Who Really Pays, and When Not to Claim

Who actually pays

Revenue, and who carries it

Kyle Pomerleau and Erica York compare how the main budget forecasters estimate tariff revenue. These include the Congressional Budget Office (Congress's nonpartisan scorekeeper), the Tax Policy Center (a Brookings–Urban Institute joint venture), the Yale Budget Lab, the Penn Wharton Budget Model and the Tax Foundation. With tariffs held roughly at 2025 levels, the forecasts put revenue at about $2.5 trillion over ten years, around $250 billion a year. Forecasts that allow for tariffs slowing the economy come in lower. On who bears the cost, the forecasters that model it agree: tariffs take a larger share of income from low- and middle-income households than from those at the top. That fits the pass-through finding. A tax that shows up in the price of goods hits hardest the households that spend most of their income on goods.

Jobs: concentrated losses, a surprising response

Rafael Dix-Carneiro and Brian Kovak review what is known about trade and jobs. Import competition can cause "sizable and highly localized" damage. Losses cluster by region, industry and occupation, recovery is slow, and the harm spreads to whole local communities. Older workers, who find it hardest to retrain or move, bear an outsized share.

That might seem to support protection, but they find a catch. When tariffs shield an industry, firms tend to invest more in machinery and earn higher profits rather than keep employment high. Protection helps the companies more than it saves the jobs. Their alternative is "smoothing policies," meaning targeted help for the people hit hardest, so that the economy as a whole can still collect the longer-run gains from trade.

Is a tariff ever the right tool?

Trade deficits: the savings arithmetic

Pierre-Olivier Gourinchas, Gene Kindberg-Hanlon, Manasa Patnam, Lorenzo Rotunno and Michele Ruta restate the standard view among economists: tariffs alone barely move the trade deficit. The deficit is a macroeconomic outcome, set by how much a country saves compared with how much it invests. If national saving and investment don't change, cutting some imports mostly shifts the gap somewhere else, for example through a stronger dollar or fewer exports. Industrial policy aimed at particular sectors runs into the same limit.

The long view

A seven-author history paper led by Miguel Acosta builds a database of US tariffs going back to 1789. Doing so meant classifying products consistently across centuries and pulling rates out of 19th-century Congressional documents. They describe three eras, the "three R's": revenue (tariffs as a main source of federal money before the Civil War), restriction (protecting new industry after it), and reciprocity (negotiated mutual cuts beginning in the 1930s). Taylor highlights one trend: tariffs have become steadily more product-specific, down to particular types of cheese, because a narrowly defined product can be used to hit some countries and spare others. The number of separately tariffed products keeps rising.

What stays open

Taylor's post only samples the papers, so the methods behind figures such as the 90% pass-through or the $2.5 trillion are not set out here. The revenue estimates assume tariffs stay near 2025 levels, which is uncertain given the legal challenges. Taylor ends with his own view. Trump "owns" the tariffs: he imposed them without Congress and offered one legal rationale after another as courts pushed back, and his promises about the benefits rested on "demonstrably false beliefs." Taylor admits a darker hope, that the costs become obvious enough to teach the next few decades why this approach doesn't work.

Two sets of weekly housing data for the week ending October 23, 2026 point the same way. Fewer people are applying for mortgages than a year ago, and more homes are sitting on the market. Bill McBride of Calculated Risk reads the pair as a sign that existing-home sales will slow over the coming months and that prices will come under pressure.

The demand side: mortgage applications

• Purchase applications (people buying homes) rose 4% for the week, seasonally adjusted, but were 18% lower than the same week in 2025 (unadjusted).
• Refinance applications rose 3% for the week but were 61% lower than a year earlier.

The weekly bounce came from rates. According to Joel Kan, the MBA's deputy chief economist, the average 30-year fixed rate fell for the second straight week, to 7.37%, its lowest in three weeks. That is still more than a full percentage point above the same week last year.

• 5.7% higher than the same week in 2025, up from 5.3% the week before;
• 1.8% lower than the same week in 2019, narrowing from 3.2% below the week before.

The 2019 comparison matters because analysts use that year as the last normal market before pandemic-era rates of around 3% emptied the supply of homes for sale. Inventory is now within a couple of percent of that baseline and closing the gap week by week.

Realtor.com, which measures listings differently, puts the yearly increase higher, at 7.7%. That is its fastest weekly growth since late January, up from about 4% in late August. Realtor.com is specific about the cause: the rise is "not being driven by a flood of new sellers." What's happening is that buyers usually pull back in the fall, and this year the homes they would have bought are staying listed and accumulating.

How the two sides connect

Inventory is a stock, not a flow. It is the number of homes listed for sale at a given moment. It rises when new listings arrive faster than homes sell. Realtor.com's explanation says the second half is doing the work here.

An illustration with made-up numbers: suppose a market gets 100 new listings a week and normally sells 100. Inventory stays flat. If sales drop to 90 because buyers are waiting out 7%+ rates, inventory grows by 10 homes a week even though no extra sellers showed up. Over a season that adds up. Sellers whose homes linger face a choice between waiting and cutting the price.

