An Empirical Examination of Capacity Costs

An Empirical Examination of Capacity Costs

Merle Ederhof

Venky Nagar

Madhav Rajan

Research Background

1. Excess capacity costs in management accounting

Evidence: HBS cases and small sample studies

2. How material are these effects?

3. Large sample evidence using confidential Census data of 151,900 plants from 1974 – 2011 spanning machinery to food and lumber

Research Questions

1. Are excess capacity costs material?

Answer: removing excess capacity costs can increase operating profit margins by 25 percent

2. Does volatility of profit margins decline after removing excess capacity costs?

Answer: Yes, std dev declines by 5 percent (significant)

3. Are there shadow costs and benefits of excess capacity?

Answer: Yes, but not large (consistent with point 2 above)

Table 1: Distribution of establishment sample across years

Year / N
1974 / 8,000
1975 / 7,600
1976 / 7,100
1977 / 6,700
1978 / 7,100
1979 / 8,400
1980 / 7,900
1981 / 7,600
1982 / 6,900
1983 / 6,400
1984 / 6,400
1985 / 6,700
1987 / 7,300
1992 / 9,100
1997 / 16,000
2002 / 15,100
2007 / 4,100
2008 / 3,400
2009 / 3,200
2010 / 3,500
2011 / 3,700
Total / 151,900

Table 2: Distribution of the establishment sample across industries within the manufacturing sector

2 SIC / Description / N
35 / Industrial Machinery & Equipment / 18,100
20 / Food & Kindred Products / 16,800
34 / Fabricated Metal Products / 15,000
36 / Electronic & Other Electric Equipment / 11,700
28 / Chemicals & Allied Products / 10,900
33 / Primary Metal Industries / 8,900
37 / Transportation Equipment / 8,300
23 / Apparel & Other Textile Products / 7,100
32 / Stone, Clay, & Glass Products / 7,100
24 / Lumber & Wood Products / 6,600
Total / 110,500

Graph 1: Size of establishments (i.e., plants) in the full sample

Graph 1a: Distribution of the log of producer inflated-adjusted total value of shipments

Graph 1b: Mean producer inflation-adjusted total value of shipments by year

Graph 2: Histogram of establishment capacity utilization

35 - Industrial Machinery & Equipment (N=18,100)
/ 20 – Food & Kindred Products
(N=16,800)
/ 34 – Fabricated Metal Products
(N=15,000)

36 – Electronic & Other Electric Equipment (N=11,700)
/ 28 – Chemicals & Allied Products
(N=10,900)
/ 33 – Primary Metal Industries
(N=8,900)

37 – Transportation Equipment (N=8,300)
/ 23 – Apparel & Other Textile Products (N=7,100)
/ 32 – Stone, Clay, & Glass Products (N=7,100)

24 – Lumber & Wood Products (N=6,600)

Graph 3: Establishment capacity utilization over time

35 - Industrial Machinery & Equipment (N=18,100)
/ 20 – Food & Kindred Products
(N=16,800)
/ 34 – Fabricated Metal Products
(N=15,000)

36 – Electronic & Other Electric Equipment (N=11,700)
/ 28 – Chemicals & Allied Products
(N=10,900)
/ 33 – Primary Metal Industries
(N=8,900)

37 – Transportation Equipment
(N=8,300)
/ 23 – Apparel & Other Textile Products (N=7,100)
/ 32 – Stone, Clay, & Glass Products
(N=7,100)

24 – Lumber & Wood Products (N=6,600)

Graph 4: Decomposition of total annual product cost into depreciation, labor, and material components for establishments over time

Graph 4b: Mean cost ratios of material (top), labor (middle), and depreciation (bottom) costs to total annual product costs across 10 size buckets

35 - Industrial Machinery & Equipment
/ 20 – Food & Kindred Products
/ 34 – Fabricated Metal Products

36 – Electronic & Other Electric Equipment
/ 28 – Chemicals & Allied Products
/ 33 – Primary Metal Industries

37 – Transportation Equipment
/ 23 – Apparel & Other Textile Products
/ 32 – Stone, Clay, & Glass Products

24 – Lumber & Wood Products

Graph 5: Ratio of estimates of establishment annual fixed costs to total costs

Graph 6: Adjusted annual product costs relative to full costs after removing various estimates of costs of unused capacity at the establishment level

Graph 7: Annual profit margins using full costs and adjusted product costs from Graph 6

Graph 8: Plots of (sales – cogs)/sales and net income/sales for COMPUSTAT firms

Table 3: Standard deviations of profit margins by industry and overall

Unadj. ln SD / Adj. ln SD / p-val
Ind 35 (N = 3,100) / -2.39 / -2.46 / 0.00
Ind 20 (N = 3,100) / -2.73 / -2.74 / 0.20
Ind 34 (N = 2,600) / -2.60 / -2.65 / 0.00
Ind 36 (N = 2,000) / -2.42 / -2.49 / 0.00
Ind 28 (N = 1,700) / -2.47 / -2.52 / 0.00
Ind 33 (N = 1,500) / -2.42 / -2.51 / 0.00
Ind 37 (N = 1,300) / -2.46 / -2.52 / 0.00
Ind 23 (N = 1,300) / -2.55 / -2.56 / 0.19
Ind 32 (N = 1,200) / -2.53 / -2.63 / 0.00
Ind 24 (N = 1,200) / -2.63 / -2.70 / 0.00
Sample (N = 26,200) / -2.57 / -2.62 / 0.00

Under Full Allocation of Capacity Costs, Is This Result Inevitable?

When production volume is low, more unused capacity gets allocated to each unit

When production volume is high, less unused capacity gets allocated to each unit

But this could be offset by variation in variable cost per unit

Table 4: Fixed effects regression of establishment output on inputs and capacity

Log(Q) / Log(Q) / Log(Q)
Capacity Utilization / 0.171*** / 0.233***
Log(equipment) / 0.073*** / 0.076*** / 0.055***
Log(equip) * caput / 0.029***
Log(structures) / 0.038*** / 0.041*** / 0.025***
Log(struct) * caput / 0.022***
Log(labor hours) / 0.267*** / 0.262*** / 0.280***
Log(hours) * caput / —0.024**
Log(materials) / 0.498*** / 0.494*** / 0.524***
Log(mat) * caput / —0.041***
Log(energy) / 0.108*** / 0.106*** / 0.101***
Log(energ) * caput / 0.006

Conclusion

1. Accounting is estimation

2. Are excess capacity costs material?

Answer: removing excess capacity costs can increase operating profit margins by 25 percent

3. Does volatility of profit margins decline after removing excess capacity costs?

Answer: Yes, std dev declines by 5 percent (significant)

4. Are there shadow costs and benefits of excess capacity?

Answer: Yes, but not large (consistent with point 3 above)