One count of workers, cut by who they work for
In 2025 Georgia counted 1,389,700 people as employed, and Geostat cuts that one count by the kind of job each of them holds. 961,100 were hired, 426,300 were self-employed, and 2,300 fell into a third category the office calls a not-identified worker. The three add to the total exactly.
Self-employment is employment, and the classification that defines it says so in a parenthesis worth reading. ICSE-93 calls a job self-employment when the pay depends on the profits of the enterprise, and it adds that own consumption counts as part of those profits. Somebody producing on their own holding for their own household is therefore holding a job.
And an unemployment rate can only see the people who hold no job at all. It counts those who are without work, available to start, and actively looking, all three at once. So however much of a country's work is self-employment, none of it can show up in that rate.
- SELF-EMPLOYMENT IS EMPLOYMENT, AND THE DEFINITION IS EXPLICIT ABOUT WHY. ICSE-93, adopted by the Fifteenth International Conference of Labour Statisticians on 28 January 1993, separates jobs into paid employment and self-employment. Paragraph 6 defines a paid-employment job as one whose basic remuneration is not directly dependent on the revenue of the unit worked for. Paragraph 7 defines a self-employment job as one where the remuneration depends directly on the profits derived from the goods and services produced, and it adds, in its own parenthesis, that own consumption is considered to be part of profits. A person producing on their own holding for their own household therefore holds a job under this classification, and holding a job is what makes someone employed.
- TO BE COUNTED UNEMPLOYED IS TO BE WITHOUT WORK, AVAILABLE, AND LOOKING. All three at once. Somebody who is self-employed is not without work, so no amount of self-employment can raise an unemployment rate, whatever anybody thinks of the work. And somebody without work who is not looking is outside the labour force and is in neither the numerator nor the denominator of the rate. This is the single fact that makes a low unemployment rate compatible with very little hired work.
- HIRED AND SELF-EMPLOYED ARE GEOSTAT'S OWN TWO CATEGORIES, INSIDE ONE COUNT. The table publishes eleven economic-status rows, and "hired" and "self-employed" are both of them subdivisions of "employed", alongside a third, "not-identified worker", that holds a few hundred people. The three add to the published count of employed people exactly, in every group and in both periods read. So the composition below is a cut of one number and not two numbers put beside each other.
- A RATE OF THE LABOUR FORCE IS NOT A SHARE OF THE EMPLOYED. 11.2% is the unemployed people of rural Georgia divided by its labour force, which is everybody there who is employed plus everybody who is unemployed. 44.8% is the self-employed people of rural Georgia divided by its employed. The denominators are different numbers and the two figures answer different questions; neither can be computed from the other without the counts, and this edition prints the counts.
- THE EMPLOYMENT RATE HERE IS ON THE POPULATION AGED 15 AND OVER, WITH NO UPPER LIMIT. That is how Geostat publishes this table. A rate built on ages 15 to 64, which is how Eurostat and many national offices publish theirs, is a different measure on a different denominator, and this edition makes no comparison with any such figure and names no other country.
- URBAN AND RURAL, AND FEMALE AND MALE, ARE TWO SEPARATE CUTS OF ONE COUNT. Urban plus rural is Georgia and female plus male is Georgia, and the identities hold in the published figures. But no published cell crosses the two, so nothing in this edition says anything about rural women, urban men, or any combination of the two cuts. The one is used as a check on the other and never as a way through it.
So this table says how the employed people of rural and urban Georgia divided between two kinds of job in 2025, what the three published rates were for each area, how the sign of both comparisons held across every published year, and nothing about why, about what any of the work is, or about what anybody doing it would rather be doing.
One count of employed people, cut by what kind of job each person holds, and a rate that can only see the people who hold none. A reading that treats self-employment as a milder form of unemployment, or that gives employment and paid employment the same name, says what this table does not.
258,700 of rural Georgia's 577,800 workers are self-employed
Of the 577,800 people employed in rural Georgia, 258,700, which is 44.8%, were self-employed and 318,800 were hired. In urban Georgia the same cut reads 167,600 self-employed out of 811,900, or 20.6%, against 642,300 hired. The rural share is more than twice the urban one.
