Home » Open thread 3/14/2025

Comments

Open thread 3/14/2025 — 32 Comments

  1. Unwearied still, lover by lover,
    They paddle in the cold
    Companionable streams or climb the air;
    Their hearts have not grown old;
    Passion or conquest, wander where they will,
    Attend upon them still.

    But now they drift on the still water,
    Mysterious, beautiful;
    Among what rushes will they build,
    By what lake’s edge or pool
    Delight men’s eyes when I awake some day
    To find they have flown away?

    W.B. Yeats, “The Wild Swans at Coole”

  2. Does anyone know if Betsy DeVos was offered anything in the Trump 2.0 administration? I kind of forgot about her, and in today’s WSJ she has a good old fashioned Letter to the Editor. And it’s great.

    https://archive.md/cTB3j

  3. I guess the Chuck Schumer agreed to pass the CR, performatively whinging on MSMBC about “making the middle class pay for tax cuts for billionaires” or whatever nonsense. AOC, of course, is revolting.

  4. Orff, Carmina Burana — Olim Lacus Colueram (3:22)
    https://youtu.be/5wJzp4QWDOQ

    12. — OLIM LACUS COLUERAM

    Cignus ustus cantat:
    Olim lacus colueram,
    olim pulcher extiteram,
    dum cignus ego fueram.
    Miser, miser!
    modo niger
    et ustus fortiter!
    Girat, regirat garcifer;
    me rogus urit fortiter:
    propinat me nunc dapifer,
    Miser, miser! etc.
    Nunc in scutella iaceo,
    et volitare nequeo,
    dentes frendentes video:
    Miser, miser! etc.

    12. — ONCE I LIVED ON LAKES

    The roasted swan sings:
    Once I lived on lakes,
    once I looked beautiful
    when I was a swan.
    Misery me!
    Now black
    and roasting fiercely!
    The servant is turning me on the spit;
    I am burning fiercely on the pyre;
    the steward now serves me up.
    Misery me! etc.
    Now I lie on a plate,
    and cannot fly anymore,
    I see bared teeth:
    Misery me! etc.

  5. Interesting piece in the Federalist today.

    America Becoming Less Christian Is A Problem Foe Everyone

    I’m almost hypersensitive to essays like this one. One of my sharpest friends, one of the few I ever talk politics with, clearly gets a little impatient with me when I talk policy – tarriffs, taxes, school choice, etc. Eventually he says something like, yes, yes, I agree, but none of that really matters. THIS is the only thing that matters.

    https://thefederalist.com/2025/03/14/america-becoming-less-christian-is-a-problem-for-everyone/

    I have a suspicion he might be right.

  6. Bauxite, this hypothetical scenario imagined by Grok demonstrates how tariffs to encourage onshoring of industry will be a net positive.

    I asked Grok: “give me a scenario where a tariff resulted in more companies moving to the United States and causing more growth to the US economy.”

