Companies in competitive industries routinely collect information about their customers through a variety of sources (em including surveys, national census, and government and private sources....
Know Thy Customer
Including these transactions is necessary to make the totals of an input-output table consistent with national income and product aggregates, but alternative accounting procedures are usually employed in empirical input-output tables, which identify these transactions with value added and final expenditures for sectoral outputs.
The current benchmark U.S. input-output table was projected using a modified version of the RAS technique.1 We used historical data and projections of gross output, intermediate sales, and intermediate demand from our U.S. macroeconomic and regional models to balance the row sums (intermediate sales) and column sums (intermediate demand) using an iterative procedure. By repeating the procedure for each historical year and over the forecast period, we projected the table out to 2000.
Turn it into
Regional Data . . .
We then converted the national table to reflect a regional economy, assuming that production technology for each industry in the region matches that of the nation as a whole. County and state models provide estimates and projections of employment for all 4-digit Standard Industrial Classification (SIC) codes. These models use data reported by the Department of Commerce in County Business Patterns, as well as data published by the Bureau of Economic Analysis and Bureau of Labor Statistics (BLS). The resulting data set replicates all data reported by the BLS, and then supplements the data with WEFA estimates. Through-out the process all adding-up
consistency is maintained (em i.e., all three-digits sum to two-digits, and all state data sums to national data by industry. Where there is a discrepancy due to nondisclosure of data (e.g., where there is only one company in that 4-digit SIC in the region and, therefore, the sum of employment at the 4-digit SIC level does not equal the employment reported at the 2-digit level), we use alternate company information databases to fill the "hole." In the absence of regional output measures, we use employment (as a proxy for output) along with regional industry productivity growth measures to calibrate the table.
The regional input-output table is a detailed accounting of flows of goods and services in the region and provides a wealth of information about the local economy. The columns of the regional table provide a detailed profile of (4-digit SIC) local industries. An individual column in the table enumerates what that industry purchases in goods and services from other industries. From this, market researchers can estimate an industry's expenditure on various inputs that consume electricity.
Getting to sales (quantity) of electricity from industry expenditure data is far more complex and labor intensive. We use a variety of sources to obtain rate information by customer class and size for each IOU in the country. At best, these rates are averages across customer classes and size. The rates are then applied uniformly to customers based on class and size in the utility's service territory, resulting in estimates of sales of electricity to each of the 4-digit SIC industries in the region. The projected input-output tables furnish the growth in electricity expenditure for each industry in the region.
And Apply it to Your Market
For our discussion