The Impact of Financial Crises and IT Revolution on Income Distribution in Korea: Evidence from Social Accounting Matrices

For presentation at WIOD Conference (2012) Groningen (The Netherlands), April 24-26, 2012 The Impact of Financial Crises and IT Revolution on Income Distribution in Korea: Evidence from Social Accounting Matrices Hak K. Pyo 1, Keun Hee Rhee** and Gong Lee*** Abstract In recent years, Korea had experienced two financial crises in 1997-1998 and 2007-2008 respectively and the IT revolution during the interval between the two financial crises. We have constructed Social Accounting Matrix (SAM) in 2000 and 2009 for Korea by combining Input-output tables used in WIOD project in Korea with data from 10-decile Urban Household Income Survey. We analyze gross income effect and income redistribution effect of financial crises and IT revolution by adopting a SAM framework following Pyatt and Round (2004) and Saari, Dietzenbacher and Los (2010). Both financial crises and IT revolution have generated larger income multiplier effect on Higher ITintensive sector but have affected negatively on income redistribution of lower income groups. 1 Director, Center for National Competitiveness and Faculty of Economics, Seoul National University, Seoul 151-746, Korea. Tel: +82-2-880-6395. Fax: +82-2-886-4231. E-mail: pyohk@plaza.snu.ac.kr. Corresponding author. ** Senior Researcher, Korea Productivity Center, E-mail:ghlee@kpc.or.kr *** Ph.D. Candidate, Department of Economics, Seoul National University, Seoul 151-746, Korea. Tel: +82-10-9586-2020, Fax: +82-2-886-4231, E-mail: tempest3@snu.ac.kr

1. INTRODUCTION Social Accounting Matrix (SAM) provides a useful framework to analyze the composition of national income and product and the process of income distribution. Even though SAM is a static analytical tool, we can examine the dynamic aspects of national income composition and its distribution by comparing two SAM over a time interval. We have shown in a recent report (Pyo, Kim and Lee (2011)) using SAM with the household micro cell by age group that exogenous injections of income to government mostly benefit the relative income of the pensioners, the age group over 60. The empirical result is consistent with that of Llop and Manresa (2004) who have found that exogenous injections of income to government mostly benefit the relative income of inactive households, which are mainly those of pensioners and have argued that such empirical evidence is important for policy making if the policies are aimed at modifying the distribution of income among economic agents. The purpose of the present paper is to analyze the impacts of IT revolution and two financial crises on the determination of national income and changes in the level of income of endogenous sectors through multiplier analysis and the structure of income distribution by constructing micro cell of household sector. As analyzed in Pyo (2000), the growth rate of real GDP which had averaged 7.1 percent during the pre-crisis period of 1993-1997 declined sharply recording 6.9 percent in year 1998 when the so-called twin crises had hit the Korean economy. The financial crisis of 1997-1998 was called twin crises because there were both domestic banking crisis and foreign exchange crisis. As documented in Pyo (2004a) and Otsu and Pyo (2009), the macroeconomic adjustment during the post-crisis period was a successful one largely due to remarkable export performance helped by Won depreciation. During the period of 1998-2007, the export performance led by IT-intensive products has helped the economy to make a sustainable growth. When the global financial crisis occurred in 2007-2008, the recovery pattern of the Korean economy has been quite similar to the recovery pattern after the 1997-1998 crises. Korean Won was depreciated against not only US dollar but also Yen, Euro and Yuan and strong export performance in IT-intensive manufactures followed. We have attempted to analyze the impact of large scale depreciations immediately after two financial crises and export drive of IT-intensive products during the two post-crisis periods by multiplier and redistribution analysis based on SAM. For the purpose, we have constructed SAM for two discrete years of 2000 and 2009 in which we have treated Capital Accounts, External Accounts and Government Accounts as exogenous sectors and four-category production sectors as endogenous sectors:(1) Higher IT-intensive (2) Lower IT- intensive with Agriculture and Mining (3) Higher IT-intensive s and (4) Lower IT-intensive s following Ha and Pyo (2004) and Pyo and Ha (2007). We have also added a micro-sam by decomposing the household sector to 10-decile units of income distribution by using KLIPS (Korea Labor Income Panel Study) dataset (1998-2008). There are major findings of the present study. The first finding is that the impact of both IT technology and two drastic depreciations have generated significant multiplier effects and redistributed income effects on Higher IT-intensity endogenous production sectors. The second finding is that the lowest income group ( 1) was the largest beneficiary group in terms of multiplier contribution and the highest income group ( 10) was the largest beneficiary group in terms of redistributed income effect. The last finding is consistent with the empirical finding by Llop and Manresa (2004) that injections of income to activities mostly benefit the relative income of the richest active households and the finding by Noh and Nam (2006).

In Section 2, we outline a SAM model of multiplier effects and redistribution effects. Section 3 presents empirical results after constructing SAM for the Korean economy of 2000 and 2009. The last section concludes the paper. 2. THE MULTIPLIER EFFECT AND RELATIVE INCOME DISTRIBUTION Following Roland-Holst and Sancho (1992) Llop and Manresa (2004) and Saari, Dietzenbacher and Los (2010), we can specify the multiplier analysis by dividing the accounts of a SAM into two separate sectors: endogenous and exogenous accounts as shown in Table 1. If we consider m endogenous sectors and z exogenous sectors, a SAM can be written as follows: 1 where A ij are partitioned sub-matrices that contain the expenditure share coefficients calculated by dividing the transactions in the SAM by the corresponding sum column. The multiplier analysis assumes that the expenditure coefficients are constant, so sub-matrices A ij assumed to be invariant over time. Income from endogenous accounts (Y m ) can be obtained as follows from the top first row equations: 2 where I is the identity matrix, M = (I A mm ) -1 is a multiplier matrix and x = A mz Y z is a vector of exogenous variables. The multiplier matrix M shows the overall effects of a unitary increase in the exogenous components on the endogenous accounts. Therefore, the element m ij of M quantifies the changes in the income of the sector I ( dy m ), i.e. gross income effect as a consequence of a unitary and exogenous injection received by the sector j( dy ). m From expression (2), the analysis of multipliers corresponding to endogenous sectors illustrates the changes in the absolute levels of income. Roland-Holst and Sancho (1992) presented an overall context for distributive incidence based on the SAM model. To identify the changes in the relative incomes 2 of every endogenous sector, expression (2) can be normalized as follows: 3 3 where e is a unitary row vector. From (3), the changes in the relative income of the endogenous 2 By relative income we mean relative to total income of all the endogenous accounts. 3 See Roland-Holst and Sancho (1992).