The two areas are one cut of a single national count, which is what makes the comparison a comparison. 811,900 plus 577,800 is the 1,389,700 employed people of the country, and 642,300 plus 318,800 is its 961,100 hired jobs. Neither figure is an estimate laid beside another estimate; they are two parts of one published number.
The remaining category is small and it is named rather than absorbed. The not-identified workers are 300 people in rural areas and 2,000 in urban ones, which is why the two shares of each area come to 99.9% and 99.8% rather than exactly 100%.
| Status in employment | Rural areas | Urban areas | Percent of rural employed | Percent of urban employed |
|---|---|---|---|---|
| Hired | 318,800 | 642,300 | 55.2% | 79.1% |
| Self-employed | 258,700 | 167,600 | 44.8% | 20.6% |
| Not-identified worker | 300 | 2,000 | 0.1% | 0.2% |
The lower rate, and the employment rate that is almost the same
Rural Georgia's unemployment rate was 11.2% and urban Georgia's 15.7%, a difference of 4.5 points. That is the rate a reader meets first, and on its own it says rural Georgia has less unemployment than urban Georgia, which is exactly what it says and all that it says.
The employment rate, which asks a different question, puts the two places 1.0 point apart. 46.3% of rural adults aged 15 and over were in work against 47.3% of urban adults. So nearly the same proportion of adults was working in the two areas while the rates that describe looking for work were four and a half points apart.
What differs is the kind of job, and the participation rate is where the two rates part company. 52.2% of rural adults were in the labour force against 56.1% of urban adults, so the rural labour force is smaller relative to its population. A rate whose denominator is the labour force and a rate whose denominator is the population are not two versions of one measure.
| Reporting group | Unemployment rate | Participation rate | Employment rate | Self-employed, percent of employed |
|---|---|---|---|---|
| Georgia | 13.9% | 54.5% | 46.9% | 30.7% |
| Urban areas | 15.7% | 56.1% | 47.3% | 20.6% |
| Rural areas | 11.2% | 52.2% | 46.3% | 44.8% |
Sixteen printed years, and the gap never changes sign
In every one of the sixteen years from 2010 to 2025 the rural unemployment rate is below the urban one, and the rural self-employed share is above the urban one. The size of both gaps moves a great deal: the rate gap is 6.3 points in 2010, 1.1 points in 2024 and 4.5 points in 2025, and the rural self-employed share falls from 62.8% to 44.8% over the sixteen years while the urban share moves within a narrow band, 20.9% in 2010, down to 16.7% in 2017 and back to 20.6% in 2025.
The series runs back to 1998 and this edition does not read across it. Between 2009 and 2010 rural employment falls from 900,500 to 504,200, rural self-employment from 745,200 to 316,700, and the rural unemployment rate rises from 7.6% to 23.5%. No annual change does that. The API carries no comment, no footnote and no methodological note of any kind, so the movement is reported and not explained, and no level in this edition is compared across those two years.
What does hold over all twenty-eight published years is the sign, because the sign is a comparison made inside each year. Rural below urban on the rate in twenty-eight of twenty-eight; rural above urban on the self-employed share in twenty-eight of twenty-eight. A change in how a survey is drawn or weighted changes levels; it does not retrospectively reverse which of two figures measured in the same year by the same survey was the larger.
