Abstract
This article re-examines competitive structure of Canadian wheat exporters for evidence of pricing-to-market by the Canadian Wheat Board (CWB), and extends the extant literature by utilizing a broader range of data over importing countries, which better facilitates the detection of pricing behaviour. Our analysis yields interesting empirical results that contrast with other recent arguments. In particular, we find that the CWB exercises mixed practices having market power in a limited extent, getting premiums with better quality of its wheat in some markets and providing wheat at discounted prices in some other destination countries.
Notes
1Figures for market shares of major wheat exporting countries from 1960/1961 to 2004/2005 by marketing year are available from the author upon request.
2The CWB is one of the largest and longest standing public export-marketing agencies of wheat and barley in the world. It is a mandatory pool, as producers are required to sell their grain to the CWB. Three pillars provide the basis of CWB operations: single-desk selling, price pooling and government guarantees.
3The EEP was initiated under Food Security Act of 1985. This program allows exporters to sell US products in targeted markets at prices below their costs by providing cash bonuses. According to the data set of Foreign Agricultural Trade of the United States provided by Foreign Agricultural Service of the USDA, among various commodities, wheat accounts for more than 80% of the total value of all EEP-assisted sales. Until 1994, EEP was applied to an average of 50 to 70% of US wheat exports.
4Wilson et al. state that price discrimination has continued even after the elimination of EEP, but at a lesser magnitude, and that it exists to the extent that the CWB can exploit market power and differentiate their product from those of competitors.
5The importing countries are Algeria, Belgium--Luxembourg, Brazil, China, Colombia, Ghana, Indonesia, Italy, Japan, Malaysia, Morocco, Mozambique, Pakistan, the Philippines, South Korea, Thailand, the United Kingdom, the United States and Venezuela.
6The OECD data set for the unit values of Canadian wheat in destination markets is available until 1997, so samples from 1998 to 2003 were obtained from the CWB. We compared the two data sets and found that earlier observations of the CWB sample coincide with the observations of the OECD sample for the same years. When splicing two different series from different sources into a series, in some cases, there could be a jump or kink in the final time series at the point of splicing. To ascertain whether there were jumps or kinks in the data, we plotted each time series of the panel data, but did not find evidence of any jump or kink during the sample period.
Table 1. Summary statistics of Canadian wheat prices and Canadian dollar values vs. importing countries' currency values
7The Breusch and Pagan LM test was performed and the test statistic is larger than the critical value of χ2 distribution with 2 degrees of freedom at the 5% level (5.99). That is, the null hypothesis of no time and cross effect was rejected, indicating that inclusion of country and time-specific effects is reasonable.
8The Hausman (Citation1978) test was performed, and the result showed that the test statistic is larger than the critical value of χ2 distribution at the 5% level. Therefore, the null hypothesis of no correlation between the effect variables and the regressors was rejected and thus the country and time effects are treated as fixed.