sectors generated by a modification in the exogenous injections are equal to: Table 1. Simplified Schematic Social Accounting Matrix 4 Expenditures Endogenous accounts Exogenos accounts Totals Factors Sectors Production Sum of (Households and activities other companies) accounts 1 2 3 4 5 Receipts Endogenous sectors Factors 1 0 0 T 13 x 1 y 1 Sectors (households 2 T 21 T 22 0 x 2 y 2 and companies) Production activities 3 0 T 32 T 33 x 3 y 3 Exogenous sectors Sum of other accounts 4 l 1 l 2 l 3 t y x Totals 5 y 1 y 2 y 3 y x 1 4 In this expression, R is defined as the m by m redistribution matrix. It shows the change (positive or negative) in the relative income of the endogenous sectors ( dy m ) caused by unitary modifications in the exogenous injections of income received( dx m ). An individual element of this matrix, r ij, determines the magnitude (positive or negative) of the percentage change in the relative income of the sector i as a result of a unitary inflow in the sector j. This way of calculating the distribution process involves a set of bilateral connections between the endogenous sectors that tell us how one account influences the relative status of another. It is interesting that, irrespective of which endogenous components are chosen in the model, the sum of the columns in the matrix of redistribution is zero. 5 This mathematical property means that the distribution process between the endogenous accounts can be interpreted as a game of winners and losers. If we take expression (4), we can identify three multiplicative components in the structure of R : 5 4 Thorbecke, E. and H.S. Jung, Journal of Development Economics 48 (1996) 279-300 5 By relative income we mean relative to total income of all the endogenous accounts.

1 The first component, b ( e' Mx), is the inverse of the total income of the endogenous sectors and 1 is a scalar. The second matrix, D I ( e' Mx) ( Mx) e', has two parts: the first part ( I ) is the initial and exogenous injection of income that activates the multiplier process, and the second part emx Mxe 1 ( ' ) ( ) ' is the matrix of the initial relative income of every endogenous sector (with a negative sign). Finally, M is the matrix of standard multipliers. Expression (5) represents matrix R in a multiplicative form and can be transformed into an additive expression. This transformation will make it easier to interpret the effects involved in the income distribution process. Specifically, we can define the redistribution matrix as follows: 6 This representation of R reveals the underlying components of the income distribution process and displays the sequential terms involved. In expression (6), b is the inverse of the total income of the endogenous sectors or the factor of normalization. The terms in the bracket show the initial and exogenous injection that starts the multiplier effect and the distribution process. Also, the matrix ( I D) tells us the endogenous sectors relative position. The last term in the bracket, D( M I), is the net multiplier effect on relative income and represents the additive contribution of the net multipliers to the distribution process. Notice that D( M I) contains the cross multiplier effects among the endogenous sectors and its effects on relative income determination. An arbitrary element (, i j ) of this matrix, which can be either positive or negative, is equal to:.. where mn ij are the components of the matrix M I of net multipliers and mn. j are the sum of the elements of the j th column of M I. The multiplier contribution to income distribution is, therefore, equal to its net multiplier minus the distribution generated by the account j to the other endogenous institutions. The additive formulae of the redistribution matrix R clarify the direction and magnitude of the changes in the relative position of the accounts. Specifically the distribution procedure among economic agents when there are exogenous inflows of income is shown as the result of combining effects with different meaning. The initial and exogenous injection received by the sectors ( I ) affects their relative status positively. The net multiplier effects on relative income D( M I) have either a positive or negative effect within the income distribution process. Finally, the initial relative income of the endogenous sectors ( I D) contributes to the changes in the relative income negatively. In particular, it is interesting to determine the contribution of the net multipliers (i.e., the matrix D( M I) ) to income distribution because this provides information about the changes in the relative position of the sectors as a result of the multiplier process. It is important to decompose the matrix of redistribution into different additive components. In our case we will see, through the elements of

matrix D( M I), a poor multiplier capacity for modifying the relative income of the endogenous accounts. 3. Multiplier Effects and Income Redistribution Effects in Korea (2000 and 2009) (1) Financial Crises and IT Revolution: An Overview In order to assess macroeconomic fundamentals during the period of 2000-2011, we have decomposed production industries into 4 sectors as shown in Table 2: (1) Higher IT-intensive (2) Lower IT -intensive and Primary (3) Higher IT-intensive s and (4) Lower IT-intensive s following Ha and Pyo (2004) and Pyo and Ha (2007). We have summarized major final demand indicators (Figure 1), exchange rate and stock market index (Figure 2) and production and export performance by industries (see Table 3). The trend in final demand indicators clearly marks two points of financial crises in 1997-1998 and 2007-2008 when both real GDP and real investment fell sharply. After the two financial crises, export not investment had led the way for recovery. The movement in exchange rate (Won/dollar rate) shows large-scale abrupt depreciation of Won and drop in KOSPI stock index just after each financial crisis in 1997 and 2007. The decomposed industrial performances of output and exports over the period of 2000-2011 clearly indicate the strong performance by higher IT-intensity manufacturing with output growth (7.8 %) and export growth (11.9 %) respectively. Table 2 Industrial Classification by IT-Intensity Main Category intensity Sub Category Main Category intensity Sub Category 1 agriculture and fishing 22 construction manufacturing sector lower ITintensity higher ITintensity 2 mining 26 transportation, storage lower ITintensity 3 food 29 real estate 4 textile, apparels, leather 32 government 5 wood 23 electicity, gas, water service 6 paper allied 24 trade sector 10 rubber and plastic 25 hotels and restaurants 11 stone, clay, glass higher ITintensity 27 communication 13 fabricated metal 28 finance, insurance 14 machinery 30 business services 16 electrical machinery 31 social and personal services 19 instrument 21 furniture and misc. manufacturing 7 printing and publishing 8 coal and petroleum product 9 chemicals 12 primary metal

15 computer and peripherals 17 electric components 18 sound, video, communication equipment 20 transportation equipment Sources: Ha and Pyo (2004) and Pyo and Ha (2007) Figure 1. GDP, Investment and Export Growth Rate: Korea 1995-2011 Source: The Bank of Korea Figure 2. Exchange Rate and Stock Price Index: Korea 1995-2011 Source: The Bank of Korea

