*This preview shows
page 1. Sign up to
view the full content.*

**Unformatted text preview: **t −1 + β5 ∗ Decrease Dummyi,t ∗ log Assets i,t
Revenue i,t
∗ log
Revenue i,t −1
Revenue i,t + β6 ∗ Decrease Dummyi,t ∗ log Employees i,t
Revenue i,t
∗ log
+ εi,t .
Revenue i,t −1
Revenue i,t The Decrease Dummyi,t variable is included in the last ﬁve terms in
model (III), meaning that these terms are activated for all periods when 60 M. C. ANDERSON, R. D. BANKER, AND S. N. JANAKIRAMAN TABLE 4
Results of Regressing Annual Changes in SG&A Costs on Annual Changes in Sales Revenue and
Determinants of Sticky Cost
Regression speciﬁcation:
log SG&Ai ,t
SG&Ai ,t −1 = β0 + β1 log Revenue i ,t
Revenue i ,t
+ β2 ∗ Decrease Dummyi ,t ∗ log
Revenue i ,t −1
Revenue i ,t −1 + β3 ∗ Decrease Dummyi ,t ∗ log Revenue i ,t
∗ Successive Decrease i ,t
Revenue i ,t −1 + β4 ∗ Decrease Dummyi ,t ∗ log Revenue i ,t
∗ Growth i ,t
Revenue i ,t −1 + β5 ∗ Decrease Dummyi ,t ∗ log Revenue i ,t
Assets i ,t
∗ log
Revenue i ,t −1
Revenue i ,t + β6 ∗ Decrease Dummyi ,t ∗ log Employees i ,t
Revenue i ,t
∗ log
Revenue i ,t −1
Revenue i ,t + εi ,t Decrease Dummy takes the value of 1 when revenue in period t is less than revenue in t − 1,
0 otherwise. Successive Decrease takes the value of 1 when revenue in period t − 1 is less than
revenue in t − 2, 0 otherwise. Growth is the percentage growth in real GNP during year t .
The reported t -statistics are based on White’s hetroskedasticity-corrected standard errors. The
random coefﬁcients model is estimated as described in table 3.
Coefﬁcient Estimates Predicted Sign
β0
β1 + β2 − β3 + β4 − β5 − β6 − Adjusted R2 Pooled Estimation
(t -statistic) Random Coefﬁcients
Estimation (t -statistic) 0.0546
(27.01)
0.5444
(56.44)
−0.2245
(−2.63)
0.2415
(8.30)
−0.0179
(−1.83)
−0.1496
(−11.38)
−0.0338
(−2.04) 0.0209
(13.23)
0.6699
(74.58)
−0.2514
(−5.59)
0.2227
(15.66)
−0.0070
(−1.78)
−0.0975
(−12.69)
−0.0143
(−1.71) 0.4103 revenue declined. As in model (I), where the degree of stickiness increases
ˆ
with the magnitude of the negative value of β 2 , the degree of stickiness increases (decreases) with the magnitude of negative (positive) coefﬁcients
ˆ
ˆ
β 2 through β 6 in model (III).
The results of estimating model (III) are presented in table 4. The estiˆ
mated coefﬁcient β 1 = 0.5444 is signiﬁcant and positive (t -statistic = 56.44)
and of similar magnitude as its value in the model (I) estimation of
ˆ
the pooled sample. The signiﬁcant and positive coefﬁcient β 3 = 0.2415 STICKY COSTS 61 (t -statistic = 8.30) indicates that the degree of stickiness is lower in revenuedeclining periods that were preceded by revenue-declining periods, consistent with hypothesis 3a that managers would consider a reduction in demand
that occurred in successive years to be more permanent. The signiﬁcant and
ˆ
negative coefﬁcient β 4 = − 0.0179 (t -statistic = −1.83) indicates that the degree of stickines...

View Full
Document