Conservatism and consensus-seeking among economic forecasters


  • Roy Batchelor,

    Corresponding author
    1. City University Business School, London, U.K.
    • City University Business School, Frobisher Crescent, Barbican, London ECZY 8HB, U.K
    Search for more papers by this author
    • Professor of Financial Economics at City University Business School, London. He is the author of books and journal articles on monetary economics, international finance, and the use of survey data in econometric modelling.

  • Pami Dua

    Corresponding author
    1. University of Connecticut, U.S.A.
    • Department of Economics, University of Connecticut, Scofieldtown Road, Stamford, Conn 06903, U.S.A
    Search for more papers by this author
    • Economics Department of the University of Connecticut. Her publications include journal articles on forecast accuracy and rationality, and on the effects of economic expectations on political attitudes.


This paper uses the track records of a panel of US economic forecasters participating in a consensus forecasting service to test for conservatism and consensus-seeking behaviour. The tests are based on a particular method-of-moments estimator, designed to allow for the heteroscedasticity and serial correlation which is inevitably present in errors from repeated forecasts for fixed target dates. Most forecasters prove to be conservative. When revising forecasts they give too much weight to their own past forecasts. Surprisingly, forecasters are not consensus-seeking but ‘variety-seeking’. When revising forecasts, they give too little weight to the known forecasts of other forecasters.