摘要：This paper expands the contributions of Goodwin and Holt (AJAE, 1999) and Goodwin and Harper (J. of Ag. and Appl. Econ., 2000), GHH henceforth, who analyze retail-wholesale-farm price transmissions in the U.S. beef and pork industries using weekly data. First, in light of advancements in unit root tests, we re-examine in a more comprehensive manner GHH’s conclusion that the weekly U.S. cattle/beef and hog/pork price series are nonstationary. The conventional augmented Dickey-Fuller test that GHH adopt has low power in discriminating against the unit root null because it does not entertain the possibility of a structure break in the deterministic trend function. Second, we examine more closely the estimation procedure surrounding the long run price linkage equation, ensuring the unbiasedness in the estimated long run price transmission coefficient. Moreover, we ascertain whether there have been structural changes in the long run price linkage equation using a new approach, which endogenously estimates the break date. Third, we employ two data sets with different frequencies, weekly data and monthly data, to gain insight into the reasons underlying price asymmetry found in GHH.
