Estimating policy effects on spatial market efficiency: an extension to the parity bounds model.

Results for wheat flowing from the surplus southern production areas of Robe and Hosaina to Addis are reported in the first two columns of table 2. Before the policy change both of these trade routes have a relatively high probability of being in regime 3, suggesting strong evidence of unexploited arbitrage opportunities. After the policy change, the probability of being in regime 3 increases further in the case of Robe-Addis and decreases slightly in the case of Hosaina-Addis. However, the probability changes for the Robe-Addis route are not statistically significant at conventional significance levels, and the reduction in regime 3 probability for the Hosaina-Addis route, though statistically significant at the 5% level, is quite small and still leaves this trade route with an 81% probability of being in regime 3 after a five-month adjustment period. Therefore, the conclusion is that both of these trade routes from wheat surplus regions to Addis Ababa experienced frequent unexploited arbitrage opportunities prior to the policy change, and that this situation remained in force after the policy change.

Columns 3 and 4 of table 2 show results for trade between Addis and the northern wheat deficit markets of Dese and Mekele (see figure 1). In the case of Dese, there is a high probability of being in regimes 2 and 3 prior to the policy change (79% and 21%), respectively, and no statistically significant change in regime probabilities after the commercial reorientation of the EGTE. Hence, the Addis-Dese route is estimated to fluctuate between situations where price differentials are above estimated transfer costs and vice versa, with no policy effects. For the Addis-Mekele route, results are very similar to those found for maize--high probability of being in regime 3 prior to the policy change (80%) but a large, immediate, and statistically significant reduction in this probability after the policy change (down to 28%). These changes are reflected in both the reduction in mean price differentials and arbitrage profits after the policy change (see table 3), and the tendency for observed price differential to be frequently above the parity bound prior to 1999, and frequently below it after 1999 (see figure 2). The likely explanation for this result (elimination of the roadblock at Alamata) has already been discussed.

The final three columns of table 3 provide results for three routes for shipping wheat to the most eastern food deficit region of Dire Dawa (see table 2 and figure 1). All three routes have a high probability of being in regime 3 prior to the policy change (77% for wheat sourced in Addis, 59% for Nazret, and 71% for Shashemene). Furthermore, point estimates of the change in regime 3 probability after the policy change are positive for all three trade routes, though changes are jointly statistically significant for only one route (Shashemene to Dire Dawa) in which changes occurred over a six-month adjustment period. The conclusion is that these trade routes experienced unexploited arbitrage opportunities prior to the policy change, and that this situation remained after the policy change had occurred.

Overall, there is evidence of widespread unexploited arbitrage opportunities in regional wheat trade prior to the policy change, suggesting the existence of frictions, which preclude sufficient trade from occurring to drive price differentials down to transfer cost. After the policy change, spatial inefficiency stayed the same or, in some cases got slightly worse. The only exception to this is the Addis-Mekele route, which experienced a significant reduction in the probability of regime 3, probably due to elimination of the roadblock at Alamata. A possible explanation for the prevalence of spatial inefficiency is that wheat traders are generally larger scale and fewer in number than maize traders, and barriers to entry are higher, opening the door for noncompetitive behavior. In terms of speed of adjustment for wheat markets, statistically significant policy effects occurred for the Hosaina-Addis route over a five-month adjustment period, and for the Shashemene-Dire Dawa route over a six-month period. So these adjustment lengths are quite consistent. The only other statistically significant policy effect for wheat occurred on the Addis-Mekele route, which was subject to the roadblock removal at Alamata. This was a highly visible policy change and its effects were estimated to have occurred immediately within a month (zero adjustment length--see table 2).

Conclusions

The standard PBM is a valuable tool for examining spatial market efficiency. However, in the context of ongoing market reforms and policy changes, the standard PBM needs further refinement to allow the effects of policy changes on spatial market efficiency to be assessed. The standard PBM has been used mostly to analyze spatial market efficiency within a given policy regime, or if policy change has been examined then the effects are assumed to be instantaneous. However, the standard PBM will be misspecified if the actual effect of policy changes on spatial market efficiency is gradual and moves through a transition period, as might be expected in many cases.

