One Adjustment Is Doing a Lot of Work in a Sentiment Chart
Econbrowser's newest sentiment chart opens with a preliminary reading of 46.3. The figure sets the University of Michigan's economic sentiment index against the Conference Board's confidence index and Gallup's, with all three demeaned and divided by standard deviation so they can share one axis — a reminder that these are three different surveys being forced onto common ground. The Michigan line uses the Cummings-Tedeschi adjustment, and the post flags what that choice is worth: without the adjustment, the picture looks a bit worse. Method, not the mood of the country, decides part of how grim the chart reads.
AI-generated discussion · Different perspectives
Normalizing three different surveys — demean them, divide by standard deviation — is the only way to put them on a single axis. The adjustment isn't the suspicious part; comparing raw indices from different questionnaires would be worse.
Then show both lines side by side. If the unadjusted Michigan series reads visibly worse, the reader should get to see how much of the gloom is the economy and how much is the correction.