A large divergence has recently opened up between survey-based measures of economic activity and “hard” data released by official statistical insitutions.

The hard information has been holding up reasonably well, examples last week including US retail sales and industrial output and Japanese and Eurozone GDP. As the first chart below shows, G7 industrial output growth has been following my soft landing scenario closely.

Surveys, by contrast, have deteriorated sharply over the last month, in some cases to an extent suggesting recession. For example, the earnings revisions ratio from the IBES survey of equity analysts plunged to a seven-year low in February – see second chart. Marked weakness was also on display in reports last week on US consumer sentiment, US small firm optimism, Japanese consumer confidence and German investor sentiment.

Under normal circumstances survey-based information provides a useful lead on hard economic data. However, sentiment weakness may have been exaggerated by negative credit market news and the Fed’s “emergency” rate cuts, which suggested the central bank expected a recession.

I am inclined to give greater weight to the hard data but will be concerned if the deterioration in sentiment is sustained and confirmed across a wider range of surveys. In the Eurozone, this week’s “flash” PMI readings will provide an important litmus test.

The extent of near-term global economic weakness is unclear but I still expect an improvement later in 2008 as recent monetary loosening offsets tighter credit conditions.