Table 3. Average Growth Rate for different periods average growth rate 2000-2007 2008-2011 2000~2011 Gross Domestic Product 5.21 3.13 4.52 Agriculture, Forestry and Fishing 1.38 0.60 1.12 Mining, Quarrying and 8.10 5.73 7.31 8.18 5.83 7.39 s 4.53 2.63 3.89 Final Consumption Expenditure 4.71 2.38 3.93 Private 4.76 2.00 3.84 Government 4.53 3.73 4.26 Gross Fixed Capital Formation 4.46 0.45 3.13 Construction 2.95-2.03 1.29 Facilities Investment 7.13 4.65 6.30 Intangible Fixed Assets 7.95 3.40 6.43 Exports of Goods and s 11.60 7.40 10.20 Imports of Goods and s 10.69 5.05 8.81 average growth rate 2000-2007 2008-2011 2000-2011 Gross Domestic Product 5.21 3.13 4.52 (1)manufacturing industries 7.76 3.60 5.20 (2)Primary and manufacturing industries 5.11 2.67 3.86 (3)service industries 6.62 4.13 5.98 (4)service industries 6.36 3.87 5.60 Exports of Goods and s 11.60 7.40 10.20 (1)manufacturing industries 11.94 7.58 10.45 (2)Primary and manufacturing industries 5.11 3.08 4.25 (3)service industries 0.46 1.39 1.92 (4)service industries 5.12 2.80 3.86 Source: The Bank of Korea (2) Macro-SAM in 2000 and 2009 with Micro-SAM In order to identify multiplier effects and income redistribution effects for the period of 2000-2009, we have constructed SAM of Korea as shown in Table 4 for year 2000 and Table 5 for year 2009 by combining Input-Output Tables and National Accounts by the Bank of Korea in respective years. In order to construct a supplementary Micro-SAM, we have used Korea Labor & Income Panel Study (KLIPS) Database (1998-2008) to decompose household sector into 10-Decile units. They are presented in Table 6 for year 2000 and Table 7 for year 2009 which show sources of income (wage, profit, business transfer, government transfer and foreign transfer income). When we compare Table 6

with Table 7, the income distribution of profit has been markedly skewed in favor of higher income deciles. For example, 40.5 percent of the profit income was distributed to the highest income decile ( of 10) in 2000 but 52.7 percent in 2009. On the other hand, the business transfer income moved to the opposite direction; the highest income decile received 73.9 percent in 2000 but only 45.4 percent in 2009. In general, the distribution structure of gross income by 10 deciles changed in favor of lower income groups. The ratio of upper 20 % income to lower 20 % income was 17.1 in 2000 and 12.1 in 2009 respectively. The wage income of the lower income deciles ( of 1 and 2) was only 3.1 percent in 2000 but improved to be 3.7 percent in 2009. It was mainly due to the welfare policies for the lower income groups by President Kim (1998-2002) and President Roh administration (2003-2007). On the household expenditure side, we have used 2001 KLIPS data for Micro-SAM (2000) and 2008 KLIPS data for Micro-SAM (2009) as shown in Table 8 and 9 respectively. In general, the distribution structure of gross expenditure by 10 deciles changed in favor of lower income groups too. The ratio of upper 20 % expenditure to lower 20 % expenditure was 4.5 in 2000 and 3.6 in 2009 respectively. The gross expenditure of the lower income deciles ( of 1 and 2) was only 8.3 percent in 2000 but improved to be 10.1 percent in 2009. Because of differences in survey items between 2001 KLIPS data and 2008 KLIPS data, we had to use the same proportions of income and expenditure in both 2000 and 2009.

Table 4 Macro Social Accounting Matrix (South Korea, Year 2000, 1 billion won) Income/ Expenditure Production Activities Production Commodities Labor Capital Household Corporate enterprise Government Combined enterprise Rest of World Error term Total Production Activities 1,155,961 236,966 1,392,928 Production Commodities 793,283 352,371 61,653 188,443 1,395,750 Labor 267,134 696 267,830 Capital 194,087 6,954 201,041 Household 267,190 82,918 25,914 10,929 7,242 34,682 428,875 Corporate 108,609 7,788 59 0 116,456 Government 51,319 19,447 25,442 19,470 53 16,258 131,989 Combined enterprise 87,105 37,441 14,772 58,774 681 5,346 204,119 Rest of World 220,342 640 9,514 5,834 244 574 15,676 252,823 Error term 56,056 231 56,287 Total 1,392,928 1,395,750 267,830 201,041 428,875 116,456 131,989 204,119 252,823 56,287

Table 5 Macro Social Accounting Matrix (South Korea, Year 2009, 1 billion won) Income/ Expenditure Production Activities Production Commodities Labor Capital Household Corporate enterprise Government Combined enterprise Rest of World Error term Total Production Activities 2,240,903 534,074 2,774,977 Production Commodities 1,727,071 575,970 170,325 279,285 2,752,651 Labor 493,686 872 494,558 Capital 310,604 20,129 330,733 Household 493,035 108,529 42,761 40,567 15,899-8,776 692,015 Corporate 207,473 14,696 19 222,187 Government 101,522 17,131 59,479 35,824 195 47,143 261,295 Combined enterprise 142,094 27,771 104,550 48,447-449 -87,144 235,269 Rest of World 494,617 1,523 14,731 14,098 684 1,938-44,016 483,577 Error term 38,368-87,144-48,776 Total 2,774,977 2,752,651 494,558 330,733 692,015 222,187 261,295 235,269 483,577-48,776

Table 6 Household Income Sources by Decile ( Year 2000 ) Equalization income by decile(2000) Gross income Wage Profit income Business transfer income Government transfer income Foreign transfer income Error term of 1 0.004 0.008 0.011 0.009 0.255 0.061 0.002 of 2 0.023 0.023 0.036 0.022 0.233 0.266 0.009 of 3 0.043 0.043 0.036 0.015 0.119 0.134 0.021 of 4 0.059 0.061 0.059 0.029 0.113 0.084 0.026 of 5 0.075 0.077 0.098 0.016 0.064 0.068 0.037 of 6 0.093 0.099 0.034 0.017 0.056 0.044 0.042 of 7 0.112 0.117 0.079 0.043 0.044 0.054 0.044 of 8 0.130 0.133 0.109 0.039 0.046 0.063 0.043 of 9 0.164 0.168 0.133 0.072 0.042 0.052 0.083 of 10 0.297 0.273 0.405 0.739 0.028 0.174 0.692 Total 1.000 1.000 1.000 1.000 1.000 1.000 1.000 Table 7 Household Income Sources by Decile ( Year 2009 ) Equalization income by decile(2009) Gross income Wage Profit income Business transfer income Government transfer income Foreign transfer income Error term of 1 0.010 0.011 0.015 0.009 0.255 0.093 0.005 of 2 0.026 0.026 0.043 0.015 0.233 0.129 0.022 of 3 0.044 0.043 0.051 0.036 0.119 0.095 0.015 of 4 0.060 0.060 0.050 0.033 0.113 0.082 0.050 of 5 0.074 0.077 0.038 0.084 0.064 0.064 0.026 of 6 0.090 0.095 0.050 0.022 0.056 0.044 0.022 of 7 0.108 0.113 0.044 0.075 0.044 0.065 0.062 of 8 0.131 0.137 0.085 0.052 0.046 0.063 0.157 of 9 0.165 0.169 0.097 0.219 0.042 0.105 0.121 of 10 0.293 0.270 0.527 0.454 0.028 0.260 0.521 Total 1.000 1.000 1.000 1.000 1.000 1.000 1.000 Table 8. Household Expenditure Distribution by Decile ( Year 2000 ) Equalization expenditure by decile (2000) Gross expenditure Consumption Business transfer expenditure Government transfer expenditure Savings Foreign transfer expenditure of 1 0.043 0.050 0.021 0.015 0.010 0.025 of 2 0.040 0.047 0.021 0.023 0.017 0.037

