A fun(?) 🧵 on how nerdy government accounting rules had a big impact on Q4 GDP. And how they reflect how wasteful the 43-day government shutdown was. TL;DR: Small reduction in nominal federal spending in Q4. But a big decline in what we got for our dollars. So real down a lot.

Nominal federal spending (consumption + investment but not transfers) down $4b in Q4 (not annualized), a 3.5% decline (annual rate). A meaningful cut (14th percentile of growth since 1948) but not that dramatic. I'm not sure how much was shutdown vs. durable cuts vs. noise.

A lot of that was back pay for people who were not working and providing services (e.g., someone in a passport office that was furloughed). So BEA estimates that the actual amount of of *real* stuff the govt got was -16.6% annual rate. 1st percentile of real growth since 1948.

Nominal down a little, real down a lot, how to square them? BEA puts it in the price index, estimating that it effectively "cost" the government 15.6% more to buy goods and services in Q4. That price increase is the 99th percentile of experience since 1948.

In fact, the price differential between federal and overall GDP was the second biggest since 1948. The only bigger one was 1949-Q1 when there was an unusually large reset of compensation for federal military personnel.

Overall this subtracted 1.15 pp from GDP growth in Q4, so it would have been 2.5% absent this but instead was 1.4%. But it will artificially add about 1pp to growth in Q1 as the govt snaps back to something more normal in terms of real stuff per dollar.