestic businesses

preferential access to a much larger market for their

exports. For a relatively small country, that market access

can significantly boost that country’s total trade flows. By

contrast, for a large country that establishes an agreement

with a smaller country, the additional trade that the

agreement creates is unlikely to be substantial in compar-

ison with that country’s existing trade flows. In addition,

the greater the degree to which PTAs reduce barriers to

trade among member countries, the greater the potential

to increase total trade flows.

The effects of PTAs other than NAFTA on total trade

have probably been insignificant, but such estimates are

subject to considerable uncertainty and may be under-

stated. Various issues prevent the models that researchers

use from precisely estimating how PTAs affect trade flows

(for more information, see the appendix). Those stylized

and econometric models may systematically understate

the effect of PTAs on trade flows, for two reasons. First,

although most models of trade are well-suited to analyze

the effect of tariff reductions, only to various degrees do

those models account for how nontariff barriers affect

trade flows. Second, because of modeling difficulties and

data issues, some of those studies do not fully capture the

effect of PTAs on trade in services—a group of industries

for which the United States has a comparative advantage

over most of its trading partners.

Agreements Other Than NAFTA. All of the preferential

trade agreements affecting the United States, except

NAFTA, have not had much potential to boost total U.S.

trade.

That is because the United States has established

PTAs mostly with significantly smaller countries or

groups of countries (see Table 1 on page 6). In addition,

trade barriers between the United States and its partner

countries were already low. When partner countries

signed agreements with the United States, average tariff

rates on their exports to the United States never exceeded

5 percent, and (with the exception of Morocco and

Jordan) the average tariff rates on U.S. exports to those

countries never exceeded 10 percent.

Economic studies find that PTAs by the United States

have raised total U.S. trade by small amounts, although

magnitudes vary by agreement. For PTAs other than

NAFTA, the increases in total trade for the United States

have probably been insignificant (in relation to the exist-

ing volume of U.S. trade). In general, they have reflected

small increases in trade with member countries, which

have been partially offset by decreases in trade with other

countries. Those findings come from the few studies

completed before the establishment of those agreements

that used stylized models of world trade to predict the

probable effects of proposed agreements.

NAFTA. The likely increases in overall trade from NAFTA

are significant but still small. The findings for NAFTA are

based on analyses done before NAFTA’s establishment

and empirical assessments completed afterward.

Empirical assessments suggest that NAFTA substantially

increased total trade flows between the United States and

Canada, and the United States and Mexico, although the

estimates of those effects vary greatly. A 2015 study

reported the strongest estimates for the effects of NAFTA

SEPTEMBER 2016 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY 11

CBO

Table 2.

Estimated Effects of NAFTA on U.S. Trade With Canada and Mexico

Percent

Source: Congressional Budget Office, using foreign trade data from the Census Bureau.

The studies cited in this table are Lorenzo Caliendo and Fernando Parro, “Estimates of the Trade and Welfare Effects of NAFTA,” Review of Economic

Studies, vol. 82, no. 1 (January 2015), pp. 1–44, http://dx.doi.org/10.1093/restud/rdu035; Maureen Rimmer and Peter Dixon, “Identifying the Effects of

NAFTA on the U.S. Economy Between 1992 and 1998: A Decomposition Analysis,” Global Trade Analysis Project (April 2015), http://tinyurl.com/gtap4657;

and John Romalis, “NAFTA’s and CUSFTA’s Impact on International Trade,” Review of Economics and Statistics, vol. 89, no. 3 (August 2007), pp. 416–

435, http://dx.doi.org/10.1162/rest.89.3.416.

a. Includes only trade in goods and does not calculate changes to imports and exports separately.

on those total trade flows.

The results suggested that

NAFTA accounted for about 34 percent of the observed

growth in those trade flows over the 1992–1998 period

(see Table 2). Two other studies found more modest

effects.

All those assessments, though, suggest that

NAFTA probably affected U.S. trade with Mexico more

than it affected U.S. trade with Canada, primarily

because more barriers to trade existed between the United

States and Mexico than between the United States and

Canada when NAFTA began.

Results from the broad literature on all trade agreements

also imply that NAFTA resulted in a large increase in

U.S. trade with Canada and Mexico, although the exact

amount of the increase is uncertain. In a 2016 report

summarizing the literature, the International Trade

Study

Period

Examined Trading Partner

Caliendo and Parro (2015) 1994–2005 Canada 6.1 144.2 4.2

Mexico 105.9 302.4 35.0

Combined 32.4 185.9 17.4

Rimmer and Dixon (2015) 1992–1998 Canada 4.8 67.8 7.0

Mexico 143.9 240.9 59.7

Combined 41.4 113.3 36.5

Caliendo and Parro (2015) 1994–2005 Canada 10.5 104.8 10.0

Mexico 127.8 180.1 71.0

Combined 44.9 126.9 35.4

Rimmer and Dixon (2015) 1992–1998 Canada 16.9 63.4 26.6

Mexico 27.9 77.6 35.9

Combined 20.3 67.8 29.9

Romalis (2007)

1994–2000 Canada -0.3 62.5 -0.5

Mexico 23.2 154.6 15.0

Combined 6.0 88.1 6.9

Caliendo and Parro (2015) 1994–2005 Canada 8.2 125.7 6.5

Mexico 117.0 240.6 48.6

Combined 38.4 157.7 24.4

Rimmer and Dixon (2015) 1992–1998 Canada 10.6 65.7 16.1

Mexico 81.8 153.5 53.3

Combined 30.9 90.8 34.1

Growth of Trade

Attributable to NAFTA Total Growth

Share of Growth

ttributable to NAFTA

(Percentage points)

U.S. Imports From Partner Countr

U.S. Exports to Partner Country

Total U.S. Trade With Partner Countr

12 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY SEPTEMBER 2016

CBO

Table 3.

Estimated Effects of NAFTA on Total U.S. Trade

Percent

Source: Congressional Budget Office, using foreign trade data from the Census Bureau.

The studies cited in this table are Lorenzo Caliendo and Fernando Parro, “Estimates of the Trade and Welfare Effects of NAFTA,” Review of Economic

Studies, vol. 82, no. 1 (January 2015), pp. 1–44, http://dx.doi.org/10.1093/restud/rdu035; Maureen Rimmer and Peter Dixon, “Identifying the Effects of

NAFTA on the U.S. Economy Between 1992 and 1998: A Decomposition Analysis,” Global Trade Analysis Project (April 2015), http://tinyurl.com/gtap4657;

and John Romalis, “NAFTA’s and CUSFTA’s Impact on International Trade,” Review of Economics and Statistics, vol. 89, no. 3 (August 2007), pp. 416–

435, http://dx.doi.org/10.1162/rest.89.3.416.

a. Includes only trade in goods.

Commission found that PTAs “across the world have led

to an increase of 30 percent to 114 percent in each part-

ner’s trade over a 10 year period after an agreement has

entered into force.”

Those estimates are varied and uncer-

tain because of the difficulty in distinguishing the effects of

PTAs from other factors that may have influenced trade

flows over that time.

Taking into account both the trade created among

NAFTA members and the trade diverted from other

countries, evidence suggests that NAFTA probably

increased total U.S. trade slightly. Considering both

those effects, analyses done before NAFTA began pre-

dicted that the agreement would slightly increase total

U.S. trade (relative to the size of total U.S. trade with all

countries).

Later empirical studies corroborated those

predictions (see Table 3). According to one study,

NAFTA was responsible for an estimated 7.5 percent of

the growth in total U.S. trade between 1992 and 1998.

Over the longer period from 1994 (NAFTA’s inception)

to 2005, NAFTA’s contribution to the growth of total

U.S. trade was 7.9 percent, according to a second study.

A third study attributed only 2.8 percent of the growth of

total U.S. trade between 1994 and 2000 to NAFTA.

Effects on the Trade Balance

Estimates of the effects of preferential trade agreements on

the U.S. trade balance are very small and highly uncertain.

By one estimate, NAFTA increased the U.S. trade balance

by an average of 0.1 percent of GDP from 1992 to 2000.

Analyses of other preferential trade agreements find mixed

results—some agreements have improved the U.S. trade

balance and others have caused it to deteriorate—but in

all cases, the estimated effect is small.

The reason is that

economic factors besides PTAs (such as national saving

and investment trends) primarily determine the balance

of trade.

Those mixed results reflect the difficulty in estimating how

PTAs affect the balance of trade.

PTAs directly affect

the trade balance through their influence on exports and

imports; they indirectly affect the trade balance through

changes in prices of goods and services, exchange rates,

and interest rates.

Different ways of modeling those

indirect effects can produce different estimates. In addi-

tion, many models that estimate the effects of PTAs

exclude trade in services (a sector for which the United

States has historically maintained a trade surplus). As a

result, those models probably overstate the negative effect

of PTAs on the U.S. balance of trade.

In any event, a country’s trade balance does not have to

rise for a country to benefit from a PTA. The benefits of

lower prices and greater variety and availability of goods,

for example, may outweigh the additional financing costs

of a larger trade deficit.

Study

Period Examined

Romalis (2007)

1994–2000 1.7 61.7 2.8

Caliendo and Parro (2015) 1994–2005 10.7 136.3 7.9

Rimmer and Dixon (2015) 1992–1998 4.6 62.2 7.5

Growth of

Total U.S. Trade

Attributable to NAFTA

Growth of

Total U.S. Trade

Share of Growth

Attributable to NAFTA

(Percentage points)

SEPTEMBER 2016 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY 13

CBO

Effects of Preferential Trade Agreements on

Foreign Direct Investment

Although preferential trade agreements can affect flows of

foreign direct investment in several ways, past agreements

have primarily encouraged flows from more-developed to

less-developed countries.

