Ch. 8: Conglomerate Firms and Internal Capital Markets 469
Schoar (2002)also used the LRD plant-level data to examine productivity of con-
glomerate firms and changes in productivity following plant acquisitions.Schoar (2002)
establishes that market valuations of single-segment and conglomerate firms track es-
timates of productivity derived from LRD data. The tracking is equally strong for
single-segment and conglomerate firms. This suggests that the conglomerate discount,
if it exists, is unlikely to be caused by investors’ inability to evaluate diversified firms’
operations as efficiently as those of single-segment firms.
Schoar also finds no evidence that conglomerates’ plants are less efficient than those
of single-segment firms. Specifically, using plant-level data she runs the following re-
gression
TFP=a+b∗DIV+c∗(plant size)+d∗(plant age),where TFP is total factor productivity and DIV is a dummy that takes on a value of 1
if the plant belongs to a diversified firm and zero otherwise. The coefficient of DIV is
positive and significant and remains so when the equation is augmented by segment-
level control variables.
LikeMaksimovic and Phillips (2001), Schoar finds that acquired plants on average
increase in productivity while the acquirer’s own plants decline in productivity. She calls
this the “new toy” effect, and argues that post-acquisition productivity of the acquirer
is on balance negative. However, as Schoar points out this time-series effect does not
cancel out the cross-sectional finding that diversified firms’ plants have a higher TFP.
An intriguing possibility raised by Schoar’s work is that a diversification discount
may arise because conglomerates pay out a higher proportion of their revenues in
salaries and benefits than standalone firms. She finds that diversified firms pay higher
hourly wage rates than similar standalone firms. Assuming that these differences do not
reflect differences in the educational level or quality of their respective workforces, the
wage difference is enough to explain a 2–3% discount for diversified firms.
4.6.2. Spinoffs
Several studies, includingGertner, Powers and Scharfstein (2002), Dittmar and Shiv-
asani (2003), Burch and Nanda (2003), Anh and Denis (2004), andColak and Whited
(2005), examine spinoff and divestiture decisions that reduce the number of divisions
that a conglomerate firm operates. These papers examine the investment efficiency of
firms before and after the refocusing decision. This approach has potential advantages
over studies that examine a sample of firms, some of which refocus and some which
do not. If it can be assumed that the severity of measurement error does not change
over time, measurement error bias that in the comparison of before and after refocusing
performance, is mitigated. These papers further argue that they have reduced omitted
variables bias by focusing on changes in value and efficiency in a single sample of
firms.
Gertner, Powers and Scharfstein (2002) examine sensitivity of segment invest-
ment to the median Tobin’sqof the single-segment firms in that segment’s industry.