"Income disparity" has become a hot issue in this election year. We have touched upon this topic in Comment 29 and Comment 29A with data up to the year of 2003. Here we have not only extended the data to 2010, but also have conducted a more detailed and accurate analysis to find the origin of this substantially widened income disparity compared to 1970s.
In the next section how to measure income disparity is discussed and then the actual income data for various income groups are presented; the data cover the time period from 1967 to 2010. Section 3 is devoted to the analysis of the data, and the last section is for some concluding discussions.
We need to determine first how to measure the growth of income disparity. Suppose there are two households, A and B. Household A has an annual income of $200,000, and B an income of $20,000. This means that Household A has 10 times more annual income than Household B. The numerical difference is $200,000 - $20,000 = $180,000. Suppose during the following 5 years the annual incomes of both A and B have grown by 10% respectively. Thus after 5 years, the annual income of Household A becomes $220,000 and the annual income of Household B becomes $22,000. The difference between the annual income of two households becomes $220,000 - $22,000 = $198,000; apparently the difference in annual income between two households has grown by $18,000. Does this mean that the income disparity between two households A and B has grown during the 5 year period? The answer is "no". We should note that after 5 years, the annual income of Household A is still 10 times of the annual income of Household B. Therefore, the income disparity between A and B during the 5 year period has not widened nor shrunk.
The above example shows that we must study the growth rate of annual income of different income groups in order to determine whether income disparity is widening or shrinking. How to measure the growth rate of the annual income of each group? A simple answer is to plot the annual income in logarithmic scale versus year in linear scale. In such a semi-logarithmic graph, the inclination of a straight line indicates that the data on the line is growing with a constant rate.
Census Bureau conducts a household survey every 10 years. The survey covers 100,000 households with various detailed questions, including the annual income of each surveyed household. Census Bureau then uses various information and estimates to come up with an income distribution for each year between the surveys. Currently the available data covers from 1967 to 2010. Census Bureau divides all the households to five groups, the bottom 0% to 20%, 20% to 40%, 40% to 60%, 60% to 80%, and 80% to 100%. On top of those 5 groups, Census Bureau throws in another group, that is, the top 5% (from 95% to 100%) of households. One may ask why not present the data of the top 1% since that is what recent political focus of the liberals. We should note that the household survey only covers 100,000 households. 1% of 100,000 means only 1,000 households. Census Bureau probably views 1,000 households as too small a statistical sample so the data is not presented. Some economists use tax return data to compile the income of the top 1% households, but a complete data set covering a long time span, like Census Bureau data from 1967 to 2010, is difficult to obtain, so we confine ourselves to the study of the data from Census Bureau only.
The red curve at the top of the graph at right is the CPI adjusted mean value of annual income of the top 5% group plotted in logarithmic scale. The blue curve right below the red one is the CPI adjusted mean value of the annual income of the group from 80% to 100% plotted in logarithmic scale, and so on until the bottom black curve that is the CPI adjusted mean value of the annual income of the bottom 0% to 20% group, also plotted in the logarithmic scale. The logarithmic scale for each group is all the same. Only the vertical positions of the curves are adjusted so that all six curves will fit into one graph paper. In such a semi-logarithmic graph, a straight line like the red line A means that data points falling on the line are growing with a constant rate determined by the slope of the straight line. Using the same logarithmic scale means that we can compare two straight lines and conclude which one is growing faster by just comparing the slopes of two straight line. For example, by comparing the red Line A with the blue Line A, we are able to say that the red Line A indicates a faster growth rate than the blue Line A since the slope of the red Line A is steeper than the blue Line A.
With the data about income distribution for each income group ready, we can now start to analyze the data and see how the income disparity has expanded in the past 40 plus years as carried out in the next section.
To analyze the data presented in the graph, we draw trend lines along each curve. A trend line is a straight line connecting two adjacent bottoms.
