Environmental Engineering Reference
In-Depth Information
The Compustat database is widely used by
researchers who apply Fan & Lang's method in
their analysis. However, Compustat database is
not always consistent with the other databases
(e.g. Hoover or Mergent Online) or annual reports
of the companies because the companies do not
announce all the industries in which they operate.
This limitation makes it impossible to calculate
vertical integration for undeclared subsidiaries.
On the other hand, only public companies have
to declare the segments and segment sales cor-
rectly. The data about the private companies may
not be accurate and this limitation may cause
incorrect calculations. In this study, the sample
is composed of public companies; additionally,
we made crosschecked with the other databases
and completed the missing data from the Hoover
industry reports and Mergent Online database.
The Fan & Lang's method measures the rela-
tionship between the primary and the other seg-
ments of the company. The primary segment is
defined as the segment which has the highest sales;
the relationship between the primary segment and
others is weighted according to the sales of other
segments. Acemoglu et al. (2009) modified Fan
& Lang's method, used equal weights for each
segment, and examined the relationship between
all segments. As explained in the data section,
Compustat data is limited with 10 segments and
this causes inaccurate vertical integration calcu-
lation for the companies which are operating in
more than 10 different segments.
Other concerns about the Compustat database
are related to the definition of segment itself.
Because of the ambiguity in definition, some
firms may disclose the segments as an aggrega-
tion of a couple of unrelated segments (Davis and
Duhaime, 1992). Furthermore, they may change
the segments and number of segments in their
disclosed reports even if there is no change in
their operations (Denis et al., 1997). This may
cause incorrect allocation of industries to firms.
The Compustat database compiles the segment
ASE, NASDAQ, and OTC. Therefore, we can
say that Compustat limits the sample to publicly
traded companies. Additional concerns about the
Compustat database can be found in Davis and
Duhaime (1992), Denis, Denis, & Sarin (1997),
and Villalonga (2004).
CONCLUSION
This study compared the vertical integration level
of sustainable and non-sustainable companies.
Literature of natural-resource-based view (e.g.
Hart (1995) and Russo & Fouts (1997)) and
natural-transaction-cost-economics theory (e.g.
Carter & Carter (1998) and Finon & Perez (2007))
propose increasing the vertical integration level for
sustainable companies. Carter and Carter (1998)
measured the vertical coordination through the
supply chain with a survey and concluded that
vertical integration increase the environmental
performance of the companies. In the literature,
the linkage between environmental strategies and
vertical integration has not been examined with
an economy-based vertical integration index. This
study attempts to fill this gap by measuring the
vertical integration with Fan & Lang's method
and trying to understand if the sustainable com-
panies tend to be more vertically integrated than
their non-sustainable counterparts. The results
demonstrate that sustainability-focused companies
in the Health Care and Industrials industries tend
to have more vertically integrated organizational
structures than their industry non-sustainable com-
petitors. There was no significant difference in the
vertical integration level of sustainability-focused
versus non-sustainability-focused companies for
the other seven industries studied.
Higher vertical integration may not be pos-
sible under some industry, product, market, and
economic conditions. Under these circumstances,
sustainable companies may prefer to increase the
social capital with its suppliers to eliminate the
