The Productivity Puzzle No One Wants to Solve

Twice a year, the Council I sit on publishes a review of Norway's productivity trends, and twice a year it gets roughly the same reception: polite acknowledgment from the press, a short segment on the evening business news, and then silence until the next release. Compare that to the reaction a half-point change in the policy rate gets, or a single percentage point move in unemployment, and you start to notice something odd. Productivity growth is, by a wide margin, the single largest determinant of whether living standards rise over a generation. It is also the topic in economic policy that generates the least political energy of any I've worked on.

I've come to think this isn't a coincidence. Productivity reform is politically unloved for structural reasons, not because people don't understand its importance, and those reasons are worth naming plainly, because until they are, the reviews will keep landing on deaf ears.

The problem has no villain

Most economic policy debates have a shape that political systems know how to process: there's a cost, there's a benefit, and there's usually an identifiable group on each side. Raise taxes on high earners, and you know who's unhappy and who's pleased. Cut a subsidy, and you can point to the industry that will complain. Productivity reform doesn't fit this shape, because slow productivity growth doesn't have a face. Nobody wakes up angry that the economy grew 1.1 percent instead of 1.6 percent. The cost is diffuse, distributed across everyone, and delayed by years which makes it one of the hardest categories of problem for any democracy to prioritize, because democracies are quite good at responding to concentrated, visible, near-term pain and much worse at responding to slow, distributed, long-term drag.

Meanwhile, the actual levers that move productivity planning reform that lets denser housing and infrastructure get built faster, competition policy that keeps incumbent firms from coasting on market power, education spending that pays off over fifteen years rather than four, R&D incentives whose returns are genuinely uncertain each of these does have a concentrated, visible cost attached to a specific, sympathetic group. Planning reform generates furious opposition from existing homeowners and neighborhood associations. Competition policy makes powerful incumbent firms very unhappy, and those firms employ lobbyists whose job is specifically to make that unhappiness heard. So you end up with an unusual and politically toxic combination: diffuse, invisible benefits and concentrated, vocal costs. Any policy area with that shape will lose in the ordinary competition for legislative attention, almost regardless of the underlying economics.

What the data actually shows

Setting the politics aside for a moment, the numbers are worth sitting with. Across most advanced economies, productivity growth has slowed meaningfully since the 2008 financial crisis, and the slowdown has persisted through a technology cycle the spread of cloud computing, then AI tools that many economists, myself included, expected to show up more clearly in aggregate output per hour worked. It largely hasn't, at least not yet, and that gap between the promise of new technology and its measured economic impact is itself a puzzle economists have been arguing about for going on a decade.

Part of the explanation is almost certainly measurement. A lot of what digital tools produce faster customer service, better search results, more convenient scheduling doesn't show up cleanly in GDP statistics that were designed to count physical output. But I don't think measurement error explains all of it, or even most of it. A more uncomfortable explanation is that new technology only translates into productivity gains once it's paired with organizational change, and organizational change is slow, expensive, and something most firms are reluctant to undertake unless competitive pressure forces the issue. Give a firm a faster tool and, absent pressure to restructure how work is actually done around that tool, it will often just do the same work slightly faster rather than fundamentally differently which shows up as a small, disappointing bump rather than the step change the technology's underlying capability would suggest is possible.

This is where competition policy and productivity growth turn out to be much more tightly linked than they first appear. An economy with genuine competitive pressure forces firms to actually restructure around new capability, because their rivals will if they don't. An economy where a handful of incumbents can comfortably absorb new technology without changing anything fundamental because market structure protects them from a rival doing it faster will systematically under-realize the productivity gains new technology makes possible. I suspect this is a meaningfully underrated part of why some economies have converted the current technology wave into growth more effectively than others, and it has very little to do with how much any given country has spent on R&D subsidies.

What I think would actually move the needle

If I had to rank the levers by plausible impact relative to political difficulty, planning and permitting reform would sit at the top for one simple reason: it's the rare productivity lever whose benefits, while still diffuse, are at least visible on a reasonable timeline. Housing gets built faster, infrastructure projects that used to take a decade of approvals take three years instead, and people can actually see the change within an electoral cycle rather than a generation. That visibility matters politically in a way that's easy to underrate when you're focused purely on the economics.

Competition policy is harder to sell but probably matters more in the long run, precisely because its effects compound. An economy that stays genuinely contestable keeps forcing the organizational adaptation that turns new technology into real output, year after year, rather than getting a one-time bump and then plateauing.

Education spending is the hardest of all to defend politically, because the payoff arrives so far in the future that it's genuinely difficult for any government to claim credit for it by the time a curriculum change shows up in workforce productivity, two or three different governments will have come and gone, none of whom will get to claim the win.

None of these are new insights, and I don't think the Council's next review will discover some previously unknown fourth lever nobody has thought of. What I'd like to see change isn't the diagnosis, which economists have gotten right for years it's the willingness of political systems to spend visible capital today on a benefit that won't be legible until well after anyone currently in office is out of it. That is a genuinely hard ask, and I don't have a clever solution for it. I just think it's worth saying plainly, rather than pretending the problem is that nobody's noticed.