Nominal GDP and the US midterms
Thank you for subscribing to Institutional Economics, a weekly newsletter about economics, finance and public policy by Dr Stephen Kirchner.
Please consider becoming a paid subscriber. Your paid subscription helps the newsletter’s free trade advocacy reach a wider audience and offsets the cost of the subscriptions and other inputs that go into its production.
A reminder that there is a standing 50% discount for those with email addresses affiliated with educational institutions (pro-tip: alumni email addresses seem to work with this feature).
Also a reminder you can find me on Bluesky @insteconomics.bsky.social.
For a Fed chair intent on offering no forward guidance, Kevin Warsh did a great job of conditioning expectations for the October FOMC meeting. The meeting delivered a steady policy outcome, as markets expected, and left expectations for future meetings largely unchanged.
His evasive posturing in the post-meeting press conference was once again indicative of a desire to hold the office, but not do the job. His performance was compared to that of a kindergarten teacher (John Authers), while the process leading up to the new meeting schedule announcement was ‘a masterclass in poor management.’
The proposal to substitute a strategy for a policy meeting once a year is part of the same diversionary exercise as the task forces that are meant to report into it. While Warsh likes to say that the Fed owns price stability, it is clear that he does not see current or expected inflation performance as mainly a monetary phenomenon requiring a monetary policy response.
The drift in US nominal stability was underscored by the Q3 national accounts release, which left nominal GDP running at 5.9% y/y. Core PCE inflation (ex-food and energy) is running at 3.1%.
That still leaves US Treasuries rich relative to nominal GDP growth, but cheap relative to the S&P 500 forward earnings yield. Mike Sandifer’s NGDP growth gap comes in at around +0.7%.
Nominal GDP and the midterms
Polymarket has a Democrat sweep at 68%, with a Republican Senate and Democrat House at 27% (regular reminder to Australian readers that the Polymarket domain is blocked by ACMA and retail OTC binary options are banned by ASIC for 10 years).
Ray Fair’s House vote share equation has the Democrats with a 50.6% vote share, with a standard deviation 2.39, implying a 60.4% probability that the Democrats get a vote share that exceeds 50%. The translation of that vote share into seats won is another matter.
Ray’s model is very old fashioned and overly deterministic, but is still a useful discipline on the more vibes-based takes on the role of the economy in the elections. I like it because the variables on the right-hand side of the equation are basically per capita nominal GDP growth relative to trend, making it the electoral analog of nominal GDP level targeting.
The other predictors are purely deterministic, at least with respect to any given election: the House incumbency effect and the Democrat House and Presidential vote share from the 2024 election yield a 51.65% share in conjunction with the deterministic constant.
The dynamics are given by the vote share elasticities on the GDP deflator and per capita real GDP growth relative to a 3.2% trend assumption over the first seven quarters of the Trump administration. Based on this week’s Q3 national accounts release, inflation adds 1.24 percentage points to the Democrat vote share, while per capita real GDP growth subtracts 2.26 percentage points...
Continue reading this post for free in the Substack app
© 2026 Stephen Kirchner
Sydney, Australia