Identifying firm heterogeneity in value added and trade for U.S. businesses.
Fetzer, James J. ; Strassner, Erich H.
This article discusses a new economic accounting framework for the
United States that describes the production and sourcing patterns for
different types of firms in an industry. The estimates show that
domestic value added by an industry as a share of output is lower for
globally engaged firms, compared with domestically oriented firms, and
that exports and imports of goods as a share of output are larger for
globally engaged firms.
THE GROWTH OF COMPLEX and increasingly fragmented supply chains in
the global economy has increased trade in intermediate goods and
services, some of which return to the sending country for final sale or
for further processing. These changes have increased the interest in
measuring the degree to which gross trade values reflect value added
from the trading partner or from intermediate inputs sourced from other
countries, including the home country. For example, Johnson and Noguera
(2012) estimate that as much as two-thirds of gross trade is comprised
of intermediate goods that pass a given customs frontier multiple times
before becoming final goods and reaching their ultimate consumer. They
also estimate that when bilateral trade flows are adjusted to include
only trade in value added (TiVA), the United States-China bilateral
trade deficit for goods and services is about 30 percent to 40 percent
smaller and the United States-Japan deficit is about 33 percent larger.
However, national statistical agencies have found that directly
measuring the degree to which trade flows contain value added from the
immediate partner country, as opposed to intermediate inputs from other
countries, is very difficult. The most popular approach to estimating
TiVA is by using input-output (I-O) coefficients for the home country
and for its major trading partners. The coefficients measure, for a
given industry, the amount of each industry's output that is
required as an intermediate input to produce one unit of output in that
industry. These coefficients come from a global supply-use table (SUT),
which decomposes total output for an industry into value added by firms
in that industry and intermediate inputs that are sourced domestically
or imported. There are multiple international efforts to develop the
statistical infrastructure to develop TiVA estimates, including efforts
by the Organisation for Economic Co-operatkon and Development, the Asia
Pacific Economic Cooperation, the World Trade Organization, and the
United Nations Statistics Division. (1)
The underlying economic accounting framework in these efforts is
the extended SUT. The key feature of this type of table that
distinguishes it from a traditional SUT is the explicit accounting for
firm-level heterogeneity. In a traditional SUT, firm heterogeneity is
captured only by the disaggregation by industry. The extended SUT
recognizes that even within an industry, there can be important firm
heterogeneity. A consensus has not yet emerged in the literature on the
most relevant firm characteristics to capture firm heterogeneity, but
the Bureau of Economic Analysis (BEA) has chosen to focus on the degree
of a firm's global engagement. (2) Researchers such as Melitz
(2003) have demonstrated that globally engaged firms are more productive
than their domestically oriented peers and have production functions
that are significantly more reliant on exports and imports. In this
research, globally engaged firms are U.S. businesses that are part of a
U.S.-based or foreign-based multinational enterprise (MNE). Roughly 90
percent of trade in goods (Bernard, Jensen, and Schott (2009)) and more
than 80 percent of trade in services (Barefoot and Koncz-Bruner (2012))
involve multinational enterprises. Accounting for firm-level
heterogeneity will yield more informative TiVA and global value chain
measures, as the aggregate production functions of industries in
national economy SUTs are decomposed into relevant sets of firm
characteristics and as differences in TiVA are ascribed to these
characteristics.
This Research Spotlight presents preliminary estimates that compare
components of total output by industry and by firms' level of
global engagement. The estimates show heterogeneity in the composition
of output among different types of firms. They allow us to answer some
important questions about the effects of globalization on the U.S.
economy, such as the following:
* Which industries are more involved in global value chains'
* How much do MNEs contribute to the domestic economy by purchasing
intermediate goods and services from other U.S. businesses'
* How much value do MNEs add to U.S. production'
* What types of firms are more likely to import and export'
Similar to this study for the United States, research that has
analyzed the components of value added for other countries within a SUT
framework has shown evidence of firm heterogeneity (for a broad group of
countries, see Piacentini and Fortanier (2015), for Turkey, see Ahmad,
Araujo, Lo Turco, and Maggioni (2013), and for China, see Ma, Wang, and
Zhu (2015)). These studies found that value added by foreign-owned MNEs
is a smaller share of output or sales than value added by domestic-owned
firms in many, but not all, of the countries studied. Exceptions include
only a few countries, such as Finland, Turkey, and for production of
processing exports, China. (3) Exports and imports as a share of output
or turnover are typically larger on average for foreign-owned MNEs than
for domestic-owned firms.
The estimates presented for the United States rely on source data
from BEA and from other sources for 2011. Internal Revenue Service
Statistics of Income (SOI) data are used to account for all firms with
operations in the United States. All data for MNEs are based on
comprehensive data on U.S. direct investment abroad and foreign direct
investment in the United States that are collected from mandatory
surveys conducted by BEA. (4) These data cover both the operations of
MNEs with headquarters in the United States (domestic-owned MNEs) and
U.S. affiliates of MNEs with headquarters abroad (foreign-owned MNEs).