Three provisions of PTAs can directly affect FDI flows:

B Lower Trade Barriers. When barriers fall, exporting

becomes less costly than direct investment, and FDI

between member countries might decline as a result.

Lower barriers can also affect FDI flows from countries

outside an agreement. For example, because of NAFTA,

FDI flows into Mexico grew significantly as a result of

foreign businesses’ setting up affiliates in Mexico to

take advantage of the preferential treatment of imports

into the United States from Mexico.

B Stronger Protections for Foreign Investors. Investment

provisions can prevent discrimination against foreign

businesses that operate in domestic markets—giving

foreign investors national treatment.

For example,

agreements have made it harder for governments to

expropriate foreign investments, and they have relaxed

rules that prevent foreign businesses in certain

industries from opening factories or branches.

B Stronger Intellectual Property Protections. Such

protections include longer durations for copyright and

patent exclusivity.

Countries with stronger intellectual

property protections tend to have higher investment

from abroad. Changes to such protections in past

agreements have not been strong enough to

significantly affect FDI.

Preferential trade agreements also can affect FDI by sig-

naling to foreign investors that a country, particularly

one with a developing or emerging-market economy, is

committed to domestic reforms. For example, Mexico’s

willingness to give FDI better protections under NAFTA

signaled to investors the government’s willingness to pro-

tect all international investments. For that reason, inves-

tors from countries outside the United States and Canada

probably were more willing to invest in Mexico as a result

of that country’s commitment to reforms under NAFTA.

PTAs have probably increased flows of foreign direct

investment from the United States to other countries

and, to a lesser extent, from other countries into the

United States. That conclusion relies on results from

studies that examine how all PTAs affect FDI flows,

because economists have had trouble estimating how

preferential trade agreements affect U.S. flows of FDI.

Studies suggest that although PTAs increase investment

flows between similarly developed countries, a much

stronger stimulus of FDI flows is evident from developed

countries to developing and emerging-market ones.

addition, PTAs probably have increased the amount of

FDI received by the United States. However, no studies

of the United States specifically (to CBO’s knowledge)

have found evidence that PTAs have encouraged FDI

flows into the United States, probably because it is diffi-

cult to distinguish the effects of PTAs from other factors

that affect those flows.

Indirect Effects of Preferential Trade

Agreements on the U.S. Economy

In CBO’s view, the consensus among economic studies is

that the small increases in total trade that have resulted

from PTAs have yielded modest, but generally positive,

indirect effects on the U.S. economy, increasing produc-

tivity, average wages, output, and consumer spending

slightly. Nevertheless, those agreements have not had

uniformly positive effects. Because PTAs encourage eco-

nomic specialization, some workers and sectors have fared

better than others.

Productivity

By boosting total trade between the United States and

its partner countries, PTAs have probably promoted a

more efficient allocation of U.S. resources and increased

U.S. productivity in the same ways that trade in general

affects productivity (as discussed above), although the

size of those effects is uncertain. There is some evidence

that productivity improved in countries that participate

in trade agreements.

However, CBO was not able to

find any studies that specifically estimate how much

PTAs have increased U.S. productivity.

Total Employment and Average Wages

Most economic evidence suggests that the total number

of workers directly affected by PTAs has been too small

to significantly affect labor market conditions nationwide.

14 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY SEPTEMBER 2016

CBO

Most of that evidence comes from studies of NAFTA, the

agreement with the greatest potential to affect U.S.

employment. Those studies concluded that NAFTA’s

effects on the size of the labor market and net changes in

total U.S. employment each year have been small.

Those findings are consistent with the economic theory

that PTAs should have little long-term effect on total

employment because all displaced workers would eventu-

ally find new employment or would have stopped work-

ing anyway. However, according to some estimates,

NAFTA contributed to many lost jobs.

Conversely, many U.S. workers have had some small

benefits as a result of PTAs. By lowering consumer prices

(primarily through their effects on prices of imported

goods) and increasing the productivity of workers (from

greater competition), those agreements have probably

increased average real wages for U.S. workers, albeit only

slightly.

If that slight increase occurred, it would have

induced more people to work, increasing the U.S. labor

supply to a small degree. To CBO’s knowledge, there is

no evidence of such an effect on the labor supply,

although if it had occurred it would have been small and

extremely difficult to detect.

Output

PTAs have slightly increased the total goods and services

produced in the United States and have altered the com-

position of production across industries. By increasing

competition and domestic productivity, PTAs can boost

real economic output. As a result of the establishment of

NAFTA, for example, annual GDP in the United States

probably increased by a few hundredths of a percent,

according to a CBO analysis from 2003.

In addition, a

2003 study by the International Trade Commission sur-

veyed the literature and found that U.S. GDP probably

increased by between 0.1 percent and 0.5 percent as a

result of NAFTA.

The International Trade Commis-

sion and

other researchers have estimated smaller effects

on real U.S. output, stemming mostly from higher pro-

ductivity, from the implementation of other trade agree-

ments.

By lowering trade barriers in the United States and other

member countries, PTAs have slightly altered the compo-

sition of U.S. economic output. Typically, U.S. sectors

that experience the largest gains in output after imple-

mentation of a trade agreement are those for which the

agreement has significantly lowered trade barriers in part-

ner countries. For example, research suggests that NAFTA

led to output gains for U.S. agricultural exporters as a

result of Mexico’s agreeing to reduce strong import

protections on corn and corn products. But other studies

indicate that U.S. production of textiles and steel was

lower as a result of NAFTA, because the United States

removed import quotas on those products from Canada

and Mexico. PTAs’ effects on the composition of output

are specific to each agreement and are generally small.

Therefore, it is difficult to draw conclusions about how

PTAs in general have altered the composition of U.S.

economic production.

Consumer Spending

By increasing consumers’ purchasing power (through

lower consumer prices) and income earned domestically,

PTAs probably increased total real consumption in the

United States. According to some estimates, NAFTA

boosted U.S. consumption by 0.4 percent between 1992

and 1998.

NAFTA also increased the variety and avail-

ability of Mexican products for U.S. consumers by offer-

ing Mexican producers wider access to U.S. markets.

Outcomes for Workers

Preferential trade agreements have hurt some U.S. work-

ers (sometimes substantially) and helped others. Workers

in low-skilled occupations or in manufacturing industries

have typically been harmed the most; those who lost their

job usually endured the most substantial hardships.

Many of those displaced workers experienced a costly

transition to a new job, and most faced lower lifetime

earnings as a result of that displacement.

Other dis-

placed workers could not find a good match in a new job

and stopped working. In addition, increased competition

resulting from PTAs has stifled wage growth in certain

occupations and industries, affecting even those workers

who kept their job. For example, one study found that

between 1990 and 2000, NAFTA decreased the cumula-

tive growth rate of wages of low-skilled workers by 16

percentage points in U.S. textiles and plastics industries

in comparison with the wages of similar workers in

industries with no decrease in tariff rates under

SEPTEMBER 2016 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY 15

CBO

NAFTA.

That study estimated that NAFTA probably

depressed average wages for all workers (even those in the

service sector) in the U.S. localities where businesses were

most sensitive to competition from Mexican imports.

Some U.S. workers have probably directly benefited from

PTAs—particularly workers in occupations, businesses,

and industries that have expanded as a result of those

agreements—although the available empirical evidence

is weak.

According to economic theory and analyses

of other countries’ labor markets, the nominal wages of

those workers should grow faster than the wages of other

workers because their services are in greater demand.

Some empirical evidence supports those theoretical pre-

dictions, but no empirical studies (to the best of CBO’s

knowledge) estimate whether any U.S. workers have seen

their wages rise as a direct result of PTAs.

Effects of Preferential Trade

Agreements on the Federal Budget

How PTAs affect the federal budget is unclear. Past cost

estimates from CBO have indicated that PTAs, once

implemented, would slightly lower federal revenues from

tariffs.

However, those results did not take into account

macroeconomic feedback—the estimated effects on the

federal budget that would arise from changes in eco-

nomic output or other macroeconomic variables.

Most PTAs have led to small reductions in federal tariff

revenue, according to CBO’s estimates. That revenue

falls for two reasons. First, imports from member coun-

tries are subject to lower tariff rates. Even though imports

from those countries increase under trade agreements, the

associated boost in tariff revenue in the United States

from greater volume typically does not offset the effect of

the lower rates. Second, U.S. consumers replace imports

from high-tariff countries with cheaper imports from

countries with lower tariffs. Partially offsetting both of

those effects is an increase in income and payroll taxes

that stems from the lower tariffs (putting aside any effect

on overall economic activity and income).

The direct budgetary effects of PTAs on tariff revenue

have been small in comparison with the total size of fed-

eral tariff revenue, according to past CBO estimates. In

1993, CBO projected that NAFTA would decrease fed-

eral tariff revenue by 2.6 percent (or between $2 billion

and $3 billion) over its first five years, but the budgetary

effects thereafter were unclear.

In 2011, CBO estimated

that the Korea–U.S. Free Trade Agreement would lower

federal tariff revenue by about 1 percent (or just over

$7 billion) between 2012 and 2021.

