Researchers' enthusiasm for estimating industry oligopoly power in developing countries is often hampered by a lack of available data. Even when available, data are often incomplete, inconsistent, too aggregated, and almost always collected by government agencies for purposes different from those of the researcher.
As a consequence, a researcher addressing the issue of oligopoly power has four options:
- Ignore the issue
- Conduct a descriptive industry study
- Develop an elaborate structural econometric oligopoly model and strip it to suit available data
- Estimate ad hoc regressions using available data
- Address the issue by generating theoretical predictions from a stylized theoretical model of oligopoly.
None of the options deliver estimates based on explicit theoretical restrictions implied by optimizing behavior of imperfectly competitive firms.
An alternative option, which this paper suggests, is to use the theoretical restrictions implied by firm optimizing behavior to specify and make inference about market power in a conjectural elasticity (CE) model when data availability is a problem. The novel feature of the model is that it is empirically implemented with observations on only two variables likely to be found in most industry statistics collected for tax purposes by governments in developing countries: Sales revenue and payroll (the wage bill).