Why Can Working Just One Hour a Week Mean You Are No Longer Statistically Unemployed?

Geoffrey Chen

When Numbers Start Making Decisions · Season One, “People on Either Side of a Threshold” · Article 1

1. What does one hour of work change?

Imagine that someone used to work twenty hours a week. Business at the shop then declines and their roster is repeatedly cut. In the week covered by the labour force survey, they are given only one hour of work.

The income from that hour does almost nothing to change their life. They still need to look for another job. They still cannot rely on their present earnings to pay the rent and meet ordinary expenses. They do not even know whether another shift will be available next week.

Yet in Australian labour force statistics, this person may no longer be classified as unemployed.

Under the Australian Bureau of Statistics definition, a person aged 15 or over is generally counted as employed if, during the reference week, they worked for one hour or more for pay, profit, commission or payment in kind. Someone who worked for at least one hour without pay in a family business or farm may also be counted as employed. Some people who performed no work during the reference week remain employed because they retained a clear attachment to a job and were only temporarily absent. ABS: Labour Force, Australia methodology

The practical difference between working zero hours and working one hour may therefore be almost invisible, while the statistical identities lie on opposite sides of a boundary.

That contrast naturally produces suspicion. How can someone with only one hour of work be said to “have a job”? Does the unemployment rate conceal the actual extent of employment hardship?

Before answering, we need to identify what the statistic is designed to measure. It is not deciding whether a person has enough income, whether their life is secure, or whether their employment problem has been solved. Its initial task is narrower: to divide the working-age population into statistical categories that can be measured consistently and compared over time.

2. “Employed”, “unemployed” and “not in the labour force” are not complete descriptions of a life

Australia’s Labour Force Survey uses a short reference period, normally asking what respondents did in a particular week. Respondents do not simply choose whichever identity—employed, unemployed or without work—best matches how they feel. Their classification is derived from a sequence of survey questions.

In this framework, unemployment does not merely mean having too little work or feeling that one has no job. A person will generally need to satisfy several conditions at once:

  • they were not employed during the reference week;
  • they had actively looked for work during the preceding four weeks; and
  • they were available to start work during the reference week.

Alternatively, they may be waiting to start a job within four weeks and have been able to start during the reference week if the job had been available then.

A person without work who did not actively seek it, or who could not start at that time, will generally not be counted as unemployed. They will instead be classified as not in the labour force. The three categories are therefore not natural descriptions of people’s lives. They form a statistical structure with an order of priority:

  1. First determine whether qualifying employment activity occurred.
  2. If it did not, determine whether the person both sought work and was available to start.
  3. If neither test is met, classify the person as not in the labour force.

International labour statistics describe this in terms of an activity principle and priority rules. Within the short reference period, qualifying employment activity that reaches the one-hour threshold takes priority. Only people who are not employed are then tested against the unemployment criteria. ILO: Labour force

“Employment” therefore initially indicates participation in work for pay or profit, not whether that work can sustain a life. “Unemployment” indicates an absence of employment activity together with active job search and availability. These are technical categories built for statistical purposes, not judgements on a person’s entire condition.

3. Why is the threshold one hour?

If one hour plainly does not represent a secure job, why not set the threshold at five, ten or fifteen hours?

The most direct reason is that labour force statistics need to capture all employment activity that actually occurred. One hour is very little, but it is still labour input and may still produce pay, services or economic output. Suppose employment instead required ten hours. Where should a person who worked nine hours be placed? They clearly did not have no work at all. Yet if they were also looking for more work and available to increase their hours, they might be placed among the unemployed. Unemployment would then combine two different conditions: having no work and having some work but insufficient hours.

The one-hour threshold solves a classification problem, not a problem of economic security. It sets a low and explicit operational floor for deciding whether employment activity occurred, allowing employment, unemployment and absence from the labour force to remain distinct.

It also supports comparison across months, years and countries. Australia has used the one-hour rule throughout the history of its Labour Force Survey, and international labour standards have long retained it. The ABS explains that employment statistics measure participation in employment, while the volume of work is measured separately through hours data. ABS: Hours worked guide

One hour is not, then, a socially agreed standard of “enough work”. It is closer to the lowest operational line that a statistical system needs to draw:

Once identifiable employment activity has occurred, record employment first. Whether the work is sufficient, stable and capable of supporting a life must be answered with other measures.

The rule is technically defensible, but only for its intended purpose. It can help a statistical agency distinguish labour activity. It cannot prove that one hour of work meets a person’s economic needs.

4. The statistical difference across a threshold may be much larger than the difference in real life

Imagine two people who each want to work thirty-eight hours a week and are both actively seeking more work.

The first performs no work during the reference week. If they satisfy the search and availability tests, they may be classified as unemployed.

The second happens to receive one hour of paid work. Their income and circumstances are almost identical to those of the first person, but that one hour may place them in employment.

The institutional distinction is sharp: one person enters the unemployment count and the other enters the employment count. The difference between their lived circumstances may be negligible.

This is a characteristic effect of thresholds. A boundary does not prove that the people on either side are fundamentally different. It temporarily cuts a continuous reality into categories so that a statistical system can operate.

The issue therefore divides into two questions.

First, is the one-hour rule suitable for deciding whether someone participated in employment activity during the reference week? Broadly, yes. If that hour genuinely meets the definition of work, ignoring it entirely would also distort the statistics.

Second, is the one-hour rule suitable for deciding whether someone’s employment problem has been solved? No. It does not measure whether their income is sufficient, whether their job is stable, whether their hours match their preferences, whether another shift will be available next week, or whether the work provides leave, career development and basic security.

