The Z Variable, and the Country That Just Broke a 40-Year Rule

Two equations get written on the board today, and Tantri is upfront that the payoff is still a class away: “Today, we are only going to set it up. Insights will come from next class.” The setup is worth having anyway, because the last twenty minutes turn it on Japan — forty years of near-zero interest rates, a carry trade that quietly funds power projects and pension funds worldwide, and the first real sign that the rule is breaking.

The setup, briefly

Workers ask for a wage based on three things: what they expect prices to be, how much unemployment there is around them, and Z — a catch-all for everything outside the firm’s control, mostly government-set: labor law, minimum wage, how strictly it’s enforced. Firms, meanwhile, set actual prices as wages plus a markup, the markup being whatever product-market competition allows them to keep. Put the two together and you get real wage — what a worker’s pay is actually worth in goods — and a natural rate of unemployment: the point where what workers expected and what firms actually paid turn out to be the same number, so nobody is surprised and nothing needs to adjust. (The full mechanics of this — including why this is the machinery underneath the Phillips curve — get their proper derivation in Session 11; this class is the first formal pass at it.)

The one piece worth sitting with here is Z, because it does something the rest of the model doesn’t: it moves wage demands without moving productivity at all. Tantri’s example is deliberately blunt: a government passes a strict minimum-wage law, and workers duly ask for more — but the firm’s markup, set by product-market competition, hasn’t changed, and neither has what the worker actually produces. “There is no equilibrium in hiring the guy. That’s the equilibrium — unemployment rate will go up.” The firm cannot simply raise prices to cover it, because — as he puts it to a room full of people who’d otherwise default to cost-plus thinking — “pricing is not cost plus… P is whatever it is. You can’t do anything within the price, you have to figure out your cost.” So the only variable left to absorb a Z-driven wage floor is the number of people employed. This, he says, is “a very, very solid empirical fact… as you tighten labor laws, unemployment rate goes up” — the reason Europe’s unemployment rate has sat persistently above America’s, and, closer to home, his own research on NAREGA finds the identical mechanism in India’s rural labor market.

Then the class turns to Japan

A student’s question about Japan’s negative-rate era — why didn’t the Bank of Japan just treat it as Japan’s permanent steady state? — pulls the whole discussion into the present tense, and this is where the class earns its keep.

For decades, Japan ran the arrangement almost too well to believe: print money at essentially zero cost, invest it abroad — equities, riskier foreign bonds — and earn something like 6% doing it. “Imagine a situation, your foreign investment gives 6% return, cost is zero for printing money.” On that spread, Japan could carry a debt-to-GDP ratio of roughly 200% without ever meaningfully taxing its own population to service it, because — in the model’s own terms — Japanese inflation expectations simply never moved. Z stayed put, wages stayed put, the natural-rate machinery never had reason to reprice anything.

That stability turned into the foundation of a trade the rest of the world leans on more than most people realize: borrow cheap and short-term in yen, invest it long-term somewhere else. Tantri’s own example is not hypothetical — Power Finance Corporation, where he sits on the board, borrows three- and five-year money from Japan and lends it out for twenty-year power-project loans. The arrangement works exactly as long as Japanese rates stay near zero.

They are no longer staying near zero. Post-COVID inflation expectations rose “anywhere in the world, in most places, and so is in Japan” — and for the first time in a generation, Japanese wages and prices are actually moving. “A generation in Japan hasn’t heard of something known as inflation. Now they are having 2% inflation.” The tell he points to is specific and startling: Japan’s 40-year government bond, for the first time, priced at a 4% yield — “unheard of for Japan.” The Bank of Japan, trying to catch up, has pushed its policy rate to 0.75%, which by his account is “not sufficient, because the inflation expectation is going even higher pace.”

The risk isn’t abstract for the people in the room. If Japan keeps tightening, the carry trade that funds long-dated loans everywhere from Indian power projects to global pension books gets more expensive to roll over — and, as he notes, dryly, about the cost to himself personally, “I will lose my board membership… these guys have borrowed a lot of money.” Whether the story ends in an orderly unwind or something rougher, his honest answer to the room was that it’s genuinely unresolved: “I know it is not a very, very clean answer… I’ll send you that paper, read that, and then we’ll discuss again.”

What You Can Now Do

Two separate but related habits, one from each half of this class. First: when you hear that a labor-market rule (a minimum wage, a hiring protection, a state jobs guarantee) is unambiguously good for workers, check what it does to Z — and remember that a rule which raises wage demands without raising what workers actually produce doesn’t get absorbed for free; it shows up as unemployment, exactly where NAREGA’s own data shows it. Second: next time a headline says “Japan is finally tightening,” don’t file it as a domestic Japanese story. Ask who is funding long-term assets with short-term yen — because the answer, once you start looking, tends to be a lot closer to home than you’d expect.

Next: /2026/02/medium-term-inflation-labor-markets-phillips-curve/ — where these same two equations get their full formal derivation, and the natural rate stops being a setup and starts being the diagnosis.