Table 1: Selected Production Function Studies
AUTHOR  EQUATION  DATA  ELASTICITY*  COMMENTS 
ASCHAUER (1989)  CobbDouglas production function and TFP regressions  time series 194985 Private Business Economy 
0.39 0.36 0.370.41 Significant 
Constant returns to scale (CRS) in all inputs, including public capital input 
MUNNELL (1990a)  CobbDouglas production function reproduces Aschauer  time series 19481987 private non farm sector  0.340.41 Significant 
CRS in all inputs; also priv. and publ. cap. coef. equal 
MUNNELL (1990b)  CobbDouglas production function  crosssect. time series 48 states 19701986  0.15  see Munnell 1991 and other references 
MUNNELL (1991)  CobbDouglas production function  crosssect. Average 19701986 states values 12 high endowm. 26 mid. endowm. 10 low endowm. 
0.14 0.11 0.22 Significant 
Returns to Scale 1.01 1.03 1.04 
GARCIAMILA AND McGUIRE (1988)  CobbDouglas production function  crosssect. time series 14 annual obs. of 48 states gross state prod. labor, capital expenditures on education and highways  Highways: 0.0450.044 Education: 0.160.072 Significant 
Returns to Scale 1.04 Cannot reject increasing returns to scale 
EBERTS (1988)  Translog production function  crosssect. manufacturing 19581978 38 Metropolitan Areas  0.04 Significant 
CRS; public and private capital substitutes public and labor complements 
HULTEN AND SCHWAB (1991a)  CobbDouglas production function with first differences.  time series 19491985 same as Aschauer  0.42 Significant 0.028 Insignificant 
() coeff. for labor 
TATOM (1991)  CobbDouglas production function including energy price, with first differences.  time series 19741987 Business Sect.  0.146 Insignificant 
CRS 
MERA (1972)  CobbDouglas production function  Japan pooled data of regions and time 3 sectors 4 classifications of social overhead capital 
0.22 0.20 (.50) 0.120.18 Significant 

FORD AND PORET (1991)  TFP regressions  USA and 11 OECD countries time series and country crosssection  Half of countries significant effect after 1960 Mixed support of Aschauer results 

HULTEN AND SCHWAB (1991b)  TFP regressions  cross section time series regional study of SnowSun Belt 19701986 Gross output value added  public capital insignif. in all regressions private capital insignif. in gross output regres. signif. in value added implying scale .88 
Table 2: Cost or Profit Function Estimates
DESCRIPTION  DIRECT EFFECT  INDIRECT EFFECT  
Author  Unit of Analysis 
Specification  Public Capital  Cost  Labor  Capital  Intermediate 
BERNDT AND HANSSON (1991)  Sweden Private Sector 19601988  Variable Cost Labor Requirement Function  Core Public Capital  Cost Savings Unclear  Shortrun complements     
DENO (1988)  USA 36 SMSA Manufacturing Industries 197078 Pooled 
Profit Truncated Translog  Highway, Water and Sewer Adjusted with the proportion of population employed by the sector  Profit increase Elasticity = .08 to .5 
Gross complements Elasticity = 0.1 to .4 
Gross complements Elasticity = 0.11 to .4 
 
CONRAD AND SEITZ (1992)  West Germany Manufacturing Construction, Trade and Transport 19601988 TimeSeries  Cost Translog and MR=MC  Total Adjusted with capacity utilization rate  Cost Savings  Substitutes  Complements  Substitutes 
KEELER AND YING (1988)  USA Trucking Industry 19601988 Regional Pooled  Cost Translog  Highway Stock  Cost Savings       
LYNDE AND RICHMOND (1992)  USA Nonfinancial Corporate Business Sector 19581989 TimeSeries  Cost Translog P = MC and CRS 
Total Federal and State  Cost Savings  Substitutes Elasticity = .45 to .49 
Complements Elasticity = .71 to .90   
LYNDE AND RICHMOND (1993)  U.K. Manufacturing sector 1966:1 to 1992:2 value added  Cost translog  Total  Cost Savings    Substitutes   
MORRISON AND SCHWARTZ (1991)  USA Manufacturing by State 19711987 Pooled by Region State specific Effects  Variable Cost Generalized Leontief P=MC  Core  Cost Savings Elasticity = .10 to .27       
DESCRIPTION  DIRECT EFFECT  INDIRECT EFFECT  
Author  Unit of Analysis 
Specification  Public Capital  Cost  Labor  Capital  Intermediate 
NADIRI AND MAMUNEAS (1991)  USA Manufacturing 12 2digit industries 19551986 Pooled Industry Specific Effects  Cost Translog CRS for Private Inputs  Total Stock Adjusted with Capacity Utilization Rate  Cost Savings Elasticity = 0 to .21 
Substitutes Elasticity = 0 to 1.4 
Substitutes Elasticity = .02 to 1.4 
Complements Elasticity =.12 to .76 
SEITZ (1992a)  West Germany 31 2digit Industries 19701989 Pooled Industry Specific Effects  Cost Generalized Leontief  Public Roads Length of Motorway System 
Cost Savings  Substitutes Elasticity = .0004 
Complements Elasticity = .03 to .04 
 
SEITZ (1992b)  West Germany 31 2digit Industries 19701989 Pooled Industry Specific Effects  Cost Generalized Leontief  Total Core 
Cost Savings  Substitutes Elasticity = .15 to .13 
Complements Elasticity = .34 to .86 
 
SHAH (1992)  Mexican Manufacturing Sector 26 3digit Industries Pooled  Variable Cost Translog  Total Adjusted with industries' output proportion  Cost Savings  Complements Elasticity = .006 
Complements Elasticity = .002 
Substitutes Elasticity = .005 