According to an article appearing in the April 16 issue of the Economist, Hal Varian (econ grad students, you know his textbook) and Google colleague Hyunyoung Choi have discovered surprisingly strong statistical correlations between certain Google searches and the movement of related economic variables.
For example, time-series search data from the Google Trends database for Ford vehicles is highly correlated with Ford's light-vehicle sales in the United States (see graphic). You can read Choi and Varian's full paper here.
Google Trends is free for now, friends. So if you are looking for potential proxies for variables you cannot precisely measure--or ones you simply cannot afford--think about using Google Trends data.
Just be sure to cite Choi and Varian.