    Let’s construct a plausible scenario where a tariff led to more companies moving to the United States, spurring economic growth, grounded in real-world dynamics and historical patterns. I’ll draw inspiration from past tariff policies and their outcomes, like the Trump-era trade actions, while keeping it hypothetical yet realistic.
    Imagine it’s 2026, and the U.S. government imposes a 35% tariff on all imported electronics—smartphones, laptops, semiconductors—aimed primarily at East Asian manufacturers in China, South Korea, and Taiwan. The stated goal is to reduce reliance on foreign supply chains and boost domestic production, echoing national security and economic resilience rhetoric. Initially, consumer prices spike—iPhones jump from $1,000 to $1,250, and Samsung laptops see similar hikes. Retailers grumble, and importers scramble.
    Foreign companies, facing a steep cost disadvantage, start rethinking their strategies. Take a fictional South Korean firm, “TechNova,” a major player in budget smartphones. Exporting to the U.S., its biggest market, now slashes its profit margins by 30% due to the tariff. Rather than lose market share to pricier imports, TechNova decides to build a $2 billion assembly plant in Texas. The state offers tax breaks and a skilled workforce from nearby universities, sweetened by federal incentives tied to the tariff policy. Construction begins in mid-2027, employing 1,500 workers directly and another 2,000 in related industries—steel, concrete, logistics—adding $300 million to the local economy in the first year alone.
    Meanwhile, a Taiwanese semiconductor giant, “ChipMaster,” follows suit. With the U.S. consuming 40% of its chips, the tariff makes exporting from Taipei uneconomical. ChipMaster partners with an American firm and invests $5 billion in a fabrication plant in Arizona, creating 3,000 high-tech jobs and sparking a mini-boom in the region—housing starts rise 15%, and local GDP grows 2.5% annually by 2028. The ripple effect hits suppliers: a Missouri-based glass manufacturer ramps up production for TechNova’s screens, hiring 400 more workers, while a California battery maker expands to meet ChipMaster’s demand, adding $150 million in revenue.
    Domestic firms also capitalize. A U.S.-based startup, “AmeriTech,” previously struggling against cheap imports, now scales up its Ohio factory to produce mid-range laptops. The tariff levels the playing field, and AmeriTech’s output doubles by 2029, boosting national electronics production by 10% and contributing $1.8 billion to GDP. Retaliation from affected countries—like a Chinese 20% tariff on U.S. agricultural exports—hurts some sectors (soybean farmers lose $500 million), but the administration offsets this with subsidies, keeping the net economic impact positive.
    By 2030, the U.S. sees a 12% increase in manufacturing jobs—about 1.5 million new positions—concentrated in tech-heavy states. GDP growth ticks up 0.8% annually from 2027-2030, driven by $20 billion in new investment from relocating firms. Inflation stabilizes as domestic supply chains mature, and the trade deficit shrinks by 5% as imports drop and exports of U.S.-made goods rise. Critics point to higher consumer costs—electronics remain 8% pricier than pre-tariff levels—but proponents argue the long-term gains in jobs and industrial capacity outweigh the hit, with real wages rising 3% due to demand for skilled labor.
    This scenario mirrors dynamics seen with Trump’s 2018 steel tariffs, where some foreign steelmakers expanded U.S. operations (e.g., Nucor’s growth), and aligns with posts on X claiming tariffs reshored auto jobs in Georgia. It’s a simplified model—real life would involve messier variables like geopolitical blowback or supply chain lags—but it illustrates how a tariff could plausibly draw companies stateside, fueling economic growth through investment and employment. Want me to tweak any specifics?

    Then I asked: “would this scenario result in higher tax revenues to the US government”

    Grok’s response:

    Yes, the scenario I outlined—where a 35% tariff on imported electronics drives companies like TechNova and ChipMaster to relocate to the U.S., spurring job creation and economic growth—would likely result in higher tax revenues for the U.S. government. Let’s break down how this shakes out across different tax streams, using reasonable estimates based on the hypothetical’s scale and real-world tax mechanics.
    First, corporate tax revenue would climb. TechNova’s $2 billion Texas plant and ChipMaster’s $5 billion Arizona fab, once operational by 2028-2029, start generating U.S.-based profits. Assume TechNova earns $300 million annually in net profit from its U.S. sales, and ChipMaster pulls in $800 million from its chip production. At the current federal corporate tax rate of 21%, that’s $63 million from TechNova and $168 million from ChipMaster yearly—$231 million combined. Add in smaller players like AmeriTech, scaling up to $100 million in profit, contributing $21 million, and you’re looking at roughly $250 million in new annual corporate tax from these relocations. Pre-tariff, most of those profits were taxed abroad or offset by import-related deductions, so this is largely net new revenue.
    Next, personal income tax gets a boost from the jobs created. The scenario pegs 1.5 million new manufacturing jobs by 2030, with an average salary of $60,000—realistic for a mix of skilled tech workers and support roles, per Bureau of Labor Statistics data. That’s $90 billion in total wages. Applying an effective federal income tax rate of 15% (accounting for brackets and deductions), plus payroll taxes (7.65% employee-side FICA), you get about $20.25 billion annually from workers—$13.5 billion in income tax and $6.75 billion in Social Security/Medicare. State income taxes (e.g., Texas has none, but Arizona’s 2.5% flat rate adds $150 million) layer on extra, though variably. Before the tariff, many of these jobs were offshore, so this is a direct gain.
    Tariff revenue itself also kicks in early. In 2026, before companies fully relocate, assume $50 billion in electronics imports face the 35% tariff—down from a higher baseline as firms adjust. That’s $17.5 billion in duties collected that year. As production shifts stateside by 2029, imports drop (say, to $20 billion), cutting tariff revenue to $7 billion annually. Over the first five years, this averages $10-12 billion yearly, peaking early then tapering—a temporary windfall that funds infrastructure incentives to lure more firms.
    Sales and property taxes grow too. The $20 billion in new investment (plants, factories) boosts local property tax bases—ChipMaster’s $5 billion fab might add $50 million annually in Arizona property taxes at a 1% effective rate. Consumer spending from 1.5 million new workers (assume 60% of wages, or $54 billion) generates sales tax— at an average U.S. rate of 7%, that’s $3.8 billion split among states. Federal coffers don’t see this directly, but it supports state budgets, reducing pressure on federal grants.
    Summing it up: By 2030, when the shift stabilizes, the federal government could see roughly $34-36 billion in new annual revenue—$250 million (corporate), $20.25 billion (personal income/payroll), and $7 billion (residual tariffs). Early years might hit $40-45 billion with higher tariff hauls. Offsetting this, subsidies to farmers ($500 million annually for trade retaliation) and tax breaks for relocating firms (say, $1-2 billion) trim the net gain to $32-34 billion yearly. Compared to pre-tariff levels, where these profits and wages were largely offshore, this is a clear uptick.
    Caveats: Higher consumer prices (electronics up 8%) might dampen spending elsewhere, shaving taxable activity by $1-2 billion. Inflation or wage pressure could also nudge tax brackets, but the net effect stays positive. Historical precedent—like the 2018 tariffs adding $20 billion in duties—supports this scale, though relocation lagged then. So, yes, higher tax revenues are probable—likely $30-35 billion annually by 2030, with a bigger bump upfront. Want me to adjust any assumptions?