However, CBO’s past estimates did not consider the ways

in which the macroeconomic effects of PTAs might other-

wise alter federal revenues and expenditures. By increasing

economic productivity, output, and income, PTAs would

probably increase federal tax revenues. The small size of

PTAs’ effects on output suggests that their budgetary

effects would also be small, however, because budgetary

feedback effects are generally a fraction of the effects on

output.

As of yet, the economics literature includes no estimates

of the effect of PTAs on the federal budget that take

macroeconomic feedback into account.

16 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY SEPTEMBER 2016

CBO

1. See, for example, Richard M. Alston, J.R. Kearl, and

Michael B. Vaughan, “Is There a Consensus Among

Economists in the 1990’s?” American Economic

Review, vol. 82, no. 2 (May 1992), pp. 203–209,

www.jstor.org/stable/2117401; and Daniel B. Klein

and Charlotta Stern, “Economists’ Views and

Voting,” Public Choice, vol. 126 (March 2006),

pp. 331–342, www.jstor.org/stable/30026756. More

recently, a 2012 Initiative on Global Markets survey

of prominent economists found that 94 percent of

the respondents agreed with the statement, “Freer

trade improves productive efficiency and offers

consumers better choices, and in the long run these

gains are much larger than any effects on

employment.” See Chicago Booth, “Free Trade”

(Initiative on Global Markets Forum, March 13,

2012), http://tinyurl.com/igm-chicago. Several of the

respondents in that survey noted that society should

do a better job compensating the workers and

businesses that lose income as a result of trade.

2. According to a review of the research on the link

between trade and economic growth, trade has

significantly contributed to economic growth in the

United States over history. See Scott C. Bradford,

Paul L. Grieco, and Gary C. Hufbauer, “The Payoff

to America From Globalisation,” World Economy, vol.

29, no. 7 (July 2006), pp. 893–916, http://dx.doi.org/

10.1111/j.1467-9701.2006.00828.x.

3. Within-industry reallocation is responsible for much

of the productivity gains from trade. Two studies of

the Canada–U.S. Free Trade Agreement found that

such reallocation significantly raised productivity in

the Canadian manufacturing sector. See Daniel

Trefler, “The Long and Short of the Canada–U.S.

Free Trade Agreement,” American Economic Review,

vol. 94, no. 4 (September 2004), pp. 870–895,

http://dx.doi.org/10.1257/0002828042002633; and

Alla Lileeva and Daniel Trefler, “Improved Access to

Foreign Markets Raises Plant-Level Productivity...

for Some Plants,” Quarterly Journal of Economics,

vol. 125, no. 3 (August 2010), pp. 1051–1099,

http://dx.doi.org/10.1162/qjec.2010.125.3.1051.

Another study found that within-industry reallocation

accounted for two-thirds of the productivity

improvements in the Chilean manufacturing sector

between 1979 and 1986. See Nina Pavcnik, “Trade

Liberalization, Exit, and Productivity Improvements:

Evidence From Chilean Plants,” Review of Economic

Studies, vol. 69, no. 1 (January 2002), pp. 245–276,

http://dx.doi.org/10.1111/1467-937X.00205.

4. According to some estimates, import growth may

be important for spurring productivity growth.

For example, see Jakob B. Madsen, “Technology

Spillover Through Trade and TFP Convergence:

135 Years of Evidence for the OECD Countries,”

Journal of International Economics, vol. 72, no. 2

(July 2007), pp. 464–480, http://dx.doi.org/

10.1016/j.jinteco.2006.12.001; and Francisco Alcalá

and Antonio Ciccone, “Trade and Productivity,”

Quarterly Journal of Economics, vol. 119, no. 2

(May 2004), pp. 613–646, http://dx.doi.org/

10.1162/0033553041382139.

5. Not all businesses profit from more open trade.

Differences in productivity drive certain businesses to

become exporters and others to exit the market when

countries open to trade. See Marc J. Melitz, “The

Impact of Trade on Intra-Industry Reallocations and

Aggregate Industry Productivity,” Econometrica,

vol. 71, no. 6 (November 2003), pp. 1695–1725,

www.jstor.org/stable/1555536.

6. One study found that increasing competition from

trade in Taiwan (China) drove the least productive

businesses out of the market. See Bee Yan Aw, Sukkyun

Chung, and Mark J. Roberts, “Productivity and

Turnover in the Export Market: Micro-Level Evidence

From the Republic of Korea and Taiwan (China),”

World Bank Economic Review, vol. 14, no. 1 (January

2000), pp. 65–90, http://dx.doi.org/10.1093/wber/

14.1.65. Another paper found that through increased

competition, trade stim

ulated a

reallocation of resources

from less productive businesses to more productive

businesses. See Andrew B. Bernard and J. Bradford

Jensen, “Exporting and Productivity in the USA,”

Oxford Review of Economic Policy, vol. 20, no. 3

(Autumn 2004), pp. 343–357, http://dx.doi.org/

10.1093/oxrep/grh020.

7. Some studies present evidence for a possible third

channel: reallocation across industries. One study

found that type of reallocation from specialization

accounted for a substantial proportion of the increase

in productivity in Switzerland between 1997 and

2009. See Ulf Lewrick, Lukas Mohler, and Rolf Weder,

When Firms and Industries Matter: Understanding the

Sources of Productivity Growth, Working Paper 469

(Bank for International Settlements, October 2014),

SEPTEMBER 2016 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY 17

CBO

www.bis.org/publ/work469.htm. However, another

study found only interindustry reallocation as a result

of trade. See Romain Wacziarg and Jessica Seddon

Wallack, “Trade Liberalization and Intersectoral

Labor Movements,” Journal of International Economics,

vol. 64, no. 2 (2004), pp. 411–439, http://dx.doi.org/

10.1016/j.jinteco.2003.10.001. Another study found

that workers take a long time to move to industries

with a comparative advantage. See Naércio A.

Menezes-Filho and Marc-Andreas Muendler, Labor

Reallocation in Response to Trade Reform, Working

Paper 17372 (National Bureau of Economic Research,

August 2011), www.nber.org/papers/w17372.

8. One study found that the reallocation of resources

may explain over 40 percent of the growth in the

productivity of the U.S. manufacturing sector

between 1983 and 1992. See Andrew B. Bernard

and J. Bradford Jensen, “Exporting and Productivity

in the USA,” Oxford Review of Economic Policy,

vol. 20, no. 3 (Autumn 2004), pp. 343–357,

http://dx.doi.org/10.1093/oxrep/grh020.

9. The literature describes some such effects as dynamic

effects. For examples of how trade can stimulate

productivity improvements among businesses, see Alla

Lileeva and Daniel Trefler, “Improved Access to

Foreign Markets Raises Plant-Level Productivity... for

Some Plants,” Quarterly Journal of Economics, vol. 125,

no. 3 (August 2010), pp. 1051–1099, http://

dx.doi.org/10.1162/qjec.2010.125.3.1051; and John

R. Baldwin and Wulong Gu, Long-Term Productivity

Growth in Manufacturing in Canada and the United

States, 1961 to 2003, Canadian Productivity Review

Series Paper 2007015e (Statistics Canada, Economic

Analysis, 2007), http://tinyurl.com/statcan-15-206-x.

Also see Alla Lileeva, Trade Liberalization and

Productivity Dynamics: Evidence From Canada,

Economic Analysis Research Series Paper 2008051e

(Statistics Canada, Analytical Studies Branch, 2008),

http://tinyurl.com/statcan-11f0027mie; Johannes

Van Biesebroeck, “Exporting Raises Productivity in

Sub-Saharan African Manufacturing Firms,” Journal of

International Economics, vol. 67, no. 2 (December

2005), pp. 373–391, http://dx.doi.org/10.1016/

j.jinteco.2004.12.002; and Jan De Loecker, “Do

Exports Generate Higher Productivity? Evidence From

Slovenia,” Journal of International Economics, vol. 73,

no. 1 (September 2007), pp. 69–98, http://dx.doi.org/

10.1016/j.jinteco.2007.03.003.

10. One study found that technology spillovers caused

by trade have significantly increased total factor

productivity over the past 135 years. See Jakob B.

Madsen, “Technology Spillover Through Trade and

TFP Convergence: 135 Years of Evidence for the

OECD Countries,” Journal of International

Economics, vol. 72, no. 2 (2007), pp. 464–480,

http://dx.doi.org/10.1016/j.jinteco.2006.12.001. In

addition, import growth has been shown to increase

productivity. Another study found that increased

openness to imports of capital goods improved

productivity. See David T. Coe, Elhanan Helpman,

and Alexander W. Hoffmaiseter, “North-South R &

D Spillovers,” Economic Journal, vol. 10

7, no. 440

(Januar

y 1997), pp. 134–149, www.jstor.org/stable/

2235275.

11. For more information, see Hildegunn Kyvik Nordås,

Sébastien Miroudot, and Przemyslaw Kowalski,

Dynamic Gains From Trade (Organisation for Economic

Co-operation and Development, November 2006),

http://dx.doi.org/10.1787/276358887412; and Susan

Stone and Ben Shepherd, Dynamic Gains From Trade:

The Role of Intermediate Inputs and Equipment Imports,

OECD Trade Policy Series Paper 110 (OECD

Publishing, 2011), http://dx.doi.org/10.1787/

5kgf17f17ks1-en.

12. Economic studies have established a connection

between degree of trade exposure and job separations.

See Lori G. Kletzer, “Trade and Job Loss in U.S.

