Several top income papers provide empirical evidence on an inverse relation between the top income series and the top marginal tax rates.30 We have constructed the top marginal tax rates in effect for the top 0.1 income individuals during the market economy periods before and after the communist era, displayed on Figure C.5 in Appendix. If we consider the communist system as an extremely high taxation of income above an income threshold, then the overall negative relationship between top tax rates and top income shares is apparent in Hungary during the overall analysed time period. However, when focusing at subperiods between 1927-1944 and 1992-2008 we do not see direct negative relation between movements in the top tax rates and shares, suggesting that the top tax rates are not the only determinants behind the changes in top income shares. We further investigate the question of the channels generating income disparities by exploiting the exogenous shock of the planned economy and its equality by design. The analyses suggests that both secured capital ownership right, and liberalized wage settings played significant role in the evolvement of the top shares. By constructing the Hungarian top income series from 1914 up till recent years, we can also study how quickly the top income shares are reverted back after the end of the planned economy to prevalent level at Western European market economies, and also that which income factor is behind the surge in the shares.
The factor decomposition of earners at the top 1 percent is displayed in Figure 3.8. As can be seen from the figure, the interwar period top income shares are high and they come from large capital
29The decreasing trend in business income from 2001 is unlikely to be due to simple reorganization of tax labels as only one main item was excluded from business income, but from several years earlier from 1996. From 2003 entrepreneurs with income below 25 million HUF could choose to declare their income in a simplified system (Egyszerűsített vállalkozói adó- EVA), in this case they are not present in the personal income tax statistics, hence we can not include their income in the top income share calculation.
30See: Saez (2004) on USA, and the Chapters on India, Japan, Argentina, Italy, Sweden in Top Incomes Global Perspective (2010).
CEUeTDCollection
Figure 3.8: Decomposition of the top 1 percent income share, 1932-2008
Note: Capital: income from capital assets, land and buildings, for 1992-2008 also realized capital gains are included. Labor:
wages and salaries and other employment income. Business: mixed income. See Table C.12 for detailed income categories.
Source:Table C.21
structures as well as land and real-estate. Approximately 40 percent of the top 1 share income is from capital, that is similar to the US trend as can be seen in Figure C.13 in the Appendix. During the planned economy, the top income shares are solely composed of labor income, since profits accrue to the unique capital owner. After the transition, income shares of both capital and labor increase in tandem, as capital begins to be remunerated and wage settings were liberalized. Based on the share decomposition in Figure C.13 the significance of capital income in the top shares grew rapidly during the last two decades, from 5 percent to 20 percent – which is in line with the USA level – among individuals in the top 1 percent. These finding are also in line with the statement that both capital and labor income factors play a significant role at increasing income inequality when market forces determine income. In the next subsection we look into this mechanism how capital and labor income creates a surge in the top income shares.
3.4.1 Capital
Capital income was significant part of the top income shares at the beginning of the 20th century.
As can be seen in Figure 3.5, in the years immediately after the introduction of the personal income tax in 1914 more than half of the income of the top 0.1% originated from capital including land, buildings and financial assets. Between 1914 and 1932 the country underwent the World War, a hyperinflation episode in 1923-24, the Great Depression and a banking crises leading to a drop in the financial asset
CEUeTDCollection
component from 18% to below 5% (see Table C.18 in the Appendix). The overall capital income component remained around 50% during the interwar period.
Figure 3.9: Capital to GDP ratios
Note: Estimates of fixed capital to GDP, net of depreciation, based on two different methodologies and on two different sources of data (see Appendix C.8). For 1995-2010 assets are calculated at market value. For 1959-1980 half of the assets such as dwellings, roads, bridges, dams, private sector assets is valued at replacement value, while the other half is valued at book value (1968 prices, and 1976 prices).
Source:Table C.23
As mentioned earlier after World War II the market economy was quickly converted into a planned economy by forced collectivization of agriculture, nationalization of the industrial companies and banks.
The state owned the right to manage all productive assets and extract the profit from them. Income from private property and capital was practically eliminated. Households had limited capacity to own real or financial assets, property rights were not secured and investment possibilities were lacking.
