In April 2025 Belarus’s statistics committee published two numbers for the gender pay gap: 26.8% before adjustment and 6% after it. Its officials explain the larger number above all by where women work: in the social sphere, with its schools, clinics and other low-paid public services. Belstat’s own table for December 2025 does not support that as the main reason: the industries women work in account for less than a third of the gap. What we cannot see is how the rest of the adjustment divides, since Belstat has published the result but not the share of each factor.
Two numbers from one office
The adjusted figure came first. On 31 January 2025, before the committee’s year-end board meeting, its chair Inna Medvedeva told journalists that the pay gap was 5.8%, BelTA reported. This was a new, experimental calculation on data for 2021. The matching unadjusted gap, in hourly pay for October 2021, was 24.3%. It follows from page 29 of Belstat’s labour booklet of 2022: 9.20 rubles an hour for men and 6.96 for women. “Today Belarus gives a woman the opportunity to receive the same wage as a man,” she said.
Two days before Women’s Day, the head of Belstat’s labour statistics department, Irina Bondarenko, went further. The experiment, she said, confirmed that there was no significant pay gap by sex. She put the unadjusted gap at 25–26% and explained it above all by employment: most women work in the social sphere, where average pay is lower. SB, the presidential administration’s daily, carried the same explanation under the headline “Why women’s pay is 25% lower than men’s, Belstat explained”.
On 30 April 2025 Belstat put both numbers in one infographic, marked as new. For October 2024, the gap in hourly pay was 26.8% unadjusted and 6.0% adjusted. The adjusted figure compares women and men with the same tenure, occupation group and education, and allows for industry. The infographic names three factors: the industry factor, workers’ education and “occupational segregation”. Two days later the state broadcaster CTV summed it up: “the main reason for the differences in income is the type of activity.”
Put the two survey rounds side by side, and a pattern appears. Between October 2021 and October 2024, the unadjusted hourly gap grew from 24.3% to 26.8%. The adjusted gap moved from 5.8% to 6.0%. So almost all of the increase sits in the part that the adjustment removes. The 2021 figure was an experimental calculation, and the method may have changed since, so this comparison is a guide rather than a measurement.
In January 2026 the committee made the adjusted measure official. Its method, approved on 23 January, defines the adjusted gap but not an unadjusted one. Belstat’s labour booklet for 2026, registered on 30 September, marks the adjusted figure “for the first time!”. It asks “Why do women earn less than men?” and answers that “men mainly work in better-paid spheres of activity”. Its booklet for 2022 gave pay by occupation group for women and men in October 2021. The new booklet gives pay by occupation group for October 2024 with no split by sex.
State media followed the same order. We checked by hand every item on the gap in state media and pro-government Telegram channels from 31 January 2025, when the adjusted figure first appeared. Of 23 items that gave a figure, 18 carried the adjusted one, and 10 of them carried nothing else. Five gave only the unadjusted figure. Four headlines, in BelTA, SB and two Telegram channels, called 5.8% simply “the difference in pay between men and women”. Seven more items gave no figure and said the gap was small or absent. In August 2025, BelTA’s headline read “Belstat: in Belarus there are no differences in pay depending on the worker’s sex”. A year later, in a BelTA interview that Grodnenskaya Pravda republished, Medvedeva said the 6% “confirms there is no significant pay gap between men and women in Belarus”.
This need not be a choice made in newsrooms. State outlets usually relay Belstat’s own releases, and Belstat’s Telegram channel announced the 2024 result with the adjusted 6.0% alone. Nor did state media hide the table. In May 2026, 1prof.by reported the 25.7% gap for December 2025. Its headline asked where most women work and where they earn more than men. The answer to the second question was one industry, real estate, by 0.5%.
Independent outlets did the opposite. Of their 24 items that gave a figure, none carried the adjusted one alone. Eleven carried both, and 13 only the unadjusted one. Zerkalo opened with 26.8% and gave the 6% further down, under the headline “Women began to earn even less than men”. Charter97 and Euroradio headlined that the gap had grown. One independent site read the change of measure as a change in pay. On 31 January 2025, Pozirk set the 5.8% against the 24.3% that Belstat gives for 2021 before adjustment. Its headline said the gap had shrunk more than fourfold in three years. Both figures describe 2021.
The table that answers back
So almost all of the increase sits in the part that the adjustment removes.
Belstat also publishes a table that lets anyone test the industry explanation. Its bulletin on wages in 2025 gives, on pages 25 to 27, the number of women and men and their average pay in every industry for December 2025. Women were 53.0% of the 3.16 million employees counted. Their average accrued wage was 2,582 rubles, against 3,474 rubles for men. That is 74.3% of men’s pay, a gap of 25.7%.
