Estimating Prices from a Natural Gas Monthly Survey

• Activity Number: 499
• Type: Contributed
• Date/Time: Wednesday, August 7, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Survey Research Methods Section
• Author(s): Samson Adeshiyan*+
• Companies: US Energy Information Administration
• Keywords: classical ratio estimator ; cutoff sampling ; price estimation ; standard error ; variance

Abstract

The U.S. Energy Information Administration (EIA) conducts a monthly survey of natural gas suppliers, which collects volume and revenue data. These data are used for estimating State level natural gas total sales volume and average price published in EIA's Natural Gas Monthly. This paper documents the methodology used to estimate total volumes and average prices. We briefly compare the form of a design-based classical ratio estimator (D-CRE) to that of a model-based classical ratio estimator (M-CRE). While these two estimators are algebraically equivalent, the D-CRE may lead to a discrepancy in practice. Since the survey is based on a quasi-cutoff sampling design, a modeled-based estimator, like the M-CRE, is appropriate. Furthermore, unlike with the D-CRE, we can obtain associated standard errors for M-CRE estimates from a cutoff-sample. Finally, we discuss correlations of components in the price estimator, whose variance is based on the Taylor series estimator of a ratio, and the covariance computation involved in calculating standard errors of price estimates. This approach, of estimating price and its standard error, can be beneficial to other EIA surveys.