Results for domestic-owned non-MNEs are computed as the difference
between the SOI data for all U.S. firms less the results for
domestic-owned and foreign-owned MNEs. SOI data are used instead of data
from BEA I-O tables because the SOI data are collected and published by
industry at the enterprise level, similar to the BEA MNE data.
Results
The components of total output show heterogeneity in value added,
imports, and exports as a share of output between the three types of
firms: domestic-owned MNEs, foreign-owned MNEs, and non-MNEs. The
patterns of the heterogeneity are consistent with other research.
In most broad industry groups, value added as a share of output is
smallest for foreign-owned MNEs and largest for non-MNEs (tables 1 and 2
and chart 1). In particular, domestic value added as a share of output
is
* Smaller for foreign-owned MNEs than for both domestic-owned MNEs
and non-MNEs.
* About 10 percentage points larger on average for domestic-owned
MNEs than for foreign-owned MNEs.
* Larger for domestic-owned MNEs, compared with foreign-owned MNEs
in almost all industries, although the share varies across industries.
* At least 50 percent larger on average for non-MNEs than for MNEs.
* Larger in labor-intensive industries such as health care and
social assistance.
* About 6-7 percentage points smaller on average for MNEs that
export goods than those that do not, although this varies by industry.
Smaller for foreign-owned firms, compared with domestic-owned
firms.
Consistent with earlier BEA studies, such as Zeile (1998), imports
and exports as a share of output are larger on average for the
foreign-owned MNEs, compared with domestic-owned firms, though these
shares vary by industry (chart 2). In particular,
* Imports and exports as a share of output are larger for
foreign-owned firms than domestic-owned firms.
* Imports as a share of output are 2-3 times larger on average for
foreign-owned MNEs.
* Exports as a share of output are largest for foreign owned MNEs
(9 percent) and smallest for non-MNEs (3 percent).
[ILLUSTRATION OMITTED]
* Imports as a share of intermediates are larger on average (23
percent) for foreign-owned MNEs than for domestic-owned MNEs and
non-MNEs (9 percent each). (5)
These preliminary results raise a number of interesting research
questions, such as the following:
* Why is value added as a share of output for foreignowned MNEs
typically smaller than that for domestic-owned MNEs'
* Why is value added as a share of output larger for domestic-owned
MNEs in manufacturing than for non-MNEs'
* What are the differences in the structure of production by
different types of firms'
* How do different types of firms engage in global value
chains'
Future research will lead to refinements in the measurement of
extended SUTs and new insights about the role of MNEs in the U.S.
economy. BEA and the U.S. International Trade Commission are currently
collaborating to produce more detailed estimates that are disaggregated
by firm type using the methodology in this paper. Looking ahead, BEA and
the Census Bureau will undertake research at the Census Bureau Center
for Economic Studies with the eventual goal of developing official
extended SUTs and other statistics that more fully measure the impact of
global value chains on the U.S. economy.
[ILLUSTRATION OMITTED]
References
Ahmad, Nadim, Sonia Araujo, Alessa Lo Turco, and Daniela Maggioni.
2013. "Using Trade Microdata To Improve Trade in Value-Added
Measures: Proof of Concept Using Turkish Data." In Trade in Value
Added: Developing New Measures of Cross-Border Trade, edited by Aaditya
Mattoo, Zhi Wang, and Shang-Jin, 187-219. Washington, DC: The World
Bank.
Barefoot, Kevin, and Jennifer Koncz-Bruner. 2012. "A Profile
of U.S. Exporters and Importers of Services." SURVEY OF CURRENT
BUSINESS 92 (June): 66-87.
Bernard, Andrew B., J. Bradford Jensen, and Peter K. Schott. 2009.
"Importers, Exporters, and Multinationals." In Producer
Dynamics: New Evidence from Microdata, edited by Timothy Dunne, J.
Bradford Jensen, and Mark J. Roberts, 513-551. Cambridge, MA: University
of Chicago Press, for the National Bureau of Economic Research.
Johnson, Robert C., and Guillermo Noguera. 2012. "Accounting
for Intermediates: Production Sharing and Trade in Value Added."
Journal of International Economics 86 (March): 224-236.
Ma, Hong, Zhi Wang, and Kunfu Zhu. 2015. "Domestic Content in
China's Exports and Its Distribution by Firm Ownership."
Journal of Comparative Economics 43 (February): 3-18.
Melitz, Marc J. 2003. "The Impact of Trade on Intraindustry
Reallocations and Aggregate Industry Productivity." Econometrica 71
(November): 1695-1725.
Ismaylov, Rita, and Ricardo Limes. 2015. "Activities of U.S.
Affiliates of Foreign Multinational Enterprises in 2013." SURVEY OF
CURRENT BUSINESS 95 (November).
Piacentini, Mario, and Fabienne Fortanier. 2015. "Firm
Heterogeneity and Trade in Value Added." Organisation for Economic
Co-operation and Development Working Paper.
Scott, Sarah P. 2015. "Activities of U.S. Multinational
Enterprises in 2013." SURVEY OF CURRENT BUSINESS 95 (August).
