Estimation of Industry Distribution of Statistical Discrepancy in National Accounts
Gross domestic product (GDP) and gross domestic income (GDI), though conceptually equivalent, differ by statistical discrepancy (SD). Currently, there are no estimates of SD by industry. Lack of such information hinders a proper understanding of the sources of inconsistency in the national account and makes it difficult to identify improvements needed to minimize SD. This paper describes and illustrates an estimation method that can correctly estimate industry distribution of SD according to the reliability of the initial estimates, and that can accurately reconcile GDI by industry account and Input-Output (I-O) account which measures GDP as value-added by industry. The reconciliation model is described by a constrained optimization model which minimizes the weighted sum of squares of deviation from the initial estimates in all components of I-O and GDI by industry accounts. The optimal solution is equivalent to estimates from generalized least square estimation. Data used in estimation are from BEA’s 1997 I-O and GDI by industry accounts
The text is part of a series Computing in Economics and Finance 2006 Number 401
Classification:
C13 - Estimation ; C61 - Optimization Techniques; Programming Models; Dynamic Analysis ; E01 - Measurement and Data on National Income and Product Accounts and Wealth