• The Korean Journal of Applied Statistics
• /
• v.22 no.1
• /
• pp.75-88
• /
• 2009

Gross Regional Domestic Product (GRDP) is regarded as an essential information to understand regional economy. However, GRDP is hardly used for establishment of regional economic plan and related statistical research due to its late and yearly publication. Therefore, it is necessary to estimate quarterly GRDP to grasp the current regional economy faster In this study, considering the comovement between GDP and GRDP for the same industry, reference series are made. Quarterly GRDP is estimated the following two steps; First, preliminary quarterly GRDP is estimated using Chow-Lin's method based on the reference series to eliminate temporal discrepancies. Second, preliminary quarterly GRDP is adjusted using Denton's multivariate method to eliminate contemporaneous discrepancies.

• Survey Research
• /
• v.10 no.1
• /
• pp.113-139
• /
• 2009

It is desirable to use a monthly benchmarked construction indicator which contains the characteristics of statistical data in an annual survey in order to analyze the cyclical phenomenon of the construction activity. The benchmarked indicator is expected to improve the data quality in terms of accuracy, consistency, comparability, and completeness. In this paper, benchmarking methodologies of compiling monthly construction indicators arc researched by using a monthly prompt data holding short - term fluctuations and an annual survey data regarded as more accurate statistics than monthly data. The benchmarking is the methodology by which a high frequency data should he adjusted in order to hold the short-term and cyclical phenomena, and the long - term trend of two data groups with ensuring the consistency of an annual summation between a high frequency data and a low frequency data. This paper considered the numerical approach like pro rata distribution method, proportional Denton method, EFL or HP filter Benchmark - to - Indicator ratio method, and the model - based approach such as Chow and Lin method, $Fem{\acute{a}}ndez$ method. Also, the benchmarked construction indicators were estimated by early mentioned benchmarking methods with practical data, and these methods were empirically reviewed and compared. In case of construction indicators with severe seasonal fluctuations and irregulars, the numerical approach seemed to be performed more correctly than the model- based approach. Among numerical methods, the proportional Denton method used in general was a little nice. The HP filter Benchmark - to - Indicator ratio method may be considered with survey errors or measurement errors in an annual survey data.