of 3 0.057 0.064 0.043 0.044 0.031 0.056 of 4 0.072 0.079 0.058 0.063 0.036 0.076 of 5 0.084 0.089 0.082 0.083 0.056 0.087 of 6 0.098 0.101 0.101 0.100 0.084 0.096 of 7 0.112 0.112 0.131 0.119 0.091 0.112 of 8 0.122 0.126 0.130 0.145 0.110 0.135 of 9 0.151 0.140 0.157 0.170 0.188 0.151 of 10 0.221 0.192 0.258 0.238 0.379 0.224 Total 1.000 1.000 1.000 1.000 1.000 1.000 Table 9. Household Expenditure Distribution by Decile (Year 2009 ) Equalization expenditure by decile (2009) Gross expenditure Consumption Business transfer expenditure Government transfer expenditure Savings Foreign transfer expenditure of 1 0.053 0.038 0.008 0.015 0.002 0.025 of 2 0.048 0.049 0.015 0.023 0.009 0.037 of 3 0.055 0.064 0.031 0.044 0.025 0.056 of 4 0.069 0.078 0.049 0.063 0.040 0.076 of 5 0.080 0.089 0.070 0.083 0.059 0.087 of 6 0.095 0.099 0.097 0.100 0.097 0.096 of 7 0.112 0.118 0.115 0.119 0.120 0.112 of 8 0.126 0.130 0.151 0.145 0.115 0.135 of 9 0.145 0.143 0.182 0.170 0.189 0.151 of 10 0.217 0.193 0.282 0.238 0.343 0.224 Total 1.000 1.000 1.000 1.000 1.000 1.0 (3)Multiplier Effects In both Table 10 and Table 11, we have shown multiplier income effects on four sectors of industries, (1) with, (2) with, (3) with and (4) with. The row average of each table represents average sensitivity of the sector when expenditure is injected from each of four sectors. In both Table 9 and 10, the multiplier effect is highest with the Sector (2) and lowest with the Sector (3). By comparing two tables, we can see that each sector s income multiplier sums in all four sectors have become larger in 2009 than in 2000. The sector (3) with had the lowest average sensitivity, 0.433 and 0.411 respectively. Therefore, we can conclude that sector (3) with low-it intensity has both the lowest average sensitivity and the lowest income multiplier effect, which is a typical nature of the non-tradable service sector.

Table 10 Multiplier Contribution ( M ) to Income Distribution in Production (Year 2000) Production Activities Income Expenditure (1) (2) (3) (4) nn (1) Production Activities (2) (3) (4) Average Sensitivity 1.791 0.770 0.639 0.661 0.965 0.447 1.527 0.426 0.438 0.709 0.172 0.170 1.188 0.203 0.433 0.563 0.553 0.579 1.639 0.833 Total Effect 2.974 3.020 2.831 2.942 Table 11 Multiplier Contribution ( M ) to Income Distribution in Production (Year 2009) Production Activities Income Expenditure (1) (2) (3) (4) nn (1) Production Activities (2) (3) (4) Average Sensitivity 1.807 0.794 0.659 0.682 0.985 0.427 1.554 0.417 0.411 0.702 0.154 0.145 1.165 0.182 0.411 0.635 0.624 0.667 1.759 0.921 Total Effect 3.022 3.116 2.907 3.034 In Table 12 and 13, we have summarized multiplier contribution of production to household income by 10-decile groups. In both 2000 and 2009, total multiplier contribution was highest in sector (3) lower IT-intensity service sector such as Construction, Transportation and Storage, Real Estate and Government. These industries have lower intermediate inputs and higher value-added and therefore, when the production activity increases, the income generation through multiplier effects is relatively larger than other sectors. In general though, the multiplier effects of production to household income have become smaller in 2009 than in 2000. In addition, we can point out that the average sensitivity by each unit of household is larger as we move to upper income units. It also became smaller in 2009 than in 2000.

Table 12 Multiplier Contribution ( M ) of Production to Household Income (Year 2000) Income Expenditure nn (1) Production Activities (2) (3) (4) Average Sensitivity 1 0.006 0.005 0.007 0.007 0.006 2 0.017 0.015 0.021 0.020 0.018 3 0.026 0.024 0.033 0.032 0.029 4 0.039 0.036 0.049 0.048 0.043 Equalization household by decile 5 0.051 0.047 0.065 0.063 0.056 6 0.050 0.047 0.067 0.063 0.057 7 0.067 0.062 0.087 0.083 0.075 8 0.078 0.073 0.102 0.098 0.088 9 0.100 0.093 0.130 0.125 0.112 10 0.224 0.206 0.275 0.271 0.244 Total Effect 0.657 0.607 0.835 0.810 Table 13 Multiplier Contribution ( M ) of Production to Household Income (Year 2009) Income Expenditure nn (1) Production Activities (2) (3) (4) Average Sensitivity 1 0.007 0.006 0.009 0.009 0.008 2 0.016 0.015 0.021 0.021 0.018 3 0.025 0.023 0.034 0.032 0.028 4 0.031 0.029 0.043 0.041 0.036 Equalization household by decile 5 0.039 0.036 0.054 0.051 0.045 6 0.045 0.042 0.064 0.060 0.053 7 0.054 0.050 0.077 0.071 0.063 8 0.067 0.062 0.094 0.088 0.078 9 0.089 0.082 0.123 0.116 0.102 10 0.186 0.170 0.244 0.238 0.210 Total Effect 0.559 0.515 0.763 0.727

In Table 14 and 15, we have presented inter-household multiplier contribution to income in 2000 and 2009. It shows that when an income is injected into the lowest income unit ( of 1), it has the largest income effect. The effect becomes smaller as we move toward higher income deciles. It is not a surprising result because the consumption expenditure to income increase is more sensitive in lower income brackets than in higher income brackets. On the other hand, if an equal amount was injected across all units of income decile, the highest income unit gains the most. It should also be pointed out that the column sum of two tables, Table 14 and 15, is the largest in of 1 in both 2000 and 2009. This is because the lower income groups pay lower taxes and pension contributions while they tend to have higher propensity to consume. Since the government sector is regarded as an exogenous account and the higher income groups pay higher taxes and pension contributions, their propensity to consume will be less than lower income groups. On the other hand, average sensitivity becomes higher as we move from lower income units to higher income units in both 2000 and 2009. It indicates that higher income units derive higher multiplier effects and marginal benefits.