Misunderstanding begins when the second question borrows the answer to the first. “Classified as employed” gradually becomes “has a job”, and this can then be interpreted as “no longer needs help”. During that transition, a technical category acquires social meanings it was never designed to carry.

5. What does the unemployment rate leave out?

The ABS is not unaware of the distance between one hour of employment and adequate employment. Labour statistics also report hours, full-time and part-time employment, the employment-to-population ratio, participation, underemployment and broader labour underutilisation.

Headline underemployment chiefly covers part-time workers who want and are available for more hours, together with some full-time workers who worked part-time hours for economic reasons. Adding unemployment and underemployment produces a broader measure of labour force underutilisation. ABS: Underutilised labour

Someone who works one hour, wants more work and is available to increase their hours may therefore disappear from the unemployment count while remaining visible in underemployment data.

Even unemployment and underemployment together cannot fully describe everyone whose capacity or need for work is unmet. Consider people who:

  • searched for a long time but eventually stopped because they became discouraged;
  • cannot start immediately because of caring responsibilities;
  • obtain fragments of platform work while their income and hours remain highly unstable;
  • work enough hours but still earn too little to live on;
  • perform extensive unpaid care, volunteer work or household labour; or
  • want to change jobs but cannot afford the risks involved in searching.

Their existence does not mean the statistics are erroneous. It demonstrates that every single indicator answers a limited question.

The unemployment rate asks: among people in the labour force, how many had no employment during the relevant period, actively sought work and were available to start?

It does not tell us:

  • how many people have enough hours;
  • how many can live on their employment income;
  • how many have stable and predictable work;
  • how many have abandoned their job search;
  • how many people’s capabilities remain underused; or
  • how many are classified as employed while experiencing acute financial pressure.

When public discussion cites only the unemployment rate, these other questions are compressed out of sight. A falling unemployment rate may reflect more people obtaining stable employment. It can also be associated with growth in part-time or fragmented work, job seekers leaving the labour force, or movement between statistical categories. Participation, underemployment, underutilisation, hours, income and job quality must be considered together before deciding what has happened in the labour market.

6. A statistical employment status does not automatically cancel benefit eligibility

Another common misunderstanding follows: if one hour of work makes someone statistically employed, do they lose JobSeeker Payment?

The ABS classification does not decide that question.

The Labour Force Survey and the social security system serve different purposes. ABS categories estimate labour market conditions across Australia and its states and territories. JobSeeker Payment applies social security rules to an individual’s eligibility and payment rate.

Services Australia states that people who are not in full-time work—including people doing part-time or casual work, people who have been temporarily stood down, and people whose hours have been reduced—may still satisfy the basic circumstances for JobSeeker Payment. They must also meet the relevant age, residence, income, assets and other requirements. Services Australia: Who can get JobSeeker Payment

Recipients with employment income generally need to report their own and their partner’s income according to their reporting requirements. If income and the other conditions remain within the rules, a payment may be reduced progressively rather than ending merely because the person is statistically employed. Services Australia: Income test for JobSeeker Payment

The same person can therefore occupy both of these positions at once:

  • employed in ABS labour force statistics; and
  • not in full-time work and still eligible for income support, subject to the social security rules and obligations that apply to them.

There is no contradiction. The two systems answer different questions.

The danger arises when a government, media organisation or member of the public carries one statistical category directly into another institutional judgement. If “employed population” becomes shorthand for “people who do not need assistance”, or if a falling unemployment rate is treated as proof that employment hardship has generally been resolved, a measure designed to describe aggregate labour activity begins making decisions it was neither designed nor authorised to make.

The Labour Force Survey produces population statistics rather than an appealable welfare, tax or legal decision about a particular respondent. Its main correction mechanisms are therefore public definitions, methodological review, statistical revisions, sample quality controls and the use of complementary measures such as participation, underemployment and hours.

But the moment statistical results are used to allocate resources, determine service capacity or claim that a policy has succeeded, the burden of explanation increases. Whoever uses the number must explain what it measures, what it omits, and whether it can genuinely support the decision being made.

Conclusion: the one-hour rule does not lie, but it is easily asked the wrong question

The claim that one hour of work can make someone employed sounds implausible because “having a job” in ordinary language often implies a meaningful amount of income, hours and stability. Employment in statistical language asks only whether qualifying employment activity occurred within the reference period.

The one-hour threshold does not declare that one hour is enough. Nor does it prove that the people on opposite sides of the line live materially different lives. It is an operational rule used to construct categories that are mutually exclusive, exhaustive and comparable.

The reasonable response is therefore neither to abolish the one-hour rule nor to dismiss the unemployment rate as fraudulent. What should be rejected is the expansion of an indicator beyond its authority:

A number suitable for deciding whether employment activity occurred should not be used to decide whether a person’s employment problem has been solved.

Modern institutions must compress reality. Without doing so, they could not produce national statistics, compare change or make policy. But every compressed number must carry the boundaries of its proper use. Employment, unemployment and underemployment are different windows onto a labour market, not complete identities of the people observed through them.

Numbers do not independently decide who belongs on either side of a threshold. Institutions and the people who use statistics choose the boundary, name the categories and interpret the result. Statistical honesty therefore depends on more than correct calculation. It also depends on refusing to pretend that a number answered a question it never measured.

Primary sources

Discover more from Geoffrey Chen

Subscribe to get the latest posts sent to your email.