Publication
The Korean Journal of Economic Studies
How Do We Utilize a New “Output-Output Model” and “Output-Output Table”?
Ho Un GimYear 2009Vol. 57No. 2
Abstract
Gim and Kim(2008a, 2008c) recently found out that there is no consecutiveconnection between the Leontief inverse ? and the total output and thatthere are some limits to computing the impact effects of the initial change ofoutput. Thus a new “output-output(OO) model” was developed by the authorsto solve the consecutive connection and overestimation problems very naturallybased on the output requirements matrix for output ? .On the basis of the latest research findings mentioned above, the specificobjectives of this paper are summarized as follows. (1) We reexamine themerits and demerits of different types of input-output(IO) models: the demandside, the supply side, and the mixed type. (2) We perform a comparativeanalysis between the IO and OO models in structure and characteristics andillustrate the usefulness and validity of the OO model. (3) We derive therelevant equations between the cause and effect variables, which are needed tocompute induced effects, multiplier effects, and linkage effects by using thenewly developed OO model and table. (4) We present some valid applicationexamples, focusing on the empirical economic analysis based on raw data of 2003 Input-Output Tables compiled by the Bank of Korea. The major findings from the empirical analysis are followed below. Wecalculated the output, employment, and income multipliers and the impact andsensitivity coefficients for final demand and output through the Leontief inverse? and the output requirements matrix for output ? . The Spearman’s rankcorrelation coefficients( ????), ? , are specially computed to observe therelevant structural economic characteristics of each sector between two types ofemployment multipliers, income multipliers and sensitivity coefficients. The???? between employment multipliers is 0.9878 and that between incomemultipliers is 0.9402 and that between sensitivity coefficients is 0.9962. Thehigher values which are closer to 1 signify that the entire priority ranksbetween multipliers(or coefficients) are almost the same throughout the wholesector.