- Objectives:
While several factors contribute to the vitality of the local retail market the most fundamental factor is the relative size of the market in terms of potential customers. The purpose of this applied study is to provide an empirical analysis of market demand threshold using stochastic frontier and count data approaches. In this study I review the underlying economic theory and empirical approaches to threshold demand analysis. The empirical analysis estimates the demand threshold for retail business using county level data from the U.S. The results from this study can be used by local economic development practitioners and entrepreneurs to retain, promote, and attract retail commercial businesses in the local communities.
- Data/Methods
The approach adopted in this study builds on threshold analysis by examining the underlying structure of demand for a range of different types of retail and service firms. The predicted number of firms for each community from each model, given its socioeconomic characteristics, can compare to the observed numbers. If the number of firms observed in a sector is greater (less) than the predicted level, then it is reasonably concluded that the sector(s) is strength (weakness) for the community. The county level data on business establishments are being collected from various sources such as, Economic Census of Retail Trade, County Business Patterns, and U.S. Census Bureau. This data will provide 3-digit North American Industry Classification System (NAICS) code on total number of retail establishments, employment, and number of establishments by employee-size. The data on socio-economic information would be extracted from various U.S. Government publications such as, City County Data Book (U.S. Census Bureau) and U.S. Community Fact Finder. The study will include twelve 3-digit NAICS rural businesses in a county.
- Results/Expected Results
It is expected that the estimated system parameter lambda in stochastic frontier model would be significant implying rural markets are over-retailed due to lower rates of acceptable return. The dominant variable in the equation, population, is expected to be significant and positive. The variable on per capita income is expected to be positive and significant. Depending on the statistical significance of the conditional mean of the dependent variable lambda in count data model it will be determined whether Poission or negative binomial model is best for our data. Comparing the expected value of retail establishments to the actual number of firms in the community, the economic development professionals can pursue retail development strategies.
- Conclusion
Increasingly potential entrepreneurs are required by financial organizations to develop detailed business plans. While the cost side of the business plan is often easier for the entrepreneur to develop and commit to writing, the potential revenue side of the business plan is often guess-work at best. Here, market threshold estimates, in conjuncture with the tools of Trade Area Analysis and a host of other market analysis tools, can begin to help the entrepreneur think through the market potential of his or her business idea.