Income of richest households vs. tax on high incomes

February 15, 2009

During the last 90 years the U.S. has experienced large changes in the share of the total national income controlled by the richest households. The (pre-tax) share of the richest 0.01% of households, for example, was under 1% in the late 1970’s, but over 5% in the late 1920’s and in the years 2005 and 2006.

Over the same period, income tax rates on high incomes varied greatly. The top income rate fluctuated between over 90% (1950’s and early 1960’s) and under 40% (1920’s and 1987 to the present day).

As is evident in the chart below, those two phenomena are strongly negatively correlated over time. The correlation factor is about -74%. Of course, correlation does not imply causation, but even if high tax rates on high incomes do not directly translate to smaller (pre-tax) income for the rich, the correlation supports the notion that inequality in income is determined to a large extent by public policy rather than by an “impersonal phenomenon, the market price”.

Data: Income data are from Thomas Piketty and Emanuel Saez, “Income Inequality in the United States, 1913-1998”; updated version and updated data files available at Saez’s website. Tax rate data are from “Personal Exemptions and Individual Income Tax Rates, 1913-2002”, an IRS publication.

Advertisement

4 Responses to “Income of richest households vs. tax on high incomes”


  1. It took me a while to tease out the data from the chart, but I get it.

    Why the line chart, as opposed to a scatter chart?

    I would think that it would be clearer that way.

  2. yoramgat Says:

    Hi Matthew,

    This is a scatter chart, but with a line added, indicating the temporal order of the observations.

    A scatter chart with points only is appropriate when dealing with a set of independent samples, but not when dealing with a time series. When dealing with a time series, ignoring the temporal dimension may be misleading or may obscure an important part of the phenomenon.


  3. OK, it’s just that in order for me to “grock” the graph, I had to mentally remove the temporal order line, and then it was very clear.

  4. yoramgat Says:

    Hmmm…, maybe the line needs to be dotted so it doesn’t dominate the picture. I’ll do that. Thanks.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s