That is McBride's chain of reasoning. High rates suppress purchase demand, which shows in the 18% drop in applications. Slower sales let listings build up, which shows in inventory up 5.7–7.7%. More supply chasing fewer buyers pushes on prices. He concludes that with 30-year rates "just under 7 1/2%," existing-home sales "will likely slow" and prices "will be under pressure."

What to keep in mind

• Weekly data is noisy. That's why McBride's purchase chart shows a four-week average (red) next to the weekly series (blue), and why the yearly comparisons say more than the week-to-week ones.
• "Pressure" isn't a forecast of a crash. The post doesn't predict how much prices will move or how fast. Inventory is still slightly below 2019, which by that benchmark is not a glut.
• Lock-in cuts both ways. It holds down demand from move-up buyers, and it also holds down supply, because owners with cheap loans don't list their homes. The post doesn't say how much each force contributes.
• Rates are the swing factor. The past week's small rebound in applications followed a small drop in rates. The post doesn't say what would happen if rates fell further.
• National numbers hide regional differences. None of these figures are broken down by region, and housing markets vary a lot from place to place.

Joachim Klement, writing on his Substack, reviews a 2026 study by Aizenman and co-authors on whether central-bank gold helps defend a currency. The question matters now because, in Klement's view, surprise moves by the Federal Reserve look increasingly likely under the new Fed Chair. Gold advocates argue that central banks should hold more gold because it protects against both a falling home currency and a falling dollar. The study's answer is narrower. Gold seems to work about as well as reserves in general, but no better than holding dollars directly, and the evidence for its effect is weak.

The problem central banks face

Many central banks, especially in emerging markets, try to keep their exchange rate steady against the US dollar. A surprise Fed rate hike makes dollar assets more attractive. Capital tends to flow out of these economies and their currencies weaken. A sharp fall raises the local cost of imports and of debt taken out in dollars, which is why central banks fight it.

How the study measured it

A rough illustration, scaling the reported rates and assuming the effect grows in step with the size of the shock: suppose two similar emerging-market currencies face a 50 basis point Fed surprise. If one central bank holds one standard deviation more total reserves, its currency would fall about 0.2 percentage points less (5 × 0.04). If the extra holdings were specifically dollars, the gap would be about 0.5 percentage points (5 × 0.1).

Results

Klement calls this intuitive. To defend a currency against the dollar, you need dollars to sell. Other currencies have to be converted first and may themselves fall when the Fed tightens.

Where gold comes in

Gold is priced and traded in dollars, so a central bank can sell it to raise dollars. In that sense, Klement notes, it is an indirect dollar holding. The researchers repeated the analysis three ways: FX reserves plus gold, FX reserves alone, and gold alone.

Gold's dampening effect came out about the same size as that of FX reserves in general. But the range of uncertainty around it was wider, so the gold effect was not statistically significant. Gold added no measurable protection against Fed shocks beyond what other reserves already provide.

What Klement concludes, and what stays open

Klement says the analysis supports no "hard and fast" conclusions, but suggests two. First, currency markets seem to treat gold as an indirect dollar asset that can be sold for dollar cash when needed. Second, gold reserves don't improve currency stability compared with simply holding dollars.

Some limits to keep in mind:

• Insignificant is not zero. Gold's estimated effect was about as large as general reserves. The problem is noise, not a clear absence of effect. A longer sample might sharpen it either way.
• Only one job is tested. The study looks at gold as a cushion against Fed rate surprises. It doesn't address other reasons central banks give for holding gold, such as protection against sanctions or reducing reliance on the dollar itself. Klement doesn't discuss these.
• The effects are small. Even dollar reserves reduce depreciation by only 0.1 percentage points per 10 basis point surprise. Reserves soften the blow but don't stop it.
• Details are missing. The summary doesn't give the sample of countries, the period studied, or how Fed surprises were identified, so it's hard to judge how widely the results apply.

For investors, the practical point is about the story often told to justify gold prices. If central banks buy gold to defend their currencies, this evidence suggests dollars do that job better. Gold works as a less reliable stand-in for dollars, not a superior shield.

Alex Tabarrok, the economist who co-writes the Marginal Revolution blog, points readers to a post by Adam Nash on one of the most misread charts in personal finance: US household wealth by age.

The chart that looks unfair

Nash writes that every year he pulls the most recent US Census data on household net worth, broken down by the age of the household. Net worth means assets such as home equity, savings and retirement accounts, minus debts. The data use the median, the middle household, so a handful of very rich families don't distort the picture.

The difficulty is that a single snapshot mixes two different things. One is inequality, meaning some people really are better off than others. The other is the life cycle: people at different stages of the same path will naturally hold different amounts of wealth. A chart sorted by age can't tell you which one you're looking at.