Manufacturing, 1979–1994,” in Robert C. Feenstra,

ed., The Impact of International Trade on Wages

(University of Chicago Press, 2000), pp. 349–396,

www.nber.org/chapters/c6198; Bernard Hoekman

and Alan L. Winters, Trade and Employment: Stylized

Facts and Research Findings, Working Paper 7 (United

Nations, Department of Economics and Social

Affairs, 2005), https://ideas.repec.org/p/une/wpaper/

7.html; and David H. Autor and others, “Trade

Adjustment: Worker-Level Evidence,” Quarterly

Journal of Economics, vol. 129, no. 4 (November

2014), pp. 1799–1860, http://dx.doi.org/10.1093/

qje/qju026.

13. See Scott C. Bradford, Paul L. Grieco, and Gary C.

Hufbauer, “The Payoff to America From Globalisation,”

World Economy, vol. 29, no.7 (July 2006), pp. 893–916,

http://dx.doi.org/10.1111/j.1467-9701.2006.00828.x.

18 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY SEPTEMBER 2016

CBO

14. That increase in purchasing power amounts to, in

effect, an increase in consumers’ real income.

15. Several studies looking at trade liberalization support

the hypothesis that increased international competition

puts downward pressure on domestic prices. For

example, see James Levinsohn, “Testing the Imports-as-

Market-Discipline Hypothesis,” Journal of International

Economics, vol. 35, no. 1–2 (August 1993), pp. 1–22,

http://dx.doi.org/10.1016/0022-1996(93)90002-F;

Ann E. Harrison, “Productivity, Imperfect Competition,

and Trade Reform: Theory and Evidence,” Journal of

International Economics, vol. 36, no. 1 (February

1994), pp. 53–73, http://dx.doi.org/10.1016/

0022-1996(94)90057-4; Filip Abraham, Jozef Konings,

and Stijn Vanormelingen, “The Effect of Globalization

on Union Bargaining and Price-Cost Margins of

Firms,” Review of World Economics, vol. 145, no. 1

(April 2009), pp. 13–36, http://dx.doi.org/10.1007/

s10290-009-0003-8; and Jan De Loecker, “Product

Differentiation, Multiproduct Firms, and Estimating

the Impact of Trade Liberalization on Productivity,”

Econometrica, vol. 79, no. 5 (September 2011), pp.

1407–1451, http://dx.doi.org/10.3982/ECTA7617.

However, one study found that after trade liberalization,

businesses in India did not fully pass on to consumers

the cost savings from increased productivity; see Jan

De Loecker and others, “Prices, Markups, and Trade

Reform,” Econometrica, vol. 84, no. 2 (March 2016),

445–510, http://dx.doi.org/10.3982/ecta11042.

16. See Nigel Pain, Isabella Koske, and Marte Sollie,

“Globalisation and OECD Consumer Price Inflation,”

OECD Journal: Economic Studies, vol. 2008, no. 1

(December 2008), pp. 1–32, http://dx.doi.org/

10.1787/eco_studies-v2008-art4-en.

17. See Pablo D. Fajgelbaum and Amit K. Khandelwal,

“Measuring the Unequal Gains From Trade,” Quarterly

Journal of Economics, vol. 131, no. 3 (August 2016),

pp. 1113–1180, http://dx.doi.org/10.1093/qje/qjw013.

18. Results from several studies suggest that increased

product variety is an important consequence of

trade liberalization. For example, see Scott L.

Baier, Jeffrey H. Bergstrand, and Michael Feng,

“Economic Integration Agreements and the Margins

of International Trade,” Journal of International

Economics, vol. 93, no. 2 (July 2014), pp. 339–350,

http://dx.doi.org/10.1016/j.jinteco.2014.03.005; and

Timothy J. Kehoe and Kim J. Ruhl, “How Important

Is the New Goods Margin in International Trade?”

Journal of Political Economy, vol. 121, no. 2 (April

2013), pp. 358–392, http://dx.doi.org/10.1086/

670272.

19. One study showed that trade in intermediate inputs

represents an important source of increases in product

variety; see Pinelopi Goldberg and others, “Imported

Intermediate Inputs and Domestic Product Growth:

Evidence From India,” Quarterly Journal of Economics,

vol. 125, no. 4 (November 2010), pp. 1727–1767,

http://dx.doi.org/10.1162/qjec.2010.125.4.1727.

Another study found that trade was a large contributor

to the increase in the variety of products in the

United States between 1972 and 2001; see Christian

Broda and David W. Weinstein, “Variety Growth and

World Welfare,” American Economic Review, vol. 94,

no. 2 (May 2004), pp. 139–144, http://dx.doi.org/

10.1257/0002828041301443.

20. For an example, see Bo Chen and Hong Ma,

“Import Variety and Welfare Gain in China,”

Review of International Economics, vol. 20, no. 4

(September 2012), http://dx.doi.org/10.1111/

j.1467-9396.2012.01056.x.

21. Several studies have found that trade raises wages for

some workers. One study found that exports have a

sizable positive effect on industry wages; see Philip Du

Caju, Francois Rycx, and Ilan To

jerow, Wage Structure

Effects of International Tr

ade: Evidence From a Small

Open Economy, Working Paper 1325 (European

Central Bank, April 2011), http://tinyurl.com/

itstheecon. Another study showed that an increase

in exporting raised wages for skilled and nonskilled

workers; see David Hummels and others, “Offshoring,

Transition, and Training: Evidence From Danish

Matched Worker-Firm Data,” American Economic

Review, vol. 102, no. 3 (May 2012), pp. 424–428,

http://dx.doi.org/10.1257/aer.102.3.424.

22. See Catherine L. Mann, “How Does Trade Affect

the American Worker?” in Is the U.S. Trade Deficit

Sustainable? (Peterson Institute for International

Economics, 1999), p. 59, http://tinyurl.com/

piie-4iie2644 (PDF, 271 KB); and Ann E. Harrison,

“Productivity, Imperfect Competition, and Trade

Reform: Theory and Evidence,” Journal of International

Economics, vol. 36, no. 1 (February 1994), pp. 53–73,

http://dx.doi.org/10.1016/0022-1996(94)90057-4.

SEPTEMBER 2016 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY 19

CBO

23. One study showed that jobs with less routine tasks

(such as communication-based occupations) tended

to see wages for their workers rise as a result of trade;

see Lindsay Oldenski, “Offshoring and Polarization of

the U.S. Labor Market,” Industrial and Labor Relations

Review, vol. 67, no. 3 (May 2014), pp. 734–761,

http://dx.doi.org/10.1177/00197939140670S311.

24. Longer-tenured workers tend to have more job-specific

skills that are less portable to other jobs. See Lori G.

Kletzer, “Trade-Related Job Loss and Wage Insurance:

A Synthetic Review,” Review of International Economics,

vol. 12, no. 5 (November 2004), pp. 724–748,

http://dx.doi.org/10.1111/j.1467-9396.2004.00479.x;

and Avraham Ebenstein and others, “Estimating the

Impact of Trade and Offshoring on American Workers

Using the Current Population Surveys,” Review of

Economics and Statistics, vol. 96, no. 4 (October 2014),

pp. 581–595, http://dx.doi.org/10.1162/rest_a_00400.

25. See Jon Haveman, The Effect of Trade Induced

Displacement on Unemployment and Wages

(Purdue CIBER Working Papers, January 1993),

http://docs.lib.purdue.edu/ciberwp/76; and Lori

G. Kletzer, “Trade-Related Job Loss and Wage

Insurance: A Synthetic Review,” Review of

International Economics, vol. 12, no. 5 (November

2004), pp. 724–748, http://dx.doi.org/10.1111/

j.1467-9396.2004.00479.x.

26. Some estimates suggest that of the displaced workers

most vulnerable to import-related job losses,

approximately two-thirds earned less in their new job

and about a quarter of those earned 30 percent less in

their new job; see Lori G. Kletzer, Job Loss From

Imports: Measuring the Costs (Peterson Institute,

2001), http://bookstore.piie.com/book-store/

110.html. Another study showed that the losses

suffered by workers displaced when companies moved

operations overseas to take advantage of lower costs

were significant—amounting to about half a year’s

earnings for those workers before they were displaced;

see David Hummels and others, “The Wage Effects of

Offshoring: Evidence From Danish Matched Worker-

Firm Data,” American Economic Review, vol. 104,

no. 6 (June 2014), pp. 1597–1629, http://dx.doi.org/

10.1257/aer.104.6.1597.

27. The wages of workers who switched occupations as a

result of a trade-related displacement declined by

between 12 percent and 17 percent; see Avraham

Ebenstein and others, “Estimating the Impact of

Trade and Offshoring on American Workers Using

the Current Population Surveys,” Review of Economics

and Statistics, vol. 96, no. 4 (October 2014), pp. 581–

595, http://dx.doi.org/10.1162/rest_a_00400.

28. See Organisation for Economic Co-operation and

Development, OECD Employment Outlook 2005

(OECD, July 2005), http://dx.doi.org/10.1787/

empl_outlook-2005-en.

29. See Daron Acemoglu and others, “Import Competition

and the Great US Employment Sag of the 2000s,”

Journal of Labor Economics, vol. 34, no. S1 (2016),

http://dx.doi.org/10.1086/682384; and Justin Pierce

and P.K. Schott, “The Surprisingly Swift Decline of

US Manufacturing Employment,” American Economic

Review, vol. 106, no. 7 (July 2016), pp. 1632–1662,

http://dx.doi.org/10.1257/aer.20131578.

30. See Chad Bown and Douglas A. Irwin, The GATT’s

Starting Point: Tariff Levels Circa 1947, Working

Paper 21782 (National Bureau of Economic Research,

December 2015), www.nber.org/papers/w21782.

