IT Spillover Effects on Korea Economy: for the Creative Economy based on IT

Objectives: The studies in the unit of nation revealed that the IT productivity is in the course of influencing GDP. Therefore, this study analyzed the TFP (Total Factor Productivity) of IT hardware and IT service industries and their productivity spillover effect on each industry. Methods/Statistical Analysis: This study used linked input-output tables to reclassify the IT industry into IT hardware and IT service, obtained the TFP of the IT hardware and IT service, and analyzed the effect of each TFP on other industries. In this way, unlike conventional labor productivity based IT productivity, it was possible to obtain TFP and price-based qualitative numerical values in terms of the productivity effect on other industries. Findings: According to the analysis, the TFP of IT hardware was 2.313 in 2000 and 3.906 in 2005, which was the highest of the TFPs of 30 industries. The TFP of IT service was 0.990 in 2000 which was the 11th position, and 0.988 in 2005 which was the 8th position. Application/Improvements: The study results bring about the following suggestions: First, the productivity of the IT industry positively influenced the Korean economy, and the influence was on the rise. Secondly, the spillover effect of the IT industry expanded from the secondary industry to the tertiary industry. The effect of IT hardware on the tertiary industry was also charged greatly in terms of importance. Thirdly, the IT industry evenly influenced overall industries, rather than particular industries, with the lapse of time.


Introduction
In 1987, Solow created the term of Information Technology Productivity Paradox meaning that the use of information technology fails to influence productivity. Since then, the dispute over the value of information technology use has been activated. Unlike IT ROI (Return On Investment), IT productivity uses production function and is related to the effect of IT capital stock on per capita gross product. Solow 1 suggested that information technology failed to influence gross product and his suggestion greatly impacted the IT industry and enterprises. In the 1990s, theses that opposed pointed out the mistakes of previous study results and presented that IT investments positively influenced productivity. The typical thesis related to such argument is introduced by another researcher 4 who studied the topic in the unit of enterprise and of industry. Aside from that, many studies have been conducted on the IT productivity of enterprises or industries in a country. In 5,6 researched the IT productivity in the unit of country and looked into the effect on GDP per laborer, and presented that IT capital stock per laborer positively influenced GDP.
The dispute over IT productivity paradox seemed to be wrapped up. However, with the development of IT, enterprises use IT in various ways, including IT specification, new forms of connection, creation of new industries, and tool for corporate innovation, beyond the past simple labor replacement and productivity improvement. The phenomenon was proved by the fact that IT investments in mid-2000 accounted for over 50% of all corporate investments. Accordingly, there has been more curiosity of the economic spillover effect of IT. Therefore, since 2000s, theses related to IT productivity have been more diversified by focusing on the effect of specified IT technology, the effect of IT workers, the effect of the intensification extent of IT application, and studies on the spillover effect of IT in firms [7][8][9][10] .
However, most studies have been conducted in the unit of enterprise. The studies in the unit of nation revealed that the IT productivity is in the course of influencing GDP. Therefore, this study used linked input-output tables to reclassify the IT industry into IT hardware and IT service, obtained the TFP (Total Factor Productivity) of the IT hardware and IT service, and analyzed the effect of each TFP on other industries. In this way, unlike conventional labor productivity based IT productivity, it was possible to obtain TFP and price-based qualitative numerical values in terms of the productivity effect on other industries.

Data
The input-output table used in the analysis has the following features. Input-output analysis is a qualitative method to analyze inter-industry correlation, first used by Leontief in his study of input-output relations in the US economy [11][12][13] . The analysis involves an input-out table whose columns show the input structure (i.e. the composition of production costs spent by each sector) and rows indicate the output structure (i.e. how much output of each industrial sector was utilized to satisfy intermediate or final demand of various sectors).
The Table 1 shows the basic structure of input-output The aim of this research is to investigate spillover effects of IT production on production in other sectors of economy, by carrying out input-output analysis the data for the analysis were obtained from the input- output table  at constant price for years 2005, 2000 and 1995, published  by the Bank of Korea in 2009 11 . The input-output table at constant price neglects effects of price fluctuation, and hence assesses economic growth and changes in industrial structure by the amount of output. This makes it suitable for analysis of qualitative changes in the industrial structure and production technology and well as for identification of individual contribution to economic growth by each factor, such as consumption, investment, export, import substation and technological advancement 11,14 . For TFP analysis, the base year for price level was readjusted from 2005 to 1995, by reverse-calculating the adjustments made in the original input-output table, where the base year was 2005.
For the purpose of this paper, IT industry was divided into IT hardware industry and IT service industry to investigate how productivity gain in each industry affects other industries. Table 2, further classification yielded 12 sectors for IT hardware, all of them under larger-sized sector 13, "electronic and electrical equipment", and 2 sectors of IT service, derived from larger-sized sector 24, "business service". The classification system was constant for years 1995 to 2005. Later, IT hardware and IT service were classified as larger-sized sectors on their own, occupying larger-sized sectors 29 and 30, respectively.  Table 3 is industries of re-classification to analysis. IT industry located to 29sector of IT hardware and IT service. So, IT industry is included in re-classified 30 industries.