After the transition into the market economy capital to GDP ratios adjusted quickly, and have been subject to the economy’s opening to international capital flows (see Figure 3.9). Changes in fixed capital formation have been remarkable, as was the program of decentralization of property ownership rights through privatization, and the transition from a single capital owner (the State) to multiple ones (the domestic and international markets). Figure 3.10 displays that the capital income share of GDP rapidly reached levels of regularity met in market economies, illustrating also that the remuneration proportion of capital in the total gross income increased substantially.
Which segment of the income distribution benefitted the most from the decentralization of capital ownership and secure private property rights? And which fraction gained the most from the increased share of capital remuneration in the overall economy? After the transition the significance of the capital
CEUeTDCollection
Figure 3.10: Proxies of capital income share of GDP
Note: Proxies of capital income share of GDP, based on two different methodologies and on two different sources of data (see Appendix C.9). For 1991-2011 the series report the capital factor share (gross operating surplus of households and firms). For 1968-1982 the series report the net income the state extracted as the owner from enterprises, i.e. profit and income tax. An alternative series includes additionally also the net of production subsides and production tax.
Source:Table C.24
component including realized capital gain at the very top of the income distribution quickly recovered, reaching even the interwar levels. Those at the top top 1% and top 0.1% received respectively more than 25% and 50% of their income from capital holdings during the years preceding the recent financial crises (see Table C.19). Meanwhile the lower fractiles received much smaller shares of their income from capital. In just two decades the significance of capital income component at the very top of the distribution became supreme, reaching comparable levels even to the USA, a country with high capital income concentration.31 It is clear from the income decomposition in Figure 3.8, that the capital income component was a strong drive behind the surge in the top income shares after the transition to the market economy. Beside capital, labor income also had a significant role in increasing top income shares, to which we turn in the next subsection.
3.4.2 Labor
To shed light on the mechanism behind the increased role of labor income in the top shares, we construct a skill premium series for the period from the 1950’s until recent years. We estimate the relative earning premium for skilled people with the difference of log average wages of intellectual and
31The capital income component including realized gains for the top 1% shares was 32.99% in Hungary, while 30.18%
in the USA in 2006. See: Table C.19 for Hungarian and World Top Income Database for USA data.
CEUeTDCollection
manual workers, (see description in Appendix Section C.7). Figure 3.11 presents the skill premium series, comparing it the labor income fraction in the top share the comovement is noticeable. Both series have a negative overall downward trend during the most of the planned economy with a jump in 1970. The skill premium series displays an upward trend from the early 1980’s that is followed by an increase in the top shares also. Both series continue to increase after the transition to the market economy.
Figure 3.11: Skilled Labor Supply and Skill Premium, 1955-2011
Note: Skilled labor supply is the percentage of the labor force with high school and university degrees, and skill premium is the ratio of log average white-collar worker wage over log average blue-collar worker wage. (See Appendix C.7 for details.) Source:Table C.22
During the planned economy a more compressed wage distribution was favoured by ideological concepts. But as part of the New Economic Mechanism the Central Planning Bureau gave the right to state-owned enterprises to exploit some margins of compensation to workers according to productivity, in effect redistributing a fraction of any potential surplus to the middle and top management. Though this wage reform was short-lived, we see a parallel increase both at the skill premium and at the top shares exactly after the introduction of the reform in 1968. The reform was reversed within few years, followed by decreasing skill premium and top income shares. From the early 1980’s wage and bonus settings of executives were delegated to enterprise level.32 This policy shift, which marks a first step to complete liberalization of the labor market, led to an increase in the skill premium (see Figure 3.11). After the transition, evidence of skill-biased technical change is prevalent, although interrupted by labor market regulations and foreign exchange crises during the mid-2000s. Our findings are in
3233/1983 Statute of Ministry 8§
CEUeTDCollection
line with Kézdi(2002), who based on micro data also documents a steady increase in skill premium for 1986-1995 as a consequence of inter-sectoral skill reallocation and dramatic jobs losses of the unskilled less educated people, and an even higher skill premium growth for the second half of the ’90s with skill biased technological change at most sectors.
The comovement between the skill premium and the top shares suggests that wage-setting decen-tralization favoring the remuneration of skills also play a role in the increase of the top income shares, beside other documented forces such as increased bargaining power of top managers, decreased top tax rates (Piketty et al. (2014)), and increased importance of capital income (Piketty (2013)).