31. Between 1960 and 2014, global trade in goods and

ser

vices grew from 25.0 percent to 60.2 percent of

global GDP.

32. D

espite the departure from the MFN principle,

GATT treaties allow participating nations to join

PTAs under certain circumstances.

33. Trade agreements prevent imports produced mostly

by nonparticipating countries from gaining

preferential treatment by setting rules for origin

requirements. Those rules specify how much of the

production process (or value added) for any imported

final good must take place within member countries

for a good to receive preferential treatment under a

trade agreement. Strong rules make it harder for

exporters in nonmember countries to use the tariff

preferences granted under a trade agreement. Rules-

of-origin requirements apply for all provisions of a

trade agreement that provide preferential treatment to

imported goods, not only tariff preferences.

34. In countries where intellectual property is granted

stronger protection, some medicines can become

considerably more costly for consumers; see Shubham

Chaudhuri, Pinelopi K. Goldberg, and Panle Gia,

20 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY SEPTEMBER 2016

CBO

“Estimating the Effects of Global Patent Protection in

Pharmaceuticals: A Case Study of Quinolones in

India,” American Economic Review, vol. 96, no. 5

(December 2006), pp. 1477–1514, http://dx.doi.org/

10.1257/aer.96.5.1477.

35. Estimates of those effects are uncertain, for several

reasons. First, data on trade flows are limited and

often unreliable. Second, existing models may not

fully capture the mechanisms through which trade

agreements affect trade flows and economies. Third,

researchers do not have reliable estimates of how

nontariff barriers affect trade flows. The appendix

discusses the sources of difficulty in assessing the

effects of PTAs.

36. The rule-based reforms included in PTAs lower trade

barriers and affect total trade, although their net

effect is unclear. For example, harmonizing

commercial standards can promote trade by making it

less costly for exporting businesses to comply with

various product safety standards in multiple markets.

But standard harmonization can also impede trade if it

raises production costs for businesses in member

countries.

37. All other U.S. trade agreements have been with

countries whose combined GDP was less than

10 percent of U.S. GDP. Even for NAFTA, the

combined GDP of Canada and Mexico amounted

to only one-fifth of U.S. GDP at the time of the

agreement.

38. Before establishing potential U.S. PTAs, the United

States International Trade Commission (ITC) employs

a stylized model of trade to project the economic

effects of those agreements. In those projections, the

agency estimates that all agreements yield small

increases to total U.S. trade. See, for example, United

States International Trade Commission, U.S.–

Colombia Trade Promotion Agreement: Potential

Economy-wide and Selected Sectoral Effects, Publication

3896 (December 2006), www.usitc.gov/publications/

332/pub3896.pdf (1.18 MB); and United States

International Trade Commission, U.S.–Korea Trade

Promotion Agreement: Potential Economy-wide and

Selected Sectoral Effects, Publication 3949 (September

2007), www.usitc.gov/publications/pub3949.pdf

(3.84 MB). Other researchers analyzing all recent

PTAs (not just those involving the United States) also

projected increases in total trade; see, for example,

Scott L. Baier and Jeffrey H. Bergstrand, “Do Free

Trade Agreements Actually Increase Members’

International Trade?” Journal of International

Economics, vol. 71, no. 1 (March 2007), pp. 72–95,

http://dx.doi.org/10.1016/j.jinteco.2006.02.005; and

Christopher S.P. Magee, “New Measures of Trade

Creation and Trade Diversion,” Journal of

International Economics, vol. 75, no. 2 (July 2008),

pp. 349–362, http://dx.doi.org/10.1016/

j.jinteco.2008.03.006.

39. For a review of empirical analyses of other effects

from NAFTA, see Justino De La Cruz, David Riker,

and Bennet Voorhees, “Econometric Estimates of the

Effects of NAFTA: A Review of the Literature” (United

States International Trade Commission, December

2013), www.usitc.gov/publications/332/

ec201312a.pdf (124 KB).

40. See Maureen Rimmer and Peter Dixon, “Identifying

the Effects of NAFTA on the U.S. Economy Between

1992 and 1998: A Decomposition Analysis,” Global

Trade Analysis Project (April 2015), http://tinyurl.com/

gtap4657.

41. See Lorenzo Caliendo and Fernando Parro, “Estimates

of the Trade and Welfare Effects of NAFTA,” Review

of Economic Studies, vol. 82, no. 1 (January 2015),

pp. 1–44, http://dx.doi.org/10.1093/restud/rdu035;

and John Ro

malis, “NAFTA’s and

CUSFTA’s Impact

on International Trade,” Review of Economics and

Statistics, vol. 89, no. 3 (August 2007), pp. 416–435,

http://dx.doi.org/10.1162/rest.89.3.416.

42. The U.S.–Canada Free Trade Agreement, implemented

in 1989, liberalized trade between those countries.

NAFTA superseded that agreement in 1994.

43. See United States International Trade Commission,

Economic Impact of Trade Agreements Implemented

Under Trade Authorities, 2016 Report, Publication

4614 (June 2016), www.usitc.gov/publications/332/

pub4614.pdf (3.39 MB).

44. An important complication of estimating trade

creation from NAFTA is identifying the import and

export growth that would have occurred without the

agreement. Doing so is difficult in part because

NAFTA began just as Mexico was growing rapidly as a

result of policy reforms during the 1980s and early

1990s. Moreover, Mexico experienced a severe financial

SEPTEMBER 2016 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY 21

CBO

crisis in 1994 that resulted in a steep depreciation of

the peso, which encouraged exports from Mexico and

discouraged imports to Mexico. An earlier review of

NAFTA found that the agreement had a positive effect

on the growth of trade flows between the United States

and both Mexico and Canada, but that effect in the

literature varied greatly for the same reasons. See Gary

Clyde Hufbauer and Jeffrey J. Schott (assisted by Paul

L.E. Grieco), NAFTA Revisited: Achievements and

Challenges (Peterson Institute for International

Economics, October 2005), http://bookstore.piie.com/

book-store/332.htm.

45. For a review of studies completed before the

implementation of NAFTA, see Mary E. Burfisher

and others, “The Impact of NAFTA on the United

States,” Journal of Economic Perspectives, vol. 15, no. 1

(March 2001), pp. 125–144, http://dx.doi.org/

10.1257/jep.15.1.125.

46. See Maureen Rimmer and Peter Dixon, “Identifying

the Effects of NAFTA on the U.S. Economy Between

1992 and 1998: A Decomposition Analysis,”

Global Trade Analysis Project (April 2015),

http://tinyurl.com/gtap4657.

47. See Lorenzo Caliendo and Fernando Parro, “Estimates

of the Trade and Welfare Effects of NAFTA,” Review of

Economic Studies, vol. 82, no. 1 (January 2015),

pp. 1–44, http://dx.doi.org/10.1093/restud/rdu035.

48. See John Romalis, “NAFTA’s and CUSFTA’s Impact

on International Trade,” Review of Economics and

Statistics, vol. 89, no. 3 (August 2007), pp. 416–435,

http://dx.doi.org/10.1162/rest.89.3.416.

49. See Maureen Rimmer and Peter Dixon, “Identifying

the Effects of NAFTA on the U.S. Economy Between

1992 and 1998: A Decomposition Analysis,”

Global Trade Analysis Project (April 2015),

http://tinyurl.com/gtap4657.

50. Most studies indicate that the U.S.–Korea Free Trade

Agreement (KORUS) lowered the U.S. trade balance.

Renan Zhuang, Jeremy Mattson, and Won Koo

estimated before the agreement’s implementation that

KORUS would cause the U.S. trade balance to

deteriorate by $7.8 billion, and the United States

International Trade Commission found that KORUS

would lower the U.S. trade balance by $362 million.

See United States International Trade Commission,

U.S.–Korea Trade Promotion Agreement: Potential

Economy-wide and Selected Sectoral Effects, Publication

3949 (September 2007), www.usitc.gov/publications/

pub3949.pdf (3.84 MB); and Renan Zhuang, Jeremy

W. Mattson, and Won W. Koo, Implications of the

U.S.–Korea Free Trade Agreement for Agriculture and

Other Sectors of the Economy (Center for Agricultural

Policy and Trade Studies, North Dakota State

University, October 2007), http://tinyurl.com/aer619

(PDF, 166 KB). The ITC has found similarly mixed

results from several other U.S. trade agreements. For

example, the ITC predicted that the U.S.–Australia

Free Trade Agreement would boost the U.S.

trade balance by $339 million over its 18-year

implementation period and that the Dominican

Republic–Central America Free Trade Agreement

(CAFTA-DR) would raise the U.S. trade balance by

$756 million over its 20-year implementation period.

Conversely, the ITC concluded that the U.S.–Colombia

Free Trade Agreement and the U.S.–Singapore Free

Trade Agreement would reduce the U.S. trade balance

by $66 million and about $108 million, respectively.

Those estimates differ not only for economic reasons

(differences in the size of tariff reductions) but also

as a result of differences in the analyses’ modeling

techniques.

51. Trade deficits are not caused by either U.S. or foreign

trade policies. Rather, they are determined by the

balances between saving and investment in the United

States and in other countries and the effects of those

balances on international capital flows. Trade policy

normally has little, if any, effect on the trade deficit

because it generally has little effect on saving and

investment, both domestically and abroad. See

Congressional Budget Office, Causes and Conseq

uences

of the Trade De

ficit: An Overview (March 2000),

www.cbo.gov/publication/12139.