| Year | Urban unemployment rate | Rural unemployment rate | Urban self-employed share | Rural self-employed share |
|---|---|---|---|---|
| 2010 | 29.8% | 23.5% | 20.9% | 62.8% |
| 2011 | 30.7% | 22.0% | 17.8% | 64.8% |
| 2012 | 29.7% | 22.4% | 17.3% | 62.6% |
| 2013 | 29.9% | 21.3% | 20.3% | 60.1% |
| 2014 | 26.1% | 18.3% | 18.3% | 59.9% |
| 2015 | 25.5% | 16.3% | 18.0% | 56.5% |
| 2016 | 24.8% | 16.9% | 16.9% | 57.0% |
| 2017 | 25.1% | 16.4% | 16.7% | 53.1% |
| 2018 | 20.4% | 17.6% | 17.1% | 49.0% |
| 2019 | 18.2% | 16.7% | 17.1% | 50.3% |
| 2020 | 20.3% | 15.8% | 18.2% | 51.7% |
| 2021 | 22.2% | 18.2% | 18.3% | 51.2% |
| 2022 | 18.3% | 15.7% | 20.3% | 48.8% |
| 2023 | 17.6% | 14.6% | 20.0% | 47.1% |
| 2024 | 14.4% | 13.3% | 21.9% | 45.8% |
| 2025 | 15.7% | 11.2% | 20.6% | 44.8% |
Women have the lower rate and less self-employment
The same table cuts the same national count by sex, and the pairing inverts. Women's unemployment rate was 11.4% against men's 15.8%, the lower rate, as rural areas have the lower rate. But women's self-employed share was 20.9% against men's 38.5%, the LOWER share, where rural areas have the higher one.
That is why this edition states its finding of the urban and rural cut and of nothing wider. If a lower unemployment rate came with more self-employment as a rule, the sex cut of the very same count would show it, and it shows the opposite. One cut of one country in one year is the whole territory of the claim.
And the third reading in the table is the one that should temper all of them. 44.2% of women aged 15 and over were in the labour force against 66.3% of men. Two unemployment rates computed on labour forces that different are each exact about their own denominator and say correspondingly less when set side by side.
| Reporting group | Unemployment rate | Self-employed, percent of employed | Participation rate |
|---|---|---|---|
| Georgia | 13.9% | 30.7% | 54.5% |
| Women | 11.4% | 20.9% | 44.2% |
| Men | 15.8% | 38.5% | 66.3% |
What these figures do not say
They do not say that rural Georgia's labour market is healthier, or that urban Georgia's is worse. The unemployment rate is lower in rural areas and that is the finding; what kind of work the employed are doing, what it pays, how secure it is and whether anybody would prefer different work are all outside everything read here. Geostat publishes the status of the job and the count of the people, and this edition read no earnings figure, so it names none.
They do not say that self-employment is unemployment under another name, or a milder form of it. Under the classification the survey uses, a self-employment job is a job: paragraph 7 of ICSE-93 defines it by the dependence of the pay on the profits of the enterprise, counting own consumption as part of those profits, and nothing in the definition turns on whether the holder would rather be hired. Nor do they say the reverse, that the two kinds of job are interchangeable. The edition prints the two counts and the definition and makes neither judgement.
They do not say that one of these figures causes the other. No published cell crosses status in employment with looking for work, so the sentence that the rural rate is low because so much rural work is self-employment is not demonstrable from anything read here, however reasonable it sounds, and it is not written. Nor is any other cause named: not agriculture, not migration, not education, not the structure of the economy, and not policy of any kind.
They do not say that rural self-employment is farming. Agriculture, forestry and fishing employed 218,300 of the 1,389,700 people in work in 2025, 15.7% of employment, while self-employment is 30.7% of the same count, so self-employment is about twice the size of the whole of agriculture. And the activity table has neither an urban and rural dimension nor a status-in-employment one: the two cuts cannot be crossed, so what rural self-employed people actually do is not in this edition at all.
They do not say that 11.2% is a record low or 44.8% a record high. Twelve of the twenty-eight published years carry a lower rural unemployment rate than 11.2%, all of them before the discontinuity the edition refuses to read across, so 11.2% is the lowest of the sixteen printed years and nothing more. The rural self-employed share is at its lowest of those sixteen years, not its highest, having fallen from 62.8% in 2010. And they do not say anything the edition did not read: nothing here is broken out by age, region, occupation, education or earnings; the two periods printed are an annual average and a single quarter and are never mixed in one sentence; no other country appears; nothing is a forecast; and no figure in this edition comes from our own platform.



