Disparate Impact

Recent news reports suggest that 80% of the layoffs in the current recession have been men.

From labor force participation rates, it seems men are about 55% of the employed, so this is an enormous "disparate impact".

For those who think simple statistical disparities are good indicators of discrimination, disparate impact has been one of the standards by which some claim discrimination has occurred.

But does anyone think men are being "targeted" with layoffs because they are men, across the entire economy? Of course not.

Instead, job losses have tended to occur in fields which have a high proportion of its jobs filled by men (such as construction or finance), while job losses have been low in fields where a high proportion of jobs are filled by women (such as nursing or education).

This is a good lesson of why using disparate impact as a standard of discrimination is often wrong.