52. The ITC, for instance, warned in its study of KORUS

that “simulation results should not be interpreted as

changes in total imports and exports, or as implying

meaningful information about the balance of trade

impact of the entire U.S.–Korea FTA.” See United

States International Trade Commission, U.S.–Korea

Trade Promotion Agreement: Potential Economy-wide

and Selected Sectoral Effects, Publication 3949

(September 2007), p. xix, www.usitc.gov/

publications/pub3949.pdf (3.84 MB).

22 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY SEPTEMBER 2016

CBO

53. Suppose, for example, that trade liberalization sharply

increased U.S. agricultural exports. Greater agricultural

output would raise GDP, which could lead to an

appreciation of the value of the dollar. That change

would tend to discourage U.S. exports and encourage

more U.S. imports. The net result could be an

increase or a decrease in the trade balance.

54. One study showed that the complexity of the

production and distribution process often determined

whether companies chose to serve those markets

through exports (if the process was more complex) or

through affiliates (if the process was less complex).

See Lindsay Oldenski, “Export Versus FDI and the

Communication of Complex Information,” Journal of

International Economics, vol. 87, no. 2 (July 2012),

pp. 312–322, http://dx.doi.org/10.1016/

j.jinteco.2011.12.012.

55. One study found that NAFTA led to large flows of

foreign direct investment into Mexico; see Maggie X.

Chen, “Regional Economic Integration and

Geographic Concentration of Multinational Firms,”

European Economic Review, vol. 53, no. 3 (April

2009), pp. 355–375, http://dx.doi.org/10.1016/

j.euroecorev.2008.05.002. More generally, Chen’s

results suggested that countries with relatively large

labor pools and those that gained preferential access

to large markets because of trade agreements were the

most significant recipients of FDI inflows as a result

of those agreements.

56. To ensure that all investment-related provisions are

properly enforced, some agreements establish an

independent authority to resolve legal disputes

between foreign entities and domestic governments.

When foreign investors have legal disagreements with

domestic governments, dispute resolution authorities

give those investors legal recourse outside the

domestic court system. That mechanism reinforces

the legal rights of investors and promotes additional

direct investment in member countries. One study

found that dispute resolution provisions in the U.S.–

Vietnam Trade Agreement led to increased FDI flows

from the United States to Vietnam; see Tim Büthe

and Helen V. Milner, “The Politics of Foreign Direct

Investment Into Developing Countries: Increasing

FDI Through International Trade Agreements?”

American Journal of Political Science, vol. 52, no. 4

(October 2008), pp. 741–762, http://dx.doi.org/

10.1111/j.1540-5907.2008.00340.x.

57. As part of NAFTA, Mexico agreed to loosen

regulations that prevented U.S. financial services

businesses from opening branches or subsidiaries in

Mexico.

58. In addition, agreements have tried to ensure proper

oversight of the rules through a reliable legal system.

59. See Jeong-Yeon Lee and Edwin Mansfield, “Intellectual

Property Protection and U.S. Foreign Direct

Investment,” Review of Economics and Statistics, vol. 78,

no. 2 (May 1996), pp. 181–186, http://dx.doi.org/

10.2307/2109919.

60. To CBO’s knowledge, there is only one empirical

analysis of the effect of U.S. trade agreements on

FDI. Although that study found that NAFTA may

have decreased total FDI among member countries, it

could not identify how NAFTA affected FDI for each

country in the agreement. See Philippa Dee and

Jyothi Gali, “The Trade and Investment Effects of

Preferential Trading Arrangements,” in Takatoshi Ito

and Andrew K. Rose, eds., International Trade in East

Asia, NBER–East Asia Seminar on Economics, vol. 14

(National Bureau of Economic Research, August

2005), pp. 133–176, www.nber.org/chapters/c0193.

Most empirical studies have found that agreements

tend to significantly increase FDI among member

countries. One study found that the stock of FDI

among trade agreement partners was 27 percent

larger, on average, than the stock of FDI among

countries without agreements—an effect driven

mainly by investment flows from developed countries

to developing countries; see Eduardo Levy Yeyati,

Ernesto Stein, and Christian Daude,

Regional

Integration and the Location of FDI, Working Paper

2 (Inter-American Development Bank, Research

Department, July 2003), http://tinyurl.com/

iadb-wp492. Another study examined the trade

agreements that created the European Union (the

Europe Agreements) and estimated that FDI stocks

between agreement members were, on average,

144 percent greater than the bilateral FDI stocks of

countries that did not join the agreement. The rise in

FDI among Europe Agreement members tended to

flow from Western European nations to Central and

Eastern European countries; see Badi H. Baltagi,

SEPTEMBER 2016 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY 23

CBO

Peter Egger, and Michael Pfaffermayr, “Estimating

Models of Complex FDI: Are There Third-Country

Effects?” Journal of Econometrics, vol. 140, no. 1

(September 2007), pp. 260–281, http://dx.doi.org/

10.1016/j.jeconom.2006.09.009. One other study

estimated that trade agreements tended to raise FDI

stocks among member countries by 170 percent above

the levels of FDI that would prevail without those

agreements over the first 10 years of enforcement. Like

the other studies, those estimates reflect large increases

in direct investments by developed countries in

developing countries; see Max Büge, Do Preferential

Trade Agreements Increase Their Members’ Foreign

Direct Investment? Discussion Paper 37/2014

(German Development Institute, September 2014),

http://tinyurl.com/die-gdi-dp37.

61. Estimates from two studies indicate that the trade

agreement between Canada and the United States

improved the productivity of Canadian businesses;

see Daniel Trefler, “The Long and Short of the

Canada–U.S. Free Trade Agreement,” American

Economic Review, vol. 94, no. 4 (September 2004),

pp. 870–895, http://dx.doi.org/10.1257/

0002828042002633; and Keith Head and John Ries,

“Rationalization Effects of Tariff Reductions,” Journal

of International Economics, vol. 47, no. 2 (April

1999), pp. 295–320, http://dx.doi.org/10.1016/

S0022-1996(98)00019-1. Another study found a

significant positive effect of NAFTA on the

productivity of Mexican businesses. See Rafael E. De

Hoyos and Leonardo Iacovone, “Economic

Performance Under NAFTA: A Firm-Level Analysis of

the Trade-Productivity Linkages,” World Development,

vol. 44 (April 2013), pp. 180–193, http://dx.doi.org/

10.1016/j.worlddev.2012.09.008.

62. Several studies emphasize that even though NAFTA

created some jobs and destroyed others, the net effects

of trade agreements on overall employment are

insignificant compared with factors such as economic

growth and technological change. For example, one

study estimated that the job displacements in the

United States associated with trade with other

NAFTA countries between 1990 and 1997 (not

identifying the specific effect of NAFTA) accounted

for less than 2 percent of all U.S. job separations over

that period. That study’s authors, after incorporating

their estimate of the number of jobs supported by

new exports to NAFTA partners, concluded that

NAFTA marginally increased total U.S. employment.

See Raúl Hinojosa Ojeda and others, The U.S.

Employment Impacts of North American Integration

After NAFTA: A Partial Equilibrium Approach

(University of California, Los Angeles, 2000),

http://tinyurl.com/ucla-ojeda (PDF, 361 KB). See

also Mary E. Burfisher and others, “The Impact of

NAFTA on the United States,” Journal of Economic

Perspectives, vol. 15, no. 1 (March 2001), pp. 125–

144, http://dx.doi.org/10.1257/jep.15.1.125; and

Willem Thorbecke and Christian Eigen-Zucchi,

“Did NAFTA Cause a ‘Giant Sucking Sound’?”

Journal of Labor Research, vol. 23, no. 4 (December

2002), pp. 647–658, http://dx.doi.org/10.1007/

s12122-002-1033-3.

63. A series of studies by Robert Scott and his coauthors

had the most pessimistic assessment of NAFTA’s

employment effects. One study from that series

estimated that trade between the United States and its

NAFTA partners was responsible for around 92,000

jobs lost, on net, in the United States every year from

1994 to 2003; see Robert E. Scott, Carlos Salas, and

Bruce Campbell, Revisiting NAFTA: Still Not Working

for North America’s Workers, Briefing Paper 173

(Economic Policy Institute, September 2006),

www.epi.org/publication/bp173/. H

owe

ver, those

estimates should be interpreted with caution, for two

reasons. First, the estimates rely on an assumption

that changes in the U.S. trade deficit can be translated

directly into changes in employment—an assumption

with little empirical support. Second, the estimates

incorporate the assumption that NAFTA was

responsible for all changes in the U.S. balance of trade

with Mexico and Canada. Because several other

factors influenced the trade balance with NAFTA

partners over that period (exchange rates and

economic growth, for example), the authors probably

could not distinguish the effect of NAFTA from other

factors that affected trade flows. For both reasons,

Scott and his coauthors probably overestimated

NAFTA’s effects on U.S. employment. For additional

commentary, see Gary Clyde Hufbauer, Cathleen

Cimino-Isaacs, and Tyler Moran, NAFTA at 20:

Misleading Charges and Positive Achievements, Policy

Brief 14-13 (Peterson Institute for International

Economics, May 2014), http://tinyurl.com/

piie-pb14-13; and Gary Clyde Hufbauer and Jeffrey

J. Schott (assisted by Paul L.E. Grieco), NAFTA

Revisited: Achievements and Challenges (Peterson

24 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY SEPTEMBER 2016

CBO

Institute for International Economics, October 2005),

http://bookstore.piie.com/book-store/332.html.

