Hard Data vs Soft Data

Hard data are measured quantities — output, employment counts, prices actually charged, goods actually shipped. Soft data are survey answers about sentiment, intentions and perceived conditions: business confidence, purchasing managers' surveys, consumer sentiment. Both are official or long-established statistics; the difference is what they record. Soft data arrive first, because asking is faster than counting, and they are usually the earliest reading available for a period. They also measure something the hard series does not: what respondents say they expect or perceive, which need not translate into the activity counted later. The two can diverge for months, and neither is the corrected version of the other. Analysts describe that gap in its own terms rather than treating the survey as a failed forecast of the measured series.