The University of Michigan’s preliminary July reading came in at 54.4, up from a final June figure of 49.5 and well above the 51.0 consensus, the highest level since February and the second consecutive monthly gain following May’s 44.8, an all time low in a series that runs back to the 1940s. All five components improved, led by roughly 20% gains in buying conditions for durable goods and in expected business conditions over the year ahead. The improvement was broad based across age, income, wealth, and political affiliation, with the strongest gains among consumers without a bachelor’s degree. Falling gasoline prices did most of the work. And yet sentiment remains about 12% below where it stood a year ago, and the index is still sitting at a level that, historically, has only appeared during severe recessions.
That last fact is the puzzle. The Michigan index matters because it is the oldest continuous measure of household attitudes in the United States, because its inflation expectations series feeds directly into Federal Reserve thinking about whether price expectations are anchored, and because consumption is roughly 70% of GDP. If households are as despairing as 44.8 implies, spending should be collapsing. It is not. The gap between what the survey says and what consumers actually do has been the defining feature of this data series since 2021, and it has now persisted long enough that treating the index as a straightforward read on economic conditions is a mistake. The survey is measuring something real. It is simply not measuring what its level appears to claim.
Start with what the survey is structurally sensitive to. Michigan is built around household finances and prices rather than employment, which is why it diverges from the Conference Board’s confidence index, a measure weighted toward labor market perceptions. Prices are salient in a way that job security is not: consumers encounter gasoline and grocery prices weekly, and the survey’s open ended questions capture exactly that salience. For three straight months more than half of respondents spontaneously volunteered that high prices were eroding their personal finances. That is a statement about the price level, not about the inflation rate. A household that absorbed a cumulative 25% rise in the cost of essentials since 2020 does not feel relief when the annual rate falls to 3%; it feels the level, every week, indefinitely. The survey faithfully records that feeling and then reports it on a scale designed to measure cyclical conditions.
The methodological problems compound this. Michigan shifted from telephone to web based interviewing, a transition that coincided with a step down in the index and that the university’s own researchers have acknowledged affects comparability with the historical series. Sample sizes are small, roughly 500 respondents in the preliminary read, producing month to month noise that markets routinely over interpret; June alone was revised from a preliminary 48.9 to a final 49.5, which is worth remembering before anyone builds a thesis on July’s 54.4 ahead of the final release at month end. And partisanship has become a first order contaminant. Respondents’ assessments of the national economy now track their party’s control of the White House more tightly than they track any economic variable, which means the aggregate index is partly an opinion poll about politics wearing the costume of an economic indicator. Michigan itself has published work noting that national estimates now align most closely with the views of political independents, a tacit acknowledgment that the partisan wings are pulling in opposite directions and largely cancelling.
None of this makes the survey useless, and the Fed is right not to discard it. Inflation expectations are the mechanism through which sentiment becomes macroeconomically real: if households expect prices to keep climbing, they bargain for wages and accept price increases, and the expectation validates itself. Year ahead expectations were still running at 4.6% in June, far above the 3.4% reading in February before the Iran conflict began, while long run expectations had come down to 3.3%. That combination, near term expectations badly unanchored and long run expectations only modestly elevated, is the single most useful thing in the report and it deserves more attention than the headline. The right way to use Michigan is as a relative measure and a diagnostic of expectations, not as a level. Watch the direction of travel, watch the inflation series, and watch the gap between what consumers say and what they spend. When that gap finally closes, it will matter enormously which side does the moving.
Disclosure
This material is provided by Gryphon Financial Partners, LLC (“Gryphon”) for informational purposes only. It is not intended as a substitute for personalized investment advice or as a recommendation or solicitation of any particular security, strategy, or investment product. Facts presented have been obtained from sources believed to be reliable, though Gryphon cannot guarantee their accuracy or completeness. Gryphon does not provide tax, accounting, or legal advice. Individuals should seek such guidance from qualified professionals based on their specific circumstances.