Labour Underutilisation in Georgia Is Double the Headline Unemployment Rate: What the LFS Misses

GILS Research Team

Editor: Prof. Giorgi Pkhakadze, MD, MPH, PhD

Georgian Institute of Labour Studies · Data & Wages

2 September 2026

In 2024 Georgia’s unemployment rate fell to 13.9%, down 2.5 percentage points on the previous year. Yet the ILO’s composite measure of labour underutilisation stood at 28.8% — more than twice as high. This analysis shows that the gap is not driven by employed people wanting more hours, but by a large reservoir of people outside the labour force who want work and are not counted as unemployed. That finding changes what labour market policy should prioritise.

1. Introduction

The unemployment rate is the most cited labour market statistic in Georgia, and for good reason: it is internationally comparable, published quarterly, and easy to communicate. But it measures only one narrow form of labour market failure. A person is counted as unemployed only if they are without work, actively searched for a job in the past four weeks, and are available to start within two weeks. Anyone who wants a job but has stopped searching, or who searched but could not start immediately, is classified as outside the labour force — and disappears from the headline figure.

This analysis asks a simple question: how much labour slack does Georgia’s headline unemployment rate miss, and where is it located? Our thesis is that the missing slack sits primarily in the potential labour force — people at the margin of the labour market — and that this has direct consequences for policy design. Programmes that create more working hours for the already-employed will do little; programmes that bring discouraged and constrained people back into work will do a great deal.

2. Concepts and Measurement

The international standards for measuring labour underutilisation were adopted by the 19th International Conference of Labour Statisticians in 2013 (ILO, 2013). They define three groups whose labour is under-used:

Unemployed persons — without work, actively seeking, and available. Persons in time-related underemployment — employed people who worked part-time, actively looked for additional or alternative work, and were available to work more hours. The potential labour force — people outside the labour force who either searched for work but were not available to start, or were available but did not search.

From these groups, four indicators are constructed. Geostat publishes all four, following the separation of the Labour Force Survey from the Integrated Household Survey in 2017 (Geostat, n.d.):

The extended labour force is the labour force plus the potential labour force. This distinction matters: LU3 and LU4 use a larger denominator, so the potential labour force raises them even though the base is wider.

3. Findings

3.1 The headline improvement is real

By conventional measures, Georgia’s labour market improved substantially after the pandemic. The unemployment rate was 17.6% in 2019, rose to 20.6% in 2021, and then fell steadily to 13.9% in 2024 (Geostat, 2025; PMCG Research, 2025). In 2024 the labour force participation rate rose to 54.8% and the employment rate to 47.1%, increases of 1.5 and 2.6 percentage points respectively over 2023 (Geostat, 2025). These are meaningful gains and should not be dismissed.

3.2 But underutilisation remains more than twice as high

Source: Geostat, compiled by PMCG Research (2025). Change and ratio columns are GILS calculations.

3.3 Decomposition: where the slack sits

The decomposition is striking. Moving from LU1 to LU2 — adding employed people who want more hours — raises the 2024 rate by just 0.6 percentage points. Moving from LU1 to LU3 — adding the potential labour force — raises it by 14.4 points. Time-related underemployment is therefore a minor component of Georgian labour slack. The potential labour force is the dominant one.

PMCG Research (2025) interprets the same data in similar terms: a relatively small segment of underemployed people seeking extra hours, alongside a considerable number of discouraged and marginally attached people excluded from the standard unemployment metric.

3.4 Who is likely to be in the potential labour force

The Labour Force Survey releases used here do not break the potential labour force down by sex or age, so its composition cannot be stated precisely. Two well-documented features of the Georgian labour market, however, indicate where it is likely concentrated.

First, women. In 2024 the labour force participation rate was 44.9% for women and 66.2% for men (Geostat, 2025). Between 2019 and early 2025 the gap averaged 21.5 percentage points with no clear trend towards narrowing (PMCG Research, 2025). A participation gap of this size, in a country where women’s educational attainment is high, is a strong indicator of constrained rather than freely chosen inactivity.

Second, young people. In 2024 unemployment in the 15–19 age group was 36.3%, the highest of any age group (Geostat, 2025). Young people at the start of their working lives are those most likely to move between searching, studying and inactivity — exactly the transitions that place people in the potential labour force.

3.5 Emigration as a confounding factor

A falling unemployment rate can reflect people finding jobs, or it can reflect people leaving. Between the first quarter of 2019 and the first quarter of 2025, the population outside Georgia’s labour force fell by 6.1%, or nearly 88,000 people. PMCG Research (2025) estimates that at most 45.8% of this decline can be explained by movement into the labour force; the remainder likely left the working-age population, possibly through emigration. Some of the improvement in headline indicators is therefore demographic rather than economic.

4. Discussion

Argument 1: The headline rate overstates labour market health

If policymakers judge the labour market by unemployment alone, they see a rate that has fallen by a quarter in four years. If they look at the composite measure, they see that more than one in four people in the extended labour force was still under-used in 2024. Both statements are true. But only the second captures the scale of the policy challenge.

Argument 2: The structure of exclusion has not changed

The ratio of LU4 to LU1 was 2.14 in 2020 and 2.07 in 2024. Underutilisation fell in roughly the same proportion as unemployment. In other words, the recovery lifted the whole distribution without changing its shape. The groups at the margin — those who want work but are not searching or not available — remain as large relative to the unemployed as they were during the pandemic. Growth alone has not drawn them in.

Counter-argument: the potential labour force is not all “hidden unemployment”

A reasonable objection is that the potential labour force includes people who could not take a job immediately — for example, those with caring responsibilities or students finishing a course. It would be wrong to describe all of them as jobless people waiting for work. We accept this. LU4 should not be read as a “true unemployment rate”.

But the objection strengthens rather than weakens the policy case. People who want to work but cannot be available — because of care obligations, distance, cost of transport, or lack of suitable part-time work — face barriers that policy can address. People who want to work but have stopped searching are, by definition, people the existing system has failed to reach. Neither group is helped by policies designed for the active job seeker.

Policy implications

Because time-related underemployment is small, measures that mainly increase hours for those already in work will have limited effect on overall slack. Because the potential labour force is large, the highest returns are likely to come from three areas: reducing barriers to availability (especially childcare and elder care, which disproportionately constrain women); proactive outreach by employment services to people who are not registered job seekers; and better monitoring, so that progress is measured against the broader indicators rather than the headline rate alone.

5. Limitations

This analysis uses annual figures for 2020 and 2024 only; intermediate years and quarterly variation are not examined. The composition of the potential labour force by sex, age and reason for non-availability is inferred from related indicators rather than observed directly. Survey-based measures are subject to sampling error, and emigration may affect the representativeness of the sample over time. A fuller analysis would use Geostat microdata to decompose the potential labour force directly.

6. Conclusion

Georgia’s labour market has improved, and the decline in unemployment to 13.9% in 2024 is real. But the composite underutilisation rate of 28.8% shows that the headline figure captures less than half of the labour that is going unused. That slack is located not among employed people seeking more hours, but among people at the edge of the labour force. A labour policy that measures success by the unemployment rate will consistently underestimate the problem and target the wrong people. Georgia should report, monitor and set targets against the broader measures it already collects.

GILS analyses are prepared with AI-assisted drafting. All figures and claims are verified against the cited primary sources and reviewed under the responsibility of the Editor.