Labour market data is politically potent, technically complex and often imperfect. Recent US events are a reminder that Australia’s own systems deserve closer scrutiny.

The announcement from the Trump administration that US monthly jobs reports might be suspended “until they can get the data and methodology in order” has generated predictable outrage. Critics point out, correctly, that the same methodology served successive presidents, including Trump himself, without complaint. The sudden declaration of “unreliability” now, when the numbers may be politically awkward, is hard to view as anything other than a pretext.

But beneath the political theatre lies an uncomfortable truth: labour market data is complex to collect, interpret and present. It is imperfect in every country, including our own. And while outright suspension is rare, the temptation to “fix” or “massage” the numbers is not confined to Washington. Canberra is no stranger to that game.

Labour market statistics are not a direct headcount of people with jobs. In the US, the Bureau of Labor Statistics relies on two entirely separate monthly surveys: a household survey capturing individual employment status and a payroll survey tracking jobs through employer records. In Australia, the Australian Bureau of Statistics uses a rotating panel survey of households, extrapolating from a sample of about 50,000 people to estimate national employment. Both systems involve seasonal adjustments, weighting for underrepresented groups and periodic benchmarking against more comprehensive administrative data.

These layers of adjustment are necessary because raw numbers can mislead. Employment spikes before Christmas and drops in January. Students and retirees drift in and out of casual work. Industries shed workers in predictable seasonal cycles. Without smoothing for these patterns, we would mistake noise for trend. But each adjustment introduces its own set of assumptions and each assumption is contestable.

The result is that “the unemployment rate” is as much a statistical artefact as it is a reflection of lived reality. Change the definition of “actively looking for work” or the way underemployment is factored in and the headline number moves, sometimes by enough to shift a political narrative. Successive Australian Governments have been perfectly willing to use this to their advantage. Expanding training programs that reclassify people from “unemployed” to “not in the labour force”, tweaking the treatment of casual and part-time work or shifting survey timing can all produce politically useful bumps in the figures.

And while the ABS operates at arm’s length from government, it works within funding and priority settings determined by the political environment. Like any statistical agency, its release schedules and chosen measures are shaped by resource constraints, evolving policy needs and technical considerations. The methodology underpinning the “headline” unemployment rate has attracted periodic criticism from labour economists, particularly for its narrow scope and its tendency to understate aspects of job insecurity.

It is tempting, especially in Australia, to dismiss the US as a place where politics corrodes institutions in ways we would never tolerate. But we should be cautious. The complexity of labour market measurement makes it inherently vulnerable to political framing, even in systems that remain technically independent. Governments everywhere have both the motive and, to some degree, the means to influence not just how data is interpreted but how it is collected and when it is released.

The deeper question is whether our current approaches to measuring labour market health are still fit for purpose. Both here and overseas, headline unemployment rates fail to capture the rise of insecure work, the gig economy, long-term skill mismatches and underemployment. Hours worked, job quality and income security matter at least as much as whether someone did one hour of paid work in the survey week. Yet, these dimensions are often relegated to secondary tables, far from the political spotlight.

Labour market data will never be perfect. Human behaviour is too messy and the economy too dynamic. But we should resist the lazy comfort of treating the monthly unemployment figure as a definitive scorecard. In Australia, as in the US, the real danger is not just in the data being “cooked” but in the public becoming accustomed to shallow, incomplete measures that suit governments more than they inform citizens.

If there is a lesson to be drawn from the current US controversy, it is not simply that Trump is playing politics with statistics, though he plainly is. It is that we ought to examine whether the way we measure and report on labour markets is genuinely up to the task in an era of economic and workplace change. Because if we do not trust the numbers, and if they do not capture the reality of work, then we risk basing policy and public debate on shadows.

The views expressed in this article may or may not reflect those of Pearls and Irritations.