Paul Krugman notes the disjuncture between what confidence is being expressed in the Trump economy as measured by surveys, and what the actual data indicate. Here are some additional observations on the disjuncture, based on yesterday’s consumption release.

First, monthly consumption has been flat since December, despite elevated consumer sentiment in the wake Trump’s election.

Figure 1: Consumption in billion Ch.2009$ SAAR (blue, left log scale), and Michigan consumer sentiment index (red, right scale). Dashed line at election. Orange shading denotes Trump administration. Source: BEA, and FRED.

Second, durables consumption is down.

Figure 2: Consumption in billion Ch.2009$ SAAR (blue, left log scale), and three month trailing moving average of the Baker, Bloom and Davis Economic Policy Uncertainty news index (red, right scale). Dashed line at election. Orange shading denotes Trump administration. Source: BEA, and policyuncertainty.com.

While durables consumption is a relatively small share of overall consumption (most is services and nondurables), movements of durables can be informative regarding expectations of the longer term future and uncertainty. (Services consumption should be approximately a random walk, on the other hand.)

The downward move in durables consumption could be explained by appeal to economic policy uncertainty, which is also elevated. Aggregate consumption, dominated by services and nondurables consumption, should not be too affected by policy uncertainty. But durables consumption should be noticeably affected — and that appears to be the case.

Or, sentiment is mismeasured, and consumption is depressed because the actual level of confidence is actually low.