A Critique of Empirical Studies of Relations Between Market Structure and Profitability Phillips, Almarin (1976), Journal of Industrial Economics, 24 (4): BADM 546, Group #1 (Meredith Blumthal, Wooje Cho, Barclay James, and Kumar Sarangee)
2 The Purpose of the Study To examine the previous empirical studies of relations between market structure and profitability Model Specification Problems Data and Measurement Problems Econometric Identification Problems
3 The Purpose of the Study Background: There have been a number of empirical studies of the relations, and the agreed argument from them is that a positive relationship exists between industrial concentration and profitability.
4 The Empirical Model The equation in question in the studies: (1) Profitability: various measures Market Structure CR: a measure of market concentration (supply) B: one or more measures of entry barriers (supply) D: the level (or rate of change) in market demand (demand)
5 Theoretical Problems Critiques to the variable, CR CR is not very sensitive to variation in N (# of firms) CR4 CR4 is a good proxy for a small N with higher value (near 1.00), but, for a large N, fails to distinguish the different performance come from different Ns (e.g. CR4 =.6 when N=7 or N=70) Fails to distinguish the relative size of the four largest firms (e.g. CR4 =.6 with.15,.15,.15,.15 vs. CR4 =.6 with ,.01,.01: source: lecture slide set #4) H (Herfindahl measure) Can solve a problem of CR4 But, still cannot tell the different performance come from different Ns (e.g. H=.2 with N=5 vs H=.2 with N=9) Since concentration can vary widely for any given N, the behavioral theories as well as the structural theories raise questions as to whether the equation (1) is properly specified
6 Theoretical Problems Critiques for the variable, B Entry barriers are a necessary condition, but not a sufficient condition for long-run economic profitability Once N reaches some appreciable number, the partial effect of B should disappear, because rivalry among existing firms replaces B. The model in equation (1) is rarely specified so as to such problems in B. Nothing in equation (1) approaches the difference between consumer and producer goods industries.
7 Data Problems: Two ways of measuring the profit variables The price cost margin method: (PCM: margin, VA: value added, W: wage and salaries, S: value of shipments for a census-defined industry) The rate of return on capital method Return on stockholders ’ equity, or Return on stockholders ’ equity plus long-term debit, or Return on assets
8 Data Problems-1 Critiques to the price cost margin method: VA-W comprise not just accounting profits, but depreciation, indirect payment, rent and the un- tabulated input purchases. S may be uncorrected for changes in inventories PCM is not a close substitute for the economic concept of markup of price over marginal costs Critiques to the rate of return on capital method: Calculated rate of return could be significantly different by various depreciation practices. Cannot distinguish expenditure in advertising and R&D
9 Data Problems-2 Critiques to CR measurement Problems in market definition come from: Product differentiation Geographic locational differences Critiques to B measurement A/S (adverting-sales ratio): absence of consistent information on full promotional expenditures C/O (capital-output ratio): low C/O for entry, but high C/O for efficient operation MES (minimum efficient scale of plant): problem in industry definition C/O seem to be preferred as an entry barrier measure Critiques to D measurement Uses the change in output over some period of time as a proxy for D: the change can occur from factors on the supply side
10 Other Problems Problems from using weighted average in regression The data for the largest firms control the observations The result can be biased if there is correlation b/w size and the dependent variable or any of the independent variables Problems from assumption of linearity in the equation (1)
11 Econometric Identification Problems Not clear whether profitability is high because of the existence of monopoly power or whether the firms are large and the market is concentrated because of differences in production efficiencies, leading to high profitability It is not clear how to interpret the data. Both monopoly explanations and efficiency explanations are consistent with the data).
12 Conclusions Market concentration may provide market power and thus be the cause of higher profitability, or, conversely, market concentration and higher profitability may be the result of superior capabilities and economic efficiencies by these firms. Better theory, better data, and above all, better econometrics are needed for better understanding relations between market structure and profitability