With that table, the gap can be split into two parts. One part comes from the industries women work in, and the other from pay differences inside each industry. Across the 19 sections of the national classification, the first part is 4.4 to 6.0 points, depending on the weights used. With manufacturing and trade divided into their branches, it rises to 5.4 to 7.8 points. Either way, more than two thirds of the gap sits inside the same industries. The table for December 2024 gives almost the same split: 4.9 to 8.0 points of a 27.1-point gap, and the table for December 2022 gives 5.2 to 8.5 points of 26.1.
The gap inside industries is large where women are many and where they are few. In health and social work, where 85% of workers are women, women earn 21.5% less than men. In finance, with 72% women, the gap is 24.1%. In mining, where women are 20% of workers, it is 26.5%, and in information and communication it is 37.7%. Education and transport are the near exceptions, at 4.6% and 6.2%. Real estate is the only industry where women earn slightly more. Zerkalo noted the same pattern in the table for 2024.
Either way, more than two thirds of the gap sits inside the same industries.
Women and men do work in different places. By the standard index of dissimilarity, 35% of women would have to change industry to match the distribution of men. With the finer split, the figure is 40%. But the split runs across both well-paid and badly paid industries. Women dominate education and health, which pay below the average, and also finance, which pays far above it. Most workers in construction and manufacturing are men, but so are most workers in agriculture and transport, which pay less than the average. So the split by sex is strong, while its effect on average pay is small.
There is another reading of the same table. Pay gaps inside industries may reflect hours, tenure, education and the jobs people choose, and none of these needs to involve unfair treatment. December pay also includes year-end bonuses, which can widen gaps for one month. In 2024, the hourly gap in October, 26.8%, was close to the monthly gap in December, 27.1%. This suggests that shorter hours are not most of the story. The two figures come from different surveys, however, so the comparison is only a guide.
What an adjustment is allowed to explain
Here Missing Voices in Economics is useful. The collection, edited by Veronika Dolar and Teresa Perry, came out from Palgrave in January 2026. It is about women in economics as a profession, and it does not mention Belarus. But its first chapter, by Marlene Kim, describes the habit of mind behind the “no significant gap” reading. Mainstream theory, she writes, explains gaps in pay and occupation by preference, skill and productivity. She quotes a senior economist, citing Julie Nelson (2020): “The word ‘discrimination’ has almost disappeared. It’s the wage gap and unexplained factors … anything other than discrimination.”
So the split by sex is strong, while its effect on average pay is small.
Two other chapters measure segregation with the index we use here. Giulia Zacchia and Izaskun Zuazu apply it to the research fields of the world’s top economists. Zacchia and Marcella Corsi apply it to the topics of Italian economists. In both chapters, segregation is a result to be measured and explained.
Belstat’s adjustment treats occupation the other way. Once women and men are compared inside the same occupation group, the sorting of women into lower-paid jobs drops out of the result. Belstat’s own infographic names “occupational segregation” among the factors. Its officials then present what remains after that factor is removed as the answer. Both numbers are true. The choice is which one gets the explanation.
Adjusted gaps are a normal tool in labour statistics. A 6% gap between women and men in the same occupation group, with the same tenure and education, after allowing for industry, is not small. Belstat’s figure is not zero, although its officials call it insignificant. What the official reading leaves aside is the larger part. By Belstat’s own two numbers for 2024, the adjustment removes about 21 points. On our split of the December table for that year, industry accounts for 5 to 8 of them. As a rough guide, the rest lies inside the same industries. It is in occupation group, tenure and education, and in the employers women work for. Belstat’s method also allows for pay levels that differ between organisations.
Both numbers are true. The choice is which one gets the explanation.
What the year’s materials leave out
On 1 January 2026, Lukashenko’s Decree No. 1 declared this the Year of the Belarusian Woman. Belstat’s plan for the year has 13 items, including sports contests, charity work and meetings with clergy. None of them concerns pay data; the one statistical item covers reviews and infographics for holidays. We read all six infographics listed on Belstat’s page for the year: the statistical portrait, demography, employment and decision-making, education and science, gender equality and time use. None gives a wage figure by sex. The page on gender equality cites the UN’s Gender Inequality Index, on which Belarus ranks 27th of 172 countries; that index does not measure pay.
Belstat did publish the pay table this year, in its routine wage bulletin for 2025. The materials made for the Year of the Woman do not repeat it. The time-use page compares a 2025 survey with one from 2015, but it gives men only as a ratio to women. The year’s twelve-page statistical booklet collects the same infographics and adds no pay figure either. Its back page says its design and printing were supported by a joint project on localising the Sustainable Development Goals, “financed by the Russian Federation”. The employment page gives one number that bears on pay. Women, it says, make up 49.7% of managers. In October 2021, women managers earned 29.5% less than men managers, by Belstat’s booklet of 2022. How women can hold half of management posts and still earn less in almost every industry is a question the published data cannot answer.
