The U.S. Department of Commerce regularly surveys businesses on their plans for capital investment. This article assesses the contribution that these surveys make to forecasts of business investment, once other economic variables are taken into account. The author finds that the surveys have only marginally improved forecasts since the 1970s. For short-term forecasts, the history of investment spending and output does more to reduce forecast errors than do the surveys. For forecasts of a year or more, the survey information is not as useful as that in the historical movements of various macroeconomic indicators.
The surveys do not cover all types of businesses or all industries, and the capital spending that respondents report does not necessarily match the concept of investment reported in the national accounts. Most significantly, the relationship between respondents’ capital spending and total investment has been changing since the 1970s. The survey was a more accurate indicator of capital purchases when the ratio of respondents’ capital spending to total investment was more stable.