Methodology
Productivity can be measured as PFP (Partial Factor Productivity) or TFP (Total Factor Productivity). (1) PFP represents the output per unit of a specific factor input, such as labor productivity, capital productivity and raw material productivity. Out of these, labor productivity is an important determiner of employment rate and quality of life. PFP is relatively, but a crucial disadvantage of being unrelated to efficiency of production process as a whole 12 .
On the other hand, TFP represents the output per unit of all factor inputs involved in the production process. Equation 2 poses a problem when calculating TFP as X(x i ) is the total input factor, an aggregate of all factors.
Two approaches have been evolved to approach this problem. Kendrick's solution, shown in equation (3), calculates the arithmetic mean of the quantity of factor inputs by using the price of each factor. On the other hand, Solow's approach, shown in equation (4), is to calculate the geometric mean by assessing contribution of each factor to total cost.
In relation to each industrial sector in INPUT-OUTPUT TABLE, the output is the total national output (X j ) and factor inputs are labor, capital (v kj X j ), and intermediary goods (α ij X i ).

(5)
Productivity analysis with INPUT-OUTPUT TABLE  involves identifying input-output relationship within a  time period shown in each row of the INPUT-OUTPUT  TABLE. Applying this to industry A, the target industry, following equations are obtained.
i(=1,….,n) represents product types and j(=1,…..,m) represents input factor types. Price of products and quantity produced are denoted by p and q, respectively. Lastly, w and f denote the price of input factors and quantity required, respectively. However, since productivity gain is measured in terms of real change in quantity, base-year measures of absolute productivity, measured at constant price at any point time, t, is expressed as following. (8) This can be converted to represent productivity gain in concrete price terms as in equation (9). By combining (6), (7) and (9) other equations can be derived.
Equation (10) shows productivity gain from one point of time to another, and (11) explains productivity gain in terms of price changes. It shows that productivity gained between two points of time at equilibrium affects product price of industry a, which redistributes the productivity gain via spillover effects and change in price of original factor inputs.
This effect, applied to industry j through input coefficient of input-output analysis, allows conversion of equations (6) to (11) into following equations.  Understanding of the above set of equations is aided by the fact that the basic presumption of input-output analysis is that production in any one sector consists of homogenous product 12 . 0 and t indicate time of price evaluation, while a ij represents technical coefficients of intermediary goods and v kj represents value-added coefficients of original input factors on the INPUT-OUTPUT TABLE.
As the primary aim of this research is to investigate how price change induced by productivity gain affects productivity in other industrial sectors, only equation (17) will be considered.
The equation shows how productivity gain in sector j is redistributed to other economic sectors, and the next variable shows the benefit gained by sector j from price changed in other sectors. In other words, productivity gain in a specific sector directly reduces costs of the same sector, and this in turn lowers the price of its product, thereby the whole economy experiences lowered costs 15,16 .

Analysis of Productivity Gain
Productivity gain for each industrial sector was analyzed by applying equations (8) and (9) (Tables 4 and 6).
The Tables 4-6 divides productivity gain into its component factors and shows price change in product, intermediary goods, labor and capital as well as economic gains from productivity gain. In 2000, IT hardware experienced rise in price of intermediary goods, labor and capital by 9,804,949, 5,565,290 and 13,049,873, respectively. Yet, the price of product has fallen due to productivity gain. In numeric terms, price of total input for the whole economy is 47,986,539, and productivity gain mounts to 76,406,650.
In 2005, price of intermediary goods has fallen by 2,987,965 while prices of labor and capital have risen by 3,182,375, and 5,270,293. Again, total rise in price of input has been cancelled out by productivity gain, and product price has fallen. In numerical terms, productivity has increased by 286,391,295 and price of total input was 280,926,591 (Tables 6,7).
On the other hand, IT Service industry was less successful with TFP lower than 1 for both years 2000 and 2005. In 2000, price of all inputs rose and productivity gain was calculated to be -93,463, resulting in price of total input of -2,604,164 (Table 4). In 2005, price of all inputs rose again, and productivity changed by -190,004, while price of total input was -5,799,013, faring even worse than it had in 2000 (Table 6).

