Measuring Celebrity

Celebrity is not fame, but it too can be measured. A version of this paper, somewhat mangled by the editor, appears in Annals of Improbable Research 12, 1 (Jan/Feb 2006), 14–15. Fame is being known. Celebrity is being known by your first name. Monica Lewinsky became famous when people learned of her presidential activities. She became a celebrity when people called her Monica. Schulman [1, 2, 3] has quantified fame. We quantify celebrity. Schulman [2] has shown that Google [4] is the best tool for measuring how well you are known. The quantity relevant to celebrity is the Google First Name Rank (GFNR), the position of the first reference to you when Googling your first name. For example, Bill Clinton has a raw GFNR of 10 because the tenth entry in a Google search for “bill” is a page about him. However, 6 of the preceding 9 entries are for non-people such as the Bill of Rights and Kill Bill. We omit such entries, giving Bill Clinton a corrected GFNR of 4. There are many more Bills than Norberts, so Bill Clinton’s GFNR of 4 indicates greater celebrity than that of Norbert Christlieb [5]. Your measure of celebrity should depend on both your GFNR and the number of people who share your first name. As Schulman observed, the visibility of our cultural stars should be measured on a logarithmic scale. Therefore, we define your Celebrity Index (CI) to equal log ( percentage of population with your first name your GFNR × constant ) where the constant depends on the unit of celebrity. Following Schulman, we define the Monica to be the unit in which Monica Lewinsky’s CI equals 0. The CI is country-specific. Jacques Chirac’s GFNR is 16 as measured by google.com, but it’s 3 on google.fr; and many more Frenchmen than Americans are called Jacques. Chirac’s CI is therefore much higher in France than in the United States. In this article, celebrity is measured only for the U.S. To measure the CI of Americans, we use data on the frequency of male and female first names from the Census Bureau [6]. Since about 49% of the population is male, the CI measured in deciMonicas (dM) is 10× log ( .49×M + .51× F GFNR × 70.86 ) where M and F are the percentages of the male and female population called by that first name. The values of M and F cannot be obtained directly from the census data, because people are not always called by the names on their birth certificates. Only .112% of Americans are named Bill, but a majority of the 2.4511% of Americans named William are called Bill. Comparing the