Table 14 Multiplier Contribution ( M ) to Income Distribution in Household (Year 2000) Income Expenditure nn Equalization household by decile 1 2 3 4 5 6 7 8 9 10 Average Sensitivity 1 1.005 0.005 0.005 0.005 0.005 0.005 0.005 0.004 0.004 0.004 0.105 2 0.015 1.015 0.015 0.014 0.014 0.014 0.014 0.013 0.013 0.012 0.114 3 0.024 0.023 1.023 0.022 0.022 0.021 0.021 0.021 0.020 0.019 0.121 4 0.036 0.035 0.034 1.033 0.032 0.032 0.031 0.031 0.030 0.028 0.132 Equalization 5 0.047 0.045 0.044 0.044 1.042 0.041 0.041 0.041 0.039 0.037 0.142 household by decile 6 0.047 0.046 0.044 0.044 0.042 1.042 0.041 0.041 0.039 0.037 0.142 7 0.062 0.060 0.059 0.058 0.056 0.055 1.055 0.054 0.052 0.049 0.156 8 0.073 0.071 0.069 0.068 0.066 0.064 0.064 1.063 0.061 0.057 0.166 9 0.093 0.090 0.088 0.086 0.084 0.082 0.082 0.081 1.077 0.073 0.184 10 0.205 0.199 0.194 0.191 0.186 0.183 0.182 0.180 0.172 1.164 0.286 Total Effect 1.608 1.589 1.574 1.565 1.549 1.538 1.535 1.529 1.506 1.480 Table 15 Multiplier Contribution ( M ) to Income Distribution in Household (Year 2009) Income Expenditure nn Equalization household by decile 1 2 3 4 5 6 7 8 9 10 Average Sensitivity 1 1.006 0.006 0.006 0.006 0.006 0.006 0.006 0.005 0.005 0.005 0.106 2 0.015 1.015 0.014 0.014 0.014 0.013 0.013 0.013 0.013 0.013 0.114 3 0.023 0.023 1.022 0.022 0.021 0.021 0.021 0.021 0.020 0.020 0.121 4 0.029 0.029 0.028 1.027 0.027 0.026 0.026 0.026 0.025 0.025 0.127 Equalization 5 0.037 0.036 0.035 0.034 1.034 0.033 0.033 0.033 0.032 0.031 0.134 household by decile 6 0.043 0.043 0.041 0.040 0.039 1.038 0.038 0.038 0.037 0.036 0.139 7 0.052 0.051 0.050 0.048 0.047 0.046 1.046 0.046 0.045 0.044 0.147 8 0.064 0.063 0.061 0.060 0.058 0.057 0.057 1.057 0.055 0.054 0.159 9 0.084 0.083 0.081 0.079 0.077 0.076 0.075 0.075 1.073 0.072 0.177 10 0.173 0.172 0.166 0.162 0.159 0.156 0.155 0.154 0.150 1.147 0.260 Total Effect 1.527 1.522 1.505 1.492 1.481 1.472 1.471 1.468 1.456 1.446

(4) Redistribution Effects When there is income generated by one account, the income generated is distributed not equally to each sector and it becomes a zero-sum game. We can analyze how income generated by each production sector gets redistributed across different income units of household. For example, we can identify the induced production activities of four endogenous sectors by increased final demand in three exogenous accounts. Then we can trace how the income generated in endogenous sectors gets redistributed among household income units. Table 16 and 17 is the matrix of such redistributed income which sums up to zero. The result is different from multiplier effects. When income is generated from the entire production sectors, the Higher IT-intensity sector has the lowest income redistribution effect. In particular, the sector s redistribution effect became negative (-1.019) while all other three sectors have had positive redistribution effects. On the other hand, when income is injected to production activity, the lower ITintensity sector has the lowest column sum like the sector s multiplier effects and it became lower in 2009 than in 2000.

Table 16 Redistributed Income Matrix ( ( ey ' ) R ) in Production (Year 2000) Production Activities Income Expenditure (1) (2) (3) (4) (1) n nn Production Activities (2) (3) (4) Average Sensitivity 0.858-0.155-0.294-0.299 0.027-0.346 0.739-0.367-0.378-0.088-0.062-0.063 0.954-0.038 0.198-0.080-0.086-0.065 0.977 0.187 Total Effect 0.369 0.436 0.227 0.262 Table 17 Redistributed Income Matrix ( ( ey ' ) R ) in Production (Year 2009) Production Activities Income Expenditure (1) (2) (3) (4) (1) n nn Production Activities (2) (3) (4) Average Sensitivity 0.910-0.112-0.252-0.259 0.072-0.512 0.605-0.537-0.575-0.255-0.040-0.050 0.968-0.021 0.214-0.027-0.045-0.006 1.065 0.247 Total Effect 0.330 0.397 0.173 0.210

Table 18 and 19 are the results of redistributed income effects of production activity in four sectors on household income by ten income units. In both 2000 and 2009, the lower IT-intensity sector has generated the largest total income redistribution effect (0.034 and 0.081). In terms of average sensitivity, the highest income group, 10, has had the largest negative sensitivity (-0.040) in 2000 but it became the only positive sensitivity (0.002) in 2009. Table 18 Redistributed Income Effect ( ( ey ' n) R nn) of Production on Household Income (Year 2000) Equalization household by decile Production Activities Expenditure (1) (2) (3) (4) Average Sensitivity Income 1-0.006-0.007-0.005-0.006-0.006 2-0.010-0.011-0.006-0.007-0.008 3-0.008-0.009 0.000-0.002-0.005 4-0.008-0.010 0.003 0.000-0.004 5-0.008-0.012 0.006 0.002-0.003 6-0.010-0.013 0.007 0.002-0.003 7-0.010-0.015 0.010 0.004-0.003 8-0.011-0.016 0.012 0.006-0.002 9-0.015-0.021 0.015 0.006-0.004 10-0.059-0.075-0.008-0.020-0.040 Total Effect -0.145-0.189 0.034-0.015 Table 19 Redistributed Income Effect ( ( ey ' n) R nn) of Production on Household Income (Year 2009) Equalization household by decile Production Activities Expenditure (1) (2) (3) (4) Average Sensitivity Income 1-0.012-0.013-0.010-0.011-0.011 2-0.013-0.014-0.008-0.009-0.011 3-0.009-0.011 0.000-0.003-0.006 4-0.009-0.012 0.002-0.002-0.005 5-0.009-0.012 0.006 0.001-0.003 6-0.009-0.013 0.009 0.003-0.003 7-0.010-0.015 0.011 0.004-0.002 8-0.011-0.016 0.015 0.007-0.001 9-0.014-0.021 0.019 0.009-0.002 10-0.018-0.036 0.037 0.024 0.002 Total Effect -0.113-0.163 0.081 0.022