    I followed that up with the potential of retaliatory tariffs. “add that countries impose punitive tariffs on US goods”

    Grok’s response:

    Let’s revise the scenario to include punitive tariffs from other countries on U.S. goods in response to the 35% U.S. tariff on imported electronics, and see how that impacts the outcome—specifically whether it still results in higher tax revenues for the U.S. government. Here’s the updated version.
    It’s 2026, and the U.S. slaps a 35% tariff on electronics imports—smartphones, laptops, semiconductors—targeting East Asian manufacturers. TechNova (South Korea) builds a $2 billion plant in Texas, ChipMaster (Taiwan) invests $5 billion in Arizona, and AmeriTech scales up in Ohio, driving 1.5 million new jobs by 2030, $20 billion in investment, and 0.8% annual GDP growth from 2027-2030. But now, China, South Korea, and the EU retaliate with punitive tariffs on U.S. exports. China imposes a 40% tariff on U.S. agricultural goods (soybeans, pork), South Korea hits American autos with 25%, and the EU adds 30% to U.S. tech and machinery exports. These start in 2027, escalating trade tensions.
    The retaliation bites. U.S. soybean exports to China, worth $12 billion annually pre-tariff, drop 50% as buyers shift to Brazil—$6 billion in lost sales. Auto exports to South Korea ($2 billion) shrink by 40%, costing $800 million. EU tariffs cut U.S. tech exports (e.g., servers, software hardware) by 30%, trimming $5 billion from a $15 billion base. Total export losses hit $11.8 billion yearly. Affected sectors—farmers, automakers—shed 50,000 jobs, and rural GDP dips 1% in states like Iowa and Michigan.
    Despite this, relocation persists. TechNova and ChipMaster still move, as the U.S. market’s size (40% of their sales) outweighs export losses elsewhere. Corporate tax revenue rises: $231 million from the big two, plus $21 million from AmeriTech, totaling $252 million annually by 2029. The 1.5 million new jobs—$90 billion in wages—generate $20.25 billion in federal income and payroll taxes ($13.5 billion income, $6.75 billion FICA). Initial tariff revenue peaks at $17.5 billion in 2026 (35% on $50 billion imports), then falls to $7 billion by 2030 as imports drop to $20 billion—a five-year average of $11 billion.
    Punitive tariffs cut into other revenues. Lost exports reduce corporate profits for firms like John Deere or Ford—assume a $2 billion profit hit, slashing corporate tax by $420 million yearly. The 50,000 lost jobs ($3 billion in wages) trim income/payroll taxes by $675 million. To cushion this, the U.S. government boosts farm subsidies to $2 billion annually (up from $500 million) and offers $2 billion in tax credits to relocating firms—total mitigation cost: $4 billion yearly.
    By 2030, federal revenue gains are: $252 million (corporate from new plants), $20.25 billion (wages), $7 billion (tariffs)—$27.5 billion total. Losses and costs: $420 million (export profit hit), $675 million (lost wages tax), $4 billion (subsidies/credits)—$5.1 billion. Net gain: $22.4 billion annually. Early years see $30-35 billion with higher tariff hauls, but retaliation tempers the peak. Without punitive tariffs, the net was $32-34 billion; now it’s $10-12 billion lower due to export damage and mitigation.
    Higher tax revenues still happen—$22-25 billion yearly by 2030—driven by domestic job and investment growth outpacing export losses. The U.S. economy grows, but slower (GDP up 0.5% vs. 0.8% annually), and rural areas lag. Retaliation dents the windfall, but the tariff’s reshoring effect holds. Want me to tweak the retaliation scale or mitigation costs?

