Wednesday, December 28, 2011

New York Times and Washington Post Articles on Congress and Wealth: Simplistic Statistical Analysis

Yesterday, the New York Times and the Washington Post published articles reporting a dramatic difference in wealth among members of Congress and the general public. Both articles show a widening wealth gap. While wealth among the general public has declined in the last two decades, it has risen for members of Congress (see Dissenting Justice analysis).

The articles both question whether wealthy individuals can properly lead the nation. The subtext, which is not so subtle in the New York Times article, implies that simply being a member of Congress creates wealth. And many readers have concluded that corruption, inside deals, insider trading, and other improper conduct explain the wealth among members of Congress.

Although these articles raise important issues, they fail because they do not analyze important issues that could explain the wealth disparities. Instead, the articles only isolate one factor for examination: the wealth of persons inside and outside of Congress. These populations, however, have many characteristics -- especially the general public.  The articles, however, do not control for any other factors that could account for a wealth disparity.

For example, a good statistical study of this issue would control for race, gender, age, educational achievement, investment behavior, and other factors that correlate with wealth (see previous article on Dissenting Justice). It is unclear whether members of Congress have greater wealth than persons with similarly situated individuals outside of Congress. Because of this failing, the articles leave the impression that membership in Congress alone explains the wealth gap.  This problem could erode the public's trust in government, which is already suffering according to opinion polls.

The Washington Post and New York Times should update these articles with more sophisticated statistical analysis. The articles both raise important questions. But, they fail to establish a solid factual foundation for engaging in an informed -- or useful -- dialogue.

Note: An updated version of my earlier discussion of this issue is now on the Huffington Post.

No comments:

Real Time Analytics