An analysis of the cause and magnitude of foreign trade zone‐induced biases in US import statistics

• DOI: http://doi.org/10.1111/twec.12575
• Published date: 01 January 2018
• Author: Alexander J. Yeats
• Date: 01 January 2018

ORIGINAL ARTICLE

An analysis of the cause and magnitude of foreign

trade zone-induced biases in US import statistics

Alexander J. Yeats

Former Principal Economist with the World Bank (retired)

1

|

INTRODUCTION

Traditionally, import tariffs incorporated a similar structure across countries and industries. Zero,

or relatively low duties, is normally applied to materials used as production inputs with tariffs

increasing or “escalating”at higher levels of fabrication. This structure was intended to give

domestic manufacturers a competitive advantage over foreign suppliers and also allow them to

achieve higher rates of return.

1

However, several multilateral trade negotiations (MTNs) such as

the Tokyo and Uruguay Rounds caused some deviations from this normal pattern. On occasion,

the MTNs produced “inverted”structures which applied relatively higher tariffs to product ion

inputs than those on intermediate products or goods for consumption. This raised manufacturing

costs and disadvantaged domestic producers vis-

a-vis foreign competitors.

To deal with “tariff inversion”and related problems, the US utilises an extensive internal net-

work of foreign trade zones (FTZs).

2

Imports into the zones are exempt from tariffs until they are

forwarded, either in their original or a more highly processed form, for domestic consumption. At

this point, importers have a unique customs option of declaring the country originating the initial

import to be its source with its identity being either the original imported good or the product

manufactured within the zone. If tariffs are “inverted,”the duty would be lower on the processed

FTZ product, so this item would be declared on customs invoices with its origin listed as the for-

eign supplier of the material input. In the case of “escalating”tariffs, the original import would be

declared and not that of the item(s) actually forwarded through customs for consumption.

3

Stated

1

These issues are often discussed using the concept of “effective”tariff protection (see Finger and Yeats, 1976; Grubel and

Johnson, 1971; or Michaely, 1977).

2

Foreign trade zones are generally acknowledged to make important contributions to the facilitation of production and trade

through several cost-saving measures and the modification of restrictive regulations. The former may include deferments on

applicable tariffs and taxes on imports until goods are transferred from the zones to domestic markets for consumption. See

Bolle and Williams (2012) for a useful comprehensive discussion of current zone operations, or the internet site of the For-

eign Trade Zone Resource Center for a history of the development of the US trade zones.

3

Several aspects of this procedure are highly problematic. First, goods may be classified as imports even when key elements

of their production occurred within US geographic territory where local domestic resources and technical expertise are cen-

tral to their manufacture. Second, this paper will show the reported origins of the consumption goods may be countries

whose export statistics do not record any matched shipments of the products to the US or any other destinations, and who,

according to production capability surveys, may not have the capacity or resources required for their manufacture.

DOI: 10.1111/twec.12575

World Econ. 2018;41:143–170. wileyonlinelibrary.com/journal/twec ©2017 John Wiley & Sons Ltd

|

143

differently, existing zone regulations can provide an important incentive to misstate the true natur e

of products actually imported and the true source of the good clearing customs.

The possible incorporation of invoicing biases in dutiable import statistics raises several impor-

tant analytical considerations. First, accurate import statistics are an essential input for many

diverse empirical trade studies ranging from analyses of the level and structure of trade barriers to

the derivation of measures of a countries “openness”to imports (such as import penetration ratios).

In recognition of the point, the World Trade Organization’s (WTO) Integrated Database makes

detailed matched tariff and import statistics available to member countries for these initiatives. Sec-

ond, attempts to project trade gains resulting from tariff or NTB reductions require accurate

detailed dutiable import statistics as their empirical base. These projections may play a key role in

the formulation of a country’s overall strategy in trade negotiations, as well as the assignment of

priorities for specific sectors or products. Since inaccuracies in essential import statistics can seri-

ously misdirect these liberalisation initiatives, this analysis will attempt to assess the overall magni-

tude of induced trade zone biases and also identify the industry sectors primarily affected.

2

|

CHARACTERISTICS OF FOREIGN TRADE ZONE

IMPORTS

Any analysis of FTZ-induced effects on trade statistics should examine factors like the overall

magnitude of zone imports, their rate of change relative to traditional trade and the diversity of

product sectors and foreign suppliers involved. While only limited information previously existed,

a major data source relating to these points recently became available. Specifically, the Interna-

tional Trade Commission now provides access to detailed data on US zone imports at its Internet

site (see www.USITC.gov) along with an extraction tool called the “trade data web”for compila-

tion of this information. The “web”can be used to analyse zone imports from partner countries at

the tariff line level from the mid-2000s to the present. Users may compile longer “historical”

records back to 1989, but these statistics may not be comparable from year to year due to changes

in the underlying classification system.

Table 1 utilised the USITC “data web”to compile statistics on zone imp orts both in total and

at the two and four-digit harmonised system level. Each HS group’s primary supplier and its

import share are shown along with the total number of countries which originated imports. In addi-

tion, a “product coverage ratio”showing the percentage of underlying six-digit HS groups in

which zone imports occur is also given. For example, in 2014 the zones imported $29.7 billion in

transportation equipment from 46 partner countries with Germany being the single largest supplier

sourcing 55% of the total. According to the product coverage ratio, zone imports occurred in 70%

of the underlying six-digit products. Overall, the data indicate 2014 zone imports of all products

totalled $299 billion which represented an increase of about $138 billion over their 2005 base

value. To put these numbers in perspective, at current levels zone imports are more than double all

US imports for consumption from Germany ($124 billion) or Japan ($137 billion), and only about

one-third less than similar imports ($398 billion) from all European Union 15 member countries.

With regard to current imports, petroleum oils, transportation equipment plus electrica l and

non-electrical machinery account for approximately 85% of all shipments into the zones. Petroleum

imports (HS 27) of $169 billion originated in 61 countries with Saudi Arabia sourcing 25% of this

exchange. Incentives to avoid extensive escalating and inverted tariffs on petroleum derivatives,

which include a broad range of distillate fuels and organic chemicals, were o ne factor responsible

for the magnitude of zone crude oil imports.

144

|

YEATS