Analysis of Inter-industry Spillover Effects on Productivity
The research also investigated how decline in product prices of IT industry due to productivity gain affected other industries. The table A-1 and Table 2 shows the inter-industrial distribution of productivity. Each row shows how much each industry affects other industries by price changes in its own product. Positive values denote a cost reduction and negative values denote cost increase. Each column represents how much cost change is incurred on each industry by price change in intermediary goods of other industries. For example, industry i lowered costs of the entire economy by value R, and in turn received cost reduction effect of value C from other industries. By simultaneously considering economic spill over from productivity gains, shown in the previous Table. 4-5, and the R value of this table, it can be determined     (Table 5) whether productivity gain of individual sectors affected other sectors via price change in intermediary goods or in final consumption. The Table 4 shows that in 2000, IT hardware industry created a benefit of 76,406,650 million KRW on the whole economy, of which 13,804,500 million KRW was an effect on other industries (  8,9).
Compared to IT hardware industry, IT service industry was found to affect all sectors more evenly, and especially on knowledge services as time went by. The correlation between IT hardware industry and IT service industry also seems to have grown closer. Table 10 presents the variance result of the analysis on the productivity spillover effect of each IT industry and on the bias of the effect on particular industries. In regard to the effect on the secondary industry in 2000, the effect of IT hardware accounted for 96.9%, down to 90.8% in 2005. The change led to a rise in the effect on the tertiary industry in 2005. In IT service, more clear changes were found. In other words, in 2000, its effect on the secondary industry reached 60.7%, but in 2005 the effect on the secondary industry sharply fell to 29.3% and the effect on the tertiary industry soared to 70.5%. In addition, the effect of IT service on the primary industry greatly increased from 0.07% in 2000 to 0.22% in 2005. However, the productivity spillover effect of IT service showed negative (-) value overall in 2005. Although the issue is required to be analyzed in another study, it is speculated that from early 2000 to mid-2000 investments in IT service had considerably increased, but since then they had failed to lead to the effect of direct cost reduction. The issue needs to be analyzed in the future research. It is impossible to deny that the IT industry now influences the tertiary industry, rather than the secondary industry, and give the effect on overall economy.

Conclusion
Both IT hardware and IT service influenced the Korean economy more greatly in 2005, than in 2000. In the case of IT hardware, its TFP soared from 2.313 in 2000 to 3.906 in 2005, which was the largest among the TFPs of all industries. The benefit that affects the overall economy increased from KRW 76,406,650 million to KRW 286,391,295, up almost four fold. TFP of IT service was 0.99 both in 2000 and in 2005, but the benefit decreased more than two fold from KRW -93,463 million to KRW -190,004million. That was attributable to increases in prices of all investment elements. The investment also increased more than twice from KRW 2,604,164 million in 2000 to KRW 5,799,013 million in 2005.
In the analysis on productivity spillover effect, what needs to be focused on is that the spillover effect of IT hardware on overall industries was found more evenly in 2005 and that the tertiary industry has become more important. IT service greatly influenced the secondary industry in 2000, whereas its productivity spillover effect on the tertiary industry sharply increased Vol 9 (46) | December 2016 | www.indjst.org   The study results bring about the following suggestions: First, the productivity of The IT industry positively influenced the Korean economy. In particular, in the case of IT hardware, its economic benefit caused by the productivity was the highest of the benefits of all industries. Secondly, the spillover effect of The IT industry expanded from the secondary industry to the tertiary industry. In the case of IT service, its effect on the tertiary industry was found clearly in 2005, compared to 2000. The importance of the spillover effect of IT hardware on the tertiary industry was changed largely. Thirdly, with the lapse of time, the IT industry evenly influenced the overall industries, rather than particular industries. The variance of each industrial weight of the productivity spillover effect decreased in 2005, compared to 2000.
This study has the following limitations: In the case of IT service, the investment increased twice in 2005, compared to 2000. According to 2005 analysis, it negatively influenced economic benefit and cost reduction in each industry. Nevertheless, it is hard to define that IT service negatively affects the economy. The reason for such results is that this study analyzed the 2000 and 2005 spillover effect of TFP on each industry with the use of linked input-output tables, and the analysis results were analyzed on the basis of the prices of investment elements. The effects made with the lapse of time were not taken into account, and price changes were analyzed at a given point. In addition, the management and application of IT service were not taken into consideration in terms of qualitative aspect.
In regard to the influence of certain investment elements, it is possible to analyze the influential results properly after some time. In particular, in the case of IT service, its outcomes appear after IT hardware capital is mature enough (Shin 2013). According to this study, IT hardware brought about its outcomes differently, depending on the extent of direct investment elements. However, in the case of IT service, its investment industry changes from the secondary industry to the tertiary industry, and the productivity is found different depending on the extent of hardware infrastructure and IT application level of each industry. Therefore, it is hard to expect the productivity spillover effect of IT service immediately.
Given that, the future study needs to use production function to analyze the influence with the lapse of time, and look into other managerial aspects than economic approach in order for productivity and efficient use of IT service.