Semester of Graduation
Master of Science (MS)
Agricultural Economics and Agribusiness
Trade data across multiple databases experience unavailability across some countries/forestry products, inconsistency, and unreliability; and these qualities manifest as discrepancies in the data. Literature provides evidence of discrepancies and inconsistencies within international trade statistics, including documented cases in which they are present within agriculture sector trade.
While researchers have worked to pinpoint factors to explain discrepancies, studies on the forestry trade databases are not as prevalent. Therefore, more research needs to be conducted to identify discrepancies within forest sector products trade data to understand the nature of discrepancies found between different bilateral trading partners.
The goal of this thesis is to identify and analyze discrepancies in forestry trade data found in bilateral trade series sets for the United States and its top trade partners of forestry products. Discrepancies will be identified using simple mathematical formulations and compared across trade flows and forestry products. Then a unique time-series trade data discrepancy analysis approach is conducted to estimate the nature of trade data discrepancies. The thesis aims to provide a framework for proceeding researchers to utilize in order to apply time-series analysis techniques to trade data discrepancies across any product and country. It also aspires to fill in gaps in the literature of trade data discrepancy analyses that examine forestry product trade data.
Results indicate that discrepancies are present between the bilateral import and export quantity statistics for industrial roundwood, sawnwood, plywood and wood chips and particles for the trade flows from Canada to the U.S, Brazil to the U.S, China to the U.S, the U.S. to China, the U.S. to Japan, and the U.S. to the U.K. Augmented Dickey Fuller (ADF) and cointegration tests reveal that each unique time series bilateral trade quantity pair exhibits unique data generating processes. Therefore, estimating the discrepant relationship between bilateral import and export quantities of forestry products does not follow one clear cut method; but most discrepancies can be estimated through simple or multiple linear regressions, vector error correction models (VECM), or auto-regressive distributed lag error correction models (ARDL ECM).
Daigle, Morgan Alyce, "A Forestry Trade Data Discrepancy Analysis" (2023). LSU Master's Theses. 5697.