    Now add increase this strategy of reshoring to multiple industries that we consider essential to our economic/military security.

    This is where Trump’s reciprocal tariff strategy comes in. Industries we are strong in just need other countries to not block our exports. Industries that we consider essential that have left need to have a tariff strategy that encourages their return.

    As to potential increased cost to domestic products, I think Trump has proposed eliminating income taxes below $150,000, which, according to Grok, pays about $415 billion in income taxes. If that threshold were reduced to incomes below $100,000 the taxes paid is somewhere between $230-250 billion, which might be the amount of tariffs collected each year.

  7. This article from the world of Astronomy/Cosmology got my attention.

    Many here are likely aware that the Hubble Space Telescope, and now the James Webb Space Telescope, have found galaxies that are surprisingly mature in the very early universe. Mature here has always meant size and shape. Now comes this, and it is really mind-boggling.

    https://scitechdaily.com/this-ancient-galaxys-unexpected-chemistry-could-rewrite-the-story-of-the-universe/

    Here is an excellent quote from the article that sums up the situation very well.

    Another instrument on the telescope – the Mid-Infrared Instrument, or MIRI – revealed something extraordinary: significant amounts of oxygen.

    In astronomy, anything heavier than helium is considered a “metal,” Helton said. Such metals require generations of stars to produce. The early universe contained only hydrogen, helium and trace amounts of lithium. But the discovery of substantial oxygen in the JADES-GS-z14-0 galaxy suggests the galaxy had been forming stars for potentially 100 million years before it was observed.

    This is the first time, that I know of, that the chemical composition of one of these very early galaxies has been observed, and it is not at all what was expected.

  8. ”…this hypothetical scenario imagined by Grok…”

    Good grief, Brian E! If I wanted to know the imaginations of Grok, I’d ask it. You don’t have to cut and paste three pages of text here.

  9. Brian E – Grok is a large language model. It literally finds patterns of words that match the pattern of words in your prompt. It is farcical to expect it to be authoritative on matters of economics. That’s just not what it does.

  10. Mike Plaiss,

    I continue to be amazed at how the Webb is upsetting the cherished theories of the Big Bang. Nothing can be better for science than a whole slew of data showing an established theory may be wrong. Exciting times.

  11. Yes, it answers the question I framed. But the principals are there. You can add a counter hypothetical (and I could have asked that), but to dismiss it by handwaving doesn’t refute the numbers.

    This isn’t complicated math, it’s arithmetic.

    All of a sudden everyone is a luddite.

  12. Mike Plais,
    I had a good preacher say one time something along the lines of some people have their head in the clouds so much they are useless on Earth. He said something along the lines of keeping one eye on Heaven and one on Earth.
    You think about the transgender push on kids. That is both a political and a spiritual thing. It has implications across the board. And the people that are voted into school boards and state legislatures, etc can effect what is pushed on kids in the classroom.