Trend lines A of various colors indicate the growth rates of the CPI adjusted mean value of the annual income of each income group from 1967 to 1974, that is, before the first energy crisis induced recession, respectively. We can see that the growth rate of the mean annual income of the top 5% group is larger than that of the 80% to 100% group, and that in turn is larger than the 60% to 80% group, and so on until the 20% to 40% group. However, the income growth rate of the bottom group, from 0% to 20%, beats even the top 5% group. That was due to "The Great Society" that has enhanced social welfare system for the very poor. The working poor in the 20% to 40% group was excluded from the improved social welfare system so the income gap between the working poor and the very poor had shrunk quite a bit during that era.
Various colored lines labeled as B are the trend lines for the mean annual income of various income groups during the energy crisis and high inflation period of mid 1970s to early 1980s. Since the listed income in the graph are adjusted for inflation via CPI index, high inflation rate naturally means stagnated income growth for every income group. Especially the lowest income group saw meaningful amount of income decrease because social welfare payments under "The Great Society" were not able to keep pace with inflation.
The most important group of lines in the graph are the lines labeled as "C". They are the trend lines of various income groups during the so called "globalization" era, from early part of 1980s to around 2007. We can see that it is during the "globalization" era the trend lines of higher income groups have substantially steeper slopes than the trend lines of lower income groups, indicating that the annual income of higher income groups grew much faster than the lower income groups. Especially the top 5%, represented by the red curve, had witnessed explosive growth of their income compared to the lower income groups.
Now let us concentrate our attention on the top 5% group, that is, the red curve. The red curve had experienced two bouts of especially strong and lasting growth. One is from 1982 to 1989, and the second is from 1992 to 2000. The first explosion was in Reagan era, and the second bout of income explosion for the top 5% group came under the watch of Clinton Administration. It is well known that a large part of the income of rich people comes from capital gains. The "junk bond bubble" of Reagan era, and the "dot-com bubble" of Clinton era were both stock market bubbles, so the top 5% enhanced their income enormously by the capital gains during those two bubbles. From 2004 to 2007 there was also a more timid and short lived rising phase of the red curve. This bout of income growth had come during the "mortgage and housing bubble" under the watch of Bush administration. However, it is much more difficult to trade houses in and out to produce capital gains compared to the case of stock market bubbles. That is why the income of the top 5% had failed to have an explosive growth rate during the mortgage and housing bubble.
The last group of lines labeled as "D" indicate how the income of various groups has suffered during "The Great Recession" of 2008 to 2009. We can see from the graph that the rate of the decline is most gentle for the income group from 80% to 100%. The rate of decline for the top 5% is more or less similar to the group from 60% to 80%. For the groups below 60% the rate of the decline became steeper as the income level drops, reflecting the severe unemployment environment.
The explicit analysis in this article shows that the real culprit of enormously widened income disparity is the "junk bond bubble" of Reagan era and the "dot-com bubble" of Clinton era. As has been pointed out in Seminar Notes, all three bubbles formed in the "globalization" era, that is, Reagan's "junk bond bubble", Clinton's "dot-com bubble", and Bush's "mortgage and housing bubble" are all the result of the runaway US trade deficit. However, Bush's "mortgage and housing bubble" is not a stock market bubble, so its contribution to the widened income disparity is much smaller than the two earlier bubbles. Therefore, the notion floated by some to claim that the widened income disparity is the result of Bush tax cut is not true.
The above mentioned seminar notes has also pointed out that the runaway US trade deficit in turn is an integrated design of so called "globalization" scheme intended to benefit politicians in power and enrich Wall Street speculators. In deed in the first two stock market bubbles Wall Street speculators leaped huge profits and contributed to the rapidly expanding income disparity. However, the mortgage and housing bubble has triggered the final collapse of the "globalization" itself. In the aftermath of the collapse, FED is creating unlimited amount of fiat money to prevent the economy to fall into another "Great Depression". The result of this money printing campaign is creating an environment of never ending zero interest rate and sluggish growth that we call "Japan Syndrome American Style" in article 17. How the income of various groups will fare in a "Japan Syndrome American Style" environment is an important issue to deserve our continued attention.