Premium
Temporal Aggregation in Distributed Lag Models
Author(s) -
Ladd George W.
Publication year - 1964
Publication title -
american journal of agricultural economics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.949
H-Index - 111
eISSN - 1467-8276
pISSN - 0002-9092
DOI - 10.2307/1236451
Subject(s) - lag , distributed lag , time lag , lag time , computer science , econometrics , biological system , mathematics , biology , computer network
Nerlove and Mundlak have shown that the estimated adjustment coefficient in Koyck‐Nerlove models is generally larger in studies using longer unit observation periods. Results from quarterly and monthly food demand analyses are consistent with their conclusion, while results from quarterly and annual analyses are inconsistent with their conclusion. The most interesting explanation of the inconsistency leads to the following hypothesis. National annual consumption of foods is predetermined and price is endogenous. The coefficient of lagged annual consumption is biased away from zero if (a) current consumption is assumed endogenous, but is unbiased if (b) price is assumed endogenous. Estimated coefficients of lagged annual consumption will be significant more frequently in case (a) than in case (b). Annual food demand analyses accept this hypothesis.