Overview

Mycroft Sowizral

This presentation will give a brief introduction to NAICS and the "backcasting" methodology. Basic concepts and a brief introduction to the algorithm will be discussed.

NAICS "Backcasting": Creation and Analysis

of a Historical Data Series

The United States is in the process of changing its industrial classification system. Over the next few years, the North American Industrial Classification System (NAICS) will replace the Standard Industrial Classification (SIC) system. The motivation is to update the system to more accurately reflect the current economy, and to have a uniform system with Canada and Mexico. Many economic series will change to reporting on a NAICS basis in the next year or two.

This conversion will cause major disruptions in many economic data series. The lack of historical data for comparison and projection purposes will cause difficulties in making policy decisions.

To address this issue, staff at the New York State Department of Labor and student interns from the UAlbany School of Public Health have developed and implemented a "backcasting" (forecasting in reverse) methodology. The goal of this methodology is to allocate NAICS codes historically to the employer universe. This methodology uses algorithms to generate NAICS codes for each firm in the previous SIC-based quarterly universe data.