Another study found that imports from NAFTA were

responsible for 10.7 percent of all job losses in the

United States between 1993 and 1999. However, that

estimate did not account for possible job creation

from increased exports under NAFTA. See Lori G.

Kletzer, “Globalization and American Job Loss:

Public Policy to Help Workers,” Perspectives of Work,

vol. 6, no. 1 (2002), pp. 28–30, www.jstor.org/stable/

23272039.

64. One review of the literature found that NAFTA

raised average real wages in the United States, but

only modestly. See Justino De La Cruz and David

Riker, The Impact of NAFTA on U.S. Labor Markets

(United States International Trade Commission,

April 2014), www.usitc.gov/publications/332/

ec201406a.pdf (135 KB).

65. For CBO’s economic analysis of NAFTA, see

Congressional Budget Office, The Effects of NAFTA

on U.S.–Mexican Trade and GDP (May 2003),

www.cbo.gov/publication/14461.

66. See United States International Trade Commission,

The Impact of Trade Agreements: Effect of the Tokyo

Round, U.S.–Israel FTA, NAFTA, and the Uruguay

Round on the U.S. Economy, Publication 3621 (August

2003), www.usitc.gov/publications/332/pub3621.pdf

(4.1 MB).

67. Analyses of other trade agreements done using

stylized models also indicate small increases in U.S.

GDP through those channels. For example, although

the ITC completed no beforehand estimate of the

effects of NAFTA on GDP, the agency’s estimates

suggest that KORUS, the U.S.–Colombia Free Trade

Agreement, and the U.S.–Australia Free Trade

Agreement had even smaller effects than CBO’s

estimate for NAFTA. See the following publications

of the United States International Trade Commission:

U.S.–Colombia Trade Promotion Agreement: Potential

Economy-wide and Selected Sectoral Effects, Publication

3896 (December 2006), www.usitc.gov/publications/

332/pub3896.pdf (1.18 MB), U.S.–Australia Trade

Free Trade Agreement: Potential Economywide and

Selected Sectoral Effects, Publication 3697 (May 2004),

www.usitc.gov/publications/332/pub3697.pdf

(1.81 MB), and U.S.–Korea Trade Promotion

Agreement: Potential Economy-wide and Selected

Sectoral Effects, Publication 3949 (September 2007),

www.usitc.gov/publications/pub3949.pdf (3.84 MB);

as well as Congressional Budget Office, The Effects of

NAFTA on U.S.–Mexican Trade and GDP (May 2003),

www.cbo.gov/publication/14461.

68. See Maureen Rimmer and Peter Dixon, “Identifying

the Effects of NAFTA on the U.S. Economy Between

1992 and 1998: A Decomposition Analysis,” Global

Trade Analysis Project (April 2015), http://tinyurl.com/

gtap4657.

69. One study described how NAFTA increased the

variety of agricultural products available to U.S.

consumers. Examples of “new” produce include

certain varieties of tomatoes and avocados imported

from Mexico. See Cathy Jabara and Brendan Lynch,

Exports and New Varieties: An Analysis of U.S.–Mexico

Agricultural Trade, Working Paper No. 16 (United

States International Trade Commission, 2006),

http://go.usa.gov/xKpxM (PDF, 450 KB).

70. See David H. Autor and others, “Trade Adjustment:

Worker-

Lev

el Evidence,” Quarterly Journal of Economics,

vol. 129, no. 4 (November 2014), pp. 1799–1860,

http://dx.doi.org/10.1093/qje/qju026.

71. See Shushanik Hakobyan and John McLaren, Looking

for Local Labor Market Effects of NAFTA, Working

Paper 16535 (National Bureau of Economic Research,

November 2010), www.nber.org/papers/w16535.

72. Ibid.

73. Workers in geographical areas whose economies have

gained the most from PTAs may also benefit. Areas

with businesses that expand as a result of those

agreements see an increase in their demand for other

goods and services to support the growth of those

expanding businesses. As a result, service-sector

workers in unrelated industries often see an increase

in demand for their labor.

74. The economics literature has consistently found that

exporting businesses tend to pay higher wages; for

example, see Andrew B. Bernard, J. Bradford Jensen,

and Robert Z. Lawrence, “Exporters, Jobs, and Wages

in U.S. Manufacturing: 1976–1987,” Brookings

Papers on Economic Activity: Microeconomics, vol. 1995

SEPTEMBER 2016 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY 25

CBO

(1995), pp. 67–119, http://dx.doi.org/10.2307/

2534772. However, the literature has been unable to

show whether exporting tends to raise wages or

whether businesses that pay higher wages are more

likely to become exporters.

75. For example, see these Congressional Budget Office

publications: cost estimate for S. 1642, United States–

Korea Free Trade Agreement Implementation Act

(October 12, 2011), www.cbo.gov/publication/42643;

cost estimate for H.R. 5684, United States–Oman Free

Trade Agreement Implementation Act (September 22,

2006), www.cbo.gov/publication/18147; and cost

estimate for S. 2610, a bill to implement the United

States–Australia Free Trade Agreement (July 30, 2004),

www.cbo.gov/publication/16161.

76. When import duties, excise taxes, and other indirect

business taxes (such as fees on businesses) are lowered

on goods and services, they tend to increase income

for workers or business owners in the taxed activity

and for others throughout the economy. As a result,

revenues from individual and corporate income taxes

and payroll taxes also tend to be higher. Increases in

such indirect business taxes would have the opposite

effect. See Congressional Budget Office, The Role of

the 25 Percent Revenue Offset in Estimating the

Budgetary Effects of Legislation (January 2009),

www.cbo.gov/publication/20110. For the latest offset

percentages, which vary annually between 25 percent

and 26 percent, see Joint Committee on Taxation,

New Income and Payroll Tax Offsets to Changes in

Excise Tax Revenues for 2016–2026, JCX-7-16

(February 2016), http://go.usa.gov/xKpkT (PDF,

17 KB).

77. See Congressional Budget Office, A Budgetary and

Economic Analysis of the North American Free Trade

Agreement (July 1993), www.cbo.gov/publication/

20868. In 2015, import tariff duties accounted for

only 1 percent of all federal tax revenues.

78. See Congressional Budget Office, cost estimate for

S. 1642, United States–Korea Free Trade Implementa-

tion Act (October 12, 2011), www.cbo.gov/publication/

42643. In addition, other CBO estimates of how trade

agreements affected U.S. federal tariff revenues also

have been small. CBO estimated that CAFTA-DR

would increase the federal budget deficit by about

$4 billion between 2006 and 2015, the Singapore–

U.S. Free Trade Agreement would increase the federal

budget deficit by about $1 billion between 2004 and

2013, and the Colombia–U.S. Free Trade Agreement

would decrease the federal budget deficit by $22 million

between 2012 and 2021. Those estimates incorporate

the assumption that some trade will be diverted from

countries outside those agreements in favor of countries

covered by them. See Congressional Budget Office,

cost estimate for S. 1307, Dominican Republic–

Central America–United States Free Trade Agreement

Implementation Act (July 18, 2005), www.cbo.gov/

publication/16993; cost estimate for H.R. 2739,

United States–Singapore Free Trade Agreement

Implementation Act (September 16, 2003),

www.cbo.gov/publication/14758; and cost estimate for

H.R. 3078, United States–Colombia Trade Promotion

Agreement Implementation Act (October 5, 2011),

www.cbo.gov/publication/42611.

CBO

Appendix:

Difficulties in Estimating How

Preferential Trade Agreements Affect an Economy

Estimating how preferential trade agreements (PTAs)

affect an economy is difficult. Trade data are often limited

and unreliable, and distinguishing the effects of PTAs from

the effects of other factors is hard. Moreover, estimates

of the economic effects of PTAs depend on modeling

assumptions.

Data Difficulties

Limited and unreliable data are major obstacles to accu-

rately estimating the economic effects of PTAs. In partic-

ular, three types of difficulties are associated with data on

international trade:

B Discrepancies in Measures of Trade Between Countries.

Trade partners sometimes report significantly different

measurements for trade between their countries. One

reason for the discrepancies is that countries typically

collect data on import flows—from which they earn

tariff revenue—much more carefully than they collect

data on export flows, but even data on imports are

often inaccurate.

B Mistakes in Reporting the Origin of Imports. Because

today’s business supply chains are global, goods often

pass through a third-party country on their way

from exporter to importer.

When that happens, the

importing country is likely to attribute those imports

to the final exporting country rather than the original

exporter.

B Difficulties in Measuring Trade in Services. Because

most services are not subject to tariffs, trade in ser-

vices—which has been a growing fraction of overall

trade in recent years—is not subject to the same

reporting requirements as trade in merchandise.

(Examples of services include financial services and tele-

communications.) Therefore, governments usually

expend fewer resources to collect data on services.

Also, because services are often intangible, tracking

those transactions is much harder than tracking trade

in merchandise.

Modeling Challenges

Researchers regularly use two types of models to estimate

the economic effects of PTAs: stylized models of world

trade and econometric models. Stylized models are based

on economic theory and calibrated to real-world economic

conditions. Using stylized models, researchers compare

1. For further discussion of approaches used to analyze the economic

effects of trade agreements, see World Trade Organization,

A Practical Guide to Trade Policy Analysis (WTO, 2012),

http://tinyurl.com/wto-unctad12 (PDF, 2.28 MB).