The EPBM developed here is a stochastic gradual switching regression model that allows the dynamic effects of policy change to be estimated. The EPBM improves the standard PBM in two ways. First, it facilitates formal statistical tests for changes in the probabilities of different trade regimes due to policy changes. Second, the EPBM allows a better understanding of the nature of transition from old to new policy regimes, including estimates of the length of the adjustment period.

The EPBM is estimated using maximum likelihood and applied to data on wholesale maize and wheat prices for several market pairs in Ethiopia to examine the effects of grain marketing policy changes implemented in October 1999. Results indicate unexploited arbitrage opportunities in maize and (especially) wheat trade in Ethiopia, both before and after the policy change. After the policy change, both maize and wheat traded between Addis and Mekele in the north showed evidence of improved spatial efficiency (consistent with trade expansion and a reduction in spatial price differentials), most likely due to elimination of the Alamata roadblock. But while the policy change had little effect on other wheat markets there were four maize trade routes (Nekemte-Addis, and the three routes into Dire Dawa), which experienced higher price differentials and an increase in unexploited arbitrage opportunities after the changes made to the EGTE in 1999. Estimated adjustment periods for the full effects of the policy change to be felt ranged from instantaneous adjustment to twenty-one months, with most adjustment periods lasting between zero and six months. These results highlight the importance of allowing for adjustment to policy changes.

Partial financial support for the study was provided by the International Food Policy Research institute and the Michigan Agricultural Experiment Station. The authors would like to thank Thorn Jayne, Eleni Gabre-Madhin, Chris Barrett, and three anonymous reviewers for helpful comments.

[Received November 2004; accepted August 2006.]

References

Asche, F., H. Bremnes, and C.R. Wessells. 1999. "Product Aggregation, Market Integration, and Relationships between Prices: An Application to World Salmon Markets." American Journal of Agricultural Economics 81:568-81.

Barrett, C.B. 1996. "Market Analysis Methods: Are Our Enriched Toolkits Well Suited to Enlivened Markets?" American Journal of Agricultural Economics 78:825-29.

Barrett, C.B., and J.R. Li. 2002. "Distinguishing between Equilibrium and Integration in Spatial Price Analysis." American Journal of Agricultural Economics 84:292-307.

Baulch, R. 1997. "Transfer Costs, Spatial Arbitrage, and Testing for Food Market Integration." American Journal of Agricultural Economics 79:477-87.

Bekele, G. 2002. "The Role of the Ethiopian Grain Trade Enterprise in Price Policy." In T. Bonger, E. Gabre-Madhin, and S. Babu, eds. Agriculture Technology Diffusion and Price Policy: Proceedings of a Policy Forum in Addis Ababa. 2020 Vision Network for East Africa Report 1, Washington DC: Ethiopian Development Research Institute and International Food Policy Research Institute, pp. 83-91.

Blake, N.S., and T.B. Fomby. 1997. "Threshold Cointegration." International Economic Review 38:627-45.

Dadi, L., A. Negassa, and S. Franzel. 1992. "Marketing Maize and Tef in Western Ethiopia: Implications for Policy following Market Liberalization." Food Policy 17:201-13.

Dercon, S. 1995. "On Market Integration and Liberalization: Method and Application to Ethiopia." Journal of Development Studies 32:112-43.

Dessalegn, G., T.S. Jayne, and J.D. Shaffer. 1998. "Market Structure, Conduct, and Performance: Constraints on Performance of Ethiopian Grain Markets." Working Paper 9, Grain Market Research Project, Ministry of Economic Development and Cooperation.

Enke, S. 1951. "Equilibrium Among Spatially Separated Markets: Solution by Electric Analogue." Econometrica 19:40-47.

Fackler, P. 2004. "Estimating the Degree of Market Integration." Unpublished manuscript. North Carolina State University.

--. 1996. "Spatial Price Analysis: A Methodological Review." Unpublished manuscript. North Carolina State University.

Fackler, P.L., and B.K. Goodwin. 2001. "Spatial Price Analysis." In B.L. Gardner and G.C. Rausser, eds. Handbook of Agricultural Economics,