When Elon Musk took over Twitter in October 2022, he bemoaned what he saw as its bloated ranks of software engineers, according to Walter Isaacson’s biography. If all 2,500 of them wrote just three lines of code each day, he reasoned, that would be “enough for a whole operating system”.

It was the prelude to mass layoffs at the social media company, now known as X. But Musk ended up asking some of the fired workers to come back – effectively acknowledging that the calculus of inputs and outputs was not as simple as he had made out.

How to even measure productivity, let alone make it better, is a big question that corporate chiefs and economists are trying to answer. Getting it right has rarely been more important. Productivity numbers may well be what settles the fight between bosses and employees about whether work-from-home really works – and determines whether artificial intelligence (AI) technologies live up to the hype or prove a bust. In the wider economy, all kinds of worries, including problems associated with ageing populations and mounting debts, could potentially melt away if only businesses can get productivity running at a decent clip.

The concept involves measuring how much output workers can produce in a given time with the available technology. The United States government’s official gauge is derived from aggregating data such as hours worked and pay across various industries. That is where the trouble starts, says Jason Furman, a professor at Harvard University and former head of the White House Council of Economic Advisers. “Productivity is perhaps the most volatile major economic statistic,” he says. “It takes an error-prone numerator – output – and divides it by an even more error-prone denominator – hours.”

For that reason, experts look beyond just quarterly and even yearly fluctuations in productivity. Some of the pandemic-era numbers are especially misleading. Early 2020, for example, saw a productivity surge unmatched since 1947 – but there was nothing to celebrate, just mass layoffs in which less-productive workers were the first to be fired. Likewise, it is too early to read much into the jump of 5.2 per cent in the most recent quarter, more than treble the recent average.

“It’s tough to measure, even in the rearview mirror,” says Adam Ozimek, chief economist at the Economic Innovation Group. “There are still, today, economic historians debating what happened to productivity growth in the 1940s.” (At this month’s annual gathering of the American Economic Association in San Antonio, one session explored the productivity impact of America’s rubber shortage during World War II.) There is plenty of timelier work, too: a recent study tries to hone productivity measurements by looking at the European debt crisis and the introduction of the 35-hour week in France.

GET BT IN YOUR INBOX DAILY

Start and end each day with the latest news stories and analyses delivered straight to your inbox.

Sign UpVIEW ALL

‘Huge’ benefits

There are several reasons why these old puzzles have taken on a new urgency. The emergence of hybrid work as the dominant model at most white-collar businesses has bosses scrambling to figure out the optimal blend of remote and in-office collaboration. And, they have to decide how generative AI can enhance worker productivity –typically a hit-and-miss business with new technologies. A recent study of 758 Boston Consulting Group analysts using ChatGPT found that there were “huge performance benefits” for some tasks, while others were more easily performed without it. Employees at tech-focused venture capital firm Seven Seven Six, headed by Reddit co-founder Alexis Ohanian, are using AI to draft press releases and automate data entry.

Beyond the boardroom, productivity may become more important as population growth in developed countries slows down, making it harder to rely on an expanding workforce for economic growth. And societies are ageing. In 1920, just one in 20 Americans was 65 or older. Now that is one in six. There is concern about how a smaller share of workers will provide for a larger share of retirees. Productivity growth at a decent clip would make the task much easier.

Another trend to account for is the economy’s shift from churning out goods to services, which complicates the calculation. It is easier to assess the productivity of a worker at an auto plant than one engaged in, for example, home care for the elderly – an industry that is projected to be among the biggest sources of US job growth in the coming decade. Some measures rely on simply asking people if they feel productive, or counting e-mails sent, or lines of code written.

At the business level, yardsticks for desk workers vary. Those in professional services, such as consultants, are measured by “utilisation” – the ratio of billable hours to total time worked, along with project outcomes and client satisfaction. For tech workers, it is all about the duelling acronyms of KPIs (key performance indicators) and OKRs (objectives and key results). KPIs, like the number of software bugs fixed, are granular and specific to each role, while bigger-picture OKRs ensure that individual goals align with the company’s broader strategy. But, if you are a middle manager at a tech giant, productivity could be gauged by how many people work under you – an incentive to overhire.

‘Feeling good’

The confusion over what productivity means is not just an abstract problem for economists; it can also foment distrust inside an organisation. A recent example: almost nine out of 10 workers polled by Microsoft in 2022 reported being productive, but just one in 10 leaders in the same survey expressed confidence that their own teams were firing on all cylinders.

That is likely because both groups have their own distinct definitions of efficiency. For chief executive officers such as Meta Platforms’ Mark Zuckerberg and Jamie Dimon at JPMorgan Chase, it typically means seeing actual butts in seats, while for employees, productivity is often a more abstract feeling of accomplishment.

That gap is fuelling arguments over work-from-home. Research on the productivity impact often focuses on small slices of the white-collar world – Indian data-entry workers, say, or Chinese call-centre staff – and cannot be generalised. Notably, not having to commute makes remote workers feel more productive, but their bosses do not recognise that as any sort of gain. A recent paper from the Federal Reserve Bank of San Francisco examined 43 industries and found that there is “essentially no relationship” between working from home and productivity.

It has all got so muddled that one efficiency expert made a splash last year with a book that simply argued that “feeling good” was the best path to productivity.

“It’s such a funny thing that clearly does not measure what it purports to measure,” says Julia Pollak, chief economist at job site ZipRecruiter. “Companies are investing time, money and thought on doing this better.” BLOOMBERG