Comments on Competition Policy and Productivity Growth Jonathan

Comments on “Competition Policy and Productivity Growth” Jonathan B. Baker American University Washington College of Law June 25, 2009

Topics n n n Significance of the research Measurement issues Specification of the model Interpretation of the results Additional analyses

Significance n n Important question Results are plausible Results are potentially important Does this evidence change prior beliefs?

Measurement Issues n Key variables n Total factor productivity (TFP) growth n n for a country, industry, year Competition policy indicators (CPI) n for a country, year

Source: BCDSV presentation, slide 12

Measurement n Measurement error in TFP growth Robustness test uses labor productivity (1) y = Akθ (2) △y/y = △A/A + θ△k/k where TFP growth = △A/A n n Measurement error in CPI

Specification of the Model n n Appropriability Technological Opportunity

Interpretation of the Results n ∆ TFP growth ≈ 0. 01% ∆ CPI n n n If CPI increases by 20%, TFP growth rate increases by about 0. 02 n n Mean TFP growth ≈ 0. 01 (std dev ≈ 0. 07) Mean (aggregate) CPI ≈ 0. 5 At the mean, from 0. 01% to 0. 03% Is this magnitude economically significant?

Additional Analyses? n Does CPI coefficient differ for: n n Rapidly growing industries? Industries with substantial imports? Patent intensive industries? More highly concentrated industries?

Conclusion Effect of CPI on TFP growth n n n plausible robust hard to measure small may be biased downward Is this evidence enough to change prior beliefs?