  13. Mike,
    Jesus talked about being ” salt and light”. It seems to me “salt” is the influence on and preservation of the culture around us and ” light” is the gospel of salvation.
    I could be wrong, but that is how I think of ” salt and light” .

  14. Speaking of patterns, it is inevitable that anything improper or untoward on the part of any Jew enters societal conversation about Jews generally. Rogan is no exception:

    Joe Rogan and the Jews

    Unfair, of course, but life isn’t fair.

  15. Brian E:

    What is the added value?

    “Look what AI (Grok) can do with my insightful questions?”

    Luddites? No.

  16. om, “what is the added value?”

    Of tariffs, or using Grok to create a hypothetical fleshing out the actual benefit to bringing back industries to the US?

    As to using Grok, there was a time when I would have put numbers to a hypothetical myself, but I don’t have the energy right now.

    I’ve always preferred to talk about issues using data– while hypotheticals make numerous assumptions, they are still instructive as to the general principals.

    If Grok is so uninformed/unreliable it should be a simple task to show where the hypothetical doesn’t represent the real world.

    Bauxite had claimed that targeted tariffs reshoring manufacturing/industries doesn’t grow the economy.

  17. om,
    synopsis: even with the expense of retaliatory tariffs, bringing manufacturing back to the US is a net positive to the economy and the government.

  18. Brian E – You’re asking a language-based model to do math. There’s not much more to say.

    If you have particular points you want to make, I’d be happy to consider them, but I’m not going to wade through a word salad made by an LLM.

  19. I asked Grok if it was good at math.

    My Capabilities
    As Grok 3, built by xAI, I’m designed to be helpful and truth-seeking, so I can tackle a decent range of math problems by reasoning through them. In your earlier question about income tax, I used 2025 tax brackets and estimated totals based on IRS data—showing I can handle applied math with context. If you give me a problem like “What’s 15% of 240?” I’d say: 0.15 × 240 = 36, and I’d be right. But if you asked for the 10th digit of ? or to factor a 7th-degree polynomial, I’d struggle without assistance.
    Are We “Good”?
    For everyday math—taxes, percentages, basic algebra—yes, LLMs like me are pretty good, especially with clear prompts. For advanced math or precision-critical tasks, we’re less reliable than specialized tools. Our strength lies in explaining and contextualizing, not raw computation. Want to test me? Throw a math problem my way, and I’ll show you what I can do!

    I think it could handle the math in this hypothetical.

  20. Economists are almost all on board with Free Trade as better than tariffs.
    Most also agree that printing money/ too much govt debt is bad and, because of hyperinflation, like Zimbabwe or Weimar, hugely destructive.

    None seem quite able to explain how Japan can, without inflation, have a debt/gdp ratio of over 250%. Economist predictions don’t have nearly as good a record as physicists, who we see either Hubble news might also have theories which make bad predictions.

    I think a trade war would be bad, but is preferable to a shooting war. Using tariffs to get other countries to drop tariffs seems fine, but moving to a higher tariff/ lower income tax world is pretty uncertain.

    Thanks, Brian E, for asking Grok for a plausible pro-tariff scenario. I’d guesstimate its likelihood, or better, at only about 10%, calling it a very positive result, with low positive being about 40%, low negative being 45%, and very negative being only 5%. 50/50 on plus minus, but a bit more high plus upside.

    Nobody knows.
    All who claim to be sure should have their prior predictions closely looked at—did they predict the 2008 financial crash, as Dr. M. Burry did (becoming a billionaire)? If not, why should we believe them more than Grok? I certainly don’t believe Grok, either, but don’t disbelieve the prediction just because it’s ai.

    As I think about reputations and who to believe, and why, lots of Nobel winning economists have been wrong on lots of policy outcomes they’ve predicted in the past. Including how bad Trump 45’s tariffs would be, when the actual outcomes looked pretty good.

    Since nobody knows, competing plausible scenarios are good, but unfortunately need a good number of words to Wade thru.

  21. Now, I send a large gratitude to writer Grunt for above generous salubrity.

  22. I have decided to RESIST Trump the Authoritarian, by not setting my watch to Daylight Saving Time. This will also save me the trouble of resetting it in November. Hope I’m not late to all my appointments!

Leave a Reply

Your email address will not be published. Required fields are marked *

HTML tags allowed in your comment: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>