Lastly we examine the redistribution effect of household income injected on household income redistributed among different income units. Table 20 and 21 are redistributed household income among different income units in 2000 and 2009 respectively. In general the total effects are positive across all income units. But within each unit, the redistribution effect is positive only to itself and negative in all other sectors. In other words, the redistributed income matrix of households has positive diagonal elements but all negative off-diagonal elements. The total redistribution effect is the largest in the highest income unit and became smaller as we move lower income units in both 2000 and 2009. The total effect was larger in 2009 than in 2000 across all units. On the other hand average sensitivity moved in the opposite direction becoming lower as we move from lower income units to higher ones. It implies when income is injected to household account, the higher income units derive higher redistribution effects. On the other hand, the redistribution effect becomes more sensitive with lower income groups because their income level are at lower level and therefore, they are affected marginally more. It is noted that average sensitivity of 10 was negative (-0.002) in 2000 but it became positive (0.049) in 2009 which implies after the global financial crisis of 2007-2008, the highest income unit ( 10) has become net gainers than net losers in redistribution of household income.

Table 20. Redistributed Income Effect ( ( ey ' n) R nn) in Household (Year 2000) Expenditure Equalization household by decile Income Average Sensitivity 1 2 3 4 5 6 7 8 9 10 1 0.992-0.008-0.008-0.008-0.008-0.008-0.007-0.007-0.007-0.007 0.092 2-0.014 0.986-0.014-0.014-0.013-0.013-0.013-0.013-0.013-0.012 0.087 3-0.014-0.013 0.987-0.013-0.013-0.012-0.012-0.012-0.012-0.012 0.087 4-0.016-0.015-0.015 0.985-0.015-0.015-0.015-0.015-0.014-0.014 0.085 Equalization 5-0.019-0.019-0.018-0.018 0.982-0.018-0.018-0.018-0.017-0.017 0.082 household 6-0.020-0.019-0.019-0.019-0.019 0.982-0.018-0.018-0.018-0.017 0.081 by decile 7-0.024-0.023-0.023-0.023-0.023-0.022 0.978-0.022-0.022-0.021 0.078 8-0.027-0.027-0.026-0.026-0.026-0.025-0.025 0.975-0.025-0.024 0.074 9-0.035-0.035-0.034-0.034-0.033-0.033-0.033-0.033 0.968-0.031 0.067 10-0.111-0.108-0.106-0.105-0.103-0.101-0.100-0.100-0.097 0.907-0.002 Total Effect 0.712 0.719 0.724 0.726 0.730 0.734 0.735 0.737 0.743 0.752 Table 21 Redistributed income matrix (( ey ' n) R nn) in Household (Year 2009) Expenditure Equalization household by decile Income Average Sensitivity 1 2 3 4 5 6 7 8 9 10 1 0.986-0.014-0.014-0.014-0.013-0.013-0.013-0.013-0.013-0.012 0.087 2-0.017 0.983-0.016-0.016-0.016-0.015-0.015-0.015-0.015-0.015 0.084 3-0.014-0.014 0.986-0.013-0.013-0.013-0.013-0.013-0.013-0.012 0.087 4-0.016-0.016-0.015 0.985-0.015-0.015-0.015-0.014-0.014-0.014 0.085 Equalization 5-0.016-0.016-0.015-0.015 0.985-0.015-0.015-0.015-0.014-0.014 0.085 household 6-0.018-0.017-0.017-0.017-0.017 0.984-0.016-0.016-0.016-0.016 0.083 by decile 7-0.020-0.020-0.019-0.019-0.019-0.019 0.981-0.018-0.018-0.018 0.081 8-0.023-0.023-0.022-0.022-0.022-0.022-0.022 0.979-0.021-0.021 0.078 9-0.030-0.030-0.029-0.029-0.028-0.028-0.028-0.028 0.973-0.027 0.072 10-0.055-0.054-0.053-0.052-0.052-0.051-0.051-0.050-0.050 0.951 0.049 Total Effect 0.777 0.781 0.786 0.789 0.791 0.794 0.794 0.796 0.799 0.802

4. CONCLUSION In the present paper, we have examined the impact of two financial crises and IT revolution on multiplier effects and redistribution effects of both production sectors and household income units in Korea for the period of 2000-2011. We have decomposed the entire production activities into four sectors by the ranking of It-intensity. As a consequence of depreciation in the post-crisis periods, there was a strong export performance in higher IT-intensity Sector. This has generated a strong exogenous impact on four endogenous sectors of production. It also has generated income redistribution effects among different sectors of production and across different income units. In general, the distribution structure of gross income by 10 deciles changed in favor of lower income groups. The ratio of upper 20 % income to lower 20 % income was 17.1 in 2000 and 12.1 in 2009 respectively. The wage income of the lower income deciles ( of 1 and 2) was only 3.1 percent in 2000 but improved to be 3.7 percent in 2009. It was mainly due to the welfare policies for the lower income groups by President Kim (1998-2002) and President Roh administration (2003-2007). In analysis of multiplier effects, we have found that the sector (3) with had the lowest average sensitivity, 0.433 and 0.411 respectively. Therefore, we can conclude that sector (3) with low-it intensity has both the lowest average sensitivity and the lowest income multiplier effect, which is a typical nature of the non-tradable service sector. On the other hand, the multiplier effects of production to household income have become smaller in 2009 than in 2000. In addition, we can point out that the average sensitivity by each unit of household is larger as we move to upper income units. It also became smaller in 2009 than in 2000. In analysis of redistribution effects, the lower IT-intensity sector has generated the largest total income redistribution effect (0.034 and 0.081) in both 2000 and 2009,. In terms of average sensitivity, the highest income group, 10, has had the largest negative sensitivity (-0.040) in 2000 but it became the only positive sensitivity (0.002) in 2009. In general the total redistribution effects are positive across all income units. But within each unit, the redistribution effect is positive only to itself and negative in all other sectors. In other words, the redistributed income matrix of households has positive diagonal elements but all negative off-diagonal elements. The total redistribution effect is the largest in the highest income unit and became smaller as we move lower income units in both 2000 and 2009. The total effect was larger in 2009 than in 2000 across all units. On the other hand average sensitivity moved in the opposite direction becoming lower as we move from lower income units to higher ones. It implies when income is injected to household account, the higher income units derive higher redistribution effects. These finding provide evidence increasing globalization of production activities in Korea and IT-intensity deepening has generated more redistribution effect in favor of higher income units. However the overall indicator of income disparity has not been worsened between 2000 and 2009 after two financial crises. When we measure it as the ratio of upper 20 percent income divided by lower 20 % income, the income disparity has been narrowed.

