The last mile of inflation is the hardest

Core PCE prices rose at an annualised 2.8 per cent in October, the fourth month in a row that core inflation has stuck near three per cent. The sceptics in the bond market warned that the 'last mile' of disinflation would be the hardest. So far they have been right.

From five per cent to a plateau

The first miles were easy by comparison. Core PCE inflation, the Federal Reserve's preferred gauge, peaked above five per cent in 2022 as supply chains snarled and pandemic savings chased scarce goods. It fell quickly through 2023 as shipping costs normalised, used-car prices dropped and goods prices flattened or fell. Those one-off reversals are now largely spent, which leaves the stickier parts of the index to set the pace.

The headline measure looks better only because energy has been cheap. Overall PCE prices rose by 2.3 per cent in the year to October, close to target, while core stayed stuck. In September officials projected core PCE inflation ending this year at 2.6 per cent. The October reading suggests they will miss even that modest goal, with two months of data left.

Services won't budge

Traded goods have seen plenty of disinflation. Non-housing services and domestic wages are still rising at rates that do not square with the Federal Reserve's two per cent target. The incoming administration has promised universal tariffs, corporate tax cuts and tighter limits on labour supply, all of which push prices up. The Fed may have to live with inflation settling well above its goal.

Services prices are mostly labour costs, and pay settles slowly. Wage growth has cooled from its peak but still runs faster than the target allows unless productivity keeps growing at its recent brisk pace. Housing adds a further lag: official rent measures track the whole stock of leases, which reprice about once a year, so they trail market rents by many months. Insurance premiums and medical fees also reset annually, carrying last year's cost increases into this year's index.

The PCE index has a quirk of its own. It includes portfolio management and investment advice fees, which are priced off the value of assets under management, so a strong stock market pushes measured services inflation up. With American equity indices at record highs this autumn, that component has added to the core reading without saying much about the price pressure households face.

Labour supply is the swing factor. Immigration surged over the past two years, adding workers faster than most forecasters expected and helping wage growth cool without a rise in layoffs. Promised restrictions on immigration, and possibly deportations, would reverse that source of slack. Fewer new workers in construction, farming, hospitality and care would push wages up in the very services that make up the stubborn part of the index.

The Fed is still cutting

The central bank has lowered its policy rate by three-quarters of a point since September, to 4.50–4.75 per cent, and futures markets lean towards another cut in December. Minutes of the November meeting, published this week, showed officials favouring a gradual path from here. Cutting while core inflation sits nearer three per cent than two risks confirming what households and firms have begun to suspect, that two per cent has become a floor rather than a goal.

The costs of a plateau fall unevenly. Savers and holders of inflation-linked bonds are protected, while households with fixed incomes, and workers whose pay lags behind, lose purchasing power each year the overshoot persists. Borrowers gain in real terms, including the federal government, whose debt is eroded by higher inflation. Bond investors lose if the extra inflation is not priced into yields, which is one reason long-dated Treasury yields have risen since the Fed began cutting.

A 1970s echo

The historical warning is the 1970s, when the Fed under Arthur Burns eased each time inflation dipped, only to see it come back higher in the next wave. A repeat on that scale looks unlikely, since inflation expectations are far better anchored today and the Fed's credibility is much greater. The risk is subtler. A central bank that grows comfortable with three per cent inflation in a period of tariffs and tight labour supply may find the rate harder to shift once firms build it into their pricing.

A gentler precedent comes from the 1990s. Under Alan Greenspan the Fed spent years with inflation between two and three per cent, and some officials argued openly for waiting for favourable shocks to bring it lower rather than forcing it down with tighter policy. Productivity growth rose in the second half of the decade and did the job for them. The present Fed may hope for a similar outcome from new technology. It would be relying on luck to deliver what policy has not.

What would change the call

Officials have their own reply. With the funds rate at 4.50–4.75 per cent and core inflation at 2.8 per cent, the real policy rate is still close to two percentage points, above most estimates of neutral. On that view, trimming rates only stops policy from tightening by default as inflation drifts down, and holding still would be the more aggressive choice.

The case against the column is that the stall is an artefact of lagging components. Rents on new leases have slowed sharply, and as those feed through, housing inflation should fall through next year and drag core PCE down with it. If monthly core readings slip back towards two per cent annualised over the winter and wage growth keeps cooling, the plateau will look temporary. If instead the first tariffs land on top of a core rate near three per cent, the Fed's projections for next year will need rewriting, and the December dots will be the first place to look.

Core inflation has stalled near three per cent. If the Fed keeps cutting into that, it will probably lock in an overshoot that lasts well into next year and beyond. Expect core PCE inflation to end this year closer to 3 per cent than to the 2.6 per cent officials projected, and the December dots to show fewer cuts for next year.