Name:

Description:

Data on quantiles of the distributions of family incomes in the United States. This combines three data sources:

(1) US Census Table F-1 for the central quantiles

(2) Piketty and Saez for the 95th and higher quantiles

(3) Gross Domestic Product and implicit price deflators from Measuring Worth. (NOTE: The Measuring Worth Web site, https://MeasuringWorth.com, often gives security warnings. The desired data still seems to be available, however.)

Variables:

A data.frame containing:

Year

numeric year 1947:2012

Number.thousands

number of families in the US

quintile1, quintile2, median, quintile3, quintile4, p95

quintile1, quintile2, quintile3, quintile4, and p95 are the indicated quantiles of the distribution of family income from US Census Table F-1. The media is computed as the geometric mean of quintile2 and quintile3. This is accurate to the extent that the lognormal distribution adequately approximates the central 20 percent of the income distribution, which it should for most practical purposes.

P90, P95, P99, P99.5, P99.9, P99.99

The indicated quantiles of family income per Piketty and Saez

realGDP.M, GDP.Deflator, PopulationK, realGDPperCap

real GDP in millions, GDP implicit price deflators, US population in thousands, and real GDP per capita, according to MeasuringWorth.com. (NOTE: The web address for this, https://MeasuringWorth.com, seems to be functional but may not be maintained to current internet security standards. It is therefore given here as text rather than a hot link.)

P95IRSvsCensus

ratio of the estimates of the 95th percentile of distributions of family income from the Piketty and Saez analysis of data from the Internal Revenue Service (IRS) and from the US Census Bureau.

The IRS has ranged between 72 and 98 percent of the Census Bureau figures for the 95th percentile of the distribution, with this ratio averaging around 75 percent since the late 1980s. However, this systematic bias is modest relative to the differences between the different quantiles of interest in this combined dataset.

personsPerFamily

average number of persons per family using the number of families from US Census Table F-1 and the population from Measureing Worth. (Note: The web site for Measuring Worth, https://MeasuringWorth.com, often gives security warnings. It still seems to work. It seems that the web site is not maintained to current internet security standards.)

realGDPperFamily

personsPerFamily * realGDPperCap

mean.median

ratio of realGDPperFamily to the median. This is a measure of skewness and income inequality.

Details

For details on how this data.frame was created, see "F1.PikettySaez.R" in system.file('scripts', package='fda'). This provides links for files to download and R commands to read those files and convert them into an updated version of incomeInequality. This is a reasonable thing to do if it is more than 2 years since max(incomeInequality$year). All data are in constant 2012 dollars.

Link To Google Sheets:

Rows:

Columns:

Files:

License Type:

References/Notes/Attributions:

Author(s)

Spencer Graves

Source

United States Census Bureau, Table F-1. Income Limits for Each Fifth and Top 5 Percent of Families, All Races, https://www.census.gov/data/tables/time-series/demo/income-poverty/historical-income-inequality.html, accessed 2016-12-09.

Thomas Piketty and Emmanuel Saez (2003) "Income Inequality in the United States, 1913-1998", Quarterly Journal of Economics, 118(1) 1-39, https://eml.berkeley.edu/~saez, update accessed February 28, 2014.

Louis Johnston and Samuel H. Williamson (2011) "What Was the U.S. GDP Then?" MeasuringWorth. (Note: Their web address, https://www.measuringworth.org/usgdp, often gives security warnings. The desired data still seems to be available there. However, it seems that the site is not maintained to current internet security standards. The data used in the current USGDPpresidents data set was extracted February 28, 2014.)

R Dataset Upload:

Use the following R code to directly access this dataset in R.

R Coding Interface: