[4_1]_2003_Anderson_et_al_Are_selling,_general_and_administrative_costs_sticky

Costs form the basis for the reported numbers in the

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: negative change in sales revenue form the basis for the reported numbers in the first row and observations with a negative change in SG&A costs form the basis for the reported numbers in the second row. Panel A: Distribution of Annual Revenue and SG&A Costs from 1979 to 1998 Mean Sales revenue Selling, general, and administrative (SG&A) costs SG&A costs as a percentage of revenue Standard Deviation Median Lower Quartile Upper Quartile $1277.09 $229.45 $5983.43 $1042.49 $87.53 $17.49 $17.51 $4.56 $447.75 $79.12 26.41% 17.79% 22.62% 13.66% 34.31% Panel B: Periodic Fluctuations in Revenue and SG&A Costs from 1979 to 1998 Percentage of Standard Upper Lower Firm-Years Mean Deviation of Median Quartile of Quartile of with Negative Percentage Percentage Percentage Percentage Percentage Percentage Decrease Decreases Decrease Decreases Decreases Change from Across Across Across Across Across Previous Period Periods Periods Periods Periods Periods Sales revenue SG&A costs 27.01% 17.45% 18.64% 10.99% 23.76% 4.38% 24.98% 15.67% 16.40% 10.07% 21.63% 3.94% We present the results of estimating model (I) for the pooled sample in table 2. The results were similar when we estimated a fixed-effects model. We initially estimated the model with changes in SG&A costs and sales revˆ enue defined for one-year periods. The estimated value of β 1 of 0.5459 (t -statistic = 164.11) indicates that SG&A costs increased 0.55% per 1% inˆ crease in sales revenues. The estimated value of β 2 = −0.1914 (t -statistic = −26.14) provides strong support for the sticky costs hypothesis. The comˆ ˆ bined value of β 1 + β 2 = 0.3545 indicates that SG&A costs decreased only ˆ ˆ ˆ 0.35% per 1% decrease in sales revenue. The fact that β 1 and β 1 + β 2 are both significantly less than one (p -values = 0.001) indicates that SG&A costs were not proportional to changes in revenue, even though this cost driver is apparently strong.5 For comparative purposes, we also estimated a model 5 Noreen and Soderstrom [1994] find that overhead costs at hospitals in Washington State were not proportional to activity. In a related study, Noreen and Soderstrom [1997] find that the average variation of overhead was about 20% of the variation in the activity driver. They suggest that the low percentage may reflect maintenance of specific capacities by hospitals. 55 STICKY COSTS TABLE 2 Results of Regressing Changes in SG&A on Changes in Sales Revenue for the 20-Year Period 1979–98 Regression specification for model (I): log SG&Ai ,t SG&Ai ,t −1 = β0 + β1 log Revenue i ,t Revenue i ,t −1 + β2 ∗ Decrease Dummyi ,t ∗ log Revenue i ,t Revenue i ,t −1 + εi ,t Regression specification for model (II): log SG&Ai ,t SG&Ai ,t −1 = β0 + β1 log + β3 log Revenue i ,t Revenue i ,t + β2 Decrease Dummyi ,t ∗ log Revenue i ,t −1 Revenue i ,t −1 Revenue i ,t −1 Revenue i ,t −1 + β4 Decrease Dummyi ,t −1 ∗ log +...
View Full Document

Ask a homework question - tutors are online