Nash removes inequality completely and asks what the chart would look like then. In his imaginary Equal Land, everyone:

• works from age 25 to 64 and retires at 65
• lives to exactly 85
• earns the same $100,000 a year
• saves 10% of income in every working year
• earns 8% a year on savings, every year
• spends 4% of the portfolio each year in retirement

Every person follows the same path. The only difference between two people is how far along that path they are.

What happens

Nash's result is that wealth in Equal Land is even more unequal by age than in the real world.

The article doesn't publish Equal Land's numbers. Here is an illustrative calculation of our own using Nash's assumptions, with savings added at the end of each year.

• Age 30 (5 years of saving): about $59,000
• Age 35 (10 years): about $145,000
• Age 65 (40 years): about $2.6 million

Total contributions are only $400,000 ($10,000 × 40). Compounding supplies most of the $2.6 million, and it pays off most at the end. In the last ten working years alone the balance grows by well over $1 million.

Retirement doesn't reverse this. A retiree earning 8% and withdrawing 4% still grows the portfolio by about 3.7% a year (1.08 × 0.96 ≈ 1.037). Someone aged 67 holds roughly $2.8 million, and at 85 the figure passes $5 million. Under-35 households, meanwhile, range from zero to about $145,000, with a middle near $59,000.

The ratio between the 65–69 group and the under-35 group comes out around 45 to 50 times, using these simplified assumptions. That is several times the real-world 9.83x. A society with no inequality at all produces a steeper age-wealth chart than the actual United States.

An analogy, offered here only to illustrate the idea: comparing a seedling with a 40-year-old oak tells you nothing about whether the forest is fair. The two trees are the same species at different ages.

Within-age versus between-age

Tabarrok is careful not to overclaim. He says this doesn't mean all, or even most, inequality comes from age. Economists can split total inequality into two parts:

Caveats the model sets aside

Equal Land is a teaching device, and its simplifications matter.

None of this weakens the main point. These factors change how big the gap is, not the fact that wealth rises steeply with age.

What it means for you

Two practical lessons follow. First, a fair benchmark for your own net worth is people your own age, not the median for everyone, and certainly not your parents. Second, the numbers show why starting early matters. In Equal Land, the first decade of saving produces only about $145,000. Left to compound for another three decades, though, that early money grows into a large share of the final total.

The article leaves open how much of the real 9.83x ratio is pure life cycle and how much reflects true differences between generations. Answering that would require tracking the same households over time, and Equal Land doesn't attempt it.

His policy almost certainly covers some of it. The question he raises is whether he should use it. In his words, owning a policy and using it "are separate decisions."

The arithmetic says claim

Crothers begins with the numbers you'd expect.

Whether claiming pays then depends on how long the surcharge lasts:

Surcharges seldom last as long as seven years, so on this table claiming wins. Most people would stop here and file.

What the table leaves out

Crothers names three costs the table doesn't count. Each one moves the decision in the same direction.

1. Losing the policy. Few insurers will cover a coastal home in a region prone to Atlantic storms, and fewer still if the home sits empty for long periods. A claim, even a small one, could lead his insurer not to renew the policy. He would then have to find new coverage in a thin market, probably at a higher price. That cost isn't capped at $500 a year. It could run on with no fixed end and could be much larger.

2. A bigger surcharge than assumed. The 25% increase is a guess, and Crothers thinks it may be optimistic for a property insurers already regard "with less than enthusiasm." If the surcharge is 50%, or $1,000 a year, break-even arrives in three and a half years, not seven.

3. A smaller payout. The $3,500 is a ceiling, not a promise. Policies often cap what they pay for outdoor items. An adjuster could also ask whether leaving a vacant property unwinterized affects coverage. Both would reduce the recovery.

Put together, his best case is a gain of $1,000 to $2,000. His bad case is losing affordable cover on a property that is hard to insure. Against those odds, he writes, he would "rather not test my insurer's patience over a loss I can afford."

A worked illustration

The principle underneath

Crothers puts his rule this way: "I keep the policy for the catastrophic loss, the kind that would genuinely hurt, and I pay for the smaller ones myself." This is the standard advice on insurance. It's for losses you can't absorb. Small losses you can cover are often cheaper to pay yourself, because each claim uses up some of your insurer's goodwill, and you may need that goodwill when something big happens.

He is open that the rule depends on his circumstances. He can absorb $5,000 without real hardship, and he calls owning a second home a privilege. For someone who can't absorb the loss, the arithmetic table alone might point the right way.

Trade-offs and open questions

The post is a personal reckoning, not a study, and several inputs are guesses. He doesn't say how long surcharges actually last with his insurer, how likely non-renewal really is after one small claim, or what his policy says about outdoor property and vacancy. Calling the insurer to ask could alert them even if he never files. The post doesn't discuss that, but many people worry about it. Readers in calmer markets, with insurers competing for their business, face different odds and may find claiming makes sense.

There's one cost he does acknowledge: the unwinterized property was his own mistake. He plans to winterize fully before leaving this time. In spring, he writes, he'll trail behind his wife, Suzie, while she picks new furniture, and he'll offer no opinions.