Justin Wolfers

Hey, folks at the White House and the Council of Economic Advisers! You guys ready to issue a correction yet?

This isn't controversial: It's arithmetic. Indeed, it's your arithmetic, but I fixed your errors.

The expected duration of this shutdown keeps rising.

Hey, that's my Econ 101 students! Talking econ! In the public square! With the great @svaneksmith.bsky.social!

Love you guys.

Think of the economy as like baking a cake. H‑1B visas supply missing ingredients -- high skill STEM workers. If Trump cuts off access: fewer projects get baked, worse recipes get used, or we import the cake. None of those make Americans better off. #TheProfessorIsIn #TeachEcon

The numbers say H-1Bs raise wages, boost patents, and drive productivity.

The story is that H-1Bs aren’t just “workers”—they’re engines of innovation. They're the strivers who crossed the ocean, and brought their fresh new ideas with them.

Economist Justin Wolfers joins Molly Jong-Fast to explore the implications of the ongoing U.S. government shutdown, particularly focusing on its effects on public data dissemination and the broader… Is the Trump Hiding Job Numbers? w/ Justin Wolfers

Brunch and pod with my pal, @mollyjongfast.bsky.social and me. We dig into the shutdown, the missing jobs report, tariffs, the reality that Trump's CEA can't do arithmetic (or apologize) and all of this econ-madness that's all around us.

Research shows:
More H-1Bs → more patents
Fewer H-1Bs → more offshoring

Your choice, America.

If H‑1Bs “took jobs,” you’d expect lower wages for U.S. grads. The best available studies show the opposite: 1990–2010 H‑1B inflows raised wages of American college grads by 4.2% and non‑college by ~2%, with native employment unchanged.

Think complements, not substitutes.

"Only the portrait of a deceased individual may appear on United States currency and securities." 31 USC 5114(b)

A shutdown is Washington burning your money to send a message. Work won't get done but workers will get paid. Same cost, less benefit.

Hey, folks at the White House and the Council of Economic Advisers! You guys ready to issue a correction yet?

This isn't controversial: It's arithmetic. Indeed, it's your arithmetic, but I fixed your errors.

Turns out the check wasn’t presidential restraint; it was Senate reality: even they won’t confirm "a partisan hack, a clown, someone who's relentlessly unqualified" to run the BLS.

That’s a relief—and a reminder that institutions still push back when norms are tested.

Boom!

September jobs numbers: Payrolls rose/fell __k. Unemployment up/down a tick to 4-point-something percent.

Economic data are noisy, yet they still confirm my priors.

Markets are responding exactly as I thought they would.

Another tough moment for the Fed. #JobsReport

Together we could find the trend,
If only Congress would let it send.
Oh someday girl, I don’t know when,
We’ll get to that Friday mornin’,
See the jobs come rollin’ in again

’Cause baby, we were born…
Born to run… with no report.
Born to guess if the jobs hold on.
With the BLS doors locked shut,
We’re drivin’ this economy in the dark.

Shutdown’s got us ridin’ blind,
No labor stats for us to find.
Baby this town rips the data apart,
It’s a death trap, it’s a stats gap,
We need a chart.

🎵sung to Born to Run🎵

In the day we sweat it out on Wall Street’s run,
Lookin’ for the numbers that never come.
Out on the wires the markets scream,
But the jobs report’s just a broken dream.

🧵

Macro vs micro: A short shutdown barely nudges GDP. But it’s still an infuriating waste—turning off services we’ve already paid for. The macro hit is a drop in the bucket; the micro pain is real for those caught up in it.

You're correct. For small growth rates, division ends up being pretty close (as your example demonstrates).

At a more personal level, lemme say...

"To the public servants who really work so hard and wake up with a sense of mission and purpose, I just want to say: I see you. And I'm grateful for you."

But I also am implacably opposed to pseudo-economists fumbling their arithmetic, misrepresenting serious research, and lying to the public.

And none of this is to say I like having a government so dysfunctional that it randomly shuts down. I don't. I think it's a horrible and pointless waste. $3.75 billion in waste is still $3.75 billion too much.

Put it all together, and the Fed and Goldman estimates suggest that each week of the shutdown will reduce 2025's annual GDP by about one-hundredth of one percentage point, and it will have no effect on next year's GDP.

(I'm not endorsing the estimate, just explaining it.)

CEA also cites an estimate from the Fed from a decade ago. Same problem.

They cite the Fed estimating a 0.2%-pt impact on this quarter's annualized growth, and ignore the rest of that sentence "AND INCREASE THE RATE OF GROWTH BY THE SAME AMOUNT IN THE FIRST QUARTER AS ACTIVITY RETURNED TO BASELINE"

Second, honesty.

CEA cites Goldman's estimate that the shutdown will subtract "about 0.2 percentage point per week once private sector effects were included."

They ignore the rest of that sentence: "and growth would rise by the same cumulative amount in the quarter following reopening"

Point is, CEA overstated the "cost" of a government shutdown by a factor of FOUR.

The whiz kids at CEA forgot the difference between an annualized rate of growth (what would happen to GDP if the growth shock were to persist all year?), and quarterly growth (what actually happened this quarter?)

First, arithmetic.

A weeklong shutdown will reduce *annualized* growth in Q4 by 0.2 %-points. So that's 0.05 %-points of actual quarterly growth. And GDP in a quarter is $30 trillion/4 = $7.5T.

So the "cost" is 0.05% x $7.5 trillion = $3.75b

NOT the $15 billion CEA claim

Trump's CEA is circulating its estimate of the economic cost of the government shutdown.

But, BREAKING: It's wrong. It makes a simple arithmetic error. And it's dishonest.

Surprising, I know. 🧵

Lotta folks evaluating the ADP jobs data are showing a version of my top chart, and the revised ADP numbers looks pretty reliable.

But for judging reliability of the latest ADP number, you want real-time accuracy. My lower chart shows that on this score, ADP is a lot less useful.

It is also unfortunate that White House leadership has decided to use American consumers as a hostage or pawn in the trade negotiations. American consumers also voted for President Trump (not sure why that's relevant).
Why aren't we expecting news on Tuesday of substantial support for our consumers?