2. Ibid. A recent paper describes the discrepancies between trade data

released by the United States and trade data released by China; see

Michael F. Martin, What’s the Difference?—Comparing U.S. and

Chinese Trade Data, Report for Congress RS22640 (Congressional

Research Service, March 24, 2016).

3. For example, when buying a good from China, the United States

records the full value of that good as an import from China.

However, with the rise of global supply chains, most of that good’s

value might reflect value added in another country (such as

Malaysia) before its arrival in China for final assembly. By

attributing the entire value of the import to China and none to

Malaysia, the United States overstates China’s importance in trade

and understates Malaysia’s. As a result, researchers might

incorrectly assume that goods produced primarily by a member

country but assembled in a nonmember country (before final

export) would not qualify for preferential treatment—when, in

fact, they would qualify.

4. For more information on the challenges of collecting data for

trade in services, see International Trade Centre, “Capturing and

Utilizing Services Trade Statistics, A Guide for Statistical

Compilers in Developing Countries” (no date), http://tinyurl.com/

intracenorg (PDF, 266 KB).

28 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY SEPTEMBER 2016

CBO

two scenarios: one that incorporates the estimated eco-

nomic consequences of the proposed agreement and

another that does not. Those models offer some of the

best available predictions of future PTAs’ potential mac-

roeconomic effects. Econometric models, by contrast, use

economic data from member countries to estimate the

consequences of trade agreements.

Both models have difficulties estimating the economic

effects of PTAs, but for different reasons. For the North

American Free Trade Agreement (NAFTA), for example,

most stylized models predicted smaller economic effects

from NAFTA before its implementation than the econo-

metric models attributed to NAFTA after its implemen-

tation.

The main reason for that discrepancy appears to

be that stylized models have underestimated the willing-

ness of consumers to substitute imports from one country

for imports from another.

Stylized Models of World Trade

The stylized models, a class of computational general-

equilibrium models, are grounded in rigorous economic

theory and use estimates based on historical data as

parameters.

They can assess the prospective effects of

PTAs and thus are particularly valuable when policymak-

ers consider the merits of possible agreements. The results

of those models yield useful predictions of the sectors

likely to expand or shrink as a result of PTAs. Stylized

models also predict how lower tariffs are likely to affect

trade among member countries and between a member

country and the rest of the world.

Although stylized models can evaluate the mechanisms

by which PTAs promote specialization and alter trade

flows, their quantitative predictions should be interpreted

with caution. Stylized models of world trade have diffi-

culty capturing the effects of nontariff provisions—such

as labor standards and intellectual property protections—

on trade flows. The structure of stylized models requires

researchers to convert all nontariff provisions associated

with a trade agreement into an equivalent tariff reduction

on specific goods or services. In other words, all nontariff

provisions need to be described in the model as if those

provisions affected prices systematically.

However, some

effects of nontariff provisions cannot be represented

appropriately as a change in tariff rates, making it hard

for those models to capture the economic effects of those

nontariff provisions.

Therefore, that required conversion

adds another layer of uncertainty to quantitative esti-

mates of PTAs’ economic effects. Furthermore, the

results from such models are particularly sensitive to esti-

mates of how consumers would substitute imports from

one country for the imports of another if the relative

prices of those imports changed. Because those estimates

vary widely, assessments of the economic impact of

changes in trade policy also vary considerably, making it

important to check the robustness of the results to

changes in those parameters.

In addition, stylized mod-

els often have difficulty capturing the costs that occur

when trade agreements reallocate economic resources.

For example, those models typically do not estimate the

5. See Timothy J. Kehoe, An Evaluation of the Performance of Applied

General Equilibrium Models of the Impact of NAFTA, Staff Report

320 (Federal Reserve Bank of Minneapolis, August 2003),

http://tinyurl.com/minneapolis-fed-sr320; and Serge Shikher,

“Predicting the Effects of NAFTA: Now We Can Do It Better!”

Journal of International and Global Economic Studies, vol. 5, no. 2

(December 2012), pp. 32–59, http://tinyurl.com/shikher2012

(PDF, 497 KB).

6. See Serge Shikher, “Predicting the Effects of NAFTA: Now

We Can Do It Better!” Journal of International and Global

Economic Studies, vol. 5, no. 2 (December 2012), pp. 32–59,

http://tinyurl.com/shikher2012 (PDF, 497 KB).

7. Computational general-equilibrium models are used to analyze

economic behavioral relationships and interactions among all

sectors of an economy—households, businesses, and

governments—in ways that are consistent with economic theory.

Stylized models of world trade are used to analyze economic

behavioral relationships and relationships among all major trading

nations. Such models allow analysts to use historical data to

simulate how those sectors would react to a potential change in

trade policy and how those reactions might alter macroeconomic

variables.

8. A recent study documents difficulties in estimating how trade

policies lower trade costs. The authors emphasize the issues

researchers confront when trying to estimate the extent to which

nontariff barriers affect trade costs. See Pinelopi K. Goldberg and

Nina Pavcnik, The Effects of Trade Policy, Working Paper 21957

(National Bureau of Economic Research, February 2016),

www.nber.org/papers/w21957.

9. This is particularly true for market access provisions in service

sectors. Trade agreements sometimes include provisions that allow

businesses in service sectors to more easily sell their services to

domestic businesses and consumers in member countries. Such

reforms have complex and potentially substantial effects on both

trade in services and investment flows between member countries,

but distilling those effects into a measure of an equivalent

reduction in tariff rates on those services is challenging.

10. For more discussion, see World Trade Organization, A Practical

Guide to Trade Policy Analysis (WTO, 2012), http://tinyurl.com/

wto-unctad12 (PDF, 2.28 MB).

APPENDIX HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY 29

CBO

transition costs borne by workers whose jobs get dis-

placed by the agreement.

As a result, estimates from

stylized models may understate the costs of establishing

trade agreements.

Researchers use stylized models to assess PTAs in two

ways, and each method has drawbacks. One way high-

lights how an economy would differ if a trade agreement

was fully implemented immediately. The other way proj-

ects economic conditions in some future year (or years),

typically in the year when a trade agreement is likely to be

fully established (usually 10 years or 15 years after the

agreement takes effect). Analyses that assume immediate

implementation can offer insight into economic relation-

ships and the general direction of the effects of PTAs on

certain economic variables. But such analyses are much

less likely to reliably estimate the size of the effects, because

PTAs take years to implement. By contrast, analyses that

project future effects of PTAs can be problematic because

they depend on the underlying economic forecast, which

is uncertain.

Econometric Models

Once a trade agreement has been in effect for several

years, researchers can use newly collected data and

econometric models to estimate its economic effects.

Those models are particularly useful for learning how

previously implemented PTAs have affected economies.

Although econometric models are also grounded in eco-

nomic theory, they require fewer structural assumptions

than stylized models of world trade. Furthermore, econo-

metric analyses can be used to estimate more detailed

effects of PTAs than most stylized models.

The results of such analyses must be interpreted with cau-

tion, however. Econometric models cannot easily sepa-

rate the effect of PTAs from unobserved or unmeasurable

factors that may affect the economies of member coun-

tries. Econometric models have difficulties capturing the

indirect macroeconomic effects of changes in trade pol-

icy. In addition, estimated effects from econometric

models are susceptible to certain biases in the data. For

example, countries that enter into PTAs differ systemati-

cally from countries that do not. If econometric models

do not control for the factors that make countries more

or less likely to enter into PTAs, that source of bias might

distort the estimated effects of those agreements.

11. See David Riker and William Swanson, A Survey of Empirical

Models of Labor Transitions Following Trade Liberalization (United

States International Trade Commission, September 2015),

www.usitc.gov/publications/332/ec201406a.pdf

(224 KB).

12. If models neglect to account for those factors, they are likely to

significantly underestimate the effects of trade agreements on trade

flows. See Scott L. Baier and Jeffrey H. Bergstrand, “Do Free Trade

Agreements Actually Increase Members’ International Trade?”

Journal of International Economics, vol. 71, no. 1 (March 2007),

pp. 72–95, http://dx.doi.org/10.1016/j.jinteco.2006.02.005.

30 HOW PREFERENTIAL TRADE AGREEMENTS AFFECT THE U.S. ECONOMY SEPTEMBER 2016

CBO

About This Document

This Congressional Budget Office report was prepared at the request of the Chairman of the House

Ways and Means Committee. In keeping with CBO’s mandate to provide objective, impartial analysis,

this report makes no recommendations.

Daniel Fried wrote the report, with guidance from Jeffrey Werling and Kim Kowalewski. Mark Booth,

Ann Futrell, Mark Hadley, Joseph Kile, Nathan Musick, and Charles Whalen provided useful

comments and suggestions on various drafts of the report. Robert Shackleton, Claire Sleigh, and

Adam Staveski fact-checked the report.

Helpful comments were also provided by Scott Baier of Clemson University; Jeffrey Bergstrand of

the University of Notre Dame; Maggie Chen of the George Washington University; Gary Hufbauer

of the Peterson Institute for International Economics; Nuno Limão of the University of Maryland;

John McLaren of the University of Virginia; Nina Pavcnik of Dartmouth College; and Dani Rodrik of

Harvard University. The assistance of external reviewers implies no responsibility for the final product,

which rests solely with CBO.

Wendy Edelberg and Jeffrey Kling reviewed the report, Christine Bogusz and Gabe Waggoner edited

it, and Jeanine Rees prepared it for publication. Maureen Costantino designed the cover. An electronic

version is available on CBO’s website (www.cbo.gov/publication/51924).

Keith Hall

Director

September 2016