References Dadush, U., Dasgupta, D. & Uzan, M. (2000), Private Capital Flows in the Age of Globalization: the Aftermath of the Asian Crisis", Edward Elgar, Cheltenham Defourny, J. & Thorbecke, E (1984), "Structural path analysis and multiplier decomposition within a social accounting matrix framework", Economic Journal Vol. 94, No. 373(March, 1984) pp.111-36. Ha, B.C. & Pyo, H.K. (2004), "The Measurement of IT Contribution by Decomposed Dynamic Input-Output Tables in Korea (1980-2002)", Seoul Journal of Economics 17, pp.511-546. Keuning, S. & Thorbecke, E (1992), "The social accounting matrix and adjustment Policies: the impact of budget retrenchment on income distribution", Chapter 3 of E.Thorbecke (Ed.), Adjustment and Equity in Indonesia Paris, OECD. Otsu, K. & Pyo, H.K. (2009), "A Comparative Estimation of Financial Frictions in Japan and Korea", Seoul Journal of Economics Vol.22 no.1, pp.95-121 : Institute of Economics Research, Seoul National University Pyatt, G.(2003) "An Alternative Approach to Poverty Analysis", Economic Systems Research, Vol. 15, No.4(June) pp.113-133. Pyatt,G., & Round, J.I.(1979), "Accounting and fixed price multipliers in a social accounting matrix framework", Economic Journal Vol.89, pp.850-873. Pyatt,G., & Round, J.I.(2003), "Multiplier analysis and the design of social accounting matrices" University of Warwick. Pyatt,G., & Round, J.I.(2006), "Multiplier Effects and The Reduction of Poverty", In Alain Janvry & Ravi Kanbur (Ed.), Economic Studies in Inequality, Social Exclusion and Well-Being (pp.233-259). Springer US. Pyo, H.K. (2004), "Interdependency in East Asia and the Post-Crisis Macroeconomic Adjustment in Korea", Seoul Journal of Economics Vol.17 No.1, pp117-152 Pyo, H.K. & Ha, B.C. (2007), "A Test of Separability and Random Effects in Production Function with Decomposed IT capital", Hitotsubashi Journal of Economics 48, pp.67-81 Pyo, H.K., Kim, D.K. & Lee, G (2012), An Analysis of Effects on National Pension through Socio- Economic Changes by Using Social Account Matrix, National Pension System Working Report 2012-02.

Roh, Y.H. & Nam, S.H.(2006), "Redistributed Income Effect of Korean Economy with Social Accounting Matrix", The Bank of Korea(2006) Roland-Holst, D. W. & Sancho, F. (1992), "Relative Income Determination In The ed States: A Social Accounting Perspective", Review of Income and Wealth Vol.38, pp311 327. Round, J.I.(2003), "Social Accounting Matrices and SAM=based Multiplier Analysis", Chapter 14 in F Bourguignon, and L A Pereira da Silva (editors) Techniques and Tools for Evaluating the Poverty Impact of Economic Policies, World Bank and Oxford University Press. Saari, M.Y., Dietzenbacher, E. & Los, B. (2010), "Modeling Impact of Higher Energy Prices on Income Distribution with Substitutions in Production and Household Sectors," U.Groningen. Stone, R. (1985), "The disaggregation of the household sector in the national accounts", Chapter 8 of Pyatt, G. and J.I. Round (Ed.) Social Accounting Matrices: A Basis for Planning Washington, D.C., the World Bank Thorbecke E.& Jung H.S.(1996), "A multiplier decomposition method to analyze poverty alleviation", Journal of Development Economics, 48 (2), pp. 279-300.

Appendix Table A. Detailed Industrial Classification Main Category Manufac turing sector intensity Low ITintensity High ITintensity Sub Category co de Standard of year 2000 Standard of year 2009 Name of Sector co de Name of Sector 1 Crops 1 Crops 1 agriculture and fishing 2 Livestock breeding 2 Animals 3 Forestry products 3 Forest products 4 Fishery products 4 Fishery products Mining of coal, crude petroleum and 5 Coal mining 6 natural gas 2 mining 6 Crude petroleum and natural gas 7 Metal ores 7 Metal ores 9 Meat and dairy products 9 Meat and dairy products 10 Processed seafood products 10 Processed seafood products 11 Polished grains, flour and milled cereals 11 Polished grains, flour and milled cereals 12 Sugar and starches 12 Other food products Bakery and confectionery products, 13 13 Beverages noodles 3 food 14 Seasonings and fats and oils 14 Prepared livestock feeds 15 Canned or cured fruits and vegetables and misc. food preparations 15 Tobacco products 16 Beverages 17 Prepared livestock feeds 18 Tobacco products 19 Fiber yarn 16 Fiber yarn and fabrics 20 Fiber fabrics 17 Apparels and other textiles 4 textile, apparels, leather 21 Wearing apparels and apparel accessories 18 Leather and fur products 22 Other fabricated textile products 23 Leather and fur products 5 wood 24 Wood and wooden products 19 Wood and wooden products 6 paper allied 25 Pulp and paper 20 Pulp and paper 10 rubber and plastic 36 Plastic products 30 Plastic products 37 Rubber products 31 Rubber products 8 Nonmetallic minerals 8 Non-metallic minerals 38 Glass products 32 Glass products 11 stone, clay, glass 39 Pottery and clay products 33 Ceramic ware 40 Cement and concrete products 34 Cement and concrete products 41 Other nonmetallic mineral products 35 Other nonmetallic mineral products 42 Pig iron and crude steel 36 Pig iron and crude steel 43 Primary iron and steel products 37 Primary iron and steel products 13 fabricated metal Nonferrous metal ingots and Nonferrous metal ingots and primary 44 38 primary nonferrous metal products nonferrous metal products 14 machinery 46 Machinery and equipment Machinery and equipment 40 of general purpose of general purpose 47 Machinery and equipment Machinery and equipment 41 of special purpose of special purpose Electronic machinery, 48 16 electrical machinery equipment, and supplies 42 Electrical equipment, and supplies 52 Household electrical appliances 46 Household electrical appliances 19 instrument 57 Furniture 51 Furniture 53 Precision instruments 47 Precision instruments 21 furniture and misc. manufacturing 58 Other manufacturing products 52 Other manufactured products 7 printing and publishing 26 Printing, publishing and Printing and reproduction 21 reproduction of recorded media of recorded media 8 coal and petroleum product 27 Coal products 22 Coke and hard-coal 28 Petroleum refinery products 23 Refined petroleum products 29 Organic basic chemical products 24 Basic chemical products 30 Inorganic basic chemical products 25 Synthetic resins and synthetic rubber 31 Synthetic resins and synthetic rubber 26 Chemical fibers 9 chemicals 32 Chemical fibers 27 Fertilizers and agricultural chemicals 33 Fertilizers and agricultural chemicals 28 Drugs, cosmetics, and soap 34 Drugs, cosmetics, and soap 29 Other chemical products 35 Other chemical products 12 primary metal 45 Fabricated metal products 39 Fabricated metal products except machinery and funiture 15 computer and peripherals 45 Computer and office equipment 17 electric components 49 Electronic components and accessories 43 Electronic components and accessories 18 sound, video, communication Radio, television and Audio, video and communications 50 44 equipment communications equipment equipment 54 Motor vehicles and parts 48 Motor vehicles and parts 20 transportation equipment 55 Ship building and repairing 49 Ship building and repairing 56 Other transportation equipment 50 Other transportation equipment

