What do Sentiment Surveys Measure?

Indices of business and consumer sentiment receive widespread media coverage and are closely watched by market economists despite their limited success as leading indicators. In this paper we ask what explains ‘sentiment’ and find that lagged economic indicators (such as changes in GDP, job vacancies and the cash rate) can explain a substantial proportion of the variation in a number of backward and forward-looking sentiment indices. This does not rule out the possibility that they may be useful for forecasting. We find, however, that when currently available economic information is appropriately ‘filtered’ from the sentiment indices, in most cases they fail even rudimentary Granger-causality tests of predictive ability. On a more positive note, we find that the Roy Morgan consumer confidence rating, NAB actual business conditions, NAB expected employment outlook over the next three months and the second question in the Roy Morgan and Westpac/MI consumer surveys all provide some, albeit small, contribution to forecasting employment growth. The second question of both consumer confidence surveys (which asks about anticipated personal financial conditions over the coming year) also appears to have some ability to predict recessions. Outside of these results there is little evidence that the surveys tell us anything we didn’t already know. Thus, there is reason to suspect that surveyed respondents’ forecasts offer little more information about the future path of the economy than a weighted average of lagged economic variables.