NCPA - National Center for Policy Analysis

Problems With Polling

September 16, 2002

Politicians use polls extensively to measure political support. However, polls have significant flaws. Observers explain that polls regularly fail to take into account basic economic considerations.

Most major polls simply ask whether a person supports expanding a program or not. However, the question does not take into consideration the added cost of the program, the unintended consequences of the program, or the success of the programs. In 1989, Gallup Polls found that an overwhelming majority supported a federal program to reduce classroom sizes to 15 students. However, a second poll found that if the cost of the program was factored in:

  • Some 31 percent of those polled refused to support the program.
  • Only 25 percent were willing to pay the full cost of the program -- a significant drop.

Nor did Gallup's questioning method consider the consequences of the program. According to the second poll:

  • When told that the federal program was only occasionally successful, support for the program dropped 35 percent.
  • When told that reducing classroom sizes could result in funding cuts for music and art, support dropped by 70 percent.
  • When told that only half of the new teachers would be qualified, only 15 percent still supported the program.

Additionally poll results varied considerably when the questions offered options. When given the choice, 83 percent preferred upgrading teacher training to reducing class sizes. A clear majority (54 percent) favored a tax-deferred education account.

Polls are often misleading, observers say, but most polling firms will continue to use simple polls and use statistical analysis to garner any extra understanding from the flawed data.

Source: Robert Weissberg, "The Problem With Polling," The Public Interest, Number 148, Summer 2002.


Browse more articles on Education Issues