Statistics, discrimination, and the presidential election
A presidential election yields a variety of numbers that, like all statistics, can be manipulated to favor one’s own position.
The errors in some of the “mandate” claims coming from victorious Republicans are perceptible to anyone with 6th grade math skills and common sense critical thinking abilities.
To make my comments on this relevant here, I shall introduce them with an employment discrimination example illustrating similarly simple but potentially effective “lying” with statistics (it’s not really lying as long as the calculations are correct; perhaps it would be more accurate to say “misleading use of statistics”).
(It is with some trepidation that I venture into Michael’s specialty of employment statistics; I invite his comments and corrections — as well as anyone else’s)
Start with this simple puzzle. An employer is charged with disparate impact racial discrimination. Undisputed statistics show the employer has twice as many African-American workers as 2 years earlier. Yet it must concede that its hiring practices over the last 2 years have had a disparate impact against African-Americans. How can this be?
Easy:
Its workforce growth has been great enough to allow hiring of many African Americans, but many more whites.
For example, suppose the workforce went from 15 African Americans and 85 whites (total 100) to 30 African Americans and 270 whites (total 300), thus hiring only 15 African Americans out of 200 new hires (7.5%).
If 7.5% of new hires is sufficiently below the percentage of qualified applicants who were African American, there has been a disparate impact. Likewise, if 7.5% of new hires is sufficiently below the percentage of qualified individuals in the available work force, there has been a disparate impact. (Right Michael?)
Note that in this example while the employer correctly and proudly proclaimed a doubling of the number of African Americans to a record number, the simple sixth grade math of looking at percentages rather than raw numbers discloses that the change was unfavorable, not favorable, with the racial diversity of the workforce backsliding from 15%African American (15 out of 100) to 10% (30 out of 300).
Now lets talk about the election — ignoring the electoral college, which most people (including me) feel is inaccurate in terms of measuring the “pulse of the electorate” (though perhaps it has other merits).
CNN, shows the popular vote as Bush 51%, Kerry 48%, and apparently Nader 1%.
(In raw numbers, it’s Bush 59,459,765, Kerry 55,949,407)
Margin of victory = 3%
Questions:
1)is that a large margin?
2)is it a “mandate”?
For answers, I quote “Will the President Now Work For Greater National Unity?” (by Scott Lilly), published by the Center for American Progress:
As White House staff told reporters on Wednesday—and as was dutifully reported on network broadcasts Wednesday night—the president not only won the largest number of votes of any candidate for the presidency in history, but he was the first candidate in years to win a majority of the popular vote.
It is true that he won the largest number of votes of any candidate in history. But it is also true that the U.S. population is growing by nearly three million people a year. Population growth is the reason that 33 of the 45 presidents elected since popular vote totals were recorded could say the same thing.
[Add to that higher turnout, perhaps]
It is true that because of active third party campaigns, no other candidate since 1988 has received a majority of the votes cast. But it is also true that only three sitting presidents in American history who were re-elected had a smaller percentage of the popular vote. Further, . . . no president has ever been re-elected with a [percentage] margin that small.
So do they believe their spin, or do they just spin away knowing to what extent they are lying with statistics?
Seems to me the most relevant of the above statistics is that “only three sitting presidents in American history who were re-elected had a smaller percentage of the popular vote.” That discloses this election to to be a rather unusual “non-mandate,” creating quite little of the “political capital” of which Bush spoke the other day.
And now I retreat from partisan comments until the next election.
Sphere: Related ContentIf you enjoyed this post, please consider leaving a comment or subscribing to the feed or by email.
Comments
No comments yet.
Leave a comment