The US government collects 13.85% (7.65% from the paycheck and 6.20% is from the employer) of the first $106,800 that employees earn.  Collecting this much tax on the first dollar that a person in poverty earns is cruel.  Compassion would dictate that we set a goal of exempting payroll taxes from everyone up to the level of poverty.

I have taken data from the US Census Bureau and plotted household income by year.  The format that I picked presents the data as a straight line, so that the full range of data supports each point, and the data can be extrapolated beyond the end points.

The horizontal x-axis is logarithmic household income.  Logarithmic was picked in part because on this axis, a 5% raise would be the same distance to the right at both the low and high income.

The vertical y-axis is percent normal probability of households that earn less than a given income.  Normal probability is the “bell curve” that you are familiar with.  This “bell curve” uses 50% at the middle, lower probability percentages going down the axis, and higher percentages going up.

Log-normal plots are often used for describing many group attributes.  Among the disparate groups represented by straight lines on the plot would be; particle size distribution, risk assessments, both hardware and living-things life-time estimates, and monetary donations.  I recommend this plot as a good visualization for making tax policy decisions.

The plot of Household Income has parameters describing the years of 1967, 1979, 1989, 1999, and 2009.  The lack-of-progress from 1999 to 2009 is difficult to see because 2009 is worse than 1999, but both line are displayed on the plot.

This plot shows that the top 30% of households are doing well and the bottom 30% are doing poorly.  This is the reason that I support giving everyone a household payroll tax exemption on the first $20,000 income earned.