SQUIRES ET AL.: JAPANESE AND U.S. SABLEFISH MARKETS 



merited if 



0, j = 0,1, 2,..., J, 



(2) 



which can be determined by imposing the param- 

 eter restriction of Equation (2) on Equation (1), and 

 testing this restricted model against the unrestricted 

 model of Equation (1) with anF-test. Nonrejection 

 of the linear restrictions or null hypothesis indicates 

 that the price in local market i depends only on its 

 own lagged values and local market characteristics. 



Short-Run Market Integration 



A price change in the central market will be im- 

 mediately and fully passed on to the ith local market 

 price if 



1. 



(3) 



This hypothesis, in addition, requires that there be 

 no lagged effects on prices in the future: 



= b. 



0, 



1,2,. ..J. 



(4) 



If both Equations (3) and (4) are accepted as param- 

 eter restrictions, then market i is integrated with 

 the central market within one time period. 



A weaker form of short-run market integration 

 will also be tested, in which the lagged effects need 

 only vanish on average: 



J J 



Z. aij + 2. 6,j = 0. 



(5) 



An additional indicator of short-run market in- 

 tegration occurs if 6,0 = 1, but Equation (4) or (5) 

 do not hold (Heytens 1986). In this case, short-run 

 market integration cannot be accepted, yet econom- 

 ic forces causing central market price changes are 

 generally being reflected in the local price level. A 

 form of integration is occurring, even though the 

 central and local markets are not being fully linked 

 in the short run; that is, changes in the price margin 

 between the central and local markets are not be- 

 ing fully passed on. 



Absence of Local Market Characteristics 



This hypothesis assumes that 



Ci = 0, (6) 



where c, is a vector if there is more than one local 

 market characteristic. Testing this hypothesis is of 

 interest when local prices are suspected to have dif- 

 ferent seasonality than the central market. In this 

 case, X,, can be defined as a matrix of dummy 

 variables. 



Long-Run Market Integration 



A long-run equilibrium is one in which market 

 prices are constant over time, undisturbed by any 

 local stochastic effects. Thus, when P,, = P,*,j = 

 2,. . .,n, Pit = Pi*, and e,, = for all t, Equation 

 (2) takes the form 



P,* = 



P,* I b,j + Xuc 



J 



1- la., 



(7) 



Long-run market integration now requires that 



J J 



Z. a,j -I- 2. 6,j = 1. 



(8) 



If this linear parameter restriction is not rejected 

 by an F-test, then the short-run process of price ad- 

 justment described by the model is consistent with 

 an equilibrium in which a unit increase in the cen- 

 tral market price is fully passed on in local market 

 prices. Markets where previous central market 

 prices and past spatial price differentials are the 

 primary determinants of local prices (rather than 

 previous local prices) are well connected in the sense 

 that supply and demand conditions in the central 

 market are communicated effectively to local mar- 

 kets. In the long run, the central market influences 

 local market prices irrespective of previous local con- 

 ditions, even though traders may fail to connect the 

 two markets through commodity flows in the short 

 run (cf Timmer 1974). Acceptance of the short-run 

 restrictions implies long-run market integration, but 

 the reverse is not necessarily true. 



If the linear restriction for long-run market in- 

 tegration is not rejected, then more efficient esti- 

 mates of the remaining parameters and more power- 

 ful statistical tests are possible if the model is 

 reestimated with long-run market integration im- 

 posed. Equation (1) under long-run integration can 

 be written in the following equivalent form (Raval- 

 lion 1986): 



343 