Zeile, William. 1998. "Imported Inputs and the Domestic
Content of Production by Foreign-Owned Manufacturing Affiliates in the
United States." In Geography and Ownership as Bases for Economic
Accounting, edited by Robert E. Baldwin, Robert E. Lipsey, and J. David
Richardson, 205-232. Cambridge, MA: University of Chicago Press, for the
National Bureau of Economic Research.
By James J. Fetzer and Erich H. Strassner
(1) For more information, see the papers presented at the
International Conference on Measurement of Trade and Economic
Globalization.
(2) For details about this research, see "Identifying
Heterogeneity in the Production Components of Globally Engaged Business
Enterprises in the United States."
(3) Processing exports are part of China's processing trade
regime, which exempts materials imported for further processing and
reexported from import duties. Value added is a larger share of output
for processing exports produced by foreign-owned firms than processing
exports produced by Chinese-owned firms because foreign-owned processing
exports rely much less on Chinese produced inputs. See table 3 in Ma,
Wang, and Zhu (2015).
(4) See Scott (2015) and Ismaylov and Limes (2015) for recent
descriptions of these data for 2013.
(5) The MNE data on imports do not provide information on whether
the imports are used as intermediate inputs or sold as final demand in
the economy. The assumption made that total imports are consumed as
intermediate goods is most plausible for manufacturing in which imports
make up 25 percent of intermediate inputs for foreign-owned MNEs,
compared with 18 percent for domestic-owned MNEs and 1 percent for
non-MNEs.
Table 1. Use Table for All Private Industries by Type of Firm, 2011
[Percentage of total output]
Domestic-owned Foreign-owned
MNEs MNEs
Domestic-owned MNEs n.a. n.a.
Foreign-owned MNEs n.a. n.a.
Non-MNEs n.a. n.a.
Total domestic intermediate
consumption and imports of
services 63 61
Total imports of goods 7 18
Total intermediate consumption 69 79
Value added 31 21
of which:
Compensation of employees 17 12
Gross operating surplus 12 7
of which:
Consumption of fixed capital 4 3
Taxes on production and imports 2 2
Total output 100 100
Non-MNEs Exports Other
of goods uses
Domestic-owned MNEs n.a. 6 94
Foreign-owned MNEs n.a. 9 91
Non-MNEs n.a. 3 97
Total domestic intermediate
consumption and imports of
services 48
Total imports of goods 5
Total intermediate consumption 53
Value added 47
of which:
Compensation of employees 22
Gross operating surplus 22
of which:
Consumption of fixed capital 5
Taxes on production and imports 4
Total output 100
n.a. Not available
MNEs Multinational enterprises
NOTE. The preliminary estimates presented in this table are provisional
and are intended only for discussion and to illustrate the types of
analysis that can be performed with this framework.
Table 2. Share of Output Accounted for by Value Added, Employee
Compensation, and Gross Operating Surplus by Type of Firm for Selected
Industries, 2011
[Percentage of total output]
Value added's share
of output
Domestic- Foreign- Non-MNEs
owned owned
MNEs MNEs
All industries 31 21 47
Manufacturing 29 22 23
Food, beverages, and tobacco
products 28 27 28
Chemicals 39 25 49
Fabricated metal products 32 28 30
Wholesale trade 14 8 20
Finance and insurance 23 23 102
Insurance carriers and related
activities 9 16 63
Professional, scientific, and
technical services 63 39 76
Mining 57 52 59
Transportation and warehousing 48 29 64
Health care and social assistance 65 57 89
Employee compensation's
share of output
Domestic- Foreign- Non-MNEs
owned owned
MNEs MNEs
All industries 17 12 22
Manufacturing 14 12 5
Food, beverages, and tobacco
products 11 12 35
Chemicals 17 12 -7
Fabricated metal products 21 19 28
Wholesale trade 7 5 10
Finance and insurance 19 18 19
Insurance carriers and related
activities 11 11 11
Professional, scientific, and
technical services 39 30 42
Mining 13 18 10
Transportation and warehousing 27 19 24
Health care and social assistance 46 38 87
Gross operating surplus's
share of output
Domestic- Foreign- Non-MNEs
owned owne
MNEs MNEs
All industries 12 7 22
Manufacturing 13 8 14
Food, beverages, and tobacco
products 14 12 -13
Chemicals 21 13 47
Fabricated metal products 10 8 -1
Wholesale trade 6 2 8
Finance and insurance 3 4 77
Insurance carriers and related
activities -3 3 48
Professional, scientific, and
technical services 22 8 31
Mining 38 30 45
Transportation and warehousing 18 9 38
Health care and social assistance 17 18 -1
MNEs Multinational enterprises
NOTES. The preliminary estimates presented in this table are
provisional and are intended only for discussion and to illustrate the
types of analysis that can be performed with this framework. Some
values for employee compensation and gross operating surplus for
non-MNE firms are negative because of differences in the data sources
used to calculate the values for non-MNE firms.
