Critique Of Arnold Harberger’s Article On Monopoly And Resource Allocation

Arnold Harberger, in his famous article called “Monopoly and Resource Allocation” innovated how Economists viewed deadweight loss, as well as how monopoly analysis was being done in general. His article was a breakthrough in monopoly research and made empiricism the ruling system of analysis. Throughout his article, Harberger applied his theories regarding deadweight loss on the manufacturing industry in the late twenties and came to a surprising conclusion. According to his findings, due to the low value of the deadweight loss compared to the national income, analysis of the economic process and the allocation of resources can be done accurately by neglecting monopoly elements. How warranted is Harberger’s conclusion however? Is that deadweight loss as low and conservative as he claims it to be to neglect monopoly elements?

Further down in this essay I will give my own criticisms and arguments to Harberger’s analysis and conclusions. Harberger wanted to know if it is possible to find a numerical value of the allocative and welfare effects of monopoly. To do that, he needed a period when the economy was relatively stable and economic transactions were recorded in detail, as close to their real value as possible. According to him, the only time when those two requirements were met was during the late twenties. Using Professor Epstein’s study around Industrial Profits of the manufacturing industry as a source of data, combined with mathematical formulas, he estimated “how much consumer welfare would have improved if resources had been optimally allocated throughout American manufacturing”. He fully based this estimation on three major assumptions. Firstly, in the long run, resources can be allocated in an industry in such a way as to yield constant returns. In other words, long run average cost are constant. Secondly, elasticity of demand in the industry in question is equal to 1. As a result, any price fluctuation would not change total revenue. Finally, the rates of return to capital are assumed to be equal to 10%.

By using these tools, Harberger concluded that less than one-tenth of 1 per cent of the national income is deadweight loss. A surprisingly low number than what was expected. However, there are many reasons why I believe his reasoning and conclusion are flawed. I believe, one of the major problems with Harbegerger’s train of thought is assuming the manufacturing industry is in a long-run equilibrium under perfect competition and yet mentioning discrepancies in profits and losses between companies. On page 77, Harberger assumes that marginal and average costs are equal but, in the long run with perfect competition, due to the high number of firms entering the industry, firms will be earning zero economic profits. Thus, there can’t be profits for some manufacturers and loses for others. Another issue with Harberger’s logic was the extrapolation of total transfers of resources from industries from Epstein’s study to the whole economy. When analyzing the 73 manufacturing industries of Epstein’s study to calculate the deadweight loss, Harberger extrapolated the figure of total allocation of resources to cover the whole manufacturing economy. I think it is worthwhile to mention that Epstein’s industries only accounted for 45% of total sales in manufacturing.

Consequently, the assumption without any evidence that the companies that make up the 55% of sales have a monopoly in their respective industries, as well as elastic demand equal to one, is a weak argument. Finally, Harberger’s estimate of the deadweight-loss is very conservative. His three main assumptions that lead his examination, heavily limit the results since it is very difficult to apply them in the real-world economy. For example, no monopolist will set a price for the elasticity of demand for all outputs to be equal to -1 and perfect competition is almost impossible to come by. On the other hand, it is undeniable that Harberger was cautious about his assessments. He extensively noted that many factors may have caused him to underestimate or overestimate his figures. Apart from that, he also acknowledged that his goal was not to minimize the effects of the monopoly, but merely to see how severely or not it affects welfare through resource allocation.