State media have covered the year at length, but seldom its pay side. From 31 December 2025 to 9 October 2026, our corpus holds 4,885 state media items that name the Year of the Belarusian Woman. By our search, three of them mention a gap in pay. Two carry one claim by Sergei Aleinik, the national coordinator for achieving the Sustainable Development Goals: there is “no difference in pay for doing the same duties”, BelTA reported. Two days later Zerkalo replied that the gap had not fallen below 20% in ten years.
Belstat’s December series is older and steadier than the hourly one. Its labour booklet of 2022 gives women’s pay as 74.6% of men’s in December 2017, 72.7% in 2018, 73.1% in 2019, 73.3% in 2020 and 71.9% in 2021. The booklet of 2024 gives wages for December 2023 that put women’s pay at 73.5% of men’s. The wage bulletins give 73.9% for 2022 (bulletin for 2022, p. 25), 72.9% for 2024 and 74.3% for 2025. So there is a published December figure for every year from 2017 to 2025. In March 2024 a deputy labour minister, Valery Kovalkov, still put the gap at “about 20–25%” and explained it by the jobs women mostly do, Zerkalo reported. When Belstat’s officials spoke about the gap in 2025, they brought the adjusted figure with them.
How sure we are
The figures taken from Belstat’s own documents are High: the tables add up, and the infographic states both gaps. The comparison of the two hourly rounds is Medium, because the first adjustment was experimental. The reading of what the adjustment removes is also Medium, because it puts two surveys side by side. The media counts are Medium: we checked each state item by hand, but a keyword search can miss items, and we left out video.
How it fits
How it fits. Belstat publishes pay tables that undercut its own explanation of the gap, and its officials presented the adjusted number as the answer. This fits what we saw in “Not the Quality of the Data”, which read the same BelTA interview of 23 August: the national figures may be sound, while the comparisons placed around them read well for Belarus. A second reading is technical catch-up, with an office adopting a standard measure and state media simplifying it. The first release under the new method, the labour booklet of September 2026, shows both numbers but not the share removed by each factor, and it drops the split of pay by occupation and sex. If Belstat publishes the factor shares, the second reading gains.
What to watch
By 31 December 2027, Belstat publishes neither its October 2024 tables of pay by sex and occupation or industry nor the effect of each factor in its adjustment. We expect this, with Medium confidence, because the method does not promise them.
By 31 December 2026, no Belstat product made for the Year of the Belarusian Woman, including its review for Mother’s Day on 14 October, gives a wage figure by sex. We expect this, with High confidence.
By 30 June 2027, the wage bulletin for 2026 shows a December gap between 24% and 28%, and the industries women work in explain less than a third of it. We expect this, with High confidence.
Method and limits
The split of the December gap uses Belstat’s published tables of headcount and average accrued wage by sex for 2025, 2024 and 2022. Each covers 19 industry sections and, in a second version, 33 sections and branches. All three tables are internally consistent: the sections add up to the totals within rounding, and the published ratios match the wages. We split the gap into a between-industry and a within-industry part with the standard shift-share method, and we report both weightings because the result depends on them. The index of dissimilarity is half the sum of the absolute differences between women’s and men’s shares across industries. December wages include year-end bonuses and are gross. The hourly figures come from a separate Belstat sample survey held in October, so we compare sizes across the two sources, not decimals. Belstat has not published the share of each factor in its adjusted gap. Its method also allows for pay levels that differ between organisations in the same industry. Our division of the removed 21 points is therefore a rough guide, not Belstat’s figure. Our reading of Belstat’s materials for the year covers all six infographics on its page for the year, the year’s statistical booklet, the plan and the review for 8 March 2025. The media count uses our corpus of Belarusian state, independent and exile outlets and Telegram channels. We searched it from 1 January 2024 to 9 October 2026 for words on pay, sex and a gap. The search returned 185 items about Belarus. We checked every state item from 31 January 2025 by hand. We read most in full and judged a few reprints by their headline and the figures our search extracted. We dropped items not about the gap, posts that only linked to other items, and three video items we could not check. Reprints in other outlets count separately. Our automatic count missed figures that stood far from the word “gap”, so we report only hand-checked numbers. We checked every independent item with a figure in the same way; five whose pages did not load were judged by their headline and figures. The Year of the Woman count covers items from 31 December 2025 to 9 October 2026 that name the year.