service sector Low ITintensity High ITintensity 22 construction 61 Building construction and repair 55 Building construction and repair 62 Civil Engineering 56 Civil engineering 65 Transportation and warehousing 59 Land transport 60 Water and air transport 26 transportation, storage Strorage and support activities 61 for transportation 29 real estate 68 Real estate agencies and rental 65 Real estate 70 Public administration and defense 69 Public administration and defense 71 Educational and research services 70 Education 32 gonerment 72 Medical and health services, and social security 71 Medical and health services 72 Social work activities 73 Sanitary services 23 electicity, gas, water 59 Electric services 53 Electric utilities 60 Gas and water supply 54 Gas and water supply 24 trade 63 Wholesale and retail trade 57 Wholesale and retail trade 25 hotels and restaurants 64 Eating and drinking places, and hotels and other lodging places 58 Accommodation and food services 27 communication 66 Communications and broadcasting 62 Communications services 28 finance, insurance 67 Finance and insurance 64 Finance and insurance 51 Computer and office equipment 5 Agriculture, forestry and fishing related services 30 business services 69 Business services 66 Research and development 75 Office supplies 67 Business services 68 Other business services 73 Culture and recreational services 63 Broadcasting 74 Other services 74 Publishing and cultural services 31 social and personal services 76 Business consumption expenditure 75 Amusement and sports activities 77 Nonclassifiable activities 76 Social organizations 77 Other services 78 Dummy sectors

Table B-1 Micro Social Accounting Matrix 25 25 (South Korea, Year 2000, 1 billion won) Income/Expenditure Production Activities Production Commodities Equalization household by decile highintensity lowintensity highintensity lowintensity highintensity lowintensity highintensity Production Activities Manufac- Manufac- lowintensitintensity high- lowintensity highintensity Manufaclowintensity Production Commodities Manufac- high- intensity lowintensity Labor Capital 1 2 3 Equalization household by decile 4 5 6 7 8 9 10 Corporate enterprise Government Combined Rest of Error capital World term ccounts 0 0 0 0 222493 113301 47934 49834 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 65562 0 499123 0 0 0 0 54841 162688 49454 40096 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 117212 0 424290 0 0 0 0 12686 12369 39668 38136 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 22509 0 125368 0 0 0 0 57226 44552 69166 141519 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 31683 0 344146 152686 77753 32895 34199 0 0 0 0 0 0 3475 3151 4226 5618 6505 7147 8132 9142 10309 13813 0 0 39018 0 0 408070 37635 111645 33938 27516 0 0 0 0 0 0 1141 1107 1530 1951 2344 2597 3029 3437 3778 4605 0 0 27600 0 0 263853 lowintensity 8706 8488 27222 26171 0 0 0 0 0 0 7254 6235 8168 9097 10106 11200 12244 12987 14316 20519 0 60733 96298 0 0 339745 highintensity 39271 30574 47466 97118 0 0 0 0 0 0 5656 6068 8800 11017 12446 14485 16225 18851 20869 28789 0 920 25527 0 0 384082 Labor 42137 28923 105275 90798 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 696 0 267830 Capital 45912 26791 56889 64495 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6954 0 201041 1 0 0 0 0 0 0 0 0 2016 885 0 0 0 0 0 0 0 0 0 0 232 2785 0 444 86 6447 2 0 0 0 0 0 0 0 0 6014 2947 0 0 0 0 0 0 0 0 0 0 565 2549 0 1927 314 14316 3 0 0 0 0 0 0 0 0 11538 2989 0 0 0 0 0 0 0 0 0 0 379 1304 0 968 742 17919 4 0 0 0 0 0 0 0 0 16294 4926 0 0 0 0 0 0 0 0 0 0 763 1230 0 605 885 24702 5 0 0 0 0 0 0 0 0 20588 8092 0 0 0 0 0 0 0 0 0 0 412 699 0 495 1285 31570 6 0 0 0 0 0 0 0 0 26329 2843 0 0 0 0 0 0 0 0 0 0 441 610 0 322 1456 32000 7 0 0 0 0 0 0 0 0 31132 6551 0 0 0 0 0 0 0 0 0 0 1102 477 0 390 1541 41193 8 0 0 0 0 0 0 0 0 35425 9071 0 0 0 0 0 0 0 0 0 0 1006 507 0 455 1484 47947 9 0 0 0 0 0 0 0 0 44911 11018 0 0 0 0 0 0 0 0 0 0 1853 458 0 379 2883 61502 10 0 0 0 0 0 0 0 0 72945 33596 0 0 0 0 0 0 0 0 0 0 19163 310 0 1258 24006 151278 Corporate enterprise 0 0 0 0 0 0 0 0 0 108609 160 160 334 453 641 787 1017 1010 1219 2006 0 59 0 0 0 116456 Government 8542 15407 13528 13842 13038 4945 0 1464 0 0 375 582 1107 1614 2120 2533 3040 3677 4334 6060 19470 0 0 53 16258 131989 Combined capital 15973 15787 31730 23614 0 0 0 0 0 0 361 627 1145 1333 2102 3137 3404 4110 7027 14195 14772 58774 0 681 5346 204119 acounts Rest of World 0 0 0 0 110816 70456 11959 27111 640 9514 148 219 327 441 508 560 656 789 879 1308 244 574 15676 0 0 252824 Error term 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 56056 0 0 231 0 56287 Total 350862 315370 348944 377753 471100 408310 218181 298159 267830 201041 18570 18149 25638 31524 36772 42447 47747 54003 62730 91296 116456 131989 204119 252823 56287 Total